The Most Flexible Sicilian

A Black Repertoire with 1.e4 cS 2.�f3 e6

Alexander Delchev
Semko Semkov

Chess Stars
www.chess-stars.com
Chess Stars Publishing
Current Theory and Practice Series

The Most Flexible Sicilian

Cover design by Kalojan Nachev

Copyright© 2014 by Alexander Delchev and Semko Semkov

Printed in Bulgaria
ISBN: 978-954 8782 97-5
Contents
t.e4 c5 2 . �f3 e6

Foreword 5
Introduction 6

Open Sicilian
Part 1 The Classical System 11
3 . d4 cxd4 4.tt:\xd4 tt:\c6 5.tt:\c3 V!ic7 6.ie2 ; 4 ... a6 5.tt:\c3 V!ic7 6.ie2
Part 2 The English Attack 45
3 . d4 cxd4 4.tt:\xd4 tt:\c6 5.tt:\c3 V!ic7 6.ie3 ; 4 . . . a6 5.tt:\c3 V!ic7 6.ie3
Part 3 The Third Rank Set-Up 77
4.tt:\xd4 tt:\c6 5.tt:\c3 V!ic7 6.ie3 a6 7.id3 ; 4 . . . a6 5.tt:\c3 V!ic7 6.id3
Part 4 The f4 System 123
3 .d4 cxd4 4.tt:\xd4 tt:\c6 5.tt:\c3 V!ic7 6.f4; 4 . . . a6 5.tt:\c3 V!ic7 6.f4
Part 5 The Fianchetto 155
3 . d4 cxd4 4.tt:\xd4 tt:\c6 5.tt:\c3 V!ic7 6.g3 ; 4 . . . a6 5.tt:\c3 V!ic7 6.g3
Part 6 Kan - the Mar6czy Bind 197
4 . . . a6 5.c4; S.id3
Part 7 Taimanov with S.tt:lbS 241
3 .d4 cxd4 4.tt:\xd4 tt:\c6 S.tt:lbS d6
Part 8 Rare Lines on 5th and 6th Move 273
3 .d4 cxd4 4.tt:\xd4 tt:\c6 5.tt:\xc6 ; 5.tt:\c3 V!ic7 6.tt:\xc6 ; 6.V!id3; 6.tt:ldb5

Anti-Sicilian Systems
Part 9 The Alapin System 289
3 . c3 dS
Part 10 KID Reversed 301
3 .d3
Part 11 Rare Lines on Move 3 315
3 .b4 ; 3 .b3 ; 3.c4; 3.tt:\c3 ; 3.V!ie2 ; 3.g3

Index of Branches 337

Bi b l i ogra phy
Books
The Safest Sicilian, Alexander Delchev and Semko Semkov, Chess Stars
2008
The Sicilian Taimanov: Move by Move by John Emms, Gloucesters
Publishers, 2012
Play the Sicilian Kan; Johan Hellsten, Everyman Chess 2008
The Complete Hedgehog: Volume 1+2 by Sergei Shipov, Mongoose Press
Experts vs the Sicilian, 2nd edition, Quality Chess, 2006
English Attack by Nick de Firmian and John Fedorowicz, Batsford, 2 004
Anti-Sicilians: A Guide for Black by Dorian Rogozenko, Gambit, 2003
Winning With the Sicilian by Mark Taimanov, Collier Books 199 1
Secrets of Opening Preparation 2 by Mark Dvoretsky and Artur Yusupov,
1996 (In Russian)
Super Tournaments 2000, Ches� Stars, 2001
Super Tournaments 2002, Chess Stars, 2003
Super Tournaments 2003, Chess Stars, 2004
Opening for White According to Anand, vol . 9, by Alexander Khalifman,
Chess Stars 2 0 07
The Complete c3 Sicilian by Evgeny Sveshnikov, New in Chess 2010
How to beat the Sicilian Defence by Gawain Jones, Everyman Chess 2011

Periodicals
Chess Informant
New in Chess

Internet resources
Databases
The Week In Chess (chesscenter.com)
10 Days (Chessmix.com)
Internet Chess Club (chessclub.com)
ChessPublishing.com forum
Chess Today

4
F or ew or d
This book offers a double reper­
toire for Black based on the moves
l.e4 cS 2.ltlf3 e6 ! . It is divided in
several parts : playing against ie2 ;
against f4 ; g3 ; the third rank set-up
ie3+id3 ; the English Attack; the
hedgehog structures.
After 3.d4 cxd4 4.lt:lxd4, I
analyse in separate chapters the
Taimanov move order with 4 . . . lt:lc6,
and the Kan with 4 ... a6. Every part
begins with a chapter "Main Ideas"
which discuses the pros and cons of
the two approaches and offers ad­
vice which one works better against
White's particular set-up. The next
chapters, called "Taimanov - Step
by Step" and "Kan - Step by Step",
present detailed theory. Finally, a
"Complete Games" chapter offers
additional annotations.
This structure should allow you
to better understand the typical
ideas which are similar in many as­
pects :
Black's pawn often remains on
d7 allowing an early activation of
the dark-squared bishop through
b4, cS or even d6 in some lines, the
.. . dS break is without loss of tempo
o n . . . d6;
The pressure on the e4-pawn
with . . . bS, . . . ib7, . . . ltlf6 is common
for both systems;
The queen goes early to c7 to
control the central dark squares and
support the bishop when it goes to
b4 or d6;
We often change the pawn struc­
ture by recapturing . . . dxc6 followed
by . . . e6-e5.
The hedgehog structures share
the same main plan - to achieve the
break . . . dS.
The concluding parts present a
repertoire against the Anti-Sicilians
like the Alapin with c3 , the King's
Indian Reversed with d3, and the
modern fianchetto with 3.g3.
It is difficult to make a repertoire
which would satisfy every taste. I
would recommend the Kan to read­
ers who : understand positional play
and prophylaxis; are not theoreti­
cians ; defend calmly and unabash­
edly.
If you like early clashes and
sharp theoretical disputes (and
have good memory!), the Taimanov
is the better choice.
The ultimate goal should be to
master both systems. Then you'll
be really unpredictable to your op­
ponents !
Alexander Delchev
April 2014
5
I ntroduction
The Taimanov System which aris­
es after the moves l.e4 cS 2.lt:lf3 e6!
3.d4 cxd4 4.tt:lxd4 tt:lc6 S.tt:lc3 Y!fc7
was my main weapon against l.e4
until 2008.
It suits well my style and it
brought me good practical results.
Another cause of my successes was
that my opponents lacked clear
models to follow since the varia­
tion was not popular at top level. I
was able to play my own schemes
and I only had to make minor im­
provements on my previous games.
It was sufficient to keep in store a
few alternatives which aimed to es­
cape home preparation and assured
me of the psychological edge. I had
always the feeling that the battle
was fought on my land. However,
the second edition of The safest
6
Sicilian brought about a burst of
popularity of my pet system. The
first players developed many new
plans, the engines also improved
dramatically. Some of my back­
up alternatives were refuted. There
was no more ground for retreat and
I had to accept deep theoretical
disputes in every game. It became
practically unrewarding to keep my
repertoire in full swing all the time.
So I decided to reserve the
Taimanov for rapid games where
there is no time for targeted home
preparation and good understand­
ing is often more important than
the objective evaluation of the posi­
tion. Thus I was able to experiment
with new ideas without spending
too much time on them.
At the same time I began my mi­
gration toward the Kan. I was ex­
tremely impressed by its flexibili­
ty and the variety of plans it offers.
The move order l.e4 cS 2 .tt:lf3 e6!
3.d4 cxd4 4.tt:lxd4 a6 is outwardly so
modest and unobtrusive that White
could choose virtually any develop­
ing scheme at his liking. The play
is not forced and both sides have
tons of options on every move. This
greatly reduces the chance of be­
ing caught on a home preparation .
 Introduction
The wide choice also means an am­
ple ground for strategic mistakes.
Even good grandmasters often mix
up the move order and allow an easy
equalisation. Look at the follow­
ing variation: l.e4 cS 2.l2:\f3 e6 3.d4
cxd4 4.l2:\xd4 a6 5.c4 l2:\f6 6.id3 l2:\c6
7.ie3?!
White played a natural devel­
oping move, but Black suddenly
leaps forward: 7...l2:\e5! 8.h3 - an­
other natural move which occurred
in 5 games (8.l2:\c3 l2:\fg4; 8.0-0
ti:Jfg4 9.if4 gS!) 8...\Wb6! and Black
already has an initiative - 9.l2:\b3
(9.a3 �xb210.l2:\b3 l2:\xd3+ ll.�xd3
'&eS12.l2:\ld2 �c7; 9.l2:\b5 icS+) 9...
'&c6! (9...ib4+ 10.l2:\c3 �d6 ll.id4
ti:lxd3+ 12.�xd3 e513.0-0-0!)
White must give up the c4-pawn
with a dubious compensation.
The best way to exploit the weak­
nesses of your opponents' reper­
toires is to master the whole range
of systems with 2...e6. That would
allow you to choose the most un­
pleasant set-up against any par­
ticular opponent. For instance, if
the White player is narrowly spe­
cialised in the English Attack, you
may choose to delay ...l2:\c6. This
approach effectively discourages
White's set-up with ie3. It is also
very good against the fans of an
early f4, the fianchetto with g3 and
the Classical system with ie2. On
the contrary, if your opponent is a
practical player who avoids main
lines, then the Taimanov move or­
der would be more unpleasant as it
narrows White's safe paths in the
opening.
I would choose the Kan against
any player who treats it with 5.l2:\c3,
and the Taimanov against fans of
the Mar6czy Bind. The Kan may
also be the better option if your op­
ponent tried to outwit you by the
move order l.e4 cS 2.l2:\f3 e6 3.l2:\c3.
The knight is committed to c3 so 3...
a6!? would be an excellent choice.
Keep in mind that the Taimanov
is a self-sufficient system while in
some Kan lines it is best to develop
the knight on c6 switching over �o
the Taimanov. One such example IS
l.e4 c5 2.l2:\f3 e6 3.d4 cxd4 4.l2:\xd4
a6 5..ie2!?
7
 Introduction

players. That means your oppo­

nents will lack a model game to fol­
low and will have to play over the
board with their own capacity.
A more complex approach is
s . . . tt'lf6 6.0-0 Vf!c7 7.Vf!e2 d6 8.c4.
Again, I propose 8 . . . g6.

Perhaps the best option here is

s .. :�c7 (or s . . . tt'lf6 first) 6.0-0 tt'lf6
7. tt'lc3 tt'lc6.
Undoubtedly, the move order
with S.id3 is the most challeng­
ing system against the Kan. White
hopes to build up the Mar6czy
Bind with c4 and keep a firm grip Ivanchuk often plays it with
on the centre. A partial consolation the fresh positional idea of . . . tt'lhS,
for Black is that he is spared of di­ . . . ig7xc3 followed up by a blockade
rect attacks against his king. I offer on the dark squares.
two anti-dotes. If you hate long ma­
noeuvring in hedgehog pawn struc­ The most dangerous plan against
tures, you can play S . . . g6 ! ? followed the Taimanov has always been the
by . . . tt'le7 and . . . d7-d5. English Attack - l.e4 cS 2 . tt'lf3 e6
3.d4 cxd4 4.tt'lxd4 tt'lc6 5.tt'lc3 Vf!c7
6.ie3 a6 7.Vf!d2 tt'lf6

Thus you will be able to elimi­

nate the centre and gradually neu­
tralise White's lead in development. In 2 013, Black has developed a
This line is not popular among top new way of meeting it - 8. 0-0-0

8
 Introduction

.>le7! 9.f3 b5 10 .g4 tt::lxd4 ! ll . bd4

. I wrote this book in first
.l!i.b7 which brings terrific results. person, but it is a collabora­
White even began to avoid it and tive work.
switched over to 8.f4, but we are Every line has been checked
well prepared against it, too. and discussed by both authors. I
play these systems with both col­
In conclusion, the Sicilian devel­ ours while Semko Semkov de­
opment with 2 . . e6 has never been
. fends exclusively Black's side -
in a better theoretical state. It has first as a practical player, then as
become extremely popular among a theoretician.
the elite so we'll be seeing it very of­ We tried to answer for you all the
ten in the years to come. questions we thought important.

Alexander Delchev
April2014

9
10
 Pa rt 1

Th e C lassic a l S yste m w ith ,i e2

In this part, I consider the always popular system with i.e2 and short
castling. This is my favourite plan when I defend White's side. No matter
what is currently in vogue, this positional system will always be one of
the most testing set-ups against Taimanov/ Kan structures. It stakes on
natural development and firm control of the centre, which allows White to
choose between a wide range of plans on both flanks.

While some typical Black's set-ups with White's pawn wedge on e5 look
outwardly cramped, this very pawn could also be White's biggest weak­
ness. A lot of games witness a sudden turn of events caused by the under­
mining move t7-f6.
The secret of Black's vitality on so little space is the excellent efficiency
of his long-range pieces. His bishops often operate on the important dia­
gonals gl-a7 and hl-a8.

11
Pa rt 1. The Classical System with ie2

M a in Ideas

In the Taimanov, White aims for the

position on the following diagram :

l.e4 c5 2 .-li�f3 e6 3.d4 cxd4

4. tLlxd4 tLlc6 5.tLlc3 V!fc7 6 .le2 a6
•

7.0-0 tLlf6 8 .le3

•

Ta imanov 1

Black will also play here . . . tt:lc6

but first he'll gain an important
tempo by hitting the e4-pawn. Now
8.�f3 is not very impressive : 8 . . .
tt:lc6 ! 9.tt:lxc6 dxc6 10.'�� e 2 �d6 ! ?
ll.g3 tt:lf6 o r lO .eS VffxeS ll.l'!el V!fc7
�hS �e7 13.1'!xe6 g6. Thus the only
way to stay in the battle for the
This is the favourite approach of opening advantage is:
players who love the Scheveningen
with White. Sometimes they use the s.gel tLlc6 9.tLlxc6 dxc6 10.e5
refined move order 6.�e3 a6 7.�e2, gds n . .ld3 c5
aiming to avoid the Scheveningen
without . . . a6 which might arise after Ka n 1
6.�e2 tt:lf6 7.0-0 �e7. However, it
also has a counter-strategy - Black
can refrain from . . . a6 by 6.�e3 tt:lf6,
inviting the opponent to a theoret­
ical duel in the sharp line 7.f4 �b4
which is, however, dubious - see
Part 2 .
Our main line i n the Kan arises
after l.e4 c5 2 . tLlf3 e6 3.d4 cxd4
4.-li�xd4 a6 5.-li� c3 Wc7 6 . .le2 b5 The current status of this posi­
7. 0-0 .lb7 tion is roughly equal. Black's rook
12
 The Classical System with .te2

is very active on the d-file, the eS­

pawn is a double-edged asset as it
needs protection.

Let's see what happens if we try

to employ the same development in
the Taimanov. In line 1, Black can
play 7 . . . b5? ! 8.4Jxc6 dxc6 9.a4!

12 . . . l'l:b8 ! 13.g3 4Jf6 14 ..tf3 g6 !

15.4Jb6 eS (or 15 . . . l'l:d8)

The most topical line is 10. �xc6

bxc6 11. �b6 gb8 12. �xc8 �xc8
13.�d4
Taimanov 2

It turns out that we missed one

tempo for .. Jl:d8 so 9 . . ..tb7? los­
es owing to the knight's sac on bS.
The only way to keep the queenside
closed is 9 . . . b4, but then White ob­
tains long-term pressure on the
weak c- and a-pawns by occupying
c5 : lO.liJbl 4Jf6 ll.liJd2 .tcs 12 . .td3
eS 13. 4Jb3 .ta7 14.h3 aS 15.�e2 0-0 Currently, 13 . . . c5 looks a bit
16 .te3 .
.
more accurate than 13 . . . �c7. White
still preserves an initiative, but
A plus for the Kan: the plan with practice has shown that Black does
. . . bS works better. not experience serious problems .

In Taimanov 1, Black should opt Do not try 6 . . . 4Jf6 7.0-0 .tb4 in

for the Scheveningen, or for 8 . . . the Kan !
.ib4 when White's most challeng­
ing i dea is to thrust his knight to­
ward b6 by 9.�a4 �e7

Here, the pawn sacrifice 10.c4

41xe4 ll.cS 0-0 12 J'kl has faded
out of fashion since Black success­
fully frees himself with . . . d6, for in­
st ance:

13
 Part 1

It may look that Black has a fa­ veningen or the transition to the
vourable version compared to Taimanov with 6 . . . �c7 7. 0-0 ll:lc6.
the Taimanov, because he direct­ The only way for Black to refrain
ly threatens to win a pawn. In my from . . . ltlc6 is 6 . . . .ib4
opinion, the opposite is true be­
cause White has not committed his
bishop to e3 yet. Thus 8.i.gS ! i.xc3 ? !
( 8 . . . i.e7! i s a wiser decision, but if
Black wanted to play the Schevenin­
gen, he could have done it anyway)
9 ..ixf6 gxf6 10.bxc3 �xc3 1U'1b1 is
dubious for Black since his king will
never find a safe haven.

A plus for the Taimanov: the plan
with . . . i.b4 is more solid.
It seems that many White play­
ers prefer to avoid the Kan with
ltlc3. That can explain the burst of
popularity of the tricky move order
S.i.e2 ! ? (instead of S.ltlc3). Then
S . . . bS 6.a4 ! is unpleasant for Black
so he should play S . . . ltlf6 ! ? 6.ltlc3,
when objectively best is the Sche-
7.eS ! 7 . . . ltldS (7 . . . ll:le4 8 . 0-0)
8. 0-0 ll:lxc3 9.bxc3 i.xc3 10 .i.a3
�as (10 . . . �b6 ll.ltlf3 �as 12 .�d6
ltlc6, but 13J''1ab1 i.xeS 14.ll:lxeS
�xeS 1S.�d2 gives White a very
unpleasant initiative) 11 .i.d6 ll:lc6
12.ltlxc6 dxc6 13.�d3 ! .
The move order with S.i.e2
shows why Black should have in his
repertoire not only the Kan, but also
the Taimanov or the Scheveningen.

Weapon of Choice

I tend to prefer the Kan (compare positions Kan 1 to Taimanov 2). It can be
learned virtually in 2 hours and Black's play is fairly simple and straightfor­
ward. On the other hand, the same is true for White, too ! The pawn struc­
ture is symmetric and it is relatively easy to exchange everything and make
a draw. The Taimanov's strategically unbalanced game should offer more
chances against much weaker opponents.

14
Pa rt 1. The Classical System with .ie2
Ta i ma nov - S tep by S tep
t.e4 c5 2 . �f3 e6 3.d4 cxd4
4.�xd4 �c6 5. �c3 f!Jc7 6.J.e2
6.i.e3 a6 7.i.e2 is the better
move order unless White is plan­
ning 8.a3 or B.ltth l. After 6 . .ie2 ,
Black may choose the Scheveningen
without . . . a6, e.g. 6 . . . ltJf6 7. 0-0 ie7
which is a worthy alternative to the
Taimanov. If Black tries to reach the
same position after 6 . .ie3 ltJf6, he
must reckon with 7.ltJdb5 f!Jb8 8.f4
or 7.f4. Most experts avoid these
possibilities in favour of 6.ie3 a6,
when 7.ie2 leads to our main line.
6 .•• a6 7. 0-0
7.£4 bS is considered in Part 4
wh ile 7 . . . ltJxd4 8.f!Jxd4 bS is possi­
ble as well.
White can also play 7 . .ie3 ltJf6
8.a3 having in mind to attack 8 . . . ie7
by 9.g4 or 9.f4 d6 10.g4. This pro­
phylactic move is certainly not dan­
gerous for players who have a good
command of the whole TaimanovI
Kan arsenal. For instance, we'll see
below that 7.0-0 bS is best met by
8.ltJxc6 dxc6 9.a4 ! . So the most nat­
ural retort to 8.a3 is:
8 . . . b5
The only flaw of this move is
perhaps the possible draw after:
9.ltJxc6 dxc6 10.f4
10.0-0 ib7 ll.f4 ie7 12 .e5 Ei:d8
13.f!Je1 ltJd7 leads to the same struc­
ture : 14.f!Jg3 0-0 15.Ei:ae1 cS 16.if3
.ixf3 17.Ei:xf3 f!Jc6= , Baramidze­
Ribli, Austria 2006.
10 ... ib7 1l.e5 Ei:d8 12 .id3 ltJd7
(12 ... �d5 13.ltJxd5 cxdS 14.i.d4 i.cS
15.c3 0-0 16.f!Je2) 13.f!Jg4 cS 14. 0-0
g6
15
 Part 1

15. 'Wh3 (or Black will advance his
queenside pawns) 15 . . . �e7 16.'Wh6
�f8 17.'Wh3 .
7 ••• t0f6
7 . . . b5 is an attempt to get a posi­
tion from the Kan. However, White
has a straightforward way of ob­
taining the initiative : 8.lt:lxc6 ! dxc6
9.a4!
White carries on the same plan
after ll . . . aS, e.g. 12 .�d3 �e7 13.'We2.
12 .�d3 eS
Alternatively:
12 . . . 0-0 13 .'We2 eS 14.tt:lb3 �g4
1S.'We1 �e7 16.�c4 c5 17.f3;t, Heber­
la-Zhigalko, Instanbul 2 0 05.
12 ... h5 13.'We2 tt:lg4 14.tt:lf3 �d6
15.);13;!;, Abu Sufian-Hossain, Dha­
ka 2013.

13.tt:lb3 �a7 14.h3 aS 1S.'We2 0-0

16.�e3

The only way to avoid seri­

ous trouble now is to keep the play
closed with :
This set-up promises White
9 . . . b4 lasting pressure because Black's
queenside pawns will be a perma­
9 . . . �b7? loses owing to the
nent cause of concern . The game
knight's sac on bS.
Heberla-Firman, Warsaw 2 0 06,
9 .. J�b8 10.e5! is also awkward.
went 16 ... �e6 17.ha7 'Wxa7 18 .�c4
Against I.Sokolov, I replied 10 . . . tt:le7
�feB (18 . . . tt:lh5 19.�xe6 tt:lf4 2 0 .'Wc4)
and went on to draw after ll.�f4 fS
19.he6 �xe6 2 0 .�adU.
1 2 .exf6 'Wxf4, but ll.axbS ! is very
unpleasant. White opens the a-file
and the endgame after ll . . . axbS
12 .'Wd6! is difficult. 12 . . . 'Wb7 13 .'Wd3
tt:lg6 14.tt:le4! is even lost.
The other recapture ll . . . cxbS
does not save the day either: 12 .tt:le4
tt:ld5 13 .'Wd4 with domination in the
centre.
10.tt:lb1 tt:lf6 n.tt:ld2 �cs

16
 The Classical System with .ie2
After 7 . . . ftlf6, White has three
main continuations: A. 8.a3 ; B.
s . @h 1; c. B . .ie3
8 . .ig5 is inconsistent. Apart from
8 . .ie7, Black can equalise with typ­
. .
ica1Taimanov means: 8 . . . .id6 9.<i>h1
L2Jxd4 10.\Wxd4 .ie5 ll .\Wd3 b5 12 .f4
.�.xc3 13.bxc3 .ib7 or the minimalis­
tic 8 . . . ftlxd4 9.\Wxd4 ic5 10 .\Wd2 b5
ll..if4 e5 12.ftld5 ftlxd5 13.\Wxd5 .ib7
14.\Wxe5+ \Wxe5 15 .he5 0-0, Brkic­
S mirin, Rijeka 2010.
A . 8.a3
Sometimes White opts for this
move in order to avoid 8.ie3 ib4.
Of course, this is a meager rea­
son for playing second rate varia­
tions. Any system against 8.<i>h1 is
even better against 8.a3. In many
Taimanov set-ups, a3 could prove
to be a superfluous move. Also in
the Scheveningen, Black faces more
problems when defending against
a4 instead of a3.
8 . . . �xd4 9.ti'xd4 .1d6
lO.�hl
10 .g3 could be attacked later
with h5, for instance, 10 . . . b5 ll . .if3
ib7 12 .\Wd3 �c8 13.�e1 .ie5 14.id2
h5+, Alexandridis-Miladinovic, Ka­
vala 2001.
10 .te5 U.ti'd3 0-0!
••• 12.f4
hc3 13.bxc3 d5 14.exd5
Stein-Suetin, Moscow 1964 saw
14.e5 �e4 15.c4 \Wxc4 16.\Wxc4 dxc4
17.hc4 .id7= . Black could aspire to
an edge with 15 . . . b5 ! ? 16.cxd5 exd5
17.ti'xd5 if5+.
14 .•. exd5 15 .te3
•
Black has the initiative. All his
pieces have good prospects. Per­
haps best is 15 .. J:'le8 ! 16.id4 ftle4
while G. Kuzmin-Andreikin, Alush­
ta 2 0 04, saw 15 . . . .id7 16.id4 ftle4
17.c4 if5 18.\We3 �feB 19.cxd5
\Wc2 = .
B . S . �hl �xd4
Black can transpose to the Sche­
veningen by 8 . . . .ie7 9.f4 d6, hav­
ing won a small opening victory.
The point is that by having commit-
17
 Part 1

ted his king to h1, White deprived

himself of one the most danger­
ous system against the Schevenin­
gen which is based on the quick ma­
noeuvre �d1-e1-g3 and Ei:ae1 while
the king remains on gl.
I suggest to stay in the Taimanov
waters. Our first task will be to find
an active occupation to the dark­
squared bishop. Besides the text, we
often see 8 . . . .tb4. Then 9.�d3 �d6
10.tt'lxc6 �xd3 ll.cxd3 dxc6 leads Bl. ll.f4 ; B2. ll.igS ! ?
to an equal endgame, but 9.-tgS ! is
more testing. Then 9 . . . �d6 10.tt'lxc6
B l . ll.f4 .ib7 12 . .ifJ
�xd1 1l.tt'lxd1 dxc6 12 .e5 tt'ld7 13.f4
as in Smislov-Razuvaev, USSR
1975, leaves Black in a cramped al­
beit solid position.

9.Ybd4 .ic5 10.�d3

Alternatives:
10 . .tf4 .txd4 ll . .txc7 .txc3
12 .bxc3 tt'lxe4 13.c4 dS=.
10 .�d2 0-0 (Black should not
provide the opponent with a lever
for an attack with 10 . . . h6) 1l .�g5
d6 (11.. . .td4 12 . .td3 .txc3 ! ? 13.bxc3
d6 14.f4 dS is more ambitious, but
risky.) 12 . .td3 tt'ld7 (12 . . . b5 13 .�g3
.tb7 14.ih6 tt'le8 = , Jakovenko-Lau­
tier, France 2 0 07) 13.f4 (13.�g3 bS
14.tt'le2 tt'leS 15.f4 tt'lxd3 16.cxd3 fS)
13 . . . id4+, Van Oosterom-Giri, En­
schede 2009.
10 •.• b5
lO . . . hS is too committing. You
could opt for such plans if you des­
perately needed a win.
18
12 . . . h5
This move enables . . . tt'lg4 and
builds up tension.
12 ... 0-0 ! ? is simple and sol­
id: 13.e5 tt'le8 14.ib7 (14.tt'le4 .te7
15.ie3 fS 16.tt'ld2 Ei:c8= , Oleksienko­
Shaposhnikov, Minsk 2 006) 14 . . .
�b7 15.a4 (15.f5 f6=) 1 5 . . . b4 16.tt'le4
.te7 17 . .te3 �c6 temporarily gains
space, but White is unable to keep
his achievement : 18.c4 bxc3 19. bxc3
dS 20.tt'ld2 tt'lc7= , Ye Jiangchuan-J.
Polgar, Prague 2 0 0 2 , or 18.tt'lg5 g6
(18 .. .f5=) 19J'!ad1 d6 20 ..td4 dxeS
2l .ixe5 tt'lf6 22 .�h3 hS+, Grischuk­
J. Polgar, Linares 2001.
 The Classical System with i.e2
13.e5
13 .i.d2 li:\g4 14.'<Mfe2 d6! restricts
White's activity in the centre.
13 .i.e3 ! ? i.e3 14.'<Mfe3 aims to
weaken Black's dark squares (d6,
c5, b6) . Therefore, 14 . . . d6! is called
for. (14 . . . b4? ! 15.li:\a4 d6 16.c3)
15 .E!ad1 (or 15.a3 '<MfcS ! 16.'<Mfd2 h4
17.E!ad1 0-0- 0 ! ) 15 . . . b4 ! 16.li:\a4 (16.
1Lle2 0-0) 16 . . . 0-0 17.'<Mfd4 (17.c3
,1!W6) 17 . . . a5 ! 18.E!fe1 E!fd8 ! = (Gross).
13 ••• �g4 14 .ib7 •
White does not get an advantage
by 14.li:\e4 i.xe4 (14 . . . i.e7 15.li:\d6+ ;
14 . . . E!b8 ! ?) 15.'<Mfxe4 E!c8 16.'<Mfe2
l/ih6 17.a4 li:\f5 18.axb5 axbS 19.i.d2
h4 20 .i.g4 '<Mfc6 = .
14 •.• '<Mfb7 15.�e4
In several games White scored
well with 15.a4 ! ? b4 16.li:le4 l:!c8
17.c4. After fixing the queenside, he
could turn his attention to Black's
king. I suppose that the best an­
swer is 15 . . . 0-0 ! ? 16.li:\e4 i.e7 17.h3
l!lh 6 , maintaining tension all over
=
the board.
15 l:!c8 ! ?
• .•
Black recognises 16.i.e3 a s the
main positional threat of the oppo­
nent. Database statistics show 61%
in Bl ack's favour after 15 . . . l:!c8 ! ?
Let's ponder over the frequently
seen 15 . . . i.e7.
White has two reasonable plans :
to b re ak through the queenside by
pushing b3-c4, or to blockade it by
c3-b4, trying to organise an attack
on the other wing. The latter is rath­
er annoying for Black.
16.i.e3
16.b3 is well met by 16 . . . l:!c8 !
(Black is trying to dissuade the op­
ponent from pushing 17.c4. In that
event 17 . . . 0-0 would find White un­
developed for active play. ) 17.i.b2
li:\h6 18.'<Mff3 '<Mfb8 19.l:!f2 li:lfS. Black
accomplished the thematic knight's
manoeuvre-see game 1 Lutz­
Grabliauskas, Pula 1997.
16 . . . l:!c8 17.i.d4 !
In the majority of games White
chooses 17.c3. Then the march of
the h-pawn is very efficient: 17 . . .
h4 ! 18.h3 (18.i.d4? ! h3 ! 19.g3 fS !
20.exf6 li:lxf6 ! 21.i.xf6 gxf6 ! ) 18 . . .
li:\h6 19.l:!ad1 li:lfS 2 0.i.f2 0-0.
White's pawn chain on the kingside
is paralysed for good and could be
attacked, e.g. 21.'<Mff3 l:!c7 2 2 . l:!d3 f6 !
17 . . . li:\h6 (17 . . . h4 18.�f3 '<Mfc6
19.'<Mfg4 '<Mfe4 20 .c3 !;!;) 18.c3 ! h4 (18 . . .
li:\f5 19.�f3 ! '<Mfc6 2 0 .l:!adU) 19.'<Mff3;!;.
White can follow up by b4, i.cS.
Now that White cannot play
16.i.e3, he probably should conti­
nue with :
19
 Part 1

18.ic3 ltlh6 ! 19.f5 tt:lfS 2 0 .�xf5

exfS 2 l .ltld6 ixd6 2 2 . exd6 cj{fB
23.�e7 �h6 24.\WfS f6 25.\Wxd7 (25.
�xd7 �g6) 25 ... \Wxd7 26.�xd7 �xc3
27.bxc3 fS ! = ;
18 .c3 ltlh6 19 .ie3 ltJfS 2 0 .id4
h4 21.\Wf3 0-0 2 2 .�d1 �fd8.

JJ2 . u . .tgs .tb7 t2 .f4 .tb4 !

16 .td2
•

16.a4 is inconsistent. When the

queenside is opened, Black can
safely castle, because he can easi­
ly organise counterplay should the
opponent attempt to assault his
king. I propose 16 . . . ie7! ? 17.axb5
axbS 18.�a5 (or 18 .h3 ltlh6 19�ie3
0-0 2 0 .ltlg3 \Wc6 = , Tirard-Delchev,
Cappelle ·1a Grande 2006) 18 . . . b4
19.id2 \Wc7 2 0.�fa 1 \Wc6 ! = . f4-f5 is
White gets an edge after 12 . . .
no longer a threat.
b4? ! 13.e5 tt:ldS 14. tt:lxd5 ixdS
15.if3 ixf3 16.�xf3 d6 17.�afl!t, as
16 . . . .te7! 17.gael
in Berg-Brynell, Gothenburg 2006.
Alternatively: 17.a3 \Wc6 ! 18.ib4
13 . .tf3
fS= , Delchev-De la Riva, Benasque
1997; 17.b3 \WdS.
13.ixf6 defines White's plans
too early. In Dolezal-El Debs, Sao
17 . 0-0oo.
. .

Jose do Rio Preto 2 007, Black put

his rook to g8 and stood well : 13 . . .
According to my analysis, Black
gxf6 14.a4 ixc3 15.bxc3 �g8 16.g3 fS
holds firmly in this position. The
17.if3 fxe4 18.ixe4 ixe4 19.\Wxe4
biggest danger for him is 18 .f5, but
\Wc6 = .
18 . . . tt:lxe5 19.\Wg3 f6 20 . .ic3 �c4
neutralises the pressure.
13 . . . .bc3 14.bxc3 gc8 15 .txf6
•

Black has successfully tested in

practice 17 . . . \Wc6 ! ? 18 .'.f9f3 fS ! = , Yu 15.a4? ! \Wxc3 16.axb5 \Wd3
Shaoteng-Ye Jiangchuan, Yongchu­ 17.cxd3 axb5 18.�fb1 ic6 was good
an 2003. Other options after 17 . . . for Black in Zezulkin-Tregubov,
\Wc6 ! ? are : Krasnodar 1998.

20
 The Classical System with .ie2
15 . . . gxf6 16.gadl = .
Black can gain space o n the
kingside by 16 . . . h5 and hope to use
White's split pawn structure on the
ot her wing.
C. 8.J.e3 J.b4
Black attacks the e4-pawn. If
White defends it, he loses the initi­
ative and could become worse:
a) 9.if3?! lt:leS 10.if4 ixc3
ll. bxc3 d6+.
b) 9.id3?! lt:leS leads to positions
from Part 3 with an extra tempo for
Black. After 10.lt:lde2, he can try to
punish immediately the opponent
by 10 . . . lt:lfg4 (10 . . .ixc3 ll .bxc3 dS= )
ll. if4 g S 12 .ig3 hSoo, but probably
the best way is to continue develop­
ment by 10 . . . 0-0 ! ? ll .h3 d6.
c) 9.f3? ! Black carries on ... d7-
d5 with excellent play. This plan
co uld be executed in different set­
ti ngs: 9 . . .ixc3 10 .bxc3 dS or 9 . . . d5
lO.exdS exdS . Most players prefer
9 . 0-0 as more flexible. Depend­
. .
i ng on White's response, Black will
ch oose whether to give firstly his
bishop for the c3-knight or push d7-
d5 and retreat to d6 in future.
Obviously White should ig­
nore the threat to e4 and coun­
ter-attack, but how?
The straightforward approach :
d) 9.f4 ixc3 10 .bxc3 lt:lxe4
11.id3 proves to be insufficient af­
ter ll . . . dS ! (ll . . . lt:lf6 12 .'1We1 is not so
clear) . Black's knight on e4 is a good
defender.
e) 9.lt:lxc6 bxc6 10.f4 ixc3
ll .bxc3 is more testing.
Now after 11. . . lt:lxe4 (In Ran­
tanen-P. Cramling, Helsinki 1991,
Black chose the modest: ll ... cS
12 .e5 lt:ldS 13 .id2 c4 14.l'!b1 0-0 = . )
12.if3 [12 . .id3 lt:lf6 13.'1Wf3 (13 .id4
cS 14 . .ixf6 gxf6 lS.fS eS+) 13 . . . d6
14.id4 l'!b8 15.'1Wg3 'it>f8 16.c4 cS
17.ic3 lt:le8 18.l'!ael fSoo, Oleksien­
ko-Swapnil, Kolkata 2012], Black is
forced to cover the critical diagonal
a3-f8 by 12 . . . lt:ld6 (12 . . . lt:Jf6 13.ic5
dS 14.'\Wd4�) . The game Pulkkinen
-Suetin, Espoo 1989, went 13.l'!b1
l'!b8 14.l'!xb8 '\Wxb8 1S.ic5 lt:lb7,
when White's initiative is bound
to evaporate : 16.'1Wd4 (16.'1Wb1 '\Wc7
17 . .ib6 '1Wd6) 16 . . . lt:lxc5 17.'1Wxc5 'IWbS
18.'1Wd6 f6 ! .
21
 Part 1
f) 9.tt'lxc6 bxc6 10.'Wd4 (10 .'Wd3
d5 11.exd5 cxdS= ; 10 .id3 dS =) 10 . . .
cS ll.'Wc4 has always been a side­
line.
That accounts for the fact that
Black has not established yet his
best defence.
I faced it in Delgado Crespo­
Delchev, Sort 200B and chose . the
natural 11.. .ib7? ! 12 .a3 (12 .if3
'WeS = or 12 . . . e5=) 12 . . . d5
12 . . . ixc3 is more popular, but
I'm afraid that the position af­
ter 13.'Wxc3 is in White's favour.
The renowned Taimanov expert
Pia Cramling chooses here 13 . . . 0-0
which is quite safe, but leaves the
opponent with a small edge after
14.f3 �feB 15.ig5 tt'leB 16.�fe1 (16.
eS dS) 16 ... 'Wc6 17.�ad1 dS 1B.exd5
exdS:t, De Ia Villa-Cramling Cor­
doba 1995. Taking on e4 is out­
right dubious : 13 . . . ixe4 (13 . . . tt'lxe4
14.'Wxg7 0-0-0 15.f3 �hgB 16."Wxf7
�dfB 17.'We7±) 14.f3 ig6 15.'Wxc5
'WxcS 16.ixc5 hc2 17.�ac1 ia4
1B.id6 'it>dB 19.�c5 tt'leB 20 .ig3
ic6 2l.�a5�, Shtyrenkov-Eljanov,
Novi Sad 19B9.
13.exd5 exd5 14.'Wf4
Black's pawn centre is dangerous
with queens : 14."Wh4 hc3 15.bxc3
22
0-0 16.id3 (16.�ab1 �feB 17.id3
'Wc6 1B .hc5 'WxcS = , Cawdery-Gre­
tarsson, Reykjavik 2014; 17 . . . c4 ! ?)
16 . . . 'Wc6 17.f3, Garbisu-G. Hernan­
dez, Villalba 1993, 17 . . . c4= .
1 4 . . . 'Wxf4 15.hf4 hc3 16.bxc3
0-0. Here my opponent chose the
straightforward 17.�ab1 ic6 1B.�b6
ibS 19.ixb5 axbS 20.�xb5 tt'le4 = .
However, 17.c4 ! would have given
White somewhat preferable chanc­
es because of the weakness of the
cS-pawn.
I think that Black should look for
improvements earlier:
11 . . . 0-0 ! 12 .if3 d6
Black is behind in development
so it is logical to keep the centre
closed. He had no problems in Na­
kar-Roiz, Acre 2013, which went
13.a3 iaS (13 . . . hc3 14.'�'xc3 ib7=)
14.b4 (14.e5 dS 15.'Wxc5 tt'ld7) 14 . . .
cxb4 15."Wxc7 hc7 16.axb4 ib7= .
White discovers the sore point
in Black's position-b6, and the cen­
tral dark squares in general. Note
that the move order 9.tt'lxc6 bxc6
10 .tt'la4 is probably inaccurate, be­
cause it offers Black additional op-
 The Classical System with �e2
tio ns as 10 . . . 0-0 (10 . . . �d6 11.tt:lb6
i\bB 12 .g3 ! looks better for White)
u . tt:lb6 !=1b8 12 .tt:lxc8 !=1fxc8 13.ha6
!"\dB 14.�d3 �d6. This line was
popular in the 70s, then it has faded
out of fashion. It is considered satis­
factory for Black, e.g. 1S.f4 eS 16.b3
exf4 17.�d4 f3 18.eS �xeS.
9 . . . .ie7!
White was threatening c4-cS.
Unfortunately, 9 . . . bS runs into
1 0 . tt:lxc6 dxc6 11.�cS ! ;t hcS (11 . . .
bxa4 12 .�b4 cS 13.�a3 ! tt:lxe4 14.�f3
s�.b7 1S.E1e1 tt:ld6 16.!=1eS±) 12.tt:lxcS;t.
The old move 9 . . . 0-0?! gives
White a terrible initiative after
1 0 . c4 ! tt:lxe4 (10 . . . �d6 ll.g3 tt:lxe4
1 2 .�f3 fS 13.cS �eS 14.tt:lb6 !=1b8
15.!'1c1�) ll.cS dS (ll . . . d6 12 .tt:lc6
VJlc6 13.cxd6 �xd6 14.�c2 fS 1S.!=1fd1
VJ1e7 16.tt:lb6 !=1b8 17.!=1acl±) 12.tt:lxc6
fixc6 13.tt:lb6 !=1b8 14.!=1c1 �aS
(14 . . . �d7 1S. tt:lxd7 �xd7 16.!=1c4±)
1 5.�f4±.
Cl. 10.c4; C2 . 10.tt:lxc6
1 0 .tt:lb6 should be answered by
10 . . . !=1b8. Then White can try to
surprise us with the tricky move
order: ll. tt:lxc6 bc6 12.tt:lxc8 �c8
13.c4 tt:lxe4 14.�d4 tt:lf6 1S.cS, but
this position is good for Black. For
instance, he can prepare . . . !=1b8-b4
by 1S . . . aS 16.�c4 tt:ldS 17.�g7 �f6
18 .�g3 !=1b4.
Cl. 10.c4 ttJxe4 ll.c5
ll.tt:lxc6 bxc6 12 .�d4 tt:lf6 13.tt:lb6
!=1b8 14.cS is a decent alternative.
Black must free himself immedi­
ately because after 14 . . . e5 15.�c4 0-0
16.!=1fd1, White has a bind - 16 . . . tt:ldS
(16 . . . dS 17.cxd6 ixd6 18.!=1acl±)
17.tt:lxdS cxd5 18.�xdS !=1xb2 19 .ic4
Wh8 20 .ib3 fS 2 1 .�d3±. So he fol­
lows up with :
14 . . . d6 1S.cxd6 hd6 16.tt:lc4
ixh2+ 17.Wh1 cS 18 .�xcS �xeS
19.hcS ic7 (19 . . .if4 is more risky
and materially unbalanced after
20.if3 �b7 21 .id6 hf3 22 .hb8
hg2+ 23. Wxg2 hb8 24.!=1ac1 if4
2S.E1c3 We7oo) 20.!=1ac1 tt:le4 and
White cannot extract much from
the a3-f8 diagonal, e.g. 2 1.�a3 ib7
2 2 . Wg1 fS 23.!=1fd1 !=1d8 = .
11 . . . 0-0 12.gc1
12 .�f3 is countered by 12 . . . tt:lgS !
23
 Part 1

The text move is more straightfor­

ward.

12 . . . gb8 ! 13.g3 (preparing a

retreat square in the event of if3
tt:lgS) 13 tt:lf6 14.if3 g6 !
•••

16. tLlb3 d6 17.cxd6 ixd6

Black aims for . . . e6-e5 so he
takes fS under control. White has
been unable to find gaps in Black's
armour. His compensation is only
sufficient for equality.
15. tLlb6
18.tLlc4
White attempted to improve
with 18.ix:c6 bxc6 19.�f3 in Kam­
sky-Nisipeanu, Sofia 2 007, where­
as John Emms suggests 19 . . . \WdB ! ?
2 0 .tt:lxc8 e4.

Alternatively: 18 .•• ie7 19.ib6 \Wd7 20.tLlc5

15.Eie1 (15.\Wb3 eS ! ; 15.\Wd3
Eld8=) 15 ... e5 16.tt:lb3 d6 17.cd6 .id6
18.ia7 Ela8 19 . .ib6 \We7 20 . .ic6 bc6
2 l.Eic6 ib4 22 . .ic5 .icS 23.tt:lbc5
Eld8 24.\Wb3 .ih3 with active play;
It is dangerous for White to
keep on queens with 2 0 .\We2 e4
2l..ixe4 tt:lxe4 2 2 .\We4 \Wh3 = , Pol­
zin-Heissler, Germany 1999 .

15.ig5 Eld8 16.tt:lb6 h6 ! ? 17.ix:h6

dS 18.Eie1 eS 19. tt:lxc6 bxc6 2 0 .\We2
tt:ld7 2l..ixd5 tt:lxcS 2 2 . tt:lxc8 ElxdS
23.tt:lxe7+ \Wxe7 24.Eied1, Brenjo­
Sarenac, Belgrade 2009, 24 . . . e4 ! = .

15 .•• e5

Karpov introduced 15 . . . Eid8 ! ?

when 1 6 . .ig5 h 6 i s the above-men­
tioned game Brenjo-Sarenac.
16 .�a4 eS 17.tt:lb3 dS 18.cxd6 Elxd6 The play is balanced. Black can
19.tt:la5 .ih3 20.Eife1 e4 is also fine. trade queens or aim for complica-

24
 The Classical System with .ie2
tions with 20 .. .'�h3 2 l..ig2 �h6
22Jle1 .ig4.
C2. 10.tbxc6 bxc6 ll.t'Llb6
Elb8 12 .t'Llc8 �c8
The position is very unbalanced
strategically. Black has more pawns
in the centre and the semi-open b­
file for his rook. He only needs a cou­
ple of moves to consolidate ( . . . 0-0,
.. . �c7, ... d5) and his pieces will be
perfectly co-ordinated. White bases
his hopes on a kingside attack. For
the moment, he has some spatial ad­
vantage and the initiative, so he must
play actively.
13 .id4
.
a) 13.e5 is less popular because
after 13 . . . ll'ld5 the bishop has to re­
treat to its initial place:
14 . .ic1 (the difference with the
main line is that after 14 . .id4 c5
Wh ite has not the e5-square for the
bishop and has to play 15.c4 cxd4
1 6.cxd5 �c5 ! 17.i.f3 i.g5 ! 18.Elb1
exd5 ! = or 17.d6 i.f8 ! .
14 . . . .ic5 !
15.c4
15.�d3 is best met by 15 . . .'�c7!
(avoiding 15 . . . 0-0 16.�g3) 16.�e4
(16.�g3 i.d4 !), when 16 . . . 0-0
17 . .ixa6 f5 18.exf6 ll'lxf6 19.�e2
ll'ld5 2 0.g3 !'lxf2 21.!'1xf2 Elf8 2 2 .i.f4
i.xf2 + 23.�xf2 e5 is playable, but
it is better to keep more tension
by 16 . . . a5 17.a3 0-0 18.i.d3 f5, e.g.
19.�c4 �a7 2 0 .c3 ll'le7 2 l .i.c2 ll'lg6
2 2 .�e2 a4 23.Wh1 f4't, Corrales­
Quesada, Las Tunas 2 009.
15 . . . ll'le7 16.Wh1
16.b3 �c7 17.i.b2 faces Black
with a choice. He can challenge the
opponent to enter unbalanced po­
sitions with 17 . . . d6 18 .ed6 i.xd6
(19.h3 .ih2 2 0 . Wh1 .ie5) 19.�d4 c5
(19 . . . ll'lf5 2 0 .�e4 i.xh2+ 2 1 . Wh1 c5
2 2 .g4 ll'ld4 23 .i.xd4 cxd4 24.f4 i.g3
25.Wg2 i.h4 26.!'1ad1 i.e7=) 20.�xg7
.ixh2+ 21 .Wh1 !'lg8 or 19.i.xg7 !'lg8
2 0 .i.c3 i.xh2+ 21.Wh1 i.e5 2 2 .i.xe5
�xeS. A more restrained approach
is to castle and try to provoke f2-
f4 before breaking in the centre
with . . . d6, for example, 17 . . . tt:lg6 ! ?
18.Wh1 0-0 19.f4 o r 1 7. . . 0-0.
16 . . .'�c7 17.f4 0-0 18 .b3. I played
here 18 . . . a5, but this move is not
necessary. It would be natural to
25
 Part 1
open the centre by 18 . . J'l:fd8 19 . .ib2
lt:lfS (or 19 . . . d6 20 . .id3 dxeS 2l.i.xe5
.id6 2 2 .�c2 lt:lg6 ! =) 2 0 J'l:f3 dS ! t .
b ) 13.�d4 i s rarely seen. I an­
swered it with 13 . . . �c7! ? 14.e5 lt:ldS
15.b3, Asrian-Delchev, Evry 2 008,
when 1S ... c5 ! would have seized the
initiative.
13 . . . c5
13 . . . �c7 is Miladinovic's pet line.
This variation is very flexible, but
with accurate play White can get a
small, but lasting edge due to his
better pawn structure. See game 2
Adams-Banikas, Porto Carras 2011
for more details.
14 . .tes gb6
15.�d3
Alternatively, 15.b3 lt:lxe4 (White
is slightly better after 15 . . . 0-0
16.�d3 d6 17 . .ib2.) 16.hg7 !!g8
17 ..ib2
1 � . .ieS.id6! 18 ..ig3 lt:lxg3 19.fxg3
.
.ieS Is equal. The game Radjabov­
Cori, Tromso 2013, went 20.!!b1
(2 0 ..ih5 l"lg7 2 1.c3 lt>e7=) 20 ... \t>e?
26
2 l..id3 !!gS 2 2 .�f3 �g8 23.!!be1
�g7 24.Wh1 aS 25.!!e3 f6= .
17 . . . if6 ! ?
1 7. . . !!d6 18 .id3 c 4 19.bc4 �c6
2 0.g3 lt:lgS stumbles into 2 l.f3 hS
2 2 . @h1 ! . Instead, a member of the
chesspublishing.com forum pro­
posed to play 17 . . . c4 ! immediate­
ly. Capturing on c4 is bad, due to
18 .hc4? �c6 19.g3 icS 2 0 .id4
lt:lgS; 18.if3 lt:lgS; 18 .id4? !!d6,
18 .ih5 �c6 19.�f3 lt:lf6 2 0 .�c6 !!c6
2 l ..if3 dS are not dangerous either.
However, White should be slight­
ly better with a sound central stra­
tegy: 18.�d4 c3 19.ic1 �c6 (19 . . .
!!b4 2 0 .�d3 icS 2 1.�h3 �c6 2 2 .a3
!!d4 23.b4 ib6 24.ie3 lt:ld2 25.hd4
.ixd4 26.g3±) 20 . .if3 dS 2l.ie3 !!b4
2 2 .�d3 lt:lgS 23.ie2.
It is safer to abandon the at­
tacking plans and opt for the good
positional idea of exchanging the
dark-squared bishops:
18.ixf6 lt:lxf6 as in game 3 Bali­
nov-Sommerbauer, Tweng 2007.
15 . . . d6 16 . .tc3 0-0 17.b3
Sadvakasov played against me
17.�g3 . It was a novelty in 2 004. I
answered 17 . . . �c6 18.!!fe1?! �xe4 !
 The Classical System with �e2

and won. Ribli suggested 18.i.f3 e5
19J�fel c4=. Later games featured
17 .d518.exd5 exd5 19.b3 !'1e6=.
. .
19...!'1d8 does not solve all the
problems: 20.�g3 (Black is intend­
ing to double his rooks on the d­
file so White must hinder this plan.
20.!'1adl leads to simplification fol­
lowing 20...f6 21.i.g3 tt:lf4 22.�f3
tt:lxe2+ 23.\Wxe2 c4=; 20.�h3 los­
es control over d6 - 20...�f6 21.hf6
tt:lxf6 22.!'1adl !'1bd6=)

17 .'e«c6!?
••

This move keeps more tension

than:

17...d5 18.ed5 White is slightly better after ei­

ther 20...i.f6 21.i.xf6! tt:lxf6 22.!'1adl
Or 18.e5 tt:Jd7 19.f4 c4 20.bxc4
or 20...f6 21.i.b2 i.d6 22.�f3 �c7
L\c6 21.�h3 tt:lb6 22.i.d3 g6 23.!'1f3 23.g3 i.e5 24.ie5 �e5 25.ic4;!;, Ba­
(23.f5 exf5 24.!'1xf5 tt:lxc4 25.e6 fxe6 rua-Muhren, Arnhem 2007.
is a desperate attack without seri­
ous chances to succeed.) 23...!'1e8oo. 20.hf6 tt:Jxf6 21.!'1adl �c7
Black can be optimistic about his
position, Nisipeanu-Delchev, Beni­
dorm 2005.

18... tt:Jxd5 19.i.e5

19.i.d2 tt:Jb4 20.�c4 !'1d6 (or 20...

ild8=) is equal: 21.i.c3 (21.hb4 cxb4
22.�xb4�xc2 23.ha6!'1xa624.�xe7
ilxa2 25.!'1xa2 �xa2, draw, Kasparov­
Anand, Frankfurt 2000) 21...!'1fd8
22.!'1fcl, Arizmendi-De la Riva, Se­ This position occurred in seve­
ral games. White commonly trip­
ville 2004, when Black should either
activate his passive dark-squared pled his heavy pieces on the d-file
and Black kept the balance by ex­
bishop with 22...i.g5 or bar the long
diagonal with 22...tt:lc6 23.i.d3 tt:Jd4. changing one of his weak queenside
pawns with ...a6-a5-a4 or ...c5-c4.
19...i.f6 However, in one of my games as

27
 Part 1

White I carried on the manoeuvre 24.�g3 g6 25..ial lt:lb6 26..id3

22.�d2! h6 23.�a5! which main­ lt:ld7=) 2l...�a4! .
tains some pull.

18 .if3
•

The endgames after 18.e5 lt:ld5

19.exd6 .b:d6 20..if3 �b5 2U!adl
�xd3 22J'!xd3 lt:lb4 23..b:b4 cxb4
or 19..if3 �b5 20.�xb5 axb5 2l..ib2
lt:lb4 22.exd6 :i!xd6 23..idl .if6 are
equal.

18 . . . lt:ld7

18...c4 is also interesting. Then 20 . . . cxb3

19.bxc4 lt:ld7 20.e5 �a4 2l..ie4 g6
22.exd6 :i!xd6 23.�e2 .if6 24..b:f6
lt:lxf6 25..id3 lt:ld7 gives Black good
counterplay so White should prob­
ably try 19.�e3 e5 with sufficient
counterplay on the queenside.
19.�e3 leads to similar positions
after 19...:1!c8 20.:1!adl (20..ie2 .if6)
20...c4 2l..id4 (2l.b4 �a4=) 21...
:i!b7 22..ie2 cxb3 23.cxb3 :i!bc7.
Black can also keeps the queenside
closed with 19...h6 20J!adl .ig5
2l.�e2 lt:le5.
19 . . . c4
Or 19...:1!c8 20.�e3 c4 2l.b4 (21.
ie2 cxb3 22.cxb3 �b7 23..ic4 :i!bc6
20...:1!c8 2l..id4 :i!b7 22..ie2
cxb3 23.cxb3 :i!bc7=.
21.cxb3 V9 b7
Black has no serious problems
here as long as he keeps control over
the c-file. His play is not based on
forced variations so you can choose
other moves, too. For example:
2l....ig5 22.�xg5 �xc3 23.:1!cl%'b4
24.�e3 �aS 25.:1!c2lt:le5 26..ie2 :i!c6
27.:1!xc6lt:lxc6 28.:1!dl :i!d8=.
After the text, the game may con­
tinue with either 22..ib2 :i!c8 23.:1!cl
:i!bc6 24.:1!xc6 :i!xc6 25.:1!cl :i!xcl+
26..ixcl= (26.�xcl tt:le5), or 22.�d4
tt:le5 23..ie2 .if6 24.�e3 :i!c6 25..ial
V9c7 26.:1!cl :i!c8 27.:1!xc6 V9xc6 28.f4
lt:ld7 29.e5 idS=.

28
Part 1. The Classical System with i.e2
Kan- S tep by S tep
t.e4 c5 2 tL!
. f3 e6 3.d4 cxd4
4.tL!xd4 a6 5.tL!c3 'f!Jc7 6 . .ie2
If White aims to get the Sche­
veningen or the Taimanov, he
should use the move order with
5.ie2! ?.
Black's normal options here are
5 . . .'�c7 6.0-0 tt:lf6 7.4:lc3 d6 or 7...
Cilc6 - see the Taimanov chapter.
A more sophisticated version of the
Scheveningen- without ...'f!!c7, may
be reached after 5...4:lf6 6.lt':lc3 d6.
White can try to be original with
6.�f3, but Black can benefit here
from having saved ...d6, e.g. 6...'�c7
7.0- 0 J.e7 8.4:lc3 0-0 9.g4 tt:lc6 10.g5
Cile8 ll.J.g2 f6!. It is nice to have e6
defended.
Black can remain in Kan waters
with 5...4:lf6, but let's also consider:
a) 5...b5?! 6.a4!
The point of White's tricky move
order. His queen's knight will reach
d2 in one move (compare it to vari­
ations with tt:lb1-c3-b1-d2!). Thus
he will be able to save J.f3 which is
not too active.
The mundane 6.0-0 is less chal­
lenging:
6...ib7 7.if3 (7.'�d3 Wic7 8.f4
tt:lf6 9.if3 tt:lc6 10.4:lxc6 dxc6 ll.eS
l"ld8 12.Wie2 tt:ldS=, A.Sokolov-Ivan­
chuk, Asnieres sur Seine 2006) 7...
tt:lc6
It is already late for 7...d6 due to
8.a4! b4 9.Wid2, when 9... d5 loses to
10.l"ld1 dxe4 11.4:lxe6 Wixd2 12.4:lc7+
@d7 13.ixd2. However, 7...Wic7 is a
good alternative: 8.a4 (8.l"le1 tt:lc6!
9.4:lxc6 dxc6 lO.eS l"ld8 ll.Wie2 tt:le7
12.4:ld2 cS=) 8...b4 9.l"le1 tt:le7 lO.eS
tt:lbc6.
8.l"le1 (after 8.a4 b4, White
has to exchange the knights with
an equal game after 9.4:lxc6 ixc6
10.4:ld2 tt:lf6) 8...4:lge7 9.4:lc3 tt:lxd4
10.'�xd4 tt:lc6 ll.Wie3 tt:leS=, Tosic­
Goldin, Vrnjacka Banja 1998.
6...b4 7.0-0!
7.e5 Wic7 8.if4 ib7 9.if3 tt:lc6
equalises: 10.tt:lxc6 ixc6 11.0-0
tt:le7 12.tt:ld2 l"lc8 13.ixc6 Wixc6
14.l"le1 tt:lg6 15.ig3 J.e7 16.tt:le4
29
 Part 1
'\1!1Txc2 17.lt:Jd6+ .b:d6 18.'\1!1Txd6 aS,
Petrik-Oral, Czechia 2006; 7..!f3
'\1!lrc7 8.0-0 .!b7 9.�e1 lt:Je7! gives
Black his typical play, Volokitin­
Markowski, Moscow 2004.
7...ib7 8.lt:Jd2 Vfic7
Or 8...lt:Je7 9.lt:Jc4t; 8...d6 9.c3
bxc3 10.'\1!1Tb3 '\1!1Tc7 ll.Wixc3 �xc3
12.bxc3. White's big lead in de­
velopment secures him an advan­
tage: 12...lt:Jd7 13.�b1 lt:Jc5 14.f3 lt:Jf6
1S ..!a3 lt:Jfd7 16.lt:Jc4t.
9.�e1
I cannot find a good way of de­
fending the b4-pawn. White's main
positional threat is to play lt:Jd2-b3,
id2. Eventually, he may open up
the queenside with c2-c3. I have
analysed:
9... d6 10..!f3 lt:Jd7
Or: 10...lt:Jf6 11.lt:J2b3 .!e7 12.i.d2
aS 13.c3; the same plan works
well after 10...lt:Je7 ll.lt:J2b3 lt:Jbc6
12..!d2 g6 13.lt:Jxc6 lt:Jxc6 14.c3;t ig7
1S.cxb4 .b:b2 16J'!b1 ig7 17.�cl '\1!1Te7
(17...�d8 18.ic3 0-0 19.ixg7 'it>xg7
20.eS;t) 18.bS axbS 19.axbS lt:JeS
20.b6 t.
11.lt:J2b3 lt:Jgf6 12.id2 lt:JcS
13 ..bb4 lt:Jcxe4 14.ia5 '\1!1Td7 1S.c4
30
ie7 16.Wic2 lt:JcS 17.lt:JxcS dxcS
18.'\1!1Tb3 .b:f3 19.lt:Jxf3 0-0 20.lt:JeS
with domination;
9...lt:Je7 10.lt:Jc4 dS ll.exdS .b:dS
12.b3±.
b) S ...lt:Jf6! ? 6.lt:Jc3 ib4?!
This attack is rather risky. You
should be ready to defend a very
passive position after 7.eS!.
The Scheveningen after 6...d6
is a sound and more complex ap­
proach. Another decent option is 6...
�c7 7.0-0 lt:Jc6 (Caruana bet Anand
with 7....!cS B..!gS lt:Jc6 9.lt:Jxc6
bxc6 in a rapid game, Zurich 2014.)
which transposes to the Taimanov
while 7...ib4 (7...bS 8.if3±) may
amazingly lead to ... the Scheve­
ningen after 8.igS!. Now Black has
nothing better but return to e7: 8...
ie7! , when the paradoxical 9..!e3!?
looks best.
Note that the popular line 8...
ixc3 9.ixf6 gxf6 10.bxc3 Wixc3
1U!b1 is rather dubious for Black:
ll ...lt:Jc6 12.lt:Jxc6 bxc6 13J'!b3
'\1!lre5 14.f4 '\1!1TcS+ 15.'it>h1�. Black is
running out of useful moves: 1S ...
hS 16.Wid2 h4 17.h3 aS 18J:l:c3 '\1!1Tb4
19.a3 Wffe7 20J'!bl.
 The Classical System with .ie2

7 .eS! 10...�aS

Alternatively: Hellsten recommends 10...�b6

bl) 7.�d3 may transpose to the ll.lt:lf3 �aS 12.�d6 lt:lc6, but 13.:gab1
Taimanov after 7...�c7 8.0-0 lt:lc6 .ixeS 14.lt:lxeS �xeS 1S.W'd2 gives
9.i'h1 lt:lxd4 10.�xd4 icS. I like White a very unpleasant initiative.
more:
ll.id6 lt:lc6 12.lt:lxc6 dxc6
7...d6 8.0-0 0-0 which is consid­
13.�d3!
ered below. S...lt:lbd7 9.lt:lb3 lUeS is
also possible. The stem game Giri-Vitiugov,
Reggio Emilia 2012, went 13.f4 ixa1
b2) 7.0-0 d6! [7...ixc3?! S.bxc3
13...�dS!? 14,:gb1 �xd1 1S,:gfxd1
ll:lxe4 9.if3!± dS 10.ixe4 dxe4
aS looks enough for a draw: 16.a4
11.ia3 is difficult to defend, for
(16.a3 a4 17.i'f2 iaS 1S.c4 idS
instance: ll...eS 12.W'e2 fS (12...
19.cS ie7 20.:gb4 :gfS=) 16...ib4=,
exd4 13.�xe4+ .ie6 14.�xb7 lt:ld7
Ramnath Bhuvanesh-Arutinian,
1S.cxd4±) 13.f3�] S.�d3 0-0 9J�d1
Dubai 2012.
The attack on g7 is harmless:
14.�xa1 �d2 1S.id3 �e3+
9.�g3 lt:lbd7 10.ih6 lUeS 1U'lfd1
16.i'h1 id7 17.�b2 bS 1S.a4 :gcs
(ll.igS lt:ldf6) 1l....hc3 12.bxc3
19.fS, when Stohl suggests 19...hS
Wic7 13.c4 lt:lcS=. Other options are:
intending .. ,:gh6, exfS.
9.igS lt:lbd7 10.lt:lb3 .hc3 ll.bxc3
'f1c7 12J�ad1 bS=; 9.id2 eS 10.lt:lb3 13...ixeS 14J�adU.f61S.�e3 ie7
1ie6 ll.a3 ixc3 12..hc3 lt:lc6=. 16.ixe7 i'xe7 17.:gd3 :ges 1S.�e4
9 ...lt:lbd7 10.lt:lb3 lt:leS 11.�d4 (11. White has full compensation for
'a'e3 bS 12.a4 lt:lc4) ll...lt:lc6 12.�e3 the two missing pawns.
ffc7 13.id2 dSt, Fedorchuk-Hillarp
Persson, Budva 2009. 6 . . . b5 7. 0-0

7...lt:ldS (7...lt:le4 8.0-0) S.0-0 7.f4 ib7 S.if3 is considered in

(S.id2 lt:lxc3 9.bxc3 ie7 10.0-0 Part 4.
Vffc7 ll.f4 d6) S...lt:lxc3 9.bxc3 ixc3
10.ia3 7. . . .ib7 8.l3 el

31
 Part 1
8..if3 could be justified only if
White is planning to sac a pawn af­
ter 8... tt:Jc6 9.tt'lxc6 dxc6 10.e5
10.'\We2 offers Black a pleas­
ant choice. At the Olympiad 2012
against Martinez Reyez, I decided
to play solidly and opted for a sym­
metric position with 10...e5 ll.a4
tt'lf6 12J'!d1 (It is better to exchange
on b5 immediately 12.axb5 cxb5 al­
though the position after 13..ig5 .ie7
14..bf6 hf6 15.tt'ld5 .txd5 16.exd5
�d6 17.c4 bxc4 18JUc1 0-0 19J�xc4
aS should be impossible to crack.)
12...ie7 13.ie3 (13.axb5 cxb5 is al-
ready totally equal) 13...0-0 14.g3
!'ladS 15.ig2, when 15....ic8 or 15...
b4 16.tt'la2 a5 is fine for Black.
Much more interesting is 10.�e2
id6!? 11.g3 tt'lf6 (or ll...tt'le7 12.ig2
c5 13.f4oo) 12.ig2 0-0 13.f4 e5 14.f5
White hopes for a kingside pawn
storm, but Black is the first to start
active actions: 14...a5 15.�f3 b4
16.tt'ld1 ic5+ 17.ie3 he3+ 18.�xe3
.ia6 19.l"le1 l"lfd8 20.tt'lf2 c5 21.l"led1
h6+, Hecht-Ravikumar, La Valetta
1980.
10...�xe5 ll.l"le1 �c7
This variation had its five min­
utes of glory when 15-years old
32
Kasparov crushed Polugaevsky at
the Soviet championship in 1978:
12.ih5 ie7 13.l"lxe6 g6 14.l"le1
l"ld8? 15.�f3 c5 16.if4 �b6 17.�g3
gxh5 18.ic7. However, two years
later Wojtkiewicz took the piece
14...gxh5! and won. The latest at­
tempt to revive this line was Nava­
ra-Movsesian, Czechia 2010, but it
turned out that Navara had noth­
ing new to show: 12.a4 tt'lf6 13.axb5
axb5 14.tt:Jxb5 cxb5 15.l"lxa8+ ixa8
16.ixa8 id6 17.g3 0-0 18.ig2 b4.
Black is fine and that was known
from a previous game.
8 ... �c6 9 .li� xc6 dxc6 10.e5
Compared to the Taimanov with
an early ...b5, Black has inserted
here 7...ib7 8.l"lel. That takes the
sting of 10.a4 due to 10...l"ld8 (10...
id6 ll.g3 ie5 12..id3 tt'lf6 13.ie3
0-0 14.ic5 l"lfd8 15.f4, Asrian-Mi­
lov, Groningen 1998, 15...hc3
16.bxc3 tt'ld7 17..ie3 c5oo) ll.id3
tt'le7! (White gets a strong attack
after ll...id6? 12.e5 he5 13.�h5
.txc3 14.bxc3 tt'le7 15.l"lb1 l"ld5
16.�h3) 12.e5 tt'lg6 13.tt'le4 l"ld5.
to ,gds ll . .id3 c5
..
 The Classical System with .ie2
In this position, White usually
transfers the queen to the kingside:
A. 12.'Wh5; B. 12.'Wg4
12.'We2 is a minor option: 12...
�'lle 7 13.a4 (13.ig5 h6 14.'Wh5 l"!d4
15.ie3 l"!d7 16.if4 c4 17.ie4 b4!
18.'Lldl 'LldS, V.Dimitrov-Vyzma­
navin, Burgas 1993) 13...b4 14.'Lle4
V'ic6 15.b3 (15.'Wh5 'Llg6 16.ifll"!d4
1 7.f3 l"!dS) 15...'Llg6 (15...'Llf5 16.ig5
;(d5 17.l"!ad1 h6! 18.if4 ie7 is also
p ossible, but I want to provoke f4 in
order to limit White's dark-squared
bishop's scope) 16.f4 ie7 17.ie3
tilh4. It is unclear how White should
improve his position from here. For
instance, the attack on c5 by 18.'Wf2
is neutralised with 18...l"!d5.
A. 12 .�h5 g6
12...c4 13.ie4 'Lle7 14..b:b7 'Wxb7
15.ig5 l"!d4 16.l"!ad1 l"!xd1 17.'Wxd1 h6
18.ih4, S.J.Solomon-Miezis, Mel­
b ourne 2013, is pleasant for White.
13.�h3
T.Horvath chose against me in
Varna 2012 13.'Wg5 with the obvious
idea to provoke further weakening
of my castling position. However,
the inclusion of 13...h6 is in Black's
favour. After 14.�g3, I could have
expanded on the kingside by 14...
c4 15.ie4 'Lle7 16.'Wf3 'Llc6 17..if4
g5 18.ig3 'Lld4 19.'We3 ib4 20.l"!ed1
ixe4 21.'Wxe4 hc3 22.bxc3 'LlfS
with complex play. I opted for the
common 14...ig7. This is not a bad
move, of course, but it gives White
more options, for instance, 15.f4
'Lle7 16.'Wf2 'Llf5 (16...0-0 17.ie3)
17.ie4 b4 18.ixb7 bxc3 19.ie4 cxb2
20.hb2 0-0=. My opponent erred
with 15.if4? 'Lle7 16.a4 and was lost
after 16...c4 17.ie4 b4 18.hb7 bxc3
19.ixa6 cxb2 20.ib5+ 'it?f8. When­
ever White's bishop leaves cl, the
idea with ...c4, followed by ...b4 be­
comes effective.
13.'We2 ig7 is rarely seen. Then
14.f4 'Lle7 15.ie4 'LldS, and 14.ig5
'Lle7 15.a4 b4 16.'Llb1 l"!d7 17.'Lld2
0-0 are roughly equal. The most
principled continuation is 14.a4 b4
15.'Lle4, but 15...�xe5! 16.g3 'Llf6
equalises.
13 .ig7 14 . .ig5
.•.
Sacrificing the central pawn by
14.a4 b4 15.'Lle4 is a mistake (15...
ixe5 16.ig5 l"!d7), but White has
two decent alternatives to the main
line:
a) 14..if4!? is a solid and unpleas­
ant approach. Black has not any­
thing better but to hold the oppo-
33
 Part 1
nent to a draw in the line 14...tLle7
(14...h6?! preserves more pieces
on the board, but 15.a4 b4 16.tLle4
.ixe4 17.gxe4 a5 18.gd1 c;f;>f8 19.�f3
tLle7 20J''!ee1 c;f;>g8 2l.h4 is obvious­
ly better for White.) 15.a4 b4 (Hell­
sten's recommendation 15...c4?
16..if1 b4 17.tLle4 tLlfS loses the c4-
pawn after 18.c3! bxc3 19.bxc3 0-0
20.g4 .ixe4 21.gxe4 tLle7 22.gxc4::t)
16.tLle4 .he4 17.gxe4 �c6 18..ig5
gd7 19..he7 gxe7.
b) 14..ie4 creates the strong po­
sitional threat of taking full control
over the dark squares, for example,
after 14...tLle7? (Hellsten) 15..hb7
�xb7 16..ih6! . It is clear that Black
should hinder this plan. I see two
sensible ways:
14....ixe4 15.tLlxe4 .ixeS 16.i.g5
gd7 17.gad1 h6 18.gxd7 �xd7
19.tLlxc5 �d6 20.i.e3 tLle7 21.tLld3
.id4 22.gd1 .ixe3 23.�xe3 tLldS;
14...b4 15.hb7 �xb7 16.tLla4
(16.tLle4?? �xe4! ; 16..ig5 gd4
17.tLla4 �c6 18..ie3 gds 19.tLlxc5
gxcS 20.hc5 �xeS leads to very
unbalanced position where Black's
chances should not be worse, in my
opinion: 2l.a3 tLle7 22.axb4 �xb4
23.�a3 �xa3 24.gxa3 0-0 25.gxa6
34
gbs 26.b3 gcs 27.c4 gbsoo) 16...
tLle7 17.tLlxc5 �c7 18.tLld3 h6 19.�f3
0-0. Black has compensation for
the pawn, for example, 20.�d1 tLlc6
2l..id2 .heS 22.tLlxe5 tLlxeS 23.�c1
gS 24.h3 gd4i or 20.i.e3 b3 2l.axb3
�xc2 22.tLlb4 �xb3 23..ic5 �xf3
24.gxf3 gd7 25.gxa6 gcs 26.he7
gxe�.
14 ... gd7!?
This move is less explored than
14...tLle7 which has a good reputa­
tion, too:
a) 15..if6 0-0 equalises easily:
16.a4 (16.tLle4 he4 17.gxe4 gd4)
16...b4 17.tLle4 he4 18.gxe4 hf6
19.exf6 tLldS 20..ixa6 tLlxf6.
b) 15.a4 b4 16.tLle4 he4 17.gxe4
.ixeS 18.ha6 .ixb2=.
c) 15..ixb5+ axbS 16.tLlxb5
�b6 17.tLld6+ gxd6 18.exd6 �xd6
19.gad1 �c6 20.�g3 h6 2l.�b8+
�c8 22.�xc8+ hc8 23.ie3 .hb2
24..ixc5 fS 25.a4 at first sight looks
promising for White, but Black can
set up coordination of his pieces
and take over the initiative, e.g. 25...
c;f;>f7 26.gd6 ges 27.c4 .ic3 28.ged1
eS 29Jk1 iaS 30..ib6 .ib4 3l.c5
 The Classical System with .ie2

.�.e6 32.a5 !'!:a8 33.1"1:a1 .ic3+. 16.ih4 .b:e4 17.tt:lxe4 ixeS

18.1"1:ad1 is not dangerous. Black only
d) 15..ie4 !'!:d4 16..b:b7
has to avoid Hellsten's recommen­
16.1"1:ad1?! !'!:xd1 17.1"1:xd1 .b:eS
dation 18...g5? which stumbles into
18.1xb7 'Wxb7 19.'We3 .id4 20.'We4
19.1"1:xd7! 'Wxd7 20.!g3 'WdS 21..b:e5
','!J'x e4 21.tt:lxe4 fS gives Black a slight
'WxeS 22.'Wa3±. Instead, he can opt
i nitiative; 16.'We3 tt:ld5=.
for a drawish endgame after 18...f5
16...'Wxb7 17.f3 h6
19.1"1:xd7 ltfxd7 20.tt:ld2 tt:lf6 21.'We3
Hellsten suggests as an alterna­
tt:lg4 22.�d3+ 'Wd6 23.h3 'Wxd3
tiv 17... tt:ld5! ? 18.tt:le4 !xeS.
e
24.cxd3 .b:b2 25.hxg4 ic3 26.1"1:e2
18.ie3 !'!:c4 19.lt:Je4 .b:eS 20.c3
.b:d2 27.1"1:xd2 fxg4 or for attrac­
.S idS. This position was reached in
tive complications after 18... 1"1:xd1! ?
the game Rosito-Hellsten, Mendo­
19.1"1:xd1 ltff8 20..id8 (20.'Wa3 b4)
za 2005. White chose here 21.b3?
20...'Wc6 21.'We3 .id4 22.'We1 f5oo.
)(xc3 22.tt:\xc3 .b:c3 23..b:c5 .ixa1
2 4 .1"1:xa1 'Wc'T-F. Hellsten gives as best
16 . . . b4 17 . .ixb7 bxc3 18 . .ixa6
21 .R:ad1 'Wc6 22.b3 !'!:xc3! 23.tt:lxc3
'Wa5 19.J.c4 cxb2 2o.gabl lLle7
.0.xc3 24.id2 id4+ 25.1tfh1 hS, fol­
(20...g5? 21.�b3) 21.�b3 0-0
lowed by ...0-0 and 'Black has no
problems'.
Black has finally castled and his
pieces are very active, for instance,
15.J.e4
22.'Wb5 �c7 23.1"1:xb2 gS 24.ig3 tt:lfS
25.f4 !'!:fd8 26.id3 tt:\xg3 27.hxg3
a) 15.a4 b4 16.tt:le4 .b:eS.
gxf4 28.gxf4 c4�.
b) 15.f4 c4 16.ie4 b4 17.hb7
does not work (17.tt:ld1 is the only
move) due to 17...bxc3+. B. 12.'Wg4 ! ? gd4

15 . . . h6 Black should resist the tempta­

.
tiOn to harass the queen by 12...h5
because this move seriously weak­
ens his kingside. White has the
better chances after 13.'We2! [13.
'Wh3 .ie7 14.b3 (14.if4 gS 15.ic1
b4 16.tt:le4 'WxeS) 14... 1tff8 15.a4 c4
16.bxc4 bxc4 17.if1 tt:\h6oo] 13...
c4 (13... tt:le7 14.ig5) 14.ie4 .b:e4
15.'Wxe4 .icS 16..ig5 tt:\e7 17.a3;!;.

12...tt:le7, on the contrary, is a fair

alternative to our main line. 13..ig5
16.J.f4 !'!:d4 14.�g3 tt:lg6

35
 Part 1

15J'!ad1 �2.c3 dxc3 23.�b4 c2 24.�c1 lt:'le7

15.lt'le2 �d7 16.�ad1 does not
change the structure. Black may
grab space on the queenside by 16...
c4 [16...h6 17.ie3 lt'le7 18.c4! bxc4
19.ic2 g6 20.lt:'lc3 (20.�xd7 l!?xd7
21.�f4 lt:'lf5) 20...lt'lf5 21.�xd7 lt>xd7
22.hf5 gxfS leads to an interest­
ing position where Black's king is
perfectly safe on the queenside.]
17.hg6 hxg6 18.�xd7 �xd7 19.lt:'lc3,
Danin-Bocharov, Taganrog 2011,
·
19...b4! ? 20.�d1 �c7 21.lt:'l a4 idS
22.b3 !e7 23.he7 l!?xe7 24.�e3
�b8 25.h3 �b5=.
15...ie7 16.hb5!
The critical variation. Alterna­
tively:
16.lt:'le2 �d8 17.hg6 hxg6=, Zel­
cic-Lanka, Graz 2009.
16.he7 �xe7 17.lt:'le4 he4
25.�e2 lt:'ldS 26.�d4 0-0 27.�exc2
�a6 or 22.�b4 lt:'le3 (22...g5! ?)
23.�xe3 dxe3 24.�d6 �c7 25.�b5+
l!?e7 26.�b4 lt>e8=.
13.�g3
White has an alternative:
13.ie4 he4 14.lt:'lxe4 �xe5 15.c3
�d3 [Hellsten suggests 15...�a4, but
theendgameafter 16.�e2 (16.b3 lt:'lf6
17.lt:'lxf6+ gxf6 18.�d1 �e4 19.�xe4
�xe4 20.a4 ig7 21.!e3 0-0=) 16...
lt:'lf6 17.lt:'lxf6+ gxf6 18.�xe5 fxeS
19.�xe5 is slightly better for White
since theh7-and c5-pawns are weak.]
16.i.f4 lt:'lf6 17.lt:'lxf6+ �xf6 18.a4 (18 .
.igS �g6 19.�ad1 c4 20.�xd3 cxd3
21.�g3 f6=) 18....ie7 19.axb5 axbS
20.�a8+ �d8 21..ie5 �gS 22.�xg5=
with mass elimination.
13 . . h5 ! ?
.

18.�xe4 �dS was balanced in Efi­

menko-Svidler, Fuegen 2006.

16... axb5 17.lt:'lxb5, Polivanov­

Zakhartsov, Lviv 2012, 17...�b6
18.lt:'lxd4 cxd4 19.he7 lt'lxe7 20.�b3
�c6 21.f3 lt:'lfS. In Chess Informant
116, Zakhartsov assesses this posi­
tion as equal. I would say, it is un­
clear. The game might continue:

36
 The Classical System with i.e2
I criticized this advance on the
p re vio us turn, but here it has at­
t ac ki n g purpose and allows Black
to seiz e the initiative:
14.h3
Or 14.f3 h4 15.\Wf2 h3 16.g3 &iJe7
1 7.il. e4 &iJc6; 14.�e4 &iJe7 15.\We3
t/1 c6 ; 14.f4 h4 15.Wf2 &iJh6 16.ie4
h3 17.\Wf3 ixe4 18.&iJxe4 &iJfS.
14 h4 15. \We3 �e7 16.b3 �c6
••.
17 .ib2 gd7 18J:!adl .ie7 19.\We2
•
�d4 20.\Wg4 0-0
The game might continue 2l.icl
&iJfS 22.ixf5 l=l:xdl 23.&iJxdl exfS
24.\WxfS .ic8 25.\Wf4 l=l:d8 26.&iJc3
ie6� 27.a3 b4 28.&iJe4 l=l:dS.
37
Pa rt 1. The Classical System with ie2

Com p lete G a m es

my opm10n, 18...Wi'b6, which has

1 . Lutz-Grablia uskas
been designed as an improvement
Pula 1 997 on 18...Wi'b8, is not so good in vew
of 19.Wxh5:t. The other option 18...
l.e4 c5 2.ll:�f3 e6 3.d4 cxd4
Vfffc6 puts the queen on a file which
4.ll:�xd4 c!Dc6 5.c!Dc3 W/c7 6.�e2 a6
is going to be opened soon. That
7.0-0 c!Df6 8.�hl c!Dxd4 9.Wi'xd4
gives White the initiative, 19J'lac1
�c5 10.Wi'd3 b5 ll.f4 �b7 12��f3
0-0 20J"lfe1 h4 21.g4 hg3 22.hg3
h5 13.e5 c!Dg4 14.hb7 Wi'xb7
Eic7 23.c4, Tischbierek-Michaelsen,
15.c!De4 �e7
1996.

I recommend 15..J'k8 in order to

19.�f2 c!Df5 20J:!dl h4
avoid 15...i.e7 16.ie3. Anyway, 15...
Eic8 16.b3 ie7 would have trans­
The opening stage is over. Black
posed to the game.
regrouped successfully and leveled
the game. His aim is to organise
16 .b3 �c8 17.�b2 c!Dh6 18 .Wi'f3
some play on the queenside after
which he will castle.

21.�d3 �k7 22 J'!fd2

18 .•. Wi'b8 !

White was threatening to win

the queen, so we must retreat. In 22 b4? !
•.•

38
 The Classical System with i.e2
Ribli assesses positively this
m o ve , but Black is still not ready for
it. 2 2 . \Wb? (or 22...\Wa7) 23.a3 aS=
. .
wou l d have been better.
Now White misses his chance for
2 3 .c4 ! bxc3 24.i.xc3;1; and Black's
pawn is unable to reach a4.
23 ti'a7 24."i'f2 "i'b7 25.ti'fJ
.•.
a5 2 6 c4 bxc3 27.hc3 a4+±
.
2 8 . 13dl? !
With this humble move, White
acquiesces in being worse. 28.b4!
was better, when 28...i.xb4 leads
to a draw by perpetual after 29.13b2
VXfa7 30.�f2 \Wa6 31.13xb4 \Wxd3
32J=Ib8+ <Jle7 33.13xh8 13xc3 (33...
VXfxe4?? 34.\Wb6) 34.tt:\xc3 \Wxc3=
3 5 . h3 \Wc1 36.<Jlh2 tt:\g3 37.13h4 tt:\fl.
28 0-0 29.�el (29.bxa4 \Wa7)
•••
2 9 axb3 30.axb3 gas 31.�f2
.• .
13a2 32 .g4 hxg3 33.hxg3 �h6
34. �c3 \WxfJ 35.gxfJ
35 .•. gb2
Afte r this move, Black's advan­
tage dissipates and White gradual-
ly seizes the initiative. 35...13c2 ! ?+
36.tt:\e4 tt:\g4 37..id4 f6 would have
kept the active rook alive.
36.�a4 ge2 37.ge3 ga2
38.gc3 gxc3 39.�c3 gb2 40.gd7
gb3 41.�e4! �a3 42. �g5
Despite the limited material and
pawns on one wing, Black still has
some difficulties because his pieces
are cramped.
42 gbs 43.�a7 ges 44. �g2
•••
�e7 45. �e4 �g4 46.�h3 f5 ! ?
47.exf6 draw.
At this moment the opponents
signed the draw. White had chances
for torturing the opponent for many
moves ahead. He should aim for a
position like that:
For instance after 47.tt:\d2;1; .if8
48..id4 ie7 49.tt:\f3 <JlfB 50.13c7 13b8
Sl..igl 13b1 52.<Jlg2 13b2 53.<Jlhl.
The vulnerability of the e6-pawn
makes the difference.
The play could continue with
53.. J"�a2 54.tt:\d4 13a6 55.tt:\c6 .ia3
56.<Jl g2 <JleB 57.tt:\b8 and Black is
still struggling.
39
 Part 1
2. Adams-Ban i kas
Porto Ca rras 04. 1 1 .20 1 1
l.e4 c5 2 . c!Llf3 e 6 3.d4 cxd4
4. c!Llxd4 tilc6 5.c!Llc3 Y!Jc7 6 .te2 •
a6 7.0-0 c!ilf6 8 .te3 .tb4 9.c!Lla4
•
.te7 10. c!Llxc6 bxc6 ll. c!Llb6 gbs
12.c!Llxc8 Y!Jxc8 13.J.d4 Y!Jc7 14.e5
tild5
15.c4!
This looks stronger than 1S.b3
aS!
It is better to delay ...c6-cS. In
some variations Black can use the
cS-square for a piece. Another ar­
gument in favour of 1S ...aS is seen
in the line 16.i.d3 0-0 17.Y!JhS g6
18.Y!Jh6 fS, when 19.c4 lt:Jb4 is equal,
due to the threat of lt:Jc2, while 19.ef6
i.f6 20.i.g6? fails to 20... d6! .
16.g3
If White takes the f4-square un­
der control with 16.Y!Jd2!? 0-0 (af­
ter 16...i.b4! ? 17.Y!JgS 0-0 18.c4
lt:Je7 19.i.d3 lt:Jg6 20.Y!Je3 a4, More­
no Carretero-Raijmaekers, ICCF
email 2009, White could not prove
any substantial advantage) 17.c4,
the zwishenzug 17...i.b4? stumbles
40
into 18.Y!Jc1! , but 17... lt:Jb4 18..!c3 cS
19.i.f3 (19.£4 d6) 19...d6 (19 ...dS! ?)
20.exd6 (20.Y!Je2 dxeS) 20....ixd6
2l.g3 ieS 22J'Ud1 ixc3 23.Y!Jxc3
a4 equalised in Moreno Camera­
Valles Moreno, Madrid 2009.
16...cS 17..!b2 0-0
The play is balanced. It would
be interesting to try 18.a4 gfd8, but
my game Magem-Delchev, France
2009, finished here in a draw. In The
Safest Sicilian I analysed in detail:
18.Y!Jd2 a4! 19.if3 lt:Jb4 20.1"lfd1
gfd8 2l.Y!Je2=. White attempted to
improve by 20.1"lfe1 axb3 2l.axb3 c4
22.ged1 cxb3 23.cxb3 gfd8 24.Y!Je2
lt:Jc6 2S.!'ld3 lt:Jb4 26.1"lc3 lt:Jc6=, Gar­
cia Corada-Van Unen, ICCF email
2009.
15 ••• c!Llf4 16.g3
16..!f3 is more popular. Black
meets it by 16...0-0! 17.ge1
17.g3 lt:Jg6 18.Y!Je2 (18.ie4 lt:JxeS!
19.hh7+ 'it>xh7 20.Y!JhS+ 'it>g8
21..beS Y!JaS 22.1�'e2 d6 23.ic3
Y!JfS=) 18...£6 19.exf6 hf6 20.hf6
!'lxf6 2l.b3 aS=, Duda-M.Andersen,
Chatawa 2010.
17... lt:Jg6! 18.cS. (18.g3 ib4+)
 The Classical System with ie2

This position first occurred (after a owing to 20.i.h5+.

different move order) in the game
19 �c6+ 20.£3 f6 21.exf6
Kasparov-Anand, Linares 2002.
•••

i.x£6 22 .�d3 �f7 23.gadl �h6

18. f6 19.ef6 if6 20..te4 and here
. .

1 proposed the novelty 20.,..txd4! ?

Black can take on c3 on move 22
2 l .�xd4 tDf4 22.!!adl tDdS with mu­
or 23, but I do not like his position.
tual chances. The play may contin­
ue with 23.�d3 (23.a3 aS 24.g3 �dB
24.W6 gxf6 25.�c3 0-0
2S ..id3 �f6 26.!!e5 g5=) 23... !!xb2
26.i.d3;t
2 4 .ixh7 'it>h8 25.i.g6 �f4 26.!!e2
:·(xe 2 27.�xe2 �gS 28.�h5 �xhS
2 9.ixh5 !!b8t=.

16 c5 ! 17.i.c3 (17.gxf4 cxd4

. .•

18.�xd4 0-0�) 17 tDh3 ! 18.<.!,lg2

•••

tDg5

White's rooks are much more

active while the bishop restricts
Black's knight and is ready to eat it
should it reach fS or c6. Black will
have to struggle for the draw.

Black's knight is heading for fl. 26 . . . �c7 27.gf2 gbd8 28.gfd2

It is the ideal place for it. The knight
will control from there the critical
squares eS and d6, protecting at the
same time Black's castling position.
19.b3 !
f5 29.i.e2 f4 30.i.d3 d6 31.i.e4
t[)f5 32.hf5 gxf5 33.g4 gf7
34.gd3 <.!,Jf8 35.h4 <.!,le7 36.h5? !
gf6 37.a3 a5 38.�el (38.b4! ?)
38 h6 39. <.!,lh3 e5 40.�e4 <.!,Jf8
.••

This novelty of Adams allows

White to consolidate while re­
taining the better pawn structure.
1 9 .�d3 �c6+ 20.f3 f6 2l.!'!adl tDf7
2 2 .exf6 ixf6 23.ixf6 gxf6 24.�c3
�?l e 7 25.!!d2 hS, Mastrovasilis­
Miladinovic, Valjevo 2011, or
1 9 .f3 f6 20.exf6 .txf6 21.�d2 .txc3
2 2 . �xc3 eS lead to complex play.
After the text, 19...f6 is dubious Black has not any counterplay.

41
 Part 1
It looks incredible that he could
hold this position.
4UM5 gbs 42 .Y;\'d3 'i:t>g7
43.Y;\'c3 ges 44,g5d2 'i:t>h8 45.b4
axb4 46.axb4 cxb4 47.Y;\'xb4
Y;\'c6 (47...Y;\'a7! �) 48,gd5 'i:t>g8
49.Y;\'b3 'i:t>h8 50 .Y;\'c3 Y;\'a6 51.c5
'We2 52,g5d2 Y;\'e3 53.Y;\'xe3 fxe3
54,gxd6 gxf3 + 55. 'i:t>h4? 'i:t>h7
56.gel? ggs 57.c6 gf4 58.gg6
gxg6 59.hxg6+ 'i:t>xg6 60,gxe3
draw.
3. Balinov-Sommerbauer
Tweng 2007
l.e4 c5 2 . ltlf3 e6 3.d4 cxd4
4.lL!xd4 ltlc6 5.lL!c3 a6 6 .ie2 •
'Wc7 7.0-0 ltlf6 8 .ie3 .ib4 9.lL!a4
• 20...Wc7
.ie7 lO. lL!xc6 bxc6 ll. ltlb6 gbs
12. ltlxc8 Y;\'xc8 13 .id4 c5 14 .ie5
• •
gb6 15.b3 lL!xe4 16.hg7 ggs
17 .ib2 .if6 ! ?
•
This might be Black's best op­
tion. It seems appealing to attack
on the g-file, but apparently White
has adequate defence even against
17...c4! ?. By exchanging the dark­
squared bishops, Black ensures the
safety of his king in the centre. It is
42
extremely difficult for White to find
a target, because the central pawn
cluster keeps his pieces at a bay.
Black has plenty of dark squares
to manoeuvre on them. It is im­
portant first to exchange a pair of
rooks to reduce White's attacking
resources.
18.hf6 lL!xf6 19 .if3 •
This allows the manoeuvre
.. .l'!gS-eS which removes the only
flaw of Black's set-up - unconnect­
ed rooks. I have investigated the
more challenging move:
19.Y;\'d2! 'i:t>e7 20.l'!adl
20.l'!fel l'!d6 2l.We3 Wc7 22.l'!adl
l'!xdl 23.l'!xdl dS 24.g3 aS.
White has no weaknesses and
his pieces are much more compact.
Nevertheless, Black's game is very
easy! It is a widely know fact that
W+ tLl often dominate W+.i. If he can
trade one ore both pairs of rooks, he
would become clearly better. The
a6-pawn will move out of the bish­
op's range to aS, and the knight will
be seeking to land on a central dark
square.
 The Classical System with .ie2

2 1 . 'tt> hl :Se5 22.ti'c3 :axel+ 23.:Sxel <lJe7

24.a3 :Sb8 25.g3 a5 26.:Sdl :Sc8
Alternatives are:
27.ti'c4 We5 28 .'1Wa6 Wc7 29.<lJg2
2 1.'1We3 �d6 22.�xd6 \Wxd6
:Sb8 30.:Sd3
2 3 . �dl \Wc7 24.c4 (24..ba6 �aS);
2l .c4 �d6 22.\We3 �xdl 23.�xdl
hS ;
2l..if3 �d6 22.\Wf4 �c8 23.'\Wa4
� xdl 24.�xdl aS 2S.c4 hS 26.h3 h4
2 7.\Wa3 �b8 28.'\Wcl a4 29.'We3 axb3
:3 0.axb3 'WaS 31.'\WgS �xb3 32.'\Wxh4
VVb4 33.'tt> h2 �b2 34.'\Wg3 �d2+;

2l. ..�d6 (or 21...'\WeS 22.�fel

V·VgS) 22.We3
Or 22.Wf4 \Wc6 23.f3 �xdl
2 4J:!xdl dS 2S ..id3 Wd6 26.Wh4
V"feS 27.Wa4 (27.�el Wd4) 27...lt:JhS
28.Wxa6 lt:Jf4 29..ifl 'We3.
2 2 ...�d4 23..if3 (23.c3 'LldS;
2 3 . �xd4 cxd4 24.Wxd4 Wxc2) 23...
,!Jg4 24.hg4 �gxg4 2S.c3 �de4
2 6 .\Wd2 �eS 27.f3 (27.f4 �dS) 27...
T\h4 28.h3 dS 29.'Wf2 �f4�.
19 . . . :Sg5 20 .:Sel Wc7 21.Wd2
30 ••• d5
Sommerbauer has shown that
he could simply stay and wait. Now
he demonstrates that Black is able
to shift forward his defence line.
31.:Sd2 Wb6 32.ti'e2 :Sd8
33.'1We5 '!Wd6 34.'\Wc3 '\Wb6 35.'1We5
'\Wd6 36.'\Wc3 '\Wb6 37.�e2 '\Wc7
38.<lJgl Wb6 39.:Sel '\Wc7 40.:Se2
'!Wb6 41.:Sel draw.

43
44
 Pa rt 2

The Engl ish Attac k

This part is devoted to the Taimanov move order l.e4 c5 2 . �f3 e6

3 . d4 cxd4 4. �xd4 �c6 5.�c3 ti'c7 6 .ie3 a6 7.%Yd2 �f6
•

I allocated only one page to the Kan since l.e4 cS 2.tt'lf3 e6 3.d4 cxd4
4 .'Llxd4 a6 S.tt'lc3 'Jfffc7 6.ie3?! is simply dubious due to 6...tt'lf6!
White is living through a crisis in the English Attack with f3. I expect
to see the focus of investigations shifting toward set-ups with an early f4.

45
Pa rt 2 . The English Attack
M a in Ideas
l.e4 c5 2 . �f3 e 6 3.d4 cxd4
4.�xd4 �c6 5. �c3 �c7 6 . .!e3
6 . . . a6 ! 7.'i;Yd2
This set-up has no analogue in
the Kan because after 4...a6 S.lik3
�c7 6.i.e3 tt:\f6 7.f3 i.b4! 8.�d2 dS
gives Black an initiative.
Thus the whole chapter deals ex­
clusively with the Taimanov move
order.
The plan with 0-0-0, f3, g2-
g4 used to be White's main weap­
on against the Taimanov, but in the
last 2 years White is struggling to
get break-even. His latest attempt
is 7.�f3 and I'm sure we'll be seeing
more of it in the coming months. Do
not miss to look at game 5 Vallejo­
Spraggett, Catalunya 2013 where I
suggest 7... tt:\f6! 8.0-0-0 :JJ.e7 9.i.e2
0-0.
46
7. . . �f6 8. 0-0-0
Black has greater choice against
8.f3. Perhaps the most practical re­
tort is to transpose to the main line
with 8...b5 9.g4 .ib7 10.g5 �xd4! .
8.f4! ? :JJ.b4 9.1J.d3 0-0 10.a3 :JJ.e7
leads to complex play with mutual
chances.
8 ••. .!e7!
8....ib4 had been the main move
for a decade. It is not refuted, but
White has found several variations
where he could play slightly better
endgames without any risk. Only in
2013 did the second players devised
a totally new approach. It is bring­
ing amazing results so far. White
still cannot recover from the shock
and seems unable to generate any
sensible ideas. Let's investigate the
position after
 The English Attack

9 .£3 b5 10.g4 (10.\t>bl 0-0! ...dS. We castle only when we are

l l . g4 tLlxd4 12..b:d4 ib7) 10
ti:Jxd4 !
.•• forced to do it or if White gives us a
tempo by playing some quite move
like lt>bl. Next, I will show the main
scenarios from the diagram posi­
tion.

1. 11. \Wxd4 ib7 12 .g5 c!Llh5

13.ih3

The point of Black's new plan.

He intends to attack the centre with
. . ..�b7 and ...dS or ...eS followed up
by . . ds.
.

The trick is to trade knights before

castling! 13 . . . gc8 ! 14.ig4 b4 15.hh5
bxc3 16.\Wxg7 cxb2+ 17.1t>xb2
Before, Black played 10...0-0 gf8 = . See game 8 Demetrio­
l l .gS tLlhS, when 12.tLlce2 tLlxd4 Schiendorfer, ICCF email 2009.
13 .�xd4 ib7? would drop the d7-
pawn. That forced Black to opt for 2. 11.hd4 ib7 12 .g5 c!Llh5
flank counterplay with 13...f5, but 13.c!Lle2 (Taking f4 and g3 under
White is somewhat better after control. 13.i.e5 �xeS=; 13.h4 tLlg3!)
1 4 .tLlg3 tLlf4 lS.exfS .hgS 16.tLle4
eS 17.�d6t, Shirov-Munoz Santa­
n a , Istanbul 2012. Of course, Black
could try to enable ...dS by 12..J'!d8,
but this tempo gives White the in­
i t i ative. He may choose, for exam­
ple, 13.tLlg3 tLlxd4 14.tLlxh5 tLlxf3
1 5 .�f2 tLle5 16.ib6, although his ex­
t ra exchange does not automatically
guarantee him an advantage.
13 gcs
••• (threatening 14...
The timely exchange on d4, fol­ .txgS!) 14.gg1 0-0 15 .th3 g6 (or
•

l owed by ....tb7, keeps the d7-pawn lS...dS 16.e5 g6) 16. lt>b1 d5 17.e5
Pr o tected and enables fast develop­ c!Llg7. Black has a good game. He will
rn ent of the queenside with ..J'k8, push his pawns to a4 and b4.

47
 Part 2

We see that White's most Here are some examples of the

straightfmward plans, which are break ...dS.
typical for the English Attack, do
not set serious problems. White
may also turn to natural develop­
ment and central play with:

3. ll .hd4 i.b7 12. i.d3

•

Then the play is not forced and

Black has various decent options. I
analyse the most consistent one - to
limit the bishop's scope with: 14.tt:le2 dS

12 e5 13. i.f2 0-0 14.g5 c!tlh5 !

.•. Watch out if White played
15.c!tld5 (stopping ...dS) 15 .hd5..• 14.tt:la4 dS 15.tt:lb6.
16.exd5 c!tlf4 17.h4 (17.ixh7!
l!ixh7 18.d6 hd6 19.1Mfxd6 E:ac8 is
roughly equal)

We must meet this by 15...dxe4!

- never allow e4-e5 when White's
knight is on b6. This pawn struc­
17 f5 ! , when 18.gxf6? tt:lxd3+
.•. ture would be pleasant for White in
would be in Black's favour. the event of 15... E:ab8?! 16.e5 tt:ld7
17.tt:lxd7. We need the knight for
Latest games show that White is counterplay from cS.
disappointed from the above-men­
tioned courses and looks for more
flexible move orders with 'i!ibl
waiting for Black to define his plan �
first. The ensuing play is not forced
and offers both sides rich possibil­
ities of playing chess. I suggest to
meet 'it>bl with . . . 0-0 and then . . J'k8
or ...ds.

48
 The English Attack

If White has already played Game 6 Volokitin -Najer

g S , we can boldly trade our light­ M uelheim 201 3
squared bishop for the enemy
kn ight since the possession of the
f4-square fully balances White's
bishop pair: 15...b4 16.tt'la4 .hd5
1 7.tt'lb6 !!ad8.

If we delay too much the break,

t rying to combine it with ...e5 first,
White may get additional options:

Analysis 16...d4! 17.tt'lxd4 0-0 18..ie2

!!fd8 19.1ltfcl g6 20.!!he1 1!tfxe5.

A more sophisticated example:

Analysis

17.ig4 tt'lf4 18.tt'lxf4 exf4 19.e5!

VJff xeS 20..id4t.

OK, we have achieved ...d5.

what's next? The short answer is:
17...d4! 18.tt'lxd4 0-0 19.hh5
we aim for ...d4!.
gxh5 20.!!hel !!fd8 21.1ltff2 (21.f4
.if3). White's king lacks defenders
Bauer-Lopez Martinez so we can even ignore the e5-pawn:
Linares 201 3 21...a5! (21...1!tfxe5 22.@bl=) 22.@bl
a4-+.

Ana lysis

15...d4! 16.tt'lxd4 0-0 17.id3

�f d8 18.@bl g6 19.!!hf1 1!tfxe5.

49
 Part 2

Even if we cannot activate our Black will be threatening to

bishop, White has not any attack.
We easily defend the 7th rank:
26.hxg6 fxg6 27.gdhl 4Jf5.
Keep in mind the following typi­
cal trap White often overlooks:
build up a �+.i battery on the main
diagonal. Moreover, the f3 and h2-
pawns are weak.
Even without queens, Black's
bishop pair and an active rook com­
monly balance the game:

Analysis

15.g5? 4Jxe4! 16.fxe4 hg5! , win­

ning the queen because of the mate
on c2.; it also works after 15.4Jb6 ·

4Jxe4! 16.fxe4 .igS ! . The position is unclear.

Going deeper into the middle­

Game 7 Robson-Wang, Vue
game, I would like to bring your at­
rapid, N i ngbo 201 3
tention to the possibility of sacrific­
ing the exchange for a pawn. This
is a good way to neutralise the en­
emy attack.

Analysis

22....if4. Black is better due to

his domination on the dark squares.

50
Part 2 . The English Attack
Ta i manov - Ste p by Step
l .e4 c5 2 .c!l:'lf3 e6 3.d4 cxd4
4 . ttl xd4 ttlc6 5. ttlc3 '1Wc7 6 .te3
•
6 . . a6
.
6 ... ttlf6 practically eliminates
the English attack as a choice for
White, but why to jump out of the
fryi ng pan into the fire?! The arising
unexplored positions bring White
much better results than the Eng­
l is h attack itself! Although there is
a vast field for improvements and
investigation, I do not see any rea­
son to deviate from the main lines.
Let me note two important
ga mes after 6... ttlf6.
Giri-Bauer, Leon 2012: 7.f4 .ib4
8 . lt:l db5 �aS 9.e5 lt:ldS 10..id2 lt:lxc3
l l .ii.xc3 0-0 12.�d2
12...f6 (12....ixc3 13.lt:lxc3 f6
14.exf6 l'!xf6 15.0-0-0 d5 16.g3, Mot­
ylev-Shimanov, Aix-les-Bains 2011)
13.exf6 l'!xf6 14.g3 d5 1S ..ixb4 lt:lxb4
16.lt:ld4 �b6 17.c3 lt:lc6 18.0-0-0.
White has a stable positional edge.
Another possible argument in
favour of 6...lt:Jf6 would be to avoid
the 7..id3 main line with 7...b6 or
7...lt:Jxd4 8..ixd4 .icS 9..ixc5 �xeS
10.�e2. Again, I doubt the sound­
ness of these deviations.
Summing up, by delaying ...a6
Black hardly improves his chances
in the main lines while the addi­
tional options of White, especially
7.f4, look really unpleasant.
7.'1Wd2
7..ie2, 7..id3 and 7.f4 are subject
51
 Part 2
o f other parts of the book. I will con­
sider here two rare altern atives:
a) 7.a3 is a particula r move or­
d e r, which throws us out of our re­
p e rtoire. B lack's biggest problem is
h i s enormous choice. Virtually all
conceivable plans are playable, but
I u nderstand that such a statement
has no practical value. Schevenin­
gen fa � s will be especially happy to
_
face this mnoc uous move. I propose
to adopt a typ ical Taima nov set-up
for consistency sake:
7...b5
8.tt'lxc6
Or 8..ie2 .ib7 9.f4 tt'lxd 4 10.�x d4
l'!c8 ll.if3 (Anticipatin g u...tt'lf6.
lU!dl tt'lf6 12.e5 tt'ld5 would be OK for
Black) ll...h5 12.h3 h4 13 .0-0 tt'lf6oo.
8...�xc6 9.ie2
9.�d4 ib7 10.ie2 l'! c8 ll.l'!d1
tt'lf6 12.e5 should be Ill et by 12...
tt'ld5! . The trick is that 13.if3
is
not dangerous in view o f 13 tt'lxe
... 3
14.�xe3 �c7 15.ixb7 �x b7
16.0-0
b4 17.axb4 hb4 18 . tt'le4
l'!xc 2
19.l'!d6 l'!c6.
9...ib7 1 0.i.f3 �c7 1 1.0 -0
i.d 6
12.g3 tt'lf6 13.�d3 i.e S
wit h com ­
plex p l ay.
52
b) 7.�f3 tt'lf6! (7...b5 8.�g3;
8.tt'lxc6 �xc6 9.id3 .ib7 10.�g3; 7...
ib4 8.0-0-0 tt'lge7 9.�g3, Chepa­
rinov-Swiercz, Yerevan 2014, are
pleasant for White) 8.0-0-0 is
trendy.
The Scheveningen approach 8...
.ie7 (8...d6 is less accurate as it al­
lows 9.tt'lxc6!? bxc6 10.g4) 9.ie2
(9.tt'lxc6 bxc6 10.g4 is already un­
impressive due to 10...d5) 9...0-0
10.�g3 [10.g4 tt'lxd4 ll.l'!xd4 (11.
hd4 b5 12.g5 tt'le8 13.l'!hg1 ib7
14.'it>b1 l'!c8 15.id3 f5 16.gxf6 ixf6
17..ixf6 tt'lxf6 18.�g2 l'!f7) ll...d5
12.exd5 tt'lxd5] 10...d6! seems the
most logical retort to White's set­
up, because his queen hampers the
march of the g-pawn. After ll.f4,
Black will complete development
with ...id7-e8 keeping the tension
in the centre. See more details in
the annotations to game 5 Vallejo­
Spraggett, Catalunya 2013.
7 c!Df6
•••
7...b5 contradicts the spirit of
this book and my own understan­
ding of the Taimanov. I think that
Black should seek active piece play
and the knight on c6 has an impor-
 The English Attack

role in that design. So I have a

t a nt b) 8.f4 ! ?
goo d reason to avoid 7 . . . b5 8.ltlxc6.
This line enjoys a burst o f pop­
ularity. However, it might be short­
8 . 0-0-0
lived if White does not find some­
thing substantial against the rather
I n the early days of the English
forced endgame which arises after:
Attack, White used to play:
8 . . . ib4 !
a) 8.f3
Karjakin-Svidler, 2014, saw 8 . . .
It was aimed against 8 . . . ltlg4,
bS , but I believe that it i s important
bu t it is already clear that the
to provoke a3 in order to have a le­
k n ight jump after 8 . 0-0-0 is bad
ver for our attack on the queenside.
d u e to 9.if4 ltlge5 (9 . . . e5? ! is even
The game went 9.e5 ltlg4 10.ig1 ib7
worse, 10.ltld5 �d8 ll .h3 ! ±) 10 .ig3
11. 0-0-0 ltlxd4 12 .ixd4 ie7 13.id3
.' : . x d 4 1l.�xd4 f6 12.f4 icS 13.�d2
f5 14.h3 ltlh6 15J�hg1 0-0 16.�e3 (16.
: f7 14.e5t, Grischuk-Needleman,
g4 g6 17.1!lb1 if3 18.Eidf1 ic6 19.g5
Khanty Mansyisk 2 005.
ltlfl 20 .h4 is better for White) 16 . . .
8 . . . b5 9.g4 E\ac8 17.1!lb1 ic6 18.ltle2 idS 19.g4
g6 20.b3 �b7 2 l .Eidfl aS 22 .ltlg3
9 .ltlxc6 dxc6 is similar to
ic4 23 .ie4 idS 24.id3 draw.
1 0 . CiJxc6 ; 9. 0-0-0 transposes to the
main line. 9.id3

9 . . . ib7

9 . . . h6 10.ltlxc6 ! [10. 0-0-0 ltleS

( i n te nd ing . . . dS) ll.h4 b4 12.ltla4
d5 13.ie2 counts on the trick 13 . . .
dxe 4 14.ltlb6 �xb6 15.ltlc6 (15.ltlxe6
' ' \' a5 16.ltlxf8 ib7). However, Black
h as th e counter-blow 15 . . . ltld3+
1 6 .Jhd 3 �c7 17.if4 e3 ! .] sets some
p r oblems since 10 . . . dxc6 might be
s l ig htly better for White because 9 . . . ltla5
. . . h 6 has provided him with a le­
ve r o n
This is a critical line for White's
the kingside. Perhaps Black
eighth move, but I prefer the dou­
s h o uld try 10 . . . �xc6 and follow up
ble-edged: 9 . . . 0-0 ! . Now 10.ltlxc6
by . . . ib7,�c7, ie7, dS.
bxc6 ll.e5 ltlg4 12 .igl dS cannot be
1 0 . g5 ltlxd4 ! transposing to the of any concern to us. The only prin­
111 ai n l i n e. (In the event of lO . . . ltlhS cipled retort is 10.a3 ie7! 11. 0-0-0
l l . l2Jxc6, our knight would be mis­ bS 12.e5 ltlg4 13.igl ib7 with com­
Pl ace d on hS.) plex play.

53
 Part 2
We can undermine the enemy
centre with .. .f6 or push . . . b4. For
instance, 14.tt:lb3 b4 15.axb4 tt:lxb4,
or 14.'We2 ttlh6. Unfortunately, I do
not know of any practical example
yet.
10 .a3
In the event of 10.tt:lde2 . tt:lc4
ll .ix:c4 'Wxc4 12 .e5 tt:le4 13.'Wd3,
Black completes development with
either 13 . . . b5 or 13 . . . d5.
10 . . .ix:c3 ll.'Wxc3 'Wxc3+ 12 .bxc3
dS (In Shomoev-Grigoriants, Ta­
ganrog 2 0 14, Black tested 12 . . . d6,
but 13.ttlb3 is pleasant for White
who has pressure on the queenside
with ttld2, a4, gbl.) 13.e5 ! (13.exd5
tt:lxdS 14.i.d2 ttlb6 15.0-0 tt:lbc4=,
Ziaziulkina-Zhigalko, Minsk 2014)
13 . . . tt:le4 14.ttlb3 ! tt:lc4 15.ixc4 dxc4
16.tt'l a5
54
16 . . . i.d7!
16 ... b5? ! 17.i.d4 fS 18.exf6 gxf6
19.0-0, Edouard-Javakhishvili, Ca­
leta 2 0 14, may be defendable, but I
see no fun for Black.
17.i.d4 gcs 18.ttlxb7 .ic6 19.tt:la5
i.dS 2 0.gb1 fS ! . White's knight
might remain stranded on aS, his
doubled extra pawn is not worth
much. White also has to worry about
possible . . . ggs, . . . gS. 2 1 .gb6 is not a
threat due to 2 1 . . . \t>d?. Perhaps best
is 2 1.ttlb7 gS= with counterplay.
8 . . . ie7
8 . . . b5 is less accurate due to
9.i.f4 �b6 when 10.ttlb3! tt:lg4
ll.i.e2 ! tt:lgeS 12.�e3 assures White
of the edge. 10.ttlxc6 'Wxc6 ll.f3 also
deserves attention.
9.f3
9.f4 forces the play so it de­
mands memorisation : 9 . . . b5 ! (9 . . .
d 6 10 .i.e2 0 - 0 i s a sharp Schevenin­
gen position) lO.eS
10 .i.d3 gives Black time to repel
the c3-knight: 10 . . . b4 11.ttla4 gbs.
The game Gashimov-Movse­
sian, Reggio Emilia 2010, went fur­
ther 12.i>b1 d6! 13.b3 (13.e5 dxeS
14.tt:lxc6 \Wxc6 lS.fxeS tt'ldS) 13 . . . 0-0
14.tt:lxc6 'Wxc6 15.ghe1 \Wc7 16.'We2
.id7 with mutual chances.
10 . . . b4 !
This move provokes the follow­
ing interesting sacrifice:
 The English Attack

15.�xd6 �xd6 16.E!xd6 E!xa2 17.\ilb1

E!aS 18.ie2 .
13.exf6 gxf6. White has a forced
way to equality, see game 4 Wei Yi­
Stukopin, Kocaeli 2013.

9 ... b5 10.g4

a) 10.lt:lxc6 dxc6 brings about a

11.lt:lcb5 symmetric pawn structure where
the chances are roughly equal.
ll.exf6 bxc3 12 .�xc3 hf6 13 .g4 10 . . . �xc6 ll .g4 d6 is often seen
h6 14.h4 (14.ig2 ib7 15J'l:hg1 E!c8 in the Scheveningen, but I believe
16.\ilb1 lt:lxd4 17.�xc7 E!xc7 18.ixd4 that White is better prepared for an
is at least equal if not slightly bet­ attack.
ter for Black, Hracek-Akesson, Par­ ll.g4 eS 12 .h4 ie6
dubice 2013) 14 . . . ib7 15.E!h2 E!c8
16.�d2 . Black has a wide choice
here. Perhaps safest is 16 . . . hd4
17.ixd4 lt:lxd4 18.�xd4 0-0 19.g5 hS
2 0 .E!f2 d5 ! = , Guseinov-Macieja, Is­
tanbul 2012. In Wijk aan Zee 2 014,
Saric tried to contest this assess­
ment against Yu Yangyi. Instead of
attacking on the kingside, he tried
to exploit Black's "bad" bishop with
2 1 .E!e1 g6 22 .E!e3 aS 23.E!b3. How­
White may claim some space ad­
ever, the Chinese GM found a good
vantage on the kingside, but he has
way to obtain counterplay with the
no real threats. For instance, after:
pawn sac 23 . . . a4 ! 24.�xa4 �cS
13.g5, Black can exchange a pair
25.E!g2 (25.E!bf3 d4 26.E!a3 d3) 25 . . .
of rooks with 13 . . . E!d8, or he can
:l aS 26.�b4 �xb4 27.E!xb4 E!xa2.
preserve more pieces with 13 . . . lt:ld7
Bl ack has no problems in this end­
followed up by 0-0-0. 13 . . . lt:lh5 ! ? is
game.
good, too. At the same time, White's
ll . . . axbS 12.lt:lxb5 YNaS king's bishop lacks prospects.
Therefore,
12 . . . 'i9b8 might be the only way
to play for a win, but it is risky as 13.ih3 looks consistent. Moty­
Wh ite has good compensation for lev-Fominyh, Ubeda 2001, went 13 . . .
t he pawn after 13.exf6 gxf6 14.lt:ld6+ E!d8 14.�h2 E!xd1+ 15.E!xd1 0-0? !
C l 4. \ilb1 d5 15.�f2 ia6 16.ib6 hbS 16.g5 lt:lhS 17.he6 fxe6 18 .'Wh3t,
17 . hb5 �b7 18.f5 E!b8) 14 . . . hd6 but 13 . . . lt:ld7 evens the chances.

55
 Part
Then 14.g5 would not make much
sense in view of 14 . . . .ixh3 15J'!xh3
0-0- 0 = , but 14.ltJd5 cxd5 15.exd5
.ixd5 16.�xd5 �d8 17.�d3 ltJb6 is
also fine for Black.
White may wait to see where the
black king castles:
13.'it>bl. Perhaps castling is not
necessary at all. 13 . . . �d8 14.�h2
�xdl+ 15.ltJxdl h6 16.i.h3 �a5 17.b3
ltJd7 gives counterplay on the dark
squares.
Eliseev chose twice 13 .�f2 . Pot­
kin answered 13 . . . a5 (I do not un­
derstand this move) 14.i.c5 0-0
15.h5 .ixc5 16.�xc5 ltJd7 17.�d6
�xd6 18.�xd6 �fc8 = . Dragun opted
for 13 . . . .ib4, but after 14.a3, he sud­
denly changed his mind with 14 . . .
!J.e7?. I would rather try t o trade
a pair of rooks with 13 . . . �d8 (13 . . .
ltJd7 i s probably less accurate due to
14.ltJd5) and decide later whether to
castle at all.
b) 10.'it>b1 ! ? is a tricky wait­
ing move which is not deprived of
venom. My recommendation is to
transpose to line A2 with 10 . . . 0 - 0 !
ll.g4 ltJxd4 12 . .ixd4 i.b7.
Black would be nastily surprised
if he tried to transpose to the main
line with 10 . . . ltJd4? ! ll.i.d4 i.b7
(ll . . . b4 12.ltJa4 �b8 13.�f2) 12 .�g5 !
when 12 . . . 0-0 loses to 13.e5+-. In
the blitz game Ivanchuk-Wang Yue,
Beijing 2 0 13, Black played 12 ... b4
13.ltJa4 0-0 14.ltJb6 �ae8 15.�g3
d6 16.ltJc4 e5 17.i.b6;!; when 17 . . .
�b8 would have hold on, but I do
56
2
not like such static positions with a
clear plan for White.
The other obvious alternative is
10 . . . i.b7 hoping for ll.g4 ltJxd4. In­
stead, White can set serious practi­
cal problems with the still untested
ll.i.f4 ! ? e5 12.ltJf5 exf4 13.ltJd5.
Without the inclusion of 1 0 . 'it>b1
i.b7, Black would have 13 ... �e5 ! ,
but now the b7-bishop i s hanging.
13 . . . �b8 ? ! keeps tha extra piece,
but 14.ltJxg7+ 'it>f8 15.ltJf5 is a mess.
Still, I would take White without
any hesitation. Look at the varia­
tion 15 . . . i.d8 16.ltJxf4 h5 17.g3 ! . It is
unclear how to get rid of the dom­
inating knight on f5. To be sure,
17 . . . ltJe7 fails to 18.ltJd6 i.c7 19.e5±.
It is better to eliminate to a near­
ly equal endgame with 13 . . . ltJxd5 !
14.exd5 ltJb4 15.d6 �xc2+ 16.�xc2
ltJxc2 17.ltJxg7+ 'it>f8 18.dxe7+ 'it>xg7
19.'it>xc2 �ac8+ 2 0 . 'it>b1 i.c6 2U:!d4
�he8 2 2 .�xf4 �xe7 23 .i.d3 h6. Black
will be staying with his rooks on the
e-file, avoiding exchanges.
10 . . . ltJe5 returns to the older
treatment of the Taimanov. Per­
haps Black should check the posi­
tion after ll.g4 (ll.ltJb3 �b8 12.�f2
ltJc4 13 . .ixc4 bxc4 is fine for Black)
11 . . . 0-0 12 .g5 ltJhS 13.f4 ltJg4 14.e5
 The English Attack
b4 15.li:'le4 .ib7 16 . .ig2 d6 17.li:'lf6 +
ixf6 18 . .b:b7 �xb7oo, but so far
White has won both games where it
was tested.
10 ••• lt:'lxd4!
A. ll . .b:d4 ; B. ll .�xd4
A. ll . .hd4 .lb7
The immediate attack in the cen­
tre with ll . . . b4 1 2 . li:'l a4 :Bb8, plan­
ning . . . dS, seems a plausible alter­
native. This branch is still unex­
plored, maybe because Black ob­
tains more than satisfactory results
in the main line. A possible continu­
ation is 13.rit>bl 0-0 14.�f2 dS lS.gS
tt:lhS 16.exd5 exdS 17.h4 :Be8.
An important juncture. White
should decide whether to continue
his initial plan, or focus on the cen­
tre :
Al. 12 .g5 ; A2 . 1 2 . 'it>bl; A3. 12 . .id3
Minor alternatives are :
a) 12 .h4. White aims to play hS
first, but this idea is too slow and
does not hamper the break 12 . . . d5
since 13.g5 li:'lhS 14.exd5 is met by
14 . . . b4 15.li:'le4 .b:dS. White's knight
stands well on e4, but we can al­
ways exchange it. The game Dam­
brauskas-Schiendorfer, ICCF email
2 0 1 0 , went 16.�e3 :Bc8 17.:Bd2 .b:e4
18.fxe4 0-0 19.:Bhh2 eS 20 . .ib6 �c6
2 l . l'!d5 li:'lf4 22 .l'!a5 h6! with an ini­
tiative on the kingside.
b) 12.�f2 0-0 13 . .id3 (13 . .ib6
�c6 ! 14.g5 li:'lh5) 13 . . . b4 ! ?
Topalov chose against Karja­
kin in Beijing 2 0 13 the more flexi­
ble 13 . . . :Bac8 14.li:'le2 d6 15.li:'lg3 li:'ld7
16.:Bhel li:'l cS 17.rit>bl :Bfd8 18.b3 eS
19 . .lb2 .ih4 ! 2 0 . :Bgl g6, with the
better game.
14.li:'le2 (14.li:'la4 dS lS.gS li:'ld7;
15.li:'lb6 dxe4 ! - never allow e4-e5
when White's knight is on b6 and
57
 Part 2
can take ours on d7. We need the
knight for counterplay from cS.)
14 ... d5 15.e5 lLld7 16.f4 aS 17.�b1
(17.f5 lLlxe5) 17 . . . a4t.
Al. 12.g5 �h5 13.�e2
a) 13.�b1 gives a tempo for cast­
ling: 13 . . . 0-0
a1) 14.lLle2 f5 15.gxf6 h£6
16.hf6 lLlxf6 17.h4 with a balanced
game. Black's pressure down the f­
file neutralises the open g-file. The
email game Salvador Marques­
Schiendorfer 2010 went 17 .. J�f7
(17 . . . l'!ac8 18.l'!g1 lLlh5) 18.l'!g1 'l;'ih2
19 .'l;'ie3 l'!c8 20 .a3 lLlhS 21.l'!g4 Wffc7
2 2 .l'!d2 'l;'icS 23.lLld4 eS 24.'l;'ib3 l'!cf8
25.l'!g5 g6 26.lLlf5 �h8 27.i.e2 gxfS
28.l'!xh5 fxe4 draw.
a2) 14.i.e3 does not hamper 14 . . .
dS. After 15 .exd5 b 4 16.lLla4 i.xdS
17.lLlb6 l'!ad8 18.lLlxd5 l'!xdS 19.'l;'ic1
l'!xd1 20.'l;'ixd1 lLlf4, Black seized the
initiative, Bauer-Ganguly Linares
2013.
a3) 14.l'!g1 takes control of g3 and
prepares h4, i.h3. It seems that 14 . . .
d S i s bad due t o 15.exd5 b4-? 16.lLle4
i.xdS 17.lLlf6+, but in fact the the-
58
matic break is still possible. Only
instead of 15 . . . b4, Black should play
15 . . . lLlf4 ! , recovering the pawn with
a satisfactory position. However,
White has no immediate threats so
why not keep the tension for a while
with 14 . . . l'!ac8 . White has not any­
thing better than 15.i.e3 (15.l'!g4 g6;
15.i.h3 lLlf4), when 15 ... d5 16.exd5
b4 is already possible.
a4) 14.i.d3 lLlf4 (14 ... e5 ! ? 15.i.f2
lLlf4, intending to capture the gS­
pawn on 16.lLle2 , may be stronger)
15.h4 l'!ac8 16.lLle2 lLlxd3 17.cxd3 eS
18.i.e3
Here 18 . . . d5 19.lLlg3 d4 2 0 .i.f2
g6 2 1.l'!c1 Wffd 7 2 2 .lLle2? (or 2 2 .f4 ! ?)
might be dangerous so Morozevich
opted for the forced line 18 . . . i.b4 !
19.Wixb4 Wic2 + 2 0 . �a1 'l;'ixe 2 = .
b ) 13 .i.e5 leads t o a sharp end­
game after:
13 . . . 'l;'ixe5 !
It is dangerous to avoid simpli­
fication by 13 . . . 'l;'ic6 since 14.a3 will
retain White's grip on the centre :
14 . . . 0-0 15.i.h3 fS 16.l'!he1 l'!ae8
17.exf5 exfS 18 .i.g2;!;;
14 . . . l'!d8 15.�b1 (15.i.h3 WicS
16.f4 h6) 15 . . . 0-0 16.i.e2 or 16.i.h3
 The English Attack
whereas Black's queen does not
control the critical square f4.
14.1Wxd7+ l!ifB 15.�xb7 h:g5+
16.1!ib1 �b8
17J:�d7
17.'\Wc6 g6 ! ? would be double­
edged while 17 . . . �c8 18.�xc8 + (18.
�b6 .if6 19.tt:le2 .ie5 2 0.l:'lg1 g6)
does not change this assessment.
17. . . �xb7 18.l:'lxb7 .if6 19.a4 (19.
'Lle2 g5) 19 ... h:c3 2 0 .bxc3 bxa4
21.h:a6 g5 2 2 . .ib5 l!ig7 23.l:'ld1, Ter
Sahakyan-J. Polgar, Yerevan 2 0 14.
Black must keep both rooks here
or White's c-pawn will run forward
quickly. Perhaps safest is 23 . . . tt:lf4
24.c4 !lacS ! , since 24 . . . tt:lg6 ! ? 25.c5
'Lle5 26.c6 tt:lxf3 27.c7 tt:lxh2 is rath­
er unclear.
c) 13.h4 ! ? counts on the trick
13 . . . 0-0? ! 14 . .ih3 tt:lf4 (14 . . . b4
15.'Lle2 e5 16 ..if2 d5 17 . .ig4 tt:lf4
18.tt:lxf4 exf4 19.e5t) 15 . .ie3 ! tt:lxh3
16.1Wxd7±, Muzychuk-Galojan, Bel­
grade 2 0 13. Black should kill the .ifl
with 13 . . . tt:lg3 ! 14.l:'lg1 tt:lxf1 15.l:'lgxf1
0-0, killing two birds with one shot
- he has solved the problem of his
knight at the edge of the board re­
moving a dangerous attacking piece
at the same time. The game might
continue 16.1!ib1 b4 17.tt:la4 (17.
tt:le2 e5 18 . .if2 d5 19.tt:lg3 l:'lfd8) 17 . . .
d 5 (or 17 . . . i.c6 18.tt:lc5 a S 19.tt:ld3
d5 2 0 .�e3 a4oo) 18.exd5 (18.e5 aS
19.�f2 i.c6 2 0 .tt:lc5 a4t) 18 . . . h:d5
19.tt:lb6 l:'lad8 2 0.tt:lxd5 l:'lxd5+.
d) 13.i.d3 is an arguable idea,
because Black can always equalise
by trading the bishop after . . . tt:lh5-
f4xd3. 13 . . . 0-0 14.tt:le2 e5 15.i.e3 g6
(15 . . . i.c5 16.h:c5 �xc5 17.1!ib1 l:'lac8
18.l:'lhf1 = ) 16.tt:lg3 tt:lf4 17.h4
17 . . . l:'lac8 (17 . . . 'Llxd3+ 18.'1Wxd3
!ladS = ) 18.hf4 exf4 19.'Lle2 d5
2 0 .�xf4 (20.'Llxf4 .id6 21.ttle2
l:'lfe8 ! ) 2 0 . . . .id6 21.e5 (21.1Wg4 dxe4
2 2 . fxe4 .ie5iii) 2 l . . .h:e5 2 2 .1Wd2
.ig7 23.f4 l:'lfe8 24.1!ibl. Dominguez­
Potkin, Havana 2012 saw here
24 . . . l:'le4? 25.h:e4±. Instead, 24 . . .
d 4 25.l:'lh2 l:'le3+ would have given
Black an initiative.
13 .•• l'3c8 14.l'3gl
14.h4 might prove superflu­
ous if Black defended with .. .f6 or
. . .f5. For instance : 14 . . . 0-0 15 . .ih3
f5 16.gxf6 (16.exf5 h:f3 17.fxe6 d5
winning the exchange) 16 . . . tt:lxf6.
59
 Part 2

14 . . . 0-0 15 . .th3 g6 (or 1S . . . dS 13 ..id3 b4 14.li:Je2 eS 1S.gS

16.eS g6) 16.�bl d5 17.e5 �g7
exd4 was pleasant for Black in
Yu-Nepomniachtchi, St. Peters­
burg 2 0 1 2 . Amonatov-Saiyn, Mos­
cow 2 0 14, showed that even if
White achieved everything he want­
ed, Black would still be fine - 13 . . .
�ac8 14.�f2 d 6 1S.h4 l!J d 7 16.gS b4
17.l!Je2 dS 18.exdS hdS.
The text move deserves atten­
tion if for no other reason than be­
cause it was played at highest level.

Black has a good game. He will

push his pawns to a4 and b4 keep­
ing the tension. White's only plan is
h2-h4-hSxg6, but then Black recap­
tures by . . . fxg6 and easily defends
the h7-pawn along the seventh rank
with . . . l!JfS. This suggests that Black
should leave his rook on f8 in or­
der to ensure the possibility of tak­
ing .. J:!xfS after hfS. Almasi-Mov­
sesian, Warsaw 2 0 13, went further
13 . . . d5 !
18.f4 b4 19.!g4 aS 2 0 .l!Jcl (or 2 0 .h4
a4 2 1.ltlg3 !cS) 2 0 . . .!a6 2 U�g2 a4
I suppose that this novelty is
2 2 .h4 !cS (clearing the 7th rank)
the best retort to White's tricky last
23.hcS �xeS 24J'%h2 �c7 (24 . . . a3
move.
2S.b3 �c3) 2S.hS. Here 2S . . . �b6
26.hxg6 fxg6 27.�dh1 l!JfS+ would Alternatively:
have been promising. 13 . . . �ac8 14.�g1 .ic6 (14 . . . b4
1S.gS �hS 16.l!Ja4) 1S . .id3 b4oo oc­
curred in game 9 Caruana-Svidler,
A2 . 12.�bl 0-0 13.�f2 Rhodes 2 0 13 ;
1 3 . . .b 4 14.l!Ja4! (14.g5 l!Jh5
13.!e3 does not prevent 13 . . . b4
1S.l!Ja4 dS 16.exd5 hd5 17.l!Jb6
14.l!Je2 d5 ;
�ad8=) 14 . . . d5 is positionally dubi­
13 .h4 �ac8 14.!d3 turned out ous since White trades his awkward
badly for White after 14 . . . eS 15.!e3 knight for an important black piece.
!b4+, Kurayan-Stukopin, Kocaeli _ 15.l!Jb6 �adS 16.eS l!Jd7 17.l!Jxd7.
2013. Black has no counterplay here.

60
 The English Attack
14.g5 ttld7 15.exd5 b4 16.ttle4
16.dxe6 fxe6 17.'�'e2 eS is rough-
ly equal : 18 .'�c4+ �xc4 19 . .bc4+
®h8 2 0 .ltl d5 hg5 2 1.i.f2 = .
16 hd5 17.J.d3 gfd8
•••
White's pieces look daunting,
but Black can quickly trade a cou­
ple of minor pieces, for instance,
... he4 followed up by . . . i.cS.
18.:1!hg1 eS defends everything.
Then Black can thrust forward his
own a-pawn.
18.�h4 g6 19.�f2 :i!ac8 2 0.h4
ttlc5 also neutralises the attack be­
cause 2 l .ltlf6 + ? loses to 2 1 . . .hf6
22.hf6 ha2 + ! 23.'>t>xa2 b3+ while
21.hc5 hc5 2 2 .ltlf6+ '.t>g7 23.�e2
ie7 (or 23 . . . �f4) is roughly ba­
lanced.
Remains 18 .h4 :i!ac8 19.h5, but
this drops the g5-pawn after 19 . . .
he4 ! 2 0.he4 .bgS.
A3. 12 .J.d3 e5
It is positionally sound to re­
strict the d3-bishop. Of course,
12 ... 0-0 cannot be bad either, for
instance, 13.h4 b4 14.ttle2 eS 15.i.f2
dS ; 13.ltle2 d5 14.e5 ttle4 15.'�e3 ltlc5.
13.J.f.! 0-0 14.g5 ttlh5!
Svetushkin-Negi, Linares 2 0 13,
saw 14 . . . b4? ! 15.lt:Ja4 lt:JhS 16.ltlb6
:i!ab8 17.i.c4 i.c6 18.lt:Jd5 hd5
19 . .bd5 aS, but White's bishop is
at least theoretically better than
Black's one.
15.ttl d5 hd5 16.exd5 ttlf4
17.h4 (17.hh7! '.t>xh7 18.d6 hd6
19.�xd6 :i!ac8 is roughly equal)
17 f5 !
•..
This move assures Black o f a
good game since 18.gxf6? lt:Jxd3 +
would give him pressure down the
f-file.
B . ll.�xd4 J.b7 12 .g5
12.'>t>b1 :i!c8 13.�d2 fails to 13 . . .
b 4 14.lt:Ja4 lt:Jxe4 15.fxe4 he4
16.ha6 �xc2 + .
12.:1!g1 :i!c8 13.:1!g2 (13 .g5 will
transpose to the main line) makes
some sense, but White abandons
61
 Part 2
the dangerous manoeuvre .ifl-h3-
g4. Black should not worry about
this attack anymore so he can pre­
pare short castling with 13 . . . .ic6
(but not 13 . . . b4? ! 14.tt::l a4 .ic6
1S.tt::l b 6 �b8 16.gS tt::l hS 17.�gd2 !
eS 18.�c4 �xb6 19.�xd7! ! , Pruijss­
ers-Zwirs, Dieren 2 0 13) 14.gS tt::l h S
1S.�f2 0-0. Now 16.f4? ! f6 or 16 ..ih3
tt::l f4 17 ..ig4 hgS are clearly better
for Black, so White should prob­
ably continue with 16.eS, but then
the simplest 16 . . . �fd8 17.tt::l e 4 he4
18.�xe4 d6 19.exd6 �xd6 equalises.
12 . . . tbh5 13 . .ih3
This is the only way to prevent
the freeing . . . d7-dS. Alternatively:
a) 13.�d2 �c8 14.hbS is only
good for a draw.
14.�g1 is innocuous due to 14 . . . eS
(or 14 . . .icS 1S.�xc5 �xeS 16.hcS
�xeS= 17.a4 b4 18.tt::l a 2 aS 19 . .ibS
ic6 20 .hc6 �xc6) 1S.�d3 (1S.�a7
tt::l f4 16 . .ib6 �c6 17.tt::l dS id6 18 . .if2
tt::l xdS=) 1S . . . .ib4.
14 . . . axbS 1S.tt::l xbS
1S . . . �c6 (1S . . . �d8 does not lose,
but White's heavy pieces are bet­
ter after 16.�a7 0-0 17.�xb7 hgS
62
18.hgS �xgS 19.tt::l c3 ! �e3 2 0 .�b4
�xf3 2 1 . �gU) 16.tt::l a7 �c7 (some
commentators claim that 16 . . . eS? !
17.�xeS �c7 i s unclear, but i n my
opinion, the endgame is obviously
pleasant for White, e.g. 18.�d4 �d8
19.tt::l b S �c6 2 0 .tt::l d 6+ hd6 2 1.�xd6
�xd6 2 2 . �xd6 0-0 23.b4 fS 24.gxf6
tt::l xf6 2S.icS ic6 2 6.c4;!;) 17.tt::l b S= ,
Dominguez-Caruana, Paris 2 0 13.
b) 13.@b1 �c8 puts· the question
to White : what is he going to do with
his queen?
14.�d3 is unaesthetic. To be
sure, 14 ... b4 1S.tt::l a4 dS 16.exdS
hdS 17.tt::l b 6 tt::l f4 (the thematic 17 . . .
�d8 fails here t o 18 .�xa6) 18.tt::l x dS
tt::l xdS 19 .ic1 would be pleasant for
him, but the immediate 14 . . . dS !
1S.exdS �d8 solves allthe problems.
14.�d2 admits that capturing
on d4 by queen was pointless. That
does not mean it is a bad move, but
it is certainly not the most princi­
pled opening idea. We have the
usual choice between . . . ic6, . . . 0-0
or the . . . dS-break. For consisten­
cy with the previous examples, let's
check the most straightforward
plan : 14 . . . b4 1S.tt::l e2
1S.tt::l a4 dS 16.exdS hdS 17.tt::l b 6
�d8 18.lLlxdS �xdS 19 ..id3 0-0
 The English Attack
leads to a typical position with
comfortable play for Black. The
bishop pair is inefficient due to the
weakness of f4. Black will be trying
to trade dark-squared bishops in
order to bring his knight into play.
15 . . . d5 16.e5 d4 ! . After this deep
counter-sac Black seized the initi­
ative in game 6 Volokitin-Najer,
Muelheim 2 0 13.
c) 13.a3 l"lc8 14.'tt> b i prevents
the . . . dS break, but gives Black time
to castle and undermine the cen­
tre with .. .fS : 14 . . . -icS 15.�d2 .ixe3
16.�xe3 0-0 17.l"ld2 (17.l"lg1 t!lf4
18.h4 l"lfd8 revives the plan with
... ds, e.g. 19.l"ld2 dS ! 2 0 .e5? d4 !
2U''l xd4 t!ldS) 17 . . .f5 ! . The pressure
down the f-file assures Black of the
initiative.
d) 13.lLle2 l"lc8 14.�d2 (14.l"ld2 ? !
dS was tested i n Inarkiev-Schloss­
er, Yerevan 2 0 14 - 15.exd5 .ixdS
16.4Jf4 4Jxf4 17 ..ixf4 �aS 18.�xg7
l"lf8 19.l"lxd5 �a4 2 0 .l"ld2 �xf4=
2l.h4? l"ld8 2 2 . .id3 .ib4 23.l"lhd1
when 23 . . . .ixd2+ 24.l"lxd2 'tt> e 7
would have been difficult for White)
14 . . . d5 1S.e5 d4 ! is considered in the
annotations to game 6. The play is
very similar to line b.
13 . . .f6 is inconsistent since our
main plan is to break in the centre
with . . . dS. White can maintain the
tension on the kingside with 14 . .ig4
(14.'tt>b 1 fxgS 1S . .ig4 4Jf6=) 14 . . . 4Jf4
1S.�d2 or 15.h4 ! ? .
1 3 . . . b4? 14.4Ja4 d S fails to
1S . .ig4.
The rook's move gains a cru­
cial tempo because besides target­
ing c2, it also enables . . . .icS. The
trade of dark-squared bishops be­
comes Black's primary strategic
goal when his knight reaches hS.
Thus, 14.l"ld2?! .icS ! would be ex­
cellent for him.
14.�g4
The only way to protect c2 while
evading . . . .icS is 14.�d2? ! , but this
retreat completely takes the sting of
.ig4 since Black can answer it with
. . . g6. So he can already push :
14 . . . b4 15.lLle2
15.4Ja4? ! dS ! (1S ... .ic6 is also
good - see game 7 Robson-Wang
Yue, Ningbo 2 0 13) 16.exd5? (16.
'tt> b 1 dxe4 17 . .ig4 h6! 18.f4 hxgS
19 .fxg5 g6+) 16 . . . .ixd5 17 . .ig4
63
 Part 2

15 . . . i.xa2 ! ! 18.4Jb6 (18.i.xh5 Then 17.�xd7? fails to 17 . . . i.c5 ! and

0-0+) 18 . . . 0-0 19.4::\ xcB E:xc8 . It White's dark squares around his
turns out that White has no defence king gape wide.
against 20 . . . i.b3- + .
16 . . . cxb 2 + 17.'it>xb2 gf8
1 5 . . . . d 5 16.i.g4 g 6 17.e5 d4 ! . Do
not allow the opponent to blockade
the centre ! 18.4Jxd4 0-0 19.i.xh5
gxh5 2 0 . E:he1 (White did not last
long after 20.E:hg1 E:fd8 2 l.�f2 a5
22 .E:d2 a4 23.g6 hxg6 24.4:\xe6 E:xd2
25.�xd2 �xeS- + , Espinosa-Hevia,
Merida 2 0 13) 20 ... E:fd8 2 l.�f2 (21.
f4 i.f3) 2 1 . . . a5 ! (21.. .�xe5 2 2 .'it>b1=)
22.'it>b1 a4--+.

14 . . . b4 15 . .bh5 bxc3 16.�xg7

Browne-Van Kampen, Reykja­
vik 2014, introduced the novelty
16.b3? ! . I guess it was an improvisa­
tion since 16 . . . 0-0 would be strong.
White is better co-ordinated, but
his unsheltered king allows Black to
maintain the balance. See game 8
Demetrio-Schiendorfer, ICCF email
2009.

64
Pa rt 2 . The English Attack

Ka n - Ste p by Step

l.e4 c5 2.ti�f3 e6 3.d4 cxd4 7.a3 bS 8 . .id3 .ib7 9.0-0 ll'lc6 is a

4.�xd4 a6 5 . � c3 �c7 6.i.e3 dream Sicilian.

After the text, Black has a wide,

but pleasant choice :

A. 7 .ib4 8. c!Dde2
•.•

8. 0-0? ! hc3 9.bxc3 d6! (9 ... d5?

lO.exdS liJxdS l l.id2 0-0 12 .�h5
g6 13 .�h4t) 10.�f3 eS ll.ll'lfS .ixfS
1 2 .exf5 ll'lbd7 13 J�;:abl dS.

8 ••• d6
This set-up is only a faint at­
tempt to reach the English Attack 8 . . . d5 9.a3 ! is not too clear.
since Black attacks first with . . . .ib4
9 . 0-0 c!Dg4 = .
and seizes the initiative. He only
has to start with :
B. 7 .ic5 8 .�e2
.!Df6 !
•••

6 •••

Or 8 . 0- 0 d6 9.�e2 liJbd7
Hellsten recommends 6...
ib4 7.c!Dde2 ll'lf6 8.a3 .iaS ( 8 . . .ie7 8 ••• d6 9 . 0-0-0 0-0.
9.liJa4 ! ?:t) 9.b4 .ib6, but instead of
10 .hb6, White can gain space by
10 .if4 ! t �dB l l.es. C. 8 ••• b5 9.0-0 J.b7 10.a3 J.e7

7.J.d3 Miladinovic played lO . . . hS ! ?

l l.h3 ll'lc6.
7.f3 does not hold the centre due
to 7 . . . .ib4 8 .�d2 (or 8.f3) 8 . . . d5. ll.f4 h5 +±.

65
Part 2. The English Attack

Complete Games

4. Wei,Yi-Stukopin 18.Wd2 �xb2 19.�e2 �f6? 20.�f3

Wch U20 Kocaeli 22.09.2013 would be a terrible mistake.

l.e4 c5 2.tt.Jf3 e6 3.d4 cxd4 17.Wd2

4.tt.Jxd4 tt.Jc6 5.tt.Jc3 Wfc7 6.�e3 a6
7.'�d2 tt.Jf6 8.0-0-0 fi.e7 9.f4 b5
10.e5 b4 ll.tt.Jcb5 axb5 12.tt.Jxb5
Wfa5 13.exf6 gxf6

17...�a2!

17...�xb2?! 18.l'l:b3 is in White's

favour: 18...�al 19.�c5!± or 18...
14.tt.Jd6+ �a2 19.�b5.

White's king is the more vulner­ 18.l:�b3 gxb2

able one after 14.c4 bxc3 15.ltJxc3
0-0. This move lets the white bishop
to b5. It would be simpler to keep
14...hd6 15.Wfxd6 Wfxa2 White passive by 18...�bl! 19.l'l:gl
16.�d3 (16.�b5 �al+ 17.Wd2 l'l:xb2. The difference is obvious af­
�xb2) 16...Wfal+ ter 20.�d3 �a2 21.l'l:xb2 �xb2. This
position is unclear, with mutual
It is better to play for attack. The chances: 22.�c5 (22.�c5 �b7) 22...
passive defence 16...f5 17.�c5 �al+ �c3+ 23.We2 b3 (23...f5 24.l'l:el f6

66
 The English Attack

2S.I!?f2 l!?f7 26.l'!e3 �d2 + 27.l'!e2 = ) 8 . 0-0-0

24.cxb3 �xb3 25.l'!b l = .

19 . .ib5 �a2 ? !

Perhaps Black did not notice

that he had perpetual check after
19 . �xhl 2 0 .hc6 dxc6 2 l.�xc6+
..

m£8 22 .i.d4 �xh2 23.�xc8+ <Jlg7

24.hf6+ <Jlxf6 25.�xh8+ l!?e7
26.�xb2 �xg2 = .
8 . . . d6
20J�xb2 �xb2 2 1 .ic5 ? !•

In my opinion, 8 . . . .ie7! is the

White misses the strong rede­
ployment 2 l .l'!cl ! �c3 + 2 2 .1!7e2 l'!g8
23.g3 eS 24. <Jlf2 hS 25.l'!dl, with an
initiativ e, e.g. 25 . . . h4 2 6.i.a4 hxg3 +
27.hxg3 e4 2 8 .�d5. After the text,
the game is balanced.
better move order, having in mind
9.�xc6 bxc6 10 .g4 dS. Let's consid­
er other alternatives :
a) 8 . . . .ib4? ! 9.ll:Jxc6 �xc6 (9 . . .
hc3 10.ll:Jd4 .ib4 ll . .if4) 10 .i.d4
hc3 ll .hc3 �xe4 12 .�xe4±.

21 �c3 +
• .• 2 2 . <Jldl .ia6 b) 8 . . . b5 9.ll:Jxc6 �xc6 10 .e5;!;.
23.ha6 (23 .hc6 �al+ 24.1!7d2
c) The central approach 8 . . .
�c3+=) 23 �al+ 24.<Jld2 �xa6
ll:Jxd4 9.hd4 e S 10 . .ie3 d6
•••

25 .td4 ggs 2 6.g3 �b7 27 .ic5

• •

�b8 2 8.�xb 8 + �xb8 29 . gal

li)c6 30.ga8+ �d8 3 1 .f5 exf5
32.hb4 f4 33.gxf4 draw.

5. Val lejo Pons-Spraggett

Cata l u nya 02.02.20 1 3

l.e4 c5 2 . �f3 e6 3.d4 cxd4

4.�xd4 �c6 5 . � c3 �c7 6 .ie3 •
looks reasonable, but White has
li)f6 7.�f3 a6 a clear plan : he pushes his pawns
to gS and hS and gains space with
I have also analysed the new ll:JdS followed by exdS. Black lacks
move 7 . . . i.a3? ! . It leads to cra­ counterplay. For instance :
zy complications, but White's in­
ll.h3 !
itiative is rather dangerous after
8.0-0-0 or 8.�db5. ll . .ie2 .ig4 12.ll:Jd5 ll:JxdS

67
 Part 2
13.'?9xg4 l"!c8 14.E!d2 h5 15.�h3 lt:)c3
16.i.b6 lt:)xe2+ 17.l"!xe2 '?9d7=.
ll . . .b5 12 .a3 ! (12 .i.d3 i.e? 13.g4
b4 14.lt:)d5 lt:)xd5 15.exd5 a5 is un­
clear) 12 . . .i.b7 13.g4 i.e? 14.c.!lb1!
(14.g5 lt:)xe4 15.lt:)xe4 d5 16.i.d3
l"!c8=)
Now 14 . . . lt:)xe4 15.lt:)xe4 d5
16.lt:)c3 d4 17.lt:)d5 hd5 18.'?9xd5
E!d8 leaves Black a pawn down af­
ter either 19.hb5+ axb5 2 0 .'?9xb5+
l"!d7 2l.i.c1 0-0 22.E!heU, or 19.�e4
dxe3 20.l"!xd8+ '?9xd8 2 l.fxe3 0-0
22 .i.d3 g6 23.'?9xe5.
The queenside attack with 14 . . .
ll:Jd7 15.h4 E!b8 16.h5 0 - 0 17.g5 i.c6
is not too efficient due to 18.lt:)d5 !
hd5 19.exd5 b 4 2 0.a4 b 3 2 l .c3
and White's threats are much more
tangible, e.g. 2 1 . . .f5 2 2 .gxf6 ll:Jxf6
23.'?9f5�.
9.i.e2 i.e7
9 . . . h5? ! 10 .h3 lt:)e5 turned out
well for Black in Shankland-Wang
Chen, Las Vegas 2 013. However, af­
ter 10.ll:Jxc6 ! bxc6 ll.c.!lb1, White's
pieces stand more harmoniously.
10.'?9g3 0-0 ll.f4
68
The first critical moment of the
game. Black should decide how to
finish his development. The en­
gines propose 11.. .d5 12.exdS· lt:)xd5
13.ll:Jxd5 exdS 14.i.f3 l"!d8 15.E!he1
i.f6, but it is difficult to defend such
a position over the board. White
has a lasting pull, for instance : 16.c3
h6 (16 . . . g6 17.f5 �xg3 18.hxg3 hd4
19.hd4 i.x£5 2 0 .i.b6 l"!e8 2 1 .hd5
i.g4 2 2 . l"!f1) 17.c.!lbl.
Still, I do not approve of Black's
last move either. This prophylax­
is was not obligatory at all. Per-
haps Spragget did not like ll . . . i.d7!
in view of 12.e5, but then 12 . . . lt:)e8
13.c.!lb1 (13.l"!hfl l"!c8 14.i.d3 ll:Jb4
15.c.!lb1 ll:Jxd3 16.E!xd3 , Wang Jue­
Lei Tingjie, Xinghua 2 0 14, 16 . . .
b5) 1 3 . . . l"!c8 i s quite solid. The
play may continue with 14.i.d3 b5
15.a3 b4 16.axb4 (16.ll:Jxc6 hc6
17.axb4 E!b8) 16 . . . lt:)xb4 17.i.e2 l"!b8
or 14.l"!he1 b5 15.i.d3 (15.a3 b4)
15 . . . g6. White's space advantage re­
mains, but he has not a pawn lever
on the kingside to use.
12.ghfl
 The English Attack
I doubt that White's rook's best
place is on fl. It might be more use­
ful from e1 or even from its cur­
rent stand. A waiting approach like
12 .@b1 .id7 (12 . . . bS 13.lt:Jxc6 �xc6
14 . .if3) 13 . .if3 looks more unpleas­
ant. Note that the plan with e4-eS
is risky, because it releases Black's
pieces : 13.lt:Jf3 �ac8 (13 . . . eS 14.fxeS
li:lxeS 1S.lt:Jd4 bS 16 . .igS .ie6 17.lt:JfS
li:lxe4 18.lt:Jxe4 .ixfS 19 . .ixe7 �xe7
20.li:lxd6 .ig6 2 1.�d2 f6 2 2 . �hd1)
14.eS lL'ldS 1S.lt:JxdS exdS 16.c3 bS
17J"lxdS dxeS 18 .fxeS .ie6t.
12 J.d7 13.e5
• .•
Vallejo should have kept the ten­
sion with 13.@b1 �ac8 14.lt:Jb3. [14.
li:lf3 lt:Jb4 1S.lt:Jd2 (1S.�d2 �aS) 1S . . .
ia4 16.�cl �aS 17.a3 �xc3 18 .bxc3
li:lc6�]
White is threatening e4-eS, e.g.
14 ... bS 1S.eS. It is tempting to stop it
with 14 . . . eS, but this pawn structure
is dangerous with opposite castles.
White answers 1S. �f2 bS 16.fS b4
17.lt:ldS lt:lxdS 18.exdS lt:JaS 19.lL'lxaS
�xaS 2 0 .g4 and he is way ahead
in the race, e.g. 20 .. .f6 2 1 .h4 .ibS
22 .gS �c7 23.�d2 .ixe2 24.�xe2
�d7 2S.g6.
It is preferable to avoid direct
clashes in dynamic positions when
you are lagging behind in develop­
ment. I would vacate the d7-square
with :
14 . . . .ie8 ! ? . Then White will need
to redeploy his pieces which will
give Black time to launch his own
attack: 1S . .if3 bS 16.a3 (16.�f2 b4
17.lt:la4 lt:Jd4 18 . .ixd4 .ixa4� 19.eS
lt:J d7) 16 . . . �b8 17.�f2 aS with mutu­
al chances.
13 dxe5
••• 14.fxe5 YlYxe5
15 .if4 ti'c5 16. �b3
•
It was better to play for a repe­
tition with 16 . .ie3 �eS (16 . . . lt:Jxd4
17 ..ixd4 �c6oo) . Now White has not
full compensation for the pawn.
16 YlYa7 17 . .ie3 b6 18.�f2
.••
�ab8 19.g4 .ie8 (19 . . . lt:Jb4 ! ?)
20 .g5 �d7 2 1.h4 f5 2 2 . c.t>bl f4
23 .icl e5 24.h5 �c5 25.h6
•
.ig6 26.hxg7+ @xg7 27 .if3 •
�b4 28. toal hg5 29.�gl h6
30.a3 a5 31.�el gfe8 3 2 . �b5
�e7 33.axb4 axb4 34. �xb4 e4
35 .ie2 gas 36.gd6 �e5 37 .id2
• •
ga4 38 .YlYc3 YlYxc3 39.�xc3 e3
40.�xa4 exd2 41.�c3 Draw.
6. Volokiti n-Najer
M uelheim 23. 1 1 .20 1 3
l.e4 c5 2.�f3 e6 3.d4 cxd4
4.�xd4 �c6 5.�c3 �c7 6 .ie3 •
a6 7.YlYd2 �f6 8 . 0-0-0 .ie7 9 .f3
b5 10.g4 �xd4 ll.ti'xd4 .ib7
69
 Part 2
12.g5 �h5 13.�bl �k8 14.YlYd2
b4 15.�e2
Or 1S.lLla4 dS 16.exdS hdS
15 . . . d5 16.e5
16 . . . d4 !
Life is brighter for Black with an
active light-squared bishop !
16 . . . �xeS? 17.lLlg3 lLlxg3 18.hxg3
�c7 19J''! h 2 opens lines to White's
long-range pieces.
The idea for the counter-sacri­
fice was not new to Najer. For in­
stance, in September, Lopez Mar­
tinez introduced it against Bauer in
a slightly different setting: 13.lLle2
�c8 14.�d2 dS lS.eS
1S . . . d4 ! . White's king il) still on
cl here. This gives Black a tempo in
some lines where he attacks the a2-
70
pawn with . . ..tdS or . . .�c4. 16.lLlxd4
0-0 17 . .id3 �fd8
It looks tempting to leave the
rook on f8 and open the f-file with
17 . . . .tdS ! ? 18.@bl g6 19.�hfl �xeS
2 0 .f4 �c7 2 1.lLlb3 f6 ! ? . However,
this decision is risky from a prac­
tical standpoint as it weakens the
castling position. White can sac a
piece, for instance: 2 2 .gxf6 �xf6
23.fS exfS 24 . .txfS 111xf5 2S.�xf5
hb3 26.axb3 gxfS 27.�dS+ @h8
28.�xfS lLlg7 29 . .td4= whereas
Black is unable to untie his pieces.
The source game Bauer-Lopez
Martinez went 18.@bl g6 19.�hfl
�xeS 2 0 .f4 �c7
2 1.�f2 .tf8 2 2 . lLlb3 �dS 23 . .tb6
�d6 = .
Kurmann-Berkes, Bad Wiessee
2 013, did hot improve anything:
17 . .tg2 �fd8 18.f4 hg2 19 .�xg2
�c4 ! 2 0 .@bl ltJxf4 2 1.hf4 �xd4
2 2 .�xd4 �xd4+. Black's combina­
tion became possible due to the hit
on a2 so we must recheck White's
idea with the insertion of @bl b4.
17.�xd4 0-0 18 . .te2
Analysis confirms that 18 . .tg2
.tcS 2 0 .�hgl �fd8 2 0 .�f2 does not
 The English Attack
leave White a pawn up (although
even then the game is even, since
the �+l!J is at least a match to a
queen+bad bishop), due to 2 0 . . .
�b6 2 l .f4 .bg2 2 2 J'!xg2 :gds 23.f5
l'kd8 24.fxe6 fxe6 25.:gg4 :gxeS.
18 gfd8 19.§'cl g6 2 0.ghel
••.
�xeS 21.f4 §'c7
22.f5 ! ?
Black came out o f the open­
ing with the better pawn structure
so White must attempt to alter it.
22 .-bhS gxhS would leave the b7-
bishop without an opponent on the
light squares so Volokitin's deci­
sion is understandable. His other
option was really grim: 2 2 .ltlb3 l!Jg7
23.E1xd8+ :gxd8 24.l!Jd4 .icS .
22 ••• e5 23.fxg6
23.f6 .if8 24.ltlb3 l!Jf4 2S . .ig4
l"la8 26 ..bf4 exf4 27.:gfl :gxdl
28.�xdl :gds 2 9.§'cl .id6 gives
Black a passed pawn up and the
bishop pair.
23 hxg6
••• 24.i.xh5 gxh5
25.�f5 i.e4
2S . . . :gxdl 26.:gxdl .ie4 27.l!Jxe7
�xe7 2 8 . :gd2 aS is probably slight­
ly more accurate. In the game Black
also achieves an opposite-coloured
bishops attack. White's defence is
quite difficult.
26.�xe7+ §'xe7 27.gxd8 +
§'xd8 2 8.ge2 i.g6
White failed to hold this ex­
tremely unpleasant position.
29 .b3 a5 30 .i.f2 '%Yd5 31.i.g3
a4 (31.. .h4 ! 32 .hh4 a4- +) 32.gd2
fff3 33.bxa4 b3 34.axb3 ffxb3 +
35.§'b2 ffxa4 36.he5 �h7
37.�cl §'e4 38.i.f6 §'el + 39.gdl
§'e3 + 40.gd2 §'gl + 41.gdl §'xh2
42.gd2 §'gl + 43.gdl §'e3 +
44.gd2 gxc2 + 45.Wxc2 hc2
46.�xc2 h4 47.gd3 ffe4 48. �d2
tfg2 + 49.�e3 �g6 50. �f4 '%Yf2 +
51.gf3 Wh2 + 5 2 .�e4 h3 53.i.e5
We2 + 54.ge3 Wg2 + 55.gf3 h2
56.i.xh2 We2 + 57.ge3 Wxh2 0-1
7 . Robson -Wang ,Yue
Rapid N i ngbo ( 1 . 1 ), 25.07.201 3
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4.�xd4 �c6 5.�c3 ffc7 6.i.e3
71
 Part 2
a6 7.YlYd2 tilf6 8 .0-0-0 j.e7 9.£3
b5 10.g4 tilxd4 ll.'exd4 j.b7
12 .g5 tilh5 13.j.h3 :!! c 8 14.YlYd2
b4 15.tila4
15 • .• i.c6 ! ?
This game is not too important
for the opening theory since Black
has yet another good option in his
possession - 15 . . . d5 . However, it
sets a milestone in the understand­
ing of the Taimanov pawn struc­
tures.
16.tilb6 :!!b 8 17.j.g4 :!!xb6 ! !
18.hb6 'exb6 19 .bh5 'ea5 • pawns.
20.'i!lbl 'exg5 2 1.'exg5 hg5
22 .j.g4 j.f4
It turns out that White has to
struggle for the draw despite be­
ing an exchange up. His kingside
pawns are vulnerable and the rooks
72
are passive. The subsequent play is
so logical and straightforward that
it does not need comments.
23 .ih3 'i!le7 24 .ifl a5 25 .id3
• • •
h5 26.h3 h4 27.:!!h gl g5 2 8 .:!! d el
:!!b 8 29.:!!g4 f6
White is looking for a way to re­
turn the exchange. 3 0 J �ggl is also
murky due to the possible raid of
Black's king to f4 (after . . . .ig3) fol­
lowd up by . . . dS. White will be un­
able to defend both the h3- and f3-
30 j.e3 31.:!! e l j.f2 3 2 .:!!fl
• .•
j.g3 33. 'i!lcl 'i!ld6 34. 'i!ld2 'i!le5
35.:!!g l d5 36. 'i!le2 d4 37 .:!! g 2 :!!h 8
38.:!!g l :!!h 5 39.:!!g 2 f5-+ 40.exf5
exf5 41.:!!4xg3 hxg3 42.:!!xg3
'i!lf4 43.'i!lf2 j.d5 44.b3 :!! h 8
45.j.e2 :!! e 8 46.:!!g l :!! c 8 47.j.d3
h£3 48.j.xt"S 'i!lxf5 49. 'i!lxf3
:!!x c2 50.:!!g4 :!! c3 + 51.'i!lf2
:!!xh3 5 2 .:!!xd4 gh2 + 53.'i!lf3
'i!le5 54.:!! e 4+ 'i!ld5 55.:!! g4 :!!h5
56.:!!g 2 'i!ld4 57. 'i!lg4 :!!h 8 58.:!! c 2
:!! e 8 59. 'i!lxg5 :!! e 5+ 60.'i!lf4
'i!ld3 61.:!!h 2 :!! e 2 6 2 .:!!h 3+ 'i!ld4
63.:!!h 8 :!!f2 + 0-1
 The English Attack
8. Demetrio-Schiendorfer
ICCF email, 1 0.09.2009
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4.�xd4 �c6 5 . � c3 'lfc7 6.J.e3
a6 7.'lfd2 �f6 8 . 0-0-0 J.e7 9.f3
b5 10.g4 �xd4 l l.'lfxd4 .ib7
12.g5 �h5 13 .ih3 �k8 14.J.g4
•
b4 15 .ixh5 bxc3 16.Wxg7 cxb 2 +
•
17.<.!.>xb2 �U'8
This forced line is critical for
Black's set-up. At first sight it looks
rather messy, but in fact the game is
completely balanced and it is even
White who should find a couple of
accurate moves. Black has only one
tangible target - the t7-pawn and
his pieces are more compact. This is
clearly seen in the line 18.Wd4 J.cS+
(McDonald-Brunello, London 2013
saw 18 ... d5 when 19. <;t> al! would
have been equal) 19.Wfxc5 'l!tlxcS
20 .hc5 :!'!xeS 2 l .h4 h6+, Rasik-Mi­
ton, Slovakia 2 0 14.
18.c3 seemingly defends the c­
pawn, but 18 . . . J.d6 (18 . . . d5 19 .Wfd4
id6 2 0 .'1!tff6 J.e7= is a repetition of
moves) would force 19.:t'!xd6 since
19.J.d4? eS cuts the g7-queen off
from its king.
lS.<;t>al d5? !
18 . . . Wfxc2 ? ! would give White
tempi for :t'!bl and :!'!bel, but 18 . . .
J.b4 ! 19.J.d4 J.c3+ 2 0 .hc3 Wfxc3+
2 1.Wfxc3 :t'!xc3 is a decent alterna­
tive.
OTB, I would choose this end­
game because I would have a lasting
initiative without any risk. White's
extra pawn would be irrelevant as
Black's rooks are very active. He
can play later .. .fS or . . . h6 gxh6 :t'!h8 .
In a correspondence game, how­
ever, Black can boldly aim for more
tangled play. His idea to take on e4
is double edged as it opens the f­
file and weakens the e8-king. This
might tell if White succeeded in
shifting his queen to the queenside.
For instance, he has the interesting
counter-strike 19 .c4 Wfxc4 (19 . . . dxe4
2 0 .fxe4 he4 21.:t'!hfl J.g6 2 2 .hg6
fxg6= ) 2 0 . :t'!cl Wfd3 2 l.:t'!xc8+ hc8
2 2 .J.f4 dxe4 23 .:t'!cl J.b7 24.Wfe5 with
an initiative, although 24 . . . Wfd5 de­
fends everything. Demetrio opts for
the most solid continuation :
19 .id4!
•
73
 Part 2
19.c4 dxe4 2 0 .fxe4 i.xe4 21.Elhfl
i.g6 22 .i.xg6 fxg6 is equal.
19 dxe4
..• 20 .fxe4 he4
2U::!hfl i.g6
22 .Ae2
This move obstructs the second
rank. 22 .i.f3 ! is more unpleasant.
The reason behind it is the manoeu­
vre 22 . . .�a5 23.c3 �xgS 24.Elf2 ,
aiming for the b-file. White's activ­
ity would be very unpleasant. In­
stead, he assumes a passive stand
which costs him the initiative.
22 'ti'a5 23 .ib2 l::! c7 24.h4
. .• • 16.e5 would be well met by 16 ...
hc2 25 .ih5 ! .ixdl 26.hf7+
• lt'le8 ! .
®d8 27.l::!xdl+ l::! d7 28.l::! c l Wd2
29.Yfe5 J.d6 30.Wf6 + l::! e 7 3 1.Wf3
'ti'xcl+ draw.
9. Caruana-Svidler
Rhodes 26. 1 0.201 3
l.e4 c5 2 . li:Jf3 e6 3.d4 cxd4
4.li:Jxd4 li:Jc6 5.li:Jc3 'ti'c7 6.J.e3 a6
7.Vd2 li:Jf6 8 .0-0-0 Ae7 9.f3 b5
10.g4 li:Jxd4 ll . .ixd4 .ib7 1 2 . ®bl
0-0 13.Vf2
74
Caruana's last move was obvi­
ously aimed against Black's the­
matic idea 13 . . . b4 14.lt'la4 dS when
15.lt'lb6 ! Elad8 (the exchange ·sacri­
fice 15 . . . dxe4 is in White's favour)
16.e5 lt'ld7 17.lt'lxd7 would face Black
with unpleasant position with­
out counterplay. However, 13 . . . d5 !
right away is possible - 14.g5 lt'ld7
15.exd5 b4.
14.l::!h gl J.c6 15.J.d3 b4
16.ti�e2
The knight i s very important for
the defence as I already mentioned
in the previous note. 16 . . . lt'ld5
17.lt'lxd5 i.xdS 18.f4 d6 19.g5 dxeS
2 0 .i.xe5 �c6 2 1.f5 offers White the
initiative, although Black's position
is quite resilient. After the knight's
retreat, the manoeuvre lt'le2-g3 is
slow. Black will trade the active d3-
bishop with 17 . . .i.b5 and then . . . d6
will neutralise the opponent's spa­
tial advantage. Obviously White
should seek a quick conclusion:
17.lt'le4 f6 18.exf6 lt'lxf6 19.i.xf6
 The English Attack
ixf6 2 0.g5 .ie5 2 1.li'lf6+ , but 2 1 . . .
l"lxf6 ! 2 2 .gxf6 i.x£6 offers u s anoth­
er example where Black's bishop +
pawn hold their own against a rook.
Black will always be threaten­
ing to build up a �+.i battery on
the main diagonal. Moreover, the
f3 and h2-pawns are weak.
16 ••• d5 17.g5
The pawn structure after 17.e5
is good for White only if his knight
is on b6. Otherwise the d7-knight
has good prospects from c5 : 17 . . .
lt:ld7 18.f4 .ia4 19 J'!d2 ltlc5 2 0 .li'lc1
a5 setting up the trap 2 1.f5? .ig5
22.l:%e2 li'lxd3 23.li'lxd3 .ih4 ! with an
initiative, for example, 24.�g2 �c4
25 ..ic5 l:%xc5 2 6 .b3 �c3 27.ltlxc5
.ib5 28.li'ld3 a4 ! .
17 ••• ltlh5 18.exd5 .ixd5
This position illustrates Black's
strategic aim in the English Attack.
If White has not anything forced,
he is even slightly worse. Black has
an open file against the enemy king
and an extra pawn in the centre. the
pawn storm . . . a5-a4, . . . b4-b3 is not
easy to stop. An important factor in
Black's favour is that queens trade
would underline the weakness of
White's kingside pawns. For in­
stance, 19.f4 .ic5 2 0 .f5 g6 2 1 .hc5
�xeS. Caruana attempts to take f4
under control.
19.gg4? ! e5! 20 .ie3 'ec6
•
20 . . . g6 ! was more cunning.
White has not an active continua­
tion as 21. li'lg3 could be met by 2 1 . . .
f5 o r 2 1 . . .li'lf4. The text does not ac­
tually threaten f3 and it wastes time.
White could exploit it with 2 1.li'lg3
when 2 1 . . .hf3? fails to 2 2 .li'lf5 l:%ce8
23.l:%c4.
22 f5
.••
A good move, but Black may
have technical problems to con­
vert his advantage due to his weak-
75
 Part 2
ened king. He had more solid op­
tions : 2 2 . . . .id6 ! 23 . .ie4 l!Jf4 24.hf4
he4 25.fxe4 exf4. Then 2 6 .e5 heS
27.l!Jd3 .id6 28.l!Jxf4 does not help
owing to 28 . . .f5 ! 29.Eih4 .ie7 30.l!Jd3
aS with domination.
23.gxf6 t0xf6 24,gggl t0h5
25.tM2 hf3 26.t0b3
Black has won a pawn, but White
set up co-ordination between his
pieces. Eventually, Svidler failed to
bring the point home.
26 .. ,gfd8 27.ttf2 gf8 28.i.h6
gf7 29.tte3 e4 30.t0d4 ttd5
76
31. t0xf3 exd3 32.ttxd3 Vxd3
33.cxd3 i.f8 34.i.d2 i.d6 35.t0g5
gxfl + 36.gxfl i.xh2 37.hb4
i.e5 38.t0e6 t0g7 39,ge1 i.g3
40.l:� e4 ges 41. t0c5 h5 42 .i.c3
gxe4 43.t0xe4 i.f4 44.�c2 t0f5
45.�dl �t7 46.i.d2 i.e5 47.b4
�e6 48.a4 t0g3 49. t0xg3 hg3
50.b5 axb5 51.axb5 �d5 52 .i.e3
i.d6 53. �e2 .ic5 54.i.g5 i.b6
55.�f3 �c5 56.�e4 i.c7 57.d4+
�xb5 58.�d5 i.g3 59.i.e3 �b4
60.�e6 h4 61.d5 �c4 6 2 .d6
hd6 63. �xd6 �d3 64 ..igl �e4
65.�e6 g5 66.�f6 �f4 67.i.h 2 +
�g4 68. �g6 �h3 69. �xg5 �xh2
70 .�xh4 draw.
 Pa rt 3

Th e Th i rd Ra n k Set-U p
I consider here set-ups with tt:Jc3, .id3, 0-0. White often follows up with
ie3, f4, Wlf3, but I also analyse �e2+id2 in the �n. Plans with f4 and long
castling are covered in Part 4.

The big flaw of White's set-up is that id3 cuts the queen's support for
the d4-knight and leaves the g4-square without control. The Taimanov way
to exploit that is a direct attack with . . . tt:Jb8-c6-e5-g4 while the bishop goes
to cS. The Kan fans prefer the flexible development . . . tt:Jb8-d7-c5 planning
to swap the d3-bishop.

77
Pa rt 3 . The Third Ra n k Set- U p
M a i n I d eas
l.e4 c5 2.tbf3 e6 3.d4 cxd4
4.tbxd4 ltlc6 5.ltlc3 Y!!c7 6 . .ie3 a6
7 .id3 ltlf6 (7 . . . b5 is dubious due to
•
8.ltlxc6) 8 . 0-0
White is playing "by the book" .
He leads out the knights first, then
the bishops. That ensures him a
lead in development which could
easily grow into a crushing kingside
attack. For instance, the Sche­
veningen set-ups with 8 . . . d6 are
very risky, so Black should better
look for concrete play on the cen­
tral dark squares.
A plus for the Taimanov: in order to
develop the bishop on d3 , White has
to play first .ie3 which rules out set­
ups with .id2 .
78
You may have noted that I did
not even mention possible long
castling. It is rarely seen since Black
·
obtains a good game after s :�e2
.id6 ! ? 9. 0-0-0 .ie5 10.ltJxc6 bxc6.
A plus for the Taimanov: Black
avoids plans with Y!!e2 , 0-0-0.
In the diagram position, Black
can choose 8 . . . ttJxd4 9.hd4 .ic5,
followed up by . . . d6, . . . e5, . . . .ie6 .
This i s a solid plan where White en­
joys a temporary initiative. Howev­
er, I think that Black should exploit
the biggest flaw of the opponent's
set-up - the g4-square :
8 • .• ltle5 ! 9.h3
9.ltJf3 ltJfg4 10.ltJxe5 ltJxe3
1UWh5 ! is analysed up to a draw
endgame. I'd like to recommend
9 . . . ttJeg4 ! ? 10 . .id2 d6 with a solid
position and mutual chances.
9 .ic5
• .•
At this point, White must decide
the further pace of the game.
 The Third Rank Set-Up
Taimanov 1
The oldest approach is to push
e4-e5 : 10 .'1'9e2 d6 ll.f4 ll'lg6 1 2 .ll'lb3
ixe3 13.'1'9xe3 0-0 14J'�ael b5
15.e5 dxe5 16.hg6 hxg6 17.fxe5
ttJd7 18.ll'ld4 i.b7, but practice has
proved that Black is fine. White's
hopes for reviving this line have
been connected lately with the ex­
tremely sharp line :
lO.'it>hl d6 ll.f4 �g6 12 .�el
0-0!? 13.f5 �e5 14.�h4 b5
15.l'U'3 exf5 16.exf5 �xf3
17.gxf3 i.b7 18.ggl i.xf3 19.�xf3
i.e3 20.�f6 hgl 2 1.�g5
Practical experience has seen
Black gradually improving his de­
fence in this position. The second
players also learned how to coun­
ter-attack.
Taimanov 2
The current state of theory is in
Black's favour.
In the Kan, Black refrains from
. . . ll'lc6 in favour of . . . ll'lbd7:
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4. �xd4 a6 5.�c3 �c7 6.i.d3
Black has sidestepped the alleg­
edly dangerous English Attack so
he might be tempted to enter the
Taimanov now. However, 6 . . . ll'lc6
can be met by 7.ll'lxc6 \Wxc6 ! ? 8 . 0 - 0
which looks balanced, but i t i s a leap
in the unknown.
6 �f6 7.0-0 (7.\We2 i.d6 ! ?) 7 . . .
.••
d6
Black's next moves are 8 . . . ll'lbd7,
9 . . . i.e7 (only to 8.a4 he plays 8 . . . b6
79
 Part 3

first) waiting for White to define his Kan 1

plan. Delay . . . bS in order to rule out
queenside activity with a2-a4, c2-
c3.

A plus for the Kan : Black's set-up is

very flexible. It offers the opponent
a wide choice and an ample ground
for positional mistakes.

White has 4 main set-ups:

See g;,�me 18 Kerekes-A. Ko­
vacs, Hungary 2 007.
1. 8.f4 �bd7 9 .Wf3 J.e7 10.J.e3

�
'

2. 8.f �bd7 9.�f3 b5! lO.cj;>hl

(lO .�el i.p7 ll.i.d2 .te7) 10
J.b7 •..

ll.Wel · ·

Kan 2

The third rank set-up is ineffi­

cient against . . . lLlbd7. It is only dan­
gerous if Black submits himself to
an attack with 10 . . . 0-0? ! l l.g4. In­
stead, we should fianchetto the
A good option here is ll . . . b4
bishop to b7 and restrict the oppo­
12 .�dl d� or 12 . . . lLlc5 13.lLlf2 dS.
nent's pawn advance by . . . g6, aim­
ing for this position:
3. s . tl\'e2 �bd7 9 . cj;>hl i.e7
10.i.d2 ! ?; 0-0 ll.f4 �c5 12 .gael
Ka n 3

or for ... h7-h5-h4 as in the fol­

lowing example :
80
 The Third Rank Set-Up

Objectively best here is 12 . . . b5 Kan 4

which, however, assumes a draw in
the variation 13.e5 ll'lfd7 14.i.xh7+
lfixh7 15.�h5+ @g8 16J'!f3 dxe5
17J!h3 f5 18.�h7+ .
12 . . . ll'lxd3 13.cxd3 b5 is passive,
but at least Black will be playing for
all three results.

4. 8.a4 b6 9 .f4 .ib7 10.�f3

�bd7 11 .id2
• ll . . . i.e7 (ll . . . g6 ! ? 12.�h3 .ig7)
12 .gael 0-0

A plus for the Kan : There is little Black is waiting for White to play
theory and the play is not forced. 13.�h3 before putting in 13 . . . ll'lc5.
You do not have to constantly watch See game 16 Cubas-Hellsten, Bue­
for new developments. nos Aires 2006.

Weapon of Choice

If you feel confident in your home preparation and regularly check modern
practice for new developments in position Taimanov 2, then the Taimanov
should be your weapon of choice. It leads to very sharp play with a lot of cal­
culations and offers Black considerable winning chances. If you started with
the Kan move order, you should try to steer the game into the Taimanov
with 6 . . . ll'lc6. The only sensible White deviation would be 7.ll'lxc6 �xc6 ! ? .

I f you are much stronger than your opponent, you may want t o choose
the Kan. It allows many different move orders and the cost of every move in
the opening is much lower than in the Taimanov. You could always outplay
your opponent in the middlegame. It would be enough to kill his d3-bish­
op with . . . ll'ld7-c5xd3 in order to secure your king, and patiently wait for an
opportunity to open the main diagonal for you b7-bishop.

81
Part 3. The Third Rank Set-Up
Taimanov - Step by Step
l.e4 c5 2.tlJf3 e6 3.d4 cxd4
4.lt)xd4 tlJc6 5.tlJc3 �c7 6 ..ie3
a6 7..id3
7...tlJf6
The other popular move is 7...
bS. It scores well and Emms recom­
mends it in his repertoire book, but
I definitely do not like the position
after 8.tt'lxc6! Wxc6 9.0-0 .ib710.a3
tt'lf6 ll.El:el!
82
Black lacks a sensible plan. The
common ll....ie712.Wf3 d613.El:adl
Wc7 (13...0-0 14.Wh3) 14.Wh3! 0-0
lS..igS h6 (15...Effd8 16.El:e3 Wcs
17.tt'ld5---t) 16.hh6! gxh6 17.e5! is
·
gloom.
Emms and Ribli advocate 11...
.id6 12..id4 eS 13..ie3 .icS. Then
14.Wf3 he3 15.El:xe3 �cS 16.Wf5!
�f8 (16...0-0 17.tt'ld5) 17.El:dl We7
18.El:f3 �g8 19.h4 h6 20.tt'ld5 hdS
2l.exd5 d6 22.c4 is by no means
fun for Black, but 14.a4! b4 lS.tt'ldS
tt'lxd516.exd5 Wxd517.Wg4 is a real
disaster: 17...0-0-0 18.hc5 WxcS
19..ie4---t ( or 19.c3).
8.0-0
Castling is by far the most fashio­
nable continuation nowadays, but
White has also tried:
a) 8.tt'lb3. This retreat is aimed
against 8.0-0 tt'lxd4 9.hd4 .icS
which isquite safe for Black. White's
idea is to organise a Scheveningen­
style attack with 0-0, f4, Wf3. Still,
wasting a tempo in a sharp posi­
tion ( even two tempi, since White
needs his knight for the attack and
often brings it back to d4 en route
to f3), reduces his chances of get-
 The Third Rank Set-Up

ting an edge. bxc6 (9 . . . dxc6 10.f4 e5 ll.f5 b5

Black should develop by 8 . . . . b5 1 2 . a4 ! ) 10.f4 e5 11.f5 ! .ie7 12 .�f3
9.f4 d6 10 .�f3 .ib7 and trade his c6- .ib7 (12 . . . 0-0 13 .g4 h6 14.h4 tlJh7
knight for the d3-bishop. 15.g5 ! +-) 13.�g3 0-0 14 . .ih6.

b) 8.�e2 , intending to castle

long, is seldom seen lately. Black's
A. 8 . . . ttJxd4 ! ? 9 . .ixd4 .ic5
most testing answer is 8 . . . .id6 ! ?
9.0-0-0 (9 .g3 .ie5 10 .tlJb3 d5 ! ) 9 . . .
ieS, when after 1 0 .tlJxc6 both cap­
tures are fine for him :
10 . . . dxc6 ll.tlJa4 tlJd7 12 .�d2
id6 (12 . . . .if6 ! ? - Miladinovic) 13.f4
eS 14 . .ic4 .ie7 15.�f2 ! = , Sax-Mila­
dinovic, Ano Liosia 1999 ;
1 0 . . . bxc6 ! ? ll. tlJa4 l'!b8 12 .b3
(12.c3 ! ?) 12 . . . d5 13 . .ic5 .id6 ! , Ro­
din-Hasangatin, Voronezh 2003.

Black's strategy is simple and

straightforward. He wants to com­
plete development and play in the
centre. Black's next steps are: . . . d6,
... e5, ... .ie6 and castling when pos­
sible.

Remember! We refrain from

Black has a number of alter­
natives here. Note that the Sche­
veningen set-up with 8 . . . d6 is very
risky against 7 . .id3 . White often ob­
tains a crushing attack with simple
developing moves like f4, �f3-h3 ,
�ael.
A. 8 . . . ttJxd4; B. 8 . . . ttJe5
8 ... .id6? ! is an inferior alterna­
tive because White quickly launch­
es a kingside attack - 9.ttJxc6 !
. . . b5 in the opening. Thus we
gain time for development and de­
prive the opponent of a target on
the queenside.
White has three main plans from
this point:
1 . He doubles or triples on the d­
file. Then we move our king to e7.
2. White splits our kingside
pawns by: 1 0 .ixf6 gf6 11.�g4 @f8 .
W e can leave our pawn o n d 7 in or­
der to have . . . .id6 or defend it by
. . . .ic6. There is nothing to worry
about as we gradually connect our

83
 Part 3

rooks and display activity on the 16 . . . 1'!g5 17.g3 aS ! ] 14 . . . 'ifle7 lS.i.hS

kingside. l'!af8 16.'iflhl �b6 .

3. White plays f4. We answer it ll • .• .ixd4 1 2 . t!fxd4 e5 13. t!fb4

by . . . eS, . . . i.e6, exf4 before castling. .ie6

Al. 10.i.e2 ! ? ; A2 . 10.i.xf6 ; A3 .

lO.i.xcS

Al. 10 .ie2 ! ?
•

This position occurred in the

very first game played in this varia­
tion. At the Black side was the God­
father Taimanov. White hopes to
use the weakness of d6.

10 d6
White might need his rook on
•••

the queenside so 14.1'!fdl is a logical

Many strong players prefer 10 . . .
alternative. Black can answer with
eS, but I have strong doubts about
either 14 . . . 'ifle7 ! ? when 15.f4 is not
the soundness of this choice. Af­
dangerous in view of lS . . . aS 16.�b5
ter ll.i.xcS �xc5 12 .�d3 bS (12 . . . d6
(16.�a3 �b6 + 17.'iflfl �b4 18 .fxe5
13J%adl 'ifle7 14.�g3±), the novelty
dxeS) 16 . . . 1'!hc8, or 14 . . . �c5 15.a3
13J�tbl! assures White of a lasting
�xb4 16.axb4 'ifle7 17.1'!a5 !!hc8
edge. For example: 13 . . . i.b7 14.b4
18.f3 i.c4 = , Dembo-V.Schneider,
�c6 (14 . . .�c7 15.a4 l'!c8 16.axb5±)
Plovdiv 2 0 0 8 .
15.a4 bxa4 16.1'!dl 0-0 17.tt:\d5 tt:\xdS
18.exd5. 14 ••• t!fc5 (14 . . . 'ifle7) 15.a3

ll.'ffd3 15.�xb7 0-0 16.a3 l'!tb8 17.�e7

!!dB= (threatening l'!d7- +) leads to
White scores poorly after a repetition.
ll.i.x£6 ! ? , but this is his most prin­
cipled continuation. The compari­ 15 • • • 0-0-0 ! ?
son with line A2 is in White's fa­
vour here because of the weak d6- Hector-Haba, Hamburg 2 0 04,
pawn. Black must defend very accu­ saw 15 ... �b4 16.ab4 'ifle7 17.1'!d2
rately: ll . . . gxf6 12 .�d2 i.d7 13.1'!adl l'!hc8 18.f3 i.c4 = . I decided to face
l'!g8 14.�h6 [14.�f4 'ifle7 15.1'!d3 the opponent with more complex
i.c6 16.a3 (16.i.h5 1'!af8 17.1'!fdl l'!g5) problems.

84
 The Third Rank Set-Up

16.gd2 'Ml4 17.ab4 d5 = . O r 12 .�h4 �e5 13.�h6 ritlg8

14.'iflh1 �g5 15.�h3 h5= .
Black has even some slight in-
itiative, M amedov-Delchev, Be- 12 • . • h5 13.�h4
nasque 2005.
13 .�h3 �f4 14.i.e2 b5 15J''l a d1
should be met by 15 . . . d6!oo. (15 . . .
A2 . 10 .ixf6 gxf6 l l.�g4!
• rifle7? lost a pawn after 16.Eld5 d6
17.Elh5, Balashov-Tregubov, Kato­
wice 1992.)

13 ••. .te7 14.f4 b 5

This is more precise than 11. riflh1

h5 and Black follows the scheme
... b5, . . ..tb7, . . . 'itle7, . . . Elag8, . . .f5 like
in the game Morozevic-Poluljahov,
Orel 199 2 . Black's game is easy. He wants
The text is a n attempt t o drag to complete development and play
Black's king to f8 instead of the bet­ . . . ritlg7, . . . Elag8, . . . riflf8 . White can­
ter place e7. not wait.

ll . . . ritlf8 15.f5 !

This is not obligatory at all. In Tzeitlin-Soffer, Tel Aviv 1994,

Black has perfect compensation
for the pawn after ll .. .'�e5 12 .'�g7
(perhaps 1 2 . 'iflh1 is more solid, 12 . . .
�g5 13 .'�e2 h5 14.f4 �g4oo) 12 . . . Elf8
13.riflh1 �g5 14.�h7 Elg8 15.�h3 b5
16.f4 �g7 17.Elf3 Elh8 18.Elg3 �f8
19.�g4 i.b7. If White plays h3, riflh2 ,
Black can shift his dark-squared
bishop to the h2-b8 diagonal.
12.riflhl
White delayed this thrust - 15.Elae1
and Black achieved the redeploy­
ment following 15 . . . i.b7 16.f5 �e5 !
17.a4 i.c6 !
The pawn structure in the cen­
tre is identical with The Poisoned
pawn variation in the Najdorf. Black
should keep control of c4 to avoid
the manoeuvre lt:ld1-e3-c4. In Fe­
dorov-Miezis, Istanbul 2 0 0 0 , White

85
 Part 3
chose 17.ll)dl at once, but could not
even equalise after 17 . . . �g7 18J''!f3
:i'!ag8 19.c3 �f8 2 0 .tt:le3 :!'!g5 2 l .:i'!e2
:!'!hg8 22 ..ic2 .ic5 23.tt:lfl d5.
18.a5 :!'!e8 19.tt:'ldl .id8 2 0.c3 �g7
2l.b4 :!'!eg8+.
15 Yfe5 16.gf3
•.•
16Jl:abl .ib7 17.tt:le2 .id6 18.tt:lg3
�e7.
White seems unable to improve
his position. We can continue with
. . . .id6, . . . �e7, or . . . �g7-h8 .
A3 . 10 .hc5 fixeS
86
ll.�hl
The plan with f4 is currently ac­
knowledged to be the only real dan­
ger for Black. Other options are:
a) ll.a4 White prevents b7-b5,
but we were not too eager to play
it anyway! ll . . . d6 12 .a5 0-0 13.�d2
.td7 14J'i:fel :i'!ac8 15.tt:la4 i.xa4
16.:!'!xa4 tt:lg4 17.c3 :i'!fd8 18 . .ifl
tt:le5= Mitkov-Zapata, Mexico City
2 0 07.
b) ll.tt:'la4 �c7 12 .c4 d6 13.:i'!cl
.id7
Now the typical hedgehog for­
mation after 14.tt:lc3 0-0= is fine for
Black since two minor pieces have
been exchanged and he has plen­
ty of room for manoeuvring. White
can open the centre with :
14.c5 0-0 15.cxd6 �xd6 16.tt:lc5
.ic6 17.�e2 :i'!ac8 ! = . See game 12
Zaragatski-Delchev, Bad Wiessee
2005.
c) l l.�d2 d6
You should remember to refrain
from . . . b7-b5 before castling safe­
ly. Although the variation ll . . . b5
12 .:!'!el tt:lg4 13 .e5 tt:lxe5 14.tt:le4 �c7
 The Third Rank Set-Up
15.�b4 d5 16.lLld6+ �d7 17.�xe5
�xd6 18.�d4 i.b7 19.a4 bxa4
20J'�e3 is not clear, in a practical
game White's game is much easier.
12 .�ad1
Solak's rook lift 1 2 .�ae1 (having
in mind �e3-f3xf6) sets concrete
problems. For instance, 12 . . . 0-0
13.Ele3 e5 14.�f3 �h8 15.�xf6 gxf6
16.�h6 �g8 17.lLld5, Solak-Safar­
li, Baku 2 0 13, gives White an ini­
tiative: 17 . . . �g7 (17 . . . �g6 18.�f8 +
Elg8 19.�xt7) 18.lLlxf6 i.e6 19.�e1
�b4 2 0 .�e3 �d2 2 1 .h3, when Black
is deprived of any counterplay. I
think that it would be wiser to re­
frain from castling: 12 . . . e5 ! 13.�e3
ie6 14.�f3 h 6 ! ? 15.i.e2 �e7 16.�g3
g5 17.�d3 �hd8. White stays pretty,
but I do not see any threat to Black
in near future.
12.�fe1 is inconsistent. Besides
12 . . . i.d7, Black gets time to castle:
12 ... 0-0 13.�ad1 e5 14.�e3 (14.i.fl
Eld8 =) 14 . . . i.e6 15.i.e2 �fd8 16.�d3
lLle8 17.lLld5 (17.�g3 �h8) 17 . . . b5 ! =
(preventing any ideas with c2-c4).
White can switch over to plans with
f4, for example: 1 2 .�h1 e5 ! 13.�ae1
0-0 14.f4 �h8 ! ? anticipating 15.lLld5
lLld5 16.ed5 �d5= as in Kotronias­
Pelletier, Athens 2 0 05.
12 . . . i.d7
White would get a slight pull in
the endgame in the event of 1 2 . . . 0-
0 13.i.e2 �d8 14.�d4 �xd4 15.�xd4
e5 16.�b4
13.i.e2 �e7 Black is just in time
to regroup successfully:
14.�fe1 �hd8 15.i.f3 i.c6 16.�e3
�d7= .
ll . . . d6!
ll ... b5 is often seen, but it is con­
nected with a wrong plan.
Black obtains a good game with
a central strategy.
1. His first aim is to prevent e4-
e5. It is best achieved by . . . e6-e5.
2. Then he should castle.
3 . He must evade some tactical
tricks, connected with a sacrifice on
f6. Moves like . . . �h8 or . . . �e5 serve
well this cause.
4. The next stage is to organ­
ise counterplay in the centre by ex­
changing on f4 and activate the
rooks - to e8 and c8 .
1 2 .f4
Berg advocated 1 2 .f3, but I do
not see any reason behind this
move. After 12 . . . 0-0 13 .�e1 eS, he
did not find anything better than
14.f4 (14.�h4 i.e6) 14 . . . exf4 15.�xf4
i.e6 16.�d2 �g5 = , Berg-Pelletier,
Crete 2 0 07.
12 .�e1 e5 13.lLla4 �c6 14.�b4
is a provocation to 14 . . . b5? ! when
87
 Part 3
15.ttlc3 followed up by a4 will be in
White's favour. Instead, 14 . . . .ie6 is
equal.
12 ... e5 ! 13.Yfel
13.'!9e2 .ig4 ! 14.'!9el .ie6 is simi­
lar to the main line.
13.ttld5 lLlxd5 14.exd5 deserves
attention.
Now safest is 14 . . . 0-0= 15.fxe5
(15.'!9d2 f5 16 . .ie2 a5 ! 17J'U3 .id7)
15 . . . dxe5, when 16.'!9h5? ! f5 17 . .ixf5
hf5 18 J''i xf5 1!9xd5 19.Elxe5 walks
into 19 . . . '!9a2+.
14 ... 1!9xd5 is risky because af-
ter 15.'!9h5 Black will have to castle
long: 15 . . . .ie6 ! ? 16.Elael 0-0-0, but
this line is also more challenging.
13 .ie6
• .•
Black is ready to take on f4 so
White should do something about
it. Waiting moves like 14.a3 do not
work: 14 . . . exf4 15.Elxf4 '!9e5 16.Elf3
ttlg4t.
It would be logical to close the
centre by 14.f5 .id7 15.'!9g3, (or
15.'!9h4 .ic6 16.Elf3 h6 17.Elg3 @f8
18.a3 b5 19.Elfl b4 2 0 . lLl h l Elb8t,
Kotronias-Filip, Eforie Nord 2 0 08)
88
but then Black's king would feel fine
in the middle: 15 . . . l:!g8 ! 16.a4 .ic6
17.Elad1 '!9b4+!.
14.gdl exf4 15.c!Lld5
White must cut Black's queen
from e5. He would be struggling af­
ter 15.Elxf4 �e5 16.'!9d2 0-0 17.Eldfl
Elac8 .
15 c!Llxd5 16.exd5 �xd5
.•.
17.gxf4 �g5 18.gd4 0-0 19.gxd6
gadS = .
B. 8 ttle5
• ••
Black's knight strives to reach
g4, for example, 9.f4 lLleg4 10 . .id2
.ic5 ll.lLlce2 e5 (ll . . . d6 12 .h3 e5 is
also possible) 12 .fxe5 d5 ! ! (only
move ! ) 13 ..if4 ttlxe5 14.exd5 lLlxd5= ,
SPIKE-H IARCS, IPCCC, Paderborn
2 007.
Our general plan is to carry on
a dark-squared strategy by . . . .ic5,
. . . d6, . . . ttlg6 and possibly . . . e5 . In
contrast to line A, Black keeps both
knights. That makes play more tan­
gled and double-edged.
Now our main line branches to:
Bl. 9.lLlf3 ; B2. 9. h3
Sometimes White tries to sur­
prise the opponent with 9.i.e 2 . I
propose you to follow the usual
plan:
9 ... .ic5 10.f4 d6 ll.@hl 0-0
12 .'!9el lLlg6 (12 ... lLleg4 ! ? has been
successfully tested in practice)
13.�g3
 The Third Rank Set-Up
Black has two interesting possi­
bilities:
a) 13 ... .ia3 14.e5 (14.lild1 .icS=)
14 . . . dxe5 15.bxa3 '1!9c3 16.fxe5 tt:ld7+;
b) 13 . . . .ib4 ! with complica­
tions that look to be in his favour:
14.f5 (or 14 . .id3 hc3 15.bxc3 eS
16.f5 exd4 17.cxd4 tt:le7 18 . .ih6 tt:lhS
19.'\WgS f6 2 0 . '\WxhS gxh6 2 1.'\Wxh6
§'xc3 2 2 .Eiad1 .id7 23.Eif3 '.th8+)
14 ... hc3 15.fxg6 hb2 16.gxh7
1!1h8 17.Eiab1 tt:lxe4 ! 18.'1!9h4 hd4
19.hd4 f5+.
Bl. 9 . tt:l£J IC!eg4 ! ?
The text i s a more complex ver­
sion of the natural set-up with :
a) 9 . . . d6 ! ? 1 0 .lilxe5 dxe5= . Even
here, the possession of the b6-
square does not give White a sub­
stantial advantage : ll.lila4 .id7
12 .c4
The position is rather closed so
Black's lag in development cannot
be punished by 12.lilb6 Eib8 13.a4
.icS 14.tt:lxd7 tt:lxd7 15.'1!9g4 '.tf8 ! .
The king can make a n artificial
castling after 16 . .id2 h6 17.a5 '.tg8
18.'.th1 lilf6 19 .'1!9h4 Eid8 = .
12 . . . .ic6 13 .'1!9c2 '1!9a5
The most straightforward way
to swap the dark-squared bishops.
In Zaragatski-Perunovic, Balaton­
lelle 2 0 0 1 , Black successfully real­
ised the idea .of obtaining a strong
knight against a bad bishop after
13 . . . tt:ld7 14.a3? ! ha4 15.'\Wxa4 .icS
16.hc5 '\WxcS 17.b4 '\Wc7 18.c5 0-0
19.Eifd1 tt:lb8 ! 20 .'\Wb3 tt:lc6 2 l . a4
tt:ld4, with an initiative. However,
White could prevent it by 14.Eifd1!
ha4 (14 . . . Eid8 15.tt:lc3 .icS 16.hc5
tt:lxcS 17. .if1) 15.'1!9xa4 .ie7 16.c5.
14.lilc3 .ic5 = , Siewert-Capuano,
ICCF 2 0 07.
b) 9 . . . tt:lfg4 has been the main
line. It's status is still OK, but Black
has no chances to win the end­
game after 10.lilxe5 tt:lxe3 1 1.'\WhS
g6 12.'1Wf3 1!9xe5 13.fxe3 f6 14.'\Wxf6
'1!9xf6 15.Eixf6 .ig7 16.Eif3 bS 17.Eiafl !
89
 Part 3
White's idea is to meet 17 . . .
j'!f8(?) with 18.j'!h3 .
17 . . . .ib7 18.j'!f7
If White retreats his knight to a
passive square, Black can already
swap one pair of rooks : 18.tt'ld1 j'!f8
19.j'!h3 j'!xf1+ 2 0 .'it>xf1 h6 2l.j'!g3
'it>f7 2 2 .j'!f3 'it>e7 23.'it>e2 .ie5t, Li
Shilong-Wen,Yang, Shandong zt
2007. The fine point is that White
is left without a plan, whereas we
can proceed with a typical Sicilian
queenside attack.
18 . . . .ie5 19.tt'le2
Or 19.tt'ld1 .ic6 2 0 .c3 aS,
Kotrotsos-Mastrovasilis, Kavala
2 0 07. Only Black can play for a win
in this position.
19 ... .ixb2 20.tt'ld4 0-0-0 2 l.c4
2l.j'!g7 j'!dg8 was good for Black
in Pilgaard-Bui Vinh, Budapest
2007.
21.. .bxc4 22 . .ixc4 .ixd4 23.exd4
.ixe4 24.j'!c1 'it>b8 25 . .ixa6 h5= ,
Persson-Coleman, ICCF 2 007.
10 . .id2 d6
90
ll.a4
White has never tried the seem­
ingly most consistent continuation
ll.h3 tt'le5 1 2 .tt'lxe5 dxe5 13.f4 .ic5+
14.'it>h1 .id4 15 .f5. I suppose that he
does not like the possibility for long
castling, e.g. 15 . . . exf5 16.exf5 .id7
17.tt'le4 .ic6 18.�e1 0-0-0+.
The first top-level game in this
line, Carlsen-Ivanchuk, Skander­
borg 2005, saw ll.tt'le2 .ie7 12 .c4
(White's knight has no business on
the kingside : 12 .h3 tt'le5 13.tt'lxe5
dxe5 14.tt'lg3 0-0 15.tt'lh5 tt'le8 ! ?
16.�e2 b 6 17.j'!ac1 .ib7 18.c4 j'!d8
19.j'!fd1 f5t, Moller-Petukhov, ICCF
2010) 12 . . . 0-0 13.j'!cl b6 14.tt'lc3 .ib7
15.�e2 tt'ld7= .
ll . . . b6
The game Sanner-Tosi, ICCF
2009, suggests that Black can allow
a4-a5 - ll . . . .ie7 ! ? . This move may
be even more accurate since it dis­
courages 12 .h3 tt'l e5 13.tt'lxe5 dxe5
14.f4 due to 14 . . . exf4 ! ? 15 . .ixf4 e5
16 . .ig5 .ie6= . Sanner chose 12 .a5
.id7 13.h3 tt'le5 14.tt'lxe5 dxe5 15.�f3
0-0 16.j'!fd1 .ic6 17.tt'la4 j'!ad8 with a
comfortable game for Black.
12 .%Ye2 il.e7 13.h3 ttle5
14. ttlxe5
14.tt'ld4 0-0 15.f4 tt'lg6 is simi­
lar to line B 2 . Instead, 15 . . . tt'lxd3
16.cxd3 e5 17.tt'lc2 left White with a
slight pull after 17 . . . .ie6 18.j'!ac1 exf4
19 ..ixf4 �b7 2 0 . tt'ld4 j'!ac8 2 l..ig5,
Mueller-Ribli, Germany 2 0 07.
 The Third Rank Set-Up
14 ••• dxe5 15.f4 0-0 16.gael
The game is balanced. It is pos­
sible to meet f4-f5 by . . . exf5, for
instance, 16 . . . �c5 + ! ? 17.'it>h l �d4
18.f5 exfS 19.exf5 �b7 2 0 .�g5 hc3
21.bxc3 l'!fe8 2 2 .hf6 gxf6 = .
I n Dembo-Delchev, Cappelle
la Grande 2006, I refrained from
this exchange : 16 . . . �b7 17.f5 :!! feB
18.fxe6 fxe6 19 .�g5 ttJd7 2 0.he7
l'l:xe7 2 1.�c4, but the e6-pa\vn is a
permanent weakness.
B2. 9 . h3 .tc5
Taimanov's pet line 9 . . . b5 has
faded out of fashion. The famous
game Fischer- Petrosian, Santa Mo­
nica 1966, went 10.f4 ttJc4 ll.�c4
j!,!fc4 12 .'1Mfd3 dS 13.e5 ttJd7 14.\MI'xc4
dxc4 15.f5 ttJxe5 16.fxe6 fxe6 17.l'!ae1
with a pull in the endgame, e.g. 17 . . .
1Lld7 18.a4 b 4 19.ttJd5.
Main branches here are :
B21. 10.ttJa4 ; B 2 2 . 10 .�e2 ; B23.
10.'it>h1
10.f4 d6 transposes to other
lines.
10 .�d2 d6 ll.f4 ttJg6 12.'1Mff2 sets
the positional trap 12 . . . 0-0 13.e5
dxe5 14.ttJxe6 (which is far from
clear after 14 . . . he6 15.hc5 ttJxf4
16.hf8 l'!xf8�). Perhaps Black
should neutralise it by 12 . . . b6 and it
is unclear what plan White actually
had in mind.
Womacka-Zakhartsov, Guben
2 0 1 1 , went further 13.ttJce2 0-0
14.c3 eS 15.ttJb3 (15.f5 exd4 16.cxd4
ttJeS) 15 . . . he3 16. �xe3 exf4 with an
excellent game for Black.
In my game Eggleston-Delchev,
Bad Wiessee 2 0 13 , White imitat­
ed activity by 13 .g4? ! , but 13 . . . e5 !
14.ttJde2 (14.fxe5 dxeS 15.ttJde2 �e6
16.hc5 bxc5 17.b3 0-0 18 .�c4 hc4
19.bxc4 \MI'a5 2 0 .\MI'f3 l'!ab8 2 l .tLld5
l'!b2 2 2 . tLl xf6+ gxf6 23.\MI'xf6 \MI'd2)
14 ... exf4 15.tLlxf4 ttJxf4 16.�xf4 �e6
17.l'!ae1 0-0 18.ttJd5 hdS 19.exd5
ttJd7! gave me a typical Sicilian
structure with a better knight and a
safer king.
B 2 1 . 1 0 . � a4 .ia7 11.c4
According to Taimanov, the
whole variation with 9 . . . �c5 gives
Black a restrained position due
91
 Part 3

to the possibility of ll.c4. This

understanding ensues from the old
passive treatment of the hedgehog
structures by Black. If the second
player aimlessly manoeuvred on
the last two ranks, he undoubtedly
would risks to get under a kingside
attack. On the contrary, active play
on the dark squares with the break­
through b7-b5 in mind, ensures him
a comfortable game.

15.Yfe2

15.a4 only weakens the b3-

Here, 11 . . . 0-0 12J:'k1 �dB seems
the most concrete continuation. It
counts on luring the opponent into
c4-c5. Then Black would obtain ex­
cellent counterplay by . . . d6. How­
ever, White can calmly pursue his
main plan for consolidation, when
the rook move to dB could prove to
be inaccurate since . . . �feB or . . . �feB
could be better. Even more impor­
tant consideration is that Black may
be able to carry out . . . b7-b5 or . . . eS
without moving his rook at all.
ll . . . d6 12.gcl .id7 13 .�c3 0-0
14.b3 Yfa5 ! ?
The queen i s eying the knight o n
c3, making � c 2 impossible.
pawn. Black continues with 15 . . . .icS
intending to meet 16.�e2 or i 6 . .ie2
by . . . �b6.
In the game Astaneh-Vl. Dimit­
rov, Vila de Marin 2 005, White
chose 15 . .ie2 . This move has at least
two drawbacks. The bishop is too
passive there because it is not aimed
at the enemy king, and it is hinder­
ing the protection of the e3-bishop
by �e1 or �e2 . Black followed up
with 15 . . . .ic5 16.a4 �feB 17.f4 ltJg6
(17 . . . ltJc6 ! ?) 18.�d2 , when 1B . . . e5 ! ?
1 9 .feS deS+ would have seized the
initiative. Dark-squared strategy
was also good, e.g. 16 . . . �b6 17.�e1
ltJc6 = .
15 . . . l"!fe8
Black's rook stands up in oppo­
sition to the enemy queen. It is clear
that White is unable to generate any
threat, see game 13 Varga-Almasi,
Budapest 2 0 04.

92
 The Third Rank Set-Up

B22. 10 .Ve2 White can prevent . . . bS, but

Black has good places for his piec­
This is the oldest set-up. It was es anyway: ll.a4 b6 12 .lt'lb3 he3
tested at highest level in the World 13.�xe3 0-0 14.gael .ib7 15.f4 lt'lg6
title match Spassky-Petrosian, 16.lt'ld2 lt'ld7 17.lt'lc4 gadS 18.�f2
Moscow 1969. dS ! , Cuenca Jimenez-Zakhartsov,
Benasque 2 0 1 0 .

ll . . .li�g6

The game Spassky-Petros ian,

showed that ll . . . lt'ld7 leaves Black's
castling position shaky under the
impact of the d3-bishop. Capturing
on d3 is sad, because we would have
to struggle without counterplay
against a possible queenside activ­
ity of the opponent.
The reason behind the text move
is to protect the bishop on e3 and ex­ 12.li:lb3 .ixe3 13 .Vxe3 0-0
pand in the centre without further
prophylaxis. White's next moves
are f4, lt'lb3 and a quick eS. The ex­
change of the dark-squared bish­
ops exposes two critical points in
Black's position - d6 and cS, but the
biggest danger usually comes from
the kingside and the f6-square.
With time, Black has learnt to
neutralise the enemy's threats. All
he needs to know is a couple of ac­
curate moves in order to finish de­
velopment. Black can also choose 13 . . . b5
since the only attempt to punish
10 . . . d6 this move order would be 14.a4,
but then 14 ... b4 15.lt'lce2 0-0 16.a5
10 . . . lt'lg6 has no positive sides gbs (preventing �b6) would be fine
after ll.lt'lb3. However, White could for him. 13 . . . b5 also provokes 14.e5
try to exploit the unnecessary re­ dxeS 15.hg6 hxg6 16.fxe5 lt'ld7
treat by ll.lt'la4 followed by c4. 17.gxf7? which loses after 17 .. .c;t>xf7
18 .�f3 + lt'lf6 ! 19.�xa8 Vb6 + 2 0mfl
ll.f4 ghs ! .

93
 Part 3
14.gael
14Jiad1 bS 15.e5 dxeS 16.hg6
is not dangerous due to 16 . . . hxg6
17.fxe5 lt:ld7 18J':'id6 b4, equalising.
14.c;!;>h1 b5 15.a4 b4 16.lt:le2 eS= .
14 b5 15.e5
•••
15.a3 loses momentum. Black
was fine in the game Exizoglou-Vl.
Dimitrov, Thessaloniki 2004: 15 . . .
i.b7 16.lt:ld4 :B:ae8 17.'Wf2 e5 18.lt:lde2
(18 .fxe5 :B:xeS) 18 ... ef4 19.lt:lf4 lt:lf4
20.'Wf4 :B:eS+.
15 dxe5••• 16.i.xg6 hxg6
17.fxe5 lt:ld7 18. lt:ld4 J.b7 19.lt:ldl
With this final touch, Black
solves the opening problems. He
is threatening . . . lt:lxeS ! In Radulov­
Suetin, Budapest 1970 , White an­
swered 20 .c3 lt:lb6 21.lt:lf2 lt:lc4 and
Black already had an edge.
B23. 10.c;!;>hl
This is another straightforward
plan. In distinction to B22 which
94
is a positional line aimed basically
at obtaining a spatial advantage by
e4-e5, here White is trying to pre­
serve his dark-squared bishop for a
kingside attack. Black's counterplay
is directed toward the knight on d4.
He should also open the e-file. Prac­
tice has seen long castling too, but
that is really playing with fire.
10 ••• d6 ll.f4 lt:lg6
12 .'Wel
A multi-purpose move. The
queen defends the e3-bishop, en­
abling the plan with lt:lb3. At the
same time it is eying the square h4
which is its ideal attacking posi­
tion. Black is at a critical juncture
and I will examine his options in
detail, but let us also see some oth­
er White's options:
a) 12 .'Wd2 0-0 13 . :B:ae1 (13.'Wf2
:B:e8) 13 . . . b5 14.a3 :B:e8. Black got
precious time for development and
stands well, see game 14 Onis­
chuk-Salov, Elista 1998.
b) 12.�f3 leads to similar po­
sitions after 12 . . . 0-0 (12 . . . b5 ! ?)
 The Third Rank Set-Up

13.Elael bS. 13 . . . i.d7 is also playa­

ble, but be sure to avoid the typical
mistake, seen in the game Reinaldo
Castineira-Comas Fabrego, Andor­
ra 2001: 14.ltJb3 .be3 15.�xe3 eS? !
(15 . . . i.c6 is essential ! ) 16.fxe5 dxeS
17.Elxf6 ! gxf6, when 18.lDd5 ! �d6
(18 .. .'�d8 19.Elfl fS 2 0 .exf5 hfS
2l.Elxf5 �ds 2 2 .�h6 Elfc8 23.lDd2±)
19.lt'la5 ! i.c6 2 0 . ltJ c4 �d8 21.lDcb6
.ixdS 2 2 . exd5± would have been un­ 16.i.g1 exfS 17.b4 �xb4 18.Eiab1
pleasant for Black. �a4 19.exf5 Elhe8+, Haznedaroglu­
Ribli, Saint Vincent 2 0 05, or 16.b4
c) 12 .f5 ltJeS 13.ltJce2 (13.�e1 �xb4 17.Elab1 �a4 18.Elb3 ltJxd3
is well met by 13 . . .i.d7) 13 . . . exf5 19.cxd3 eS 2 0 . ltJ c3 �aS 2 1 .lDc2
14.exf5 0-0 15.i.g5 ltJed7. i.c6+, Stoumbos-Javakhishvili, Li­
nares 2005.
It seems that Black is safe in this
Perhaps after 13.f5 ltJeS, White
position. White is unable to break
should opt for the more restrained :
through his defence until he is hold­
ing the critical squares f6 and h6. 14.ltJce2 �b6 (14 . . . exf5 15.exf5
At the same time, the open e-file 0-0-0 16 .b4 Eihe8 is unclear) 15.�g3
and the weak square e4 ensure suf­ 0-0 16.b4 �xb4 17.c3 ! �b6 18.i.h6
ficient counterplay. See game 15 ltJfg4oo.
Kotronias-De la Riva, Bled 2 0 0 2 .
13.f5 ltJe5 14.Yfh4
12 ••• 0-0 ! ?
14.�g3 is harmless in view of
12 . . . i.d7 i s a decent alternative. 14 . . . b5 or 14 . . . �h8. Or 14.g4? dS !
Then 13.f5 (13.ltJb3 he3+
14 b5
14.Vtfxe3 0-0 15.Eiad1 Elad8 is a
•••

calm, typical Sicilian game.) 13 . . .

ILleS 14.�h4 0 - 0 15.Elf3 exfS 16.exf5
li:Jxf3 17.gxf3 i.c6 is a better version
of the main line, because Black's
queenside has not weaknesses, see
game 10 Vatter-Zakhartsov, Bad
Wiessee 2 0 1 1 .
I n this line, Black can also cas­
tle long: 14 . . . �b6 15.ltJce2 0-0-0.
White has failed to meet it ade­
quately so far:

95
 Part 3

15.gf3 15 ...tt:Jxf3 16.gxf3 tlle8 is little

I'm still unsure what sign to tack
onto this move but it is undoubt­
edly the most challenging contin­
uation. Macedonian GM Mitkov's
idea (it was first employed against
the set-up with ...b5 by Nedev) is
truly amazing and it faces Black
with considerable practical prob­
lems. Computers think that the ex­
change sacrifice is not entirely cor­
rect, but over the board humans do
not cope very well with the heavy
calculations that are vital to survive
the attack.
Lately White began to try other
moves, too:
15.tllce2 exf5
Houdini 4 suggests 15...Wa7
(Hector-Rasmussen, team ch.
explored. Perhaps Black fears the
closed centre which facilitates the
opponent's attack. A possible line is
17.:ggl g6 18.:gg4 Wb7 19.Wh6 :ga7!
20.b4= hb4 21.:gh4 f6 22.fxg6 Wg7
23.tllc6 :gc7 24.gxh7+ r4;>h8 25.tt:Jxb4
:gxc3 26.:gg4 Wxh6 27.hh6 :gf7
28.id2 :gxd3 29.cxd3 r4;>xh7.
16.exf5 c!Llxf317.gxf3 .ib7
Adla and Glavina suggest 17...
d5? which offers White an edge af­
ter 18.:ggl We5 19.tlldl id7 20.c3±.
19.r4;>h2 loses to 19...We7 20.Wg3
We5 2l..if4 tllg4+.
19 .ie3 20.�f6 .ixgl 21.iWg5
•..

DEN 2008, saw 15...Wb6?! 16.b4!

ix:b4, when 17.fxe6 fxe6 18.tll xe6
Wxe3 19.tllxf8 r4;>xf8 20J�xf6+ gxf6
21.tllf4! would have torn Black
apart) preparing to meet 16.b4 by
16...ib6, but I do not like to distract
my strongest piece from the king­
side in such a crucial moment of the
game. White might get a somewhat
better version of the main line with
16J'U3 exf5 17.exf5 tt:Jxf3 18.gxf3 We7
19.ih6 ib7 20.:gg1 ix:f3+ 2l.tt:'lxf3.
This crazy position first arose
16.exf5 :ge8! (Belov-Volkov, in Stojanovski-Jovanic, Sarajevo
Moscow 2009 saw 16...ib7 and 2006, but it gained popularity af­
here 17.ih6 keeps the initiative) ter the game 11 Carlsen-Vachier
17.ih6 We7 18.Wg5 Wf8 19.Wxf6 Lagrave, Cap d'Agde 2006. It went
gxf6 20.ix:f8 r4;>xf8=. 2l...Wc6 22.tlle4 f6 23.Wxgl. In The
Safest Sicilian I proposed the nov­
15 exf5
•..
elty 23...Wd5!. It was tested in the

96
 The Third Rank Set-Up
game Hamilton-Foulds, ICCF 2 0 1 0 .
Apparently, Black is only slightly
better after it. Perhaps he should
try to play for a win by:
21 . . . ti'c5 2 2 .�xgl
22."�'xgl 'Wxgl+ 23.Wxgl l:Uc8
should be in Black's favour because
White's knights have not access to
the e6-square. Besides, Black can
also play 22 .. J'Ue8 or 22 . . . Wh8 as
shown below.
22 . . .!:�ae8 ! ?
I borrowed this set-up from ·
game 10 Vatter-Zakhartsov, Bad
Wiessee 2 0 1 1 (where the pawn is on
b7) . Its idea is to refrain from . . . f6
as long as possible. Instead, Ev­
ans-Veen, ICCF 2 007, went 2 2 . . .
f6 23.'1Wg4 d S 24.�f3 �feB 25.Wg2
�adS 26.a3 �d7 27.�d4 �eS 28.�e6
'We3 29.lt:ldl ti'el 3 0 .lt:lf2 �e7 31.h4
�5xe6, draw.
After the text, Black has more
chances to convert his material ad­
vantage. His plan is to give up a rook
for a knight and play with a rook+2
pawns vs. 2 minor pieces in an open
position, for example: 23.�h4 (23.
lt:le4 'WdS) 23 ·... �e5 24.lt:lf3 �xfS
25.'Wg4 �hS+.
97
Pa rt 3 . The Third Ra n k Set- U p
Ka n - Step by Step
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4.�xd4 a6 5 . � c3 Wc7 6 . .id3
White has defined his set-up and
we should make our choice right
now. The Taimanov is a tempting
option against i.d3 so we can an­
swer:
a) 6 . . . tt:lc6. White's only reason­
able deviation is 7.�xc6, when all
three recaptures are possible.
7 ... dxc6 8.0-0 e5 9.f4 tt:lf6
10.\t>hl i.d6 11.f5 is a bit cramped
for my taste.
7 . . . bxc6 8 . 0-0 ttlf6 9.�e2 d5 is a
popular set-up. White's game is too
straightforward and easy. Moreo­
ver, Black has practically lost a tem­
po on . . . a6, compared to the pure
98
Taimanov move order with 4 . . . tt:lc6
5.tt:lc3 Wffc 7 6.tt:lxc6 . thus my vote
goes for the unexplored and flexi­
ble continuation:
7 . . .'\Wxc6 ! ? 8. 0-0
8 . . . b5
Grischuk tried in a blitz game 8 . . .
tt:le7 ! ? 9.l"'el tt:lg6 10 .e5 �c7 1l.V!fh5
b5 with mutual chances.
9 .�e2 Wff c 7! ? (9 . . .i.b7 10 .a4 b4
11.ttld5 ttlf6 1 2 .ttlxf6+ gxf6 should
be slightly better for White) After
the text, ttld5 is no longer a threat.
Both opponents are on their own
-
I do not know of any practical ex­
ample.
b) 6 . . . b5 is beyond the scope of
our repertoire. I advocate to de­
lay the fianchetto in favour of . . . d6,
. . . tt:ld7.
 The Third Rank Set-Up
A. 7.'<!9e 2 ; B. 7. 0 - 0 ; C. 7.f4
7.f4 bS ! is considered in Part 4.
7. .te3? ! is totally innocuous
here. Black leads out his bishop to
b4 or cS and obtains a good game:
7. . . .tb4 8 . 0 - 0 hc3 9.bxc3 d6!
(9 . . . d5? lO.exdS ttJxdS ll ..td2 0-0
12 .'<!9h5 g6 13.�h4;!;) 10 .'<!9f3 eS
ll.ltlfS hfS 1 2 .exf5 ttJbd7 13J3abl
d5 ;
7. . . .tc5 8 .�e2 (8.0-0 d6 9.'<!9e2
tt:Jbd7) 8 ... d6 9 . 0-0-0 0-0.
7 . . .b5 is also possible, but the
play is very sharp after 8.f4 .tb7 (8 . . .
b4? stumbles into a very interes­
ting new idea: 9.ltla4 .tb7 10.�f3 d6
11.0-0 ttJbd7 12 .a3 ! bxa3 13J3xa3 ! e5
14.fxe5 dxe5 15.ltJf5 with a strong
attack) 9.'<!9f3 ltlc6 1 0 . 0-0-0 b4
ll.ltlce2 h5.
A. 7.�e2 J.d6 !
7. . . d6 offers White very sharp
Scheveningen type options as 8.f4
or 8.g4. The text is familiar to the
Taimanov players. It significantly
narrows the opponent's choice.
8.J.d2
a) Fortunately, 8.ltldb5? axbS
9.ltlxb5 is not winning. In fact, Black
is better after 9 . . . '<!9a5+ 10 . .td2 .tb4
l l.hb4 �xb4+ 12 .c3 �cS 13.e5
ltldS 14.ltJd6+ �f8 15.�h5 (15.'<!9f3
fS 16.hf5 ltJf6 17. 0-0 �xeS) 15 . . . g6
Somehow Black's king escapes
and White has not even a draw:
16.�h6+
Or 16.hg6 fxg6 17.'<!9f3+ �e7
18 .�f7+ �d8 19.�g7 l'!e8 20 .'<!9f7
ltlc6 ! ! 2 1.�xe8+ �c7 2 2 . 0-0 ( 2 2 .
� h 8 ltJf4 23.0-0 ttJxe5 24.ltJe4 �ds
25.f3 b6+) 22 . . . ttJxe5 23.ltJe4 '<!9c6+.
16 . . . �g8 17.h4 (Black's minor
pieces are very strong following
17 . .te4 l'!a6 18.�g5 l'!xd6 19.exd6
b5, e.g. 2 0 .hd5 .tb7 2 1 .�d8+ �g7
2 2 .�xh8+ �xh8 23 .hb7 �eS+t)
17 ... ttJc6 18 .h5 ttJxeS and White re­
signed in the blitz game Kasim­
dzhanov-Rublevsky, Moscow 2 007.
Alternatively:
b) 8.ttJf3 ltlc6 9. 0-0 0-0 10 .h3
ttJe5 ll.ltlxe5 he5 = .
c ) 8 .h3 ttJc6 9.ltlxc6 (9.ltlf3 b6
1 0 . 0-0 .tb7 l l . .td2 ltle5) 9 . . . dxc6
10.0-0 0-0 ll.f4 e5.
99
 Part 3
8 ••• �c6 9.�f3 (9.llJxc6 dxc6
10.g3 bS) 9 0-0 10.0-0
••.
10 b5 lU!ael �g4 12 .h3
•••
�ge5 13.�xe5 �xe5 14.f4 �xd3
15.cxd3 f5 !
There is no reason to allow
15 . . . b4 16.�a4. The text address­
es Black's main problem - the pas­
sive light-squared bishop. The play
might continue 16.Eicl �b7 17.�e3
�dB = .
B . 7.0-0 d6
The majority of players prefer:
a) 7 . . . �c6, but then 8.�xc6 bxc6
or 8 . . . dxc6 are too easy to play with
White. There is nothing wrong with
these lines, but I would like to put
the opponent in a situation where
he would have tougher decisions to
make.
I would like to forewarn you
about the popular line:
b) 7 . . . �c5?! 8.�b3 �e7. Hellsten
devotes a whole chapter on it in his
book, but I believe that White's at­
tack is very strong after:
100
9.f4 d6 10.e5 ! ? dxeS
10 . . . llJfd7 ll.exd6 �xd6 12.llJe4
(Topalov beat Svidler with 12 .�e3 ! ?)
12 . . . �b6+ 13.@h1 llJf6 14.%Vf3 llJbd7
15.�e3t .
1 1.fxe5 llJfd7
11 . . . %Vxe5 12 .�f4 %Vh5 13.�e2
�g6 (13 . . . %Vh4 was refuted in the
game Kryvoruchko-Smirin, Plovdiv
2008, which went 14.g3 %Vh3 15.llJe4
eS 16.llJd6+ ! ) 14.h4 (Stohl suggests
14.Eif3) 14 . . . h5 15.�g5 and Black is
helpless against the threat of �d3,
for example: 15 ... llJg4 16.he7 @xe7
17.%Vd4± or 15 . . . llJg8 16.�d 3 fS
17.�e2 llJc6 18.llJa4±, Paragua-Bil­
guun, Jakarta 2 0 13 .
12.%Vg4 g6
13.�h6! llJxeS 14.%Vg3 llJbd7
White also has full domination
in the event of 14 . . . llJbc6 15.�e4 �d8
16.�f4 f6 17.hc6+ bxc6 18.llJe4 0-0
19.@h1, e.g. 19 ... %Vg7 2 0.he5 fxeS
2 l .Eixf8+ �xf8 2 2 .%Vd3 Eib8 23.Eif1
%Ve7 24.�d 2 + - .
15.Eiae1 �f8 16.hf8 @xf8
17.@h1 @g7 18.ltJe4 Eif8 19.llJg5 %Vd6
2 0 .llJd4+-, Zinchenko-Filip, Paleo­
chora 2 0 1 0 .
 The Third Rank Set-Up
Main branches now are :
Bl. 8 .a4, B 2 . 8.�e2, and B3. 8.f4
followed up by �f3 or l2Jf3.
Bl. 8.a4 b6 9.f4
9.�f3 l2Jbd7 10 .i.d2 i.b7 1 l.�g3
g6 12J'!fe1 counts on the trap 12 . . .
1g7? ! 13.l2Jdb5 axbS 14.l2Jxb5 �b8
15.tt:lxd6+ c;!(f8 16.i.b4 c;!;>g8 17.i.b5�.
Black can parry the threat with 12 . . .
liJhS ! . A more sophisticated version
of the same idea is 1U!fel. Then af­
ter ll . . . g6, White can put his queen
on h3, followed up by i.h6. Kozul
answered this move order with 11 .. .
liJeS ! ? 12 .�h3 (12 .�g3 hS ! ?) 12 . . .
h5 (12 . . . i.e7 ! ? = ) and quickly seized
the initiative after 13.tt:lb3 (13.a5 !
1e7 14.axb6 �xb6 1S.tt:lb3 l2Jfg4
16.1e3 l2Jxe3 17.�xe3 �xe3 18J'%xe3
li:lxd3= ) 13 . . . i.e7 14.f4 l2J eg4 15J''1e2
d5. You should remember to ignore
a4-a5 since . . . bS? ! provokes danger­
ous sacrifices.
9 . . . J.b7 10.ti'f3
a) 10.f5 eS ll.tt:lb3 demands pre­
cision from Black. He must an­
ticipate the pawn storm with g4-
g5 which would be possible after
l l . . . i.e7 12 . .ie3 tt:Jbd7 13.�f3 0-0
14Jiad1 [14.g4? dS 1S.tt:lxd5 (15.
exd5 h6 16.h4 e4) 15 ... l2Jxd5 16.exd5
li:lf6] 14 . . . .ic6 ? ! 15.g4 h6 16.h4 tt:lh7
17.g5 ! hxgS 18.hxg5 hgS 19.hg5
li:lxg5 2 0 .�g3 l2Jh7 2l.�f2 t.
Since this is the only serious
threat of White, it would be wise to
neutralise it in advance by 14 . . . h6 ! .
Then 15.h4 will be met by 15 . . .h5
while 1S.�f2 i.c6 16.g4 could be an­
swered with 16 . . . l2Jh7 17.�g3 �d8 .
In a blitz game, Ivanchuk even de­
nied g2-g4 altogether by playing h7-
h5 himself, but this plan abandons
short castling and involves more
difficult decisions from Black in fu­
ture.
b) 10 .�e2 tt:Jbd7 ll . .id2 .ie7
12 .�ae1 is somewhat underestimat­
ed.
Now it is rather risky to castle
under the fire of all White's piec­
es: 12 . . . 0-0 13.e5 dxeS [13 . . . tt:ld5? !
14.tt:lxd5 ! (14.tt:lxe6 fxe6 15.�h5 g6
16 . .b:g6 lD7f6 17.exf6 l2Jxf6 18 .�h6
hxg6 19.�xg6+ c;!;>h8 is rough­
ly equal: 2 0 .l2Je4 he4 2l.�xe4
l2Jxe4 = , or 20.�xe6 �g8 2 1.�h6+
l2Jh7 2 2 .�f2 .if8oo) 14 . . . hd5 15.c4
i.b7 16.f5t] 14.fxe5 .icS 1S.�f4 g6
16.�h4 tt:lh5 17.c;!;>h1 l2Jg7 18 . .ih6 .ie7
19.�h3 . Black can unload the ten­
sion with 19 . . .�xe5 20 .�xe5 tt:lxeS
2 l.�xe5 .if6 22 . .if4 heS 23 .he5
f6 24.i.c7 bS 2S.l2Jb3t, but the two
minor pieces might soon take con­
trol of the board.
I suppose that it would be safer
to kill the d3-bishop:
101
 Part 3
12 . . . lLlc5 13.b4 lLlxd3 14.cxd3
0-0 lSJ'kl �d7. I discuss this pawn
structure in the annotations to
game 16 Cubas-Hellsten . The dif­
ference is that White's queen is on
e2 here. This is perhaps in White's
favour as his queen is not under the
X-ray of the b7-bishop, and the d4-
knight has a retreat square on f3. In
result, Black lacks the counterplay
with . . . i.e7-f6xd4, followed up by
. . . f7-f5. However, White also has
not a clear plan. For instance, 16.g4
lLle8 (threatening with 17 . . . d5)
17.E1bl .if6. If White opts for a wait­
ing game with 16.Whl E1fc8 17 . .ie3 ,
we can adopt the same set-up I rec­
ommend against 1 0 .�f3 : 17 . . . lLle8
(planning ... �d7, ... .if6) 18.f5 eS
19.lLlf3 bS 2 0 . axb5 axbS 2l.E1fd1
lLlf6=.
lO • .• lLlbd7 ll .id2
•
a) 1l..ie3 has no advantages over
ll . .id2 . The bishop only hampers
the break e4-e5 and it does not de­
fend the c3-knight in the event of
b2-b4. Black should follow the same
plan as in the main line:
ll . . . .ie7 12 .E1ae1 0-0 13 .�h3 (13.
g4 lLlc5 14.g5 lLlfd7) 13 ... lLlc5 14 . .if2
(14.e5? fails to 14 . . . dxe5 15.fxe5
�xeS) 14 . . . E1fe8 15.Whl. The play is
balanced. Black has several ways of
manoeuvring further.
One straightforward plan is to
swap the d3-bishop. For instance :
15 . . . �d7 16.lLlb3 lLlxd3 17.cxd3 eS=
18.hb6 �xh3 19 .gxh3 .ic8 2 0 .f5
E1b8 2 l.a5 .idS.
102
b) ll.Wh1 is a bit slow. Black can
follow the main plan with ll . . . .ie7
12 . .id2 0-0 13.�h3 lLlc5 14.b4 lLlxd3
(14 . . . lLlxe4 15.lLlxe6 is slightly better
for White) 15.cxd3 E1fc8.
ll ... g6 ! ? is also possible. Que­
sada-Arencibia, Santa Clara 2 007,
coninued 12 . .id2 .ig7 13.E1ae1 0-0
14.�h3 E1ae8 15.lLlf3 lLlcS with mu­
tual chances.
u .te7
...
Black has another, much sharp­
er set-up which was tested only in
the game Antal-Wallner, - Austria
2003:
ll . . . g6 ! ? 12 .�h3 .ig7 13.a5
Black's main idea is seen in the
variations 13.f5 gxfS ! 14.exf5 eS fol­
lowed up by long castling, or 13.lLlf3
0-0 14.f5 gxfS ! ? 15.exf5 eS 16.�g3
Wh8 where the open g-file might be
in Black's favour. The text conceals a
trap - 13 . . . b5? 14.hb5 ! so we castle:
13 . . . 0-0, and if 14.axb6 �xb6!
15 . .ie3 lLlg4 = . Critical is:
14.f5 gxfS (14 ... exf5 15.exf5 lLl e5)
15.exf5 eS 17.lLlb3 bS
 The Third Rank Set-Up
17.g4 b4 18.lt'ld1 e4 19 .i.e2 ltJdS
20.gS i.eS ! ( 2 0 . . . �h8 21.f6 llJ 7xf6
22 .gxf6 .b:f6 23.gxf6 gg8+ 24.�f2
lt:lxf6 2S.lt:Je3±). Black has an initia­
tive, e.g. 2 1.c4 bxc3 2 2 .lt'lxc3 e3.
12 .gael 0-0
12 ... lt:J c5 13 .b4 lt:Jxd3 is also play­
able, but the text is more ambitious.
Black waits for White to play 13. �h3
before putting 13 . . . ltJcS. I explain
the subtleties of this position in the
annotations to game 16 Cubas­
Hellsten, Buenos Aires 2 0 0 6 .
B2. 8.�e2 lilbd7
8 . . . bS? ! is a mistake before
White has committed himself with
f4. After 9.a4 ! b4 10.lt'la2, followed
up by c2-c3, White opens up the
queenside to his favour.
9.�hl ! ?
9.f4 �b6 ! ? i s a tough nut to crack.
The complications after 10 . .ie3
V9xb2 ll.ltJdbS axbS 1 2 .ltJxbS gas
13.gfb1 gxbS 14.gxb2 gxb2oo might
lead to a drawish position follow-
ing 1S.�d2 ltJcS 16 . .id4 gb6 17.�aS
lt:Jfd7 18.-b:cS dxcS 19 . .ibS gxbS
2 0.�xbS .id6 2 1.eS !J.c7 2 2 .h4. It is
true that Black cannot disentangle
his minor pieces, but on the other
hand, White also is unable to break
the fortress.
The other alternative to the main
line with . . . i.e7 is the set-up with
9 . . . bS? ! 10.i.d2 .!b7 1l.�h1 (ll.gae1
!J.e7 12.�h1! b4 ! ) ll . . . .!e7
White gets an edge with 12 .b4 !
0-0 13.a4 bxa4 14.gxa4 lt:Jb6 1S.gaa1,
when 1S . . . dS 16.eS lt:Je4 17.lt:Jxe4
dxe4 18 . .b:e4 .b:e4 19.�xe4 �c4
2 0.c3 retains a healthy extra pawn.
The delay of f2-f4 aims to dis­
courage both . . . bS and . . . �b6.
9 ..id2 achieves the same effect, for
instance : 9 . . . bS? ! 10 .b4 i.b7 ll.a4.
However, the set-up with 9 ... g6
is quite possible since the bishop
would be passive on d 2 .
9 . . . .!e7
9 . . . bS (9 . . . .ie7 1 0 . .id2 transpos­
es to the main line) 10 .a4 b4 1l.lt'la2
�b7 12 . .id2 aS 13.c3 bxc3 14.lt:Jxc3
ltJcS 1S . .ibSH.
103
 Part 3
9 . . . g6 10.f4 ig7 11.fS tileS 12.tt::l f3
is slightly better for White who has
a clear plan on the dark squares,
for example, 12 ... tt::l c6 13.�f2 id7
14.\1;l/h4t or 12 ... tt::l hS 13.\1;lff2 tt::l xf3
14.\1;lfxf3 ieS 1S.ih6t.
10.id2 ! ? 0-0
10 ... tt::l c5 ll.f4 bS? loses to 12 .eS.
See game 17 Delchev-Castellanos
Rodriguez, Leon 2012.
ll.f4 �c5 12.gae1
Or 12 .eS tt::l fd7 13J'!ae1 tt::l x d3
14.cxd3 dxeS 1S.fxeS ih4 !
Black is at a crossroads. Perhaps
his strongest move is:
104
a) 12 ... bS which, however, ac­
cepts a draw in the variation 13.eS
tt::l fd7 14.ixh7+ (14J'!f3 tt::l xd3
1S.cxd3 dxeS ! 16.fxeS ih4) 14 . . .
lt>xh7 1S.�hS+ lt> g 8 16J'!f3 dxeS
17J'!h3 fS 18.\1;l/h7+ .
Instead, 13.b4 tt::l x d3 14.cxd3
ib7 1SJ'k1 �d7 is fine for Black. See
game 19 Moen-Topalov, ECC Rho­
des 2 0 13, about this structure.
b) 12 . . . tt::l x d3 is passive, but at
least Black will be playing for all
three results.
13.cxd3 bS
The life-long Kan player, GM Ve­
likov, chose 13 . . . \1;l/d8 14.g4 h6, but
this is really too provocative. White
has a serious initiative after 1S.gS
hxgS 16.fxgS tt::l d 7 (16 . . . tt::l h 7) 17.h4.
14.g4
Alternatives are :
14.eS dxeS 1S.fxeS tt::l d 7 16J'!c1
\1;lfb6 17.ie3 ib7 18.tt::l fS �dB = ;
14.�f2 b4 1S.tt::l c e2 aS 16J'!c1 (16.
�g1 eS 17.tt::l fS ixfS 18.exfS �c2+)
16 . . . \1;l/d7 17.tt::l c 6 ib7 18.tt::l x e7+
�xe7= ;
14.�c1 \1;lfd7 1S.a3 ib7.
14 . . . b4 1S.tt::l d 1 dS 16.eS tt::l d 7
 The Third Rank Set-Up

The continuation of White's at­ V9xb2 is dubious: ll.ltJcbS axbS

tack is not trivial. Our defence is 1 2 .ltJxb5 �b4 (12 . . . �a5 13.�ab1 �xbS
based on the potential power of the 14.hb5 V9c3 15.�b3 V9c7 16 . .id4±)
b7-bishop : 17.f5 .ib7 18J'U3 (18 .g5 13.ltJc7+ 'it>d8 14.ltJxa8 �aS 15.�ad1
exfS 19.lljxf5 d4+ 2 0 .'it>g1 .ic5) 18 . . . �xa8 16.e5 ltJdS 17 . .id4t. Instead,
.ih4 19.ltJf2 V9b6 2 0 . .ie3 .ie7. Black follows his main plan :
The same defensive set-up could 9 . . . .ie7 1 0 . .ie3 ( . . . V9b6 was al­
be employed against 14.a3 .ib7 ready an option : 10 .�g3 �b6
15.g4 (15.b4 �feB 16.g4? dS 17.e5 ll . .ie3 V9xb2 1 2 . ltJ de2 V9a3 13.�xg7
\Wb6 ! ) 15 . . . d5 16.e5 ltJd7 17.b4 (17. �g8 14.V9h6 bSoo; 10 .g4 V9b6 ll . .ie3
Eic1 �ac8) 17 . . . V9b6 18 ..ie3 f6+. �xb2 12.ltJce2 ltJcS)
Note that Black's counterplay is
often based on ... V9b6 when White
cannot find a decent way to protect
his knight on d4. Shift the queen to
f3, and the same structure would be
better for White due to the possibil­
ity of b4, ltJc3-e 2 . Black would risk
to remain without any plan.

B3. 8.f4 10 . . . b5 !
Do not castle under attack in
This move is not too flexible
this set-up ! The queen on f3 and the
as White commits himself on the
bishop on e3 do not generate any
kingside. That rules out plans on
threats in the centre, but they are
the other wing so Black can safely
well set for 10 . . . 0-0?! ll.g4. There­
answer with 8 . . . b5. We saw in line
fore, we should complete develop­
B2 that the early fianchetto was du­
ment and safeguard the kingside
bious due to a4, but after f4 this
before moving our king there.
break is no longer effective. Howev­
er, I recommend to follow the same ll.a3
move order as in line B2 for consist­
ll.g4? is neutralised by ll . . . h6.
ency sake.
The same defence works well in the
event of 1 1.�ae1 .ib7 12 .g4? (or 1 2 .f5
eS 13.ltJb3 hS) 12 . . . h6.
More interesting is 1 2 .'it>h1,
The most popular follow up al­
when 12 ... g6 13.a3 ltJcS transposes
though White should not worry
to ll.a3.
about . . . �b6 . Thus he could play:
Another very interesting plan is
a) 9.V9f3, when 9 . . . V9b6 10 . .ie3 to push ... h5-h4, for instance :

105
 Part 3
12 . . . h5 ! ? 13 . .ig1 h4 14.'�h3 tt:\h5
It is unclear how White could
display activity here while Black has
plenty of useful moves as .. J!c8/d8,
. . . @f8, . . . g6 or .. J!h6. See game 18
Kerekes-A.Kovacs, Hungary 2 0 07.
ll . . . .ib7 12 J!ae1 g6
Simple and good. Now fS is im­
possible, 13.g4 is still bad due to
13 . . . h6, so White has nothing better
than wait.
13.@h1
13 . . . tt:\c5
The combination of . . . g6 + . . . hS
is not perfect, but it is playable: 13 . . .
h S 1 4. .ig1 h 4 15.h3 tt:\h5 16.tt:\de2 ,
Saltaev-Ilincic, Elista 1998. At this
point 16 .. J!c8 (instead of 17 . . . e5?
18.tt:\d5±) 17 . .id4 0-0 would have
been unclear.
106
14 . .ig1 0-0
Black's king is absolutely safe
and nothing can stop the break . . . e5,
e.g. 15.V9g3 eS; 15.'�'h3 eS. Suess­
Spassky, Dortmund 1973 went:
15.V9e2 e5 16.tt:\b3 (16.fxe5 dxe5
17.tt:\b3 tt:\e6) 16 . . . tt:\xb3 (16 . . . exf4 ! ?
17J!xf4 E!ae8) 17.cxb3 exf4 18.E!xf4
tt:\d7= .
b) 9.tt:lf3 has been out of fashion
for quite a while. It was designed to
provoke . . . e5? ! and use the opened
f-file to sac the exchange on f6,
e.g. 9 . . . i.e7 10 .V9e1 e5? ! 11.@h1 0-0
12 .fxe5 tt:\xeS 13.tt:\xe5 dxe5 14 . .ig5
.ie6 15.'1Mfh4 E!fe8 16.i.xf6 hf6
17.E!xf6 gxf6 18.E!f1. Although 18 . . .
�dB ! would tame the first attack­
ing wave, White keeps an initia­
tive. I think that Black has various
promising plans provided he re­
frains from . . . e5. I would also avoid
an early . . . tt:lcS since Black's posi­
tion after b4 tt:\xd3 is solid, but pas­
sive. The most straightforward re­
tort to White's scheme is:
9 ... b5 ! 10 .@h1 (10 .�e1 .ib7
1 1..id2 .ie7) 10 . . . .ib7 11.V9e1
ll . . . b4 12.tt:\d1 dS or 12 . . . tt:\c5
13.tt:\f2 dS 14.e5 tt:\fe4 15 . .ie3 ie7
 The Third Rank Set-Up

16.id4 0-0 17.�e3 , Arnason-Pli­ 13.�h3

ester, New York 1989, 17 .. Jl:ac8 = .
In Thipsay-Lomineishvili, Kishi­
Black can also keep the centre
nev 1995, White chose 13.b4 l'l:fe8
fluid by ll . . . ie7 12 .id2
14.�h3, when the thematic 14 . . .
12 .e5 dxeS 13.fxe5 tt:lg4 is good
e S 15.tt:lf5 if8 solved the opening
for Black: 14.�g3 hS 15.h3 tt:lgxeS
problems.
16.LDe4 (16.�xg7 0-0-0) 16 . . . h4
After the text, 13 . . . tt:lc5 14.b4
17.�xg7 0-0-0 or 14.tt:le4 tt:lgxeS
tt:lxd3 15.cxd3 l'l:fc8 is a solid, but
15.�g3 fS 16.�xg7 0-0-0 17.�xe7
passive approach. A more straight­
fxe4 18.tt:lxe5 exd3 = .
forward attempt is 13 . . . b4 14.tt:la4
12 . . . 0-0 (12 . . . tt:l c5 13.b4 tt:lxd3
(14.tt:ld1 tt:lcS) 14 . . . d5 15.e5 tt:le4, but
14.cxd3 0-0 15J'l:c1 �d7 16.a3 l'l:fc8 =)
White retains some pull with 16.c4
13.b4 (13.tt:ld1 tt:lcS 14.tt:lf2 l'l:ac8 ;
g6 17.ixe4 dxe4 18.b3.
13.e5 dxeS 14.fxe5 tt:lg4 15.�g3 fS+)
13 . . . tt:lb6 followed up by . . . tt:lc4. I like the flexible 13 . . . l'l:fe8 ! ? . It
prepares . . . eS so White should antici­
9 . . . J.e7 10.�f3 pate it with 14.e5 dxeS 15.fxe5 LDxe5
16.if4 id6 17.tt:ldxb5 axb5 18.tt:lxb5
10 .�e2 0-0 11.id2 tt:lcS is con­ tt:lxd3 19.LDxc7 tt:lxf4 20 .LDxe8 tt:lxh3
sidered in line B 2 . 2 1 .tt:lxd6 ic6 22 .l'l:e3 lDgS 23.l'l:c3
l'l:a6 24.b4 id7. The long forced se­
1 0 . . . b5 ll.J.d2 J.b7 12.gael quence is over. In such sharp unbal­
anced positions, it is easier to play
Black has counterplay after with the two minor pieces.
12 .b4 0-0 13.a4 bxa4 14.l'l:xa4 l'l:fc8
15.l'l:aa1 tt:lb6 16.�h3 tt:lc4.
If White retreats his knight to b3 c. 7.f4 b5
now, or on the next move, we follow
the main scheme with . . . 0-0, . . . l'l:e8.

12 . . . 0-0

7 . . . d6 might turn into a loss

of tempo if Black later plays . . . b4
followed up by . . . d6-d5. It also rules

107
 Part 3

out the active development . . . .icS.
The only drawback of 7 . . . b5 is that
it allows e4-e5, but we'll see that
the arising positions are entirely ac­
ceptable.
8.e5
13 .ttlg3 tLlcS 14 ..if3 g6 15.a3 hS, in­
tending to meet 16.f5 by 16 . . . h4
17.fxg6 hxg3 (17 ... fxg6 18.ttlge2oo)
18.gxf7+ 'it>d8 19.�e1 �xh 2 + 2 0 .'it>g1
�h8 2 1.�xg3oo. However, it is more
natural to complete development
with :

Alternatives may lead to the ll . . . .icS 1 2 .'it>h1 0-0 13 .�d3 fS

Taimanov: 14 . .if3 with mutual chances.

a) 8.'�f3 .ib7 9 . .ie3 ttlc6 trans­

poses to the Taimanov with f4 - see
Part 4.
Only 9.g4 is of independent sig­
nificance. The game may conti­
nue 9 . . . ttlc6 ! (9 . . . b4 10.ttlce2 dS
ll.eS ttle4 12 .f5 ! is rather unclear:
12 . . .exf5 13.e6 .id6) 10 . .ie3 (10.
ttlxc6 �xc6 ll.gS .ib4 12 ..id2 .ixc3
13 ..ixc3 ttlxe4 14 . .ixg7 �g8 15 . .id4 In the game Edouard-Iturriza­
tLlxgS) lO . . . hS which is again a ga, Benasque 2 0 0 9 , Black even in­
Taimanov side line. tercepted the initiative after 14 . . . g6
b) for 8.�e2 .tb7, see Part 4, 15.a3 �c8 16.i.d2 bxa3 17.bxa3 ttlb6
game 26 An.Martin-Delchev, Be­ 18 . .ixb7 �xb7, but 14 . . . a5 15.�b1
nasque 2 013. ttla6 also deserves attention .

8 . . . b4 9.�a4 ! ? b) 9.ttlcb5 axbS 10.exf6 gxf6

ll.ttlxbS (11.�e2 .tcS) l l . . . �b6
a) 9.ttlce2 ttld5 10 . .te4 .tb7 is not 1 2 .�f3 i.b7! ? (12 . . . �a5 13 .i.e3 i.cS=)
popular. 13 .i.e3 �xbS.

At first White tried 11.�d3, but c) 9.ttle4 ttlxe4 10 ..ixe4 i.b7

maybe Iordachescu's suggestion 11.�f3 .ixe4 12 .�xe4 ttlc6 13.ttlxc6
ll . . . d6 (ll . . . .te7! ? 1 2 . 0-0 ttlc6=)
12.exd6 .ixd6 13.f5 ttld7 14.fxe6 tLlcS
15.�f3 0-0- 0 ! scared him. Indeed,
Black has serious initiative. Lately,
White tested:
11. 0-0 which is aimed against
ll . . . d6, e.g. 12.exd6 .ixd6 13.f5 eS
-
14.ttlb3±. Instead, Black may ex­
periment with ll . . . aS 1 2 .'it>h1 ttla6

108
 The Third Rank Set-Up
This position is equal :
13 . . . �b7 (hoping to keep an asym­
metric pawn structure with . . J'l:c8)
14.ie3 (14.�f3 dxc6 15.ie3 ie7
16.0-0 = ; 14.c4 bxc3 15.bxc3 ic5
16J'%b1 �xc6 17.�xc6 dxc6 18.'it>e2
0-0-0=) l"!c8 15.0-0-0 �xc6 16.'Wxc6
l"1xc6 17.l"!d3 ie7 18.l"!hd1 l"!c7 19 .id4
f6 with typical counterplay;
13 . . . �xc6 14.�xc6 dxc6 15.'it>e2
0-0-0 = ;
13 . . . dxc6 14.ie3 ie7 15. 0-0-0
0-0 16.�c4 c5 17.l"!d2 'Wc6 18.l"!hd1
l"1fe8= . Black has counterplay on the
queenside.
9 ••. �d5 10 .i.e4 ib7 11.0-0
ll . .• �c3 !
I n practice, Black has been
struggling so far. He has tried 11 . . .
ie7? 12 .it)f5± and :
ll . . . g6
This is Houdini's first line at
depth 23. Rublevsky also put his
fate in it against Bologan. In my
opinion, such a development is too
slow. I tried to refute it by opening
the queenside wit 1 2 . c4, but Black
holds on after 12 . . . 'Wxc4 13.b3 �c7
14.'Wf3 it)c6 15.ib2 l"!c8 16.ixd5
it)xd4 17.ixd4 ixd5 18.'Wf2 'Wc6
19.tlJb6 l"!c7 2 0 .ie3 �b7=. The most
unpleasant retort is probably:
12.l"!f2 ! (preparing to occupy the
c-file by c4, l"!c2) 12 . . . ie7 13.ixd5
ixd5 14.c4 bxc3 15.it)xc3 ib7
16.ie3 it)c6 17.l"!c1 tlJxd4 18.'Wxd4.
White has lasting pressure on the
queenside, for example : 18 . . . ic6
19.it)a4 0-0 2 0 .tlJb6 l"!ab8 2 1.l"!fc2
�d8 2 2 .l"!d2 l"!b7 23.b3.
It is better . to trade quickly
White's active pieces:
1 2 . �xc3
Or 12 .bxc3 ixe4 13.cxb4 it)c6
14.c3 it)xd4 15.'Wxd4 ixg2 16.'it>xg2
'Wc6+ 17.l"!f3 �xa4oo.
12 . . . bxc3 13.ef3
13.ixb7 cxb2 14.ixb2 (14.l"!b1
bxc1� 15.�xc1 l"!a7 16.if3 ic5
17.l"!d1 0-0=) 14 . . . �xb7 15.l"!b1 it)c6
16.'it>h1 (16.l"!f3 ie7 17.l"!b3 �c7
18.it)xc6 �xc6 19.id4 0-0=) 16 . . .
ie7 17.f5 (17.ia3 �c7) looks dan­
gerous, but Black has sufficient de­
fence :
109
 Part 3

17 . . . ll:\xd4 18.hd4 �e4 19.�d3

(19.f6 gxf6 20.exf6 l:'lg8 2 l .�d2
i.d8oo) 19 . . . �xd3 20.cxd3 with
equal endgame:
20 . . . h5 2l.g3 l:'lc8 2 2 .l:'lb6 0-0
23.l:'lxa6 l:'la8 24.l:'lxa8 l:'lxa8 =
25.fxe6 dxe6 26.l:'lal l:'la3 27.i.b2?
l:'lxd3 28.a4 i.b4;
20 . . . exf5 2l.l:'lxf5 0-0 2 2 . l:'lb7
l:'lfd8 23.l:'lfl <;!{f8 = .
Black has no problems here:
1 3 .b:e4
••• 14.�xe4 cxb2 19.i.e3 l:'ld5 2 0 .c4 ( 2 0 . l:'ltbl i.c5=)
15.hb2 ll:\c6 16.ll:\xc6 �xc6 2 0 ... l:'ld3 2 l . <;!;>f2 i.e7 2 2 .l:'ltbl l:'lhd8
17.�xc6 dxc6 18.i.d4 0-0-0 23.l:'lb3 l:'lxb3 24.axb3 l:'ld3 = .

110
Pa rt 3 . The Third Ra n k Set- U p

Com p l ete G a m es

1 0. Vatter-Zakhartsov
Bad Wiessee, 04. 1 1 .201 1

l.e4 c5 2 . �f3 e6 3.d4 cxd4

4.�xd4 �c6 5.� c3 ftc7 6.J.e3 a6
7 .i.d3 �f6 8. 0-0 �e5 9.h3 J.c5
.

lO.chhl d6 11.f4 �g6 12 .ftel J.d7

It is logical to open the g-file for

the rooks instead of playing 25 . . .
g6, but why not take o n f6 with the
same goal. The endgame after 25 . . .
gxf6 ! ? 26.lt:Jxf6 Wxg2+ i s difficult
for White since his knights have not
13.f5 �e5 14.fth4 0-0 15.gf3 any stable outposts. In this line the
exf5 16.exf5 �xf3 17.gxf3 J.c6 pawn is probably better on b7 than
ts.ggt J.xf3+ 19.�xf3 .txe3 on b5 - Black has no weaknesses.
20.�xf6 hgl 2 1.�g5 ftc5
22.�xgl gae8 2 6.ftf3?

This is the same position from 26.'1Wg5 was the only move. Then
my main line, only the pawn is on 26 . . . E:e5 27.fxg7+ E:xg7 28 .'1Wd8+
b7 instead of b5. E:g8 29.ftf6+ is a draw so Black
should play 26 . . . 1Wxg5 27.lt:Jxg5 gxf6
23.�e4 �d5 24.ftg2 chh8 28.lt:Jxf7+ @g7 29.lt:Jxd6 E:el. With­
25.f6 out the d-pawn, Black can hardly

111
 Part 3

win this endgame though. 'ecS, but Carlsen's play is nonethe­

less impressive.
26 J:!e6
•. 27.fxg7+ gxg7
28.�e2 21 ••. 'ec6 2 2 . �e4

This position is won for Black. 2 2 .�e4 \Wc5 23 .�a8 fails to 23 ...
Strongest is 28 . . . \Wxa2, but Zakhart­ .ie3.
sov's decision to leave the queen in
the centre is more practical. 22 ••• f6 23.'exgl gae8?

In The Safest Sicilian, I pro­

28 \We5 29.'ef4 'exb2
posed to centralise the queen with
•••

30.<.!.>h2 \Wxa2 31.�xd6 'ed5

32.�xf7+ ? gxf7 0-1
1 1 . Carlsen-Vachier Lag rave
Cap d' Agde 2006
l.e4 c5 2.�£3 e6 3. t0c3 t0c6
4.d4 cxd4 5.�xd4 'ec7 6 . .ie3 a6
7 .id3 �f6 8. 0-0 tOeS 9.h3 .ic5
•
23 . . . 'ed5 ! White should defend the
f5-pawn with 24.ll'ld4 !'1£7 25.ll'le6
l'!e8 2 6.\Wg2 <.!.>h8 and Black is some­
what better. The modern under­
standing is that Black should not
weaken the e6-square.
24.Vd4 d5 25.ltlc5 ge7
26.<.!.>g2 g6 27.t0e6 gxe6 28.fxe6
'exe6 29.a4 !

lO.<.!.>hl d6 11.f4 �g6 12.'eel 0-0

13.f5 �e5 14.'eh4 b5 15.gf3 exf5
16.exf5 �xf3 17.gxf3 .ib7 18.ggl
.ix£3+ 19.�xf3 .ixe3 20.'exf6
.ixgl 2 1.�g5

Black's queenside pawns are

weak, his queen is passive. Still,
nothing terrible has happened yet.

29 bxa4 30.'exa4 gbs 3 1 .b3

•••

gb6 3 2.Va5 <.!.>g7 33.c!Lid4 \Wd6

This was one of the first games
that reached this crazy position. I
mention it since it gives an excellent
idea of White's possibilities. Later
Black found improvements, as 2 1 . . .
34.c3 h5?
This natural move turns out to
be dubious. 34 . . . 'it>f7 35.b4 fS was
a better set-up. White stands im-

112
 The Third Rank Set-Up
pressively, but everything in Black's
camp is protected.
35.h4 <tt> h 6 36.b4 gbs?
Humans hate to stay passive­
ly and too often lose because they
strive to force events instead of
miserably repeating aimless moves.
Now 36 ... c;f{g7 was holding the po­
sition, but Vachier-Lagrave had a ·
counter-attack in mind.
37.ti'a6 ti'f4? 38.c!i�f5!
A beautiful mating combination
out of the blue. Black resigned.
1 2. Zaragatski -Delchev
Bad Wiessee 2005
l.e4 c5 2.�f3 e6 3.d4 cxd4
4. c!Dxd4 �c6 5 . � c3 ti'c7 6 . .ie3 a6
7..id3 �f6 8.0-0 �xd4 9.hd4
1c5 10.hc5 ti'xc5 ll.�a4 Wc7
12 .c4 d6 13.gcl 0-0 14.c5 .id7
15.cxd6 ti'xd6 16.�c5 .ic6
17.ti'e2 gac8
Black holds the position thanks
to the tactical trick 18 .e5 ti'dS 19.f4
ti'd4 20.ti'f2 Wxf2 2 1 . c;f{xf2 = , J.
Moreno-Tregubov, France 2000.
This endgame might be White's
best option though. Zaragatski re­
lies on the symmetric structure and
avoids taking any committing deci­
sions, but this is always a dubious
approach .
ts.gfdt ti'e5 t9.f3 gfd8
2 0.Wf2 h5!
In such pawn structures, Black
has two major plans. One is to push
gS. This is a good defensive set-up.
· I chose to gain space on the kingside
by . . . h5-h4. White should have pre­
vented it by playing h4 himself.
2t.Afl Wgs 2 2 .b4? ! h4 23.gds
gds 24. �h3 �hs
It is amazing how quickly one
can get into trouble in a seeming­
ly simple and safe position. White
would not have had problems with­
out rooks, but he cannot exchange
them.
25.gc5 e5 26.�d4 �f4 27. <tt> h t
.ia4 2 8 . �f5 gdl 29.g3 �d3
3o. gc8+ <tt> h 7 3 1 . ti'e2 �c1 3 2 . ti'f2
.ib5 33. �e3 �d3 34.Wgl gel
113
 Part 3

35.a4 .ixa4 36.f4 exf4 37.c!Og2 19 axb5+ 2 0 . �c2 b4 2 1 . c!Oe2

•••

�at 38.gxf4 c!Oxf4 39.�c5 Wg4 \&b6 ! 2 2 .�dl? ! (The rook is bet­
40.c!Oxf4 9f3+ 0-1 ter on the open file : 2 2 . �fcl+.) 22••.

�a5 ! 23.�cd2 .ic6 24 .ibl �f5 !+

•

25.Wg3 .ie4 26.�d6 Wa5 27.We3

hb1 28.�xbl c!Od5 29 .Wd4 Wxa2
1 3. Varga-Aimasi 30.�el h6
Budapest 2004

l.e4 c5 2 . c!Of3 c!Oc6 3.d4 cxd4

4.c!Oxd4 9c7 5.c!Oc3 e6 6 .ie3 c!Of6
•

7 .id3 a6 8.0-0 c!Oe5 9.h3 .ic5

•

10.c!Oa4 .ia7 ll.c4 d6 12 .�kl .id7

13.c!Oc3 0-0 14.b3 9a5 15.9e2
�fe8=

Black could have taken anoth­

er pawn : 30 . . . �xb3 ! since 31.tt:lg3
�gS 3 2 . tt:le4 �xg2 ! 33.'t!lxg2 tt:lf4+
34.'t!lgl �f3 35.tt:lf6 gxf6 36.\&e4
tt:lh3 37.'t!lh2 tt:lgS 38.�e3 �c8-+
loses at once. It is incredible, but
Black went on to lose from the dia­
gram position. 1 : 0 on move 76.
16.c!Of3 J.xe3 17.Wxe3 c!Oxf3 +
18.9xf3 b5?
1 4. Onischu k-Salov
A good idea at a wrong time. El ista 1 998
Black is still slightly unco-ordinat­
l.e4 c5 2.tt:lf3 e6 3.d4 cxd4
ed, so 18 .. J�ad8 ! ? would have been
4.tt:lxd4 tt:lc6 5 . c!Oc3 \&c7 6 .ie3 a6
better.
•

7 .id3 tt:lf6 8 . 0-0 �e5 9.h3 .ic5

•

lO.'t!lhl d6 11.f4 c!Og6 12 .\&d2 0-0

19.cxb5?
13.�ael b5 14.a3

White misses his tactical chance

19.e5 ! dxeS 20 .tt:le4 with a very an­
noying initiative. The c-pawn sud­
denly becomes dreadful. Nei­
ther 20 . . . \1*/dS 21.c5±, nor 20 . . . tt:le4
21 .\&e4 g6 22 .\1*/eS \&a2 23 . .ie4 �ac8
24.hb7 �b8 25.\&c7 bxc4 26.bxc4±
suits Black.

114
 The Third Rank Set-Up
14 . . . e5 lS.fxeS dxeS 16.l!Jf5 .be3
17.l!Jxe3 is slightly better for White.
14 . . . .ib7 fails to lS.fS so Black
defends the e-pawn.
14 . . . �b6 ! ? is a good alterna­
tive : lS.eS l!JhS 16.�f2 l!Jhf4 17.b4
ix:d4 18 . .bd4 l!Jxd3 19.cxd3 �c6
20.ed6co.
15. �b3 .ixe3 16.ti'xe3 ib7
17.e5
This is the main plan. 17.f5 exfS
18 J§:f5 dS or 17. l!Je2 eS 18.f5 l!Je7
transform the centre in Black's fa­
vour.
17 dxe5
••• 18 .fxe5 �d7
19 . .ixg6 fxg6 20.!:ie2 !
A flexible move which makes
way for the other rook to dl.
A critical moment of the game.
Salov plays very well endgames
so he gladly allows the trade of
queens. I think that 2 0 .. J'!ac8 main­
tains a dynamic balance without
much risk, because Black's pawn
structure on the kingside is nearly
unassailable: 21.l!Jd4 ( 2 1.E:dl l!Jb6+;
21.�d4 l!Jb6 22.lt:ld2 E:cd8=) 21 . . .
liJb6 22 .l!Jdl = .
21.ti'c5 !:iac8 2 2 .ti'xc7 !:ixc7
23.�a5 .ia8 24. �gl lM8 25.!:idl
gxdl+ 26.�xdl �f7 27.�f2 �c4
28.�xc4 !:ixc4 29.c3 g5 30.!:id2
�e7 3 1.!:id6 J.c6 3 2 .!:id4 J.d5
33.�d3
Sooner or later White should
swap the rooks ant the arising end­
game should be · drawn : 33.E:xc4
bxc4 34.l!Jdl lt>d7 35.l!Je3 lt>c6
36.Wf2 lt>cS = . In the game, White
has been stubbornly avoiding this
decision until it became too late.
33 a5 34.\t>f2 lt>d7 35.�el
••.
'it>c6 36. �e3 h6 37.g3 'it>b6
38. �d3 !:ic8 39.a4 !:if8 40.axb5
'it>xb5 41.�e2 !:if5 42.g4 !:if7
43.�el g{4+ 44.'it>e3 !:if1 45.�c2
!:if3 + 46.�d2 .ib3 47.!:ib4+ axb4
48.�d4 + 'it>c4 49.�xf3 bxc3 +
50.bxc3 �d5 51.�e3 J.c4 52 .h4
gxh4 53. �xh4 �xe5 54. �f3+
'it>d5 55.�d2 J.b5 56.�f3 J.a4
57.g5 h5 58. �f4 J.c2 59.�e5 J.f5
60.�f3 �c4 61.�h4 .ig4 62 .g6
�xc3 63.�e5 'it>d3 64.'it>d6 'it>e4
65. �e7 �f4 0-1
1 5. Kotron ias-De Ia Riva
B led ol 2002
l.e4 c5 2 . �f3 �c6 3 . d4 cxd4
4.�xd4 e6 5. �c3 ti'c7 6.J.e3 a6
115
 Part 3

7 .td3 �f6 8.0-0 �e5 9.h3 J.c5

• 2 0 . l:U'3
lO.<.!>hl d6 ll.f4 �g6 12.£5 �e5
13.�ce2 exf5 14.exf5 0-0 15 .tg5 •

�ed7

2 0 . . . .lf8 ! ?

Black anticipates l'!g3 - 20...

16.�el .tb7? 2 1.l'!g3 J.d6 ( 2 1 . . .ll:Jh5 2 2 .�h5
J.d4 23.l'!g4 i.b2 24.c3�) 2 2 .l'!g7!
Or 16.c3 dS ! , in order to keep the �g7 23.i.h6 �h8 24.ll:Jf3 with a
bishop on the f8-a3 diagonal in the crushing attack.
event of b4.
21 .lf4 �b6 2 2 .J.e3 ( 2 2 .l'!g3
•

16 . . . h6 17 .td2 d5 18.ti'h4
•
White stakes all on the kingside
attack. A more positional approach
is to trade dark-squared bishops,
trying to underline the awkward
stand of the bishop on c8. Still,
he has enough weaknesses on the
queenside to worry about: 18.i.f4
i.d6 (18 . . . ll:Je5 19.�h4 �e7 2 0 . l'!ael)
19.�g3 i.xf4 2 0 .�xf4 �b6 2 1 . ll:lb3
l'!e8 2 2 .l'!ael lLleS 23.ll:Jed4 ll:Jxd3
24.l'!xe8 ll:Jxe8 25.cxd3 i.d7=.
lt?h8) 22 . . . ti'c7 23.�f4? ! (23.J.f4
looks more realistic.) 23 . . . �c5
24.b4? �ce4
Black is aiming to prevent at any
cost an attack against his king. He
could have tried to seize the initia­
tive by 24 . . . ll:Jxd3 25.cxd3 (25.ll:Jxd3
ll:Je4 26.�g4 �c3 27.l'!e2 aS) 25 . . .
�c3 26.ll:Jde2 �xb4 27.i.d4 ll:Jh7
28.ll:Jd5 �d2 t.
25.tl:�d5 �d5 26.i.e4 J.b7
27.a3 :Bad8 draw.

Black must be very careful about

his king. e.g. 19 . . . i.e7 20.ll:Jc3 ll:JcS
is likely to lose to 21.i.xh6 gxh6
22 .�xh6 ll:Jxd3 23.cxd3�. Instead,
Black has an interesting way to re­
group with 19 . . . ll:Je4 ! ? 2 0 .i.f4 �b6,
for example 2 1.i.xe4 l'!xe4 2 2 .ll:lb3 = .
116
 The Third Rank Set-Up
Ka n G a m es
1 6. C u bas-Hel lsten
Buenos Aires (7), 09. 1 0 .2006
l.e4 c5 2.c!Llf3 e6 3.d4 cxd4
4. c!Llxd4 a6 5 . c!Ll c3 ffc7 6.J.d3
�f6 7.0-0 d6 8 . a4 b6 9.f4 J.b7
lO.YlYf3 c!Llbd7 11 .id2 J.e7 12 .gael
•
12 ••• c!Llc5
The idea of killing the most dan­
gerous white piece, the d3-bish­
op, cannot be wrong. I believe that
the ensuing position, where Black
has the bishop pair and no appar­
ent weaknesses, is roughly equal.
Still, 12 . . . 0-0 13.�h3 lLl cS is per­
haps the more accurate move or­
der. Black has made an obligatory
move while in the symmetric pawn
structure after . . . lLlxd3 the queen is
in no way better placed on h3 than
on f3. It even hampers a possible
plan with g2-g4-g5, h2-h4. Black
only has to watch out for sacrifices
on e6. The critical question is: can
Black exchange the d3-bishop at
all? Let's check 14.e5 lL\xd3 15.exf6
ttJxel 16.fxe7 .bg2 17.exf8�+ l:!xf8
18.�g3 .bfl 19 . .bel .ic4 2 0 .b3
.idS 2 1 .lLlxdS exdS 2 2 .lLlfS f6 = . At
a second glance, 14 . . . dxe5 ! ? lS.fxeS
lL\xd3 16.exf6 lL\xel 17.fxe7 �xe7
18.l:!xeloo may be even more attrac­
tive. So let's focus on:
14.b4 lLlxd3 15.cxd3 l:!fc8
Black's b7-bishop is current­
ly without prospects. On the other
hand, the c3-knight is also not par­
ticularly impressive. The evaluation
of this position depends on White's
ability to gain space on the kingside.
If he achieves g4-g5, he would get
some initiative. However, 16 .g4
would let in the blow 16 . . . d5 17.e5
.bb4. The inclusion of 16.<it>hl �d7
does not enable 17.g4 dS 18.e5 due
to 18 . . . lL\e4 ! . We see that the con­
trol of b4 is vital for White's setup.
Therefore, the most logical move is:
16.l:!bl �d7
This is the ideal place for Black's
queen in this structure - it is eye­
ing the a4-pawn while protecting
e6 and c6. Again, 17.g4 lL\e8 18 .g5
is well met by 18 . . . d5 ! 19.e5 g6 and
White cannot generate any active
idea. On contrast, Black has plenty
of play. For instance, 2 0 . l:!fcl .ic6,
when White is unable to keep the
queenside closed anymore because
of the hanging pawns on the fourth
117

rank. 20J:!b3 lt:lg7 2 1.!%a1 is useless
due to 2 l . . . lt:'lf5 ! 2 2 .lt:'lxf5 exfS 23.d4
!%c4 24.lt:'le2 .ic6 - the dozing bish­
op enters the play with a great ef­
fect. Obviously, White should post­
pone active plans:
17.!%fc1 lt:'le8 18.'it>h1 .if6
We see a reliable set-up for
Black. After 19 . .ie3 bS ! 2 0 .a5 (20.
axbS axbS 21.lt:'ldxb5? .ia6), he can
activate the b7-bishop by 20 . . . hd4
21.hd4 fS, or delay this plan in fa­
vour of doubling on the c-file first.
To take stock, 12 . . . 0-0 looks fine
for Black.
13.b4 c!Dxd3 14.cxd3
15.g4
In my explanations to move 12,
I recommended the set-up with
118
Part 3
. . .'?9d7, . . . lt:le8. It is quite good here,
too, despite the rook being still on
f8 : 15 . . . lt:le8 ! ? , when 16.g5? would
be a mistake due to 16 . . . d5. Af­
ter 16.!%c1 '?9d7, White also lacks a
good continuation. Perhaps White
should defend the b4-pawn : 16.!%b1
.if6 17.lt:lce2 '?9d7 with sufficient
counterplay.
Black can even unbalance the po­
sition by the interesting queen sac­
.
rifice: 15 . . . d5 ! ? 16.e5 hb4 17.lt:lcb5
�cS 18 .!%c1 axbS 19.!%xc5 hcS
20 . .ie3 lt:ld7 21.axb5 !%a3 2 2 .!%d1
!%fa8oo.
16.!%cl '?9d8 17.lt:lce2 !%c8 ? !
Hellsten i s constantly avoiding
concrete decisions. 17 . . . a5 18.b5
lt:lcS 19.lt:'lc6 hc6 2 0 .bxc6 lt:lxa4
2 1.d4 bS would have been much
more enterprising.
18.g5 !%xc1 19.!%xcl g6 20 .h4
0-0
20 . . . e5? !
Despite his mundane play so
far, Black still has a good game.
He had to secure the cS-square for
his knight by 2 0 . . . a5 ! . After 2 1 .b5?!
lt:lcS 2 2 . lt:'lc3 f6, White would have
 The Third Rank Set-Up

too many pawns hanging. Instead,

Black hands the opponent the d5-
square for his knight.

21 .!Llc6 hc6 2 2 . gxc6 exf4

•

23.d4 ge8 ? !

It i s amazing how passively Hell­

sten treats this position . He had to
seek counterplay with 23 . . . f6. After 9.i.d2
the text, White is already much bet­
ter and he went on to win (not with­ By delaying f4, I discourage an
out mistakes though). early . . . b5 which I would attack by
a4. However, an ever more sophisti­
24 .!Llxf4 .!Llb8 25.gcl .tf8
•
cated move order is probably 9.<.!?hl
26 .!Lld5 �g7 27.gc7 gf8 28.YU4
•
which is also aimed against 9 ... g6.
ixd4+ 29 .�e3 �e5 30.Y5g4 + -
9 .te7 10.<.t>hl! .!Llc5 11.f4 b5?
•..

This move should be played after

�ael, e.g. 1 1 . . . 0-0 12.�ael b5. Now I
could have achieved a decisive at­
tack with 12 .e5 lLJfd7 (12 . . . dxe5
13 .fxe5 .!Llfd7 14.�xt7 <.t>xt7 15.YlYh5+
g6 16.hg6+ hxg6 17.YlYxh8 lLJf8
18.�fl+ <.t>e8 19.YlYg7 '\1;;'d 7 2 0 .i.g5
i.b7 2 1.�t7) 13.lLJd5 ! ! , but I was fol­
lowing the general principle which
30 b 5 31.a5 (31.axb5 axb5
assumes that . . . b5 should be ex­
•.•

32.h5+-) 3 1 Y5e8 3 2 . .!Lle7+

ploited by a break on the queenside.
• • •

�h8 33.�f4 �d4+ 34.<.!?g2 YlYd8

Here is another example from my
35 .!Lld5 �g7 36.�d2? (36.h5 ! )
practice:
•

36 f5?? (36 . . . Y5e8) 37.gxf6 �6

.•.

38.i.h6 ggs 39.Y5e6 i.e5 40 .YlYt7 Delchev-Rasu lov

1g7 41.hg7+ 1-0 Edirne 201 3

17 Delchev-Castellanos Rodriguez
Leon 06. 1 1 .201 2

l.e4 c5 2 .!Llf3 e6 3 . d4 cxd4

•

4 . .!Llxd4 a6 5 .!Llc3 YlYc7 6.�d3 .!Llf6

•

7.0-0 d6 8 .Y5e2 .!Llbd7

119

10.b4 ! i.e7 ll.a4 bxa4 12J'!xa4±
�c8 13.tt'le5 �c7 14.t['lc4 d6 15.lDa5
i.c8 16J�a3 tt'lf6 17.tt'la4 tt'lfd7 18J'k3
�d8 19.i.e3 i.f6 20J'k4 0-0 2 1.f4
�e7 22 J'!el e5 23.f5 i.g5 24.i.f2
�d8 25.i.fl. Black is totally stran­
gled and he cannot prevent 26.�d5.
12 .b4 tt'lxd3 13.cxd3
14JUcl !
I seize the initiative. Black
should be able to hold on, but he
must endure a lasting pull.
14 �d8 15.a4! bxa4 16.tt'lxa4
..•
!:!c8 17.!:!xc8 ti'xc8 18.!:!cl ti'd7
18 . . . �b8 ! 19 .i.e3 0-0 2 0 .lDb6
�e8 was preferable.
19.tt'lb6 fid8 20.lLlc4 0-0
21.lLla5
Perhaps 2 1.i.e3 was more un­
pleasant. Black cannot find good
places for his pieces.
21 • •• ti'd7 2 2 .lLlxb7? !
In this pawn structure, Black's
only concern is his light-squared
120
Part 3
bishop which has not bright pros­
pects. I decided to exchange it in
order to gain full control over the c­
file, but it does not bring substan­
tial dividends. Instead, I should
have produced a second weakness
in the opponent's camp by 2 2 .f5 ! ,
when 2 2 . . . e 5 23.tt'lxb7 ti'xb7 24.lDc6
i.b7 d5 25.i.g5 would give me a clear
positional edge.
2 2 ti'xb7 23.lLlc6 !:!e8 24 .tel
• •• •
.idS?!
24 ... i.f8 ! 25.�c2 d5 ! 26.e5 tt'lg4
would have equalised. After the
text, Black cannot easily unpin his
knight:
25.i.h4 ! e5 2 6.fxe5
It was better to play 2 6.ti'f3 im­
mediately. The f4-pawn is impor­
tant since it prevents . . . h6, . . . g5
26 dxe5 27.ti'f3 ti'd7 28.h3
•••
!:!e6? ! (28 . . . h6 ! ) 29. lLlxd8 ti'xd8
30.d4 ! !
I t turns out that 30 . . . exd4 31.e5
�xeS 32.i.xf6 gxf6 33 .�g4+ �g5
34.�c8 wins a rook.
 The Third Rank Set-Up

30 . . . fi'd7 31.d5 gb6 3 2.ti'c3 \Wxc2 23.�xd6+ <Jlc7 24.lt:lb5+ with

'fYe8 33 .bf6 gxf6 34.fi'c7 h5
• a perpetual (24 . . . 'it>b8? 25.\Wd4). It
35.d6 1-0 is easy to understand White's wish
to kill the hS-knight.

ts ••• b4 t6 .ixh5 gxhs

•

1 8. Kerekes-A. Kovacs
H u ngary, 09. 1 2.2007

l.e4 c5 2.�f3 e6 3 . d4 cxd4

4.ctlxd4 a6 5 .id3 <tlf6 6.0-0 d6
•

7.ie3 �bd7 8 . � c3 ti'c7 9 . s!?hl

1e7 10.f4 b5 11.fi'f3 .ib7 1 2 . gael
h5

Black has active pieces. After

17.lt:lb1, he could evacuate the king
from the centre by 17 . . . 'it>f8 and fol­
low up by . . . g6 or . . . gh6.

17.� a4 �f6 18.c3 fi'a5?

This mistake mars the nice

White's set-up is entirely g4-g5
oriented. Now he has to revise his
strategy and think up a way to ex­
ploit the weakness of the h-pawn.
13 .igl h4 14. ti'h3 �h5 15 .ie2
• •
game. The simple 18 . . . lt:lxe4 would
have ensured a solid edge thanks to
the threat of . . . gh6, . . . lt:lg3+. White
could have now levelled the chances
with 19.ltlb6 ! gbs 20.cxb4.

19.b3 �xe4 20.cxb4 ti'd5

This is a difficult position for 2 1 . �b6
both sides. Black has some tempo­
rary initiative, but he must play con­
crete chess to keep it. For instance,
15.a3 0-0-0 is very interesting (of
course, 15 . . . gcs is also possible, but
the fact that it does not threaten an­
ything worries me) . After 16.lt:lde2
tLlg3+ 17.lt:lxg3 hxg3 18.�xg3 ih4
19.�xg7 he1 2 0 .gxe1 fS, White is
lucky to find 2 1 .hb5 ! axbS 2 2 .lt:lxb5

121

21
�g3 + 22.hxg3 (22 .�xg3
• •.
hxg3+) 22 ti'xg2 + 23.�xg2
• •.
hxg3+ 24 .ih2 gxh2+ 25. �g1
•
gxg2 + 26.�h1 gb2 + 27.�g1
gb1# 0-1
1 9. Moen -Topalov
ECC Rhodes 20.1 0.201 3
1.e4 c5 2 . �£J e6 3.d4 cxd4
4.�xd4 a6 5.�c3 �c7 6 . .id3
ic5 7.�b3 ie7 8.f4 b5 9.a3 ib7
10.0-0 d6 ll.�h1 �d7 12 .id2
�gf6 13.ti'e2 0-0 14.�d4 gac8
15.gae1 �c5 16.b4 �xd3 17.cxd3
17 ••• tM7 18.gcl gfe8 19.�b3
g6
The engines like 19 . . . e5, but I
suppose that Topalov did not like
20 .fxe5 when the b7-bishop would
be biting on granite until the rest of
its days. The plans with .. .fS or . . . d5-
d4 are more flexible.
20.ie3 it'S 21.gc2 �h5
122
Part 3
2 2 .. .f5 is already a threat. 2 2 .g4
lt:Jg7 would only add oil into the fire
as . . .fS would gain in strength.
Moen decides to trade a cou­
ple of rooks along the c-file, but his
move weakens the f4-pawn and en­
•
ables another break in the centre :
2 2 .lUc1 d5 (22 . . . f5 was also
possible) 23.exd5 (23.e5 d4) 23 .•.
exd5 24. �f2 d4
A typical sacrifice on gener­
al considerations which unleashes
all Black's pieces. Concrete calcula­
tion shows that 24 . . . �f5 ! 2S.d4 �d3
26 . .id2 lt:Jg7 was very strong. The
knight arrives in the centre with a
decisive effect. Anyway, Moen opts
for the wrong capture on d4 and his
position goes down quickly.
25. �xd4? gcd8 26. �e4 .ixe4
27.dxe4 gxe4 2 8 . �b3 �e6
29.gc3 ges 30 .id2 �xf4 3 1 .�d4
•
gxd4 32.!xf4 gxf4 33.�xf4 .ih6
34.�xh6?? (34.'�fl .ixcl 35J!xcl
'\!!!fe 3 36.l'!al '\!!!f c 3 37.'\!!!f c l �xcl+
38.l'!xcl l'!e3+) 34 �e1+ 35.gxe1
•••
gxe1# 0-1
 Pa rt 4

The f4 System
The early f4 is coming into fashion in the Taimanov. White then takes on
c6, castles long and tries to maintain a firm grip on the centre by a3, :B:hel .
Against the Kan, this plan i s ineffective and i t i s rarely seen.

123
Pa rt 4. The f4 System

M a in I d eas

l.e4 c5 2 .!D f3 e 6 3.d4 cxd4

• the f4-pawn ensures this advance
4 . .!Dxd4 a6 5 .!Dc3 YlYc7 6.f4
• at any time. However, it is effective
only in connection with an attack on
h7. Therefore, the queen would be
best placed on h3. Another draw­
back of �e2 is that the queen takes
the natural retreat square of the
lt:Jc3 in the event of . . . b4.

8 .!Df6 9 . .ie3 (9.g4? ! lt:Jc6!

•••

10 . .ie3 hS) 9 . . . .!Dc6 1 0 .0-0-0 (10.

g4 hS ! )
White's last move did not devel­
op a piece so Black can spend time
on a pawn move, too :

6 ••• b5 7 .id3
•

The absence of a knight on c6

narrows White's safe path because
of the threat of . . . .ib4.

A plus for the Kan : easy to learn.
7 •••.ib7 s.ef3
s.ee2 lt:Jf6 is fine for Black, see
game 26 An.Martin-Delchev, Be­
nasque 2013. White commonly puts
his queen on e2 in order to support
e4-e5. That has no sense here since
The play has transposed t o a
minor line of the Taimanov. Thus
Black side-stepped the most un­
pleasant positions with tt:Jxc6 �xc6
where Black should practically lose
a tempo on retreating to c7 later.
A plus for the Kan : Black plays an
improved Taimanov.

124
 The f4 System

In the diagram position, Black

can simplify with 10 . . . lt:Jxd4 1l.i.xd4
icS or maintain the tension with :

10 b4 ll.c!i�ce2 lL! a5 1 2.g4 d5

•••

13.e5 lL!d7 14.'i!?b1 lL! c4 15 .Ac1 •

Bb6 16.ghfl .Ac5

Black has good counterplay on

the queenside thanks to . . . aS, . . . i.a6.
The Taimanov theory is much
more complex. White can get rid of
his hanging knight on d4 by taking
on c6, and then comfortably devel­
op the rest of his pieces.
White's main set-ups are:
1.e4 c5 2 . ltl f3 e6 3.d4 cxd4
4. tLlxd4 tLlc6 5.lL!c3 'ffc 7 6.f4 a6
We are familiar with this struc­
ture from the previous chapter.
· Commonly, Black swaps his knight
for the d3-bishop with 12 . . . lt:Jb4 and
gets a safe albeit passive position. I
advocate a sharper way which may
also lead to the same knight's ma­
noeuvre, but after White has com­
mitted himself with g4.
12 .•• 0-0 13.g4

If White refrains from the pawn

storm (13.'i!?hl), Black redeploys his
forces with . . . lt:Jd7, . . . i.f6. See game
22 Belikov-Zakhartsov, Alushta
2005.

13 ••• tLlb4 ! ? 14.g5

A critical position arises after

A. 7.tLlf3 .Ac5 8 .Ad3 b5 9.e5
• 14.e5 lt:Jxd3 15.exf6 lt:Jxb2 16.fxe7
d5 (9 .. .f5 ! ?). Eife8! 17.i.d4 l:!xe7

B. 7 .Ae3 b5 8 .id3 .ib7 9 .ltlb3

• •

�f6

9 . . . lt:Ja5 ! ? now or on the next

move is easier to play, but it gives
the enemy a free hand in the centre.

10.0-0 d6 n.'fff3 .Ae7 12.eh3

125
 Part 4

In my opinion, Black has good

practical chances here.

13 . . . .ib4 ! ? 14 . .id4 0-0oo 15 . .ixf6

gxf6 16.tt::l e 2 l'!fd8 17.c3 .ic5 with
double-edged game.

Planning a counter-attack with n . . . Wc7! ? 12 .Wd4 h5

. . . .idS, . . .f6. The play is double­
edged.

C. 7.J.e2 b5 8 .�xc6 Wxc6

9.Af3 Ab7 10.Ae3 (10 .e5 Wfc7
11.tt::l e4 tt::l h 6) 10 J�c8
••

This is meant to enable 13 . . . tt::l f6.

If White answers 13.h3, then we
can vacate the h5-square by 13 . . .
h 4 14.0-0 tt::l f6, see game 24 Gof­
shtein-Tregubov, Montpellier 1998.

My suggestion is to keep the D. 7.�xc6 Wxc6 8.J.d3 b5

queens by retreating to c7 (although 9.We2
. . . Wfc4 is by all means playable) and
lead out our dark-squared bishop to
b4 or c5. Plans with . . . d6 are passive.

ll.a3

An illustration of my previous
note is the variation l l.Wid3 Wfc7 ! ?
12.0-0-0 tt::l f6 13.'kt>b1 (13 .g4 b4
14.tt::l e 2 Wc4)

126
 The f4 System

This has been the main line in
the Taimanov with f4 lately. The
question where to develop
knight is still open. It is tempting
to put it on the more active square
f6, but then White will have eS. So
Black should play a few waiting
moves first, like 9....ib7!? 10..id2
I think that Black has good
chances here, but it is difficult to
the judge without practical tests.
The plan with ...lLle7 is not so de­
manding, but it offers White a tem­
porary initiative:
9 ....ic5 10 .id2 .ib7 11.0-0-0
•

l"lc8 ll.a3 .ie7 12.0-0-0 lLlf6. Now lLle7 12.a3! 0-0

critical is 13.g4! dS 14.exd5 lLlxdS
15.:B:hf1 lLlxc3 16..ixc3 0-0 17.g5
l"lfe8! 18 ..ixg7 @xg7 19.�h5 �h8
20.f5 exfS 2l.:B:xf5 �g6 22.'\1;!fg4 '\1;!fd6
23 ..ixb5 '\1;!fe6 24.'\1;!fd4+ @g8 25 ..id7
�c4 26..ixc8 '\1;!fxd4 27.:B:xd4 .ixc8
28.:B:e5 .ie6 29.:B:a4 .id6 30.:B:e2 @g7

A topical position. I analyse it in

detail in the annotations to game
21 Nepomniachtchi-Vitiugov, Nizh­
ny Novgorod 2013. I do not see
compelling reasons to avoid it with
Black.

Weapon of choice

The answer is very easy, because we have an unconditional winner - the

Kan move order is definitely preferable.
The f4-Kan theory is a subset of the Taimanov. Black plays ... lLlc6 only
after ...bS, ....ib7. Thus he can recapture on c6 by the bishop, avoiding the
loss of tempo after ...�c7xc6-c7. White's choice is narrowed to plans with
queenside castling and double-edged positions.

127
Pa rt 4. The f4 System

Ta i manov - Step by Step

l.e4 c5 2 .c!b £J e6 3.d4 cxd4 7.lbb3 bS 8 . .ie3 transposes to

4. c!Dxd4 c!Dc6 5 .!Dc3 fic7 6.f4
• line B.

A. 7.c!D£J .ic5

7 . . . b5 first is also possible, since

8 . .ie3 .ib7 9 . .id3 c!DaS suits Black,
for instance, 10.0-0 lbc4 ll . .id4
lbxb2 1 2 .lbxb5 axbS 13.hb2 lbf6.

8 . .id3

8 .e5 gives Black good counter­

6 ••• a6 ! play after 8 . . .f5 (8 . . . d5 ! ? ) 9 . .id3 bS
10 .fie2 fia7, Hector-Zvjaginsev,
6 . . . lbxd4 has come into fashion Berlin 1993.
lately. I do not like too much this
option because it allows White to 8 . . b5
.

castle long and obtain some initia­

tive, for instance, 7.fixd4 a6 8 . .ie3
bS 9. 0-0-0 .ib7 10 . .id3 l"lc8 1l.c;t>b1 ! .
See game 2 0 Yemelin-Rodshtein,
Skopje 2013.
Commonly, White trades knights
himself, but then we take by queen,
putting pressure on e4.

A. 7.lDf3 ; B. 7 . .ie3 ; C. 7 . .ie2 ; D.

7.lbxc6

128
 The f4 System
9.e5
9.�e2 tt::'l d 4 ! 10.tt::'l x d4 i.xd4 pro­
vides Black with time for complet­
ing development: ll.tt::'l d 1 (or ll . .id2
ib7 12 .e5 fS ! ) ll . . . tt::'l f6 12 .c3 .icS
13 .e5 tt::'l d S 14.tt::'l f2 .ib7 (14 . . . f5 ! ?)
15.tt::'l e 4 0-0 16.tt::'l xc5 �xeS 17 . .id2
l"!ac8 18 .vtff2 vtfxf2 + 19.'i!lxf2 f6t,
Dvoirys-Tregubov, Rostov 1993 .
9 . . . d5
9 .. .f5 is positionally more unbal­
anced, but it brings excellent results
after 10 .g4 tt::'l g e7 or 10 .vtfe2 �a7.
10.exd6
10 .�e2 �a7 underlines the
weakness of the diagonal g1-a7.
10 . . ..ixd6 11. 0-0 c!Df6 1 2 . c!D e4
�xe4 13 . .ixe4 .ib7 14.fiel 0-0+.
B. 7 .ie3 b5 8 .id3 .ib7
• •
This retreat looks like a second­
rate move which defies the principle
of quick development. Why do I pay
attention on it?
Actually, this set-up conceals a
lot of venom. White avoids the pin
from cS which would be strong af­
ter 0-0. He also escapes the trade of
dark-squared bishops after . . . tt::'l x d4
followed up by . . . .icS. This exchange
is commonly good for Black in the
Taimanov/ Kan since after . . . d6,
. .. eS, the remaining .ib7 is obvious­
ly more active than its white coun­
terpart on d3. Another argument in
favour of the exchange on d4 is that
it reduces White's attacking forces.
At first this · system brought
White good results as it deprived
Black of his thematic Taimanov
plans and put him in a Schevenin­
gen structure. With time, the sec­
ond players have learned to han­
dle this set-up, but we can still face
it in practice. Occasionally, White
chooses it in order to avoid long
theoretical lines and keep more dy­
namic in the position.
9.�f3 tt::'l f6 is the main line of our
Kan chapter.
9 . . . c!Df6 !
9 . . . tt::'l a5 ! ? i s easier t o play, but
this is true for both sides.
White commonly answers with
10.tt::'l x a5
10.0-0 ! ? tt::'l c4 1l . .id4 tt::'l b 2 1 2 .tt::'l b 5
abS 13 . .ib2 tt::'l f6 14.\We2 b4 15.'i!lh1
.ie7 16.e5 tt::'l dS 17.f5 g6oo
10 . . . '\WxaS ll.a3 !
Securing the position of the c3-
knight. On g3 it would be harassed
129
 Part 4
by the h-pawn: 11.0-0 b4 12.ltJe2
icS 13.ix:cS �xeS+ 14.�h1 tt:lf6
1S.tt:lg3 hS ! (One move later this
move would be impossible: 1S . . .
d6? 16.'1We2 ! h S 17.eS± h 4 18.tt:lfS.
Castling short under the fire of all
White's pieces is hopeless : 16 . . . 0-0
17.eS dxeS 18.fxeS tt:ld7 19.l'!ae1--+.)
16.�£3 (Here 16.e5 fails to 16...h4-+)
16 . . . h4 17.tt:le2 l'!b8 ! ? (protecting
the bishop which is hanging in the
variation 17 . . . h3 18.g4 dS 19.eS and
there is no 19 . . . d4) 18.l'!ad1 h3 19.g4
dS 20.tt:lg3 dxe4 2 1 .tt:lxe4 tt:lxe4
22 .ix:e4 he4 23.�xe4 l'!c8+.
11 . . .�c7 12. 0-0 icS 13.icS �cS
14.�h1 tt:lh6 1S.eS 0-0! 16.ie4 �b6
17.�d3 ix:e4 18.tt:lxe4 tt:lfS 19.l'!f3
l'!fb8 20 .l'!dl b4 2 l .a4 l'!a7= . White
does have a space advantage in this
line, but it should not bring him di­
vidends.
10.0-0
We see the result of 9 ... tt:lf6 -
White is forced to renounce plans
with queenside castling as lO .�fJ
allows lO . . . dS ! 11.exdS (Or 11.eS
d4 12.exf6 dc3+, Venci Popov­
Miladinovic, Korinthos 1999) 11 . . .
tt:lb4 12.0-0 tt:lbxdS 13.tt:lxdS hdS
130
14.�g3 hS ! � . Black can play h4 and
castle short, but he could also de­
velop the bishop to d6 and hide the
king to the other flank in the event
of �xg7.
lO .eS has been tested in only
one game : 1 0 . . �b4 11.tLlbS? ! (11.
tt:le4 tt:lxe4 12 . .ixe4 dS 13.exd6
ix:d6 14.0-0 0-0 is about equal, e.g.
1S.tt:lc5 .ixc5 16.hc5 tt:le7=.) 11 . . .
axbS 1 2 .exf6 gxf6 13.0-0 tt:l e 7 ! head­
ing for fS, Hector-Lindberg, Orebro
2013.
10 . . . d6
Again, lO . . . tt:laS ! ? is a viable op­
tion. It leads the game into a typical
Taimanov track: ll.tt:lxaS
Or 11.eS b4! 12.tLlbS axbS 13.exf6
tt:lc4 14.id4 gxf6 with counterplay
on the g-file.
l l . . . �aS 12.a3 !
12 .eS i s again premature due to
12 . . . b4 13.exf6 bxc3 14.fxg7 hg7
1S.b4 �dS 16.l'!f2 l'!g8 17.hh7 id4 !
18.hg8 he3 19.�dS idS 2 0 .ih7
fS.
12 . . . �c7!
We know already this manoeuvre
 The f4 System
from the comments to move 9 .
Black keeps all his options open.
He can switch over to Sheveningen
structures with . . . d6, or trade his
dark-squared bishop through cS.
Even . . .b5-b4 is worth considering.
In the event of 13.'i!?h1 hS ! ? (aim­
ing for . . . h4-h3) 14.�f3 lt:\g4 15 . .id4
icS, Black has some initiative.
13.e5 lt:\dS 14.lt:\xd5 hdS 15.a4
(15.b4 .ie7 16.�d2 0-0 17.a4 E:fc8)
15 ... .ic5 16.�e2 0-0 ! is also fine for
him.
ll.�f3 J.e7
12.ti'h3
12 .g4? ! runs into 12 . . . h5 !
12 .a3 is a consistent move. How­
ever, it slows down the attack. Black
should play immediately 1 2 . . . b4
13.axb4 lt:\xb4 in order to keep an
eye on the d3-bishop : 14.�h3 (14.
'i:la5 0-0=) when 14 ... 0-0?! allows
15.e5 ! . Therefore, it is better to pre­
vent it by 14 . . . e5 or 14 . . . E:b8 ! ? (va­
cating a retreat square on a8 in the
event of lt:\aS).
12 . 0-0
..
Most players prefer to attack im­
mediately the enemy bishop on d3
with 12 ... lt:\b4. You should keep in
mind this move as a backup line in
your repertoire. Its downside is that
White's b3-knight gets an excellent
place on d4 from where it supports
White's play on both flanks. The
critical position arises some moves
later (on the next diagram).
13.E:ac1 0-0
In my game against Baklan at
the Olympiad in Istanbul 2000, I
underestimated the importance of
the c-file and chose 13 . . . lt:\d7? ! 14.a3
lt:\xd3 15.cxd3 �d8 16.f5 eS, when
17.lt:\d5 ! hdS 18.exd5± underlined
the weakness of the c6-square.
13 . . . E:ac8 is a worthy alternative
which often transposes to the main
line.
14.a3 lt:\d3 15.cd3 �d8
15 . . . �d7 deprives the f6-knight
from its best retreat square, but it
is playable.
16.lt:\d4 E:c8
This position is roughly equal,
but it is difficult to win it as Black.
White is very solid in the centre and
he can manoeuvre without risking
131
 Part 4

to get worse. For instance, he can forces with . . . tt'ld7, . . . �f6. See game
double the rooks on the c-file or 22 Belikov-Zakhartsov, Alushta
play '\Wg3, <;t> hl. I prefer to avoid it 2 0 05.
if possible. I will consider below the
advance of the g-pawn : 1 3 . . . tt'lb4 ! ?

17.g4
In Nedev-Lautier, Chalkidi­
ki 2 0 0 2 , was 17.tt'lf3 d5 18.e5 d4 ! ?
19.tt'lxd4 tt'ld5 2 0 .tt'lxd5 '\Wxd5 2 1.'\Wf3
'\Wd'Too .
17 . . . tt'ld7 18.g5
The game Negi-Mamedov, Dubai
2004, saw 18.b4 �f6 19.tt'lce2 E:xc1
20.E:xc1 e5? 21.tt'lf5 exf4 2 2 .tt'lxf4±.
A possible improvement is 2 1 . . .
'\Wb8, planning . . . E:c8, . . . �d8-b6.
The stem game Kozakov-Ata­
lik, Lvov 2 0 0 0 , saw another move
order: 13 . . . tt'ld7. It allows Black to
avoid the piece sacrifice after 13 . . .
tt'lb4 14.e5, but offers the enemy the
option of 14.f5 ! ? exf5 15.tt'ld5 �d8
16.gxf5 tt'lce5 with unclear play ac­
cording to Atalik. I think that unless
practical tests prove 13 . . . tt'lb4 14.e5
to be dubious, Black should delay
the manoeuvre tt'lf6-d7.
14.g5

18 ... tt'lc5 19.E:cd1 E:e8 ! ?

The crucial question here is how
to evaluate the position after 14.e5
Black should not admit the ene­
tt'lxd3 15.exf6 tt'lxb2 16.fxe7 E:fe8 !
my knight to f5. The game Baklan­
17.�d4 E:xe7
Tregubov, France 2003, saw 19 . . .
e5? ! 2 0 .tt'lf5 tt'le6 2l.tt'lxe7+ , when a
draw was signed. Instead, 2 1.�g3 !
would have been unpleasant.

Let me add that the popular

lately 12 . . . h5 is not all roses. After
13. <;t> h1, I'm at a loss to offer a good
move : 13 . . . tt'lg4 14.�g1 g6 (14 . . . g5
15.�e2±; 14 . . . tt'lb4 15.�e2) 15.�e2 ;
13 . . . g6 14.a3 ! ? tt'lg4 (14 . . . E:c8 15.f5
gxf5 16.exf5 e5 17.�g5) 15.�g1 �f6
16.E:adtt.
13.g4
If White refrains from the pawn
storm (13. <;t> h1), Black redeploys his
In my opinion, Black has good
practical chances here. The gaping
main diagonal and the semi-open c­
file will be a constant source of con­
cern to White. Black will soon play
. . . e5 to open the position. Add to
that your opponent's surprise. He
may have read annotations stating

132
 The f4 System

that 14.e5 was winning, but instead

finds himself faced with serious tac­
tical problems. Summing up, an in­
teresting struggle is ahead : 18J�ae1
ti'lc4 19.ltle4 e5 2 0 .fe5 lt:lxe5 ! oo .

It i s time t o take stock. On move

12, I recommended 12 . . . 0-0 instead
of 12 . . . lt:lb4. That could cost us even
a piece should the opponent braced
himself for 14.e5. However, only
two moves later the situation seems
nearly the same as in the comments
to move 1 2 . What has Black gained
by his risky move order?
Let us examine White's options :
15.a3 is outright bad due to 15 . . .
ti'lxd3 16.cxd3 ltl c5 17.lt:lxc5 deS+.
After 15J3acl, instead of trans­
posing to the game Baklan­
Tregubov by 15 . . J:l:ac8 16.a3, we
have 15 . . . lt:lc5 ! 16.f5 (16.lt:lxc5 dxc5
17 . .ie2 1'l:ad8 18.f5 ef5 19.ef5 c4�)
16 . . . exf5 17.exf5 lt:lbxd3 18.cxd3
�d7+, Bjerring-Plachetka, Copen­
hagen 1990.
In my unfortunate game against
Baklan I had something like this in
mind, but the untimely . . . lt:ld7 al­
lowed White to attack with the f­
pawn, leaving the g-pawn on its ini­
tial square.
Another plus is that we have
provoked the weakening advance
g2-g4.
15 . . . 13ae8 ! ?
Atalik's recommendation.
15 ... ltlc5 might finish with a re-
. petition of moves after 16.f5 exf5
17.ltlxf5 (17.exf5 lt:lbxd3 18.cxd3
'!Wd7t) 17 . . . lt:lbxd3 18.cxd3 lt:lxd3
19 . .id4
19.lt:lxg7 is dubious since Black
gets a terrible initiative after 19 . . .
lt>xg7 2 0 .g6 hxg6 2 1..ih6 'it>g8
2 2 .hf8 \Wc5+ 23. 'it>h1 1'l:xf8 24.'\Wxd3
lt>g7.
19 . . . lt:le5 2 0 . 1'l:ac1 \Wd8 21.1'l:cd1
'\Wc7 ( 2 l . . .f6 ? ! 2 2 .\Wg2t) 2 2 .1'l:cl.
16.f5
Or 16.1'l:f3 .id8 17.'\Wh5 g6 18.\Wh6
f5 ! 19.gxf6 .ixf6oo.
16 ... exf5 17.lt�xf5 (17.exf5? !
lt:lxd3 18.cxd3 ltl e5t) 17 . . . �e5oo.
C. 7 .ie2 b5
•

Another critical moment for

Black's plan is: If White wants to take on c6, he
should do it before Black played
.ib7.

133
 Part 4
8.loxc6
An alternative is 8.i.e3 .tb7 9. 0-0
(9 . .tf3 .!DaS ! 10.0-0 .!Dc4 ll.i.cl .!De7
12.@h1 .!Dg6, Hou,Yfan-Movsesian,
Khanty-Mansiysk 2 011), but 9 . . . -tcS
solves all the problems : 10 .i.f3
Or lO . .!DfS .he3+ (10 . . . .!Dce7
ll ..hcS �xeS+ 12 .�d4 �xd4+
13 . .!Dxd4 b4 14 . .!Da4 he4 1S . .!DcS .!Df6
16 . .!Dxa6 .!DedS 17.a3 @e7 18 . .!Dxb4
.!Dxb4 19.axb4 .!DdS 2 0 .g3, draw, Ni­
sipeanu-Grabliauskas, Berlin 1997)
ll . .!Dxe3 �b6 12 .�d2 d6 = .
10 . . .d 6 ll.@h1 �b6 12 . .!Dce2 (12.
eS? ! hd4 13.hd4 .!Dxd4 14.hb7
�d8 1S . .te4 fS 16.i.d3 dxeS 17.fxeS
.!Dh6+) 12 . . . .!Df6 13.c3 0-0 14.b4
.hd4 1S.cxd4 (or 1S . .!Dxd4 �c7
16.�c1 .!Dxd4 17.cxd4 �e7 18.�d3
�ac8=) Shabalov-Rohde, Philadel­
phia 1996, when 1S . . . .!Db4 ! ? 16.�b3
aS would have faced the opponent
with the task of finding compensa­
tion for the pawn.
8 �xc6 9 .if3 .ib7 10 .te3
•.• • .
lO .eS �c7 11 . .!De4 is typically met
by ll . . . .!Dh6 12 .i.e3 .!DfS 13.i.f2 �d8
134
with 14 . . . d6 coming (even to 1S.g4).
10 . . . �c4 is a fair alternative. It
commonly leads to a minimal edge
for White in the endgame after
11.�d3 �c8 12 .a3
1 2 . 0-0-0 .!Df6 13.eS is parried
by 13 . . . .hf3 14.gxf3 �xd3 1S.�xd3
b4 ! = , e.g. 16 . .!D a4 .!Dd5 17 . .!Db6 .!Dxb6
18.hb6 �b8 19 . .tc7 �b7 2 0 . .td6 aS,
Balogh-Vitiugov, Plovdiv 2 0 1 2 .
12 . . . .!Df6 13 .i.d4 ! �xd3 (Svidler­
Hracek, Yerevan 1996, saw 13 ...
dS? when 14.�xc4 ! dxc4 1S.eS .!DdS
16. @f2 would have been pleasant
for White.) 14.cxd3 .tcS 1S.hcS
�xeS and Black should gradually
equalise.
ll.a3
This is the most popular move,
but this prophylaxis is in no way
obligatory. White also chooses :
a) 11.�d4 �c7 (This plan is con­
sistent with the main line.) 1 2 . 0-0
hS (enabling ... .!Df6) 13.�ad1 .!Df6
14.h3 h4 1S.�f2 .tcS 16 . .hcS �xeS
17.hcS �xeS= .
b ) 11.�d3 �c4 i s solid, but pas­
sive, as seen in the comments to
10 . . . �c4. I would like to advocate
for:
ll . . . �c7 ! ? . It leads to sharp play
where White might easily overes­
timate his chances. Independent
variations arise after:
 The f4 System
12.0-0-0 t!Jf6 13.'it>b1 (13.g4 b4
14.t!Je2 �c4)
13 . . . �b4 ! ?
Zapata successfully tried 1 3. . .
d6, but it cuts the scope of the dark­
squared bishop.
14.�d4 0-0oo 1S.hf6 gxf6
16.t!Je2 :t!fdS 17.c3 �cS (17 . . . �e7)
18.:t!hfl 'it>hS 19.g4
How to evaluate this position? If
we ask Houdini, it tends to prefer
White's side. In my opinion, Black
has full-fledged counterplay with
his bishop pair.
19 . . . b4 (19 . . . dS 2 0 .exdS hdS
21.hdS :t!xdS 2 2 .�f3 :t!xd1 + 23.:t!xd1
aS 24.fS e5fZ) 2 0 .eS (20.cxb4 hb4
21.�d4 �e7 2 2 .1'!cl �d6) 20 . . . dS !
Black sacrifices a pawn in order
to keep the enemy pieces passive. I
have analysed from this point:
2 1 .exf6 bxc3 2 2 .�xc3 aS. An im­
portant resource. The pawn is head­
ing for a3, the b7-bishop obtains an
active place on a6. Over the board, it
is easier to play with Black. 23.t!Jc1
�b6 24.�e2 �e3 2S.�e1 a4 26.�d3
hc1 27.:t!xc1 a3 2 8.�d2 axb2
29.:t!xcS :!'!xeS 30.�xb2 �e3 31.�e2
�e4+ 3 2 .'it>a1 d4= ;
21.cxb4 hb4 2 2 .exf6 aS 23.t!Jd4
(23 .gS :t!d6 24.1'!cl �a6) 23 . . . �cS
24.t!Jb3 �xf4oo
24 . . . �d6 2S.:t!c1 �b6 26.:t!xcS
:!'!xeS 27 . .ie2 a4 forcedly leads to a
draw endgame with opposite col­
oured bishops : 28.�d4 �c7 29.:t!c1
�c6 3 0.�bS axb3 31.:t!xc6 bxa2 +
3 2 .'it>a1 �bS 33.:t!xcS+ �xeS 34.fS
�bS 3S.�d3 �eS 36.fxe6 fxe6 37.gS
�a7= .
2S.t!JxaS �b4 26.t!Jxb7 �xb7
27.�e4 :t!bS 2S.b3 �a7 29.hh7
:t!xb3+ 30 .�xb3 :t!bS 31.�c2 :t!xb3 +
32 .hb3 �b6 33.gS �b4 = .
c ) Once again, ll.eS misses the
goal - ll . . . �c7 1 2 . 0-0 (12 .t!Je4
t!Jh6 13.0-0 �e7 14.a4 bxa4 1S.:t!xa4
0-0=) 12 . . . t!Je7 (12 . . . t!Jh6 13.t!Je4
�e7 14.c3 0-0 1S.�e1 �dS 16.t!Jg3
�f3 17.:t!f3 d6=, Milos-G. Hernan­
dez, Bogota 1992) 13.hb7 (13.t!Je4
t!JdS) 13 . . . �xb7 14.�d3 t!Jf5 1S.�f2
13S
 Part 4
1S . . . hS (1S . . . dS ! ? may be sim­
pler.) 16.lLJe4 tie7 17J%ad1 0-0! (it
is better to avoid dark-squared
blockade after 17 .. .'�c6 18.c3 0-0
19.b4 ! ) 18.'�f3 (18.c3 l:l:fd8=) 18 . . .
�c6 (mind the check!) 19.l:l:d2 d6
20.exd6 lLJxd6 2 1.lLlxd6 .hd6= , Pla­
chetka-Doda, Novy Smokovec 1971 .
d) 11.0-0 ticS 12 .�e1 (12 .�d3
b4 13.lLle2 lLlf6 ! 14.lLJd4 �b6oo)
12 .. .'grb6 (12 . . . lLJe7) 13 . .hcS �xeS+
14.i>h1 lLJe7= .
ll . . . �c7!?
The queen's retreat anticipates
e4-e5 and maintains the tension.
12.�d4
After 1 2 . 0-0 ticS (12 . . . lLJe7 might
transpose, but it offers White addi­
tional options: 13.'grd2 ! lLJg6 14.l:l:ad1
ticS 1S ..hcS �xeS+ 16.'it>h1 0-0 ! . It
turns out that Black has good com­
pensation for the pawn following
17.�xd7 �b6.), Black completes
development unimpeded : 13 . .hc5
�xeS+ 14.i>h1 lLJe7 1S.�d2 when
1S . . . lLJg6 16.l:l:ad1 0-0 was men­
tioned above, but Black also has the
natural 1S . . . l:l:c7= .
12 . . . h5
12 . . . tie7! ? is worth testing, too :
13.0-0-0 (or 13.0-0 lLJf6 14.l:l:ad1
0-0 1S.eS .hf3 16.gxf3 lLJhS 17.�d7
.icS?, Ghizdau-Matera, USA 1976)
13 . . . lLJf6 14.l:l:d2 (14.eS? tif3 1S.gxf3
lLJdS) 14 . . . d6 ! oo (In the source game
136
Lukin-Novgorodskij , St. Petersburg
1996 there was 14 . . . 0-0?! 15.e5 fix£3
16.gxf3 lLJe8, when 17.'�xd7! would
have been clearly better for White).
1S.g4 (1S.i>b1 0-0) 1S . . . lLJd7 16.l:l:hd1
eS 17.�d3 exf4 18 . .hf4 lLJeS with
chances for both sides.
Black's main problem in the f4-
line is how to develop his king's
knight. Its most active stand is f6,
but the enemy could attack it with
e4-eS . That's why Black should pre­
pare for the knight a square where it
could jump from f6. The text move
enables lLJg8-f6-g4.
13.h3
This prophylaxis is positional­
ly risky since the weakness of the
g3-square often turns to be fatal
for White. At the same time, it is
the most testing continuation. Now
Black must waste a tempo on h4 in
order to find another place for his
knight - on hS. Besides, on h4 the
pawn is an easy target. The other at­
tempts are not too convincing:
a) 13 . 0-0 lLJf6 14.i>h1 .icS
1S.�xc5 �xeS 16 . .hcS l:l:xcS = .
 The f4 System
b) 13.0-0-0 tt:lf6 14J'!d2 (14.
eS? ! tt:lg4 15.!!d3 hf3 16.gxf3 tt:lxe3
17.�xe3 dS is obviously better for
Black.) 14 . . . tt:lg4 ! (In Brajovic-Ra­
jkovic, Herceg Novi 2 0 0 2 , was 14 . . .
ia3? 15.e5 'IM!xc3, when 16.�xc3 !
l'!xc3 17.hb7 !!xe3 18;exf6 i.cS
19.fxg7 !!g8 2 0 .ha6 !!e4 2 1.@b1
would have earned White a small
edge.) 15.i.g1 eS 16.fxe5 tt:lxeS
17.tt:ld5 �b8?.
13 h4 14.0-0 �f6
••.
Black has good prospects. For
15.e5 (15.!!ad1 is altogether dubi­
ous due to 15 . . . ha3 16.tt:lxb5 axbS
17.bxa3 0-0.) 15 . . . tt:lh5, see game
24 Gofshtein-Tregubov, Montpel­
lier 1998.
D. 7.�xc6 ti'xc6 8 .Ad3 bS
•
8 . . . i.c5 is Taimanov's pet move.
It often leads to the same positions
as 8 . . . b5. You can choose it in or­
der to avoid variations with short
castling for White. However, they
look pretty harmless as you'll see in
the comments on the next move. At
the same time, Black should reck­
on with the more dangerous answer
9.�g4 g6 10 .i.d2 (10 .e5 fS ll.exf6
ltJxf6 1 2 .i.g6 @d8 ! 13.'\MfgS i.e7+)
lO . . . dS ! ll.exdS exdS 12.'\MfgS tt:le7
13.0-0-0oo.
9.i.e3 (9.0-0 i.cS ! ?) 9 ... .ib7
10.0-0 tt:lf6 leads to unclear play
with mutual chances: ll.a3 i.cS
12 .hc5 (12 .�e2 he3 13.�xe3 dS
14.exd5 tt:lxdS 15.tt:lxd5 �xdS 16.!!f2
0-0=) 12 . . . '\MfxcS 13.@h1 d6 14.'1Mfe2
eS 15.ttld5 ttlxdS 16.exd5 0-0 17.fxe5
dxeS 18.'\MfxeS draw, Ni Hua-Ye
Jiangchuan, Yongchuan 2003.
Black can develop his king's
knight on e7 or f6 :
Dl. 9 . . . ic5; D 2 . 9 . . . ib7
Dl. 9 ••• J.c5 10 .id2
•
9.i.e3 i.b7 1 0 . 0-0 h4= (10 . . . tt:lf6
ll.hcS 'IMfxcS 1 2 .@h1 b4 reaches a
well known position where both
knight's retreats, to b1 or d1, lead
to equality.) , see game 23 Stuer­
zenbaum-Taimanov, Bad Zwische­
nahn 2 003.
10 .ib7
.••
10 ... ttle7 allows ll.eS J.b7 12.ttle4;!;.
11. 0-0-0
The rare ll.a4 scores extremely
well. Black's task is not easy at all.
If you are not prepared for a prin­
cipled fight after ll . . . b4 1 2 .ttld5,
137
 Part 4
you may try to surprise your oppo­
nent with ll . . . lt:Jf6 ! ? 12.axb5 axbS
13J''!x a8+ haS 14. ixbS (14.e5
'Wxg2) 14 ... 'Wb7. Black does not win
a pawn here, but his pieces are ex­
cellently placed.
Let's return to ll . . . b4 1 2 .ll::l d 5
'Wd6 13.lt:Je3 '\Wxf4 14.lt:Jc4
White has a certain compensa­
tion for the pawn. 14 . . . '\Wh4+ 15.g3
Wie7 16.ll::l a5 is messy. More ambi­
tious is:
14 . . . '\Wc7 15.Wig4 dS ! ?
15 . . . 1!if8 16 . .if4 is unclear.
16.exd5 fS 17.'We2 .b:dS 18.lt:Je3
.ixe3 19.'\Wxe3. This position was
reached in Wang Li-Sadorra, Sub­
ic Bay 2009. Black would have had
a comfortable game after 19 . . . lt:Jf6
20.ixb4 l'!b8 2 1..ic3 lt:Je4.
ll . lt:Je7 12.a3 !
..
The most topical continuation .
White prevents . . .b 4 and waits to
see his opponent's plan.
12 .Wih5 is the older approach.
Black may transpose to the main
line by 12 . . . 0-0 when 13.a3 seems
obligatory (Papp-Szabo, . Buda­
pest 2008, saw 13.g4 b4 14.ll::l e 2 aS
138
15.1!ib1 a4 and Black is ahead in the
pawn race). However, he may also
choose long castling, using the awk­
ward position of the enemy queen:
12 ... g6 ! ?
13.'\WgS
From here, the queen is eying
the bishop on cS thus making 13 . . .
b4? impossible d u e t o 15.lt:Ja4±.
In contrast, 13.'\Wh4 allows 13 . . . b4
when White has two possibilities :
14.'\Wf6 0-0-0 15.ll::l e 2 dS 16.ll::l g3
(16.e5 stumbles into 16 . . . d4 ! ) 16 ...
'Wa4 ! ? 17.®b1 dxe4 18 .b3 '\Wd7
19 . .ie4 .id4 2 0 .'\Wh4 .ixe4+;
14.lt:Je2 f5 15.lt:Jg3 0-0-0 16.l'!hel
'1Wb6 17.1!ib1 l!ib8 18.l'!e2 l'!c8+.
13 . . . 0-0-0 ! ?
1 3 . . .£5 i s a worthy alternative.
The game Rublevsky-Papin, Tagan­
rog 2011, went on 14.exf5 gxfS
15.'\WhS+ ®dB 16.g4 b4 17.gxf5 exfS
18.ll::l e 2? '\Wg6+. Instead, 18.l'!he1
was the only move, when simplest
is 18 . . . 1!ic7 19.ll::l e 2 Wig6 = .
14.l'!he1 fS 1 5 . .ie3 l'!de8 16 . .ixc5
'WxcS 17.exf5 h6 18.'\Wh4 gxfS 19.g4
fxg4 with mutual chances, Rublev­
sky-Alekseev, Irkutsk 2 0 1 0 .
 The f4 System
12 . . . 0-0
12 ... 0-0-0 does not promise
Black adequate counterplay. After
13.'.t>b1 g6 14.1":1he1 fS 15.g3, White's
pieces are extremely well co-ordi­
nated in the centre.
I suppose that this sharp posi­
tion will be extensively tested in the
following years. I analyse it in de­
tail in the annotations to game 21
Nepomniachtchi-Vitiugov, Nizhny
Novgorod 2 0 13 .
D 2 . 9 . . . .tb7! ? 1 0 . .id2
White could try to use the pos­
sibility of castling short, but prac­
tice is favourable for Black: 1 0 . 0 - 0
ttJf6 ! ll.a3 (ll.eS b 4 12.exf6 bxc3+)
ll . . . i.cS 1 2 . '.t>h1 i.d4 ! 13.lLla2 d6
14.a4 bxa4 1S. lLlb4 '\Wb6 16.1":1xa4 aS !
(improving on Hector-Plachetka,
Gausdal, 1989, which saw 16 . . . 0-0)
17.i.b5 (17.c3 i.cS 18 .i.b5 '.t>e7)
17 . . . '.t>f8 ! 18.c3 axb4 19.1":1xa8 ha8
20.cxd4 i.e4+.
10.a3 i.e7 ll.i.e3 is another rare
alternative : 11 . . . 1":1c8 1 2 . 0-0 lLlf6
13.e5 lLldS 14.lLlxd5 '\WxdS = . The
game Lanin-Vitiugov, St. Peters­
burg 2 007, soon ended in a draw af­
ter 15.a4 i.cS 16.hc5 '\WxcS 17.'\Wf2
'\Wxf2 + 18.1":1xf2 i.c6 19.axb5 axbS
2 0 . 1":1a5 b4 ! 2 1.1":1d2 '.t>e7 2 2 . '.t>f2 1":1a8
draw.
10 . . . .te7! ? ll.a3
Or 11. 0-0 b4 12.lLld1 lLlf6 13.lLlf2
0-0t.
It is good to make a useful move
before leading out the knight. Af­
ter l l . . . lLlf6 1 2 . e5 lLldS 13.i.e4 b4
14.axb4 i.b4 1S.lLlxd5 hd2 16.'\Wxd2
exdS 17.i.f3, White is a pawn up al­
beit in a sharp position.
1 2 . 0-0-0
12.0-0 lLlf6 is equal : 13.'.t>h1 0-0
14.1":1£3 dS 15.exd5 lLlxdS 16.hh7+
(16.lLlxd5 '\WxdS 17.1":1h3 g6) 16 . . .
'.t>xh7 17.1":1h3 '.t>g8 18.'\WhS f6= or
13.e5 lLld5 14.lLle4 fS= .
12 . . tt)f6 13.g4 !
.
139
 Part 4

White has a number of other op­ 2 0 .tt:le4 i.e7 21 . .ic3 0-0 = . Perhaps
tions. Black always meet them with critical is 16. 'i!lb1 which is yet to be
. . . b4 : tested.

a) 13.'i!lb1 13 . . . . d5

This is aimed against the the­

13 . . . b4 14.axb4 .b:b4 15.g5 .b:c3
matic 13 . . . b4 14.ab4 i.b4, when
16 . .b:c3 tt:lxe4 17 . .b:g7 :!! g 8 18.i.e5
15.l!:ld5 i.d2 is without check. Nai­
was better for White in M . Leon
ditsch-Grigoriants, Moscow 2008,
Hoyos-Y.Hernandez Guerrero,
went 16.tt:lf6 gf6 17J�d2 V9a4 18.V9e3
Toluca 2009.
V9b4 19J�hd1 :!! cS and here 20.V9g3 ! ,
eyeing b8, would have been quite
14.exd5 (14.e5 tt:ld7) 14 tt:lxd5
unpleasant. We can answer with the
• • •

15J::!hfl tt:lxc3 16.hc3 0-0

no less cunning:
13 . . . V9b6 ! which in its turn neu­
tralises White's main idea of 14.g4
owing to 14 . . . b4 15.axb4 .b:b4 16.g5
.b:c3 17 . .b:c3 :!!xc3 18.gxf6 gxf6oo.
b) 13.:!!hf1 b4 14.axb4 .b:b4
15.g4 0-0 (15 . . . d6 leads to more
complex play) 16.g5 .b:c3 17 . .b:c3
tt:lxe4 18 . .b:e4 V9xe4 19.V9xe4 .b:e4
20.:!!xd7 :!!fd8. This endgame should
be a draw due to the opposite col­
oured bishops. Black has completed develop­
ment, but there is a fair chance that
c) 13.:!!h e1 ! ? b4 14.axb4 .b:b4
White is winning by force. The good
15.g4 d6
news is that the most consistent con­
tinuation of White's play, 17.f5, is
not dangerous due to 17 . . . i.g5+ (An
only move. 17 . . . :1!fe8? 18.f6 .b:a3 is
very dubious as White gets a terri­
ble attack with 19.i.e5 i.cS 2 0 .b4
.b:b4 2 1.fxg7.) 18 .i.d2 (18.'i!lb1 eS is
balanced: 19.h4 .b:h4 2 0 .'�h2 i.f6oo
or 19.f6 e4 2 0 .fxg7 :!!fe8 21.V9f2 V9g6
22 . .ie2 e3 23 .'�g3 i.e4 24.:!! d 6 :!! e 6
25.:1!d7 i.f6 = ) 18 . . . .b:d2 + 19.�xd2
Jovanovic-Timofeev, Budva
exfS 20 . .b:f5 :!! c 7.
2009, saw further 16.e5 �eS 17.g5
tt:ld7 18.i.e4 V9c7 19 . .b:b7 V9xb7 Critical is:

140
 The f4 System

17.g5 gfe8 ! 27J"!xd4 hc8 2 8J"!e5 i.e6 2 9J"!a4

i.d6 3 0.ge2 <tlg7
Emms' suggestion 18 . . . W/b6? los­
es at once to 18 .i.d4 i.c5 19.hg7.

18.hg7 <tlxg7 19.1lfh5 gbs

20.f5 exf5 21.gxf5 1lfg6 2 2 .1lfg4
ffd6

I think that Black has decent

play here, but it all depends on the
evaluation of the following hazy
ending: 23.hb5 (23.i.e4 W/b6
24J"!xf7+is a perpetual check af­
ter 24 . . . <tlxt7 25.1lff5+ <tlg8 26.i.d5+
i.xdS 27.1lfxd5+=) 23 . . . W/e6 24.W/d4+
c;!;>g8 25.i.d7 W/c4 26.hc8 W/xd4
Next, Black will play 3 1 . . .gas
and then he will attack the enemy
kingside pawns with his king and
bishop pair. I believe that over the
board, Black's game is easier.

141
Pa rt 4. The f4 System
Ka n - Step by Step
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4. �xd4 a6 5.�c3 'f!Jc7 6.f4
This move is seldom seen, be­
cause it offers Black a wide choice
of good plans. It is better to await
. . . lt:Jf6 first as in the variation 6 . .td3
lt:Jf6 7.f4. In the diagram posi­
tion, Black can transpose to the
Taimanov with 6 . . . lt:Jc6, but that
would not be the best decision. It is
more logical to fianchetto the bish­
op to b7 first. Then he could recap­
ture on c6 by . . . hc6. As for the g8-
knight, it can go to e7 in some lines.
6 . . . b5 7 . .td3
The only reasonable alterna­
tive to this development is 7 . .te2
142
(7.a3 .tb7 8 . .td3 transposes) 7 . . . .tb7
8 . .tf3 . The simplest retort is 8 . . . b4
(8 . . . lt:Jc6 is also good : 9. �xc6 hc6;
9 . .te3 lt:JaS 1 0 . 0-0 lt:Jc4 ll . .tcl .tc5
12 . 'it>h1 lt:Je7 13.�d3 �b6 1 4 .lt:lce2
0-0 15.b3 lt:ld6 16 . .tb2 f6+, Rojo
Huerta-Rajkovic, Madrid 1994)
9.lt:Ja4 lt:Jf6 10.�e2 lt:Jc6 11.lt:lb3 lt:Ja5!
1 2 . 0-0 .te7.
7 . . . .ib7 8.�f3
a) 8.�e2 lt:Jf6 is fine for Black,
see game 26 An . Martin-Delchev,
Benasque 2 0 13. Of course, 8 . . . lt:Jc6
9.lt:Jxc6 �xc6, transposing to the
Taimanov, is also possible, but it is
hardly the sternest test of White's
set-up.
b) 8.0-0?! .tcs pins the d4-
knight: 9 . .te3 �c6 ! (there is no rea­
son to give White an initiative after
9 . . . �b6 10.lt:Jce2 lt:Jf6 ll.c3 ! lt:Jxe4
12 .b4 .tf8 13.a4) 10.lt:Jce2 (10.�g4
lt:Jxd4 11.�xg7 lt:Jxc2 12 .hc2 he3+
13.'it>h1 b4 14.lt:ld1 .tb6 1S . .td3 lt:Jf6,
when 16.�xh8+? loses to 16 . . . 'it>e7
17.�g7 �g8 18 .�h6 �xe4-+) 10 ...
lt:Jf6 ll.c3 lt:Jg4 12 . .tc1 0-0 13.'it>h1
eS = , Skage-Macieja, Tromso 2 0 07.
 The f4 System

8 .lt�f6 9 .ie3
.• . 14.lt:Jd1 hb2 1S.lt:Jxb2 �c3 + 1 6 . .id2
�xb2 17.0-0 lt:Jf6+) 14 . .id4 b4
9.g4? ! is premature. 1S.h:f6 gxf6 16.lt:Je2, when 16 . . . dS !
17.lt:Jg3 !!h4 18 .fxe6 �xe6+ would
Even the most straightforward
have gained the edge.
retort 9 . . . b4 10.lt:Jce2 dS l l.eS lt:Je4
gives Black counterplay: 1 2 .lt:Jg3 ll . . . lt:Jg4 12 . .ig1 lt:Jxd4 13 .h:d4
liJcS (12 . . . lt:Jc6 13 . .ie3 .icS 14.lt:Jb3 ! .icS ! ?
he3 1S.�xe3 lt:Jxg3 1 6.hxg3 d4
1 3. . .e S 14.fxeS .icS 1S.hcS �xeS
17.�g1 lt:JxeS 18.fxeS �xeS+ 19.�d1
16.0-0-0 0-0 is unclear.
hh1 2 0.�xh1 !!b8 2 1 . a3 bxa3
22.!!xa3 �xg3 is a total mess) 13.0- 14 . .ixcS (14.hg7 !!h7 1S . .if6
0 lt:Jxd3 14.cxd3 .icS 1S . .ie3 �b6 lt:Jxf6 16.gxf6 .id4t) 14 . . . �xcS
16.�f2 lt:Jc6 17.lt:Jge2 0-0 = . After a 1S.�e2 fS
swap of rooks on the c-file, the d3-
It turns out that 16.h3? loses to
pawn will become quite sensitive,
16 . . . b4 17.lt:Jd1 (i7.lt:Ja4 �d4) 17 . . .
e.g. 18.!!ac1 !!feB 19.!!c2 aS 2 0 . !!fc1
fxe4 18 . .ixe4 �c4 + , but 16.gxf6
hd4 2 1 .lt:lxd4 lt:Jxd4 2 2 .h:d4 !!xc2
lt:Jxf6 17.a3 0-0-0 also looks very
23.!!xc2 �bS.
pleasant for Black.
However, I think that Black can
b) 1 0 . lt:Jb3 is well met by 10 . . . dS !
aspire to the edge with 9 . . . lt:Jc6 !
ll.exdS (ll.eS d4 ! ) ll . . . lt:Jb4 12 . .id4
1 0 . .ie3 (10.lt:Jxc6 �xc6 ll.gS .ib4)
lt:Jxd3+ 13.�xd3 lt:JxdS with the
lO ... hS, transposing to 1 0.g4.
bishop pair. The play may continue
14 . .ieS �c8 or 14.lt:JxdS .ixdS 1S.O-O
9 lt:Jc6 10. 0-0-0
.id6.
. ..

a) It is still early for 10 .g4 due to c) 10.0-0 lt:Jxd4 ll.h:d4 .icS

lO . . . hS ! 12 . .ixcS �xeS+ 13.�h1 b4 ! offers
Black easy play on the central dark
squares by . . . d6, . . . eS.

ll.gS
Andriasian beat Potkin in Bel­
gorod 2 0 1 0 with ll.gxhS? ! lt:JxhS
12.lt:Jxc6 �xc6 13.f5 lt:Jf6 (13 . . . .ia3 ! ? 10 b4
...

143
 Part 4

This leads to more complex play
than 10 . . . tt:\xd4 ll.hd4 .icS. At this
point, 12 .hcS 'W'xc5 13 .'W'g3 (13.a3
b4 ! ) 13 . . . lt:lhS 14.V9g4 (14.'W'f3 lt:lf6)
19.'W'f2 .ia6 2 0 .b3 Elb8 2 1.hc4 dxc4
2 2 . .ie3 cxb3 23.cxb3 'W'c6 = .
17 a5 18.b3 �a3+ 1 9 . .b:a3
• • •

14 ... g6 1S.f5 b4 16.tt:\a4 'W'e3 + 17.'i!lbl bxa3 20.£5 0-0-0 2 1.�e3 g6

.ic6 is balanced. 2 2 .fxe6 fxe6 23.'i!lal ghf8
The problem line is 12 . .ieS V9b6
13.hf6 gxf6 14.fS. Perhaps Black
should answer 14 . . . eS followed up
by . . . hS, . . . 0-0-0. This Rauzer struc­
ture is too static. It may be solid,
but the potential appearance of op­
posite coloured bishops is a herald
of a draw.

ll.lt:lce2 �a5 12.g4 d5 13.e5

�d7 14. 'i!lbl �c4 15 .icl �b6 •

16.l:�hfl .ic5
Black has good counterplay on
the queenside thanks to . . . aS, . . . .ia6
activating the passive bishop. The
play may continue :
17.�f2
Or 17.�b3 aS 18.�xcS tt:\xcS
Both sides have fulfilled their
plans and the position is balanced.
White's first rank is permanent­
ly weak so he lacks a clear plan. A
possible continuation is 24.c3 'i!lb8
2S.lt:lf4 (2S.h4 Elxfl 26.Elxfl .ia6
27.hS Elc8 28.hxg6 hxg6 29.ha6
�xa6 30.Elbl Elf8 31.Eldl 'i!la8 3 2 .�f4
hd4 33.cxd4 gS= ) 2S . . . Elc8oo.

144
Pa rt 4. The f4 System

Com pl ete G a m es

7.Ybd4 a6 8 .ie3 b5 9 .0-0-0

20. Yemelin-Rodshtein •

.ib7 10 .id3 �c8 l l.<i>bl c!Llf6

Skopje 1 4.03.201 3 •

Some good players opt for 11 . . .

l.e4 c5 2.c!Llf3 e6 3 . d4 cxd4
WfcS 1 2 . Elhe1 Y;l/xd4 13 .hd4 ltle7,
4.c!Llxd4 c!Llc6 5 . c!Ll c3 Wfc7 6.f4
·

but this endgame is slightly better

c!Llxd4
for White after 14.a4 ltlc6 1S.�f2
bxa4 16.ltlxa4 ltlb4 17.ltlb6 ltlxd3
18.Elxd3 Elc7 19.f5, Shabalov-El
Debs, Sao Paulo 2009.

12.e5 c!Lld5 13.c!Lle4 c!Llxe3 14.Wxe3

This is an attempt to remedy

Black's basic problem in the main
line 6 . . . a6 7.ltlxc6 Wfxc6 when the
queen often has to return to c7. The
text effectively discourages plans
with �e2-f3 and leaves White a
very narrow path to follow. Thus 14 . . . �d5
you can prepare a decent repertoire
against the f4-system virtually for a 14 . . . �e7 should not change the
few minutes (by reading the anno­ course of the game as 15.Elhfl .idS
tations to this game). The flip side 16.Y;l/e1 transposes. Instead, An­
is that White's task is even easier. driasian-A.Horvath, Cappelle la
He can reach a pleasant position by Grande 2 0 1 2 , saw 15.Wg3 g6 16.h4
making the most natural develop­ hS 17.Elhe1 (17.ltld6+ hd6 18.exd6
ing moves. Y;l/xd6 19 .hb5 �dSoo) 17 . . . .id5 (17 . . .

145
 Part 4
he4 1S.he4 't!?fS=) 1S.lt:lf6+ hf6
19.exf6, when 19 . . . 't!?dS ! would have
been unclear.
The text is threatening 15 .. .'�a5.
15.ti'el IJ..e7 16.�m. b4 17.ti'e2
a5 18.c4!
This way White neutralis­
es the enemy pawn storm on the
queenside, but the weak second
rank will be a constant cause for
concern. Still, it is the best option.
1S . .ia6 l'!bS 19.lt:ld6+ hd6 20.l'!xd5
.icS 21.l'!ddl 0-0 2 2 .g4 d6 would
have been completely balanced.
18 ••• .he4
A difficult decision. Rodshtein
acknowledges the fact that he is
slightly worse and should struggle
for the draw.
After 1S . . . hc4 19.l'!cl hd3 +
20.ti'xd3 '<!Mb7 21.l'!xcS+ '<!MxcS 22 .l'!cl
'<!Mb7 23.lt:ld6+ hd6 24.1!Mxd6 't!?dS
25.1!Mc5 '<!MaS 26.'t!?al± Black would
be without useful moves.
1S . . .bxc3 19.lt:\xc3 .ic6 20 .f5 l=!bS
(20 . . ..ib4 21.lt:\b5 '<!Mb6 2 2 . l'!f4±)
21.lt:\b5 '<!Mb7 22.a4i is also boring to
defend.
19.ti'xe4 g6 20.g4 ti'c6 2 1.ti'e2
146
21 ••• ti'c5 ? !
Black's counterplay is based on
. . . a4-a3 when any open file would
be in Black's favour. The correct
move order was 2 1 . . .a4 ! 2 2 .f5 a3
23 .b3 l=!gS as in the game.
2 2 .f5 a4 23.13f3? !
White misses the opportunity
to activate his bishop with 23 . .ic2 ! ,
e.g. 23 . . .'�c6 24.'�f3;!;.
23 • • • a3 24.b3 13g8 ! ? 25.fxg6
25.f6 .idS 26.1!Md2 should be a
draw after 26 . . . 1!Mxe5 27.hg6 l'!c7
2 S.l'!el '<!Mb2 + 29.1!Mxb2 axb2 30 . .ih5
l'!gS 31.'t!?xb2 h6. The opposite col­
oured bishops become a strong lev­
elling factor without queens.
25 • • • hxg6 26.13dfl
26 • . . ti'd4
Black is right to give up the f7-
pawn, but he should have done it
by 2 6 . . . 't!?dS ! 27.l'!xf7 't!?c7 with ex­
cellent counterplay. White's extra
pawn is not worth much. On the
contrary, Black will use the open
 The f4 System
f-file to his benefit. After the text,
White retains some tine edge in the
endgame which, however, should
be drawn.
27.lU4 ti'c3 28.g4f3 ti'd4
29,gf4 ti'c3 30.gcl ti'b2 +
31.ti'xb2 axb2 3 2 . gcfl gcs (32 . . .
Elh8 =) 33,gxf7 gxe5 34,gb7 ge3
35.hg6+ �d8 36.!Uf7 gxg6
37,gxe7 gd3 38.geg7 gxg7
39.gxg7 gb3? (39 . . . 1':\di+ ! = )
40.�xb2 (40.Eig5 ! Elxh2 41.Eib5
Elh1+ 42. �xb2 Elh2 + 43.i>b1
Elh1+ 44.'it>c2 Elh2+ 4S. i>d3 Elxa2
46.'it>e4±) 40 .. ,gxh2+ 41.i>bl
�c7 42.ggs ghl + 43. �c2 gb2 +
44. �d3 gxa2 45.c5 d5 46.cxd6+
�xd6 47,gbs gg2 48.�b4 gg3+
49.�e2 �e5 5 0 . i>f2 gc3 51 . �g2
gd3 5 2 . �h2 gc3 53. i>g2 gd3
54.�h2 gc3 55,gb5 + �f4 56.g5
e5 57.g6 gc6 58.g7 draw.
21 . Nepomn iachtchi-Viti ugov
N izhny N ovgorod 1 1 . 1 0.201 3
l.e4 c5 2 .c!Li f3 e6 3.d4 cxd4
4.c!l:lxd4 c!l:lc6 5. c!l:lc3 ti'c7 6.f4 a6
7. c!l:lxc6 ti'xc6 8 .id3 b5 9 .ti'e2
•
i.b7 10 .td2 .ic5 11. 0-0-0 c!l:le7
•
12.a3 0-0
The opposite castles give Black
more chances to fight for a win than
12 . . . 0-0-0. Although both White's
bishops point to the kingside, it is
not easy to mount an attack. For in­
stance, 13.g4 may face a counter-at­
tack in the centre with 13 . . . d5. Then
14.e5 ti'c7 (enabling . . . c!l:lc6, . . . b4)
1S.f5? c!l:lc6 would favour Black.
In practice White had tried
13.Eihf1 f5 14.g4, when in both games
Black replied with the mundane 14 . . .
Elae8?! and only after 1S.i>b1, did he
push 15 . . . d5. [1S . . . .id4 16.gxf5 exfS
17.e5 d6 18.a4 looks nice for White :
. 18 . . . bxa4 19 . .ic4+ 'it>h8 20.exd6 l!Jc8
2 1.ti'd3 ti'xd6 ( 2 l . . .l!Jb6 2 2 .ti'xd4
l!Jxc4 23.Eig1 E\f7 24.Eige1 Elxe1
25.E\xeU h6) 2 2 . l!Jxa4 .ie4 23 .ti'b3
ti'f6 24 . .ic3 hc3 2S.ti'xc3t] Smirin
continued against Soffer, Plovdiv
2 0 1 2 , 16.gxf5 exfS 17.e5 d4 18.l!Ja2
l!JdS 19.l!Jb4 with a positional edge.
The rook is misplaced on e8. More­
over, it may be needed on a8 to sup­
port an attack with . . . aS, . . . b4. Thus I
reached to the conclusion that Black
should immediately define the cen­
tre by 14 . . . d5 ! 15.gxf5 exfS (1S . . . d4 ! ?
may transpose to 1 S . . . exf5).
Now 16.e5 d4 17.l!Ja2 is ineffec­
tive due to 17 . . . ti'h6 , anticipating
l!Jb4 and preventing �hS. The dif-
147
 Part 4

ference is seen in variations like

18.i.c4+ @h8 19.i.b3 l'!fc8, intend­
ing . . . as.
16.exd5 lt:lxd5 17.lt:lxd5 Wxd5
18.i.c3 l'!fe8 19.i.e5 Wfl is roughly
balanced.
This analysis led me to the
insidious move 14.@bl (instead
of 14.g4). Then 14 . . . l'!ae8? ! 15.g4
would transpose to Smirin-Soffer
so we should play 14 . . . d5 ! 15.exf5
exf5 when the only way to set Black
problems is 16.b4
23.Wg4 i.c8 24.f5 Wd5 25 . .icl
i.g7 2 6.l'!dfl l'!f7 (26 .. ,l'!xd3 27.cxd3
Wb3 + 2 8 . @al �xd3 29 .�dl Wxdl
3 0 .l'!xdl hf5=) 2 7.h3 i.d7 2 8 . l'!f4
( 2 8.i.b2 gxf5 29.hf5 i.xf5 3 0 .l'!xf5
l'!xf5 31.l'!xf5 l'!el+ 32 .i.cl �e6=)
2 8 ... l'!xd3 29.cxd3 �b3+ 3 0 .@al
Wxd3 3 1.i.b2 i.xfS 32 .�f3 �xf3
33.1'!4xf3 l'!e7 34.g4 i.c2 35.h4 h 6 = .
Nepomniachtchi tries t o im­
prove White's play by discouraging
. . .f5.

16 . . . i.d6? ! 17.Wf2 (with lt:lc3-e2- 13 .�h5 f6

d4 in mind) 17 ... i.c7 18.lt:le2 i.b6
19 . .ie3 he3 2 0 .Wxe3;!; is strategi­ 13 . . .f5?! is really bad after 14.g4
cally grim despite White's weak­ g6 15.�g5 d5 16.l'!hel l'!fl 17.gxf5
ened castling position. exf5 18.exd5±, but 13 . . . b4 ! ? 14.axb4
hb4 15.l'!hel is unclear. Instead,
Fortunately, Black can employ Vitiugov chooses to remain passive
Kasparov's favourite method of de­ on the kingside.
fence - a positional pawn sacri­
fice: 16 . . . i.d4 ! 17.Wxe7 l'!ae8 18.�g5 14.g4 gac8 15.gbet 'ml6 16. @bt
hc3 19.i.xf5 i.f6 2 0 .�g3 d4 with i.d4 t7. �e2 .if2 ts.gfl gf7
excellent compensation. Here are
some illustrative variations : 21.l'!f2
(21.Wb3+ Wc4�; 2 1 . . .Wd5 2 2 .Wxd5+
i.xd5 23.g3 l'!e2 24.h4 l'!h2 25.i.d3
i.c4�) 2 1 . . .g6 22 .i.d3 l'!e3 !
This rook splits White's position
in two. I do not see how he could
improve:

148
 The f4 System
Black's set-up is very resilient.
Nepomniachtchi admits that fur­
ther pawn advances are not promi­
sing and starts a redeployment of
his pieces.
19.'�h3 tDg6 20.i.cl .ic5?
The critical moment of the game.
The f2-bishop must have stayed on
its place in order to bar the f-file.
A possible follow up was 2 o ... a5
2l.tt:'lc3 b4 2 2 . tt:'lb5 b3 23.cxb3 (23.
c3 :!'i:ff8 24.a4 :!%cd8) 23 ... :!%xcl+ ! ?
24.:!%xcl tt:'lxf4 with fair compen­
sation. Here is an illustrative line :
25.'�f3 .ie3 26.:!%cdl fS 2 7 . .ic2 fxe4
28.he4 dS 29 . .id3 d4 30.'�g3 .ia6
3l.a4 tt:'lxd3 3 2 . :!%xf7 �xf7 33 J'!xd3
hbS 34.axb5 'W'xb5= .
I f White shifted his bishop to b2,
Black should plug the main diago­
nal by . . . eS, e.g. 2 1 .b3 'W'cS 22 . .ib2
e5= 23 .\1;lfg2 .ie3 24.fxe5 tt:'lxeS.
21.e5! b4
It turns out that 2 1 . . .fxe5 2 2 .fxe5
:!%xfl 23.:!%xfl :!%f8 fails to 24.tt:'lf4. The
game is practically over now.
2 2 .axb4 J.xb4
23.f5! exf5 24.i.xf5 tDf8
25.exf6 g6 26.J.d3 gxf6 27.b3
gxfl 28.gxfl tDe6 29.i.b2 gf8
30 . gdl gf3 3 1.Vh4 J.c6 3 2 .i.c4
i.b5 33.tDd4 gf7 34.tDxe6 hc4
35. tDg5 ge7 36.'e'h3 J.e2 37.ti'g2
ges 3 8.ti'd5 + 1-o
22. Beli kov-Zakhartsov
Alushta 2005
l.e4 c5 2 . tDc3 e6 3 . tDf3 tDc6
4.d4 cxd4 5. tDxd4 ti'c7 6. tDb3
a6 7.J.d3 tDf6. 8. 0-0 b5 9.f4 d6
10.fif3 .ib7 ll ..ie3 .ie7 12 .fih3
0-0
13.�hl? !
Playing on the basis of gener­
al considerations in a sharp posi­
tion often implies losing the initi­
ative. Besides, I am not sure at all
that the king is better off on hl. Crit­
ical is 13.g4 tt:'lb4 ! ? which is consid­
ered in "Step by Step". This game il­
lustrates Black's possibilities when­
White is waiting without commit­
ting himself with pawn moves.
13 . . . gfe8 14.gael tDb4 15.a3
149
 Part 4
tilxd3 16.exd3 tild7 17.tild4 tileS
18J:�dl
Black has consolidated his de­
fence and could think about some
activity on the queenside. The most
consistent move in that sense is :
18 . . . tt:la4 ! .
Essentially, exchanges are in
Black's favour so he should have
grasped the occasion. The point is
that White's knight cannot escape
the swap because the pawn on b2 is
hanging (19 .tt:lce2 tt:lb2 2 0 J''1 d2 tt:la4
2 1.f5 e5 2 2 .tt:lf3 d5) . After 19.tt:la4
ba4 2 0 J'kl �d7 2 1.f5 e5 2 2 .tt:lf3
!'lacS+, White's attack would be
over. Instead, Black played the non­
committal move :
18 ••. .if6?!
I had a similar position (with the
rook still on f8) in Shanava-Delchev,
Kusadasi 2006. It arose following
the move order 14 . . . tt:lb4 15.a3 tt:lxd3
16.cxd3 tt:ld7 17.tt:ld4 tt:lc5 18.l:"ladl
.if6 . My opponent chose the pas­
sive 19.tt:lf3 when 19 . . . hc3 ! would
have been fine for Black: 2 0 .bxc3
l:"lae8 2 1.f5 exf5.
150
19.f5 ! tre7 2 0 . tilf3
2 0 .fxe6 fxe6 2 1 . tt:l ce2 ! maintains
the tension. White chooses a wrong
setup for his knights and gradually
loses ground.
2 0 .. -l:�aeS 21.he5 gxe5 2 2 . d4
ge4 23 .fxe6 fxe6 24.gfel b4
25.axb4 gxb4 26.ge2 g6 27.trg3
.ig7+ 28.trel ge4 2 9.trd2 gf8
30.trd3 geeS 31.dS e5 32. tila2
gf7 33.'llYe 3 trf8 34.trb3 .ih6+
35.gf2 �h8 36.trd3 Wd8 37.Wb3
�e7 38. tile3 trc5 39.gffl. gcf8
40.tila4 tre7 41.We3 a5 !
Finally Black finds the decisive
improvement of his position. The
dozing bishop on b7 comes into play.
42.gal .ia6 43.gfel gbs
44.h3 gb4?
White is so passive, that Zakhar­
tsov loses vigilance for a moment.
Now 45.tt:lc5 ! ! could have turned
the tables, when already Black
should be careful - 45 . . . .ic8 ( 45 . . .
�c5 46.tt:le5 lt>g8 47.�xc5 dxc5
48.tt:lf7 mf7 49 .l:"la5±) 46.tt:le6 �xc3
47.bxc3 l:"lc4 48 .l:"la5 he6 49.dxe6
l:"le7 50 .l:"ldl l:"lxe6 51.l:"la8 lt>g7 52 .l:"la7
mg8 = . All that could have been pre­
vented by the simple 44 . . . \t>gB+.
45.We6? .ib5 46.trxe7 gxe7
47. tilc3 gxb2 48.tildl ge2
49.ggl a4 so.ga3 gee2 51. tile3
g2c3 52.gxe3 gxe3 53.tilg4 �el
54.tilg5 hg5 0-1
 The f4 System

23. Stuerzenbaum-Taimanov 24. Gofshtein-Tregubov

Bad Zwischenahn 2003 Montpellier 1998

1.e4 c5 2 . � £3 e6 3.d4 cxd4 1.e4 c5 2 . �£3 e6 3.d4 cxd4

4.tL!xd4 �c6 5.�c3 Wc7 6.f4 a6 4. �xd4 �c6 5.�c3 Wc7 6.f4
7.lL!xc6 Wxc6 8 .J.d3 J.c5 9.We2 �xd4 7.Wxd4 a6 8 .te3 b5 9 . .te2
•

b5 10 .te3 .th7 11.0-0 b4 12. �d1

• J.b7 10.a3 l"!c8 n�.tf3 h5 12 .h3

1Vt:lb1 is slow. Black can attack

immediately the centre - 12 .. .f5
13.tt'ld2 (13.J.xc5 Wxc5 14. r;t>hl fxe4
15.J.xe4 J.xe4 16JWxe4 l'l:c8) 13 . . . tt'lf6
14.J.xc5 Wxc5 15. r;t>h1 fxe4 16. tt'lxe4
tt'lxe4 17.J.xe4 .be4 18. \Wxe4 = .

12 ... �f6 13..b:c5 Wxc5+ 14.�12

0-0 15.c;!.>h1 e5
White anticipates . . . tt'lf6 which
would be attacked by e4-e5. Black
must seek another retreat square
for the knight.

12 . . . h4 13.0-0 �f6 14.e5

After 14.l'l:ad1 J.e7 15J:U2 d6

16.!Ud2, Black can calmly castle.

A typical move for the Taimanov, 14 . . . �h5 15 .tb7

•

performed by the guru himself!

Black fixes a target on e4. Now
16.fxe5 \Wxe5 17.tt'lg4 tt'lxg4 18. \Wxg4
l"lae8 19.'\Wd7 J.xe4+ ( 2 0 . l"lae1 f5)
would have been the lesser evil.
16. �g4? ! �xg4 17.Ybg4 f5
18.We2 (18.exf5 e4+) 18 . . . fxe4
19.J.xe4 d5+ 20 .if3 e4 2 1 .tg4
• •
The enemy knight will be
cramping White's kingside from
g3 so Anand preferred to kill it on
the spot: 15.J.h5 l'l:h5 16.l'l:ad1 J.c5
17. \Wd3 J.xe3 18. \Wxe3 \Wc5 19.l'l:d4.
Black had good play after 19 . . . f6�,
Anand-P. Cramling, Lugano 1988.

l"!f6 22 .id7 Wd6 23.J.a4 l"!xf4

• 15 ... Wb7 16.Wd3 �g3 17.l"!fe1
24.l"!xf4 Wxf4 25.l"!fl Wd6 26 .ib3 • Wc6 18 .J.f2 J.c5
l"!f8 27.l"!d1 We5 28.h3 e3 29.c3
bxc3 30.bxc3 l"!f2 31.Wd3 c;!.>f8 Black's bishop has more pros­
3 2 J�e1 Wg5 33.l"!fl d4 0-1 pects than its White counterpart, so

151
 Part 4
it would be good to keep it. Tregubov
recommends 18 . . . l2Jf5 19.E!:e2 \Wc4
20 .\Wf3 (20.\Wxc4 E!:xc4 2 l.l2Je4 dS+)
20 . . . d5+.
19.l3ad1 .ixf2 20.'i!?xf2 '1Wc5+
21.ee3 a5 22 .l3d2 f5 !
Black i s consistently paralys­
ing the kingside. Now White had to
simplify to four rooks endgame with
23.exf6 gxf6 24.1Mfxc5 E!:xcS 25.l2Je4
l2Jxe4 26.E!:xe4 = , but apparently he
underestimated the dangers to his
own king.
23.l3d4? 'i!?e7 24. 'i!?g1 b4
25.axb4 axb4 26.l2Ja4 exc2
27.l3xb4 l3c4
The fewer heavy pieces on board,
the more difficult is the defence of
the first rank. Therefore, 27 . . . E!:b8+
would have been more consistent.
28.l3xc4 exc4 29.c!Db6 eb4
30.l3d1 l3b8
The safe haven of Black's king
makes the difference in this posi­
tion. 30 . . . 'i!?t7! 3l.E!:d7 'i!?g6+ would
have underlined his advantage. The
pawns are unimportant.
152
White misses the chance to trade
queens by 32 .�a7 E!:b7 33.\WcS 'i!?f7
34.\Wb4. The ensuing endgame
could be saved: 34 . . . E!:b4 35.l2Jc5
l2Je2 36.'i!?fllt:if4 37.b3+. Now all is
over.
32 l3d5 ! - +
••• 33.�a7 l3d1
34.c!Df6 'i!?d8 35.�a8 'i!?e7 36.Wa7
'i!?d8 37.eas 'i!?c7 38.ea7 'i!?c6
39.Wa6 'i!?c5 40.Wc8 'i!?b5 0-1
25. Kosanovic-Banikas
Korinthos 1999
1.e4 c5 2 . c!Df3 c!Dc6 3.d4 cxd4
4. c!Dxd4 Wc7 5.c!Dc3 e6 6 .ie2 a6•
7.f4 b5 8. c!Dxc6 Wxc6 9 .if3 .ib7
•
10 .ie3 l3c8 ll.a3 Wc4 1 2 .Wd3
•
c!Df6 13.0-0 .ic5 14.hc5 Wxc5
15.'i!?h1 e5 16.l3ae1 (16.g4 h5 ! ?)
16 0-0
•••
17.g4 .!DeS ! 18.f5 c!Dd6! 19.l3e2
c!Dc4 2 0 .c!Dd1 l3fd8 2 1 .ig2 d5-+
•
2 2 .g5 dxe4 23.Wh3 c!Dd2 24.g6
fxg6 25.fxg6 h6 26.ru:7 c!Df3 27.c!De3
l3c6 28 .l3d7 Wb6 29.l3xd8 Wxd8
30.Wg4 l3d6 31.l3f2 eb6 0-1
 The f4 System

lLlc6 14.ia3 lLld4! 15.lLld6+ V!Jxd6

Kan Games 16.hd6 lLlxe2 17.@xe2 �g8 with a
pleasant endgame.

26. Martin Gonzalez-Delchev 9 •.• �c6 10.�f3

Benasque 06.07.2013
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4. �xd4 a6 5.�c3 V!Jc7 6.f4 b5
7..id3 �b7 8 .V!Je2 �f6
White's play looks normal, but
in fact this move puts him on the
defensive. 1 0.ie3 b4 11.axb4 ltJxb4
1 2 . 0- 0 ie7 and 1 0.lLlxc6 �xc6 are
balanced.

10 .•• b4! ll.�a4 bxa3 12.0-0

lLlb4

The computer will defend suc­

cessfully after 12 . . . axb2 ! 13.hb2
ie7 14.e5 lLld5, but against an oppo­
nent with a lower rating than mine,
I prefer the most solid approach.
9.a3
13.bxa3 �xd3 14.cxd3 �c6
White prevents . . . b4 followed by
... d5 which is possible after 9.�d2.
For instance, 9 . . .b4 (9 ... .ic5 1 0 .�b3
ie7 ll.e5 lLld5 1 2 . lLl e4 lLl c6 is also
possible, 13.c3 0-0 14.0-0 g6, plan­
ning .. .f5 . However, White retains
some space advantage.) 10.lLld1 [10.
ltJa4 d5 (or 10 . . . ltJc6 ! ? ll.lLlf3 d5
12 .e5 lLle4=) ll.e5 lLle4 12 .ie3 lLld7
13 .0-0 lLldc5 14.ltJxc5 hc5 15.lLlb3
ie7= , Bologan-Rublevsky, Dort­ 15.V!!c 2
mund 2 0 04] 10 . . . lLlc6 11.lLlxc6 �xc6
12.lLlf2 .ic5 13.ie3 0-0 14.g4 he3 15.lLlc3 ie7 16.id2 0-0 17.�fc1
15.�xe3 d5 16.e5 ltJe4+, Cuartas­ (17.d4 d5) 17 . . . �fc8 and 15.lLlb2
Kogan, Catalonia 2 0 13 . ic5+ 16.ie3 he3+ 17.�xe3 �b8
maintained equality.
Critical should b e 9.e5 b4
The queen is misplaced on the
10.exf6? ! [10.lLle4 lLlxe4 ll.he4
open c-file and I could try to ex­
ixe4 (ll . . . lLlc6 ! ) 12 .V!Jxe4 lLlc6= is
ploit it with 15 . . . i.d6 ! ? 16.e5 hf3
known from the �f3 set-up. ] 10 . . .
17.�xc7 hc7 18.�xf3 lLld5 19.id2
bxc3 ll.fxg7 hg7 12.lLlf5 if6 13.b3

153
 Part 4
0-0. White has weak pawns in the
centre.
15 ••• gcs 16.c!l)e5?
White neglects development.
16 ..ie3 .id6 17.e5 .ix£3 was only
marginally better for Black. Now
I'm in complete command.
16 J.b5 17.'�xc7 �c7 18.c!l)b2
•.•
J.c5+ 19.'.t>hl 0-0
Following the same lazy ap­
proach of avoiding concrete deci­
sions. l9 . . . c!l)xe4 2 0 .a4 c!l)f2 + 2 U!xf2
.ixf2 2 2 . axb5 axb5 brings about a
serious edge because my rooks have
invasion squares. My quiet move is
very unpleasant for White, too.
20.a4 .ic6 21.c!l)xc6 gxc6
Amazingly, White is practically
lost here. The only way to keep the
material even would be 2Vt:ldl, but
then 22 . . . d5 23.e5 lt:Jd7 or 22 .. J'!b8
would assure me of a total domina-
154
tion. My opponent decided to part
with a pawn right away. The rest of
the game is irrelevant to the open­
ing.
2 2 . c!l)c4 .id4 23.gbl c!l)xe4
24.c!l)a5
Test your self!
24 . . . lt:Jg3 + ! 25.hxg3 !!cS produc­
es a rare mating threat in a simple
endgame. My continuation 1s no
less effective though.
24 c!l)f2 +
••• 25J�xf2 gxcl +
26.gxcl .ixf2 27 .g3 gbs 28. '.t>g2
.ie3 29.gc7 .ib6 30.gb7 gxb7
31.c!l)xb7 '.t>f8 32.c!l)d6 J.c5 33.c!l)c4
'.t>e7 34.c!l)d2 .ib4 35.c!l)b3 '.t>d6
36.g4 f5 37.gxf5 exf5 38.c!l)d4 g6
39.c!l)b3 '.t>d5 40.'.t>f3 .id6 41.h3
h6 42.'.t>e3 g5 43.fxg5 hxg5
44. '.t>f3 .ib4 45. '.t>g3 d6 46. '.t>f3
'.t>e5 47.'.t>g3 .iel + 48.'.t>f3 .ih4
49.'.t>e3 g4 50.hxg4 fxg4 51.'.t>e2
.idS 52.'.t>e3 .ib6+ 53.'.t>e2 '.t>f4
54.c!l)d2 d5 55.c!l)b3 g3 56.d4
'.t>e4 57.a5 hd4 0-1
 Pa rt S

The Fian chetto

I offer a choice in the Taimanov: l.e4 c5 2.liJf3 e6 3.d4 cxd4 4.ltJxd4 a6
5.ltJc3 'Wffc 7 6.g3 a6 7 ..ig2 d6, which is well tested and solid, and the more en­
terprising flank attack 7 ... h5 - see the diagram. I propose a lot of new ide­
as and correct some wrong suggestions of other authors.
The fianchetto has no venom against the Kan - Black can fight for the
initiative.

155
Pa rt 5. The Fia nchetto

Main Ideas

The fianchetto system i s perhaps The first problem of Black

the most universal weapon of White
against the Sicilian. It is very solid,
but at the same time flexible . By
planting the bishop on the main di­
agonal, White significantly restricts
Black's counterplay. The good con­
trol of the centre allows him to
choose between a large number of
plans ranging from calm position­
al play on the queenside with a4-
a5 to a pawn storm on the oppo­
site flank. Unexperienced play­
ers underestimate the 6.g3 sys­
tem because White's wins are rare­
ly quick and spectacular. Howev­
er, the opposite is even more true.
Black should be very accurate with
his move order. Moreover, memo­
rising the first 12-15 moves does not
help much since decisive clashes in
this system usually occur deeply in
the middlegame. So second players
should know typical plans of coun­
terplay, which are illustrated with
practical examples in the "Complete
Games" chapter.
is how to complete develop­
ment.
The straightforward approach
does not work:
l.e4 c5 2 . �f3 e6 3.d4 cxd4
4. �xd4 �c6 5.�c3 §'c7 6.g3 a6
7 . .ig2 tt:Jf6 8.0-0 !J.e7 9 .:!�e1
My advice is to avoid this
position.
White is better after 9 ... 0-0
10.tt:Jxc6 dxc6 ll.e5 or 9 ... d6
10. tt:Jxc6 bc6 ll.e5 ! dxe5 12.Elxe5.
So we cannot comfortably devel­
op with . . . tt:Jf6, . . . !J.e7 and . . . d6. I of­
fer two thoroughly different solu­
tions in the Taimanov:

156
 The Fianchetto
A. 7 ••• d6 8.0-0 J.d7
Looking at the diagram position
you may ask: "OK, we have defend­
ed c6, but how do we castle?"
9.gel J.e7 10.tLlxc6.hc6 11.fig4
(line A61)
We simply do not castle at all !
ll h5 ! 12 .�e2 h4 13.a4 hg3
•••
14.hg3 'i!?f8!
Black has a good game. His only
concern is how to transfer the queen
to the kingside. The open h-file
will be a constant source of threats
against White's king. Keep in mind
the manoeuvres . . . �c7-d7-e8-�g8-
h7 (after g6 and lt>g7) and . . . �c7-d7-
g4 (e6) after a preliminary . . . e6-e5.
15.a5 gc8 16.J.e3 tLlf6oo.
See game 28 Fressinet-Del­
chev, Calvia ol. 2 004.
Now I'll try to arm you against
the dangerous plan of a kingside
pawn storm. It has been topical
lately. White's concept is to retreat
the knight from d4 to e2 or b3 and
push f4, g4, g5. The key moment
is that the f6-knight is deprived of
its natural square d7 while on e8 it
would be rather clumsy. Attempts
of reviving it by f6 or f5 would only
aggravate Black's situation. My ad­
vice is to hinder White's attack with
· . . . h6. Then the blockade on the
dark squares could be reinforced by
. . . tLlh7 and even g5 or e5. The play is
very sharp and interesting:
7 d6 8 . 0-0 J.d7 9 . tLlb3 b5
•••
10.a3 tLlf6 ll.'i!?hl
Alternatives are ll.i.e3 tLle5 or
ll.g4 h5 ! ? 12 .g5 tLJg4.
ll ••• J.e7 12.f4
In short, here is the general sce­
nario against White's knight re­
treats :
We answer 9 . . b5, then develop
.
our kingside and put our rook on b8
157
 Part S

or c8 before castling. If White plays draw looks the most probable re­
i.e3, we attack it with . . . lt:\e5-c4. sult:
Else, we play . . . h6 and only then can
we castle, having in mind to meet g4 7 d6 8 . 0-0 .id7 9 .�xc6 .bc6
•.•

by ... lt:\h7. The break . . . d5 may work 10.a4 �f6 ll.a5 .ie7 12 . .ie3 0-0
in some variations, but if White's
bishop is on e3, refrain from it! See
game31 Ponomariov-Dubov and
game 30 Guseinov-Vasovski.

A possible continuation is 12 . . .
1!b8 13.'\We2 0 - 0 14.g4 h6

Black has not serious problems

if he plays patiently for equalisa­
tion.

13 . .ib6 (13 .1!e1 lt:\d7 14.lt:\a4

1!ae8 ! 15.lt:\b6 f5 provides
counterplay.) 13 . . . ti'b8 14-l�el
�d7 15 . .id4 .if6 = .
Now 15.i.e3 lt:\h7 will restrain
White's attack while 15.h4 d5 16.g5 I f you want t o force your oppo­
(16.e5 lt:\e8) 16 . . . lt:\xe4 ! (do not nent to think on his own from the
open up the h-file ! ) 17.lt:\xe4 dxe4 very beginning, I recommend you
18.i.xe4 f5 19.i.g2 \!;>t7! 2 0.i.e3 1!h8 to pay attention to the following
gives good counterplay. daring variation:

This should be enough to feel B. 7 . . . h5 ! ?

confident against 6.g3. Should the
opponent surprise you with oth­
er plans like b3, or lt:\ce2 etc., just
remember that you are playing the
Sicilian ! Typical counterplay on the
queenside by . . . b5, . . . 1!c8 even be­
fore castling should be at least a
playable option.

Yet, the picture would be incom­

plete without mentioning the very The early pawn raid faces White
dry positional line A62 where the with concrete problems and seizes

158
 The Fianchetto
the psychological initiative. If White
hoped for a solid position where he
could play on the queenside with­
out any risk, he would be drastical­
ly thrown out of his zone of comfort.
To be sure, an early counter-attack
with Black is double-edged, but cur­
rent practice is encouraging. This
variation is still at an experimental
stage and it offers a vast ground for
innovations.
Main continuations are :
a) 8 .h3 h4 9.g4 i.cS ! 10.lLlb3 i.a7
11.'1Mfe2 lLlge7 12.f4 bS 13 .i.e3
13 . . . g5 ! ?
Black gains control over the dark
squares.
b) 8 . 0-0 h4 9 J'!e1 hxg3 10.hxg3
d6 ll. lLlxc6 bxc6
We have seen a similar posi­
tion Oine A61), but Black had a bish-
op on c6. Here, the bishop can go to
b7 while the king will feel quite safe
on the queenside. The play is dou­
ble-edged.
The Kan is much easier to play
because Black solves the queenside
development problem with :
l.e4 c5 2 . c!Llf3 e6 3 . d4 cxd4
4. c!Llxd4 a6 5.c!Llc3 ftc7 6.g3 J.b4!
A plus for the Kan :
a n easy queenside development.
A glance in the database re­
veals that White is quite far below
the 50% which is certainly unusu­
al for a modern opening. Of course,
Black can confidently transpose to
the Taimanov by 6 . . . lLlc6, but objec­
tively, the bishop sortie is the better
choice. We can then put the queen's
knight on d7 and push . . . eS. This
set-up is known from the Najdorf,
but here our dark-squared bishop is
not on e7 which is undoubtedly in
our favour. For instance :
7. c!Llde2
7.i.d2 lLlf6 8.i.g2 lLlc6 9.lLlb3
gives Black a pleasant choice be­
tween 9 . . . 0-0 ! ? 1 0 . 0-0 d6 11.f4 hc3
159
 Part S

12 .hc3 eS and 9 . . . �e7 10.0-0 0-0 !

ll.f4 (ll.g4 dS 12.exd5 exdS)

This is the point of 9 . . . ll'lbd7.

ll . . . dS ! 12.exd5 ll'lb4 with an in­
itiative.
7 . . . �f6 8 . .ig2 d6 9.0-0 �bd7
10.h3 .ic5 ! ll.g4 h6
By not castling, Black forestalls
the enemy attack on the kingside.
Thus he gains time to complete his
queenside development and pre­
pare a break in the centre or a coun­
ter-attack against the enemy king.

Weapon of choice

The Kan should be the weapon of choice against fianchetto fans. Without a
knight on c6, Black is more flexible. By attacking the c3-knigt with 6 . . . �b4,
he practically forces the enemy to retreat his active knight from d4. As a re­
sult, White is deprived of the unpleasant option of ll'ld4xc6 which is a cause
of constant concern in the Taimanov.

160
Pa rt 5 . The Fia nchetto
Taiman ov - Step by Step
l.e4 c5 2.�£J e6 3.d4 cxd4
4.�xd4 �c6 5.�c3 �c7 6.g3 a6
This move is not obligatory, but
it is more flexible than the alter­
native 6 . . . d6 which allows 7.lt:\db5
'Wb8 8 .a4. This position is practical­
ly unexplored. The game Romero
Holmes-Wolff, Palma de Mallorca
1989, continued 8 . . . .ie7 9 . .ig2 lt:\f6
10.0-0 0-0 11.�e2 l"1d8 ? ! 12 . .ie3 a6
13 .lt:\a3 d5? 14 . .ib6 l"1d6 15.a5±.
Black could improve with 11 . . .
a6 12.lt:\a3 b 6 13 . .ie3 �c7, when
14.g4 ! ?oo starts the typical attack
on the kingside, but in an original
set-up with a knight on a3 . (White
should not delay this advance since
Black would be ready with . . . d5 !
- 14.f4 l"1b8 15.l"1ad1 l"1e8 16.g4 d5 !
or 14.f4 l"1b8 15.g4 d5 16.exd5 lt:\xd5
17.lt:\xd5 exd5 18 .hd5 l"1e8;:;) .
There i s one more question that
still has not a· definitive answer:
what does Black benefit from de­
laying . . . a6. Let's check 7 . .ig2 lt:\xd4
8.�xd4 .id7 9.0-0 lt:\f6 10 . .ig5 .ie7
The trade of knights gave White
some spatial advantage in the cen­
tre. The main battlefield will be the
left wing so the queen's rook should
stay on a1 while the other one goes
to d 1 :
ll.l"1fd1 ! ? a 6 12 . .if6 gf6 13.�d2
h5 14.h4 l"1c8 15.l"1acloo, Kaidanov-J.
Horvath, Calcutta 2 0 0 0 . Black has
active rooks, but his pawn chain is
damaged. The h5-pawn would be a
tangible weakness in an endgame.
In my opinion, White's game is
somewhat easier.
161
 PartS
7.i.g2
Black has at least four different
set-ups against the fianchetto. In
line A. 7...d6, I offer you a well test­
ed and flexible approach, based on
the Scheveningen pawn structure.
Line B. 7...h5!? introduces a rare at­
tack which is nearly unexplored. It
will face your opponent with a dif­
ficult choice, Statistics confirm that
White's defence is not trivial. He
scores only 41% although the sam­
ple of only 61 games is too small to
be representative.
I rejected 7...l2Jf6 8.0-0 l2Jxd4
9.�xd4 i.c5 10.i.f4 d6, because this
is one of the most drawish variation
in the opening theory! Black practi­
cally has not won a single game at a
higher level. A typical position aris­
es after 1l.�d2 h6 12.l"i:ad1 e5 13.i.e3
�e6 14.hc5 dxc5. White will then
play l2Jd5 (or f4 first), recapture on
d5 by the e-pawn, and occasionally
will convert his tiny advantage.
7...l2Jf6 8.0-0 �e7 is a playable,
but passive alternative. Black's
problem is that after 9.l"i:e1!,
162
he cannot complete develop­
ment without considerable conces­
sions in space since 9...d6 runs into
10.l2Jxc6 bc6 1l.e5! dxe5 12.l"i:xe5.
This position was topical 40 years
ago. Exhaustive testing proved that
Black is unable to compensate for
the split queenside pawns.
A. 7 ..• d6 8.0-0
8.l2Jxc6 bxc6 9.0-0 l"i:b8 10.l2Ja4
�e7 ll.c4 c5 12.�e2 �f6 13.�e3 �d7
(13...l2Je7!? 14.f4 l"i:b4 15.lLlc3 �d4)
14.lLlc3lLle7 15.f4 �d4 16.hd4 cxd4
17.l2Jd1 0-0 18.b3 a5=, Smirin-Jan­
jgava, Vilnius 1988.
8...i.d7
The only downside of this move
is that the bishop occupies a square
which might be useful to the king's
knight. We'll discuss that issue lat­
er- see the line 8...�d7 9.lLlde2lLJf6
10.h3 �e7 11.g4. However, the alter­
natives have major drawbacks.
a) 8...l2Jf6 9.l"i:e1 �d7 10.l2Jxc6,
when the options are: 10...hc6
ll.l2Jd5 hd5 12.exd5 e5 13.f4 �e7
14.�e2±; 10...bxc6 ll.l2Ja4±, fol-
 The Fianchetto
lowed by c4, and the bishop is mis­
placed on d7. Both evaluations are
not final. They just reflect the cur­
rent state of theory, but the play is
quite tangled.
b) 8 . . . lt:lxd4. Black wants to ex­
change the d4-knight before it ran
away! 9 .�xd4 .id7
10J'!d1
10 . .if4 leads to unclear conse­
quences after 10 . . . .ic6 ll.lt:ldS exd5
12.exd5 .id7 13J'!fe1 lt:le7 14J!e3 f6
15J!c3 �d8oo.
10 ... lt:lf6 (perhaps Black could de­
lay the knight move with 10 . . . .ic6 ! ?)
ll . .igS .ie7 12 .hf6oo.
The early exchange on d4 leads
to a practically new branch of the­
ory which is waiting to be explored.
However, I think that White's play
is natural and easy.
In the diagram position, White
has tried nearly all possible moves
and plans. I'll examine the most
reasonable of them:
Al. 9 . lt:l de2 ; A2 . 9.lt:lb3; A3.
9 . .ie3 ; A4. 9J!e1; AS. 9 .a4 ; A6.
9.lt:lxc6
9.lt:lce2 is totally harmless in
view of 9 ... lt:lf6 10.b3 (10 .c4 lt:lxd4
1l .�xd4 .ie7 and Black easi­
ly achieves . . . bS later) lO . . . bS with
good play.
Al. 9 . lt:l de2 ·
White saves his knight from ex­
change and transfers it to g3 after a
preliminary h3, g4. Practical results
are quite alarming for Black. Com­
monly the second players do not
pay due attention to White's threats
and quickly fall under attack, like in
the following game :
Inkiov-Tsorbatzoglou, Chania
1994
l.e4 c5 2.lt:lf3 e6 3.d4 cxd4
4.lt:lxd4 lt:lf6 5.lt:lc3 lt:lc6 6.g3 d6
7 . .ig2 .id7 8.0-0 .ie7 9.lt:lde2 0-0
1 0 .h3 a6 ll.g4 �c7 12 .lt:lg3 bS 13 .g5
lt:le8 14.f4 i!d8 15.<it>h1 b4 16.lt:lce2 f6
17.h4±
163
 Part S
17 . . . fxg5 18.hxg5 <;!;>h8 19.f5 ltJeS
20.ltJf4 dS 2 1.W/h5--t.
In my opinion, the whole set­
up with the knight residing on e8
is very passive and leaves the op­
ponent a free hand on the king­
side. The other typical approach is
to meet g4 with h6 and gS. That is a
double-edged plan, but Black's king
in the centre is considerably more
vulnerable than his White counter­
part.
I think that Black should em­
ploy a defensive design which is
frequently seen in the hedgehog
system (the plan with f4, g4) . He
prevents g4-g5 with h6 and when
White plays f4, the f6 knight goes to
h7! Then Black could safely castle
and expand on the opposite wing.
I only hesitate about the early b7-
b5. It is risky, but enterprising. You
must be sure to push it before White
played g4, in order to force the an­
swer a3 . Otherwise the c3-knight
will replace its teammate on e2.
Let's substantiate these reflec­
tions with concrete variations :
9 . . . b5 ! ? 10.a3 �f6 ll.h3 J.e7
12 .g4 h6 13.f4
164
13.�g3 0-0 14.f4 ltJh7 leads to
the same position.
13 . . . �h7!?
Black sets a flexible defence line
on the kingside.
13 ... 0-0 14 . .ie3 l:%fd8 15.g5 hxgS
16.fxg5 ltJh7 17.h4oo gives White the
initiative in view of the threat of
. . . �f4 followed by . . . g6.
After the text, it is unclear how
White could expand further. The
only way is apparently h3-h4, but
Black is able to hinder it. For in­
stance :
14.W/e1 0-0 15.h4 leaves c2 un­
guarded after 15 . . . b4. Following
16.lZld1, Black could take an edge
with 16 . . . bxa3 17.g5 (17.bxa3 gS ! )
17 . . . axb2 18.hb2 hxg5 19.fxg5 ltJe5.
The quiet 14 ..ie3 0-0 15.ltJg3 (15.
W/el �d8 16.l:%dl l:%c8) 15 ... l:%fd8 also
looks appealing to Black who could
combine his counterplay with the
manoeuvre ltJh7-f8-g6.
A2 . 9 . �b3 ! ?
This has been the most topical
line lately. White's concept is sim­
ilar to line Al, but the knight on b3
 The Fianchetto
is a bit distant from the kingside.
On the other hand, White does not
need h3 in order to push g4. White
could also play a4-a5 to restrict
queenside counterplay. So I pro­
pose to start with :
9 . . . b5
I consider similar positions
without the insertion of 9 . . . b5 10 .a3
in the annotations to game 31 Po­
nomariov-Dubov, Tromso 2 0 13 .
10.a3
10.f4 b4 1l.tLle2 tLlf6 1 2 . a3 e5
should be fine. The early . . . b4 is not
so good against a developing move
as 10 . .ie3, since White could open
the c-file in his favour with c2-c3 .
However, .ie3 also has drawbacks.
It provokes 10 . . . tLle5 ! ? , when ll.tLld2
tLlf6 12 .h3 Ek8 13.f4 tLlg6 looks quite
safe. 10 . . . tLlf6 11.f4 gcs is also possi­
ble and White will probably trans­
pose to the main line with 1 2 . a3 be­
cause 12 .g4? could be put to the test
with 12 . . . h5! 13.h3 hxg4 14.hxg4 b4
15.tLla4 e5.
10 . . . tLlf6 ll.�h1
White does not seem to have
more useful moves since ll . .ie3 will
face again ll . . . tLle5. ll.g4 is com­
mitting as Black has not castled yet.
The standard retort would be ll . . . h6
12.f4 .ie7 13.'1We2 gbs, but ll . . . h5 ! ?
looks also very tempting: 12 .g5 tLlg4
13.�h1 g6 14.f4 V:Ma7.
ll.f4 i.e? 12 . .ie3 gcs (I'm not
sure about this move. In many
variations the rook stands better on
b8.) might transpose to the main
line after 13.'it>hl. An independ­
ent variation is 13.�e2 b4 14.axb4
tLlxb4, when 15.h3 0-0 16.g4? stum­
bled into 16 . . . d5 in Beshukov-Tre­
gubov, Krasnodar 1998. As a rule,
we should always contemplate . . . d5
against g4, even when it is a real
pawn sacrifice.
ll . . . .ie7 12.f4
12 . . . gbs
From this square, the rook sup­
ports the b-pawn. It also prevents
possible .ib6 if White set the bat­
tery .ie3-�f2 , and indirectly hits the
b3-knight (with the threat of . . . b4,
. . . tLlxb4, . . . tLlxc2) .
12 . . .1':k8 i s a good alternative.
See game 31 Ponomariov-Dubov.
Another legitimate option is
immediate castling. After 12 . . . 0-0
13 .g4 h6 14.h4 (14 ..ie3 tLlh7 ! ) 14 ... d5
15.g5 (15.e5 tLle8 16 ..ie3 gbs 17JU3
f6 18.exf6 tLlxf6, with counterplay)
15 . . . hxg5 16.hxg5 tLlxe4 17 . .ixe4
dxe4 18 .V:Mh5 is not dangerous as
White is behind in development:
165
 Part S
18 . . . g6 19 .�h4 f6 2 0J'U2 fxgS
2 1.�h6 .if6 2 2 J'�h2 .ic8 23.tt'lxe4
�g7 24.tt'lxf6+ !!xf6 = .
13.�e2
After 13 . .ie3, the most practical
approach is to forestall possible g4
and h4 with 13 . . . h6 ! . Then we can
already castle since we would have
. . . tt'lh7 to g4.
13 . . . 0-0 14.g4 h6 is more risky:
a) 15.h4 tt'lh7! (15 . . . d5? fails to
16.g5 ! hxgS 17.hxg5 tt'lxe4 18 . .ixe4 !
dxe4 19.�h5, with attack) 16.g5
hxgS 17.hxg5 g6oo.
b) 15.g5 hxgS 16.fxg5 tt'lh7
is considered in game 30 Gusei­
nov-Vasovski, Struga 2009. In this
pawn structure, we push . . . g6, pos-
sibly . . . lt>h8 , and seek counterplay
with . . . gxhS. It is important to leave
our kingside pieces as they are -
eyeing gS. We can even win the gS­
pawn at some point with . . . �d8.
It would be a mistake to redeploy
them in a passive defensive stand.
13 . . . 0-0
I do not see what we could gain
166
by delaying castling any longer as
White has plenty of useful moves.
For example, 13 . . ..ic8 14.g4 h6
15 . .ie3 tt'ld7 16.�f2 tt'lb6 17.tt'ld4
tt'lxd4 18 .hd4 0-0 19.�g3 with ac­
tive pieces.
Of course, in the above line Black
can refrain from . . . h6, allowing the
thematic plan gS, �hS. However, I
like too much to play it on the white
side so I cannot recommend it for
Black.
14.g4 h6 15 .ie3
•
15.h4 can be attacked with 15 . . .
dS ! ? 16.g5 (16.e5 tt'le8), when i t is
extremely important to avoid open­
ing the h-file by 16 . . . hxg5 ? ! 17.hxg5,
because White's game is a little bet­
ter after 17 . . . tt'lxe4 18.tt'lxe4 dxe4
19 . .ixe4 fS 20 . .ig2 g6 2 1.!!f3 ! , e.g.
2 1 . . .!!f7 2 2 .!!h3 !!h7 23.!!xh7 lt>xh7
24 . .id2 !!h8 25.\t>g1 \t>g8 2 6.!!e1 eS
27.fxe5 .ie6 28 .�f2 . Instead, Black
should continue 16 . . . tt'lxe4 ! 17.tt'lxe4
(17 . .ixe4 dxe4 18 .gxh6 fS 19.!!g1
!!f6 2 0 .�g2 g6) 17 . . . dxe4 18 . .ixe4
f5 19 . .ig2 . l!ff7! 2 0 . .ie3 !!h8, with
counterplay.
Amazingly, I have not any prac-
 The Fianchetto

tical example with this particular .ie7 11.tt:lb3 bS 12 .g4 hS 13 .gS, when
position in my database. Black has 13 . . . tt:lg4 is with tempo.
two radically different approach­
es here. The first one is to carry on
... dS and play in the centre :
1S . . . b4 16.axb4 tt:lxb4 17.tt:ld4
dS 18.eS (18.gS hxgS 19.fxgS tt:lxe4
20.tt:lxe4 dxe4 2 1.he4 g6) 18 . . .
tt:lh7 19.�f2 f6. This may b e not
the most exact move, but it gives
us immediate counterplay. 2 0 . exf6
tt:lxf6 2 1 .h3 eS 2 2 .fxeS tt:le4 23 .1klre1
1"!xf1+ 24 . .txf1 �xeS 2S.tt:lxe4 dxe4
26 . .ig1 .id6 27.1"!d1 .ic7 2 8 . .ig2 tt:ldS
29.1klfxe4 1klrxe4 3 0 . .ixe4 tt:lf6 = .

The other alternative i s more in­ 10 .�e2

teresting: 1S ... tt:lh7! ? . I analyse the
same position (only the rook is not
on b8, but on c8) in the comments
to game 31 Ponomariov-Dubov.
In our case, the game may con­
tinue 16.�f2 aS 17.h4 b4 (17 . . . gS ! ?
18.hxgS hxgS 19.eS d S 2 0 .tt:lcS .ic8
21.tt:ld3oo) 18.axb4 axb4 19.tt:le2 eS
20 .gS exf4 (After 20 . . .hxgS 2 1.fxgS
g6, we reach the same pawn struc­
ture as in game 30 Guseinov-Va­
sovski.) 21.tt:lxf4 hxgS 2 2 .tt:ldS �d8
23.hxgS .ie6 with a complex, bal­
anced game.
10.f4 .ie7 11.tt:lb3 bS and 10. tt:lb3
are considered in line A2 .
10.tt:lxc6 transposes to line A62 .
1 0 . . . i.e7 ll.f4
From the other options, only
ll.a4 0-0 12 .1"!ad1 is of independent
significance. Black equalises with
typical Sicilian methods : 12 . . . tt:lxd4
13 .hd4 eS 14 . .ie3 bS 1S.axbS axbS
16.-igS 1"!fc8 17.�d3 .ie6 18.tt:lxbS
�xc2 19.tt:ld6 1klfxd3 2 0 .1"!xd3 1"!cb8=.

ll . . .l'k8 !

A3. 9 .ie3
• If you read line A2 carefully,
A normal developing move
which keeps all the options open.
However, it is too tentative for a
sharp opening as the Open Sicil­
ian. Black gets time for develop­
ment. Another downside of the text
is seen in the variation 9 . . . tt:lf6 10.f4
you should have noted that we pre­
fer to develop the queenside before
castling. Do not forget that White
can always retreat the knight. For in­
stance, 12 .tt:lb3 bS 13.a3, when sim­
plest is 13 . . . h6 14.'i!th1 0-0 1S.g4 tt:lh7.
12 . gadl b5 13.a3 0-0

167
 Part S
A devoted Sicilian player could
only dream about such position.
Black is all set up for expanding on
the queenside. 14.fS lt:Jxd4 1S.hd4
eS 16.i.e3 aS 17.i.gS i.c6 would be
roughly equal. White can play lt:Jb3
or allow . . . lt:Jxd4, for example :
14.\t>hl tl:\xd4 15.h:d4 e5
16.i.e3 i.e6
The chances are even. Medina­
Tal, Palma de Mallorca 1966, went
17J:'lfe1 i.c4 18.�d2 aS.
A4. 9,ge1
The play usually transposes to
other lines. I examine here some
specific variations.
9 . . .i.e7 10.tl:\b3 b5 ! ?
I recommended the same active
approach against the 9.lt:Jb3-line.
The current position hides a slight
nuance - instead of a bishop on e3
White has the rook on el. The latter
supports the threat of eS, but also
makes Black's counterplay with
. . . hS more efficient because in that
case the rook is needed on fl.
168
Most of the games feature 10 . . .
lt:Jf6, when ll.a4 follows the idea of
the previous retreat. White wants to
bind the left wing by aS. If now Black
answers ll . . . b6 12 .£4 0-0 13 .g4 (13.
<i>h1 enables another set-up: 13 . . .
l'!fd8 14.g4 .ie8 1S.gS lt:Jd7) 13 . . . .ic8
14.gS lt:Jd7, we'll have a typical Sch­
eveningen position. However, a4-
aS is not so dreadful: 11 . . . 0-0 12.aS
lt:JeS 13.h3 l'!ac8 14.g4 h6 with excel­
lent play. White is unable to break
trough the dark-square blockade on
the kingside.
ll.a3 (ll.f4 fails to ll . . . b4
12.lt:Je2 lt:Jf6) ll . . . ttlf6 12.f4
White's strategy is a failure.
Black can choose between 12 . . . l'!d8
and the more enterprising, but risky
12 . . . l'!c8. In both lines 13.eS does
not work. 13.g4 is not much bet­
ter due to 13 . . . hS 14.gS (14.h3 hxg4
1S.hxg4 gS ! ? 16.fgS lt:Jh7. This vari­
ation would have been bad against
the i.e3-l'!f1 setup ! ) 14 . . . lt:Jg4.
The big difference between the
two rook moves is seen in the vari­
ation 13 . .ie3. White renews the
threat of g4. With the rook on d8,
 The Fianchetto
Black could castle 13 . . . 0-0 since
14.g4 is met by 14 . . . �c8 15.g5 llJd7.
12 .. J'k8 13 ..ie3 practically for­
ces Black to adopt the defence with
13 . . . h6 14.g4 llJh7 15.ygd2 (15.�f2
llJf8) 15 . . . g5oo with very sharp play.
In game 27 Hoogendoorn-Lau­
tier, Algarve 1999 Black realised a
dubious mix of both plans.
AS. 9 . a4
9 ••• llJf6
I have noted before that the
exchange 9 . . . llJxd4 10 .ygxd4 llJf6
ll . .ie3 (ll . .igS .ie7 12.l:!fd1 l:!c8 = )
.ie7 12 .a5 leaves Black without
· llJb2 18.e5 �c3+.
much counterplay.
lO. llJb3
The bad news now is that 10 . . .
b6? ! ll.f4 .ie7 12 . .ie3 followed by
g4 leads to a known variation from
the Scheveningen with a clear ex­
tra tempo for White who commonly
spends time on .ie2-f3-g2 .
Black is in a similar tempo-down
situation after 10 . . . llJ a5? ! ll.llJaS
�as 12 . .id2 �c7 13.a5.
Now, here is the good news :
Black can ignore the threat of a4-a5 !
10 J.e7 ll.a5 (ll.f4
• .. llJaS)
ll llJe5! 1 2 .f4 �c4
• •.
13.yge2
13.g4 is risky when Black's king
is still on e8. After 13 . . . h6 14.yge2
( 14.h4 bS) 14 . . . l:!c8 15.'i!?h1 g5, Black
could try to gain the initiative count­
ing on the excellent position of his
knight on c4 (15 . . . llJh7 is worth con­
sideration too, e.g. 16.ygf2 0-0 17.h4
g5oo). Serper-Epishin, Tashkent
1987, saw further 16.fxg5? ! (16.h3
gf4 17.l:!f4 llJh7oo) 16 . . .hxg5 17.hg5
After the text move, Black should
decide how to face White's pawn
storm on the kingside. I think it is
good to raise a pawn barrage on the
dark squares:
13 . . . e5 ! ?
Remember that castling and re­
treating the knight to e8 is clearly
in White's favour. More interesting
is the old suggestion of GM Ceba­
lo to sacrifice a pawn with . . . d5. His
169
 Part S
analysis in Informator 33/312 exa­
mines (by transposition) : 13 ... 0-0
14.g4 d5 ! 15.e5 (15.exd5 l:l:fe8 ! ) 15 ...
tt'le4 16.tt'lxe4 dxe4 17 . .ixe4 i.b5
18.i.d3 l:l:ad8�. However, no one has
tested it for more than twenty years
already.
14.f5 �!k8
14 . . . h6 is also possible, intend­
ing to meet 15.g4 by 15 . . . b5 16.axb6
�xb6 17.c;th1 l:l:c8 .
15.g4 h6
I like Black's position. In the
game Trichkov-Peschardt, Lyngby
1991, White attempted to fulfil his
main strategic goal with 16.h4 �d8+
17.g5 hxg5 18.hxg5 tt'lh5 19.�g4 tt'lf4
20.g6, when 2 0 . . . l:l:h4 ! + would have
been sobering.
A6. 9. ttlxc6
White often resorts to this ex­
change after 9.l:l:e1 i.e7. Perhaps
such a move order is more accurate,
because it enables an interesting
positional sacrifice against 1 0 .tt'lxc6
bxc6. White could follow up with
ll.e5 ! ? dxe5 12 .�e2 or ll.tt'la4 l:l:b8
(ll . . . l:l:d8 12 .�e2 i.c8 13.e5 d5 14.c4;!;
Timman-Marjanovic, Zagreb 1985)
12 .e5 dxe5 13.b3 tt'lf6 14.i.b2;!; .
9 . . . .b:c6
9 . . . bxc6 is probably underrated.
It leads to complex play:
170
A61. 10.l:l:e1 ; A62 . 10 .a4
VVe have reached the key po­
sition of the g3-system. . Perhaps
White players like it so much be­
cause one does not need any home
analysis in order to reach and play
it. White has some spatial advan­
tage, better development and no
weaknesses. This is one of the safest
line in the Open Sicilian ! Still, Black
keeps refining his defence and cur­
rently has no considerable theoreti­
cal problems.
A61. 10.gel
VVhite does not renounce the a4-
a5 plan. On the contrary, most of
the games transpose to line A62 in
the next moves. However, the text
move requires from Black concrete
knowledge of some sharp variations
which I examine in the next lines.
10 ••• i.e7!
10 . . . tt'lf6 ll.tt'ld5 .ixd5 12.exd5 e5
13.f4 i.e7 14.�e2 with some pull.
 The Fianchetto
Now ll.a4 leads to A62 .
ll . .if4 i s a n overt proposition of
a draw: ll . . . ll'lf6 (you can keep some
tension at your own risk by 11 . . . e5
12 . .ie3 lt'lf6 13 . .ig5 ll'lg8 14. \Wd2 h6
15 . .ie7 lt'le7) 12 .e5 dxe5 13 . .ixe5
'Wb6 14 . .id4 \Wc7 15 . .ie5. Then 15 . . .
'Wc8 1 6 . .ixf6 .ixf6 17.lt'le4 .ixe4=
looks like a dead draw, Shevelev­
Lerner, Israel 2003.
White's only hope to punish the
opponent for his particular move
order is :
11.\Wg4 h5! 12 .We2 h4
12 . . . b5? ! is a serious mistake.
In this system, White's main plan
is based on a queenside expan­
sion. Black's pseudo-activity goes
half way the opponent's intention.
White will easily break-through by
a4 with an edge.
13.a4 hxg3 14.hxg3
A fashionable position . I play
it with both sides ! White is going
to push aS and introduce a minor
piece on b6. Then he will try to ad­
vance the c- and b-pawns.
Black's counterplay is not evi­
dent. In all events he will display
activity on the opposite wing. First-
. ly he should decide where to evac­
uate the king. Currently . . . <;t>f8, in­
tending . . . g6 and . . . <;t>g7, looks pref­
erable to castling.
The activation of Black's queen
is another urgent matter. On the
left wing it is constantly harassed
by the enemy's bishop, rook and
knight, so it has to find a way to
the other half of the board. You can
see in my game against Fressinet
the manoeuvre . . . �c7-d7-e8 having
in mind . . . �g8-h7 after . . . g6 and
. . . <;t>g7. Another route for the queen
is . . . \Wc7-d7-g4 (e6) after a prelimi­
nary . . . e6-e5.
As a whole, Black has fair chan­
ces. See the annotated games 28
and 29 in the "Complete Games"
section in order to better under­
stand the options of both sides.
I'll confine myself here only to the
problem of the best move order in
the diagram position.
14 . . . <;t>f8 !
14 . . . ll'lf6 is a normal, frequent-
171
 Part S

ly seen move. Its downside is that Delchev, Calvia ol 2 0 04 . See the

White could answer 1S.i.f4 eS "Complete Games" chapter.
16.i.gS \Wd7 17.i.xf6 i.xf6. Black
may hold this position, but his win­
ning prospects are null, for instance A62 . 10.a4 �f6 ll.a5 .ie7
after 18J!ed1 (18.b4 ! is better) 18 . . . 12 .i.e3 0-0
i.d8 19.aS.
12 . . . lt:\d7 gives White a tempo for
15.a5l'k8 c4 as in Leko-Zvjaginsev, Wijk aan
Zee 199S, which went on:
Black should evade the trap 1S . . . 13.lt:la4 0-0 14.lt:\b6 Ei:ae8 1S.c4!
g 6 16.i.e3 lt:lf6 17.i.b6 \WeB 18 J'!ad1 The text move keeps the hit on e4,
@g7 19.lt:ldS ! which is based on so White must include Ei:e1 before
the hanging state of the e7-bish­ heading with the knight for b6.
op. Against Berescu, I answered
19 . . . i.xdS 20.exdS eS 21.c4 i.d8,
when 2 2 .b4 ! i.xb6 23.axb6 aS 24.cS
would have given White an initia­
tive. 20 . . . lt:\xdS 21.i.xdS i.f6 2 2 .�f3
\Wd7 23 .i.d4 i.xd4 24J'!xd4 does not
solve the problems either. This ex­
ample explains the reason behind
the otherwise awkward continua­
tion 1S . . . Ei:e8 ! ? 16.i.e3 lt:lf6 . I dis­
cuss it in game 29 Timman-V. Mi­
lov, Biel 199S.
13 . .ib6
16 .ie3 �f6oo.
•

I have not seen any new ideas in

this variation lately. White is very
solid, but Black has no weaknesses.
A popular alternative is:
13.Ei:e1 lt:ld7 14.lt:\a4 Ei:ae8 ! . This
is is Black's only active plan. It is
enough for obtaining sufficient
counterplay:
1S.lt:lb6 fS 16.lt:\xd7

Alternatives are :
a) 16.exfS i.xg2 17. @xg2 Ei:xfS ;
This rich double-edged position
occurred in game 28 Fressinet- b) 16.i.d4, when 16 . . . f4 ! ? looks

172
 The Fianchetto

best. 16 . . . e5 is not bad, but it offers Black's position is somewhat

White the d5-square - 17 . .ie3 f4 cramped so exchanges should re­
18.tt:ld5 .b:d5 19.�xd5 r4lh8 2 0 . .id2 lieve his defence.
b5 ! (White develops a very strong
initiative after 20 . . . fxg3 2l.hxg3 16 .ixf6 c!Dxf6
• 17.c!D a4 c!Dd7
�xc2? ! 2 2 J'!ecl �xb2 23J'%abl �d4 18.�d4 .b:a4 = .
24 . .ie3 !;!;) 2 l . axb6 �xb6 2 2 .�b3
and White even has an initia­
tive after 22 . . . fxg3 23.�xb6 gxh 2 + B . 7 . h5 ! ?
..

24.r4lxh2 tt:lxb6 25.r4lg3 l'%c8 2 6 . .ie3

tt:lc4 2 7 . .icl l'%c6 = .

1 6 . . . �xd7 17.�d3 (17 . .ib6? ! f4 ! t)

17 . . . fxe4 18 . .b:e4 .b:e4 19.�xe4
d5 = . Also equal is 19 . . . .if6 2 0.�b4
d5 2 l.c3 l'%c8 2 2 .l'%ad l = .

13 ••• �b8 14J:�el

14.�e2 is similar: 14 ... tt:ld7

15 ..id4 �c7 16.l'%fdl l'%ac8 17.b3
if6 = , Sharafiev-Zakhartsov, Kazan Black aims t o create tension on
2010. the kingside and hinder the enemy
plan of castling short and then ad­
vancing the g-pawn. Lately, . . . h5
has become very popular in the Naj­
dorf, but in that opening it has only
restrictive functions. In our case,
this advance pursues much more
active goals : we hope to open the
h-file and use it for an attack. We
may castle long or leave the king in
the centre (f8 seems a safe place).
Should White answer 8.h3 h4 9.g4,
we'll base our play on the weakness
14 c!Dd7
of the squares f4 and g3 .
•••

Jeroen Bosch recommends this

One way or another, Black will
attack in his electronic book SOS
exchange a couple of minor pieces
File 2: Sicilian Taimanov, but with
and hold the ensuing positions.
a totally wrong argumentation.

15 .id4 Af6
Main branches now are :
•

Bl. 8 .h4? ! ; B 2 . 8.h3 ; 8.B3. 8.0-0

173
 Part S
8.lt:lf3 and 8.lt:lb3 lack any sense.
8 . .igS counts only on 8 . . . .ie7?
which is a positional mistake before
White has castled short since he can
exchange on e7 and follow up with
'Wd2 , 0-0-0.
However, White's bishop on gS
will give us a valuable tempo for . . .f6
at some point. Reasonable options
are :
8 . . . lt:lge7 intending 9.'Wd2 lt:lxd4
10.'Wxd4 'Llc6 ;
8 . . . f6 9 . .if4 eS 10.lLlc6 dc6 1l . .ie3
i.g4 12 .i.f3 E:d8 13.'We2 lt:lh6 = ;
8 . . . lLleS 9.'We2
The most principled continu­
ation now is 9 . . . i.a3 10.lt:ld1 i.b2
ll.lt:lb2 'Wc3 12 .'Wd2 'Wxb2. Let the
opponent prove his compensation
for the pawn. 13 .0-0 lt:lc4 14.'Wd3
lt:leS 1S.'Wd1 f6 16.i.e3 (16.i.f4 h4
17.lt:lf3 lt:lf7oo) 16 . . . h4. In the dia­
gram position Black can also try 9 . . .
h4 ! ? .
B l . 8.h4? ! lt:lf6
We'll try to prove that the inser­
tion of h4-hS favours Black. Let's
compare some of the main lines
without these moves :
174
7 . . . lt:lf6 8.0-0 i.cS 9.lt:lxc6 dxc6
10.lt:la4 .ia7 ll.c4 and:
7 ... hS 8.h4 lt:lf6 9.0-0 .icS
10.lt:lxc6 dxc6 ll.lt:la4 .ia7 12 .c4?
'Wxg3 ! - the g3-square is weak.
Or 7 . . . -icS 8.lt:lb3 .ia7 9.�g4 'it>f8
- White has riot 'Wg4 at all. Anoth­
er argument is that White's play is
mostly based on f4, when the g4-
square will become a gaping hole.
Therefore, I advocate:
I do not like 8 . . . i.cS 9 .lt:lb3 i.a7
10.i.f4 lt:leS in view of ll . .ixeS �xeS
1 2 .f4 'Wc7 13.'We2 d6 14.0-0-0:t
where d6 is very weak. We need to
improve this idea.
9.0-0
9 . . .i.c5 10 . .!i)xc6
10.lt:lb3 i.a7 ll.i.f4 lt:leS is al­
ready fine for Black. 1 0 .i.e3 is also
innocuous due to 10 . . . lt:l eS. 10.i.f4?
presents another example of the
dividends Black can reap from the
insertion of h4-hS : lO . . . eS ll.lt:lb3
exf4 1 2 .lLlxcS fxg3.
 The Fianchetto
10 dxc6 ll.Wd3 �g4 1 2 .�a4
•••
.ia7 13 .c4 b5
I prefer Black here.
82. 8 .h3 h4 9 .g4
· generously presenting tempi. White
9 ••• .ic5 !
White's kingside dark squares
are weakened. Jeroen Bosch con­
siders only:
9 . . . ltlge7? ! . It looks consistent to
put a knight to g6, but this devel­
opment is a bit slow, and, moreo­
ver, passive. White castles long and
takes the initiative :
10.ltlde2 ! ltlg6 11 .�e3 !
Bosch completely missed this
idea. ll.f4 b5 (ll . . . �c5 12 .e5 d5
13.exd6 hd6 14.li:Je4) 12 .�e3 ltla5
13.b3 �b7 14.Wd2 !lc8 15.0-0-0
�c6 ! is double-edged.
I have analysed in the diagram
position :
a) ll . . . ltla5 1 2 .b3 b5 13.Wd2 �b7
(13 . . . �a3 14.0-0 0-0 15.!lad1) 14.a4 !
(14.0-0-0 !lc8 15.f4 �c6 ! ) 14 . . . b4
15.ltla2 ltlc6 16.0-0-0 �e7 17.f4. I do
not see clear counterplay for Black
in this position.
b) 1 1 ... ltlce5 12 .b3 �a3 13.f4 ltlc6
is rather extravagant. Black is too
can calmly develop, or open the
centre with 14.e5 d5 15.exd6 hd6
16.ltle4 �e7 17. 0-0, intending f4-f5.
As a whole, White achieves in
these lines comfortable and har­
monious development. His bishop
stays excellently on e3.
9 ... �c5 addresses exactly this is­
sue since 1 0 .�e3 Wb6 would be po­
sitionally fine for Black after 11.ltla4
V!1/a5+ 1 2 .c3 hd4 13.hd4 ltlxd4
14.Wxd4 e5 or ll.ltlce2 ltlxd4 (11 . . .
�xb2 1 2 .ltlxe6 �b4+ 13.c3 dxe6
14.cxb4 �xb4+ 15.�d2oo) 12.hd4
hd4 13.Wxd4 �xd4 14.ltlxd4 ltle7
15.0-0-0 ltlg6 16.ltle2 �e7 17.f4 d6.
In both lines White remains with
a "bad" light-squared bishop and
kingside pawns fixed on the same
colour.
10. ltlb3
If White's knight has to retreat,
it is better to do it with tempo.
10.ltlde2 b5 11.0-0 �b7 1 2 .�f4 d6
175
 Part S
13.�d2 4Jge7 14.l'!adl ttJeS is even
slightly better for Black.
10 . . . J.a7 11.�e2 4Jge7 12.f4
12 . .ie3 4Jg6 13 . .ixa7 l'!xa7 would
make a horrible hole on f4.
12 •.• b5
12 . . . 4Jd4 13.4Jxd4 .ixd4 14 . .id2
d6 15.0-0-0 .id7 is also possible, but
I prefer to complete development.
13 . .ie3
13.e5 .ib7 14.4Je4 0-0 leads to
a curious position where White is
overextended and should be wor­
ried about possible .. .f6. 15.4Jd6
does not help in view of 15 ... 4Jd4
and 15.c3 b4 apparently passes
Black the initiative.
13 . . . g5 ! ?
This i s by n o means the only
move, but I chose it since it best il­
lustrates Black's strategic goal in the
7 . . . h5 variation. Now the eS-square
is under his full controL The weak-
176
ness of g3 is also essential. The play
may continue 14.l'!fl .ixe3 15.�xe3
gxf4 16.l'!xf4 4Jg6 17.l'!f2 �g3, or
14 . .ixa7 l'!xa7 15.fxg5 �g3+ 16.�f2
�xf2 + 17.\t>xf2 l'!g8 = .
B 3 . 8.0-0 h 4 9.gel
9.4Jxc6 dxc6 (9 ... bxc6 10.4Ja4
d6 is also an option) 10 . .ie3 hxg3
ll.hxg3 4Jf6 1 2.4J a4 4Jd7 13.c4 bS
14.4Jc3 bxc4 gives Black counterplay
on the queenside.
9 ..• hxg3 10.hxg3 d6 .
I would like to lead out the bish­
op, but lO . . . .icS? (Bosch's recom­
menndation ! ) 11.4Jxc6 ! bxc6 (11 . . .
dxc6 12 .e5) 12 .e5 ! i s horrible for
Black. He is positionally lost here.
11.4Jxc6 !
If White misses the opportunity
to trade knights, he might not have
a second chance. For instance, after
ll . .igS (as Radulski played against
Ivanisevic), Black has ll . . . ttJeS or
even ll . . . .ie7 12 . .ixe7 4Jcxe7 ! ? .
The same applies t o ll.a4 ttJeS,
although 11 . . . 4Jf6 is also possible.
Then the thematic 1 2 .4Jxc6 bxc6
13.e5 dxeS 14J'!xe5 can be attacked
by 14 . . . �xe5 ! 15 . .ixc6+ .id7 16 . .ixa8
4Jg4 17 ..if4 �hS where the open h­
file is a nice asset - 18.�f3 is only
move as 18 .�d2 J.cS 19.4Jdl eS
20 . .ie3 �h7! ! is rather unpleasant.
ll . . .bxc6
 The Fianchetto
12 .e5
It is very interesting to compare
this position with the one which aris- .
es after 7 . . . lt'lf6 8.0-0 i.e7 9J'!e1 d6
10.tt:lxc6 bxc6 11.e5 ! ± . If we counted
the developed pieces, it might struck
us that Black appears to be clear two
tempi down ! In fact, the absence of
the knight from f6 takes the sting
of White's pawn trust since Black is
not obliged to capture. If we delve
deeper, we'll notice that the raid of
the h-pawn was not just a waste of
time. It developed a heavy piece -
which is now pointed at the enemy
king! In most variations Black will
castle long or leave his king on f8.
Only future will tell how substantial
Black's counterplay is.
Let's investigate what happens if
White refrains from 12 .e5 :
a) 12 . lt'l a4 i.b7 13.b3 (13.c4 cS
14.e5? does not work against the
open h-file ! ) 13 . . . i.e7! ?
Perhaps this i s the most accurate
move order because after 13 . . . lt'lf6
White may contemplate 14.i.g5.
Note that the knight's development
to f6 has no alternative. The typical
manoeuvre . . . tt:lg8-e7-c6 leads to a
passive position : 13 . . . 4Je7 14.i.b2
cS 15J�l!d2 tt:lc6 16J'!ad1 0-0-0 (16 . . .
f6 17.i.c3 a5 18.tt:lb2, heading for c4)
17.c4±. White then carries out b4.
14.c4 cS 15.i.b2 (if White delays
this move, hoping to meet . . . tt:lf6
by i.gS, then we'll occupy the main
diagonal with 15.'�e2 i.c6 16.tt:lc3
i.f6 = , e.g. 17.i.b2 Wf8 18.lt'ld1 hb2
19.lt'lxb2 eS) 15 . . . 4Jf6 16.'�e2 i.c6 ! ?
17.tt:lc3 Wf8 .
b) Another version o f the same
set-up was tested in the blitz game
Malakhov-Grischuk, Moscow
2 0 1 3 : 12 .i.e3 tt:lf6 [Black may also
cover the b6-square with 12 . . . i.b7
13 .lt'la4 cS (13 . . . i.e7? 14.i.b6 �b8
15.e5), but then White can refrain
from c4 and swing the a4-knight
to that square, starting with 14.b3.]
13.tt:la4 �b8 14.c4 cS. It might seem
that White has won a tempo (by
having provoked . . . �b8) , but the
bishop is misplaced on e3 and he
will probably have to redeploy it.
15.b3 i.e7 16.�e2 i.b7 17.tt:lc3 lt:Jd7
18.�ad1 �d8 (18 . . . \t>f8) 19.i.c1 lt'le5
(19 . . . i.f6) 2 0.i.b2 lt'lc6 2 1 . tt:lb1 i.f6
2 2 .hf6 gxf6+.
177
 Part S
12 . . . d5 13.�a4 .ib7
We must prepare to take on c4
or we would risk to remain without
any counterplay: 13 . . . lt::Je 7 14.c4 lt:JfS
1S . .igS l'l:b8 16.l'l:cl W!a7 17.a3t.
14.c4 dxc4! 15.�c3 �e7
16.fta4
16 . . . 0-0-0 !
16 . . . l'l:d8 17.W!xc4? lt:JfS was fine
for Black in Musakaev-Malakhatko,
Pavlodar 20 12 : 18 . .ie3 [It is under­
standable that White does not wish
to let Black's rook to d4: 18 . .igS l'l:d4
19.Wffe 2 cS 2 0.l'l:ad1 (20 .lt:Je4 .ie7
2 1.he7 l!txe7 2 2 .l'l:ac1 he4 23 .he4
W!xeS 24.W!xa6= ) 20 . . . .ie7 2 1 .he7
l!txe7 22 .hb7 W!xb7 2 3 .lt:Je4 W!dS,
with a firm grip of the centre.] 18 . . .
lt:Jxe3 19.l'l:xe3 .ie7 2 0 .l'l:d1 l'l:xd1+
(20 ... 0-0) 21.lt:Jxd1 0-0 2 2 . l'l:b3 aS
23.lt:Je3 .ia8 24.lt:Jg4, draw.
However, 17 . .igS ! l'l:d7 (17 . . .
l'l:d4 1 8 . .ie3) 18.W!xc4 i s better for
White: 18 . . . lt:JdS 19.l'l:acl ! or 18 . . .
lt:Jg6 19.l'l:ad1! l'l:xd1 2 0 . lt:Jxd1 l'l:hS
21.Wig4 l'l:h8 2 2 .Wfff3t when 22 . . .
lt:Jxe5?? 23.l'l:xeS W!xeS 2.4 .W!xc6+ is
a very economic mate.
178
17.ftxc4 c5 18.�e4 �f5 19.i.e3
I have also analysed 19 . .igS l'l:d4
2 0 .�c3 .idS !
A curious position where Black
cannot move his dark-squared bish­
op, but White cannot make pro­
gress either:
2 1 . l'l:ad1 �c6 2 2 . l'l:xd4 lt:Jxd4
23.�aS (23.b3 l'l:hS) 23 . . . 'i!tb8 24.b3
Wffb 7oo;
2 l . l'l:ac1 W!c6 2 2 .�f3 'i!tb8 23.b3
W!b7 24.W!g4 l!ta7 2S.l'l:cd1 W!bS.
19 . . . �xe3 2o.gxe3 �xeS ! ?
The most principled continu­
ation. Sprenger-Avrukh, Sibenik
2 0 1 2 , went 20 .. J�d4 2 1.�c2 c4
2 2J'l:d1 �xeS 23 Jl:xd4 �xd4. Here
24.lt:JgS hg2 2S. l!txg2 W!dS+ 26.lt:Jf3
l!tb7 27.l'l:c3 .ic5 28 Jl:xc4 .ib6 may
be equal, according to the comput­
er, but perhaps it is easier to play
with White whose king is more safe.
21.gb3 gd4 2 2 .�d6+ hd6
23.hb7+ 'i!td7 24.ftxa6 l!te7
25.i.c6 g6 =
The king returns home via the
dark squares.
Pa rt 5 . The Fia nchetto
Kan - Step by Step
l.e4 c5 2 . ttlf3 e6 3.d4 cxd4
4. ttlxd4 a6 5.ttlc3 YNc7 6.g3 J.b4 !
The possibility for this active
development makes the Kan a per­
fect weapon against the fianchetto
fans. White has not displayed any
new ideas for a long time. We can
simply look at a couple of Smyslov's
40-years-old games to set us ready
for the battle.
In the diagram position, White
can defend the c3-knight with :
A. 7.tt'lde2 or B . 7.�d2
A. 7. ttlde2 ttlf6 8.J.g2 d6 9 . 0-0
9.�f4 e5 1 0.�g5 provokes 1 0 . . .
tt'lbd7, but Black had exactly this
set-up in mind anyway! Van Riems­
dijk-Smyslov, Buenos Aires 1990 ,
went further 11.0-0 h6 12 .�d2 �c5
13.tt'la4 �a7 14.�b4 tt'lc5 15.hc5
when 15 . . .�xc5 16.tt'lxc5 §'xeS
17.tt'lc3 �e6 18.tt'ld5 hd5 would
have offered Black the better pros­
pects. In this typical pawn struc­
ture, the knight is more mobile than
the bishop. ·
I advocate to delay castling in or­
der to leave White without a clear
plan. However, 9 . . . 0-0 is a viable
option: 1 0 .h3 tt'lbd7 ll.g4
After this advance, Black's main
task becomes to prepare . . . d5. He
may use his king's rook ( .. J'ld8) or
leave it on f8 to support a pawn
break by the f-pawn as in Antonov­
Donchev, Sofia 1984, which saw
ll.a3 �a5 1 2 . Wh1 l"!b8 13.g4 b5 14.f4
�b7 15.g5 tt'le8 16.l"!f3 (16.f5 exf5)
16 . . . l"!d8 17.l"!g3 §'b8 18 .b3 d5 and
Black intercepted the initiative.
179
 Part S

In Ree-Darga, Las Palmas 1973,

Black played ll . . J'!d8 immediate­
ly. The game continued 12 .gS ll:Je8
13.f4 bS 14.h4 i.b7 1S.a3 i.cS+
16.�h1, when 16 . . . dS !+ would have
made clear who was in command.

10.h3

White may try to harass our

bishop with : Here, 1S . . . .ie6 . 16.ixcS dxcS
a) 10.ll:Ja4 eS ll.a3 i.cS 12.ll:JxcS 17.ll:Jc3 :!'!d8 18 .ll:JdS ixdS 19 .exdS
ll:JxcS 13 .ll:Jc3 �d6 was equal in Filippov-Stell­
wagen, Yerevan 2 0 0 6 : 2 0 .:!'!fe1 0-0
2l.c4 ll:Jd7 2 2 .f4 :!'!deS 23 .fxeS :!'!xeS
24 . .ih3 fS = . A simpler solution is
1S . . . .ig4 16.ixcS �xeS 17.h3 he2
18 .�xe2 0-0 = .

1 0 . . i.c5 ! ll.g4 h 6
.

Similar positions arise in the

Najdorf when White plays ll:Jc3-
dSxe7. They are known to be fine
for Black - all his pieces are active
and he has the better control of the
centre. The game Katalymov-Kara­
sev, Naumburg 2 0 0 2 , went 13 . . . i.e6
(13 . . . h6!?) 14.i.gS ll:Jcd7 1S.�d2 h6
16.i.e3 ll:Jb6 17.b3 :!'!c8= .
This i s the point o f 9 . . . ll:Jbd7.
b ) 10.i.d2 i.cS ! ( mind the trap By not castling, Black forestalls
10 . . . 0-0? ll.ll:JbS) ll.ll:Ja4 the enemy attack on the kingside.
Thus he gains time to complete his
After ll.ll:Jf4 0-0 12 .�e2 ll:JeS
queenside development and pre­
13.:!'!ad1 bS 14.ll:JhS ll:JxhS 1S.�xhS
pare a break in the centre or a coun­
ib7, Black owns the initiative,
ter-attack against the enemy king.
Medina-Smyslov, Las Palmas 1972 .
White's task in the diagram posi­
ll . . . .ia7 12 . .ib4 ll:JcS 13.ll:JxcS tion is not trivial. His main idea
!xeS 14.�d2 eS 1S.:!'!ad1 of pushing gS is no longer enticing

180
 The Fianchetto

while the advance of the f-pawn is �e6 2 1 .flf4 flc4 2 2 .�b4;1;) 18.l2Jd3
at least double-edged : 12 .1!ih1 �b8 l2Jc4 19.flg5. White is slightly ahead
13.f4 b5 14.a3 �b7 15.f5 exf5 16.�xf5 with his attack.
h5 ! ? 17.g5 l2Jg4.
The stem game Micheli-Smys­
lov, Venice 1974, went instead : Bl. 9 ••• 0-0 ! ?

12.a4 �e5 13.1!ih2 fle7 (13 . . . This move i s consistent with

.!Llg6 ! ? ) 14.�g3 g5 line A. We do not waste any tem­
po on retreating the bishop to e7,
Black's dark-squared strategy but opt for a sound positional idea
has triumphed. - to build up a pawn barricade on
the central dark squares. Our lead
in development practically rules
B . 7 .id2
• lDf6 8.J.g2 �c6 out any White's plan with early
9.�b3 pawn advances -since both 10 .g4 or
1 0 .f4 will be strongly met by 10 . . . d5 !
ll.exd5 exd5.
10 .fle2 or 10 .a3, besides 10 . . .
J.xc3 ll.J.xc3 e 5 , could also face
1 0 . . . J.e7 ! ? , transposing to sidelines
of line B 2 .

10. 0-0 d6 11.f4

Bl. 9 . . . 0-0 ! ? ; B 2 . 9 . . . J.e7

Miezis' pet line 9 . . . h5 10.f4 d6

ll.a3 J.xc3 12 .J.xc3 e5 is an inferi­
or version of line Bl. White's bish­
op pair is restricted, the kingside
pawn storm is stalled. Still, I do not
like this set-up. It is too passive -
Black's pieces have mostly defen­ ll . . . hc3
sive functions. The game Garnelis­
Miezis, Kaunas 2 0 09, showed a Black may not need ll . . . b5 at all.
good, active approach for White : Guseinov-Ghaem Maghami, Fu­
13.fld2 ! a5 14. 0-0-0 a4 15.fxe5 dxe5 jairah City 2 0 1 2 , went 12 .g4 l2Jd7
16.lDc5 0-0 17.h3 lDaS (17 . . . �d8 13 .g5 and Black returned to the ad­
18.flg5 .!Lld4 19 .J.xd4 exd4 2 0 .lDd3 vance of the a-pawn : 13 . . . a5 14.a3

181
 Part S

hc3 1S.bxc3 (1S . .bc3 i.b7 16.�f3 17.b3 b5 18.lU3 b4 19.J.d2 J.a6
b4 17.i.eloo) . Here lS . . . eS 16.fS a4 2 0 . gh3 gfc8
17 . .!Llcl .!LlaS would have been the
sternest test of White's attacking
potential.

12 .hc3 e5 13.f5 a5

It is against common sense

to make weaknesses on the flank
your opponent will attack. Still,
13 . . . h6 is an interesting option.
The game Kudrin-Dobrov, Mos­
cow 2004, went 14.i.d2 aS 1S.a4
bS (1S . . . �d8 ! ? 16.g4 .!Llb4 ! intend- White has achieved everything
ing 17 . . . dS) 16.axbS '?9b6+ 17.@hl he was aiming for. And still
. he can-
�xbS 18.g4 i.b7 (18 . . . a4 ! ? 19 .gS not break through !
hxgS 2 0.hgS t!Jh7 2 l.i.h4 i.d7
22 . .!Lld2 '?9xb2 23 . .!Llc4 �d4+) 19.gS 21.�h5
hxgS 2 0 .hgSoo. Perhaps White
should try something more ener­ Or 2 1 .f6 g6 2 2 .�el .!Lld4 23 .�h4
getic as 14.g4 aS 1S.h4 a4 16.gS .!Llh7 hS- + .
17 . .!Lld2 hxgS 18.hxgS .!LlxgS 19 .�hS
f6 20 . .!Llf3 although White's attack
is not too impressive.
Instead, I consider an immedi­ 2 2 .f6 .!Lld8 23 .fxg7 @xg7+ or
ate counter-attack with the a-pawn. 2 2 .g6 fxg6 23.fxg6 t!Jd4- + favour
Black can also start with . . . 13 . . . bS. Black.
The following variations are far
from being encompassing. I only 22 . . . .hd3 23.�d3 ttld4
aimed to show the defensive stand 24.c3 bxc3 25.hc3 ttlb5 2 6.Ael
against White's most direct attack­ ti'b7!
ing plan.
Black's chances are preferable,
14.g4 e.g. 27.f6 �c2 2 8 .i.d2 �ac8 29 .fxg7
@xg7 3 0 .i.h3 �e8.
14.a3 gives a tempo for a break
in the centre : 14 . . . �d8 1S.g4 h6
16.h4?! a4 17 . .!Lld2 dS 18.exdS .!LlxdS B2. 9 Ae7
••.

19 . .bdS �xdS 20 .gS .!Lld4+.

At first sight Black has lost a
14 ... a4 15.g5 ttld7 16.ttlcl a3 tempo with the bishop manoeuvre.

182
 The Fianchetto

In fact, we'll see later that he might

have even gained one, as in the var­
iations where he plays . . . lt:\c6-e5-
c4 and White retreats his bishop
to cl. A more important argument
in favour of the inclusion of . . . i.b4
is that White has no longer the op­
tion of capturing on c6, followed
by e4-e5. We know that this threat
is the crux of the opening battle in
the Taimanov. Without it, Black
has green light for developing the
queenside pieces while delaying
his castle. In more general terms,
White has one piece less in the cen­
tre and that allows more aggressive
play from Black.
10.0-0
White often connects the place­
ment of his knight on b3 with the
advance of his a-pawn :
a) 10.a4. In practice Black au­
tomatically answers 10 . . . b6, . . .i.b7,
. . . d6, . . . lt:\d7 and then defines the fu­
ture of his king. If White has weak­
ened his kingside, . . . 0-0-0 becomes
a plausible option: I consider in de­
tail this setup in game 33 Forcen­
Korneev, La Roda 2 0 1 2 .
However, I would like also t o of­
fer you as a surprise weapon a new
idea, which tries to exploit the con­
crete move order:
10 . . . 0-0 ! ?
ll.a5 lt:\e5 1 2 . 0- 0 ( 1 2 . lt:\ a4 lt:\c4)
12 . . . lt:\c4 13 .i.c1 d6 14.g4
14.'�e2 looks pointless due to
14 . . . i.d7, when 15.g4 (15.f4 i.b5)
could be met by 15 . . . d5 16.exd5 Eife8
·
17.g5 (17.dxe6?· ! he6 18.h3 Eiad8�)
17 . . . lt:\xd5 18 .hd5 exd5 19.lt:\xd5,
when 19 . . .'�xh 2 + 2 0 . \tlxh2 i.d6+
2 l.i.f4 Eixe2 2 2 .hd6 lt:\xd6 23.lt:\b6
Eid8 is roughly equal since 24.lt:\xd7
Eixd7 25.Eiad1 Eixc2 26.Eid3 \tlf8
would be even pleasant for Black.
14.f4 leads to a position from
the Taimanov where it is White to
move ! We can borrow the set-up
from the game Bartel-Akopian, Port
Erin 2 0 0 6 : ll . . . Eid8 12 .'�e2 Eib8
(12 . . . d5 13.e5 lt:\d7 14.hd5 lt:\ dxe5 is
possible, but why not await g4 while
preparing typical counterplay with
. . . b5). 13 .g4 d5 !
14.g4 d5 15.exd5 Eid8 16.h3

Now ll.f4 will b e strongly met

by ll . . . d5 12.exd5 lt:\b4. We often
meet this sacrifice in this chapter.
So White should carry on his plan :

183
 Part S
I think that Black has decent
counterplay here. For instance,
16 . . . b5 ! ? 17.axb6 c!bxb6 18.id2
lt'lfxd5 19.lt'lxd5 c!bxd5 2 0 .ia5 lt'lb6
21.ha8 (21.�e2 ib7 2 2 .�e3 l!d6oo)
2 1 . . .l!xd1 22.l!fxd1 h5 23.l!d3 hxg4
24.hxg4 �b8 25.if3 ib7oo. The
queen is powerful against the un­
protected king.
b) 10.f4 0-0
10 . . . d6 is also a good move which,
however, transposes to my alterna­
tive suggestion in the annotations
to game 33 Forcen-Korneev after
ll.a4 b6 or game 32 Radulski-Fi­
lev after 11.0-0.
Instead, ll.g4 commits White
on the kingside too early. Black has
several attractive retorts. He only
should not castle, for instance : 11 . . .
h 6 12 .h4 b 5 13.g5 lt'ld7 14.�e2 b4
15.lt'la4 a5 16.ie3 i.a6t when the
c!ba4 is a cause of concern to White,
Sutovsky-Podzielny, Essen 1999, or
11...b5 12 .g5 lt'ld7 13.�f3 (13.�h5
g6 14.'�'h3 lt'lb6 15.l!d1 id7 16.0-0
0-0-0 ; 13 . 0-0 lt'lb6 14.f5 lt'lc4 15.ic1
id7oo) 13 . . . b4 14.c!ba4 a5 15.ie3
ia6.
10 ••• 0-0 !
Black commonly delays castling
with 10 . . . d6. Then he finishes his
queenside development, prepares
to carry on the break . . . d5, and only
at that point does he define the po­
sition of his king. See game 32
Radulski-Filev, Plovdiv 2 0 1 2 , for
details. This game is also interest­
ing with the French pawn structure
which arose after 14 . . . d5 15.e5. You
184
should know what to do in it since
you may get it in several different
ways. My suggestion is more sim­
ple and straightforward.
ll.f4 (ll.g4 d5 1 2 . exd5 exd5)
ll ••• d5 !
Amazingly, this sacrifice has not
been tried in this particular position.
However, the same idea is known as
deserving attention in several oth­
er settings. Now 12 .e5 c!bd7 leads
to a pawn structure which is unan­
imously assessed as fine for Black.
For instance, 13.lt'le2 lt'lb6 14.lt'lbd4
c!bc4 15.icl id7 16 .c3 (16.b3 c!bxd4
17.c!bxd4 lil a3) 16 . . . l!ac8t .
12.exd5 lt'lb4 13 .dxe6
I suspect that the unassuming
move 13.l!c1 is a wiser option: 13 . . .
c!bbxd5 14.c!bxd5 exd5= .
13 .b:e6 14.@hl gadS 15.f5
• • •
.tc4 16.gel b6
Black has full compensation for
the pawn, e.g. 17.if4 id6 18 .hd6
l!xd6 19 .�c1 l!fd8.
Pa rt 5 . The Fia nchetto
Comp l ete Games
27. Hoogendoorn-Lautier
Al garve 1999
l.e4 c5 2 . �f3 �c6 3.d4 cxd4
4.�xd4 ti'c7 5.�c3 e6 6.g3 a6
7.�g2 d6 8 . 0-0 �d7 9J3el �e7
tO. �b3 b5 u.a3 �f6 t 2 .f4 gds
(12 . . . �c8 13 . .ie3 h6 14.g4 �h7
15.�d2 gSoo) 13 .g4? !
Too optimistic. 13 . .ie3 is more
prudent. Now Black can choose be­
tween 13 . . . h5 ! 14.h3 hxg4 15.hxg4
gS+ or the more restrained 13 . . . .ic8 .
Instead, he mixes up ideas with :
13 h6 14.h3 �c8 15.�e3 �d7
. • •
16.�d2 �b6 ? !
Black overlooks a tactical blow.
It was time for 16 . . . 0-0 17.�f2 .if6
18.�d1oo.
17.ti'd3?!
White misses the opportunity
for 17.ltJd5 ! exdS 18.exd5 �c4 (18 . . .
ltJ b 8 19.�d4 + - i s a double-attack! )
19.�c3 0-0 2 0 .dxcM.
17 . • • �c4 18 .�cl g5? !
Apparently Lautier overestimat­
ed the eS-square. 18 . . . �b6 19.'it>h1
.ib7 2 0 .a4 ltJb4t was safer and bet­
ter.
19.a4! gxf4 20.axb5 �b6
2 1 . 'it>hl axb5 22.�f4
22 .if6
• • •
The weakness of bS does not let
Black enjoy the eS-outpost. 22 . . .
.ia6 23 .�g3 b 4 24 . .if1 h S 25.ltJa4
�b7 2 6.g5 eS 27 . .icU is double-
185
 Part S

edged, but White's king is relative­

28. Fressinet-Delchev
ly safer.
Calvia ol 2004

23.�d2 � 6e5 24.�xc4 bxc4

l.e4 c5 2 . �f3 e6 3 . � c3 �c6
25.'%M2 �g7 26.i.e3 �c6 27.�a5
4.d4 cxd4 5. �xd4 �c7 6.g3 a6
h5 28.gxh5 �xh5 29 .%Ye2
7.�g2 d6 8 . 0-:-0 i.d7 9. �c6 �c6
�h8 30.�eal i.b7 31.�b5 �a8
lO.�el i.e7 11.�g4 h5 12 .�e2 h4
32.�a8 + i.xa8 33.�b6 %Yd7
13.a4 hxg3 14.hxg3 �f8 !
34. �b5±

White has won the first battle.

His rook is extremely mobile. How­
ever, he gradually begins losing
ground in the next moves.

34 i.c6 35.�a3 (35.lt:Jd4 i.b7

•••

36.'�'d2±) 35 c3 36.�b8+?
•••

This is a double mistake. White

15.a5 �c8 16 . .ie3 �f6 17.i.b6
exchanges his best piece instead of
pushing his passed pawn : 36.b4±.
White's main problem is that
he does not have a clear plan. In
36 �e7 37.�xh8 .ixh8
fact, he must wait for Black to push
•••

38.bxc3?
. . . e5. Only then can he aim for lt:Jd5
followed by a pawn storm on the
The turning point of the game.
queenside. Meanwhile he is risking
38 .b4 %Yb7 39.b5t would have re­
to lose the initiative, for instance :
tained some edge.
17J�edl e5 18 .i.b6 %Yd7 19.�d3
g6 2 0 .�adl c;t>g7 2 1 . �1d2 (21.b3
38 �b7 39 .ig5+ �d7 40.�cl
�e8 2 2 .lt:Jd5 .ixd5 23.exd5 i.d8+;
•.• •

�f6 41.%Ye3 i.d8 42 .�e2 i.e7

2 1.f4 �g4 2 2 .�1d2 W/e2 23.�e2 lt:Jd7
43.c4 �g6 44.�b5 �h4 45.�d4
24 . .if2 .idS 25.b4 i.e7 26.�ed2 lt:Jf6
YlYbl 46.�fl? (46.%Yd20 i.a8 47.c5
27.i.f3 i.d7+)
lt:Jg2 48.c6 .ic6 49.lt:Jc6 lt:Jh4+)
46 i.b7 47.i.e3 �xfl+ 48.i.xfl
•••

i.xe4+ 49.�gl e5 50.�b5 hc2

51.�f2 f5 52 .i.a7 f4 53 .ie2 �c6
•

54.�c3 �b7 55.�d5 .idS 56.c5

�xa7 0-1

The moral of this game is that

one good square (e5) does not com­
pensate for bad piece co-ordination.

186
 The Fianchetto
2 1 . . .'�e8 ! Remember this ma­
noeuvre ! The queen is heading for
h7. The play might proceed with :
2 2 .b3 Preparing tLl dS and c4.
2 2 .f4 only helps Black's plan -
2 2 .. .'�g8 23 J�d6 .M6 24J:'!:d6 tLld7
25 ..ie3 'r9h7 26 .'�f2 'r9h 2 .
2 2 . . .'�g8 23.f4 'r9h7� 24.fe5 (24.
.if3 tLld7 25 ..ie3 tLlcS = ) 24 ... de5
25.'r9f3 (25.tLld5 h:dS 26.exd5 .ib4
27.!'!:dl 'r9h2 28. 1t>fl !'!:hS 29.!'!:f3 e4)
25 . . . tLle4! 26.tLlxe4 fS 27.'�e3 \Wh2
28. 1t>fl fxe4 29 . .ie4 \Wh3+.
These variations illustrate
Black's hidden possibilities.
17.f4 is more dubious than it is
enterprising. We blockade the cen­
tre with 17 ... e5 18 . .ib6 V9d7 (18 . . .
\Wb8 19.!'!:adl exf4 2 0 .gxf4 .idS
2l.h:d8 !'!:xd8 2 2 .e5 h:g2 23.1t>xg2
\WeB+ is also good) 19.!'!:adl and here
19 . . . \Wg4 2 0 .'r9xg4 is roughly equal,
but the engines find a concrete way
for Black to intercept the initiative:
19 . . . exf4 ! 2 0 .gxf4 !'!:h4.
17 ••• Wd7!
Black had played previously only
17 . . . 'r9b8 which is a rather deplor­
able fate for the strongest piece ! I
have better prospects for it.
18.gadl We8 ! 19.J.d4
Inkiov played here 19.f4 when
19 . . . e5 is again the best retort. Then
if White continue 2 0.f5, we can
change plans withe 20 . . . tLld7 2 l..ie3
W/d8 2 2 . !'!:al (22 .b4 tLlf6) 22 . . . J.g5.
White's queenside will be vulnera­
ble.
19 ••• e5 20 .ie3
•
Black's bishop is hardly bet­
ter on b6: 20 . .ib6 g6 2 l.f4 (2 l.tLld5
hdS 2 2 . exd5 .idS 23 .J.d8 'r9d8 = )
2 1 . . . tLld7 2 2 .J.e3 W/d8 ! 23 .b4 lt>g7
24.tLld5 (24J'!:fl Wg8 ! 25.tLld5 hdS
26.!'!:xd5 'r9h7 2 7.fxe5 tLlxeS 28 . .id4
'r9h2 29. 1t>f2 !'!:h3) 24 . . . hd5 25J�xd5
W/c7 26.!'!:d2 tLlf6�. As you see, Black
can shift the direction of his attack,
combining threats along the h-file
and against the weak aS-pawn.
20 ••• g6
A consistent move, but the ma­
noeuvre 20 . . . J.d7! is stronger. The
point is that 2 l.J.g5 .ie6 2 2 .hf6
.ix£6 23.!'!:xd6 fails to 23 . . . .ie7
24.!'!:d2 J.b4 so Black keeps more
pieces on the board.
After the text, White can reach
a drawish position with 2 l..ig5
lt>g7 2 2 .!'!:d3 !'!:hS 23 .h:f6+ lt>xf6
24.tLld5 + . The aggressive 2 l.f4 ? !
lt>g7 2 2 . !'!:fl would allow the thema­
tic manoeuvre 22 . . . '?9g8 ! with coun­
ter-attack.
2 1 . tLld5 .ixd5 2 2 . exd5 Wa4!
187
 Part S
Black's queen is extremely agile !
Instead of going to h7, it hits targets
on the other wing. As long as 23 J�al
�xc2 24.l'!acl �fS 2S.l'!xc8 �xeS
26.l'!cl �d7 27.�c4 Wg7 28.�c7
�d8+ covers all invasion squares,
White is forced to give up the aS­
pawn and hope for a counter-attack
against eS.
23.c3 �a5 24.f4
24 ••• Wb5?
A horrible mistake which might
have cost me the point. The simple
24 . . . �c7 2S.fxeS dxeS 26.d6 hd6
27.�f3 Wg7 28.�xb7 aS would have
retained the pawn with a good posi­
tion. Even more consistent is:
24 ... ltJd7! 2S.fxeS [Or 2S.�f2
fS ! 26.fxeS (26.g4 �h4) 26 . . . ltJxeS
27.�b6 �bs 28 .�d4 �f6+.] 2S . . .
ltJxeS 26.�d4 �c7 27.heS (27.
l'!fl Wg7 28.l'!del l'!ce8+) 27 ... dxeS
28.�xeS �cS 29.l'!d4 l'!hS+.
The rest of the game is a desper­
ate struggle which was eventually
rewarded with a draw.
25.fxe5 �xe2 26.gxe2 tLlg4
27.exd6 hd6 28.J.f4 gds
188
29.hd6+ gxd6 30.c4 Wg7
31.c5 gd7 3 2 . d6 tLlf6 33 .b4 ghs
34.ge7 �e7 35.dxe7 a5 36.hb7
axb4 37.gcl b3 38.c6 b2 39.gbl
tLle8 4o.gxb2 ges 41.i.c8 gxe7
42 .J.d7 tLld6 43 . Wg2 f5 44.gd2
tLlb5 45.gds tLlc7 46.gas Wf6
47.ga7 tLlb5 48J'�b7 tLld6 49.gbs
ge2 + 50.ci!;>f3 gc2 51. ci!;>e3 ci!;>e5
draw.
29. Tim man-V. Milov
Biel 1995
l.e4 c5 2 . tLlf3 e6 3.d4 cxd4
4.tLlxd4 a6 5.tLlc3 Wc7 6.g3 tLlc6
7.J.g2 d6 8 .0-0 J.e7 9.gel i.d7
lO.tLlxc6 hc6 11.Wg4 h5 12 .We2
h4 13.a4 hxg3 14.hxg3 ci!;>f8
15.a5 tLlf6 (1S . . . l'!e8 ! 16.�e3 ltJf6)
16.�e3 ges
This strange-looking move aims
to ensure a good retreat square for
the queen on c8 while protecting
the e7-bishop against tactical hits
as ltJc3-dS.
As a positive side effect, Black
threatens to break in the centre by
. . . dS. 17.�gS or 17.l'!edl could be met
by 17 . . . 'itlg8.
 The Fianchetto
17 .id4
• 2 0 .f4 �f6 2 1 .if3 g6 2 2 . <;f:;>g2
Timman tries to carry on e5
with tempo. Most games continue
17 . .tb6 �c8 and then:
a) 18 J:�edl e5 19.b4 .td7 2 0.Eia3
g6 2 l .lt:'ld5 lt:Jxd5 2 2 . exd5 <;f:;>gf+,
Gallagher-Bosiocic, Olbia 2 0 0 8 ;
b) 18.b3 g6 19.lt:Ja4
19 . . . d5 ! 2 0 .e5 lt:Jd7 2 l ..td4 .tb4
22 . .tc3 .te7 23 . .td4 .tb4 24 . .tc3 .te7
25.W'e3 <;f:;>g7 26 .W'd2 Eih7 27 . .td4
Eieh8 2 8 .lt:Jb6 lt:Jxb6 29 .axb6, Salda­
no-Tsuboi, Campinas 2 0 11, when
29 . . . Eih2 , intending . . . E18h7, . . . W'h8
should win.
17 • •. W'c8 18J�edl :!!h 7
18 . . . e5 19 . .te3 W'e6 would have
been better, in order to avoid the
capture on f6.
19 .ie3
•
White admits that he has no
plan.
19 ••• �d7! ?
Preparing to meet 2 0 ..tf4 by
20 . . . lt:Je5
•
<;f:;>g7 23J:!hl geh8 24.gxh7 +
gxh7 25.ghl gxhl 26. <;f:;>xhl e5
27.<;f:;>g2 .id7! ?
White's king i s weaker s o Black's
threats are more dangerous. Now he
wants to transfer the bishop to e6.
28. �d5 �xd5 29.exd5 .ib5
30 .W'dl .if6 31.b3 .id7 3 2 . c4
.ih3 + 33. <;f:;>f2 exf4 34.gxf4 W'h8+
35.W'hl
Black has gained the initiative
and could have played for a win
with 35 . . . <;f:;>g8 ! threatening 35 . . . .td4
or 35 . . . .tc3. Instead, he prefers to
repeat the moves.
35 ••• �c8 36.W'dl ti'h8 draw.
30. Guseinov-Vasovski
Struga 12.08.2009
This game is a good example of
what Black should not do - instead
of staying active and generate con­
stant threats, he opted for a defen­
sive set-up on the kingside. This is
rarely a good approach in the Sici­
lian
l.e4 c5 2 . lt:Jf3 lt:Jc6 3 . lt:Jc3 d6
4.d4 cxd4 5. lt:Jxd4 e6 6.g3 .id7
7 .ig2 lt:Jf6 8 . 0-0 .te7 9 . lt:Jb3 a6
•
10.f4 W'c7 11 .ie3 b5 12.a3 gbs
•
13.<;f:;>hl 0-0 14.g4 h6 15.g5 hxg5
16.fxg5 �h7 17.W'g4 �e5 18 .'t!fg3
189
 Part S
It may seem that Black has an
easy game thanks to the outpost
on e5. In fact, White's plan is much
more clear. He will push the h-pawn
up to h5 (and even h6 sometimes)
and then he will be looking for a way
of swinging a knight to the kingside.
Black must quickly generate con­
crete counterplay before it became
too late. It would be a bad idea to
stay passively, as the course of this
game will teach us.
A logical continuation would
be to target the weak pawn on c2
with 18 . . . a5 19.h4 b4 20 . axb4 axb4
21.E1a7 E1b7 2 2 .E1xb7 '!Wxb7 23.lLldl
E1c8, with an initiative. The text
move is not bad either.
18 . . . �c4 19 . .icl g6 20.h4
gfc8? !
Black embarks on a wrong rede­
ployment. His kingside pieces were
perfectly placed, hitting g5. For in­
stance, h4-h5 would often cost
White a pawn due to . . . �d8. There­
fore, he should have activate his
queen with 20 .. .'�b6 ! , denying the
d4-square to the b3-knight. Then
the straightforward 2 1.h5 �h8
22 .E1f3 would be strongly countered
190
by 22 . . . gxh5 23.�h4 lLle5 24.E1h3
�g'T+, when the h-file will open in
Black's favour.
21.�e2
I would take the chance to cen­
tralise the knight at once with
21.lLld4 lLle5 2 2 . lLlf3 lLlxf3 23.'1Wxf3
.ie8 24.E1f2 . The exchange of one
pair of knights underlines the fact
that Black practically plays without
a piece.
2 1 . . .if8 2 2 .h5 .ig7? ! 23.�h4
.
�f8? !
Black has achieved his goal, but
he will soon discover that his bish­
op will be useless on the long di­
agonal. It only takes away the g7-
square to its king. He still could
have reconsidered and captured on
h5 to get some breathing space on
the kingside.
24. �bd4 a5
It is already late for 24 . . . gxh5
25.'1Wxh5 .ie8 26.c3 lLle5 27.lLlf4--t.
25.b3 �e5 26 . .i.e3 ? !
 The Fianchetto
26.h6 ! .ih8 27 . .ie3 would have
been clearly better for White as
Black cannot untie himself. For in­
stance, 27 . . . ltJc6 would run into
28.h7+ ! ltJxh7 29J'!xf7 ! . After the
text, instead of 26 . . . gxh5 ! , Black
falls into the same trap�
26 ltJ c6? 27.h6 hd4 (27 . . .
•••
.ih8 28.h7+ ! ) 28. ltJxd4 ltJh7
29.'@f2 i.e8 30.ltJf3 ltJd8 3U�adl
.ic6 3 2 . c!Llh2 '@e7 33. c!Llg4 gc7
34.gxd6 f5 35.gxf6 '@xd6 36.i.f4
'@d7 37.'@g3 gbb7 38.c!Lle5 '@e8
39. c!Llxg6 gf7 40.c!Lle7+ c!>h8
41.'@g7+ 1-0
31. Ponomariov-Dubov
FIDE World Cup, Tromso 2013
l.e4 c5 2 . c!Llf3 e6 3.d4 cxd4
4. c!L!xd4 c!Llc6 5.c!Llc3 a6 6.g3 '@c7
7.i.g2 d6 8 . 0-0 i.d7 9. c!Llb3 c!L!f6
10.f4 i.e7 11.<!>hl
11 ••• 0-0
Black chooses to await g4 and h4
in order to strike back in the cen­
tre with . . . dS. Another popular ap­
proach is to make yet another de-
veloping move before defining the
king's position. Since sooner or lat­
er Black will have to at least threat­
en with . . . dS, it looks logical to play
ll .. J'l:d8. However, on this square
the rook may be awkward when
White plays .ie3, '@f2 (g1) hitting
b6. For instance, 12 . .ie3 0-0 13.g4
h6 (13 . . . d5 14.e5 lLle8 15.'\tlfe2) 14.h4
dS 15.e5 ltJeB 16.g5 hxgS 17.hxg5
g6 18.l'l:f3 c!>g7 19.l'l:h3 l'l:h8 2 0 . l'l:xh8
c!>xh8 2 1.�g1, winning a crucial
tempo for '@h2 + .
Let's correct this idea by 1 1.. .l'l:c8
1 2 .i.e3
12 .g4 ? ! runs into 12 . . . h6 13.'@e2
eS+, but 12 .�e2 is a fair alternative.
Perhaps this is the maximum Black
can extract from the delay of 0-0 -
White has to play two quiet moves
like c!>h1 and '\tlfe2 . Then and it is
time to castle. In Alekseev-Rod­
shtein, Legnica 2 0 13, he persisted
with his tricky move order, but 12 . . .
e S 13 .i.e3 0 - 0 14.f5 put him under
attack. The game went 14 . . . b5 15.a3
b4 16.axb4 ltJxb4 17.g4 h6 18 .g5
hxgS 19.hg5 '\tlfc4 2 0 .'\tlfd2 l'l:fd8
2 1.l'l:g1 <i>h7 2 2 . .if3 + -.
12 . . . b5 13.a3 h6 14.�e2 0-0 15.g4
ltJh7! 16.'@f2 .if6
191
 Part S
Black is OK here :
17.h4 i.xc3 18.bxc3 liJe7 19.g5
hxgS 20.hxg5 f6 ! ? ;
Alekseev-Cordova, Khanty-
Mansiysk 2010, saw 17.gad1, when
I think that Black should have opt­
ed for a dark-squared blockade
with 17 . . . i.xc3 18.bxc3 e5 19.f5 liJe7.
White is not ready for a serious at­
tack. After 2 0.h4 liJf6 2 l..if3 (21.g5
lt:lg4 22 .'l;l/e2 hS) 2 l . . .d5 ! 2 2 . exd5
liJexdS 23.g5 lt:lxe3 24.'\1�'xe3 hxgS
25.hxg5 lt:lh7 26.f6 .ifS+, he is clear­
ly over-extended.
My conclusion is that 11.. .gc8
deserves attention.
12 .g4 h6 13.�e3
Remember this position ! When
White has completed development
and his bishop is on e3, we should
take the threat of h4 very seriously
and address it by 13 . . . lt:lh7! follow­
ing the example of the previous di­
agram. The . . . dS break is no long­
er effective !
13 . . . !1!ac8 14.h4 d5?
It was high time for 14 . . . lt:lh7. The
position after 15.g5 hxgS 16.hxg5
'l;l/d8 17.'l;l/h5 ? ! eS 18.liJd5 exf4 is
192
sharp, but White lacks immediate
threats. For instance : 19.i.xf4 i.xgS
2 0 .i.xd6 liJe7 or 19. liJxe7+ liJxe7
2 0 J�xf4 lt:lg6 2 1. :gf2 liJeS.
White can fight for the initia­
tive with 17.gf3 , when 17 . . . g6 ! ? , in­
tending . . . rtlg7, looks best. Instead,
17 . . . e5 18.gg3 exf4 19 . .ixf4 i.xgS
2 0.i.xg5 liJxg5 2 1.'l;l/xd6 liJh7 2 2 .�h6
�f6 23 .'l;l/h2 liJeS 24.liJd5 'l;l/d8 25.c3
leaves White with some edge.
15.g5 !
15.exd5 lt:lb4 i s extremely de­
manding for White while Black's
game is much easier.
15.e5 lt:le8 16.g5 hxgS 17.hxg5
g6 is double-edged - 18.gf3 lt:lg7
19.'l;l/g1 liJfS 20 . .ib6 �b8 21.ltlc5
i.xcS 2 2 .i.xc5 gfd8 23 . .ib6 gesoo.
15 ... hxg5 16.hxg5 t0xe4
17. t0xe4?
17.i.xe4 ! dxe4 18.'l;l/h5 .ie8
19.gf2 gives White an overwhelm­
ing attack. After the text, Black took
over the initiative.
17 . . . dxe4 18.he4 f5 19.gxf6
gxf6 2 0 , g:f2 t0b4 21.�g4 .ic6
 The Fianchetto

2 2 . gg2 .if8 23 .ixc6 ti'xc6

• castle only when he is ready to meet
24 .id4 gh6 + 25.®gl �d5?
• g4 by . . . dS. Long castling is also pos­
26.gel �e7 27.c3 gds 28 .ie3 • sible so White should think twice
gds 29.�d4 ti'd7 3o.®f2 gdhs before flinging forward his kingside
31.gegl �d5 32 .ti'f3 gb3 33.gg3 pawns. On the other hand, switch­
gxg3 34.gxg3 ti'f7 35.�e2 .ie7 ing to restrictive play with ll.a4
36.gg2 .if6 37.®gl .ie5 38.ti'e4 tLleS ! ? (ll . . . b6 is the common retort)
�xe3 39.ti'xe3 .id6 40 .ti'd4 1 2 . a5 tLl c4 13.tLla4 0-0 is not very
.ic7 41.ti'e4 .ib6 + 42. �d4 ti'h5 enticing for White.
43. ®f2 gf6 44.®e3 .ic7 45.ti'xb7
.ixf4+ 46.®d3 .ih6 47.®c2 e5 ll.f4 b5 1 2 . a3
48.ti'd5 + ®h8 49.ti'as + ®h7
50 .ti'e4+ ®h8 51.�a8 + ®h7 12 .e5 dxe5 13.fxe5 tLlxe5 14 . .ixa8?
52 .ti'e4+ ®h8 53 .�a8 + ®h7 �a7+ is clearly better for Black.
54.ti'e4+ Draw.
12 . . . .ib7 13.�e2 0-0

It makes sense to wait one more

Kan Games move before castling - 13 .. J'M8 ! ? .

14.g4 d 5 15.e5? !
32. Radulski-Filev
Plovdiv 04.02.2012 The critical line is, of course,
15.exd5 tLlxdS 16 ..ixd5 (16.tLlxd5
l.e4 c5 2 . �f3 e6 3 . � c3 a6 exdS 17.�f2 d4 18.Eiad1 Eife8�) 16 . . .
4.d4 cxd4 5. �xd4 ti'c7 6.g3 .ib4 exdS 17.tLlxd5 �d8 18.tLlxe7+ tLlxe7
7 .id2 �f6 8 . .ig2 �c6 9 .�b3 .ie7
• 19.tLla5 .idS 2 0 . Eiad1
10.0-0 d6 ! ?

If Black consolidated, he would

This move shows that Black has have fair chances against the weak­
put aside the idea of pushing . . . dS ened enemy king. But he has to
early. Instead, he will fianchetto the make a couple of accurate moves:
bishop, put a rook on c8 or d8 and 20 . . . Eic8 ! [20 . . . Eie8 21.Eife1 �b6+

193
 Part S
2 2 .�f2 �g6 23 .h3 .ie4 24 . .ib4 lLldS
2S.fS (2SJ�d4 ltlxb4 26J'�exe4 �xe4
27.f5 �f6 28.�xe4 ltlxc2 = ) 2S . . .
�h6 26.'>th2 ltlxb4 27.axb4 '?9f6
28.�e2t] 21..ic3 (21..ib4 '?9b6+
22.�f2 ltlg6 23.�xdS ltlxf4 24.'?9f3
lLlxdS 2S.�xdS �xc2 26.'?9f3 = ) 2 1 . . .
�b6+ 2 2 .�f2 �g6 23 .h3 i.a8ii;.
15 .c!bd7
••
16.:!U3
White would be unable to hold
his queenside with solid central­
ising moves. For instance, after
16 . .ie3 ltlb6 17.lLlcS, he'll face 17 . . .
d4 ! . H e also should reckon with the
threat of . . . gS at some point. There­
fore, Radulski makes a right choice.
He tries to attack h7. He is not
afraid of 16 . . . gS since after 17.�h3
gxf4 he may draw by 18.hf4 lLlcxeS
19.gS f6 20.�xh7= or gamble with
18.�d3 fS 19.exf6 ltlxf6 2 0 .lLlxdS
exdS 2 1.hdS+ 'lth8 22 . .ic3 .id8oo.
To be fair, in the latter line Black
could neutralise the attack with
22 . . . lLleS 23.-ixeS �xeS 24.hb7
�ad8 2S.�fS '?9xfS 26.gxfS �g8+
27.<kt>fl ltlg4. This sharp endgame is
probably balanced. Filev correctly
refrains from forcing the play.
194
16 ltlb6 ! 17.gh3 ltlc4 18.g5
•••
g6 19.ltldl a5 20.c3
20 ••• '?9b6 +
White's attack has reached a
dead end and Black can already
launch a counter-attack. The gener­
al rule dictates to strike in the cen­
tre and, indeed, 20 . . . a4 2 1.ltlc1 d4 is
perhaps the best way to exploit the
enemy weaknesses. The flank strike
20 . . . b4 is also a plausible alterna­
tive. Instead, Black takes a seem­
ingly inexplicable decision to seal
the queenside.
21.i.e3 ltlxe3 2 2 .'?9xe3 '?9xe3+
23. ltlxe3 a4?
A horrible move. Did Filev forget
that he had a light-squared bishop? !
23 . . . �fc8 24.ltlg4 b 4 would have
preserved a strong pull. Now 24.ltlc1
followed by ltld3 would have bal­
anced the game. You could guess the
reason behind Black's actions when
you knew the ratings ! He is more
than 2 0 0 points underdog so he just
wanted to make a draw! The rest is
unrelated to the opening. White was
stubbornly trying to win his worse
position and eventually lost.
 The Fianchetto
24.li�d2 ti)a5 25. ti)g4 �g7
26.gdl gabS 27.ti)f6 gbs 2 8 . tl)f3
h5? 29.i.fl ti)c4 30 .b:c4 bxc4
• better on c8, protecting e6 and leav­
31.�fl i.a8 3 2 . gd2 gb7 33.f5
exf5 34. ti)xd5 .ic5 35,gh4 gd7
36.ti)f4 ge7 37.ti) d5 gd7 38.ti)f4
ge7 39.�e2 gbs 40.e6 gb6
41.ti)e5 gbxe6 42. ti)xe6+ gxe6
43,gxc4 gxe5+ 44.�fl .ic6
45,gb4 f4 46.c4 .i.e3 47,gd6
i.e4 48,gb3 gxg5 49,gxe3 fxe3
50.�e2 gg2 + 51. �xe3 f5 52 .h4
gxb2 53.�f4 gc2 54,gd7+ �f6
55.gd6+ �f7 0-1
33. Forcen Esteban-Korneev
La Rod a 07.04.2012
l.e4 c5 2 . ti)f3 e6 3.d4 cxd4
4.ti)xd4 a6 5.ti)c3 Y!lc7 6.g3 .ib4
7. .id2 ti)f6 8 . .ig2 ti)c6 9 .ti)b3 .i.e7
10.a4 b6
11. 0-0
ll.f4 d6 12 .g4 h6 13.Y!le2 is more
aggressive. White is planning long
castling. The mundane 13 . . . .ib7 is
hardly the best answer although
14. 0-0-0 (14.h4 d5) 14 . . . lt:Jd7 (14 . . .
lt:Jb4 15.g5 hxg5 16.fxg5 lt:J d 7 17.g6
fxg6 18.lt:Jd4 lt:Jc5 19 .Y!lg4--.) 15.�b1
lt:Jc5 16 ..ie3 lt:Jxb3 17.cxb3 b5 is un­
clear. The bishop probably stands
ing the b-file open for the rook. So
I like :
13 . . J%b8 ! ?
Now 14.0-0 lt:Ja5 i s fine for Black,
e.g. 15.gad1 Y!lc4 16.lt:Jxa5 W/xe2
17.lt:Jxe2 bxa5 18 .b3 lt:Jxg4 19 .h3
lt:Jf6 20 ..ixa5 .ib7t ; 14.0-0-0 lt:Jd7
15.Wb1 lt:Jc5 is an improved version
of the above-mentioned line. The
game Malisauskas-Grabliauskas,
Vilnius 2009, went:
14.h4 lt:Jb4 15.0-0-0 h5 (15 . . .
b5 ! ? ) 16.gxh5, when 16 . . . lt:Jxh5
would have been double-edged.
I recommend the more concrete
move 14 . . . lt:Ja5 ! ? 15.g5 (15.0-0-0?
lt:Jxb3+ 16.cxb3 b5) 15 ... hxg5
16.hxg5 �xh1+ 17 . .ixh1 lt:Jd7
18.lt:Jxa5 bxa5 with a tangled po­
sition, e.g. 19 .b3 e5 ! , exploting the
hanging state of the c3-knight -
2 0 .f5? .ixg5.
ll . . . .ib7 1 2 .f4 d6 13.g4
13 . .ie3 transposes to a side line
of the Taimanov. Black's most natu-
195
 Part S
ral plan is to prepare the break . . . d5
with 13 . . . ltlb4. For instance, 14.�e2
(14.a5 b5 15.�b6 �c4) 14 . . . 0-0 15.g4
d5 16.e5 ltld7 17J�ad1 l'!ac8 18.l'!d2
g5.
15.�f2 is dubious due to 15 . . . ltlg4
16 . .ixb6 �b8 17.�e2 he4t 18 . .ixe4
�xb6+ 19.'it>h1 d5.
It is safer to sit tight on the
kingside with 15.l'!fdl d5 16.e5 ltld7
17 . .if2
This structure is generally good
for Black. He can continue simply
17 . . . l'!ac8 18.ltld4 ltl c6 = .
13 ••• h6 14.�f3
White aims to shift the queen to
an active position on h3. If he adopt­
ed a waiting game with 14.�e2,
Black could prepare a counterat­
tack with . . . g5, for instance : 14 . . .
ltld7 15.�e3 g5 ! ? 16.f5 ltlce5 17.ltl d4
�c4oo, or 14 . . . ltlb4 15.a5 (15.l'!f2 d5)
15 . . . l'!c8 (15 . . . b5? 16.ltlxb5) 16.axb6
�xb6+ 17 . .ie3 �c7 18.l'!fd1 d5 19.e5
ltld7 20 .l'!d2 g5t.
14 .tbd7 15.�h3 0-0-0 !
••
Black should be satisfied with
the opening. After the inevita-
196
ble . . . g5, he will control the critical
square e5.
16.Ae3 g5 17 .f5 i.f6 18.a5
b5 19.fxe6 fxe6 2 0.Ab6 �xb6
21.l'!xf6 'ee7 2 2 . l'!xh6
Or 2 2 .l'!af1 ltlc4 23J'�t7 �e8
24.l'!lf6 l'!d7 25.l'!xd7 'it>xd'l+
26.l'!xh6 l'!xh6 27.�xh6 �e7.
22 -efg7
.•• 23.l'!xh8 l'!xh8
24.ti'g3 ltlc4
Black has a huge advantage
thanks to the gaping dark squares
in the enemy camp. The rest of the
game went in a severe time trouble
and at some point White even be­
came better. in the endgame.
25.�d1 � 6e5 26. �d4 �g6
27.-efc3 �f4 2 8.-tfl 'efh7 29 .h3
'exe4 30.�f2 'efe3 31.�xb5
ti'xc3 32. �xc3 �e5 33.<.t>h2 c.t>d7
34.<.t>g3 d5 35.b4 l'!c8 36.�a4
l'!xc2 37.h4 �e2+ 38.he2 gxh4+
39.<.t>xh4 l'!xe2 40.�c5 + c.t>c6
41.'it>g3 .ic8 42.g5 c.t>b5 43.l'!b1
�g6 44.<.t>f3 l'!e5 45.�cd3 l'!f5+
46.'it>e3 l'!xg5 47.l'!c1 i.d7 48.l'!c7
.ic6 49. �d1 l'!g3 + 50.<.t>d2 d4
51.l'!g7 e5 0-1
 Pa rt S

Kan - the Ma roczy Bin d

This part covers a huge material so I separated the Kan from the Taimanov.
Actually, I analyse here both l.e4 c5 2 .4Jf3 e6 3.d4 cxd4 4.4Jxd4 a6 5.c4 and
5.i.d3 which aim to establish pawn control of the centre with c4.
After 5.c4 4Jf6 6.4Jc3 i.b4, Black's dark-squared bishop is not caged be­
hind a d6-pawn and it can split White's queenside.
5.i.d3 commonly leads to Mar6czy Bind structures, but the character of
play greatly depends on the move order. I offer two alternative approach­
es for Black:
Line Bl. 5 . . . g6 followed by . . . 4Je7 aims to eliminate the pawn centre with
... d7-d5. This is play for equalisation.
Line B 2 . 5 . . . 4Jf6 6.0-0 'fic7 7.�e2 d6 accepts the Mar6czy Bind. I advo­
cate to fianchetto the bishop on g7 from where it exerts pressure on White's
queenside and enables the positional threat of . . . i.xc3 .

197
Pa rt 6. Ka n - the M a r6czy B i n d
Main Ideas
The previous parts taught us that
Kan's utmost flexibility brings ex­
cellent dividends in all the systems
where White puts early his queen's
knight on c3. However, this flexi­
bility has its cost - it allows White
to build up the Mar6czy Bind. This
part should learn you to cope with
White's most cunning retort to the
Kan:
l.e4 c5 2.�f3 e6 3.d4 cxd4
4. �xd4 a6 5.i.d3
We are getting a taste of our own
medicine ! The enemy exploits the
lack of t2Jc6 and develops in a flexi­
ble way which keeps all his options
open. Depending on our set-up, he
may then return to t2Jc3 and long
castling, or opt for the restrictive
198
c2-c4. Black has tried in practice at
least five distinct schemes in retort.
I do not cover the 5 . . . �c5 6.tlJb3
�a7 system, because it does not suit
my taste. White trades dark-squared
bishops via e3 and the weakness of
d6 forces Black to enter Kalashni­
kov-type structures with . . . eS. They
are too rigid and look passive to
me. I have always preferred plans
with . . . dS (or .. .fS) in the Sicilian.
Besides, White is usually well pre­
pared against 5 . . . �c5 and we would
hardly surprise him in the opening.
I will offer you instead a choice
between two approaches. Both are
based on the fianchetto of the dark­
squared bishop, but the first one
aims to quickly eliminate the cen­
tre with . . . t2J e7 and . . . d7-d5 while
the second one assumes a rather
unusual hedgehog structure where
the g7-bishop longs to split White's
queenside pawns by taking the c3-
knight. I think that the surprise ef­
fect should be at your side.
In practice 5 . . . g6 and 5 . . . �c5
bring similar results.
 Kan - the Mar6czy Bind

Let's step back a little and see
what happens if White plays imme­
diately S.c4 tt:lf6 6.tt:lc3 .ib4
6.tt:lc3 .ig7 7 . .ie3 tt:le7 8 .Wfd2
should not scare us out of our main
idea: 8 . . . d5 9.tt:lb3 ! ? dxe4 ! ? 10.tt:lxe4
0-0 ll . .ih6 .ixh6 ! 12 .Wfxh6 tt:lfS
13.'?9d2 aS ! . A thematic method of
seizing space on the queenside by
tempo.

7 . .id3 tt:lc6 8.tt:lxc6 dxc6 9 .e5 '?9a5

is equal - 10.exf6 .ixc3+ ll.bxc3
Wfxc3 + so the top players prefer:

7.'?9d3 tt:lc6 8.tt:lxc6 dxc6

I give a detailed analysis of
the endgame which arises after
14.0-0-0 a4 15 . .ic4 '?9xd 2 + . Re­
member to avoid rook exchanges !
Instead of contesting the d-file, we
should put our heavy pieces on c8
and aS.

6 ••• .ig7 7.ft1b3

Alternatives do not hinder the

Here both 9.Wfxd8+ and 9.e5
freeing . . . d7-d5 - 7 . .ie3 (7 . .ie2 dS !
lead to balanced endgames where
8.exd5) 7 . . . tt:le7 8.tt:lc3 (8.0-0 dS
White has the bishop pair, but his
9.exd5 exdS 10.tt:lc3 0-0 ! ) 8 . . . d5
queenside is severely crippled, for
cuts across White's plans to torture
instance : 9 .'?9xd8+ 'it>xd8 10.e5 (10.
us in a Mar6czy Bind structure.
!d2 eS) 10 ... .ixc3+ ! 1l.bxc3 tt:ld7
12.f4 b6 13 . .ie3 c5 14 . .id3 'it>c7 15 . .if2
!b7 16.0-0 fS ! with counterplay.

Let's now return to the more so­

phisticated S.!d3 . I propose to play
for equalisation with:

5 .•• g6 ! ? 6.c4

199
 Part 6
Black's play is straightforward
and easy - 9.exd5 exd5 10.0-0 0-0
n.:gel tt:lbc6 12.tt:lxc6 bxc6 13..ig5
.ie6=.
7...tt:le7 8.tt:lbc3 tt:lbc6 9.0-0 0-0
This is a very interesting posi­
tion with mutual chances. White
has discouraged ...d7-d5 due to
the option of playing .ig5, but his
knight left the centre (and offered
us a tempo!). That gives us a free
hand on the kingside. Our offensive
could begin with ...f7-f5, leaving the
d-pawn on d7. The c8-bishop will
enter play from the queenside af­
ter ...b6, or, even better, ...b5! See
game 34 Ermenkov-Kotsur, Dubai
2000, which featured 10..ie2 f5!.
10.if4 will be met by 10...d5.
10 ..ig5!? h6 11..ih4 g512 ..ig3
ltle5 13.f4 gxf4 14 ..ixf4 ll:l7g6
15..ie3
200
Black's pieces are very active so
he can afford to sacrifice a pawn
with 15...b5! 16.cxb5 tt:lxd3 17.'\Wxd3
axb5.
You see that two hours of study­
ing suffice to start playing 5...g6.
If the idea of getting a symmetric
pawn structure with an early ...d5 is
revolting to you, my backup solu­
tion is:
5 .•• ltlf6 6.0-0 �c7
After this clever move which
awaits ...d6, we cannot escape the
Hedgehog. 7.c4 is a positional mis­
take due to 7...tt:lc6! 8.ie3 (8.tt:lxc6
dxc6 9.tt:lc3 e5- see game 35 Eich­
horn-Delchev, Bad Wiessee 2013)
8...tt:le5 9.h3 b6. Fortunately, the
queen move rules out the most dan­
gerous plans with ie3, �d2, :gdl.
Even more importantly, .id3 can­
not retreat to fl. We should use that
to trade the bishop with ...tt:lb8-d7-
e5xd3.
7 ..• d6 8.c4
We should be happy to see the
double-edged 8.f4 g6 9.tt:lc3 .ig7.
 Kan - the Mar6czy Bind

White's f-pawn cannot return to de­
fend e4 anymore. The only way to
justify its advance would be lO .eS
dxeS ll.tt:ldbS axbS 1 2 .tt:lxb5 Wffe 7
13 .fxe5 0-0 14.exf6 .ixf6, but our
king is safe and the e- and f-pawns
control the centre.
I analyse in detail the position af­
ter 1 2 . . . b6 13J'l:fdl �b7 14.�f2 !%ac8
lS.�fl !%fd8 (or 15 . . . !%fe8) 16.b4 Wffb 8
17.tt:lb3 - see game 38 Simacek­
Akopian, Plovdiv 2 0 1 2 .

8 ••• g6 ! 9 . tt:lc3 i.g7

M y conclusion i s that we should

not allow it at all !

The fianchetto has certain plus­

es : Our plan is to take the bishop,
1. It limits the scope of the d3- but let's give White the option of
bishop. displacing it to bl! Most practical
2. The bishop is more active on games went from this point 14.b3
g7. It takes an active part in the bat­ �b7 lS.i.bl? ! !%fc8 16.Wfff2 !%ab8
tle for d4 and eS enabling ideas like 17.tt:la4
. . . tt:ld7-e5 or . . . eS followed up by
. . . �g4.
3. A very interesting positional
idea is to cripple White's queenside
by . . . hc3, followed up by . . . eS,
. . . tt:ld7-b6.
Of course it has downsides,
too : Black spends an extra tem­
po, weakens the d6-pawn and the
dark squares around his king. Now
Note the rooks on b8 (protecting
White can try to put pressure on d6
b6) and c8 (hitting c4) . In the "Step
or follow a typical hedgehog plan :
by Step" chapter I consider two ac­
tive plans for Black. They involve
A. 10 .i.e3 0-0 lU�acl �bd7! shifting our queen to the kingside
(not ll . . . b6) 12.f3 �e5! (the b8-rook takes over the defence

201
 Part 6

of b6 ! ) followed up by . . . g6-g5. 13 . . . .ixc3 ! 14.bxc3 f6

In the event of:
14.�d2 , we have no choice
but kill the bishop: 14 . . . tt:lxd3
15.�xd3 tt:ld7! 16.b3 gds 17.�d2
i.b7 18.tt:lde2 .if8 19. �hl gac8
20.i.g5 ges
It is easy to play this position
with Black. See game 37 Leko­
lvanchuk, Mukachevo 2 007, for
more explanation.
Here is another version of the
same plan:
10.l2Jf3 0-0 ll . .if4

White has not a clear plan for

improvement.
B. Lately, White p refers more
energetic plans. He tries to pinpoint
our weakness on d6 with l2Jf3, l':l.dl,
.if4. They all lead to a similar pawn
structure:
ll . . . lLlhS ! 12 . .ie3 .ixc3 ! 13.bxc3
eS ! . See game 40 Grischuk-Ivan­
chuk, Beijing 2 0 13.
The idea of giving up our dark­
squared bishop sometimes works
even without the benefit of splitting
the enemy's queenside pawns :

lO.l"ldl 0-0 ll. l2Jf3 l2Jc6 12 .h3 Simek-Kanovsky

l2Jd7! (targetting the c3-knight) P rag ue 2011
13 . .ie3 (or 13 . .if4 .ixc3 14.bxc3 e5,
see game 36 Todorovic-Caruana,
Rijeka 2 010)

19 . . . .ixc3 2 0 . l':l.xc3 e5 21. f5 tt:lf4

22.fxg6 hxg6 23 . .ibl l2Jf6 24.'.We3
b5�.

202
 Kan - the Mar6czy Bind

On a final note, Black should not be scared of S.�d3 . This move is not
even White's first choice against the Kan. While the Mar6czy Bind is rather
restrictive, it is far from the true spirit of the open Sicilians. The l.e4 players
are commonly not big fans of long positional manoeuvring. Even when they
choose S.�d3 because it is supposed to be the "best" answer to 4 . . . a6, they
tend to treat it over-aggressively. That gives us fair chances for full-fledged
counterplay.

2 03
Pa rt 6. Ka n - the M a r6czy B i n d
Step by Step
l.e4 c5 2.�f3 e6 3.d4 cxd4
4.�xd4 a6
This chapter deals with White's
approach based on the Mar6czy
Bind. He either plays A. 5.c4 imme­
diately, or delay it by B. 5.i.d3 .
A. 5.c4 �f6 6. �c3
6.i.d3 ltlc6 gains control over the
dark squares and White often fails
to even equalise. Here are two ex­
amples:
7.ltlxc6 dxc6 8.ltlc3 e5 9 .'!We2
i.c5 10 .i.g5 h6 11.i.h4 i.g4 12.f3 i.e6
13.0-0-0 'WaS 14.i.c2 b5 15.i.b3 i.d4
16.i.e1 'Wb6+, Savchenko-Caruana,
Moscow 2010;
204
7.ltlc2 ltle5 8.0-0 'Wc7 9 . ltlc3 b6
10.ltle3 i.b7 ll.f3 i.d6+, Kokol-Kor­
neev, Nova Gorica 2 014
6 . . . i.b4
If you aim for the Hedgehog, you
should choose 6 . . . 'Wc7 7.a3 b6 8.i.e3
.ib7 9.f3 .ie7 10 . .ie2 0-0 1U'1c1 d6.
This structure is not part of the
opening, but rather of the middle­
game.
In the diagram position, White
has tested nearly all the legal moves.
I will focus on:
Al. 7 . .id3 ; A2 . 7.'Wd3 ; A3. 7.'Wc2 ;
A4. 7.'Wf3
Minor alternatives are :
 Kan - the Mar6czy Bind
a) 7.eS? lLl e4 8.�g4 (8.�f3
�aS 9.�xe4 hc3+ 10.'tt> d 1 hd4
11.�xd4 lLlc6; 8.�c2 lLlxc3 9.bxc3
i.e7 10.�e4 d6 ll.i.f4 �aS 12 .i.d3
lLld7) 8 . . . lLlxc3 9.a3 i.f8 10.bxc3
�aS (10 . . . d6+) 11.�g3 d6 12 .exd6
hd6 13.�xd6 �xc3+ 14.'tt> d 1 (14.
i.d2 �xa1 + 1S. 'tt> e 2 �b2 16. 'tt> e 3 i.d7
17.i.b4 �c1+ 18.�e2 �gS) 14 . . . �xa1
1S.i.d3 i.d7+, Enevoldsen-Portisch,
Munich 19S8.
b) 7.i.d2 0-0 ! 8.eS ! (8.i.d3 ? !
dS ! t) 8 . . . hc3 ! 9.hc3 lLle4
Black's defence is based on the
exchange sacrifice 10.i.b4 d6 ! (10 . . .
�b6 ll.hf8 'tt> xf8 offers Black
enough material, but a bad posi­
tion after 12 .E1b1 �aS+ 13.b4 �xeS
14.tLlf3 �c3+ 1S.lLld2 �d4 16.lLlxe4
�xe4+ 17.i.e2 �xg2 18 .i.f3 �g6
19 .�d6+ ) 11.�e2 �b6 12 .lLlc2 lLlc6
13.a3 dS ! . The stem game Korchnoi­
Furman, Leningrad 19S7, went fur­
ther 14.f3 lLlcS 1S.�e3 d4 16.lLlxd4
lLlxb4 17.axb4 �xb4+ = .
10.�c2 lLlxc3 1l.�xc3 dS ! (avoid­
ing ll . . . d6 12 .E1d1 ! ) 12.E1d1 (12 .exd6
�xd6 was thoroughly tested . . .back
in 19S7 ! )
12 . . . bS ! White's lag in develop­
ment prompts for energetic actions.
13.cS
13.cxbS axbS 14.-hbS E1xa2
1S.O-O i.d7 is balanced. White can
try to capture the pawn later, for
instance, 14.a3 i.d7 1S.lLlxbS lLlc6
16.f4 (16.ie2 lLlxeS 17.�xeS hbS
18 .-hbS �aS + ) 16 . . . �b8 17.lLld6
f6+, or 1S.hbS hbS 16.lLlxbS lLld7
17.f4 E1c8 (17 ... �h4+t) 18.�d2 �b6
19.lLld4 tLl cS 2 0 .0-0 lLle4+. 1S.i.d3
�b6 gives Black a "free" initiative.
13 . . . �c7 14.f4 f6 1S.ie2 fxeS
16.fxeS �xeS 17.lLlxbS = .
d ) 7.lLlc2 hc3+ 8.bxc3 �aS
(This is the most straightforward
equaliser. For more complex play,
you should opt for 8 . . . lLlc6 9.id3
d6 1 0 . 0-0 �c7 ll.f4 0-0 12 .lLle3 b6
or even 12 . . . bS ! ? . ) 9.�d3 lLlc6 (9 . . .
�eS ? ! 10.f3 dS 11.�d4;J;) 10 .i.e2 (10.
i.a3 dS) 10 ... �eS ll.f3 dS 12.exdS
exdS 13.f4 dxc4 ! = 14.fxeS cxd3
1S.exf6 dxe2 16.fxg7 E1g8 17.ih6 f6
18.'tt> x e2 'tt> f7 .
e) 7.f3 0-0
7 ... �c7 is often seen, but it is
connected with the passive set-up
2 0S
 Part 6
. . . tt'lbd7, . . . b6. The queen move does
not support the freeing . . . d7-d5.
For instance, 8 . .ie2 0-0 9.0-0 dS? !
is not enough due to 10.cxd5 .ixc3
ll.bxc3 exdS 12 .i.g5 tt'lbd7 13.tt'lf5
dxe4 14.'�d4t.
8 . .ie2 (8 . .ie3 does not hamper
8 . . . d5, but Black may also consi­
der the positional plan 8 . . . .ixc3 + ! ?
9.bxc3 d6. I t is effective when
White's bishop took away the e3
square putting an obstacle to the
manoeuvre tt'ld4-c2-e3.) 8 . . . tt'lc6
9 ..ie3 dS. Of course, White
can also play 9.tt'lxc6 bxc6 10 .e5
tt'JeB, but this is a fine version
of the Taimanov. Black will get
counterplay with .. .f6.
Al. 7 .id3 c!bc6
• Ahmad-Ni Hua, Macau 2 007) 10 . . .
This is the well known equaliser.
Its only apparent drawback is that it
leads by force to a drawish position
with opposite-coloured bishops. A
reasonable alternative is:
7 ... V!ic7 8.tt'lde2 tt'lc6 9.a3 .ie7
10 . .if4 d6 ll.b3 tt'leS 12 . .ic2 .id7
13.a4 c!bg6.
I have also analysed 7 ... e5
8.c!bde2 (8.tt'lf5 0-0 9.tt'le3 .ixc3+
206
10.bxc3 d6 ll ..ia3 V!ic7 12 ..ic2 Eid8)
8 . . . h6 ! (This move order aims to
avoid 8 . . . d6 9.0-0 h6 10.c!ba4 ! 0-0
ll.a3 .iaS 12 .b4 i.c7. 8 . . . .ic5 ! ? pur­
sues the same goal.)
9.0-0 i.cS ! ? 10.tt'ld5 · d6. The
game transposed to Perunovic-Ga­
jewski, Legnica 2 013 - ll.b4 ia7
12 . .ie3 tt'lxdS 13.exd5 .ixe3 14.fxe3
aS ! with counterplay, for example,
15.a3 0-0 16.Eic1 c!Dd7 17.'1Wd2 (17.
cS axb4 18.axb4 dxcS 19.bxc5 l"!aS
2 0 .c6 tt'lf6 = ) 17 . . . b6.
8 . c!Dxc6
a) 8.tt'lc2 .ixc3+ (8 . . . i.c5) 9.bxc3
d6 10.0-0 (10.i.a3 0-0 ll.f4 V!ic7
1 2 . 0-0 :!!.d 8 13.�f3 bS 14.cxb5 axbS
15.ib2 tt'l a5 16.tt'ld4 tt'lc4 17.ic1 eS+,
�c7 11.f4 0 - 0 12 .i.a3 l"!d8 13 .�e2 b6
14.l!ae1 .ib7 15.tt'le3 l!ac8.
b) 8.tt'lde2 i.cS ! ?
The idea of this retreat i s t o an­
ticipate tt'la4, followed by a3. 8 . . . d5
unloads the centre and should grad­
ually equalise. 8 . . . �c7 ! ? 9.a3 .ie7
was mentioned in the comments to
the previous move.
9.0-0 d6 (9 . . . tt'lg4 ! ? 10 .h3 tt'leS)
 Kan - the Mar6czy Bind
10 .h3 (10 .a3 0-0 ll.b4 i.a7
12 .i.f4 it)hS=) 10 . . . it)eS (or JO . . . h6)
ll.a3 i.d7 (Miezis' idea ll . . J'�g8 is
consistent due to the pawn lever
White provided with h3, but it is
still a little dubious. After 1 2 .b4 i.a7
13.i.g5 h6 14.i.f4, Black has to play
14 . . . it)fd7, because 14 . . . g5 15.he5
dxeS 16.c5 favours White.) 12 .b4
i.a7 13.i.f4 V!Jc7oo.
c) 8 .i.e3 it)eS! 9 . 0-0 (after 9.it)f3,
best is perhaps 9 ... it)xd3+ 10.V!Jxd3
bS ! 11.0-0 i.b7, destroying the ene­
my centre) 9 . . . V!Jc7 10Jk1 it)eg4.
d) 8.a3 hc3 + 9.bxc3 it)eS
9 ... V!Ja5 10.0-0 V!Jxc3 is dubious:
ll.it)xc6 dxc6 (ll . . . V!Jxa1 1 2 .V!Jb3
dxc6 13.i.b2 V!Jxf1+ 14.hfl 0-0
1S.V!Jg3 it)e8 16.c5) 1 2 J'l:b1 e5 13 .i.b2
V!JaS 14.V!Jc1 h6 1S.i.c3 V!Jc7 16.f4�.
10.0-0 (10.f4 it)xd3+ ll.V!Jxd3
dS) 10 . . . V!Jc7
ll.f4 it)xd3 12.V!Jxd3 d6. Black is
only one move shy of castling and
the pawn structure is favourable
for him. The only way to generate
some threats is 13.f5 eS 14.it)c2 bS
1S.i.g5 bxc4 16.V!Je2 it)d7 17.f6 gxf6
18.hf6, but both 18 .. J'!g8 19.V!Jh5
it)xf6 2 0 J'!xf6 l:!g6 21.l:!f2 i.b7 and
18 . . . it)xf6 19.l:!xf6 i.e6 2 0.it)e3 l:!g8
give Black good piece play while the
extra pawn does not get in his way.
8 dxc6 9 .e5 (or Black plays
•••
. . . eS himself)
9 V!Ja5 10.exf6 .ixc3+ ll.bxc3
•••
V!Jxc3+
12 .id2
•
12 .V!Jd2? is a good news for
Black. With an extra exchange plus
a couple of pawns he can always
give back a part of the loot: 12 . . .
V!Jxa1 13.0-0 V!Jxf6 14.i.b2 [14.i.a3 e5
aiming for a long castle - 15.l:!e1 (15.
V!Jb4 V!Jd8 16J'!e1 V!Jc7 17.f4 i.e6 18.f5
0-0-0) 15 . . . i.e6 16.i.d6 l:!d8 17.c5
l:!xd6] 14 . . . e5 (14 . . . V!Je7) 15.l:!e1 (15 .
.ixe5 V!Jxe5 16.l:!e1 V!Jxe1+ 17.V!Jxe1+
i.e6) 15 . . . V!Jh6+.
2 07
 Part 6
12 . . . ti'xd3 13.fxg7 gg8 14 .ih6
• I do not like 7 ... 0-0 due to 8.e5 ! ,
Yfc3 +
14 . . . �e4+ is similar: 15.Be2 (15.
@fl �h4 16.Bc1 f6 17.\t>g1 e5 18 J%b1
�f5 19J%xb7? 0-0-0 20J%a7 �xh6-
+ ) 15 ... �xe2+ 16.\t>xe2 e5 17.f4 (17.
c5 f6) 17 ... exf4 18.\t>d3 �e6 19J%hfl
b5 2 0 J!xf4 hc4 + = .
17 ••• �e6
17 . . . b5 is also possible, but the
text is more forced. The follow­
ing mass elimination leads to a
completely balanced endgame :
18.:t'!:xb7 0-0-0 19J�ia7 �h4 2 0.:t'!:a8+
\t>c7 21.:t'!:xd8 :t'!:xd8 2 2 .h3 �xc4+
23.�xc4 hc4+ 24.\t>e1 f5= , Vande­
voort-Kveinys, Charleroi 2 0 07.
A2 . 7.ti'd3
This is perhaps the most chal­
lenging continuation. White pre­
pares e5, �f4, 0-0-0 and �g3 .
7 .ti�c6
••
208
for instance : 8 . . . tt:\g4 9.�g3 �a5
10 .�e2 h5 ! ? (10 . . . tt:\xe5 ll.tt:\b3 �c7
12 .�h6 tt:\d3+ 13.hd3 �xg3 14.hxg3
gxh6 15.:t'!:xh6 f5 16. \t>e2) 11.�xg4
hxg4 1 2 . 0-0± or 8 . . . tt:\e8 9 .�f4 ! d6
10.tt:\c2 hc3 + 11.�xc3:t.
7 ... d5 is playing for two re­
sults only: 8.exd5 exd5 9.�e2 0-0
10.0-0 tt:\c6 ! ll.tt:\xc6 bxc6 12 .�g5
a5 (Rublevsky's idea) 13.cxd5 �a6
14.�c2 !t.
7 . . . �c7 is the most popular move,
but it leads to the Hedgehog after
8.a3 �e7 (8 . . . hc3 + 9.�xc3 tt:\xe4
10 .tt:\b5 ! is the point of White's idea
- 10 . . . axb5 ll.�xg7 :t'!:f8 12 .�h6 �c5
13.f3 ! , N.Ninov-Pikula, Lazarevac
1999) 9.�e2 b6.
Black can try to stay in Taimanov
waters with 9 . . . tt:\c6 1 0 . 0-0 tt:\xd4
11.�xd4 �d6 12 .\t>h1 �e5, but White
retained some edge after 13.�e3
0-0 14.tt:\a4 b5 15.tt:\b6 :t'!:b8 16.tt:\xc8
:t'!:fxc8 17.f4 �d6 18.e5 �c5 19 .�f3:t,
Leko-Bologan, Beijing 2 0 1 2 ; or
9 . . . 0-0 1 0 . 0-0 tt:\c6 11.\t>h1 :t'!:d8 12.f4
d5 13.cxd5 exd5 14.e5 tt:\e4 15.tt:\xc6
�xc6 16.f5, Giri-Landa, Eilat 2 0 1 2 .
8 .fi � xc6 (8.�d2 0 - 0 ) 8 ••• dxc6
 Kan - the Mar6czy Bind
A2 1. 9.�xd 8 + ; A2 2 . 9.e5
A21. 9.�xd8+ �xd8 10.e5
Or 1 0.�d2 e5.
10 • • • hc3 + !
Black has played s o far only 1 0 . . .
ttJ e4? ! when ll.a3 ! hc3 + 12 .bxc3
b5 13.h4 ! gives White strong pres­
sure on the kingside, for i n stance :
13 . . . ttJxc3 14J '!h3 ttJ a4 15J!g3 l'l:g8
16 .�h6 g6 17.�g5 + . This exam­
ple suggests that we should try to
blockade the position rather then
open it against White's bishop pair.
ll.bxc3 tlJd7 1 2 .f4 b6 13 .ie3
c5 14 .id3 �c7 15 .if2
• •
White hinders . . . l=l:d8 which
would be possible after 15. 0-0 l=l:d8
16.l'l:ad1 tlJf8 = . After the text, Black
should not wait passively because
White will carry on f4-f5, for in­
stance: 15 . . . ttJf8 16.0-0 tlJg6 17.l=l:ad1
tlJe7 18.g4 �d7 19.f5 h5 2 0 .�g3 h4
21.�f4±.
The solution is to blockade
the centre with . . . f5. White will
probably not allow it, but the var­
iation 15 . . . �b7! 16.0-0 f5 ! 17.exf6
ttJxf6 18.l'l:ae1 l'l:ad8 19.�b1 l=l:d2 ! is
roughly equal.
Black's rook is so active on the
second rank; that White should
be careful to maintain the bal­
ance : 2 0 . l'l:xe6 l'l:e8 21. l'l:xe8 ttJxe8
•
2 2 .l'l:e1 tlJf6 ! ( 22 . . . ttJd6 might be
drawish, but only for computers -
23.hh7 l'l:xa2 24.h4 a5 25.f5 tlJe4
26.l'l:xe4 he4 27.f6 l'l:a1+ 2 8.'iflh2
gxf6 29.he4 f5 30.�5 a4 31.h5 a3
3 2 .h6 a2 33 .h7 l'l:h1+ 34.'it>xh1 a1�+
35.'it>h2 �xc3 = ) 23 .h3 a5 24.g4 (24.
a3 'it>b8 25.g4 a4 2 6.g5 tlJh5=) 24 . . .
�a6 = .
A2 2 . 9 .e5 �xd3 !
White's spatial advantage is
more substantial with queens. His
centre is a bit shaky, but once he
consolidated, his prospects for a
kingside attack would be consid­
erable - 9 . . . ttJd7? ! 10 .�g3 (10.�f4
�a5) 10 . . . �a5 11.�f4 hc3+ (11 . . .
tlJf8 12 .�d2) 12 .bxc3 tlJf8 .
At the surface, it seems that
Black has a good game. For exam­
ple:
209
 Part 6

2 0 1 0 saw 13 . . . 0-0 14.�fl ! eS, when

instead of 15.f5, White had 15.fxe5.

13.�e3 tlJg6 14 . .ig3 b6 (14 . . . tlJe7=

15.�d2 tlJfS 16 . .if4 0-0 17 ..id3 �dB
1B.�d1 bS) 15 . .ie2 cS 16 ..if3 (16.0-
0 0-0 17.�fd1 tlJe7=) 16 . . . �a7 17. 0-0
.ib7; 13 ..ie2 tlJg6 14.0-0 0 -0 = .
14.f5 0-0 15JU1 .ixc3 !
16 .ixc3 c!tlg4 17.<i>e2 c!tlh6=
However, the stranded cB-bish­
•

op kept bothering me and a deep­

er analysis confirmed my fears. In
A3. 7.�c2 0-0 !
the diagram position, White has the
awkward manoeuvre:
This move is dubious after 7.�d3
13.h4 ! tlJg6 14.�h3 ! ! i.d7 (14 . . . h5
since White can meet it by 8.e5 tlJg4
15 . .ie2 ) 15 . .ie2 hS 16.�d1 cS 17.i.g5
9 .�g3. Against 7.�c2 , it is the most
i.c6 1B.�d2 i.e4 19.�h1 and he is
principled retort. It has no decent
dominating all over the board. It is
alternatives if we want to avoid the
better to kill the enemy's strongest
main Hedgehog lines which arise
pieces right away.
after 7 . . . �c7 B.a3 .ie7.

10.hd3 c!tld7 ll.f4 Note that 7 . . . tlJc6 B.tlJxc6 dxc6

9.e5! is dubious for Black: 9 . . . tlJd7
n ..tf4 .tas 12 . .ig3 .ic7 13.f4 f6 (9 . . . tlJg4 10 . .ie2 tlJxeS ll.0-0�; 10 .a3
14.0-0 fxeS 15.fxe5 .txeS equalis­ .icS ll.tlJe4 .td4 1 2 .i.f4 �b6 13.�b1
es: 16.�ae1 i.xg3 17.�xe6+ <i>dB c5 14.b4 0-0 15 . .ie2 tlJxe5 16. 0-0oo)
1B.hxg3 tlJf6 19.�ee1 �eB 2 0 . tlJ e4
tlJxe4 21.�xe4 �xe4 2 2 .i.xe4 'i!;e7
23 . .txh7 .ie6 24.b3 �dB. Black's
king will invade the queenside via
the dark squares.

u . . . f6 12.exf6 ttlxf6 13.i.d2

e5!

Kovacevic-Miezis, Nova Gorica

210
 Kan - the Mar6czy Bind

10.a3 ! All the other moves allow the

I n his survey for New in Chess
Yearbook 1 0 1, Ninov gives as a main
line 10.f4 f6 ll.exf6 �xf6 12 .i.d2
0-0 13.0-0-0 lt'l c5 14J"l:e1, but then
14 . . . eS ! takes over the initiative.
It is much stronger to play
13.i.d3 ! when 13 . . . �h4 + ? ! is po­
sitionally grim, but 13 . . . g6 ! holds
firmly, for example, 14.0-0-0 eS
(or 14 . . . lt'l cS 1S.E:he1 eS 16.fxeS �e7
17.a3 i.xc3 18 .�xc3 E:d8 19.E:e3
break . . . dS :
a) 8.i.e2 dS= 9.exdS exdS l0.0-0
dxc4 ll.E:d1 �c7 12 .i.gS (12.lt'le4
lt'lxe4 13.�xe4 i.d6 14.lt'lf3 lt'ld7
1S.�xc4 �xc4 16.i.xc4 lt'leS) 12 . . .
lt'lbd7= .
b) 8.i.d3 dS ! 9.cxdS exdS 10 .eS
E:e8 11.0-0 !'!:xeS 12.lt'lf3 E:hS ! .
s .tog4 9.f4 .ic5
••

lt'lxd3+ 2 0 . E:xd3 E:xd3 2 1.�xd3

The quieter 9 . . . d6 ! ? is also possi­
i.fS 2 2 .�e3 E:e8 23 .i.c3 �h4 24.b3
ble - 10.lt'lf3 dxeS ll.h3 lt'lh6 12 .fxeS
�xh 2 = ) 1S.a3 i.xc3 16.i.xc3 (16.
lt'lfS 13.a3 JJ.e7 (13 . . .i.cS 14.i.f4
�xc3 exf4 17.�xf6 lt'lxf6 18.i.xf4
lt'lc6 1S.E:d1 �e7 16.�e4 f6) 14.i.f4
lt'lg4 19.E:hfl i.fS 2 0 .i.d6 lt'le3
lt'lc6 1S.E:d1 �aS 16.i.d3 f6, Ninov­
21.i.xf8 E:xf8 2 2 .i.xf5 gxfS 23.E:f3
Shchekachev, Bethune 2 0 0 0 .
lt'lxd1 24.'.t>xd1 E:f7=) 16 . . . �xf4+
17.'.t>b1 lt'l cS 18.E:hfl �xf1 19.i.xf1 i.fS
2 0 .i.d3 lt'lxd3 2 1.'.t>a1 E:ad8 2 2 .�e2
lt'lf4 - the endgame after 2 3.�e1
lt'lxg2 24.�g1 E:xd1+ 2S.�xd1 lt'le3
2 6.�d2 lt'lc2 + 27.ri!?a2 lt'ld4 2 8.hd4
exd4 29.�xd4 is a draw. If White's
king goes to b6, Black starts push­
ing his passed pawn on the g-file.

lO . . . �aS (10 . . . i.e7 ll.f4 f6

12.exf6;!;) ll.f4 i.xc3 + 12 .bxc3 ! ( 1 2 .
�xc3 �xc3 + 13 .bxc3 b6 14.i.e2 i.b7 10.lt'lb3
1S.O-O cS= ) 12 . . . 0-0 (12 . . . bS 13 .i.e3
lt'lb6 14.E:bU) 13.i.d3. White has an 10.lt'lf3 lt'lc6 ll.h3 lt'le3 12 .i.xe3
attack and the queen cannot help i.xe3 13.g3 dSi.
much from aS. If Black attempts to
close the kingside, he lands in a dif­ 10 •.• i.f2 +
ficult position - 13 . . .fS 14.0-0 lt'lcS
1S.E:d1 E:d8 16.i.e3 E:xd3 17.E:xd3 It would be a sin to miss this
lt'lxd3 18.�xd3 bS 19.E:dl±. check. Still, 10 . . . i.a7 ll.cS (ll.i.e2
fS) ll . . . b6 12.lt'le4 (12 .i.d3 fS 13.exf6
8.e5 lt'lxf6 14.lt'le4 bxcS 1S.lt'lxf6+ �xf6

211
 Part 6
16.hh7+ 'i!ih8+) 12 . . . bxc5 is possi­
ble, too.
11 . . . '%lfh4 ! ? was played in Nanu­
Markowski, Plovdiv 2008.
12 .h3
Vuckovic-Pikula, Zlatibor 2 0 0 6
ended u p badly for Black after 12 . . .
Wfh4 ! ? 13.c5 (13.tt'le4 fS) 13 . . . '%lff2 +
14.'i!id1 tt'le3+ 15.he3 Wfxe3 16.'%lfe4
i.xc5 17.tt'lxc5 Wfxc5 18.i.d3 g6 19 .h4
dS 20 .'%lff3, but the engines assure
me that after 2 0 . . . '%lfe7! 2 l .h5 tt'lc6
only Black has an attack. I do not
have any reason to doubt this as­
sessment, but common sense and
my practice of Taimanov/Kan play­
er suggest a simple and safer solu­
tion:
12 •.• tt'lh6+
This is my fifty cents in the the­
ory of this variation. Now Black's
castling position is unassailable
while the enemy king in the centre
is a ridiculous sight. Possible con­
tinuations are : 13.g4 tt'lc6 14.i.g2
(14.f5 Wfh4 ; 14.tt'le4 aS ! ) 14 . . . d6t.
A4. 7.'%lff3
This move has obvious pluses
- Black cannot trade queens like
in the 7.'%lfd3 variation. Compared
to 7.'%lfc2 , the queen is more active
on the kingside. Its drawback is not
212
so conspicuous. White releases the
pressure on d6 and allows :
7 .ixc3 +
• ..
Exploiting the fact that 8.'%lfxc3
drops a pawn. 7 . . . d6 8.tt'lc2 ! hc3 +
is not enticing since the g7-pawn is
already hanging.
8.bxc3 d6
9 . .ia3
Piece attack on the kingside is
ineffective. In this pawn structure,
Black's knight pair is more power­
ful than the bishops.
a) 9.'%lfg3 0-0 10.i.h6 tt'le8, fol­
lowed up by . . . eS. Alsina Leal-Yu
Yangyi, Moscow 2011, saw 10 .i.d3
e5 11.tt'lf5 hf5 1 2 .exf5 tt'lbd7! 13.0-0
'i!ih8 14J'l:d1 l'!c8 15.i.e2 WfaS 16.�e3
dS+.
b) 9.i.e2 ! ? 0-0 1 0 . 0-0 tt'lbd7.
Black has no compelling reason to
define the centre at this point with
10 . . . e5 ll.tt'lfS hfS 1 2 .exf5 although
it is possible - 12 . . . tt:Jbd7 13 .'%lfxb7
tt'lcS 14.�f3 WfaS. He can quietly
complete development with natural
 Kan - the Mar6czy Bind
moves while White's play is not tri­
vial. Perhaps he will have to return
to the plan with .ia3.
9 ••• 0-0! lOJ;dl
10 .lL'lb3 �e8 (escaping the pin and
preparing . . . d6-d5) ll.�d1 (ll . .id3 ? !
ltJ c 6 1 2 . 0-0 ltJ eS ; ll . .ie2 ltJ c 6 1 2 . 0-
0 lL'leS 13 .�e3 bS 14.c5 dS 15.�fd1
.ib7 16.f4 ltJc4 17 . .hc4 bxc4+) 11 . . .
d S 12 .e5 ltJe4 1 3 . .id3 'IWgS ! 14.0-0
ltJc6 15 . .he4 dxe4 16.'1Wxe4 �xeS is
equal.
10 e5! ll..!Df5 hf5 12.exf5
•••
.!Dc6 ! 13.J.e2 (13 . .hd6 ltJd4 ! ) 13 •••
'flta5 1 4 .ixd6 c!Dd4 15.'fltd3 c!Dxe2
• and Black's play is clear and easy -
16.'fltxe2 'fltxc3 + .
Black has seized the initiative,
Dominguez Perez- Ivanchuk, Nice
2010.
B. 5.J.d3
BL s . . . g6; B 2 . s . . . lL'lf6
Bl. 5 . . . g6
If you think that this move of­
fends the basic strategic princi­
ples of chess, bear in mind that it
was employed in the past by fin­
est positional players like Smyslov,
Vasiukov and recently by Kamsky,
Bacrot. In contrast with the Hedge­
hog waiting approach, Black aims to
open the centre early with . . . ltJg8-
e7 and . . . d7-d5 or .. .fS . I would like
to warn you right away that this is
straightforward play for equalisa­
tion. In line B 12. 6.c4, we often see
mass elimination and symmetric
positions where White's initiative
gradually evaporates. On the other
hand, there is little theory to learn
he makes obvious moves and lets
the opponent sweat about how to
generate tangible threats.
If you want to keep more ten­
sion, you can always postpone . . . dS
in favour of . . . d6 and enter a non­
standard hedgehog with . . . ltJe7 and
. . . g6. However, although practice
experience has been favourable for
Black, according to my database, I
feel that White has the upper hand.
Bll. 6.lL'lc3 ; B12. 6.c4
Minor alternatives are :
6 . .id2 ig7 7 . .ic3 eS 8.lL'lf3 lL'lc6
9.0-0 ltJge7 10.ltJbd2 dS+ ;
6.lL'ld2 .ig7 7.ltJ2f3 d6 8.0-0 ltJf6
9.'1We2 0-0 10.�d1 lL'lbd7 11.lL'lb3 �c7
12 .c3 b6 13.h3 ib7 14 . .if4 e5 15 . .ig3
�feB, Bauer-Bologan, Germany
1998.
213
 Part 6
6.0-0 �g7 will most probably
transpose to the main lines, e.g.
7.�e3 tiJe7 8.c4. Instead, the game
Sevian-Stripunsky, Ledyard 2 0 14,
introduced the novelty 7.tiJf3 tiJc6
8.c4 tiJge7 9.tiJc3 0-0 10 .'�b3
We can follow the main plan
with 10 . . . d5 ! (of course, you can also
create havoc on the board with 10 . . .
f5 1l.�e3 b S 12.cxb5 axbS 13.hb5
fxe4 14.tiJxe4 dS) ll.cxdS exdS
12.exd5 tiJaS [12 . . . tiJxd5 13.�g5 (13.
�xdS hc3) 13 ... tiJf6 14J'Ud1 �aS = ] .
B U . 6.�c3 i.g7 7.i.e3
a) Another plan is to lead out the
bishop to a more active square as f4
or gS :
7.tiJb3 tiJc6 8.0-0
8.�f4 is typically parried with
S . . . tiJeS.
Smyslov answered 8.�e3 with
8 . . . d5 9 .exdS exdS and took over the
initiative after 10 .�c5?! b6 1l.�a3
tiJge7 12.0-0 0-0 13.tiJe2 �c7 14.c3
tiJeS 15.tiJbd4 tiJc4, Silman-Smys­
lov, Lone Pine 1976.
8 ... tiJge7 (8 . . . d6 is more solid,
but it allows White to stay in his
comfort zone)
214
9.�g5
9.�f4 tiJeS 10 .�e2 fS ! is a the­
matic break. Lutz-Bischoff, Es­
sen 2 0 0 1 went ll.�gS tiJf7 12 .�e3
(12 .he7 Wfxe7 13.exf5 gxfS 14.�h5
0-0 15.f4 bS) 12 . . . fxe4 13.tiJxe4 dS
14.tiJc3 tiJfS 15.�c5 b6 16.�b4 tiJeS
with excellent piece play.
9 . . . 0-0
Undoubtedly, this is the most
flexible approach. 9 . . . h6 weakens
the g6-square although White has
not much after 10 .�h4 (10.�f4 tiJeS
is familiar from the previous exam­
ple. The insertion of . . . h6 brings
about new possibilities for kingside
play: ll.Wid2 gS 12 .�e3 tiJg4 13 .�d4
�c7 14.g3 hd4 15.tiJxd4 d6.) 10 . . .
dS ! ?
10 . . . 0 - 0 ll.f4 d S 12 .e5 i s un­
pleasant. Houdini 4 suggests the
extravagant 10 . . . hc3 ll.bxc3 gS
12 .�g3 eS claiming full equality.
However, in this pawn structure the
knight should stay on f6 so I'm a bit
reserved about this line. 10 . . . 0-0
1l.f4 fS ! 1 2 .exf5 (12 .g4 fxe4 13.he4
�b6+ 14.�f2 Wfc7) 12 . . . gxf5 is stra­
tegically highly unbalanced. Black
is behind in development, but the
strong pawn control of the cen­
tre should keep him safe until he
 Kan - the Mar6czy Bind
regrouped: 13.�f3 'fffc 7 14.Wfd2 b6
15.i.f2 l!Jg6 16.�d1 l!Jce7 17.i.e2 i.b7
18.�g3 i.c6 19.i.d4 �f7 20 . .ih5 �h7.
Black has counterplay against the
f4-pawn and down the g-file.
11.exd5 exdS. Now 12.f4 0-0
13.f5 Wfb6+ 14.i.f2 Wfc7 15.fxg6 fxg6
gives Black a good game mostly due
to the eS-square for his minor piec­
es. On the other hand, 12 .Wfd2 0-0
13.�fe1 i.e6 is only slightly better
·
for White.
10 .Wfd2 (10.f4 fS) 10 . . . d6 11.�ad1
'ffc 7 12 .i.h6 i.xh6 13.Wfxh6 bS
14.f4 (14.a3 l!JeS), Wiwe-Stripun­
sky, Copenhagen 2 0 07. Here, 14 . . .
b 4 15.l!Je2 fS assures Black o f full­
fledged counterplay, e.g. 16.�h1
fxe4 17.i.xe4 i.d7.
b) 7.l!Jf3 l!Jc6 8.i.f4 dS 9.exd5
(9.e5? f6 10.'fffe 2 fxeS 11.l!Jxe5 l!Jge7
12. 0-0-0 0-0+, Li Chao2-Wang
Chen, Chengdu 2 011) 9 . . . exd5
10.0-0 l!Jge7 11.�e1 0-0 is a typical
position for this line.
The game Martinovic-Strikovic,
Zlatibor 1989, went 1 2 .'fffd 2 (12 .h3
i.e6) 12 . . .i.g4 13.l!Je5 i.e6 14.l!Ja4
l!JxeS 15.i.xe5 i.xeS 16.�xe5 Wfd6
and a draw was signed a few moves
later.
7 . . . ti)e7 8.'lfd2
8.f4 dS 9.e5 l!Jbc6 10.l!Jce2 f6
11.exf6 i.x£6 1 2 . 0- 0 0-0 13.c3 is sa­
tisfactory for Black since he can
easily develop his light-squared
bishop. For instance, 13 . . . e5 solves
the problem at once while 13 . . . 'fffc 7
14.l!Jf3 i.d7 is also possible.
8 . . . d5
The main idea of this system.
Black opens the centre and gradual­
ly neutralises the enemy initiative.
9.exd5 l!JxdS lO.l!JxdS 'fffxdS 11.0-0
0-0, Leko-Vyzmanavin, Leon 1993,
was extensively tested in practice
and proved to be innocuous for
Black. The plan with long castling
seems more dangerous :
9.ti)b3 ! ? dxe4 ! ?
9 . . . l!Jbc6 10 .i.h6 i.xh6 ll.'fffx h6
l!Jg8 12 .Wfg7 (12 .Wfd2 l!Jf6) 12 .. .'�f6
13.'�xf6 l!Jxf6 14.exd5 exdS offers
White only a symbolic advantage, in
my opinion. However, the text leads
to more complex play.
10. ti)xe4 0-0 ll . .ih6 .ixh6!
215
 Part 6
Black has played only ll . . . tt:'lbc6
so far when 12 . .ixg7 c;!,lxg7 13.h4 fS
14.tt:'lgS? h6 was fine for him. How­
ever, 14.tt:'lc3 ! h6 1S.O-O-O looks
rather dangerous. I propose to play
more concretely:
12.Yfxh6 �f5 13.ti'd2 a5 !
A thematic method of seizing
space on the queenside by tempo.
14.0-0-0
14.g4 (14.a4 b6) 14 . . . a4 1S.tt:'lbcS
b6 16.gxfS is strongly met by 16 . . .
exfS 17.Yfc3 fxe4 18.tt:'lxe4 tt:'ld7oo.
14 . . . a4 15 . .ic4 Wfxd2 + !
16. �bxd2 .id7
My analysis suggests that this
endgame is balanced. We only have
to avoid trading ANY rook. Instead,
we keep them on the queenside - c8
and aS. White cannot improve his
position. For instance :
17.tt:'lf3 .ic6 18.ghe1 a3 19.b3 bS
2 0 . .id3 b4;
17.tt:'lf6+ c;!,lg7 18.tt:'lxd7 tt:'lxd7
19.tt:'le4 tt:'lb6 ! [19 . . . tt:'leS 20 . .ie2 a3
216
21.b3 gadS 2 2 . gxd8 ( 22 .c3 gds
23.gd2) 22 .. ,gxd8 23.c3 ! (23.gd1
tt:'ld4 24 . .if1 tt:'lec6 2s.gd2 b6=) 23 . . .
tt:'lh4 24.ggu] 2 0 . .ie2 gas ! (we play
a middlegame ! ) 21. a3 ( 21.tt:'ld6? ! a3
2 2 .b3 tt:'ldS 23.tt:'lxfS+ exf5 24.c4 tt:'l c3
2s.gd2 ges 26 . .if3 bSt) 2 1 . . . ges
22 ..id3 gc8 = .
B12. 6.c4 i.g7 7. �b3
a) 7 . .ie3 tt:'le7 8.tt:'lc3
Or 8.0-0 dS 9.exdS exdS 10.tt:'lc3
0-0! - there is no reason to rush
with taking on c4 although 10 . . . dxc4
ll . .ixc4 0-0 12 .ge1 tt:'l d7 ! ? i3 . .igS
tt:'lf6 14.gc1 h6 1S ..if4 bS 16 . .ib3 ges
holds on.
8 . . . dS (or 8 . . . 0-0 9 . 0-0 dS)
The plan with . . . d6 may look
more flexible, but in my opinion it
is inferior to the text. For instance :
8 . . . 0-0 9.0-0 d6 10.Wfd2 ! tt:'lbc6
(10 . . . tt:'ld7 n.gfd1! with pressure
on d6 is clearly better for White.
He has a clear plan with gael, b4,
tt:'lb3, cS) ll.tt:'lxc6 ! tt:'lxc6 12 .gad1 e5
13.tt:'lds tt:'ld4 14.f4 .ie6 15.gf2 gcs
16.gdfl with an initiative. There­
fore, if Black wants to play a hedge­
hog structure, he should choose the
set-up with 5 . . . tt:'lf6 - see line B 2 .
 Kan - the Mar6czy Bind
9.exd5
Spassky-Panno, Manila 1976,
saw 9.cxd5 exdS 1 0 . 0-0 0-0 1L�c1
tt:ld7 ! ? 1 2 .tt:lxd5 (after 12.tt:lf3, we can
choose between 12 . . . dxe4 13.he4
tt:lf6 14 ..ic5 �xd1 1S.�cxd 1 ! �e8
16J�fe1 tt:lc6 17.hc6 bxc6 18.�xe8+
tt:Jxe8 19 . .id4 .ig4 20.hg7 hf3 ! =
and 1 2 . . . tt:Jf6 13 . .ic5 �e8 ! = ) 1 2 . . .
tt:JxdS 13.exd5 tt:lf6 1 4 . .ic4 tt:JxdS
15. �b3 tt:Jxe3 16.fxe3 when 16 . . .
�e7! evens the game.
9 . . . exd5 10.0-0 0-0 11.�e1 (11.
�c1 tt:ld7 ! ; ll.cxdS tt:lxdS 12.tt:lxd5
�xdS 13.�c1 �xa2 14 ..ic4 �xb2
is very interesting. White has am­
ple compensation for the missing
pawns. Perhaps it is safer to decline
the gift with 13 . . . tt:Jd7.) ll . . . tt:Jbc6
12.tt:lxc6 bxc6= 13 . .ig5 .ie6
White can grab a pawn, but he
will be unable to keep it for long after
14.cxd5 cxd5 1S.tt:lxd5 �xd5 16.he7
�feB 17 . .ia3 �aS, e.g. 18 .�e2 �a4
19.h3 �ab8 2 0 .ha6 ha2 .
b) 7.tt:le2 tt:Je7 8.tt:Jbc3 0-0 9 . .ig5 ! ?
This i s the only way to prevent
. . . dS. Commonly White plays 9.0-0
dS 10.cxd5 exdS ll . .igS dxe4, un­
loading the center - 12 .he4 (12.
tt:Jxe4 tt:Jbc6) 12 . . . �xd1 13.�axd1
tt:Jbc6 14.�fe1 h6 1S . .ie3 .ifS= .
9 . . . tt:Jbc6 1 0 . 0-0 h6
My analysis suggests that Black
equalises after 10 . . . �a5 11.�d2 dS
12.exd5 exdS 13;b4 ! (13.tt:Jxd5 �xd2
14.hd2 tt:lxdS 15.cxd5 tt:leS 16 . .ie4
tt:lc4 17 . .ic3 �e8 18.tt:lg3 tt:Jxb2
19.�acl �b8 2 0.hg7 lt>xg7 21. �c7
bS 2 2 .d6 .ie6 23.f4 tt:Jc4 24.f5 gxfS
25.hf5 ltJxd6) 13 . . . �xb4 14.cxd5
tt:lxdS 1S.tt:lxd5 �xd2 16 .hd2 �d8,
but the text is more interesting. It
aims to brea� the connection be­
tween .igS and �d2 .
l l . .ih4 �a5 12 .a3 (now the queen
will be pinned after 12 .�d2 dS) 12 . . .
�h5 13 . .ig3
13 . . . d5 14.cxd5 exdS 15.exd5
tt:JxdS 16.hg6 fxg6 17.tt:lxd5 .ig4
18.tt:Jdc3 gS. The bishop pair offers
compensation for the pawn.
c) 7 . .ie2 dS ! 8.exd5 exdS 9.tt:lc3
tt:Je7 10 . .ig5
10.0-0 0-0 ll . .ie3 dxc4 12 .hc4
tt:Jbc6 ; 10.tt:lf3? ! d4 11.tt:le4 f5 12.tt:lg3
tt:Jd7 13.tt:lxd4 (13.0-0 tt:lcS 14.�e1
0-0 1S . .if1 tt:lc6 16 ..ig5 �b6) 13 . . .
tt:Jb6.
2 17
 Part 6
10 . . . dxc4 11.'?;Va4+ .id7 12 .'?;Vxc4
lt:lbc6 13.lt:Jxc6 hc6 14.0-0 0-0
15J''� a d1 V9c7
Black has equalised, since
16.!1fe1 (16.lt:Je4 V9e5) 16 . . . !1ae8
17.lt:Je4 V9a5 18.ltJf6+ hf6 19.hf6
is well met by 19 . . . lt:Jd5 2 0 .i.g5 ( 2 0 .
.ih4 lt:Je3 21.fxe3 ge4 2 2 .!1d4 gxh4)
2 0 . . . lt:Jc3 ! .
7 ••• lt:Je7 8. lt:Jbc3 ltJbc6 9. 0-0
9 ..ie2 0-0 10.0-0 fS transposes
to game 34 Ermenkov-Kotsur.
9 . . . 0-0
10.i.g5
White often attempts to gener­
ate pressure down the d�file with
218
10 . .ie2 . We should counter-attack
the centre with lO . . .fS ! - see game
34 Ermenkov-Kotsur, Dubai 2 0 0 0 .
10 . . . h 6 ll.i.h4 (ll.i.f4 dS
12.exd5 exdS 13 .c5 ltJ eS) ll . . . g5
12 . .ig3 ltJe5 13.f4
Caruana opted for 13 .i.e2 against
Grandelius in Porto Carras 2 011.
The game went 13 . . . lt:\7g6 (13 . . .f5 ! ?)
14.�d6 ! (Seizing space. Or 14.V9d2
b6.) 14 . . . ltJf4 15.hf4 gxf4. It is dif­
ficult to understand what happens
here.
At first sight White is much bet­
ter, because the black bishop looks
stuck on c8. On the other hand, the
other one - on g7 - is extremely
powerful and its pressure over the
enemy queenside would be tangi­
ble in an endgame. Black's main
task should be to bring the sleeping
bishop into play. The simplest way
to achieve it is to sacrifice a pawn
with . . . bS. A finer point is the right
timing for this break. I join Grande­
lius' assessment that Black should
not be too hasty. Anyway, White's
only way to prevent it is 16.c5, but
then 16 . . . �g5 17.ltJd2 !idS threaten­
ing . . . .if8/.ie5, . . . d6, easily equal-
 Kan - the Mar6czy Bind
ises : 18.'it>h1 ltlc6 (18 . . . .tf8) 19.ltla4
.teS 2 0 .�d3 d6. The same method
works against 17.'it>h1 - 17 . . . lt:lc6.
Caruana decides to play on the
kingside. This is a really danger­
ous plan as the g-file might open in
White's favour:
16.ltld2 �gS (16 ... b5 was already
possible) 17.'it>h1 (17.lt:la4 b5 18.ltlb6
!!a7) 17 . . . 'it>h7 (17 . . .b5) 18.!!g1
White is fully prepared for an
offensive so I think now is the best
moment to distract him with:
18 . . . b5 !
Grandelius postponed it with
18 . . . !!g8 , but I'm afraid that 19.lt:la4
would have been unpleasant. Caru­
ana played inste.ad 19.!!ad1 i.f6
(19 . . . b5 ! ?) 2 0 .lt:lf3 lt:lxf3 2 1.i.xf3 (21.
gxf3 ! ?) when 2 1 . . .�e5 ! was called
for: 2 2 . lt:la4 (22 .�d3 bS 23.c5 !!a7
24.ltle2 b4 25.ltlxf4 aS 2 6.g3 .ta6
27.�e3 i.e7=) 22 . . . b5 23.ltlb6 �xd6
24.!!xd6 !!a7 25.!!d2 d6 2 6.!!c1 !!c7
27.!!cc2 !!c6 28.lt:lxc8 !!exeS = .
19.cxb5 axbS
I believe that Black has full com­
pensation for the pawn. Here is an
illustration:
a) 20.ltlf3 �hS 21.lt:lxe5 (21.
g4 ! ? lt:lxg4 2 2 . !!g2 fSoo 23.lt:le5 !!a6
24.�b4 d6 25.lt:lxg4 fxg4 26.!!xg4
�cS 27.�xc5 dxcS 2 8.hb5 !!a'Too ;
21.ltld4 �h4 2 2 .f3 lt:l c4oo) 2 1 . . .�xe5
2 2 .�xe5 .txes 23.hb5 d6 24.!!gb1
24 . . . .ta6 ! . White is unable to
hold his extra pawn due to the ac­
tivity of the black long-range pieces
- 25.a4 !!feB 2 6.i.xa6 !!xa6 27.!!a3
!!b6�.
b) 2 0 .lt:lxb5 lt:lg4 2 1.i.xg4 �xbS
2 2 . !!gb1 �b8 23.�xb8 !!xb8 24.i.e2
!!xb2 25.!!xb2 i.xb2 26.!!b1 i.c3
27.ltlf3 dS= .
c ) 2 0.!!ac1 .ta6 2 1.a3 b 4 2 2 . axb4
.txe2 23.lt:lxe2 lt:lg4 24.!!gf1 .tes
25.�d3 !!g8 2 6.g3 �hS with an at­
tack.
13 . . . gxf4 14 . .bf4 �7g6
The dark-squared strategy 14 . . .
�c7 15.�d2 (15.i.e3 b6 16.c5 bxcS
17.i.xc5 d6 18 . .te3 fS 19 .!!c1 lt:lg4)
15 . . . lt:l7g6 16.i.e3 lt:lxd3 17.�xd3
lt:leS is also possible, but we have
more aggressive intentions in store.
15 . .ie3
219

15 . . . b5!
Our bishop pair will be very effi­
cient in an open position. Black has
only tried 15 . . . b6 and 15 .. .'?9c7 so far.
16.cxb5 lilxd3 17.fixd3
axb5iii 18.lilxb5 .ixb2 19 . .ixh6
.ixal 20 J:�xal fih6+ 21.'.thl Aa6
22.Vc3 e5
Black is fine here. He only
should avoid Komodo TCECs re­
commendation 23.i.xf8
24.a4 c;!;>g7 since White gets a
strong attack with 25.lil3d4 ! ! exd4
26.4Jxd4 c;!;>f8 27.4Jf5±. Instead:
23 . . .l'�xf8 ! (23 . . . .txbs 24 . .tb4 V9f6
25.a3 lt:Jf4 is a decent alternative)
24.4Jc7 .tb7 25.4Jd5 .txds 26.exd5
220
Part 6
Vb7 27.fif3 fS. Black's mobile pawn
centre assures him of the initia­
tive. Perhaps 23.4Jc7 is more real­
istic, when 23 .. J:!ac8 24 . .txf8 V9xc7
25.Vxc7 l':ixc7 2 6.i.h6 l':ic2 is close to
the draw, but still White must make
it. Black maintains an attack due to
the threats of . . . .tb7, .. .fS, . . . lt:Jh4.
82. 5 . . . lilf6 6.0-0 fic7
7.fie2
Here is the big difference with
line Bl. If White wants to put in c4,
he has to move the queen first, but
that rules out dangerous plans with
i.e3, V9d2 , l':idl. To be sure, 7.c4 is
also possible, but I consider it as a
positional mistake. We answer 7 . . .
lt:Jc6 ! 8 . .te3 (8.4Jxc6 dxc6 9.lilc3 eS
- see game 35 Eichhorn-Delchev,
Bad Wiessee 2 013) 8 . . . 4Je5 9 .h3 b6!
c;!;>xf8? 9 ... .tc5 ! ? in Taimanov-style is
also possible : 10.lilc3 d6 ll.l':icl 0-0
12 .i.e2 h6 13.f4 (13.a3 i.d7 14.b4
.ta7 15.'?9d2 Eiac8 16.f4 lt:Jg6 17.4Jf3
i.xe3+ 18.V9xe3 Eifd8 19.Eifdl lt:JhS)
13 . . . 4Jg6 14.f5 lt:J e5 15.V9el i.d7 16.b3
exfS 17.exf5 lt:Jc6.
 Kan - the Mar6czy Bind

10.f4
This attack has no venom with­
out the light-squared bishop, but
after 10.lt:lc3 .ib7 1Uk1 (ll.f4 lt:lxc4
12 . .ixc4 �xc4 13J'k1 lt:lxe4 ! 14.lt:lcb5
axb5+) ll . . . .ie7 (ll . . . i.c5 ! ?) White
will have to push f4 anyway.

10 . . . lt:lxd3 11.�xd3 d6 12.lt:lc3

.ie7 13.l'!ad1 (13.l'!acl 0-0 14.f5 .id7
15.b3 l'!ac8 16.a4 �b7=) 13 . . . 0-0
14.b3 .ib7 15.f5 e5 16.lt:lf3 bS ! with
counterplay.
7.lt:lc3 and 7.<;t>h1 d6 8.f4 lt:lbd7
9.lt:lc3 transpose to the Third Rank
. tre. The game · went 15 . . ..ig7 16.a4
set-up.
7 . . . d6 8 .c4
White may keep us guessing
about his queenside plans with :
a) 8.f4. Then 8 . . . lt:lbd7? ! will
be met by 9 .c4 g6 10 .f5. An inter­
esting reply is 8 . . . e5 9 .lt:lb3 (9.lt:lf5
.ixf5 10.exf5 .ie7; 9.lt:lf3 .ig4 10.lt:lc3
lt:lbd7) 9 . . . i.e7 10 .lt:lc3 0-0 (10 . . .
lt:lbd7 11.g4) ll.f5 b5 12 .g4 b4 13.lt:ld5
lt:lxd5 14.exd5 lt:ld7 15.g5 .id8oo.
Perhaps it is best to fianchetto
the bishop :
8 . . . g6 9.lt:lc3
9.f5 .ig7 10 .fxe6 fxe6 ll . .ic4 is
bad due to ll . . .'�c5 12 . .ie3 lt:lg4
13.b4 �h5 14.h3 lt:lxe3 15.�xe3 lt:lc6
16.c3 .ih6 17.�g3 �e5+.
ll.lt:ldb5 axb5 12.lt:lxb5 �e7
13 .fxe5 0-0 14.exf6 .ixf6 15.c3, Wei­
Artemiev, Chongqing 2 0 13 . Black
should be fine - his king is safe and
the e- and f-pawns control the cen­
.id7 17 ..if4 e5 18 .b4 lt:la6 19.l'!ae1
when 19 . . . l'!fe8 2 0.i.c4 .ie6 2 1..ie3
l'!ac8 2 2 .i.xe6 �xe6 would have
been balanced. I suppose that 15 . . .
.id7 i s slightly more accurate. Pos­
sible continuations are 16 . .ie3 .ig7
17.'�f2 lt:la6 or 16.a4 l'!d8.
b) A rare set-up is 8 .b3 g6 9 . .ib2
.ig7 10.lt:ld2 0-0 ll.a4 (ll.@h1 b6
1 2 .f4 i.b7 13.l'!ae1 lt:lbd7 14.e5 lt:lh5
15.lt:lc4 dxe5 16.fxe5 .idS 17.lt:le3
i.xe5 18.lt:lxd5 exd5 19.lt:lb5 axb5
2 0 .i.xe5 lt:lxe5 21.�xe5 �c5 2 2 .b4
�xb4 23.l'!b1 �h4 24 ..ixb5 l'!xa2+,
Dutreeuw-Psakhis, Vienna 1991)
ll . . . b6

9 . . ..ig7 10 .e5 (or 1 0 .@h1 0-0

ll ..id2 b6 12 .l'!ae1 .ib7 13.lt:lf3
lt:lbd7oo, T.Bauer-Farago, Hungary
1993) 10 . . . dxe5

221
 Part 6
It is unclear what White should
do from this point. The game Geller­
Huebner, Wijk aan Zee 1975 went
12.lt:lc4 lt:lbd7 (12 . . . d5 ! ? 13.exd5
lt:lxd5=) 13J'l:fe1 i.b7 14.�d2 Ei:fd8,
preparing ... d6-d5. 14 ... Ei:ad8 would
have been even better.
8 g6 ! 9.lt:lc3 i.g7
•••
The main branching point of
this line. White must define his
plans. In practice, he achieves best
results by attacking the weakened
d6-pawn with B21. 10 .Ei:d1 or B 2 2 .
10.lt:lf3 (planning i.f4). The third
plan is to enter a standard hedge­
hog with B23. 10 .i.e3. The latter
is considered not very challenging
and scores close to 50%.
B21. lO,gdl 0-0 ll.lt:lf3 lt:lc6
12.h3
Alternatively:
a) 12.i.f4 faces 12 . . . e5 13 .i.e3
.tg4 14.h3 lt:ld4 15.i.xd4 .txf3
16.�xf3 exd4 17.lt:ld5 lt:lxd5 18.cxd5
�b6= , Svidler-Ivanchuk; Monte
Carlo 2005.
222
b) 12 .i.c2 is an original ma­
noeuvre. The game Shirov-Wang,
Shanghai 2 0 1 0 , illustrates its main
idea: 12 . . . lt:ld7 13.i.e3 lt:lde5 14.i.b3
i.d7 15.�d2. The c4-pawn is de­
fended while the d6-one is ripe for
collection. I recommend the more
active retort 13 . . . lt:la5 ! 14.i.b3 i.xc3
15.bxc3 b6 ! ?oo or 15 . . . e5 16.�d3 lt:lc5
17.i.xc5 dxc5 = .
12 . . . lt:lg4, preventing i.e3, i s also
possible - 13.i.f4
Naiditch's 13.lt:le1 looks over­
ly ambitious. We can take up the
gauntlet - 13 . . . lt:lge5 14.f4 lt:ld7
when the c4-pawn suddenly turns
to be very weak after 15.lt:lf3 lt:la5
16.i.d3 (16.i.b3 lt:lb6) 16 . . . lt:lc5.
13 . . . lt:lge5 14.i.b3 (14.Ei:ac1 lt:lxf3 +
15.�xf3 e5 16.i.e3 lt:ld4) 14 . . . i.d7
when Shirov's move 15.�d2 from
the above-mentioned example fails
to 15 . . . lt:lxf3+ 16.gxf3 lt:ld4.
12 . . . lt:ld7! 13 .te3
•
13.i.f4 provokes . . . e5, but this
moves is on Black's agenda anyway.
After 13 . . . i.xc3 14.bxc3 e5, the pawn
structure is similar to the main line.
See game 36 Todorovic-Caruana,
Rijeka 2 0 1 0 .
Only 13.i.d2 prevents 13 . . . i.xc3
(although the engines still like this
capture ! ) . Then 13 . . . b6 14.Ei:acl
i.b7 15.i.e3 leads to positions from
line B23 with an extra tempo for
Black. He can follow up with 15 . . .
lt:lc5 16.i.b1 Ei:fe8 17.b3 Ei:ad8, hav­
ing . . . t7-f5 in mind.
 Kan - the Mar6czy Bind

after 12 . .id2 0-0 13.l'!ac1, the knight

would be a bit awkward on c6. By
all means this continuation has no
advantages over the main line.)
12 . .id2 ! (12 . .ie3 .ig4 13.h3 ltJd4 ! = )
12 . . . .ig4 13.�e3 and White can or­
ganise an offensive on either flank.

l l .if4
•

ll.h3 eliminates the idea of . . . eS

followed up by . . . .ig4, but I noted
13 .••.ixc3 ! 14.bxc3 f6 above that it was not so attractive
anyway. We follow our main plan
A critical position for this line. In enjoying a clear extra tempo - 11 . . .
my opinion, it is very easy to play it ltJc6 1 2 . .if4 ltJhS ! 1 3 . .id2 (13 . .ie3
with Black. The last move is the first .b:c3 14.bxc3 eS 15.c5 ltJf4 16.hf4
link of a plan which aims to ensure exf4 17.cxd6 �xd6 18.e5 '?ffc S) 13 . . .
secure control of the dark squares. b 6 14.l'!ac1 .ib7 1S . .ie3
It includes . . . b6 and . . . ltJc6-d8-t7 if
needed. Nakamura opted for 14 . . .
b 6 against Dominguez i n Wijk aan
Zee 2 004, but 1S.�d2 ! iDeS 16 . .ic2
ltJaS could have turned out badly
after 17 . .ih6 ! l'!e8 18.�f4 eS 19.�f6
ltJe6 2 0 .ltJh2.
See game 37 Leko-Ivanchuk,
Mukachevo 2 0 07, for more expla­
nation.
White is threatening with ltJdS,
but we are just in time to prevent it
B 2 2 . lO.tD£3 0-0 with 15 . . . ltJeS ! . Then:
a) 16.ltJxe5 dxeS 17.g3 (17 . .ib1
10 . . . ltJc6 is less flexible as it de­
ltJf4 18.'?ffc 2 '?ffxc4 = ; 17.ltJa4 ltJf4
fines the placement of this knight
18.�c2 .ic6 19.ltJxb6 l'!ad8) 17 . . . f5.
too early while it would have been
more useful on d7 in some lines. b) 16.l'!fd1 ltJxf3 +
Black wants to seize the d4-square,
O r 1 6 . . . ltJxd3 17.'?ffxd3 l'!ad8=
but practice suggests that it might
be insufficient to balance the game. 17.�xf3 fS ! . Black has seized the
For instance, ll . .if4 eS (ll ... ltJhS, initiative: 18 . .ib1 l'!t7! 19.�e2 f4
intending . . . .b:c3, is possible, but 2 0 . .id2 f3 2 1.gxf3 l'!af8-+.

2 23
 Part 6
ll . . . l0h5!
It is already late to return to the
plan with ll . . . ll:\c6 12 J:!ac1 e5? in
view of 13 . .ig5 .ig4 14.ll:\d5.
12 .J.e3
12 . .ig5 would give an extra tem­
po for .. .f6 or . . . ll:\e6. For instance,
12 . . . .ixc3 13.bxc3 e5 14.ll:\h4, Shi­
rov-lvanchuk, Motril 2008, 14 . . .
ll:\g7 15.g3 .ih3 16.ll:\g2 ll:\ e 6 1 7. .ih6
l'!d8 18.l'!fd1 .ixg2 19. 'ii> xg2 ll:\d7oo.
12 ..id2 would mark a turn in
White's strategy. He does preserve
the Mar6czy Bind's pawn structure,
but his pieces need a serious re­
grouping. That gives us more than
enough time for completing com­
fortably development. Black's best
set-up is perhaps 12 . . . b6 13 .l'!ac1
.ib7 14.l'!fd1 (Firat-Delchev, Skop­
je 30.03.2014, saw 12 . . . ll:\d7 13.l'!ac1
b6 14.b3 .ib7 15 . .ib1?! l'!ac8 16.ll:\e1
l'!fe8 17.f4 f5+) 14 ... ll:\d7 15 . .ie3
(or 15.h3 l'!ac8 16 . .ie3 '<MI'b8 17 . .ib1
l'!fe8 18.'<MI'd2 .if8 19.b3 .ia8 2 0 .i.h6
i.e7 2l.i.g5 ll:\hf6 2 2 .i.f4 ll:\e5 = ,
Vavulin-Artemiev, Moscow 2 0 14)
15 . . . l'!ac8 16 . .ib1 l'!fe8 17.'<MI'd2 .if8
224
18.b3 �b8 19.i.h6 .ie7 2 0 . .ig5 .if8,
Kryvoruchko-Nikolov, Rethymnon
2011. It is good to preserve the dark­
squared bishop which protects the
d6-pawn and the kingside. At the
same time, it usually plays an im­
portant role in a counter-attack
with . . . d5 by hitting b4 or by going
to d6.
12 . . . hc3 ! 13.bxc3 e5!
We have obtained the same
pawn structure as in the previous
line, only the knight is on h5. This
"small" difference forces changes
in Black's plan since after 13 .. .f6,
the knight would have no prospects
whatsoever. The text opens the c8-
bishop which is assigned to kill the
white knight. It is better to eat it
on f3 rather on d5. See game 40
Grischuk-Ivanchuk, Beijing 2 013.
B23. 10 ..ie3 0-0
This line leads to the Hedge­
hog in its "nearly" best version for
White. Why is it less popular then
the previous two lines then? Kritz
even accompanies the move 10 . .ie3
 Kan - the Mar6czy Bind
with a question mark ! ! It does takes
guts to stigmatise in such a way the
white side of a hedgehog! In my
opinion, this is the most principled
and unpleasant weapon against the
Kan and I can only be too happy
that it is underrated. I suspect that
l.e4 players usually do not feel com­
fortable in these structures. They
commonly prefer more straightfor­
ward plans which involve a kingside
.
attack. In our case, Black's king is
well protected by the fianchet­
toed bishop so the direct assault
has little chances to succeed. After
ll.f4 l2Jbd7 12 J:!ac1 b6 13.'it>h1 �b7;
14.£5 would be a mistake, because it
opens the e-file to Black's favour. In
practice, White turns to the thema­
tic plan with b4, but then why had
he played f4? ! - 14.b4 E:fe8 15.l2Jb3
E:ac8 16.a4, when the best set-up is
16 . . . E:b8 ! 17.a5 bxa5 18.l2Jxa5 �a8 .
lU�acl �bd7! (not ll . . . b6)
12.f3
White may try to counter my
plan of an early . . . l2Jd7 -eS by 1 2 .E:fd1
llJeS 13.l2Jf3, but we get counterplay
with 13 . . . l2Jfg4 14.�g5 b6 15.l2Jxe5
ttJxeS 16.b3 �b7 17.�b1 fS .
Amazingly, but this position
seems utterly unexplored. I see
more than 30 games in my data­
base, but none of them can serve us
as a model to follow. Even the young
Karpov looks hesitant in his actions.
Look at the game Musil-Karpov,
Ljubljana 1975 : 12 . . . b6 13.V9f2 �b7
14.E:fd1 l2Jh5? ! (What is this? Did he
intend to push .. .fS?) 15.�fl E:ac8
16.E:c2 (16.b4 ! ) 16 . . . l:!fe8 (not at all ! )
17.E:cd2 (17.b4 ! ) 1 7. . . �f8 18.l2Jdb5? !
axbS 19.l2Jxb5 �b8 20 .l2Jxd6 l2Jc5?
2 1.l2Jxe8 and Black is in very bad
shape.
Shipov has written 1100 pages
on the Hedgehog. A great prose, but
no answers, unfortunately. Or even
worse - wrong answers. He men­
tions only two games (where White
did not even choose the best con­
tinuations) where Black is conspic­
uously worse.
The overwhelming majority of
players choose the set-up with . . . b6
and even score more than 50%, but
they clearly did not know what to do
against the plan with 12 . . . b6 13.E:fd1
�b7 14.�f2 E:ac8 15.�fl E:fd8 (or
15 . . . E:fe8) 16.b4 �b8 17.l2Jb3.
I analyse in detail this position
in the annotations to game 38
225
 Part 6
Simacek-Akopian, Plovdiv 2 0 1 2 .
M y conclusion i s that we should not
allow it at all !
I see only one possible way to ex­
ploit White's set-up with i.d3-�e2
- the bishop cannot retreat to fl.
Therefore, our first task should be
to play quickly . . . tt:lbd7-e5, aiming
to trade it for the bishop if White
answers �d2 . So we start with :
12 ... �e5! 13J:�fdl
Now we have a choice. I spent
some time on:
13 ... i.d7. It is a logical develop­
ment since the bishop would sup­
port . . . b5. In practice White invar­
iably answers :
a) 14.b3? ! when, besides 14 . . .
�xd3 , we can maintain the tension
with 14 . . . �a5 ! ? . Then 15.i.b1 �fc8
16.�d2 �e8 would allow us to push
. . . b5 (if White played a4, he would
never achieve b4 due to the weak­
ness of the c4-pawn). 15.�d2 tt:lxd3
16.\Wxd3 �fd8 is similar to line b, so
I analysed mainly 15.�b2 (threat­
ening i.f1 or b4) 15 . . . tt:lxd3 16.�xd3
�feB 17.�de2 (17.b4 �c7 18.tt:lde2
226
i.e8 19.i.d4 e5 2 0 ..ie3 �xc4; 17.a4
i.e8) 17 . . . b5 18.b4 �d8 19.�xd6
�xc4 2 0 .�cd1 �e8 21.e5 tt:l d5
2 2 .tt:lxd5 exd5 23.�6xd5 .ic6.
b ) 14.�d2 ! tt:lxd3 15.�xd3 �fd8 !
15 . . . �fc8? ! looks more active, but
the long forced variation 16.tt:lb3
.ie8 (16 . . . tt:le8 17.c5) 17.�xd6 �xd6
18.�xd6 tt:ld7 19.c5 ! i.f8 2 0 .c6 ! bxc6
2 1.�d2 c5 2 2 .tt:la4 c4 23.�dc2 tt:le5
24.tt:lb6 cxb3 25.�xc8 �xeS 26.�xc8
tt:ld3 27.�a1 tt:lb4 2 8 . axb3 �c2
29.�xa6 tt:lxe3 3 0 .tt:ld6 .id7 31.�a7
hd6 32.�xd7 faces Black with a
difficult ending - the .i + tt:� gener­
ally perform poorly against a �+
passed pawn.
Here 16.tt:lb3 is less effective
as 16 . . . i.e8 17.i.f4 �b6+ 18.@h1
�d7! ! 19 .�xd6 �xd6 2 0 . �xd6 tt:le5
2 1 . �xd8 �xd8 2 2 .i.xe5 he5 offers
Black full compensation for th e
pawn - 23.c5 f5= . Black plays on,
ignoring the enemy's extra pawn .
tt:l a5 is always met by . . . i.c6. I took
my time to explore this endgame
very deeply. The verdict is that only
Black can win it.
So far so good. I proved that
this set-up was possible. However,
 Kan - the Mar6czy Bind

returning back to move 16, my ini­ with .ifl, b4, lt'lb3.

tial enthusiasm began to evaporate. H aving shifted the bishop to b1
I could not set up decent counterplay gave us time for an active redeploy­
against best play: 16.b3 V9a5 17.lt'lde2 ment:
.ie8 18.i.f4 ! (to drag the bishop to
15 . . . !1:fc8 16.�f2 E:ab8 17.lt'la4
c6, 18 . .id2 V9c5+ 19.'k!?h1 bS; 18.'k!?h1
bS) 18 . . . .ic6 19 . .id2 !

Note that I propose to put the

rooks on b8 (protecting b6) and c8
Black has to abandon his dream
(hitting c4). We can already breathe
of . . . bS and consent to a waiting
freely as the enemy is left without
stand with 19 . . . !1:d7 2 0 .'k!?h1 ( 2 0 .
an active plan. In contrast, we have
lt'ld4 E:c8 21.lt'la4 V9d8 2 2 . lt'lxc6 E:xc6
at least two promising ideas based
23 . .ie3) 2 0 . . . �d8 .
on kingside play with . . . gS :
This analysis made me explore
a) 17 . . . lt'lfd7! 18.a3 (hoping for
the same structure with a bishop on
.ia2 , b4) 18 . . . lt'lc5 19.lt'lc3 V9e7! (re­
b7.
locating the queen to the kingside)
2 0 . .ic2 .if6 (20 . . . 'k!?h8 ! ? 2 1.�g3
13 b6
fS) 2 1 .h3 (21.'1!9e2 g5 2 2 .b4 lt'lcd7
. . .

2 3 ..ib3 E:c7=) 2 1 . . .g5 2 2 .�g3 'k!?h8.

Most practical games went from
this point 14.b3 [14 . .ib1 lt'lxc4 is bal­
anced - 15.lt'la4 (15 . .if2 .ib7) 15 . . .
b S 16.b3 e S 17.bxc4 bxa4 18.lt'lc2
.id7 19.�d2 E:fc8] 14 . . . .ib7 15 . .ib1 ? !
(15.�f2 could face 1 5 . . . d5 ! 16.cxd5
lt'lfg4) . We politely allowed the
bishop to live on and it ran away!
Naturally, we could have traded
on d3, but I think that Black has
more chances to win if he kept both
knights. My idea of playing . . . lt'ld7- b) 17 . . . .ia8 18.lt'le2 lt'led7 19 ..id3
e5 was not so much aimed against �d8 2 0 . lt'l ac3 lt'leS 2 1 .'1!9e1 V9e7
the bishop, but against the plan 2 2 . lt'ld4 lt'led7 23 . .ifl (23.b4 lt'leS)

227
 Part 6

lt:\e8 (24.b4 lt:\eS) 24.a3 .ieS 25.b4

%l/f6 26.%l/d2 gS 27.lt:\ce2 h6 2 8 . lt:\g3
%l/g6 29 . .id3 lt:\ef6 with an interest­
ing game.
However, we must also prepare
for White's best retort:

14.Wd2 lt:\xd3 15.Wxd3 lt:\d7!

t6.b3 gds t7.ed2 J.b7 t8.lt:lde2
J.f8 19.'kt>hl gac8 2 0 .ig5 ges
•

White is very solid indeed, but You know what to look for - . . . bS,
at least he has not a clear plan for or . . . dS. He might wait with . . . �b8,
improvement. Black should be pa­ . . . .ia8, . . . lt:\c5, . . J3:c7 - all that typi­
tient and wait for his chance. cal hedgehog jazz.

228
Pa rt 6. Ka n - the M a r6czy B i n d
Comp l ete Games
34. Ermenkov-Kotsur
Dubai 30.04.2000
This game was annotated back
in 2 0 0 0 for Chess Base Magazine
by our friend and established Chess
Stars translator, GM E. Ermenkov.
He has been a devoted Sicilian play­
er with both colours throughout his ·
entire career so it would be interest­
ing to compare his notes:
l.e4 c5 2 . ttlf3 e6 3.d4 cxd4
4. ttlxd4 a6 5 .id3 g6 6.c4 i.g7
• ll.exj5!? in his wonderful book
7.lt:�b3 ttle7 8.ttlc3 ttlbc6 9. 0-0
0-0 10 .ie2 •
'This move is the point ofWhite's
set-up. After the bishop's retreat,
White is preventing . . . d5 and in­
tends simply !i.f4 followed by Wff d2,
'il.fdl, 'il.acl, fi.h6, etc. with a strong
positional bind. ' (EE) lO .fi.gS is cer­
tainly more dangerous for Black.
10 • • • f5 (! - A. D.)
'I am very suspicious toward
this type of "Dutch Defense" in the
Sicilian. ' (EE) The break . . .fS is one
of the main ideas of the set-up with
. . . ttle7. Black is going to destroy the
enemy centre by taking on e4 and
pushing . . . dS. Therefore, White's
next move is cpnsistent and popu­
lar.
ll.c5
'John Nunn recommends here
"BTS-3 ". ' (EE) Well, after ll...ltJxfS
12 .fi.f4 bS ! nothing will remain from
the Maroczy Bind pawns. Philidor
would have been happy to see first­
ly . . . fS, then . . . bS ! White is walking
on thin ice - 13.cxb5 axbS 14 . .ixb5
ltJ cd4 - 15. ltJxd4 ltJxd4 16.fi.d6 ltJxb5
17.ltJxb5 'il.fS 18.a4 .ia6.
White can discourage . . . bS with
ll.a4, but ll . . . fxe4 12.ltJxe4 dS
solves the opening problems. (en­
gines are fond of the more flexible
12 . . . b6)
ll . .if4 fxe4 12 . .id6 e3 ! 13 .fxe3
'il.xfl+ 14 . .ixfl ltJfS is also pleasant
for Black.
Only the paradoxical and some­
what ugly move ll.f3 ! ? keeps con­
trol of the centre. Black should
229
choose the right moment for . . . f4.
He may play it at once and follow
up by his initial idea of . . . d7-d5.
ll fxe4
• •.
'll . .. b6 has been tried in prac­
tice, but White was a little better
after 12.cxb6 Wfxb6 13. i.e3 WfdB
14:f4. ' (EE)
14.Wffd 2 fxe4 15.tt:Jxe4 dS 16.tt:Jec5
tt:Jfs 17.i.g5 '?9d6 18J�ac1 h6 19.if4,
Psakhis-Oratovsky, Tel Aviv 1993,
would have been fine for Black af­
ter 19 . . . e5 ! , but Ermenkov's propo­
sition 14.f4? is simply bad due to
14 . . . fxe4 15.tt:Jxe4 tt:JdS. Therefore,
ll . . . b6 is a viable alternative to the
game.
12. tt:Jxe4 b6
�t first sight what Black is do­
ing seems very logical. He wants to
get rid of the pawns on e4 and c5
and then using the "d" and "e" pawn
tandem to prevail in the fight for
the centre. This plan, however is too
ambitious to succeed against the ex­
cellent development of White. ' (EE)
In fact, White has no advantage
neither in the centre, nor in devel­
opment. I would say, Black's pieces
are clearly more active which is bal­
anced by the hole on d6.
13.tL:ld6 !
This horse is coming here to stay
and graze till retirement. . . ' (EE)
13 ••• bxc5
230
14.c!lJxc5
14 . .te3 hb2 15J�b1 i.eS 16.hc5
Wffc7 17.g3 tt:Jds 18.tt:Je4 d6 19 .i.a3,
Li,C.-Grandelius, Malmo 2012,
does not look as an improvement.
After 19 . . . a5, White would have
been struggling to prove sufficient
compensation for the pawn.
14 ••• c!lJd4? !
'But not 14 . . . ttJf5 15. ttJxcB' (EE)
14 . . . tt:Jf5 ! assures Black of a fine
game. Black should not delay this
move as the pin from gS would
be awkward. Ermenkov's retort
15.tt:Jxc8 has no venom since 15 . . .
Wffx c8 16.Wffx d7? loses t o 16 . . . tt:Jcd4.
Black rules in the centre after
16.i.g5 d6 17.tt:Ja4 Wffb 7 18.!k1 tt:Jcd4
19 .b3 tt:Jxe2+ 2 0 .Wffx e2 tt:Jd4.
The possession of the d4-square
ensures full-fledged counterplay in
the other lines, too :
15.tt:Jxf5 l'!xfS 16 .ie3 l'!dS 17.'?9e1
tt:Jd4 18 .l'!cl aS 19.i.g4 a4;
15.ig4 tt:Jxd6 16.'?9xd6 tt:Je5 17.h3
tt:Jc4 18.Wffd 3 tt:JeS;
15.tt:Jce4 tt:Jcd4.
15.J.d3?!
 Kan - the Mar6czy Bind
It was essential to provoke weak­
nesses with lS . .!gS ! h6 16 . .!h4 gS
17 . .!g3 l2Jef5 18 .i.d3.
15 .'�c7 16.l2Jce4 J.e5 (why
••
not 16 . . . lLlf5 ! ) 17.J.e3 !
17 ••• l2Jdf5
17 . . . l2J ec6 ! was simpler: 18.lLlxc8
E:axc8 19.'it>hl dS 2 0 . lLlc5 �d6
21.lLlxa6 .ixh2 2 2 .�g4 .if4. Also
17 . . . lLlf3+ 18.gxf3 .ixd6 19.lLlxd6
�xd6 deserves attention.
18J�cl c!Llc6?
18 . . . �b8 was unclear. It seems
that Kotsur forgot his kingside.
19. lLlxf5 exf5
'19 . . . gx.f5 does not look attrac­
tive after 2 0 . ltJgS with the threat
� h5, check or no check with
. . . i.xh2+ while 19 . . . E:x.f5 2 0 . ltJg3
followed by 21. � g4 and Black will
not survive this attack. ' (EE)
After the text, Black's centre is
ruined and my heart is aching to
watch his poor king left to the mercy
of the opponent's pieces. Although
Ermenkov went for unnecessary
complications, he succeeded in fi­
nishing the game in style:
20.{4 .!g7 2 1 . lLld6 h6 2 2 .g4
'it>h8 23.J.bl a5 24JU'2 J.a6
25.gxf5 gf6 2 6 . c!Ll e4 gxf5
27.c!Llg3 gf6 2 8.�g4 ge6 29.f5?!
gxe3 30.f6 c!Lle5 3 1.fxg7+ 'it>xg7
3 2 . c!Llf5 +
32 'it>h8? (32 . . . 'it>h7 was more
•••
resilient) 33.�dl �b6 34.c!Llxe3
�xe3 35.gc3 �a7 36.�d6 c!Llg4
37.gc5 ! ges 38.gf8+ gxf8
39.�xf8 + 'it>h7 4o.�t7+ mhs
41.�e8+ mg7 42 .�xg6 + mf8
43.�d6+ 'it>t7 44.�c7 1-0
35. Eichhorn-Delchev
Bad Wiessee 30.10.2013
l.e4 c5 2 . c!Llf3 e6 3.d4 cxd4
4. c!Llxd4 a6 5 .id3 lLlf6 6. 0-0 �c7
•
7.c4? !
231
 Part 6

This is a frequent positional mis­

take. White should delay c4 until
. . . d6 as now his dark squares pass
under Black's control.

7 �c6 ! 8. �xc6 dxc6 9.�c3

• •.

e5 10.ti'c2 ! ? (threatening ltla4)

10 .ic5 1U��b1 a5 12.a3
•.•

White intends to push b4. Then 18.b4

the bishop cannot survive long on
d4 in view of ltle2 - 12 . . . 0-0 13 .b4
axb4 14.axb4 i.d4 15.ltle2 i.a7 16.c5
i.e6 17.h3 b6 18.i.e3. Therefore, it
is important to hinder b4, at least
for a while.
12 • •. ti'e7! 13.�a4 .id4 14 .ig5
•
My bishop is well established in
the centre, but now I must deal with
a new positional threat - 14 . . . h6
15.i.xf6 ti'xf6 16.c5 i.e6 17.ltlb6. To
be fair, the engines think that 17 . . .
:B:a7 (17 . . . :B:d8 =) would b e fine for
Black, but I doubt that an organic
creature would voluntarily isolate
his rook in such an exemplary way.
It seemed to me better to keep the
queenside fluid, but my idea has a
serious flaw.
White deludes himself that he is
the active side, but his advance only
produces a weakness. The fact that
my setup on the queenside is defen­
sive does not automatically mean
that White has an advantage there.
He could have stayed with i.e2,
:B:bdl.
18 axb4
..• 19.gxb4? ! ti'a7
20.ti'b2?
Who would believe that the
game will be over in two moves? !
2 0 .i.e2 i.e6 21.c5 would b e even -
2 1 . . .b5 2 2 .ltlb6 ltld7 23.ltlxd7 ti'xd7
24.ti'c3 ti'c7 25.a4 bxa4 2 6.:i'!xa4
:B:b1 + 27.i.fl. However, White per­
sists with his idea of attacking b6.
This leads him to a tragic end:

14 ••• b6 15.gfe1 20 ••• �d7 2 1 .ifl? c5 0-1

•

15.ti'd2 i.e6 16.i.e3 :B:d8 ! seizes

the initiative, but 15.c5 ! b5 16.ltlb6 36. Todorovic-Caruana
:B:b8 17.a4 ! b4 18.i.xf6 = would have Rijeka 06.03.201 0
effectively petrified the left wing.
1.e4 c5 2 . �f3 e 6 3.d4 cxd4
15 ••• 0-0 16 .ie3 .ixe3 17.gxe3
• 4. �xd4 a6 5 .id3 �f6 6.ti'e2
•

gbs d6 7.0-0 g6 8.c4 i.g7 9 . � c3 0-0

232
 Kan - the Mar6czy Bind

to.gdl %Yc7 1l.h3 c!Llc6 12.c!Llf3 tt:l aS 2 1 . tt:ld5 hdS 2 2 . cxd5 bS is

c!Lld7 13.J.f4 also possible) 2 0 .i.cl ( 2 0 . .ih6 l'!fc8
2 1 .tt:le3 tt:ld8 2 2 . tt:ld5 hdS 23.cxd5
gS 24.h4 tt:lfl+) 20 . . . tt:la5 2 l.tt:le3 bS
2 2 . tt:l d5 hdS 23.cxd5 tt:lc4.
16 . . . tt:lc5 17 . .ic2 'fff e 7 18.tt:ld2

13 ••• .bc3 !

This exchange is perhaps the

main strategic aim of Black's set­
up. It is possible in various settings,
Black has some initiative, Karja­
for instance, with a white bishop on
kin-Ivanchuk, Amber-blindfold Nice
e3, or a black knight on h5. The aris­
2008.
ing pawn structure is so static, that
White practically remains without
16.c!Llh2 c!Llc5 17.i.c2 f6
any active plan. He will push f2-f4-
f5 and that is about all. Black, for his
part, will seek to destroy the other
enemy knight since it is a dangerous
piece around his king. See also the
next game.

14.bxc3 e5 15 .ih6 ges

•

The idea of this strange move is

to keep the dB-square for the ma­
noeuvre . . . tt:lc6-d8-t7.
Ivanchuk chose in an earlier
game the more trivial 15 .. J'!d8 when
Karjakin answered 16.'\We3.
The manoeuvre tt:lf3-h2-fl-e3-
d5 is fairly slow and Black gets
counterplay with . . . bS - 16.tt:lh2 tt:lcS
17 . .ig5 l'!f8 18 . .ic2 (18.tt:lg4 hg4)
18 . . . .ie6 19.tt:lfl f6 (it is safer to plug
the hole on f6, but 19 . . . l'!ac8 2 0 .tt:le3
ts.ef3
18.tt:lg4 hg4 ! is absolutely safe
for Black. White might try to fix a
little his pawn structure by trading
the knight on dS after 18.tt:lfl tt:ld8 !
19.tt:le3 i.e6 2 0 .tt:ld5? ! , but sudden­
ly the h6-bishop finds itself in a pre­
dicament - 20 . . . hd5 2 1 . cxd5 gS.
White falls in the same trap in the
game :

233
 Part 6

18 ••• ti'e7 19.�fl J.e6 20.�e3 when White's only sensible idea is
�d8 ! to sacrifice a pawn with 16.l2Jd4 �e8
17.c5 ! l2Jxc5 18.l2Jb3
The threat . . . g5 is extremely an­
noying. Todorovic braces himself
for a desperate sacrifice which has
no chances to succeed.

21.�£5? gxf5 22.exf5 .i.d7

23.gxd6 �h8 24.gadl J.c6
25. ti'h5 �f7 26.gxc6 bxc6 27 .i.e3
•

gf8 28.fih4 gadS 29.hc5 �dl +

30.hdl fixeS 31.ti'xf6+ �g8
32 .ih5 �d6 33.fig5+ �h8 0-1
• White has full compensation for
the pawn and unpleasant initiative.
Perhaps Black should simplify to an
37. Leko-lvanchuk ending with 18 . . . b6 19.l2Jxa5 bxa5
Mukachevo g/10'+1 0" 2007 2 0 . �ab1 .ib7 2 1.�c4 �e7 2 2 .�d4 e5
23 .fixd6 �xd6 24.�xd6 .ixe4.
l.e4 c5 2 . �f3 e6 3.d4 cxd4 This example explains why in
4.�xd4 a6 5 .id3 �f6 6.0-0 fic7
• practice Black met 15 . .ic2 by 15 . . .
7.fie2 d6 8.c4 g6 9.�c3 .ig7 l2Jc5 16.l2Jh2 (at first the engines of­
lO.gdl 0-0 ll.�f3 �c6 12 .h3 fer 16 . .ixc5 dxc5 17.e5, but White
�d7 13.J.e3 .ixc3 14.bxc3 f6 has not any advantage after 17 . . .
l2Jxe5 18.l2Jxe5 �xeS 19.�xe5 fxe5
2 0 J:!e1 �dB 2 1 . �ad1 .id7 2 2 . �xe5
.ic6= ) when instead of 16 . . . e5
17 . .ixc5 dxc5 18.l2JfU, I prefer 16 . . .
l2Ja5 17.l2Jg4 e5.

15.�d2 gds (15 ... l2Jd8 retains

more tension) 16 .ic2 �c5
•

White has an extra tempo in

comparison to the previous game
since . . . e5 is undoubtedly a useful
move. For instance, White's knight
can reach g4 as in the line 15 . .ic2
l2Jd8 16.l2Jh2 l2Jf7 17.l2Jg4 with some
initiative. Black's knight could hit
the c4-pawn instead - 15 . . . l2Ja5

234
 Kan - the Mar6czy Bind
First critical position. 17.hc5
dxcS 18.�e3 eS 19.�xc5 i.e6
20 J':1abl assures White of a tiny
edge although his doubled extra
pawn has little practical value. For
instance : 2 0 . . J'!xdl
Grimberg-Kveinys, Bad Woer­
ishofen 2 0 1 1 , went 20 .. J'!ac8 2 U : 1b6
with a quick draw. 2 1.�b6 is more
testing. After 2 1 . . .�xb6 2 2 J'!xb6
l'!xdl+ 23 .hdl lt:ld8 24.lt:ld2 'i!if7,
Black's knight is passive on d8.
2 1.hdl l'!d8 2 2 .i.e2 �as
23.�xa5 lt:lxaS 24.<±>fl l'!d6 2S.l'!b4
b6, planning . . . lt:lb7.
Leko's move is more ambitious,
but it allows Black to consolidate.
White loses his main trump - the
initiative.
17. �d4 �aS 18 .We2 eS 19.�f3
.ie6 20. �d2 Wt7 21.f4
The second critical position.
Black has already a clear advan­
tage. He should take on f4 in order
to make another target on e4, and
then on c4. However, we cannot be
exigent to a rapid game.
21 . . . l'!ac8 2 2 .l'!abl �xc4
23.f5 gxf5 24.exf5 .idS 2S.�xc4
hc4 2 6.Wg4+ 'i!ih8 27.l'!b6 .idS
28 .Wh4 .ic6 29.l'!b4
The weakness on f6 gives White
enough conterplay and the game
was eventually drawn.
29 . . . l'!g8 30.l'!g4 We7 31.Wh6
dS 3 2 . c;!,>h2 �d7 33 . .ib3
33 . . . Wt7 34 .if2 l'!xg4 3S.hxg4
•
l'!g8 36. 'i!ih3 l'!g7 37 .ih4 c;!,>g8
•
38.l'!d3 'i!if8 39.l'!g3 'i!ie8 40.gS
�cS 41 .idl .id7 42 . .ihS .ixf5+
•
43 .'i!ih2 l'!g6 44 .ixg6 Wxg6 •
4S.Wxg6 + hxg6 46.gxf6 c;!,>t7
47.l'!e3 �e4 48. 'i!igl bS 49.c;!,>fl
gS SO.l'!xe4 .ixe4 Sl.hgS .ig6
S2. c;!,>e2 c;!,>e6 S3.c;!,>e3 c;!,>f5 S4 . .ih4
c;!,>g4 SS .if2 c;!,>gs S6.c;!,>d2 'i!fxf6
•
S7 .icS .ibl S8.a3 c;!,>f5 S9.'i!ie3
•
.ie4 60.g3 .ig2 61 ..ib4 .ih3
62 .icS .ig4 63 .ib4 c;!,>e6 64.c;!,>d2
• •
c;!,>d7 6S .iaS c;!,>c6 66 .ib4 d4
• •
67.cxd4 exd4 68 .iaS c;!,>cs •
69. c;!,>c2 c;!,>c4 70 .'i!ib2 c;!,>d3
71 .ib4 c;!,>e2 72 .'i!icl d3 73 .ic3
• •
.ie6 74 . .ib4 c;!,>f3 7S. c;!,>d2 .if5
76 .iaS c;!,>xg3 77 . .ib4 'i!if3 78 .iaS
• •
c;!,>e4 79 . c;!,>ct c;!,>e3 80 .id2 + c;!,>d4
•
8t.c;!,>b2 c;!,>cs 82 . .iaS c;!,>c4 83 .id2 •
.ig6 84 .iaS Draw
•
235
 Part 6

38. Simacek-Akopian
can simply ignore the gift with
Plovdiv 22.03.2012
18.exd5 exd5 19.c5 !

This game is an illustration of

what position we should avoid at
any cost in the Hedgehog.

l.e4 c5 2.�f3 e6 3.d4 cxd4

4.�xd4 a6 5.J.d3 �f6 6.0-0 Y;Yc7
7.'e;Ye2 d6 8.c4 g6 9 . � c3 J.g7
10.J.e3 0-0 lUUdl b6 (I recom­
mend ll . . ttlbd7-e5xd3 ! ) 12.l:'!acl
.

J.b7 13.f3 �bd7 14.ti'f2 Perhaps this is White's best way

White has achieved the most
unpleasant redeployment of his
pieces. They are all directed to the
queenside to support b2-b4 and
then c4-c5 or a2-a4-a5. Since b6 is
the sore point of Black's set-up, it
makes sense to defend it with 14 . . .
:!! ab8 15.J.fl J.a8 16.b4 l'l:fe8 17.ttlb3
J.f8. Akopian prefers to keep b8 for
his queen.
14 ••• l:'!ac8 15.J.fl l:'!fd8
of meeting . . . d5 even when grab­
bing the pawn is possible. It is bet­
ter to have a positional rather than
a material advantage. After the text,
19 . . . bxc5 2 0 .ttlxc5 ttlxc5 2 1.bxc5 J.f8
22 .ttla4 Y;Yc7 23.J.d4;!; would open the
b-file in White's favour. The pawn on
a6 and the b6-square are weak.
It is more consistent to keep the
flank closed and hope to use the c4-
square : 19 . . . b5 2 0 .ttl a5 ttle5 2 1.J.d4,
but White is clearly on top in view of
the threat a3-a4. 2 1 . . . ttlc4? simply
drops a pawn after 2 2 .ttlxb7 ti'xb7
23.J.xf6 J.xf6 24.ttlxd5 + - .
Instead o f 17 . . . d 5 , Eingom
played against E. Geller 17 . . . J.f8, but
White could safely take on b6.
16.b4 Wb8 17. �b3 .ia8

Black's main active plan is . . . d5

so putting the king's rook on a po­
tentially open file is more consistent
- 15 . . . 1'!fe8. However, after 16.b4
f;Yb8 17.ttlb3, the break-through in
the centre 17 . . . d5 does not solve the
opening problems because White
236
 Kan - the Mar6czy Bind

I feel that White is clearly bet­ '!Mfb8 27.a3, nor 19 . . . �e8 2 0 .b5 axbS
ter. He has not a direct break­ 21.cxb5 �cc8 2 2 .lt:Jd4 dS 23.lt:lc6
through since 18.c5 dxcS 19.ha6 hc6 24.bxc6 �xc6 25.i.b5 are rosy
i.b7 2 0.hb7 '1Mfxb7 2 1.bxc5 bxcS for Black.
22.lt:l xc5 lt:JxcS 23J'!xd8+ �xd8 Instead, White tries to repeat
24.hc5 stumbles into 24 . . . lt:Jg4 ! = , moves ! That suddenly unlocks the
but h e can patiently wait for the best dozing power of the black pieces.
timing, for instance, with 18.a3 i.f8
19.lt:Ja4 �c6 2 0 .�bl ! - thus he pre­ 19.lt:ld4? �c7 2 0 .tt:J b3 b5 ! ?
vents counterplay with . . . b5. Then
he can double the rooks on the d­
file or win the b6-pawn by b4-b5.
Note that White has another
thematic plan in his disposal - to
produce a passed pawn with 18.a4
i.c6 (Johnson-West, email Argenti­
na 1997, saw 18 . . . d5 ! ? when 19.exd5
exdS 2 0 .c5 ! is again the best retort)
19.\Wa2 lt:JeS 2 0 . a5 bxa5 21.\WxaS,
White has n o longer i.a7 s o this
followed by b4-b5, but it would be
break is already possible. In the en­
more effective when Black's pieces
suing hand-to-hand combat Aka­
went to passive places, for instance,
pian outplays his opponent al­
i.f8 .
though White should still be better.
For instance, 21.cxb5 ! axbS 2 2 .lt:lb6
18.tt:la4 l!k6 (18 . . . b5? 19 .i.a7)
�xcl 23.�xcl dS 24.lt:lxd7 lt:lxd7
25.exd5 hd5 26.lt:ld4 looks promi­
A critical position. The b6-pawn
sing. Then 26 . . . \Wa8 leads to a grim
is doomed in view of the threat b4-
ending after 27.lt:lxb5 (27.�c2 i.c4 ! )
b5. The only question is when to
2 7 . . . '\Mfxa2 28.\Wxa2 ha2 29.lt:la7,
take it. The immediate 19 .b5 axbS
but 26 . . . i.e5 ! 27.f4 i.d6 28 .a3 lt:Jf6
20 . cxb5 is not accurate since 20 . . .
keeps Black in the game.
�xcl 2 1 . �xcl d S offers considerable
counterplay. Instead, White can im­
21.ltlb6 bxc4 22.fucc4 fucc4
prove his position with 19.a3 �dc8
23.hc4 ltle5? !
20 .�b l !
More straightforward i s 19.�bl 23 . . . lt:Jxb6 or 23 . . . d5 24.exd5
when neither 19 . . . b5 2 0 . cxb5 axbS lt:JxdS 25.lt:Jxd5 hd5 2 6.hd5
21.i.a7 \Wc7 ( 2 1 . . .�c2 2 2 .hb8 lt:Jf6 = are stronger. After the text,
�xf2 23 .hd6 �xfl+ 24.@xfl bxa4 24.ha6 ! was possible: 24 . . . he4
25.lt:Ja5±) 2 2 .hb5 �c2 23.�bcl �xcl (24 . . . d5 25.lt:Jxa8 '!Mfxa8 26.b5 dxe4
24.�xcl '!Mfb7 25.i.b6 lt:Jxb6 2 6.\Wxb6 27.�xd8+ '1Mfxd8 2 8 .b6 exf3 29.gxf3

237
 Part 6
'\1;Yd5 30.@g2 tt'lfg4 31.b7 tt'ld7oo)
25 . ./tJaS tt'ldS 26 . ./tJxdS .bdS 27.b5
'\1;Ya8 28.�b6 ge8 29.a4 tt'ld7? .
24 .te2 d5 25. tt'lxa8 �xa8
• 11
26.exd5 tt'lxd5 27.J.c5 tt'ld7
The game is equal (28 .�d4), but
White blunders and he can resign
immediately:
28.a3?? tt'lc3 29.gd3 tL!xc5
30.tt'lxc5 i.d4 31.gxd4 tt'lxe 2 +
32 .fbe2 gxd4 33. c!Dxe6 gd6
34 .!Dc5 eds 35.ee4 gd4 36.tt'lb7
•
gdt+ 37.@f2 egs 38.h4 er6
39 . .!Dc5 @g7 40.c!Dd3 gal 41.h5
gxa3 42 .!De5 gxh5 43.c!Dd7 ga2 +
•
44.@gl �b2 0-1
39. Y.Gruenfeld-Psakhis
Tel Aviv 1991
l.e4 c5 2 . .!Df3 e6 3.d4 cxd4
4.c!Dxd4 a6 5 .td3 tL!f6 6.0-0 �c7
• tt'lxcS 18 . ./tJxcS bxcS 19 . .ba6 .ba6
7.�e2 d6 8.c4 g6 9.tLlc3 i.g7
10 . .te3 c!Dbd7 ll.f4
White assumes the obligation
to play all over the board. It is clear
that this move does not fit in well
with the queenside plan b4, tt'lb3.
Moreover, the hanging e4-pawn
demands the bishop to remain on
d3 so White has not the usual pres­
sure on d6. What does White gain
in return? Frankly, I'm not too sure.
Perhaps the advance f4-f5, but it
should be supported by all White's
heavy pieces to be a serious threat.
238
As a whole, White often turns out
to be overextended and his practi­
cal results are deplorable.
.•• 0-0 12.gacl
12.gae1 b6 13.@h1 i.b7 14.i.f2
was countered by 14 . . . e5 ! 15.fxe5
tt'lxeS, Salles-M .Nikolov, Marseille
2013.
12 b6 13. @hi
•.. .tb7 14.b4
gfe8 15.tL!b3 gac8
Both sides has made the easy
moves. It is time for White to de­
fine his plans. 16.c5 dxcS 17.bxc5
2 0 .�xa6 c4+ should be rejected by
obvious reasons. 16 .i.g1 only weak­
ens the f4-square. Black can exploit
it with 16 . . . ./tJhS, threatening .. .fS,
. . . dS or . . . �hd8-h4. Even . . . i.xc3
followed up by . . . eS is an interesting
option. Remains the move I men­
tioned in the previous game :
16.a4 ebs 17.a5 bxa5
18. c!Dxa5 .ta8 19.gbl
White has made some progress,
but his further offensive is stalled.
Black can trade a pair of knights
 Kan - the Mar6czy Bind
with 19 . . . lt:lg4, but Psakhis opts for
a more aggressive approach.
19 . . .li�h5 20. 'ed2 'ec7 2 U�bdl
This natural move proves to be
dubious due to the tactical possibil­
ity 2 1 . . .lt:lc5 ! ? 2 2 .�c2 hc3 23.'exc3
lt:lxe4. However, Psakhis prefers to
maintain the tension.
21 . . . gbs 2 2 .tl:l a4
22 . . • f5
After 2 2 . . . lt:lhf6 ! , White can­
not retain his front line intact. He
must push 23.c5 dxcS 24.hc5 lt:lxc5
25.lt:lxc5 \Wa7 where Black's bishop
pair has more potential than the
knights on aS and cS. 22 .. .f5 is also
consistent, though.
23.c5
23.exf5 exfS 24.c5 dxc5 25.lt:lxc5
(25.bxc5 :gxe3 2 6.�xe3 \Wxa5
27.\We6+ \t>h8 2 8.\Wxd? :gd8) 25 . . .
lt:lxc5 2 6.hc5 \t>h8 27.ha6 :gbd8
28.�f2 :gxd1 29.:gxdl lt:lxf4 would be
roughly equal, but White was afraid
to open the e-file and the diagonal
to the aS-bishop.
23 dxc5
• . • 24.ll:lxc5 c!Llxc5
25 . .h:c5 fxe4 2 6.J.e2 gbd8
27.tlfe3 gxdl 28 . .h:dl ttlf6+
29.J.a4 c!Lld5 30. 'eel gcs 31. 'exe4
'ef7 32.J.b3 J.c3 33.J.d4 .h:d4
34. \Wxd4 c!Llxf4 0-1
40. Grischuk-lvanchuk
Beijing 05.07.2013
l.e4 c5 2 . c!Llf3 e6 3.d4 cxd4
4.c!Llxd4 a6 5 . .id3 ttlf6 6.0-0 'ec7
7.'ee2 d6 8.c4 g6 9.ttlc3 .ig7
10.c!Llf3 0-0 ll . .if4 c!Llh5 12 . .ie3
.h:c3 13.bxc3· e5 14.gfdl
Compared to the previous game,
this position is easier for Black. He
cannot attack immediately the c4-
pawn with . . . lt:lc6-a5, but in return
he is not lagging behind in develop­
ment. The game now enters a ma­
noeuvring stage where Ivanchuk's
primary goal is to blockade firmly
the c4-pawn and prevent any white
activity on the kingside. That's why
he puts his queen's rook on d8 rath­
er than activate it from cB.
14 ... .ig4 15.h3 .h:f3 16.�xf3
c!Lld7 17.J.fl gadS 18.J.g5 f6
19.Ah6 c!Llg7 2 0 . gd2 c!Llc5 21.gadl
239
 Part 6

'9c6 22.gds gf7 23.'9e3 �ge6 ates the king in the centre in ad­
24.g3 gfd7 25 .ig2 • vance. However, his king does not
find a safe haven on e7. 26 . . . :1!t7 was
more natural.

27.h4 �e7 28.c5

2 8.g4 �xc4 29 .g5 is a better way

of seeking counterplay.

28 . . . �axc5 29.h5 '9a4 (29 . . .

lt:J a4 ! ) 30 .ie3 b 6 31.'9g4 ggs
• ·

White is on the defensive. He 32 .hxg6 hxg6

must constantly think of the weak
pawns on e4 and c4. For instance,
2S . .ie2? would have lost materi­
al after 25 . . . �c7. The text, howev­
er, leaves his queenside even more
vulnerable. Black should find a way
of redeploying his knights from a
blockading to attacking set-up. That
is a delicate matter since White will
also obtain chances with f4. Still,
staying nicely will not win the game
so he should seriously consider the White is not without threats.
manoeuvre 25 . . . lt:Jc7 26.:1!5d2 tt'la8 ! ? Now 33.f4 would have added oil into
targeting the c4-pawn. The text is the fire. Instead, Grischuk makes a
also consistent. couple of non-committal mundane
moves which tip the scales back in
25 . . . �a4 26.'9f3 �f7? ! Black's favour. The ending of the
game looks heavy influenced by a
The game has entered a more mutual time trouble.
concrete stage. The c4-pawn looks
doomed so the question is what aa ..if3 gdds 34. �g2 ghs
White will be able to do on the (34 . . . g5 ! ) 35.'9xg6 gdg8 36.�f5
kingside. Thus I can understand �g7 37.�g6 �e8 38 .�f5 �g7
Ivanchuk's reasoning - he evacu- 39.'9g6 �e8 40.�f5 Draw.

240
 Pa rt 7

Taimanov with 5.� b5

This part considers two different systems. Line A deals with SveshnikovI
Kalashnikov structures where White tries to build on his full control of the
d5-square after l.e4 c5 2.lt:lf3 e6 3 .d4 cd4 4.lt:ld4 lt:lc6 5.lt:lb5 d6 6 .i.f4 e5.
Line B considers the Taimanov version of the Mar6czy Bind: 6.c4 lt:lf6
7.lt:l lc3.

241
Part 7. Tai manov with 5 .'2J b5
Main Ideas
l.e4 c5 2 .!0£3 e6 3.d4 cd4 4.c!Od4
• C. lonescu-D. Heinbuch
c!Oc6 5.c!Ob5 d6 6.c4 c!Of6
In this system, Black gets a hedge­
hog of a highly improved breed. In
order to achieve the desired pawn
structure, White's knight under­
takes a time-wasting walk along the
route d4-b5-a3. These tempi are
just enough for Black to complete
development and counter-attack in
the centre before the opponent con­
solidated.
There is a widely accepted opini­
on that in this type of positions
time is not too important. We can
often see games where both sides
are aimlessly manoeuvring for 2 0 +
moves without committing t o any
concrete action.
My first advice to Black is:
Do not wait passively!
242
Berlin 1 988
Beware this setup! The hedgehog
structure is not bullet-proof. White
has a clear plan for a queenside
pawn storm.
In the diagram position, White
has a very unpleasant idea. Beside
preparing c4-c5 or a4-a5, he could
capture on eS and base his play on
the clumsy bishop pair.
19.he5 ! dxeS 2 0.c5 ! l'!fd8 2l.�el
l'!xdl 22 .l'!xdl l'!d8 23.cxb6 hb6
Here instead of 24.ltla4 l'!xdl
25.�xdl he4 ! with counterplay,
White should have exchanged the
rooks by 24.l'!xd8, followed by
25.ltla4±.
This design works only when
Black's dark-squared bishop has
left the a3-f8 diagonal. It is espe-
 Taimanov with S.lt::J bS
dally efficient if White's knight is
on b3, but as we'll see later, from
a3 it also could be activated via c4.
In the latter case, however, White
needs his queen on the gl-a7 diag­
onal and additional positional ad­
vantages like more active rooks.
I could add more practical ev­
idence on this theme, but I hope
you have grasped the essential -
White has a dangerous plan against
Black's entrenchment. Th erefore,
we should hinder it.
It was supposed that Black had a
universal set-up against this version
of the hedgehog. In many games the
second players half-automatical­
ly line up their forces by the follow­
ing scheme, regardless of the oppo­
nents actions:
7.lt::J lc3 a6 8 .lt::J a3 b6 9 .i.e2 i.b7
1 0 . 0-0 lt::J b 8 ll.f3 lt::J b d7 12 .i.e3 i.e7
13 .�d2 �c7 14Ji:fdl 0-0 lS.�acl
�ac8 16.i.fl
16 . . . V9b8 17.V9f2 i.d8 (In his book
about the hedgehog Shipov calls
this move "The Saemisch manoeu­
vre") 18.<;!;>hl �e8 19.�c2 !J.c7 2 0.�gl
<;!;>h8 2 1. �cd2 �g8 ? ! (see game 46
Daly-Tregubov, Cappelle la Grande
2000)
M y second advice is: Forget
about this plan !
There are many reasons for this
verdict. The most important one
is that I consider the whole idea of
putting the bishop to c7 dubious. It
only facilitates c4-c5.
I'd like to cheer up Black players
right away - they have other good
plans!
The slogan of Black's cam­
paign should be: . . . d6-d5 !
White is practically unable to
prevent this break-trough. Even if
Black forgot his lesson and failed to
achieve it in the opening, he should
be manoeuvring with this break in
mind. Sometimes it could be com­
bined also with . . . b6-b5, but this
queenside activity is usually less
efficient and often only provides
White with a juicy target.
In some positions . . . d6-d5 leads
to a mass elimination and proba­
ble draw. In "Step by Step" I sug­
gest, whenever possible, more com­
plex alternatives, at your own risk
of course. In the "Complete Games"
you can see examples which could
243
 Part 7
give you some insights about the
possibilities of both sides.
Sometimes . . . f7-f5 deserves at­
tention against passive White's
play.
Analysis
l.. J!f8 ! ? 2 .Wffl (protecting c4
and hoping for ll::l c 2-d4) 2 . . .f5 ! .
Analysis
l.. .<�h8 ! meeting 2 .f3 or 2 .b3 by
2 . . .f5.
Let's go to the variations now.
For more details check "Step by
Step" chapter.
In the event of i ll::l 5 c3, Black
refrains from . . . a6 and follows
244
the scheme . . . b6, . . . .ib7, .. J'k8 (or
. . . �d7, �fd8).
7 a6 8 . c!Ll a3 b6 9 .ie2 .ib7
••• •
10.0-0 (10 .f4 dS ! ) 10 .ie7 11 .ie3
••• •
(ll.f4 0-0) 11 0-0
•••
This is the system's "tabia".
White should decide how to con­
trol the centre. f3 is solid while f4
suggests a more risky approach
and possibly (but not obligatory!) a
pawn storm on the kingside.
12.fib3 is a modification of
the f3-plan. The queen move is
aimed at restricting the opponent's
choice. White hinders . . . dS indeed,
but Black obtains good play on the
dark squares. 12 . . . ll::l d 7 13.�fdl lLlc5
14.�c2 .if6 lS.�acl �e7!
The queen moves away from the
X-ray of the dl-rook. . . . d6-d5 is al-
 Taimanov with S.ltJbS

ready on the agenda: 16.l2Jabl l2Jb4 !

17.�d2 1'%fd8, threatening d6-d5.
Pay attention to the modern
13.1'%adl. It demands a different ap­
proach - 13 . . . l2J c5 14.�c2 \Wc7!
15.l2Jabl l2Jb4 16.�d2 1'%ad8 !

Now 13.1llli3 is the only move to

hinder . . . d6-d5. We should then aim
for the following set-up: 13 l2Jd7
•••

14.1'%adl 1'%c8 15.1'%d2 �c7 16.1'%cl

l2J ce5

Black's plan is . . . dS or . . . fS : 17.b3

dS ! 18.exd5 exdS 19.i.f4 \WeB ; 17.f3
dS ! .

After 12.f4 1'%c8 13.1'%cl 1'%e8

14 .tf3 l2Ja5, we reach a critical po­
•

sition for this line.

The game enters a manoeuvring

stage where Black's task would be
to keep the a3-knight pinned to
protect the c4-pawn. A possible
counterplay is . . .1'%f8 followed up by
. . . f7-f5.

After 12.1'%cl, we set up our

White aimed all his forces at pre­ rooks to c8 and e8 and push . . . dS:
venting . .. dS, but left c4 under-pro­ 12 l2Je5 13.�d4
•••

tected. Ironically, 15.�d3 dS ! is still

possible while 15.'1We2 \Wc7! 16.b4 13.f3 1'%e8 14.\Wd4 l2Jed7 15.l2Jc2
l2Jxc4 is another sharp option. (15.1'%fdl dS ! ) lS ... dS !

13 l2Jed7 14.1'%fdl 1'%e8 15.l2Jc2

•••

12.f3 leaves the e3-bishop hang­ 1'%c8 16.f3 d5! ? 17.cxd5 exd5
ing so 12 1'%e8 ! is a natural retort.
••• 18.exd5 .td6 !

245
 Part 7
Black plays all over the board.
His main threats are . . . !!eS or . . . bS.
These lines should be enough to
arm you sufficiently against White's
attempt of imposing slow manoeu­
vring play in a Mar6czy Bind struc­
ture.
Sveshnikov structures
6.i.f4 e5 7 .te3 a6 ! ? 8.til5c3 tilf6
•
is a different story.
White has provoked . . . eS, but
lost tempi. This may prove to be .
very dangerous for him in the line :
9.i.g5 i.e7 10 .hf6 i.xf6
•
ll.tild2
We sacrifice a pawn and seize
the initiative with:
246
11 0-0 ! 12. tilc4 i.e6 13.Vbd6
• • •
Vfe8 ! .
9.i.c4 i s perhaps the most an­
noying line - for both sides ! White's
positional aim is to exchange 2 or
3 minor pieces through ds· and ob­
tain a lasting edge due to the bet­
ter dark-squared bishop and more
active rooks. Even without any mi­
nor piece the pressure on d6 usually
ensures an advantage. In this sce­
nario Black's situation would be ag­
gravated if his b-pawn advanced to
bS since White could create a sec­
ond weakness on the queenside by
pushing a4. The best solution is to
develop quickly:
9 ••• .ie7 10.tild5 tild5 ll.J.d5
0-0
Now Black threatens to trade his
bad bishop with . . . i.gS.
 Taimanov with S.liJbS

12 .'i'd2 J.e6 13. tilc3 �k8 You should memorise the fol-
14.0-0 'i'c7! lowing move order:

9 J.e6 10. �c4 b5 ll.�b6

•••

gbs 1 2 .�bd5 �g4 13.J.e2 �xe3

14. �xe3 J.e7 15. � cd5 J.g5 16.0-0
0-0 17.c3

It is essential to avoid . . . b7-b5

or . . .f7-f5. Instead, we play . . . llJ aS
and patiently manoeuvre on the
queenside.

9 .llJd2 ! ? is White's only fresh This position commonly aris­

idea in the last years. es from the Sveshnikov. It is rare­
ly seen as Black has a decent game.
He may play for equalisation
with 17 . . . llJe7 18.'i'd3 he3 19.�xe3
Wic7 (Rogozenko), or seek compli­
cations with .. .fS - 17 . . . g6 18.a4 fS
19.axb5 axbS 2 0 J:!a6 W/c8 2 U:!b6
1:!xb6 2 2 . llJxb6 W/b8 23.exf5 gxfS
24.llJbd5, Shukh-Frolyanov, Nabe­
reznye Chelny 2010, 24 . . . Wib7!oo.

247
Pa rt 7. Ta imanov with S . ttJ bS

Step by Step

l.e4 c5 2.c!Df3 e6 3.d4 cxd4 14.@xd2 g6+) 13 . . . \Wc1 14J3c1 i.e7

4.c!Dxd4 c!Dc6 5 .!Db5 d6
• end the endgame is in Black's fa­
vour: 15.ll'ld5 ll'ldS 16.ed5 g6 17.c3
ll'lfS+.
10 . . . i.e7 ll.b4
ll.c3 i.e6 12 .\Wa4 @d7 13 .h4 \Wg6
14.\WbS (14.h5 �g5 15.�b5 \t>c8+)
14 . . . @c8+ is not appealing either.
11.. .i.g4 12 .f3 i.e6 13 .b5 ll'ld4
14.b6 \We3 15.i.e2 ll'lh5- + .

7 a 6 8 . c!D 5c3 c!Df6

• •.

A. 6.i.f4 ; B. 6.c4

A. 6 .if4 e5 7 .ie3
• •

7.i.g5? \Wxg5 8.ll'lc7+ @dS

9.ll'lxa8 counts on the surprise ef­
fect. If Black decides to win the
knight by b6, White gets some play
by taking on b6 and attacking that
pawn with the queen. Perhaps Black's
best approach is to win the knight by
his king, but firstly to develop with
Al. 9.i.c4 ; A2 . 9.i.g5 ; A3. 9.ll'ld2
9 ... ll'lf6 ! ? . The game Grafl-Jansa,
Fuerth 2 0 0 2 , continued :
9.ll'la3 ? ! is well met by 9 ... 1'!b8 !
10.ll'ld2 10.ll'ld5 ll'lxd5 ll.exd5 ll'le7 12 .c3
tt'lf5 = .
10 .ll'lc3 allows 10 ... ll'ld4 ! ll.h4
(ll.a4 d5 ! --+ ; ll.h3 d5 ! --+ ) · n . . . \Wf4 9.i.e2 probably aims t o trans­
12 .i.d3 i.g4 13 .\Wcl (13 .\Wd2 \Wxd2+ pose to the Sveshnikov with 7.ll'ld5

248
 Taimanov with S.lt'lbS
after 9 . . . .ie7 lO.lt'ldS lt'lxdS ll.exdS
lt'lb8 1 2 . 0-0 0-0 13.c4 lt'ld7 14.lt'lc3
f5 where Black is actually fine. We
can also choose the independent
line 9 . . . .ie6 ! ? 1 0 . 0-0 (lO .lt'ldS i.xdS
ll.exdS \1;ll a 5+ 1 2 .lt'lc3 lt'le7 13.0-0
lt'lexdS 14.lt'lxd5 �xdS 15.�xd5
lt'lxd5 16 . .if3 lt'lxe3 17.fxe3 �b8) 10 . . .
�c8 ll . .if3 (ll.lt'ld2 lt'le7 ! ) ll . . . lt'le7
12 .�e1 (12.lt'ld2 dS) 12 ... lt'lg6 13.lt'ld2
.ie7 with mutual chances.
9.lt'ld5 does not fit in with the
move .ie3 since White �11 lose a
tempo after 9 . . . lt'lxd5 lO.exdS lt'le7
ll.c4 lt'lf5 = , for instance, 12 . .id2 .ie7
. under control.
13 . .id3 0-0 14.0-0 lt'lh4.
Al. 9.J.c4 J.e7!
Black solves the opening prob­
lems with natural, typical Sicil­
ian methods. 9 . . . .ie6 lO .lt'ldS lt'lxe4
ll . .ib6 �gS 1 2 . 0 - 0 ! i.xdS 13.f4 exf4
14.hd5 is dangerous for him.
10. �d5 �d5 l l.J.d5 0-0
Black would be happy to trade
his bad bishop with . . . .igS. That
would be possible after:
1 2 . lt'l c3 .igS 13 .J.g5 \1;llg5 14.0-0
.ih3 15. �f3 .ie6 16.�ac1 �ac8
17.'i!th1 h6 18.a3 lt'laS 19.'11;Yd 3 �e7
2 0 .�fd1 �fd 8 = , Sulskis-Tregubov,
Koszalin 1999;
1 2 . 0- 0 .igS 13.�e1 .ie6 14.lt'la3
he3 15.�xe3 �c8 16.c3 (16.�d3
\1;lfc7 17.he6 fxe6 18.�xd6? lt'ld4),
J.Geller-Grachev, Moscow 2 0 14,
when simplest is 16 ... �c7;
12.lt'la3 .igS (12 . . . 'i!th8 ! ? 13.0-0
fS is sharper) 13.lt'lc4 he3 14.lt'lxe3
lt'le7.
Therefore, White should take gS
12 .�d2 .ie6 13. � c3 �k8
14.0-0 �c7!
It is essential to avoid b7-b5 as it
only provides White with a lever on
the queenside, e.g. 14 . . . b5? ! 15.a4 b4
16.lt'la2 ! ;t aS 17.lt'lc1 i.xdS 18 .�xd5
lt'ld4 19 .i.xd4 exd4 2 0 .lt'ld3�.
After the text, Black has no prob­
lems :
15.a4 lt'la5 = , Bacrot-Tajmanov,
Cannes 1998 ;
15.�ac1 lt'l a5= 16.�fd1 lt'lc4
17.hc4 �xc4 18.a3 �c6 19.lt'ld5
i.xdS 2 0 .\1;llx d5 �fc8 2 l .c3 g6 2 2 .f3
249
 Part 7

�xd5 23J"�xd5 f6 24.'i!ff2 'i!ffl 25.g4 13 ••• ti'e8 !

'i!fe6 26J!cd1 fS, Wedberg-Gobet,
St. Martin 1991.
A2. 9 .i.g5 i.e7 10.J.xf6
• 14.lt:Jd5 .ixdS 15.�xd5 �dB 16.�c5
It is very tempting to keep the
bishop for a while with 1 0 .lt:l d 2 ? !
intending to play lt'lc4 first and only
then exchange on f6. That would al­
low White to achieve the lt:Jd5-lt:Je3
set-up. However, 10 . . . .ig4 ! ll . .ie2
.ie2 12 .lt:Je2 d5= spoils this plan -
see game 41 .
10 .•• .i.f6 ll.tL!d2
I faced this idea as White many
years ago in my game Delchev- Ber­
nal Moro, Hyeres 2 0 0 1 , which went
lt'ld4? ! 17.i.d3 �e6= . I underesti­
mated it at that time, but later I re­
turned to it and my wife tested it on
the Black side in 2006. White im­
mediately erred - 14.lt'le3? ! �dB
15.'1Wc7? �d7 16.ti'b6 idS 17.�c5
.ie7 18 .ti'b6 .ib4+ and went on to
lose, Lillo Castan-Djingarova, Illes
Medes 2 0 07. Since no one seems
to have entered this position .ever
since, here is our "family" analysis :

14. tL!d5

14.'\Wd2 �dB 15.lt'ld5 (15.�c1

ti'e7) 15 . . . lt'ld4 16.lt:Jce3 (16.lt'lxf6+
gxf6 17.�c3 ti'c6) 16 . . . .ig5 17.c3 (17.
h4 i.h6) 17 ... i.xd5 18.exd5 �xdS+.

14 .•• hd5 15.ti'xd5

11 ••• 0-0 ! 12.tL!c4 i.e6 13.ti'xd6 15.exd5? lt:Jd4 16.lt:Je3 i.g5 17.�d1
(17.i.d3 e4) 17 .. .f5 18.c3 �f6.
13.lt:Jxd6 �e7t is dubious for
White. 15 .•. gds 16.ti'c5
White's setup does not make
much sense if he does not take the
pawn at all : 13.lt:Jd5 bS 14.lt'lce3 (14.
lt:Jxd6 i.xdS 15.�xd5 vtfc7) 14 ... i.g5
15.i.e2, Ponomariov-Dubov, Trom­
so 2013. This is the same position
as in line C, but it is Black to move !
13.lt:Je3 vtlb6 14J!b1 �d4 (14 . . .
i.gS ! ? 16.lt'lcd5 �aS+ 17.b4 vtld8)
15.i.d3 lt:Jb4 gave Black an initia­
tive in Hartman-Shulman, - Stock­
holm 1998.

250
 Taimanov with 5.ll'lb5
16 ••• .ie7!
16 . . . '1We6 17.ll'le3 (17 . .id3 �d4 ! ;
17.c3 'Wg4) 1 7. . . �d4 1 8 . .ic4 �c8
19.c3 1i.e7 is roughly equal.
17.�b6
17.'1We3? b5 18.ll'ld2 (18.lt'la3 f5)
18 . . . lt'lb4-+ .
17 ltlb4 1 8 .id3 (H3 .�c1 .ig5
• • • •
19.lt'le3 �a4 ; 1 8 .lt'le3 'Wd7- +) 18
•••
gxd3 !
18 . . . �d4 wins the queen, but
White holds on: 19.0-0 'Wb8 2 0 .a3
( 2 0 .c3 �xd3 21.cxb4 �d4) 20 . . . �c8
21.axb4 .idS 2 2 .�c5 �xeS 2 3.bxc5 = .
19.cxd3 ltlxd3 + 2 0 . �e2
Only move. 20.'it>f1 .ic5 2 1.'1Wc7
( 2 1.�xb7 �d8) 2 l . . .�b5 ; 2 0 .I!?d2
'Wd7 2 1.�hd1 �c8 .
20 ••• J.c5 2 1.Yfxb7
2 1.'1Wc7? loses to 21.. .ltlf4+ 2 2 . �fl
( 2 2 . 1!lf3 f5 ; 2 2 .�d2 �b5) 2 2 . . . Yfe6
23.'1Wxc5 �c8 24.�d1 h6 25.�b4 �xc4.
Black has a strong initiative. His
queen is eyeing both d4 and h4, f2 is
cracking. White can stay in the game
with best defence only. It seems that
he has a wide choice, but this im­
pression is deceptive :
2 2 .�hd1 ll'lf4 + 2 3 .1!lf1 �h4 ;
2 2 .�b3 ll'lxf2 2 3 .�hf1 ll'lxe4;
2 2 .ll'le3 ll'lf4 + 23.1!?f3 Yfh4 24.h3
he3 2S.@xe3 (25.fxe3 lt'lh5) 25 . . .
lt'lxg2 + .
Only 2 2 .g3 lt'lxf2 23.�d5 allows
White to consolidate - 23 . . . '1Wc8
24.lt'lxe5 ll'lxh1 25.�xh1 .ib6 2 6.�c6
Yfh3 27.\!?f3 .id4 2 8 .'1Wd7 �hS + .
Black regains the material while
the enemy king will remain in the
draught for long.
A3. 9. ltld2 ! ?
I guess that this move owes its
current popularity to the variation
9 . . . 1i.g4 ? ! 10 . .ie2 he2 ll.'Wxe2 d5
12.exd5 lt'lxd5 13.lt'lxd5 '1Wxd5 14.0-0
fie? 15.�g4 which is slightly better
for White. We'll disillusion the op­
ponent with :
9 ••• .ie6 ! ?
251
 Part 7
This move is based on the pawn
sac 10 . .ig5 .ie7 11.tt:lc4 0-0 ! - as in
line A2 .
I suppose that 9 . . . l!Jg4 10.tt:ld5
l!Jxe3 ll.tt:lxe3 .ie7! also equalises -
12 . .ic4 0-0 13.0-0 .ig5 14 . .id5 .ixe3
15.fxe3 Wfc7 16.l!Jc4 (16.Wfh5 l!Jb4
17 . .ib3 .ie6) 16 . . . l!Je7 17.tt:lxd6 .ie6
18.l!Jf5 .ixf5 19.exf5 WfcS.
10 .li)c4
• b5 n . .!L)b6 gbs
12 .li)bd5 .li)g4 (12 . . . .ie7 13.a4 ! )
• 'it> h 8 which left the fS -rook unde­
13.i.e2
White cannot exploit our move
. . . bS by 13.a4 since 13 . . . b4 14.tt:lb1
l!Jxe3 15.l!Jxe3 Wfc8 16.tt:ld2 .ie7
17 . .id3 0-0 was nice for Black in
Eid-Al Modiahki, Casablanca 2 0 0 2 .
1 3 . .ic1 may b e questioned by
13 . . . .ie7 14 . .ie2 l!Jf6. White's bishop
might be more active on d3, moreo­
ver, we have sidestepped the varia­
tion 12 . . . .ie7 13.a4 ! .
13 .li)xe3 14 .li)xe3 i.e7
.•• •
15 .li)cd5 i.g5 16.0-0 0-0 17.c3
•
This position commonly aris­
es from the Sveshnikov. It is rare­
ly seen as Black has a decent game.
252
He may play for equalisation with
17 . . . l!Je7 18.�d3 .ixe3 19.tt:lxe3
Wfc7 (Rogozenko), or seek compli­
cations with the thematic break
. . . f5 - 17 . . . g6 18.a4 fS 19.axb5
axbS 2 0 J'l:a6 (20.exf5 gxf5 2 1 .f4
exf4 22 .tt:lc2 'it>h8) 2 0 . . . Wfc8 2 U'l:b6
l'l:xb6 2 2 .tt:lxb6 �b8 23.exf5 gxfS
24.tt:lbd5, Shukh-Frolyanov, Na­
bereznye Chelny 2 0 1 0 , 24 . . . �b7! oo
(in the game, Black answered 24 . . .
fended in the line 25.hb5 ! �xbS
26.l!Jc7 Wfb3 27.Wfxd6).
B . 6.c4 .li)f6
6 . . . a6? ! is not outright bad, but it
is certainly inaccurate. In some var­
iations this move is not necessary,
for example after 7.l!J5c3 , as noted
in the next paragraph.
7 .li) lc3
•
7.l!J5c3 is rarely seen. Black usu­
ally saves a tempo on a6 and ob­
tains a good game. White is unable
to employ efficiently his b1-knight
- 7 . . . .ie7 8 . .ie2 0-0 9.0-0 b6 10 . .if4
(or 10 . .ie3 .ib7 11.tt:ld2 dS= 12.exd5
exdS 13.cxd5 l!Jb4) 10 . . . .ib7 ll.l'l:e1
l'l:c8 (ll . . . Wfd7 12 .tt:ld2 l'l:fd8 13.l'l:c1
l'l:ac8 14.a3 h6 15 . .ig3 l!Jh7 16.'it>h1
tt:ld4 17 ..id3 l!Jf8 18.tt:lfl tt:lg6 is ano­
ther interesting set-up) 12 . .if1 l!Je5,
see game 42 Kavalek-Liberzon,
Amsterdam 1977.
The rematch Fischer-Spassky
199 2 , drew attention to 7.l!J5c3 .ie7
8.g3. Black chose the logical plan of
 Taimanov with S.liJ bS
preparing . . . bS by . . . a6, . . . i.d7, but
it is a bit passive. Perhaps the at­
tack on c4 is more effective : 8 . . . b6 !
9.i.g2 i.b7 1 0 . 0- 0 0-0 ll.b3 (ll.f4oo
�c8 12 .b3 a6 13 .i.e3 tl:\d7 14.tl:\d2
i.f6 15.�c1 tl:\cS) 1l.. .�c8 (ll ... tl:\d7 ! ?
12 .i.b2 tl:\ c5 13.tl:\a3 ll.J e S i s a n orig­
inal idea. Black was slightly better
after 14.�e2 tl:\ed3 15.�ad1 tl:\xb2
16.�xb2 a6+, Nemeth-Markus,
Szombathely 2 0 03) 12 .i.b2 �c7
13.tl:\d2 �b8 14.�c1 a6 15.�e2
tl:\d7 16.�fd1 (16.a3 bS ! ) 16 . . . tl:\c5
17.tl:\f1 bS !
7 • • • a6 8.c!ila3
8.tl:\d4 is completely innocuous
as Black can equalise at once with
8 . . . d5 9.exd5 exdS 10 .cxd5 tl:\xdS
ll.tl:\xc6 bxc6 12. tl:\xd5 cxdS. How­
ever, I would recommend 8 . . . i.d7 ! ,
aiming t o push . . . bS - 9.tl:\c2 [9 .i.e2
�c8 10.0-0 (10.tl:\c2 tl:\e5 ll.tl:\e3 gS
12 .b3 bS 13.a3 �b6 14.�b1 i.c6 15.f3
g4) 10 . . . tl:\a5 11.b3 bS] 9 . . . i.e7 10 .i.e2
0-0 ll.i.f4 (11.0-0 �c8 1 2 .f4 bS
13.cxb5 axb5 14.hb5 tl:\xe4 15.tl:\xe4
�b6+ 16.tl:\e3 �xbS 17.tl:\xd6 hd6
18.�xd6 �fd8) ll . . . tl:\e5 1 2 . 0-0 �c8
13.tl:\e3 bS ! 14.cxb5 axbS 15.a3 �b6,
Palac-Rublevsky, Aix-les-Bains 2011.
8 . . . b6
This move order is hardly bet­
ter than 8 . . . i.e7 followed by 9 . . . 0-0.
I chose it for the main line because
it is more popular lately. Initial­
ly it was designed to counter f4-
plans with an early . . . dS. Black was
afraid that otherwise White would
achieve 9.i.e 2 , 10.f4 and ll.i.f3. The
text move allows to meet 9.i.e2 i.b7
1 0 .f4 by 10 . . . d5 ! with initiative, e.g.
ll.cxdS exdS 12 .e5 d4 13.exf6 dxc3
14.�xd8 tl:\xd8 !+.
Let's suppose, however, that
Black played 8 . . . i.e7 instead of 8 . . .
b 6 . After 9.i.e2 0 - 0 1 0 . 0-0 b6 ll.f4,
my analysis proves that ll ... dS is
perfectly possible:
a) 12 .e5 tl:\e4 ! (this is better than
12 . . . -tcS ? ! 13.<±>h1 tl:\e4 14.i.f3 tl:\xc3
15.bxc3 tl:\e7 16.cxd5 exdS 17.tl:\c2
i.fS 18.tl:\d4 i.e4 19.f5;!; or 14 ... tl:\f2
15.�xf2 hf2 16.cxd5 exdS 17.hd5
i.b7 18.tl:\c4 i.cS 19 .i.e3t, Valle­
jo Pons-Shavtvaladze, Wch U16
Oropesa del Mar 1998) 13.tl:\xe4
dxe4 with initiative.
b) 12 .cxd5 exdS 13.exd5 (13.e5
i.cS 14.<±>h1 tl:\e4 15 .�d5 tl:\f2 16.�f2
�ds 17.tl:\ds i.f2+) 13 . . . tl:\d4t.
c) 12 .exd5 i.c5+ 13.<±>h1 tl:\d4
253
 Part 7
14 . .ie3 lt:lxe2 15.'1'9xe2 exd5 16.hc5
bxc5 17.cxd5 lt:ld5 = .
I n the game A . Karpov-G. Kaspa­
rov, World Championship Moscow
(12) 1985, the young challenger sur­
prised the chess world with 8 . . . d5? !
which was proclaimed "novelty of
the year". The game went on with
9.exd5 exd5 10.cxd5 lt:lb4
Kasparov scored 1,5 out of two
with this move and created a mem­
orable evergreen, but the initial en­
thusiasm quickly evaporated. White
discovered that he could continue
developing by ll . .ie2 . Then neither
ll . . . .ic5 12 . .ie3 ! .ie3 13.'1'9a4 lt:ld7
14.'1'9xb4 .ic5 15.'11*/ e 4;J; nor ll . . . lt:lfxd5
12.0-0 .ie6 13.'11*/ a4 b5 14.lt:laxb5 ! ?
axb5 15.hb5--+ i s enticing for Black.
9 .ie2 J.b7 10.0-0 (10.f4 d5 ! )
.
10 .te7!
.•.
10 . . . ltlb8? ! aims to draw the sting
of plans with f2-f4 or '11*/b 3 (which is
met by . . . lt:lbd7) . I do not recom­
mend this manoeuvre because of
three reasons :
1. I think that White's most dan­
gerous plan is based on f2-f3 so eve­
ry effort of Black should be aimed
against it;
254
2. I find the manoeuvre
. . . lt:lc6-b8-d7 over-prophylactic. I
prefer a more concrete approach
with . . . d7-d5 in mind. Black should
redeploy his c6-knight indeed, but
e5 is a better place to go than b8.
Then in some variations Black can
find a better use of his centralised
knight than retreating it to d7. In
some variations, especially when
White places his queen on b3, Black
has . . . lt:lb4 (after . . . ltlc5 first) and
. . . d5.
3 . In fact Black is unable to avoid
f4-plans. Moreover, he obtains a
passive version of them after 10 . . .
lt:lb8? ! ll.f3 lt:lbd7
12 . .if4 ! ? lt:le5 13.'11*/d 2 .ie7 14J:Ud1
'11*/c7 15J:!aci 0-0 16.'it>h1 l:!ac8 17 . .ig3
l:!fd8 18.f4 lt:lg6 19 . .if3 - see game
45 Brodsky-Kruszynski, Pardu­
bice 1993. Notice that in this exam­
ple Black's choice is considerably
restricted while White could also
opt for 17 . .ie3 which is even a big­
ger source of concern for the second
player.
ll.i.e3
a) ll.f4 0-0 (ll . . . d5 12.cxd5 exd5
13.e5 .ic5+ 14.'it>h1 lt:le4 15.lt:lxe4;J;)
 Taimanov with S.tt'lbS
12 .i.f3 !lc8 13.i.e3 transposes to the
main line.
Another plan for White is :
b) 11.ll'lc2 0-0 12. tt'le3
This setup is aimed against the
freeing break . . . d6-d5. White hopes
to regroup his forces later by the
scheme b3, tt'le3-c2-d4, i.e3, return­
ing to the usual queenside plans.
Black should not get lulled by the
seemingly unambitious play of the
opponent. Perhaps his best set-up
includes the following moves : 12 . . .
W!c7 13.Wid2 !ladS 14.!ld1 !lfe8 15.b3
(15.f3 dS) 15 . . . tt'le5 and if 16.f3 , then
16 . . . d5 ! 17.exd5 tt'lg6 ! .
11 . . . 0-0
Bl. 12 .£4 ; B2. 12. '!Wb3 ; B3. 12.£3 ;
B4. 12. !lc1
12.Wffd 2 is an innocuous alterna­
tive. Black answers 12 . . . tt'le5 13.£3 ,
when 13 . . . d5 ! ? 14.exd5 exd5 15.cxd5
i.xd5 16.f4 tt'lc6 17.!lad1 tt'lb4 18.tt'ld5
tt'lfdS or 18 .i.d4 !le8 is equal. More
tangled is 13 . . . !le8 ! ? - see game 47
Rabrenovic-Delchev, 2 0 05.
White has three major plans in
the diagram position. The attempt
for a kingside attack with f4 faces an
energetic counterplay in the centre.
I propose to attack c4 after prelimi­
nary . . . !lc8 and . . . !le8. In most cases
Black solves his problems with the
help of the break . . . d6-d5, followed
by . . . i.e7xa3. This exchange ensures
an excellent outpost on c4 to the
aS-knight. The variations are very
sharp and nearly untested.
12.Wffb 3 has been Karpov's fa­
vourite choice. He used to win this
position with both sides against any
opposition, including Kasparov.
I propose a modern set-up which
gives Black comfortable play.
White's most dangerous plan is
examined in lines B3 and B4. It is
aimed at bolstering up the centre
with f3 and rooks on c1 and d1. Then
both bishops retreat to safer places
while the king goes to hl. Finally,
with all possible prophylaxis done,
White plays b4 and looks for possi­
bilities to expand on the queenside
by a4-a5 or c4-c5. Although the po­
sition is dynamically balanced, if
Black misses the right moment to
generate counterplay, he risks to get
in a murky situation.
It is difficult to show a clear­
cut counter-plan for Black because
the position allows many differ­
ent move orders. I'll try to provide
you with concrete variations where
Black equalises by pushing . . . dS at
the right moment. However, you
should be acquainted with the main
plans of Black in the event you land
in a position which requires long
and patient manoeuvring.
255
 Part 7
B1. 12.f4
In the early days of the hedge­
hog structures, White believed
that he should punish the oppo­
nent for his "passive" play by mat­
ing him. Practice has shown that
Black has enough counter-chanc­
es due to the unstable centre and
weaknesses on c4 and e4. However,
if White consolidated and brought
his knight to the centre by the ma­
noeuvre t!Ja3-b1-d2 , he would have
the better chances. Then he would
be able to choose pawn storms on
the queenside by b4-a4-a5 or on the
opposite wing by g4-g5.
12 . . . gcs
Another possible design is to re­
deploy the c6-knight to d7 via b8
and adopt a waiting strategy, but
I'm against such an approach when­
ever we have another clear plan.
13.gcl
13 .i.f3 is less accurate due to
13 . . . t!Ja5 14.�d3 (or 14.�e2 d5
15.exd5 ha3 16.bxa3 !!xC4+) 14 .. .
d5 ! (A good idea of Cramling. 14 . . .
256
�c7? ! stumbles into 15J'!acl t!Jd7
16.b4 ! when Black must retreat un­
der the blow 17.t!Jd5 ! ±) 15.cxd5 b5
16.t!Je2 exd5 17.e5 t!Je4 = .
Black can also transpose t o the
main line with 13 . . . !%e8 14.!k1. In­
stead, 14.vtfd2 t!Ja5 was balanced in
game 43 Radulski-Iotov, Sunny
Beach 2009. I also consider in the
annotations 14.�e2 e5 ! ? .
13 . . . ges 14 . .if3 �a5
Black scores well without for­
cing the events, e.g. 14 . . . t!J d7 (or
14 . . . i.f8) 15.vtfe2 vtfc7 16.!!fd1 vtfb8.
The downside of this approach is
that White is controlling firmly the
centre and can gradually build a
kingside attack with g4.
A crucial position for this line.
White is unable to prevent a clash
in the centre :
15.�d3
Alternatively:
a) 15.�e2 vtfc7 16.b4 (The game
Mannion-Lautier, Barcelona 1993,
went 16.t!Ja4? ! he4 17.hb6 vtfc6
and Black has a pleasant game. The
 Taimanov with 5.tt'lb5
exchange of the central pawn for
b6 is good for White only without
queens. Karpov proved that against
Kasparov in a famous game from
the World title K-K series. 16.�f2 is
also harmless - 16 . . . �b8 ! 17.hb6
tt::l xc4 18.tt'lxc4 l':l:xc4oo.) 16 . . . tt::l xc4
17.tt::l xc4 �xc4 18.�b2 �d3 19.hb6
tt::l x e4 2 0 . l':l:fd1 �xc3 2l.l':l:xc3 tt::l x c3�.
I proposed this queens sac in
The Safest Sicilian, but there are no
candidates to test it.
b) 15.tt'la4 tt::l d 7 16.�d2 (16.b4
d5} 16 . . . tt'lc5 17.tt'lc3 tt::l d 7 18.�e2
(18 . .ie2 tt'lf6) 18 . . . �c7 19.l':l:fd1 tt::l xc4
2 0 .tt'lcb5 axb5 2l.tt'lxb5 �b8 2 2 . l':l:xc4
l':l:xc4 23.�xc4 l':l:c8�.
15 . . . d5 ! 16.cxd5 (16.exd5 ha3
17.bxa3 tt::l x c4) 16 . . . b5 17.e5
17.tt'lcxb5? ! axb5 18.�xb5 l':l:a8
19 . .ib6 �d7+ ;
17.id4 b4 18.e5 tt::l xd5 19.tt'lxd5
hd5 2 0.l':l:xc8 �xc8 2 1.hd5 bxa3t.
17 ... .txa3 18.bxa3 c!Dxd5
19. c!Dxd5 hd5 20.gxc8 �xeS
21 . .lxd5 exd5 22.Yfxd5 c!Dc4
23.J.f2 c!Dxa3 24.gdl c!Dc4 25.h3
Yfe6 = .
B 2 . 1 2 . §'b3 c!Dd7
12 . . . tt::l a5? ! 13.�xb6t led to a bet­
ter endgame in Karpov-Kasparov,
1984.
a) 13.f3 is inconsistent. Black
could continue as in the main line
or switch to ideas from line B3 by
12 . . . l':l:e8.
b) 13.l':l:ad1
White enables the manoeu­
vre tt'l a3-b1 · and leaves the other
rook on f1 from where it could sup­
port f4. The general crisis of the f4-
plans had swept the queen's rook
move out of fashion, but Naiditch
and K.Szabo have been trying to re­
vive it lately. Perhaps the most logi­
cal approach is to underline the ab­
sence of White's rook from c1 by
preparing . . . d5 :
13 . . . tt'l c5 14.�c2 (14.hc5 bxc5
15.�xb7?? drops the qeen after 15 . . .
tt::l a5 ! ) 14 . . . �c7!
Naiditsch-Antoniewski, Berlin
2 013, went 14 . . . if6 15.tt'lab1 �e7
16.a3 aS 17.tt::l b5 l':l:fd8 18.tt'l 1c3? ! a4
19.f3, when 19 . . . tt'la7 neutralises
White's pressure. Critical, howev­
er, is 18.b3 ! ? , planning tt'l c3-a4. The
b6-pawn would be a sore point.
15 . .!Dab1
K.Szabo-Ribli, Paks 2 0 14, saw
the tricky 15.�d2 when 15 . . . tt::l b4 is
already dubious due to 16.f3 - set­
ting up the trap 16 . . . l':l:ad8? 17.tt::l cb5 ! .
257
 Part 7
Black answered 15 . . J�fe8 (15 . . .
l:'lad8 ! ?) 16.ltlc2 l:'lad8 17.b4 ltld7
18.l:'lc1. The knight does have more
prospects on c2, but it also hinders
the rook to defend the c4-pawn.
Ribli could have exploited that with
the thematic 18 . . . ltlce5 19.ltld4 ltlf6
20.£3 d5! with an initiative.
15... tt'lb4 16. Y!fd2 EfadB!
Now 17.a3? fails to 17 ... ltlb3,
the other moves allow ... d5, for in­
stance: 17.b3 d5 ! 18.exd5 exd5
19 . .if4 %Yc8 ; 17.f3 d5 ! . The only way
to hinder the break is:
17.if4, but then we can attack
the centre with .. .f5 - 17 . . . @h8 18.b3
(18.f3 fS) 18 .. .f5 19.exf5 l:'lxf5 2 0.ig4
l:'lf6 2 l..ig3 l:'lg6 2 2 .%Ye2 if6oo.
c) 13.l:'lacl ltlc5 14.'Wc2 if6
15.gfd1 transposes to the main line.
258
13 ••• tt:lc5
In reserve, Black has another
good set-up: 13 . . . l:'le8 14.l:'ld2 l:'lc8. If
White persists with his initial idea
he could quickly find himselflost a �
in Prie-Semkov, B UL-FRA m. So­
fia 1 9 9 0 : 15.'Wd1 Wic7 16.'Wf1 ltlce5
1 7.l:'lad1 ltlf6 18.£3, when 1 8 ... d5!
19.exd5 .ha3 20.d6 .hd6 21.l:'lxd 6
lLJd5+ would have crowned Black 's
strategy.
More accurate is 15.l:'lac1 ltlce5
16.Wd1 ltlg6 ! (16 . . . 'Wc7? 17.f4 ltlg6
18.e5±) 17.�f1 (17.f4 %Yc7 18.e5 d5
19.cxd5 .icS) 17. . . �c7 18J!ddL
White protects the c1-rook
against central breaks. The game
Psakhis-Ro m anishin, Szirak 1986,
went 18 . . . %Yb8 19.f3 if6 2 0 .ltJ ab1
idS 2 1 . @h1 draw, but Black could
have played more actively - 18 . . .
ltJf6 19.f3 'Wb8 ! ? 2 0 .'Wf2 d 5 2l.cxd5
id6 !+.
14. §'c2 (14. .hc5?! bxc5 15. V9b7??
ltla5) 14 .if6 ! 15 .!3acl
•.• •
a) With the rook still on a1,
15. ltl ab1 is dubious. Black could
counter-attack with 15 ... ltJ b4 ! 16. 'Wd2
aS (16 ... �c8 ! ?) 17.f3 %Ye7 18.i.f4 i.e5
 Taimanov with 5.lLlb5
19 . .ig5, Gajsin-Zakharstov, St. Pe­
tersburg 2 0 0 2 , 19 .. .f6 ! ? 2 0 .i.e3 f5
2l.exf5 exf5t.
b) Another dubious alternative
is 15.f3 i.e5 16.�d2 �h4 17.g3 'Wf6t.
c) 15.�ab1 creates the threat ofb4
so Black has to blockade the pawn :
15 . . . lL\b4 16.�d2 i.e5 ! (it is good to
provoke f4. In Averbakh-Poluga­
evsky, Palma de Mallorca 1972 was
16 . . . hc3? 17.bc3t) 17.f4 i.c3 18 .bc3
lL\e4 19 .�b2 lLlc6 = .
15 ••• Yfe7!
The queen feels superbly here.
This set-up faces White with prob­
lems. The more provocative 15 . . .
i.e5 16.lL\ab1 'Wh4 17.g3 �f6 18.�d2
'Wg6 19 .f4 lL\xe4 2 0 .lLlxe4 �xe4
2 l .'i!lf2 'Wf5 2 2 .g4 'Wf6 23 .g5 'Wf5
24.'i!le1 'We4 25.fxe5 'Wh1+ 26.i.fl
lL\xe5 27.�f2 b5 ! also gave Black
good chances in lvanchuk-Wang,
Beijing 2 0 13 .
16.c!Llabl
Alternatives are:
16.f3 �fd8 17.�d2 lLlb4 leads to
similar positions - 18.lLlab1 (18 .�e1
d5 ! +) 18 . . . d5 !+.
16.�d2 �fd8 17.�e1 ! (17.f3 lLlb4
18 . .if4 i.e5 19.i.g5 f6 2 0 .i.e3 d5
21.cd5 ed5 2 2 .lLld5 lLld5 23 .i.c4 lLla4
24.i.d4 bS 25 . .ib3 'i!lh8+) 17 . . . lLlb4
18.lLlab1 d5= , Magem Badals-Rubio
Mejia, Linares 2 0 13 .
16 c!Llb4 ! 17.Yfd2 gfd8
•••
Threatening . . . d6-d5. Black
gains the initiative after:
18 .J.d4
18.f4 aS ! + (attacking e4) 19.i.f3
lL\cd3 ! (Almasi) 2 0 .hb6 �db8 ! + ;
18 .�e1 d5+, Pokojowczyk-G.
Kuzmin, Sochi 1976.
18 • • • J.e5 !
18 . . . hd4 19.�xd4 �g5 is good,
too : 2 0 .�d2 h6 2 l .�xg5 hxg5
2 2 .a3 lLlc6 23.lLld2 lL\ d4 24.i.f1
lL\db3 25.lL\xb3 lLlxb3 2 6.�bl 'i!lf8
27.f3 'i!le7+ ; 2 0 .f3 lLlbd3 2 1.hd3 e5
2 2 .�f2 lL\xd3.
19.Yfe3 c!Llc6 2 0 .ixe5 dxe5
•
259
 Black has a clear edge. Svetush­
kin-Ducarmon, Roquetas de Mar
2014, went further 21.tt'ld2 �d4
(21.. .a5) 22 .tt'lf3 �adS 23 ..ifl f6.
B3. 12.f3
White starts building the ideal
set-up against the Hedgehog. How­
ever, Black is able to hinder his plan.
The rook's X-ray along the e-file
cocks the trigger of the break . . . d6-
d5. Instead, 12 . . . tt'lb4 13.�b3 ! d5
leaves Black lagging behind in de­
velopment - 14.cxd5 exd5 15.�fd1
Wfc7 16.�ac1 ! .ic5 17.lt:Ja4 .ie3
18 .Wie3±.
13.Wb3
The alternatives run into . . . d6-
d5 with a strong initiative:
13.Wie1? ! d5 ! + ;
13.�c1?! i s even worse : 13 . . . d5 !
14.cxd5 exd5 15.tt'lxd5 tt'lxd5 16.exd5
tt'lb4+;
13.Wid2 could also be met by 13 . . .
d5 ! = 14.cxd5 exd5 15.exd5 tt'lb4.
260
Black's knight has found a more ac­
tive stand than e5 ! We see here one
more reason to delay the manoeu-
vre . . . tt'lc6-e5. 16.d6 ! (16.tt'lc4 tt'lfd5+)
16 . . . �xd6 17.Wffx d6 hd6 18.tt'lc4
.if8 19 . .if2 tt'lbd5 2 0 . �fe1, Rindlis­
bacher-Huber, Legnica 2013, b5 ! = ;
13.tt'la4? ! was invented by a
computer: 13 . . . tt'ld7 14.tt'lxb6 tt'lxb6
15. Wffb 3, Comp Mephisto Genius -
Loeffler, The Hague 1997. Here the
human missed 15 . . . tt'ld7! 16.Wffb 7
tt'lc5 17.�c6 (17.hc5 dxc5 18.Wffx c6
Wffd 4+ 19.�f2 Wffb 2+) 17 . . . �c8+.
The classic book on the Scheve­
ningen by Kasparov/Nikitin, 1984,
mentions this as the best move,
without any analysis though. It took
several decades to the mere mortals
to reach to the same conclusion.
14.�fd1 was the old main line -
14 . . . tt'lc5 15.Wffc 2 .if6 16.�d2 !
16.�ac1 .ie5 17.�d2 Wffh 4 18 .g3
hg3 19.hxg3 �xg3+ is a draw by
perpetual, but Black can play on
with 16 . . . tt'lb4 17.�d2 d5= .
16.�ab1 tt'lb4 17.�d2 .ie5 ! (in­
tending . . . Wffh 4) 18.f4 hc3 19.bxc3
tt'lxe4 2 0 .�b2 tt'lc6 ! oo .
 Taimanov with S . .!tJbS
16 . . . i.e5 ! 17.ltJc2 (17J�acl �h4 !
18.g3 hg3 19.hxg3 �xg3 + = ) 1 7. . .
�h4 18.g3 hg3 = .
White hopes to revive his a3-
knight with .!tJ ab1 so we must tie it
up with the defence of the c4-pawn.
14 . . . ltJc5 15.�c2 ltJb4 16.�d2
�b8 (16 . . . �c7 17.ltJab5) is accepta­
ble, but a bit passive. White could
continue 17.ltJab1 (17.ltJc2 .!tJxc2
18 .�xc2 i.c6 19.Ei:d2 bS+!) 17 . . . Ei:d8
18.i.f2 ! (threatening a3) 18 . . . i.f6
(18 . . . d5 19.exd5 exdS 2 0 .a3 ltJb3
2 1.�e3 i.cS 2 2 .�g5 ltJc2 23.ltJxd5±)
19.a3 ltJc6 2 0 .�c2 ( 2 0 .b4? ! ltJb3
2 1.�d3 .!tJbd4) 20 . . . a5 2 1 . ltJb5 a4
2 2 . ltJ 1c3 ltJ a7.
15.�c2 should be met by 15 . . .
.!tJceS ( 1 5 . . . d 5 looks good only at
a cursory glance. After 16.cxd5
exdS 17 . .!tJxd5 ha3 18 .bxa3 ltJb4
19 .�xc8 ! , White has a nasty initia­
tive - 19 . . . �xc8 2 0 . axb4 bS 2 1.Ei:c1
�b8 2 2 . Ei:fd1 f5 23.ltJc7 Ei:e7 24.ltJe6
fxe4 25.f4 Ei:xe6 26.Ei:xd7 i.c6 27. Ei:a7
Ei:d6 28.Ei:xa6 i.e8 29.Ei:a7 Ei:d7
3 0 .Ei:a5;!;) 16.�d2 ltJf6 .
15 . . . �c7
15 . . . ltJa5 ! ? targets the c4-pawn
- 16.�d1 (16.�c2 dS ! 17.Ei:fd1 ha3
18.bxa3 ltJxc4) 16 . . . �c7. White has
not a rook on c1 to defend c4 if the
a3-knight headed to the centre.
17.Ei:c2 ! ? - the only way to untie
the a3-knight. Alternatives are :
17.b4 dS (17 . . . ltJc6 18.b5 ltJa7
19.c5 bxcS 2 0 .bxa6 i.c6 21. ltJ ab5
ltJxbS 2 2 .ltJxb5 hbS 23 .hb5 Ei:ed8)
18.ltJab5 (18.c5 ltJf6) 18 ... axb5
19.ltJxb5 �b8 2 0 .bxa5 bxaS ;
17.b3 Ei:ed8 18.ltJab1 i.f6 19.Ei:c2
�b8 2 0 .�d2 ltJ c6 2 1.Ei:d1 .!tJcS ;
17.c;t>h1 Ei:ed8 (17 . . . ltJe5 18.ltJa4
ltJd7 19.b4 ltJc6 20.b5 .!tJceS 21.bxa6
ha6 22.ltJc3 .!tJxc4 23.hc4 hc4
24.ltJcb5 hb5 25.ltJxb5 �c4 26.ltJxd6
hd6 27.Ei:xd6 .!tJeS) 18.b4 (18.b3 dS)
18 . . . ltJc6 19.�b3 .!tJceS 20.Ei:c1 .if6 .
17 . . . �b8 ! 18.Ei:cl (18 .�d2 ltJc6 !
19.ltJa4 i.a8 2 0 .b4 dS h) 18 . . . ltJc6
19.�d2 (19.ltJc2 .!tJceS ! 2 0 .b3 bS;
19.ltJab1 .!tJ ceS 2 0.ltJd2 .ic6 2 1.f4
ltJg6 2 2 .b4 .if6 23.�b3 bS) 19 . . .
.!tJceS and i t i s unclear how White
can improve from this point.
261
 Part ?
We have mobilised our forces
while the enemy's pieces are still
tangled on the queenside. Or next
task is to prepare an active plan on
the kingside, but not to rush with
it ! We should patiently manoeuvre
and await the best timing for .. .fS
or even gS. Of course, White should
also hinder . . . dS.
16.�dl - anticipating 16 . . . lLlce5?
17.f4. 16 .. J''kd 8 and 17.lLlc2 does
not help in view of 17 ... lLla5 while
17.lLlabl is passive - 17 . . . lLlce5 18.b3
lLlf6.
16 . . . lLlce5 17. YlYdl lLlf6 18.l:&dc2
9b8 19.Whl .ia8 20.�gl lLlfd7
2U�d2 gf8
I do not see how White can make
any progress. If he defends the c4-
pawn with 22 .�fl, we can make a
sortie out of the trench - 2 2 . . .f5 ! ?
23.exf5 E1xf5�.
You should remember
Black's primary aim - the
break d6-d5. If it is impossi­
•.•
ble, we keep pressure on c4.
262
B4. 12. gel ! ?
The old Kasparov/Nikitin book
gives this as White's main option.
Although it has drifted out of fash­
ion, this move greatly restricts
Black's options� By delaying f3,
White deprives the opponent of tac­
tical tricks along the e-file.
12 . . . lLle5
. . . E1e8 is not an obligatory move !
The text is more straightforward -
by hitting e4 Black is trying to drag
the pawn to f3. Then he could try to
carry on . . . d6-d5.
In the event of 13.f3, we should
activate the rook - 13 .. J'�e8 and
White can hardly prevent a break in
the centre:
14.�d4 lLled7 15.lLlc2 (15.E1fdl
dS ! ) lS . . . dS ! 16.cxd5 exdS 17.exd5
i.d6 18.E1fdl bS ! t ;
14.i.d4 lLlc6 15.i.f2 d S 16.cxd5
exdS 17.exd5 lLlb4 = , Beliavsky -
Liberzon, Baden 1980 .
Only 14.YlYb3 hinders . . .d S , but
after 14 . . . lLlfd7 15.E1fdl �c7 16.�c2
 Taimanov with S.lLlbS
Eiac8 17.'1Wd2 Eied8, White is still fac­
ing the problem of how to activate
the a3-knight.
13 . . . tiled7 14JUdl
14.f3 is attacked with 14 . . . d5 ! =
15.exd5 exdS ! (15 . . . �c5 ? ! 16.�d2
�e3 17.'1We3 edS 18.Eifd 1 ! ;t Karpov­
Tukmakov, URS-ch 1971) 16.cxd5
(18.Eifd1? Eie8 ! 19.'!9d2 dc4+) 16 . . .
�cS 17.'!9d2 �e3 18.�e3 EieS 19.�f2
llJdS 2 0 .lLld5 �dS draw, Tseshkovs­
ky-Geller, Las Palmas 1976 .
14.b4 i s also premature d u e to
14 . . . a5.
14 ••• Eie8
It is possible that 14 . . . �c7 is
more precise. The text move aims
for 15.f3 dS !
15.tilc2
15 . . . Eic8
15 . . . 1!9c7 occurred in my game
Morovic Fernandez-Delchev, Pula
2 0 0 0 , but Black is unable to al­
ter the hedgehog formation after
16.'i!th1 ! Eiac8 17.f3 �b8 18.�gl.
White has prevented ... d6-d5
and preserves his spatial advan­
tage. This does not implicitly mean
that he has an edge, but at least
Black should switch to defence. His
main task is to hinder b4, lLla1-b3
and eventually '!9d2 . In the game, I
managed to keep the balance with
18 . . . Eic7 19.lLla1 Eiec8 2 0 .lLlb3 lLlcS
21.lLld2 lLle8 2 2 .'!9e3 h6 23 .�fl �gS
24.1!9e1 �c6 25.�e3 he3 26.�xe3
llJa4 27.lLlxa4 ha4 2 8.Eie1 �d7
2 9.Eicd1 bS 3 0 .b3 �a7 31.'!9a7 draw.
16.f3
White should not delay this move
much longer. 16.b4 1!9c7 17.lLla3 �b8
18.lLlab1 lLleS 19.�f4 lLlc6 2 0 .�xb6
eS 2 1.�g5 lLld4 2 2 .�a5 lLl e4 was in
Black's favour, Tseshkovsky-Ribli,
Riga 1979 .
Instead of 17.lLla3, White had
17.f3. Then 17 . . . d5 ! ? (17 . . . 1!9b8 ! ?)
equalises after 18.exd5 exdS 19.�f4
(19.lLlxd5? lLlxdS 2 0 . cxd5 �f6
2 1 .�d3 hb2+) 19 . . . �d8 2 0 .�fl
dxc4 2 1 .hc4 = .
1 6 . . . d5 ! ?
16 . . . 1!9c7 transposes to the above
game Morovic Fernandez - Delchev.
263
 Part 7
17.cxd5 exd5 18.exd5
18.lDxd5?! lDxdS 19.exd5 .!f6
20.\Wd2 .!b2 2U!bl hc3 2 2 .\Wd3
iaS+.
18 . . . .id6 !
264
This position is unanimously as­
sessed as promising for Black The
game played between Leningrad
Chess Club-Central Chess Club in
1971 seems to prove that evaluation :
19.�d2 bs 2 0 . a3 �c7 2 1 . @hl
( 2 1.f4 lDcS�) 2 l . . .�b8 2 2 .g3 \WaS
23. @g2 lDeS�.
The e-mail game Nasybullin­
Bern, ICCF 2006, saw 19.�h4 bS
2 0 . .!d3 l:!eS 2 1..!f4 (21.lDe4 tDxe4 is
equal - 2 2 .\Wxd8 l:!xd8 23.fxe4 ttJcS
24 ..!f4 ttJxd3 25.l:!xd3 l:!xe4 26.hd6
l:!xd6 27.lDe3 l:!d7=) 2 l . . .l:'!h5 2 2 .�g3
\Wb6+ 23.@hl hf4 24.�xf4 lDxdS
25.lDxd5 l:!xd5 26.lDb4 l:!xc1 27.�xcl
l:!d6 28.h3 lDf6 = .
Pa rt 7. Ta i m a n ov with 5 . ltJ b5

Com pl ete G a m es

White overestimates his position

41. Almasi-lvanchuk
and makes an active move while he
Polanica Zdroj 2000
had to think about how to equalise,
probably with 15.tt'lc3 ! ? .
1.e4 c5 2.tt:\t3 tt:\c6 3.d4 cxd4
4.tt:lxd4 e6 5.tt:\b5 d6 6 . .U4 e5 15 • • • dxe4 16. tt:\xe4 .ie7 17.c5
7 .ie3 tt:\f6 8 . .ig5 a6 9.tt:\5c3 .ie7
•

10.tt:ld2 .ig4 11 . .ie2 he2 12.tt:\xe2
gy.
12."�'e2? tt:\xe4 13.tt'ldxe4 .igS+
loses a pawn.
Following the same risky strate­
17 �c7 18.W/b3 gad8 19.gfd1
• • •

g6 2 0 .W/c4 tt:\a5 21.�4 Wg7

12 • • • d5 = 2 2 . tt:\ 2c3?

According to Chess Stars book

Super Tournaments 20 0 0, White
could have maintained the balance
by 2 2 .tt'ld6 tt'lc6 23.�a3 b6 24.�a6= .

22 tt:lc6 ! 23.W/c4 f5 24.tt:\d6

• • •

.ixd6 25.cxd6 gxd6 26.gxd6

�xd6+ 27.gd1 tt:\d4 28.f4 b5
29 .�d3 gds 3o.wh1 h4!
13 .ixf6 .ixf6 14.0-0
The knight endgame is won.
•

In a later game, White played 31.tt:\e2 tt:\b5 ! 32 .�xd6 gxd6

14.tt'lc3 dxe4 15.tt'ldxe4 .ie7 draw, 33.gxd6 tt:\xd6 34.fxe5 tt:\c4
Fontaine-Lautier, FRA-ch 2 004. 35.b3 tt:\xe5 36. tt:\f4 a5 37.h4
<llf6 38.<ll g 1 tt:\g4 39.tt:ld3 <ll e 6
14 . • • 0-0 15.c4? ! 40.tt:\c5+ Wd6 ! 41. tt:lb7+ Wc7 0-1

265
 Part 7

c4-c5. It is the main reason I do not

42. Kavalek-Liberzon
recommend the stand with . . . gS in
Amsterdam 1977
the hedgehog. I prefer .. .fS or even
. . . h5-h4 instead. Of course, White
l.e4 c5 2 . �f3 e6 3.d4 cxd4
can try to redeploy his bishob, but
4.�xd4 �c6 5.�b5 d6 6.c4 �f6
that costs tempi.
7.�5c3 J..e 7 8 . .ie2 0-0 9.0-0 b6
10.i.f4 J..b 7 lU:!el �k8 12.J.fl
�e5

In the game Lopez Martinez-Flu­

White's plan is l'!c1, a3, b4, then
i.g3 and f4 in order to remove the
restraining pressure on c4. Sprag­
get got an edge in two of his games
following this scenario. Our task
should be to safeguard the eS­
knight with . . . g7-g5. 13.ll:\a3 would
have been passive, for instance, 13 . . .
W!c7 14.l'!c1 W!b8 15.h3 W!a8 16.Wie2
i.a6 17.ll:\cb5 i.b7 18.ll:\c3 = .
via Poyatos, Barcelona 2 0 1 2 , Black
opted for a waiting game with :
15 . . . Wfc7. This is possible, but
practice shows that White does not
defend well in the Hedgehog. It is
better to attack him even if that in­
volves taking risks. The above game
went:
16.b4 �b8 17.a3 l=!fd8 18.�h5
ll:\g6 19 .h3
19.ll:\f3 ! ? ll:\f6 2 0 .�xg5 was a bet­
ter try, ( 2 0 .Wfh6 ll:\g4 =) but Black
has sufficient counterplay after
20 . . . ll:\xe4 2 l.�h6 ll:\xg3 2 2 .hxg3 dS !
19 . . . \!tg7 2 0 .�d1 ll:\deS 2 1 . ll:\b5
i.f6 2 2 .ll:\b3

13 ••• �fd7! 14.J..g3

Or 14.l'!c1 a6 15.i.g3 gS 16.f3

\!th8.

14 ••• g5 !
This plan is positionally sound
with a white bishop on - g3 . Stayed
it on e3, White would have the awk­
ward plan i.d4 followed up by b4,
Black could wait further with
22 . . . ll:\f4, but he decided to throw
more forces against the enemy king:
22 . . . h5 ! ? 23.f3 (White obviously
panicked. 23.W!xh5? loses to 23 . . . g4,

266
 Taimanov with S.tt:lbS

but he should not have made weak­

nesses. 23.�d2 was unclear.) 23 . . .
.ic6 24.tt:l5d4 .ia4 2S.c;t>hl

2S.�d2 .ixb3 26.tt:lxb3 dS ! , plan­

ning . . . tt:lxf3+, is also better for Black.
After the text, 25 . . . g4 ! 2 6.hxg4 (26.
f4 h4! 2 7 ..ih2 g3) 2 6 ... -igS would
have given a decisive attack.

23 �g6
16.f3 �h8 17 . .if2 �g8
.••

Black's pieces are very menac­

ing, but 23 . . . g4 is still ineffective in
view of 24.f4. He had to bring more
power to the attack with . . . ti'h6 or
. . .fS . Liberzon's move is not bad ei­
ther. Apparently Kavalek under­
stood that passive waiting was
doomed and he made a desperate
attempt to free his pieces with :
Black has some prospects for an
24.f4 gxf4 25.tt:lxf4 �h6
attack, but White's position is very
26.�g3 �xg3
solid. The advance . . . g4 is not an
imminent threat - yet. After 18 .b4
26 . . . 2"1f8 ! 27 . .id4 fS would have
.if8 ! 19 . .ie3 �f6, the play would be
kept more tension, but Liberzon
unclear. Instead, Kavalek consider­
simply grabs a pawn.
ably weakens his castling position :

27.hg3 h4 28 .if2 .ixe4

18.h3 �g7
•

29.c5? dxc5 30.bxc5 .ixc5

31 . .ixc5bxc5 32. c!bc4 ti'f5 33. fid2
Another interesting resource
�f6 34.c!bxe5 c!bxe5 35.fid8 +
was 18 . . .�f8 19 .b4 fS when 2 0 .exf5
�g7 36.g3 c!bf3+ 37.�xf3 .ixf3
exfS 2 1 .tt:ld5 is dubious due to 2 1 . . .
38 .id3 fie5 0-1
.ixdS 2 2 . cxd5 l"1xcl 23 .�xcl g4 !
•

24.f4 gxh3 25 .fxe5 hxg2 .

19.b4 h5 2 0 . a3 figS 21.�e3

�h7 (2l . . . g4 2 2 .f4 tt:lg6 23.tt:le2 eS
24.f5 tt:lf4) 2 2 .tt:le2 �cg8 ! (22 . . .
g 4 was premature due t o 23.f4
tt:lf3+ 24.tt:lxf3 gxf3 25.2"1xf3 ti'xe4)
23.�cc3
43. Radulski-lotov
Sunny Beach 10.06.2009
l.e4 c5 2 . c!bf3 e6 3.d4 cxd4
4. c!bxd4 c!bc6 5. c!bb5 d6 6.c4 c!bf6

267
 Part ?

7. �5c3 a6 8.�a3 j.e7 9.J.e2 b6

10.0-0 j.b7 n.f4 o-o 12 .if3 gcs
•

13.j.e3 ges

16 . . • d5 ! 17.exd5

White cannot justify the place­

ment of his queen on f2 by 17.f5 be­
14.�M2 cause he has only one check after
17 . . . .ixa3 18.fxe6 fxe6 19.bxa3 lL!xc4
14J''k l is a more natural move 2 0.�f7+ 'i!th8 with an edge.
since it protects White's most sen­ After the text, Black would be on
sitive point - the c4-pawn. top after 17 . . . .ixa3 ! 18.bxa3 lL!xc4
19.J.xc4 !!xc4 2 0 . dxe6 !!xe6 2 l..id4
14 • • • �a5 15.�f2 !!g6 2 2 .g3 �c8.

lSJ'!acl is already late - 15 . . . d5 17 ,exd5? ! 18.gadl

• •

16.cxd5 J.xa3 17.bxa3 lL!c4 18.�f2

exdS 19.!Udl �d7 2 0 .e5 lL!g4+. 18.lL!xd5 simplifies to a near­
ly equal position although Black
15 •.• �d7!? retains some initiative after 18 . . .
J.xa3 19.bxa3 lL!xc4 2 0 .J.xc4 !!xc4
15 . . . lL!xc4 16.lL!xc4 !!xc4 17.e5 2 1 .lL!xb6 l'!ce4 2 2 . lL!xd7 �xd7! ? .
lL!e4 18.lL!xe4 J.xe4 19.J.xb6 �d7
20 .J.xe4 l'!xe4 is roughly equal, but 18 .ixa3
• • • 19.bxa3 lL!xc4
Iotov decides to maintain the ten­ 20.j.d4
sion.
White hopes to win the dS­
16.j.e2 pawn. 2 0 .J.xc4 dxc4 2 1.J.xb6 �xb6
2 2 . l'!xd7 �aS 23.�d2 .ia8 was rath­
White is already on the defen­ er dull for him.
sive. 16.lL!a4 lL!xc4 17.lL!xc4 l'!xc4
18.lL!xb6 lL!xb6 19 . .ixb6 �a8 gives 20 • • • lL!d6? ! ( 2 0 . . . lL!f8+) 2 1.f5?
Black an initiative - 2 0 . !!fel l'!ec8
21.b3 l'!c2 2 2 .l'!e2 l'!cl+ 23.l'!xcl 2 1..if3 lL!cS 2 2 .J.xc5 bxcS
l'!xcl+ 24.l'!el �c8--+. 23 . .ixd5 .ixdS 24.lL!xd5 lL! e4 25. �b2

268
 Taimanov with S.lt:lbS
kept White in the game. Instead, he
blunders material.
2 1 fucc3 2 2.hc3 c!Lle4 23.Wd4
•••
c!Llxc3 24.Wxc3 fuce2-+ 25J::�d 2
l::!xd2 26.Wxd2 We7 27.l::! e 1 'tYc5+
28.'tYf2 �f8 29.h3 .ic6 30.l::!b 1
.ib5 31.l::!b4 Wxf2 + 32.�xf2 .ic4
33.l::!b 2 �e7 34.�e3 �d6 35.�d4
b5 36.a4 c!Llb8 37.axb5 axb5
38.�e3 c!Llc6 39.�f4 .d4 40.g4
�d5 41.f6 gxf6 0-1
44. Sandu-Potkin
Sautron 27. 1 0.2007
1.e4 c5 2 . c!Llf3 e6 3.d4 cxd4
4. c!Llxd4 c!Llc6 5. c!Llb5 d6 6.c4 c!Llf6
7.c!Ll 1c3 a6 8 . c!Ll a3 .ie7 9 .ie2 0-0
• with full compensation - 17 . . . d3
(9 ... b6 10. 0-0 i.b7 1l.f4 dS) 10.0-0
b6 ll.f4 .ib7 12 .if3 l::! c 8 13 .ie3
• •
l::! e 8 ! 14.We2
We saw in the previous game
that 14.�d2 lt:l aS was awkward for
White so �e2 looks logical. It has
two major drawbacks, however.
The queen falls under the X-ray of
the e8-rook, and the d4-square re­
mains without control. Potkin pin­
points both with his excellent move :
14 ••• e5 ! ? 15.Wf2
White cannot keep the centre
closed since 15.f5 lt:ld4 16. �d3 stum­
bles into 16 . . . b5 ! 17.cxb5 dS+. Alter­
natively, 15.lt:ld5 lt:ld4 16.i.xd4 (16.
�f2 lt:lxf3+ 17.�xf3 i.xdS 18.cxd5
exf4 19.i.xf4 bS) 16 . . . exd4 17.Elfd1
lt:lxd5 18.cxd5 i.f6 is balanced. Black
can play for a win though with 17 . . .
i.f8 ! ? 18.Elxd4 lt:lxdS 19.cxd5 bS
2 0 .Eld2 g6. His dark-squared bish­
op has no opponent.
15 c!Lld4 ! (15 . . . exf4
••• 16.i.xb6
�d7 17.i.e2 gS) 16.l::! a d1
16.i.xd4 exd4 17.lt:ld5 (17.�xd4?
dS) 17 . . . lt:lxd5 18.cxd5 i.f6 is equal,
but Black can also sacrifice a pawn
18.Elfd1 bS 19.cxb5 lt:lxdS 20.exd5
axbS 2 l . Elxd3 �aS 2 2 .lt:lc2 .idS
2 3 .Ela3 �b6 24.lt:ld4 .if6 25J'l:d3
:!! c 4�.
16 c!Llxf3+
••• 17.Wxf3 exf4
18 .ixf4
•
18 •.• 'tYc7
Black has won the opening bat­
tle. This open Sicilian would have
269
 Part 7
been satisfactory for him even with
White's pawn on c2 and a knight
on d4 instead of a3. Now 18 .. .'�d7 ! ,
heading for e6, would b e i n his fa­
vour (19.l!Jd5 �d8 2 0 .!!fe l bS), but
the text is also good enough.
19.c.th1 .if8 (19 .. .'�'d7! ) 2 0 . l!J d5
l!Jxd5 21.exd5 'e;Yd7 22 .'e;Yd3 g6
23.b3
23 •.• f5?
It seems that Potkin could not
find an active plan. Had Black a
knight to occupy eS, this move .
would have been normal, but here
it only weakens the e6-square. He
should have displayed activity on
the queenside with 23 ... b5 and then
. . . !kS, targeting the c4-pawn.
24.�fe1 'e;Yg7? ! 25. l!Jc2 (25.
'e;Yg3 ! ) 25 . . . 'e;Yb2? (2S . . . b5 =)
26.l!Jd4? (26.�e3 ! + -) �xe1+
27.�xe1 �e8 28.�g1
It was more natural to trade
rooks in order to invade Black's
camp through the e-file - 28.�g3
l'!xel+ 29.hel �cl 30 .�e2. After
the text, Black is OK, e.g. 28 . . . l'!e4.
He only must control the e-file. The
270
rest of the game is full of mistakes
and is irrelevant to the opening.
28 . . .'llY f2 29 .ig3
• �e3?
30.hf2 �xd3 3 1 . l!Jf3? (3U3el)
31 . . . b5 !+ 3 2 . �c1 .ih6 33.�b1 �c3
34. c.tg1 c.tf7 35. c.tfl bxc4 36.bxc4
.ic8 37 . .id4? �xc4+ 38.�b6 .if8
39.c.te2 .ie7 40.c.td3 �a4 41 .ib2 •
�xa2 42,gbs .id7 43.�b7 .ic8
44.�b8 f4 45.c.tc4 .ig4 46.l!Jd4
gas 47.h3 �c5 + 48. c.td3 .if5+
49.c.te2 .ie4 50.tbe6 �xd5 0-1
45. Brodsky-Kruszynski
Pardubice 1 993
1.e4 c5 2 . l!Jf3 e6 3.d4 cxd4
4.l!Jxd4 l!Jc6 l!Jb5 d6 6.c4 l!Jf6
7.l!J 1c3 a6 8 . l!Ja3 b6 9 .ie2 .ib7
•
10.0-0 l!Jb8 ll.f3 l!Jbd7 12 .if4 •
l!Je5 13.'e;Yd2 .ie7 14.�fd1 'e;Yc7
15.�acl 0-0 16.c.th1 �ac8 17.�g3
White opts for f4, but 18 .'e;Yel in­
tending �gl, �e3 , l'!cl-c2-d2 and fi­
nally b4 is also promising .
17 . . . �fd8 18.f4 tt:lg6 19 .if3
•
19 . . .'llY b 8 20 .�e2 l!Jd7 21.tt:lab1
 Taimanov with S.lt:JbS

J.f6 2 2 . tt:ld2 i.c6 23.b4 a5 24.a3 L u p u lescu-Perunovic

'ea8 25.lt:Jdbl J.e7 26.i.g4 tt:lgf8 Kavala 2004
27.J.f3 tt:lg6 28.f;Ye3 Y;Yb8 29.J.e2
'ea8 30.i.f3 'eb8 3 1.J.e2 f;Ya8
3 2 .l:�c2 'eb8 33.ga2 tt:lf6 34.J.f3
axb4 35.axb4 b5?

Black is the first to lose patience.

36.cxb5 hb5 37.e5 tt:ld5

38.tt:lxd5 exd5 39.e6 J.c4
40.exf7+ rJlxf7 4L ga7 rJJ f8
42.tt:ld2 d4 43.�xd4 .if6 44.�gl 24J''&b l gd4 25.b5 aS 2 6.c5 ! �f8?
'exb4 45.tt:lxc4 gxc4 46 .id5 • 27.cxb6 .ixb6 28 .lt:Jc4 .icS 29 .'ecl
gd4 47.gf7+ rJle8 48.gxf6 gxd5 �d8 3 0 .b6 lt:JhS 31.lt:Jxa5 gd2
49.gel + tt:le5 50.gf5 g6 51.gf6 1-0 32 .lt:Je2 1-0

Let's return now to our main

game. White can obtain the better
46. Daly-Treg u bov
chances even without b4. That is
Cappelle Ia Grande 2000
possible thanks to the breakthrough
24.c5 ! bxcS 2S.lt:Jc4 ggd8 26.�xc5t.
l.e4 c5 2 . tt:lf3 e6 3.d4 cxd4
In the game, White missed this op­
4. tt:lxd4 tt:lc6 5. tt:lb5 d6 6.c4 tt:lf6
portunity and played too tentative­
7.tt:l lc3 a6 8 . tt:l a3 b6 9 .ie2 J.b7
ly to lose in the end :
•

10. 0-0 tt:lb8 ll.f3 tt:lbd7 12 .ie3 •

J.e7 13.'ed2 'ec7 14.gfdl 0-0 24.tt:lc2 ggd8 25.b4 gd2

15.gacl !i!ac8 16 ..ifl 'eb8 17.'ef2 26.gd2 J.d8 !
J.d8 18.rJlhl !i!e8 19.1i!c2 .ic7
2 0.'egl rJlh8 21. 1i!cd2 !i!g8 22 .if4
•
Black's first step is correct. The
tt:le5 23.he5 ! dxe5 bishop should control the squares
b4-c5.

27.a3 'ec7 28.tt:lal ! ? g5? !

It is useless to discuss the pros

Let's compare this typical posi­
tion to another model game :
and cons of this plan. If so many
players like it, apparently it has
some value, but to me it is only pro­
ducing weaknesses around Black's
king. The game will be decided in
the centre so 28 . . . lt:Jd7 29.lt:Jb3 .igS
3Q.gdl fS� looks more adequate.

271
 Part S
29.�b3 g4 30.tfe3 J.e7 31.J.d3?
White is lingering again. 31.lLla4
gxf3 32.gxf3 lLld7 33.c5 was called
for - 33 . . . bxc5 34.lLl axc5 lLlxcS
35.lLlxc5;!;.
31 ggs 32.lLla4 lLld7 33.gc2
•••
�c6 34.lLlc3 gxf3 35.�xf3
f5+± 36.�fl f4 37.lLld2 (37.c5 !
bxcS 38.lLla5) 37 a5+ 38.gb2
.••
lLlf6 39.lLlf3 �g4 40.�b1 axb4
41.axb4 �e8 42.�d1 tfh5 43.c5
bxc5 44.bxc5 hc5 45.h3 J.c8
46.�c2 lLle3 47.�xc5 �xd1
48.gbs J.d7 49.gb7 gds so.ga7
.ie8 51.J.c4 J.f7 52 .ga2 'i!lg7
53.he6 lLlf2+ 54.gxf2 he6 0-1
4 7. Rabrenovic-Delchev
Bijelo Polje 2005
1.e4 c5 2 . lLlf3 e6 3.d4 cxd4
4. toxd4 �c6 5. lLlb5 d6 6.c4 lLlf6
7. lLl1c3 a6 8.lLla3 b6 9.J.e2 Ab7
10. 0-0 J.e7 ll . .ie3 0-0 12 .�d2
�e5 13.f3
I knew that now 13 . . . d5 was
roughly equal, but I decided to
maintain the tension.
272
13 ges 14.gfd1 gcs 15.tfd4
• • •
�ed7 16.tfd2 tfc7 17.gac1 tfb8
18.lLlab1 gcd8 19 .!fl h5!?
•
This flank assault is usually bet­
ter than g5. It targets the g3 square.
20.tff2 d5!
Eventually, I discovered that
beating about the bush was not get­
ting me any closer to the point.
21.exd5 exd5 2 2 . �xd5 lLlxdS
23.cxd5 Ad6 24.hb6?
The abrupt change of the game
course clearly confused my oppo­
nent and he commits a decisive
mistake. 24.g3 h4 ! 25.'i!lg2 hxg3
26.hxg3 is more resilient although
it would be difficult to defend after
26 .. Jlxe3 ! ? 2 7.�xe3 hg3 threaten­
ing . . . lLlf6-h5, or 26 . . . lLlf6 27.lLlc3 bS.
24 hb2
• . . 25.'i!lh1 Ag3
26.tfd4 lLlxb6 27.�xb6 gd6
27 . . . .if4 28.!k2 �eS was winning
at once.
28.tfb4? �d8 29. lLlc3 f5
30.gd4 tfg5 0-1
 Pa rt S

Ra re Lines on
5th and 6th move

273
Pa rt 8. Rare Li nes on 5th a n d 6th Move

Main Ideas

l.e4 c5 2 . �f3 e6 3.d4 cxd4 A. 5.�xc6

4.�xd4 �c6
Great ideas are simple, they say.
This one might have been too, but
when put to the practical test, it just
does not work. Black has different
ways of defending against the at­
tack. When White's opening initia­
tive begins to fade, it becomes clear
that Black's powerful pawn centre
is a weighty factor in the ensuing
struggle.

5 . . . bxc6 6 . .id3 �f6 7.0-0

If you were a novice in chess,
your first thought in this position
might be : 'OK, how to mate the guy
quick and neat? �c4 is apparently
pointless, so I move the bishop to
d3, the queen to hS or g4 and throw
all the rest against Black's king.' If
you were also smart enough, you
would notice that the knight on d4
was hanging and anyway it falls out
of the whole design. At this point, 7 . . . d5
you already know what to do. Of
course wasting precious tempi on When you choose this move, you
retreating the clumsy knight is out should know how to survive against
of question, so you part happily a direct kingside attack. Black has
with it by: at least two other setups which are

274
 Rare Lines on 5th and 6th Move

probably not worse than the main 8.�d2 ti'c7! 9.f4 .la6
line. One of them is connected with
... e6-e5, the other one - with ... d6.
The latter is calm and safe, but not
too popular. Black embraces the phi­
losophy of the hedgehog structures.
Hidden behind the pawn wall along
the sixth rank, he is awaiting for the
best moment to determine his plan.
It could be an advance of any one of
the three central pawns.
After castling, Black will push
Our defence after 7 . . . d5 is based the a-pawn to a4, see game 48
on two main principles : Hou,Yifan-Dzhumaev, Kuala
1. We destroy the most danger­ Lumpur 2 0 1 0 .
ous enemy piece which is the d3-
bishop. That could be done in some More practical examples :
variations by . . . i.c8-a6, but we of­
Martin-lllescas
ten have a better solution - the ma­
Dos Hermanas 2004
noeuvre . . . tt.:Jf6-d7-c5xd3 :
Burger-Ai burt
Philadelphia 1989

The attack a5-a4xb3 creates a

weakness on b3 :
18 . . . axb3 19.axb3 l':ia2+.
Position after ll . . . tt.:Jc5.
Analysis
Should your opponent allow it,
take the bishop and push a- and c­
pawns. That will ensure excellent
counterplay.
2. If White attempted to secure
his bishop with c3, we either take
control of the centre with . . .'�c7,
. . . i.d6, . . . e6-e5, or trade our main
enemy with . . . i.a6.

275
 Part S
If White tries to stop the march
of the a-pawn with a knight, Black
should counter that by . . . ll::l d 7-b6.
B. 5.c!Llc3 V!Jc7 6.c!Llxc6 bxc6
7 .id3 c!Llf6
•
Everything said about the move
S.ll::l xc6 is valid here, too. Having
committed his knight to c3, White
has not a wide choice.
There is one important tactical
nuance which you should take into
account. The queen on c7 could be
attacked by ll::l bS ! That's why you
should refrain from dS before de­
veloping the bishop to e7:
In the diagram position,
8.V!Ie2 is notjust a transposition
(compared to 8.0-0 i.e7) :
Black has a number of interesting
options here except of 8 . . . d5? in view
of 9.exd5 ! cxdS lO.ll::l b S! V!Jb8 ll.g3
276
e5 12.i.f4 !
Perhaps 8 . . . i.e7 ! ? is best, since
9.e5 ll::l d 5 1 0 . 0-0 0-0 is balanced.
C. 5.c!Llc3 Wc7 6.Wd3
This move has occurred in only
one game, Vallejo Pons-Adhiban,
Caleta 2 014. I propose:
6 c!Llf6 7.c!Lldb5 Wb8 8.Wg3 d6
•••
9 .ie3 (9 . .if4 ll::l h S) 9 a6 10.c!Lld4
• •••
e5 l l . c!Llb3 .ie6 with a normal Sici­
lian position.
D . 6.c!Lldb5 Wb8 7 .ie3
•
There is some mystification
about this system. It has the repu­
tation of "very" interesting, proba­
bly even better for White, but it is
seldom seen. I suppose that com­
mon sense reminds us that a queen,
 Rare Lines on 5th and 6th Move
even together with a pawn, is usual­
ly inferior to 3 pieces.
7 . . . a6 ! 8 . .ib6 axb5 9 .�xb5
.ib4 10.c3 .ia5 ll.�c7 Wxc7!
12 . .ixc7 .ixc7 13.Wg4 .ie5 ! ?
Protecting indirectly g 7 and pro­
viding Black with time to castle.
Most importantly, we avoid weak­
ening f6, which is a major downside
of 13 . . . g6.
14.f4 �f6 15.Wf3 .ic7
Black has full compensation for
the queen. His only concern is how
to activate the light-squared bish­
op. That could be achieved by ei­
ther . . . b6, . . . .ib7 (beware the X-ray
of the queen from f3 ! ) or . . . .id7 af­
ter advancing the d-pawn.
277
Pa rt 8. Ra re Li nes on 5th a n d 6th M ove
Step by Step
l.e4 c5 2.tt:\f3 e6 3.d4 cxd4
4. tt:\xd4 tt:\c6
A. S.lt:lxc6 ;
B. S.lt:lc3 V!ic7 6.lt:lxc6;
C. S.lt:lc3 V!ic7 6 .lt:ldb5 ;
D. S.lt:lc3 V!ic7 6.V!ff. d3
For S.c4 lt:lf6 see Part 11/Cl
A. 5. tt:\xc6 bxc6 6 .id3 •
6.c4 is not flexible. It offers
Black an ample choice. Perhaps
simplest is 6 . . . lt:lf6 7 . .id3 dS= , but
6 . . .-icS 7 . .id3 eS 8 . 0-0 lt:lf6 is also a
good option. Black has clear play on
the dark squares.
6 . . . tt:\f6 7.0-0
278
7 .•• d5
Sometimes Black chooses plans
with e6-e5, but this set-up does not
provide many possibilities for ad­
vancing the central pawns. On the
contrary, the c6-d5-e6 wedge is
quite mobile. Black could play c6-
c5, d5-d4 or undermine White's eS­
pawn with . . . f6.
If you do not like the type of po­
sitions arising in the main line, I
propose you to consider:
7 . . . d6
The play is calmer and Black is
running less risks of getting mated
than after 7 . . . d5. I have not chosen
it for the main line mostly because it
is significantly less popular.
 Rare Lines on 5th and 6th Move
Game 50 Diez del Corral­
Korchnoi, Palma de Mallorca 1968,
is a typical example:
8.b3 i.e7 9.i.b2 e5 (White was
threatening 10 .e5;t) to.l2Jd2 0-0
ll.l2Jc4 i.e6 12 .�e2 l2J d7. See the rest
in the "Complete Games" section.
If White plays c4, Black should
follow the plan with . . . e5, . . . i.e6 .
The other common set-up with
. . . l2Jf6-d7-e5-c6 is a bit passive :
8.c4 i.e7 9.l2Jc3 0-0 10 .�e2 e5
ll.h3 i.e6 1 2 .f4 exf4 13.i.xf4 l2Jd7;
8.c4 l2Jd7 9.l2Jc3 i.e7 10 .�c2 l2Je5
ll.i.e2 c5 12 .i.e3 0-0 13J!ad1 �b6
14.f4 l2Jc6, Meij ers-Ksieski, Leu­
tersdorf 2 0 0 1 .
The game Kholmov-Korch­
noi, Riga 1970, shows the flexibili-
ty of 7 . . . d6. Black can always return
to the . . . d5-plan in proper circum-
stances:
8.�e2 i.e7 9.l2Jd2 l2Jd7 1 0 .f4 l2Jc5
ll.l2Jf3 0-0 12 .i.e3 d5= .
Perhaps the most testing plan is
8.f4 i.e7
9.cit>h1
White's best set-up was shown
in the game Lein-Tal, Kiev 1964:
9.l2Jd2 0-0 10.I!th1 e5 11.fxe5 dxe5
12.l2Jc4 �c7 13.�f3 i.e6 14.i.d2 l2Jd7
15.i.c3 f6 16.l2Je3 l'l:fe8 17.i.c4;t. Note
that the bishop goes to c3 in order
to bind Black with the defence of
the e5-pawn and then White trades
the light-squared bishops. Howev­
er, 9.l2Jd2 is inaccurate due to 9 . . .
e5 ! 10 .fxe5 ( o r 10.l2Jc4 d5 ! ll.exd5
e4 ; 10.l2Jf3 0-0 ! and White can­
not achieve the set-up of the game
Lein-Tal.) 10 . . . l2Jg4 ll.l2Jf3 �b6
1 2 . \!thl l2Jf2 13.l'l:f2 �f2 14.exd6
i.xd6 15.i.xb5 cxb5 16.�xd6 �f1
17.l2Jg1 �c4 18 .�e5 i.e6oo . After the
text, Black has to prevent 10 .e5. If
he plays 9 .. :e5 himself, then 10 .fxe5
dxe5 ll.l2Jd2 i.e6 1 2 .�f3 will prob­
ably transpose to the game Lein­
Tal which is not too appealing. Re­
mains:
9 ... �c7 10.l2Jd2 0-0 11.�e1 l2Jd7
with unclear play.
Let's return to 7 . . . d5 :
White often plays 8.e5, nurtur­
ing hopes for a kingside attack. This
is a lost cause, since Black can easily
eliminate the bishop on d3. Without
it, any direct attack should fail - 8 . . .
279
 Part S
tLld7 9.�e2 (9.f4 limits the scope of
the c1-bishop. We'll trade the oth­
er one with 9 . . . h.6 or 9 . . . tLlcS, fol­
lowed by . . .JJ.e7, . . . g6 and .. .fS) 9 . . .
JJ.e7 10.tLld2 0 - 0 ll.f4 (ll.tLlf3 ! ? is
more realistic, when Black follows
the same design - with ll . . . tLlcS)
1l.. .tLlcS 12. c;!;>h1 aS+ 13.tLlf3 ttJxd3
14.cxd3 c5 1S.JJ.e3 JJ.a6 16J:Ucl �b6
17J:k2 a4+, Burger-Alburt, Philadel­
phia 1989.
8.tLlc3 is possible, but inconsis­
tent. It leads to positions that com­
monly arise after another move or­
der - l.e4 cS 2.tLlf3 e6 3.d4 cxd4
4.tLld4 tLlf6 S.tLlc3 ttJc6 6.ttJxc6. The
knight is misplaced on c3, because
it has no prospects there and should
be redeployed, probably to a4. This
plan is harmless for Black. He
could oppose it by the manoeuvre
. . . tLlf6-d7-b6 which is Black's univer­
sal retort to tLla4.
Let's see some typical examples :
8.tLlc3 JJ.e7
9.eS tLld7 10 .�g4 (10.�e1 0-0
ll.f4 fS) 10 ... ttJxeS 11.�xg7 �g8
12 .�h6 �b8 13.tLle2 �b4.
9.b3 0-0 10 .�e2 tLld7 ll.tLla4
tLlb6=, Hector-Lautier, Malmo
1998.
280
9.�e1 ttJd7 ! ? (avoiding �gS or
the variation 9 . . . 0-0 10.if4 tLld7
ll.exdS cxdS 1 2 -l!J bS;t) 10.if4 (10.
�hS tLlcS) 10 ... d4
In normal circumstances, this
set-up is not too flexible, but here it
is with tempo.
ll.tLla4 eS 12 .ig3 hS (12 . . . 0-0)
13.h3 h4 14.ih2 0-0 1S.c3 dc3
16.tLlc3 icS 17.�fl id4 18 .�c2 tLlcS
19 .ic4= Delgado-Jobava, Havana
200S.
After 8.tLld2, we reach a posi­
tion which might be familiar to the
French-playing people - l.e4 e6 2 .d4
dS 3.tLld2 cS 4.tLlgf3 cxd4 ·s .ttJxd4
tLlc6 6.tLlxc6 bxc6 7.id3 tLlf6 8.0-0.
White is trying to be too clever and
keep all his options open. However,
it has its cost - he cannot play eS
without the help of his f-pawn after:
8 • • • �c7!
8 . . . JJ.e7 is less principled. 9 .�e2
Alternatively, White can choose
to maintain the central tension
with b3 and c4, where Black push­
es his paWns to a4 (and even to a3
in some situations), and cS. Anoth­
er possible setup is based on b3 and
a4 which allows Black to trade the
light-squared bishops through a6.
These plans are less dangerous,
because White plays on the wing
where Black is stronger - 9.�e2
0-0 10 .b3 aS ll.a4 �e8 12 .ib2 �b6
13.�ad1 ia6 14.�fe1 hd3 1S.�xd3
�c7 16.eS tLld7 17.�e3 = , Kofidis­
Semkov, Heraklio 1993.
9.�e1 0-0
 Rare Lines on 5th and 6th Move
10.c3 !
lO .eS li:ld7 ll.c3 fS ! i s excellent
for Black - 12.exf6 h:f6 ! ? 13.�h5
g6 14.h:g6 hxg6 15.�xg6+ !ig7
16.�xe6+ �f7 17.�xc6 �b8 18.�xd5
!ib7 19.�d3 �b6 with a strong at­
tack, Senff-Nielsen, Germany 2 008.
10 ... li:ld7 ll.!ic2 ! with unclear
play.
9.f4
9.c4 abandons the centre - 9 . . .
!id6 10 .h3 0 - 0 1l.�e2 .ib7 1 2 .b3
.ih2 + 13.'itlh1 .ie5 14.�b1 c5 15.exd5
exdS 16.ltlf3 , Kofidis-Ehlvest, Ko­
motini 199 2 , 16 . . . .id6 ! with initia­
tive.
9 • .. !ia6 10.e5 lbd7 11.lbf3 .ie7
White does not have any ac­
tive plan, see game 48 Hou,Yifan­
Dzhumaev, Kuala Lumpur 2 0 1 0 .
B . 5.lbc3 �c7 6 .lbxc6
This exchange has a bad repu­
tation, because the knight on c3
has not clear prospects. Sometimes
White even returns it back to b1
in order to redirect it to d2-f3, or
moves it to a4 in order to enable c4.
6 ••• bxc6
6 . . . dxc6 has its adepts, but I do
not like such symmetric pawn for­
mations in the Sicilian. They are dif­
ficult to win.
7 .id3 lbf6 8. 0-0
•
Attention! 8 .�e2 is not just a
transposition :
281
 Part 8
Black has a number of inter­
esting options here except of
8 ... d5?
This move has never been pun­
ished in practice, but it is inferior to
8 . . . ie7, 8 . . . id6, 8 . . . d6 or 8 . . . e5. The
reason is :
9.exd5 ! cxdS (or 9 . . . tt:Jxd5 10.0-0
id6 11.g3 tt'lc3 12 .bxc3 0-0 13J:'1dl±)
IO .tt:lbS ! �b8 ll.g3 eS 12.if4 where
I have not a good advice for Black:
12 . . . tt:Je4 13 .he4 dxe4 14.�c4 exf4
15.tt'lc7 @e7 16.0-0-0+- or 12 . . . tt:Jd7
13.0-0 (13.0-0-0±) 13 . . . ie7 14J:'1fe1t.
Perhaps 8 . . . ie7 ! ? is best, since
9.e5 tt:ldS 10. 0-0 0-0 is good for
Black.
More challenging is 8 . . . id6 ! ? us­
ing the fact that White has not cas­
tled. This continuation has passed
the test in a couple of games.
8 • • • J.e7
There is not a single reason to
play d7-d5 unless White threatens
eS. 8 . . . d5 9.ed5 is not dangerous,
but why to hand the opponent ad­
ditional options connected with the
exchange on dS. Still, in Stenzel-So­
fia Polgar, Kona 1998, Black com­
pleted development after 9 . . . . exd5
(9 . . . cxd5 10 .tt'lb5 with initiative)
10J�e1 ie7 11.�f3, when 11 . . . 0-0=
would have been safe enough.
9.Y!Ye2
Black should be only happy to
face a head-on "attack" of the type
of 9.f4 dS IO .@h1 0-0 ll.eS tt:ld7
12.:!=1£3. He could parry it by 12 . . .f5
282
13.exf6 tt:lxf6 or by the typical 12 . . .
tt:lcS, since 13 .hh7 @xh7 14.:!=1h3
@g8 15.�h5 fS is not really a threat.
Note that I do not recommend to
trade light-squared bishops by pre­
paring ia6. 1t is better to kill White's
bishop by the knight ( . . . tt'lf6-d7-c5),
when ib7 will rule along the main
diagonal after . . . c6-c5 and . . . d5-d4.
Black's most flexible answer to
9.:!=le1 is 9 . . . d6. Then 10.f4 0-0 ll.eS
tt:ldS 1 2 . tt'le4 dxeS 13 .fxe5, Kaka­
badze-Schlosser, Baden-Baden
1993 , is better for Black due to 13 . . .
:!=ld8 ! .
9 . . . d5
9 . . d6 1 0 .f4 tt:Jd7 offers Black
.
nice statistics in the database.
10.b3
Black can meet I O.f4 0-0 ll.@h1
with the typical setup ll . . . :!=le8
12 .i.d2 i.b7 13.:!=lae1 tt:Jd7 14.e5 (14.
tt:Ja4 tt'lb6) 14 ... tt:Jc5.
10 0-0 ll.i.b2 l;e8 12.l;ael
••.
J.b7 13.e5 tt:Jd7
I prefer Black's chances in this
typical position, in view of his mo-
 Rare Lines on 5th and 6th Move
bile pawn centre. See game 49
Martin Perez-Illescas Cordoba, Dos
Hermanas 2 0 04.
C. 5.�c3 'tYc7 6.'tYd3
This move has occurred in only
one game, Vallejo Pons-Adhiban,
Caleta 2 0 14, but when a 2700 play­
er employs it, we should pay atten­
tion. There followed 6 . . . a6 7.lt:lxc6
�xc6 8.�g3 bS 9.a3 i.b7 10 .i.d3
lt:lf6 11.0-0 with a solid edge. Where
went Black so wrong? ! First of all,
capturing by queen on c6 is dubious.
In the event of 7 ... bxc6 8.�g3, White
has a clear extra tempo (spent on
. . . a6) in comparison to line B. That
does not automatically promise him
an edge. In fact, this position of­
ten arises following another move
order. It is even considered to be
roughly equal after 8 . . .�xg3 9.hxg3
E:b8. We should not avoid such a
structure at all costs, but I propose
to get a better version of it by:
Now 7.i.g5 a6 (7 . . . lt:lxd4 8.�xd4
i.cS 9.\!;!fd2 a6 1 0 . 0-0-0 bS ll.eS!
lt:lg4 1 2 .lt:l e4 \!;!!x eS 13.E:e1 is really
dangerous for Black.) 8 .lt:lxc6 bxc6
is a blank spot in theory, but gS is
hardly the best place for the enemy
bishop. For instance, White can­
not trade queens with 9.\!;!fg3 due to
9 . . . i.d6 - the gS-bishop "protects"
the g7-pawn. Of course, White can
castle first, but 8 .0-0-0 i.e7 9.�g3
�xg3 10.hxg3 lt:lxd4 ll.E:xd4 i.cS
1 2 . E:d2 lt:lg4 13.f3 f6 14.i.f4 lt:leS is
balanced. Perhaps White should
try:
7.�db5 ti'b8 8.'tYg3 d6
We know that the g3-queen is an
obstacle to a kingside pawn storm
with g4-g5 :
9.�e3 (9.i.f4 lt:lhS) 9 • • • a6
10.lt:ld4 e5 ll. �b3
11.lt:\xc6 bxc6 1 2 . 0-0-0 i.e6
13 .i.e2 \!;!!b4 14.a3 \!;!fb7 1S.@b1 i.e7
16.f4 E:b8 17.i.cl 0-0 18.f5 i.d7.
ll �e6 12.0-0-0 (12 .i.e2 �b4
• • •
13.E:c1 dS) 12 �e7 13.�e2 0-0
••.
14.�h6 �e8 15.�d5 @h8 = .
283
 Part S
D. 6.�db5 Vfb8
7 .ie3
•
White can try to substantiate his
sixth move by 7.a4, making room
for the knight on a3. This varia­
tion is seldom seen, so the arising
positions are unexplored. I think
that Black's safest way is to restrict
White's knight with 7 . . . ll'lf6 8 . .id3
(note the trap 8 . .ig5 a6 9.ll'la3?
ll'lxe4 10.ll'lxe4 Yfe5+) 8 ... a6 9.ll'la3
d5 10.0-0 .ie7 11.exd5 exd5= .
�. .ic5 i s too risky: 9 . 0-0 h5
10.It>h1 a6 ll.ll'la3 ll'lg4 12 .f4+±.
7. . . a6
7 ... ll'lf6 8.f4 d6 is also playable,
but I do not see any reason to avoid
the text move.
8 .ib6
• l'!xf4 19 ..id3 l'!f2 ! +.
It is shameful to retreat back with
8.ll'ld4, but this might be a realistic
approach. Then Black plays normal
Taimanov, enjoying an extra "half
tempo". If the queen returns to c7,
we'll have the usual variations, but
284
Black should try to benefit from the
tempo with 8 . . . ll'lf6.
8 . . . axb5 9 . �xb5 .ib4 10.c3
.ia5 11.�c7+ Vfxc7! 12 . .ixc7 .ixc7
13.Vfg4
13 ••. .ie5 ! ?
This move casts a shadow of
doubt on the whole White's set-up.
13 ... 1t>f8 hampers the normal
development of Black's pieces, but
it also has enough fans. The game
Kornev-J. Geller, Moscow 2007
saw further 14 . .id3 ll'lf6 15.�e2 h5
16.0-0 g5 17.a4 b6 18 .b4 .ib7 with a
lively game.
14.f4 �f6 15.�f3
15.�xg7? fails to 15 . . J%g8 16.'\1�'h6
l'!g6 (16 ... ll'lg4) 17.�h4 l'!g4 18.�h6
15 . . . .ic7 16 . .id3
16.e5 opens up the main dia­
gonal in Black's favour: 16 . . . ll'ld5
17 . .ic4 ll'lce7 (18 . .id3 0-0 19.0-0 b5
2 0 . .ixb5 l'!b8 2 1.a4 .ib6+ 2 2 .lt>h1
 Rare Lines on 5th and 6th Move
i.b7t ; 18.i.b3 0-0 19.0-0 b5 2 0 .i.c2
i.b'Too) 18.g3 0-0 19.0-0-0, Savicevic­
Rajkovic Srebrno Jezero 2 013, 19 . . .
E:a5 ! intending . . . b5, . . . i.b7oo.
16.i.c4 provokes 16 . . . d5 ! , which
is good enough - 17.exd5 exd5
18 .�e3 + ! i.e6 19.f5 dxc4oo. However,
16 . . . 0-0 17.e5 ll:\e8 18.0-0 f6 19.�h5
g6 2 0.�h6 ll:\g7 was also pleasant
for Black in Alcazar Jimenez-Antoli
Royo, Mondariz 2 0 04.
16 • • • e5!
In the current situation 16 ... b6
runs into 17.e5 ll:\d5 18.i.e4 while
16 . . . d5 17.e5 ll:\d7 18.0-0 0-0 19.a4
is unclear.
17.f5 0-0 18 .g4? !
White can prevent d7-d5 by
18.0-0 E:d8 19.c4?, but then 19 . . .
i.b6+ 2 0 . @h1 i.d4 gives Black a to­
tal domination on the dark squares.
18 . . . d5 19 .g5 dxe4 20 . .ixe4
�xe4 2 1.ftxe4 g6+.
The bishop pair enters the play
with decisive effect. White certainly
can improve some variations in this
analysis, but I like Black's position.
285
Pa rt 8. Rare Li nes on 5th and 6th M ove

Com plete G a m es

12 . . . .b:d3 13.cxd3 0-0

48. Hou,Yifan-Dzhumaev
14.J.b2?!
Lumpur 11.04.2010

Hou is still hoping for - an at­

l.e4 c5 2.ttlf3 e6 3.d4 cxd4
tack with f4-f5. To be sure, 14.�e3
4. ttlxd4 ttlf6 5 .id3 ttlc6 6. ttlxc6
cS lSJ'kl \1t/b7 followed up by . . . a5-
•

bxc6 7.0-0 d5 8.ttld2 Y!lc7 9.f4

a4 is pleasant for Black, but at least
the bishop is not on the semi-open
b-file. It is instructive to watch how
helpless White looks in this struc­
ture.

14 . . . c5 15.'i!lhl a5 16.Y!le2 '&b6

17. tbd2 a4

Black can allow c2-c4 - 9 . . . �e7

lO .eS ltJd7 ll.c4 aS 1 2 . 'i!lhl g6 13 .b3
a4 14.�b2 0-0 lSJ:'kl '&b6+, Yap­
Andersson, Rome 1985, but the text
is perhaps more accurate.

9 . . . .ia6 10.e5 tbd7 11.tbf3 .ie7

12 .b3
White will have to play this move
anyway so 12 .�e3 '&b7 would not
change the character of the game.
Besides, Black may trade the sec­
ond bishop, too, with 12 . . . �c5.
White has made all the "pro­
grammed" moves . . . and has land­
ed in a gloom position. The follow­
ing display of activity is just a des­
peration. Dzhumaev does not give
his very talented opponent any tac­
tical chances.

286
 Rare Lines on 5th and 6th Move
18.f5 exf5 19,gxf5 a3 ! 20 . .icl
'We6 ! 21.gh5 g6 2 2 . gh3 'Wxe5.
The rest is agony. 23.'Wxe5 ti)xe5
24.c!t:lfl .if6 25,gbl gfe8 26 .id2
•
ti)g4 27,gf3 .id4 28 .h4 ge2
29.h4 gbs 3o .bs .ih2 3t.d4
c4 32 .ib4 ti)f2+ 33.<.t>h2 .ixd4
• 10.b3
34,gxa3 .ie5 + 35. <.t>gl ti)d3
36 .id2
• .id4+ 37.<.t>h2 ge4
38. <.t>g3 .if2 + 39.<.t>f3 .ixh4
40.g3 .ie7 4l, ga7 .ic5 42,ga6
ti)e5+ 43. <.t>g2 ge2 + 44.<.t>h3
tt)f3 45 ..ib4 ti)gl+ 46. <.t>h4 ge4+
47.g4 .if2 + 48. ti)g3 h5 49,gxgl
gxg4+ 50. <.t>h3 .ixgl 51.a4 ges
52 .id6 c3 53,gc6 gc4 54,gxc4
•
dxc4 55.a5 c2 56 .if4 .ie3 0-1
• 10 . . . 0-0 n . .ib2 ges 12.gael
49. Martin Perez-lllescas
Dos Hermanas 2004
l.e4 c5 2 . ti)f3 ti)c6 3.d4 cd4
4.ti)d4 e6 5.ti)c3 'Wc7 6.ti)c6 bc6
7 .id3 ti)f6 8 .'We2 d5?
•
Black's last move is an instruc­
tive mistake. All other reason­
able continuations are better, for
instance 8 . . . .ie7! ? 9.e5 (or 9 . 0-0
transposing to the game) 9 ... tt:ld5
10.0-0 0-0; 8 ... .id6 ! ? using that
White cannot answer 9.f4.
9. 0-0? 9.exd5 ! cxdS lO .tt:lbS !
�b8 ll.g3 eS 12 . .if4i. 9 . . . .ie7
This set-up is sluggish and gives
Black clear counterplay with a5-a4.
10.f4 0-0 n . <.t>hl )"!e8 12 . .id2 .ib7
13.)"!ael is more natural, although
the same plan as in the game works
fine : 13 . . . tt:ld7 14.e5 (14.tt:la4 ti)b6)
14 . . . ti)c5.
.ib7 13.e5 ti)d7 14.ti)bl? !
White is playing as if time in
chess were for nothing. b3 is com­
monly connected with 14.tt:la4,
when Black responds with 14 ... tt:l b6!
14 . . . ti)c5 15.ti)d2 ti)xd3
16.'Wxd3 (16.cxd3 aS+) 16 . . . a5
17.c4 a4 18,ge3 axb3 19.axb3
ga2+
Black's game is developing by it­
self. Mundane play led White to a
hopeless position thus early in the
battle.
2o .gh3 h6 2 t .ic3 gds 2 2 . �t
•
gda8 23.ti)f3 c5 24 .id2 dxc4-
•
+ 25 .ih6 gal 26.�c2 cxb3
•
27.�xb3 gxfl 28.<.t>xfl gxh6
29,gxh6 ga1 30. <.t>e2 .ia6 0-1
287
 Part S

�c5 15.gfel f6
50. Diez del Corrai-Korchnoi
Palma de Mallorca 1 968
By bolstering eS, Black prepares
dS.
l.e4 c5 2.�£3 �c6 3.d4 cxd4
16.�a5 Wc7 17.i.c4 i.c4
4. �xd4 e6 5. �xc6 bxc6 6 . .td3
tilf6 7.0-0 d6
17 . . .l'!ad8 ! 18 . .be6 lt:lxe6 would
have been fine for Black since
This move is no better than 7 . . .
19.c4? ! runs into 19 . . . d5 2 0 . cxd5
dS, but I suspect that i t may b e more
cxdS 21.exd5 lt:lf4+.
unpleasant to the opponent. Indeed,
White usually takes on c6 in order
to have a clear-cut plan of a king­
side attack. Instead, he will have to
brace himself for a positional fight
in the centre with the d6-pawn as
the prime target.

Now Black solves the problem of

his "bad bishop" with a little tacti­
cal trick:

8.b3 i.e7 9.i.b2 e5 18 . . . �e4 19.ge4 d5 20.gg4

Against other setups, Black could
arrange his pawns differently, for in­
stance he could play . . . cS, followed
by the manoeuvre lt:lf6-d7-b8-c6.
Another possibility for Black is to
delay disclosing his plans and com­
plete development first. Then . . . d6-
d5 could be on the agenda again if
meanwhile White undertook some
dubious manoeuvre.
10.tild2 0-0 ll.�c4 .te6
12 .e'e2 tild7 13.gadl ge8 14.§'d2
dxc4 2 1.ti'h6 i.c5 22.Wf6 cxb3
23.axb3 ge7 24.ti'g5?
White would have had suffi­
cient counterplay with 24.�f3 !!f8
25.�e2 i.f2 26.�hl = . Black has an
extra pawn, but the need of defend­
ing it makes his major pieces too
passive. Instead, White suddenly
crumbles down.
24 . . . gf8 25.gfl Wd7 26.gc4
.tf2 27.�hl gef7 28.Wh5 Wd2
29.gc6 i.b6 0-1

288
 Pa rt 9

Ala pin 3.c3 d 5

289
Pa rt 9 . 3 . c3 d5

Main Ideas

Sveshnikov's pet line with c3 is de­ 3 . . . d5 4.exd5

servedly considered to be rather
timid. You'll hardly see a top-level
GM playing it regularly. The main
lines are depressively equal and
deeply explored. I have always had
this problem - how to beat 2 2 0 0-
2300 rated opponents after:
4.e5 li:lc6 S.d4 is a sideline of the
French Defence. I analyse in "Step
by Step" 4 . . . d4 S . .id3 .id7, planning
. . .ic6
4 . . .Ybd5 5.d4 li:lf6

l.e4 c5 2.�f3 e6 3.c3

If White played 2 .c3, then both

2 . . . li:lf6 and 2 . . . d5 3.exd5 �xdS 4.d4
li:lf6 S.li:lf3 .ig4 would have offered
more chances to complicate White's
life. In the German league I even
played 2 . . . b6 and 2 . . . d6. On both
occasions I failed to equalise in the
opening, but I went on to win easily White has many different move
very instructive games. orders from this point. You should
In The Safest Sicilian, I proposed remember several important points :
my main repertoire with . . . li:lf6. It
1. When to take on d4;
is still a fine choice. After 2008, I
developed for my 1900-2300 stu­ 2 . When to play ... li:lc6 instead of
dents a repertoire, based on . . . dS. . . . .ie7;
It is easy to play against an isolated
3. When and where to retreat the
pawn. You can learn the basics very
queen from dS ;
quickly. True, there are a number
of drawish lines, but winning with 4. Where to develop the dark­
Black is not always obligatory . . . squared bishop.

290
 Alapin: 3.c3 d5

The following short lines an­ D. 6 . .ie2 ltlc6 7. 0-0 cxd4 ! 8.cxd4
swer these questions: .ie7 9.ltlc3 'Wd6

A. 6.l!)a3 ll:lc6 ! 7.ll:lb5 (7 . .ie2

'Wd8 ! ; 7 . .ie3 cxd4 ! ) 7 . . . �d8 8.dxc5
hc5

I would like to bring your atten­

tion to the following important line:
lO.ll:lbS 'Wd8 ll . .if4 ll:ld5 12 .i.g3 0-0
13.i.c4 a6 14.hd5 exd5 15.ll:lc7 :i!b8
16.ltle5 (which is allegedly better for
B. 6 . .ie3 cxd4 7.cxd4 .ib4+ !
White according to Sveshnikov)
8.ll:lc3 0-0 9 . .id3 b6

16 . . . .if5 ! = .
C. 6 ..id3 .ie7! 7.0-0 0-0 ! 8 . .ie3
cxd4 9.cxd4 b6 I did my best t o keep the "Step
by Step" chapter as succinct as pos­
sible. It will provide you with more
details on the plans of both sides.
Playing 3 .c3 d5 does not demand a
lot of theory, but you must be pre­
pared for long endgames or IQP po­
sitions. In short, the better players
are likely to prevail since it is rare
to meet crushing novelties.

291
Pa rt 9 . 3.c3 dS

Step by Step

l.e4 c5 2.�£3 e6 3.c3 d5 lLlxf3 + 12 .�xf3 lLlxc6 13.i.e4 �c8 =)

7 . . . lLl c6 8 . lLlf3 \Wc7 9.�e2? ! does not
keep the extra pawn because the cl­
bishop is hanging - 9 . . . lLlxe5 ! .
Even more interesting i s 6 . . . i.c6 !
(instead of 6 . . . cxd4). The i.d3 is
hanging so White has nothing bet­
ter than 7.0-0 = .

6.i.e4 i.c6

4.exd5

I suggest to meet 4.e5 by 4 . . . d4

since the rare French which arises
after 4 . . . lLlc6 S.d4 might be unfamil­
iar to you.
Our pawn on d4 is immune since
S.cxd4 cxd4 6.i.b5+ i.d7 7.lLlxd4? ! 7.d3
hbS 8.lLlxb5 a6 9.lLl5c3 lLlc6 regains
7.\We2 gives a tempo for devel­
the pawn with an edge - 1 0 . 0-0 (10.
opment - 7 . . . lLle7 8 .d3 (8.0-0 he4
\We2 \Wd4; 10.f4? \Wh4+ ll.g3 \Wh3)
9.�xe4 �dS ! lO.�xd5 lLlxd5 11.cxd4
10 . . . lLlge7 ll.lLla3 lLlxeS 12 .d4 lLl5c6
cxd4 ! 12.lLlxd4 lLlb4 13.lLlc3 lLld3 ! = )
13.lLlc2 g6 ! .
8 . . . he4 9.\Wxe4 dxc3 10.bxc3 lLlbc6.
Or S.lLla3 lLle7! 6.i.d3 lLlg6 7.0-0
i.d7. So White plays : 7 . . . lLle7 8.0-0. Black has two in­
teresting options here which need
S.i.d3, hoping for i.e4. I advo­
testing:
cate to anticipate this idea with :
a) 8 . . . lLlg6 9.cxd4 (9.i.g5 �d7
s . . . i.d7! when 6.cxd4 cxd4 10.cxd4cxd4 ll.lLlbd2 h6 12 .hg6
7.lLlxd4? ! (7. 0-0 ! i.c6 8.lLla3 lLle7 fxg6 13.i.h4 lLla6 14.a3 gS 15.i.g3
9.lLlc2 lLld7 10.lLlcxd4 lLlxe5 ll.lLlxc6 lLlcS) 9 . . . cxd4 10.�el i.e7 ll.lLlbd2

292
 Alapin : 3.c3 dS
0-0 1Vt:lb3 .be4 13.l'!xe4 tl:lc6
14.�e2 l'!c8 ! ?
We'll sacrifice a pawn. The con­
trol of the c-file and our active piec­
es ensure sufficient compensa­
tion - 15 . .if4 �b6 16.tl:lbxd4 (16 .
.ig3 l'!fd8 17.h4 aS 18.h5 tl:lf8) 16 . . .
tl:lxf4 17.l'!xf4 tl:lb4 18.a3 tl:ldS (18 . . .
%!/a6 19.tl:le1 l'!c7 2 0 .l'!d1 tl:ldS 2 l . l'!g4
l'!fc8�) 19.l'!e4 l'!c7 2 0.%!/d2 E:fc8.
b) 8 ... .be4 9.dxe4 tl:lbc6 10 .�b3
�d7 ll.tl:lbd2 tl:lg6 12 .tl:lc4 i.e7
13.l'!d1 0-0 14.i.e3 is difficult to as­
sess.
Both 14 . . . d3 ! ? and 14 . . . tl:lh4 lead
to dynamically balanced positions.
4 . . . �xd5 5.d4
5.tl:la3 tl:lf6 6 .d4 transposes.
6.tl:lb5 is dubious - 6 . . . �d8 7.d4 a6.
Sveshnikov mentions that the
delay of d4 "might" be in White's
favour after 6.i.e2 i.e7 7.0-0 0-0
8 .tl:lb5, but in fact it is the opposite.
Following 8 . . . �d8, White has no
other way to justify his strategy, but
play 9.a4 tt:Jc6 and, again, 10.d4 is
the only sensible continuation.
White cannot prove in practice
any advantage in this system so
he has tried nearly all legal moves.
I will focus on:
A. 6 . tl:l a3 ; B . 6.i.e3 ; C. 6.i.d3 D.
6.i.e2
Rozentalis also employs 6.a3
which is best met by 6 . . . tt:Jc6 7.i.e3
i.e7 8.i.d3 (8.dxc5 %!/xd1+ 9 . c;!{xd1
tl:lg4) 8 . . . 0-0 9.0-0 b6 10 .%!/e2
i.b7=, Rozentalis-Volokitin, AUT
2014. After ll.l'!d1, Black can isolate
the d4-pawn and follow the tested
plans, or keep on the tension with
ll . . . l'!fd8.
Less popular alternatives are:
a) 7.i.e3 cxd4
293
 Part 9

7. . . lt'lg4 is less explored. White 10 . . . i.e7

commonly prefers to keep the bish­
I do not like 10 . . .f6, but 10 . . . �b6 is
op - 8.i.g5 [8.i.c4 ! ? lt'lxe3 9.fxe3
a solid alternative. However, the po­
'\Mfd8 10. 0-0 i.e7; 8.lt'lb5 lt'lxe3
sition after ll.i.c4 ! i.c5 1 2 . 0-0 lt'lxd4
9.fxe3 '\Mfd8 10.d5 exdS 11.�xd5 i.e6
13.lt'lxd4 0-0 (13 . . .i.xd4 14.i.xd5)
12 .'\Mfxd8+ �xd8 13.0-0-0+ �c8
14.i.xd5 exdS 15.lt'lb3 i.e6 16.lt'lxc5
14.lt'lg5, Iordachescu-Potkin, Ser­
�xeS 17.i.e3, albeit drawish, leaves
pukhov 2008, 14 . . . i.xa2 ! 15.c4 i.b3
Black zero winning chances.
16J�ld5 i.e7) 8 . . . i.d7 9.lt'lb5 (9.h3
lt'lf6 10 .i.d3 cxd4 ll.lt'lb5 l'!c8 12.0-0 ll.lt'lxc6 bxc6 12 .i.xe7 �xe7
dxc3 13.lt'lxc3 '\Mfa5 14 . .txf6 gxf6 13.i.d3 (13 .i.c4 0-0 14.0-0 i.b7
15.l'!e1 i.e7; 9 .i.e2 cxd4 10.lt'lb5 l'!c8 15.l'!e1 l=!ad8 16.�c2 c5= ) 13 . . . 0-0
ll.cxd4 i.b4+ 12 .lt'lc3 0-0 13.0-0 14.0-0
'\Mfa5=) 9 ... l'!c8 10 .h3 lt'lf6 . Here ll.c4
'\Mre4+ 12 .i.e2 lt'lxd4 13.lt'lbxd4 cxd4
14.0-0 was equal in Tiviakov-Lenic,
Trieste 2008. ll.dxcS i.xcS 12 .i.e3
i.xe3 13.lt'ld6+ �f8 14.f:xe3 deserves
attention although Black neutralis­
es the initiative with 14 . . . l'!d8.

8.lt'lb5 '\Mfd8 9 .lt'lbxd4

Black has a good centre and an

excellent knight. He can follow up
with 14 . . . l'!d8 (14 . . . lt'lf4 ! ?) 15.�a4
l'!b8 16.l'!ab1 c5 when all his pieces
will be active.

b) 7.i.e2 �d8 ! 8.lt'lc2

9 . . . lt'ld5 ! (9 . . . lt'lxd4 10 .i.xd4 i.e7

ll.i.d3 is dangerous) lO .i.gS
10.lt'lxc6 bxc6 ll.i.d4 f6 12 .i.c4
i.d6 13.0-0 0-0 14.l'!e1 - Black con­
trols the centre. He has a number
of good options: 14 . . . �h8 ! ? 15.i.e3
'\Mfc7 16.'\Mfc2 l=!d8; 14 . . . �c7 15.i.b3
(15.'\Mfc2 �h8 16.l'!ad1 eS 17.i.e3 i.e6
18.i.cl l'!ad8) 15 . . . l'!d8 16.We2 eS ; 8.0-0 cxd4 9.lt'lb5 dxc3 ! (The
14 . . . e5 15.i.e3 i.e6. most testing. 9 . . . .tcs is balanced

294
 Alapin: 3.c3 d5
- 10.lt.Jbxd4 lt.Jxd4 ll.lt.Jxd4 0-0.)
10.�xd8+ �xd8 ll.lt.Jg5 �e7
12.bxc3 h6 13.�a3+ (13J:'!d1 �d7
14.lt.Jf3 a6) 13 . . . �e8 14.lt.Jc7+ �d8
15.lt.Jxa8 hxg5.
8 . . . �e7 9 . 0-0 0-0 10.dxc5
10 .�g5 cxd4 ll.lt.Jcxd4
12 .Ele1 lt.Jxd4 13.�xd4 .ic6= .
10 . . . .ixc5 ll.'�xd8 (ll.�g5 e5)
ll . . . E!xd8 12 .�e3 .ixe3 13.ttJxe3 b6= .
7 Wfd8 8 . dxc5 hc5 9.Wfxd8 +
. . •
�xd8 10 .if4 •
a) 10 .b4 �b6 ll.lt.Jd6 (11.�f4
lt.Je4) 1 1 . . .�e7 1 2 .lt.Jc4 lt.Je4! 13.lt.Jxb6
axb6
14.b5 (14.�b2 Eld8 ! 15.a3 f6 was
slightly better for Black in Lastin
-Krasenkow, Dagomys 2 008)
14 ... lt.Jb8 15.a4 lt.Jd7 16.lt.Jd2 lt.Jxd2
17.�xd2 Eld8 18.�c2 lt.Jc5 19 .�e3
Ela5 2 0 .�e2 e5= , Grekh-Kononen­
ko, Odessa 2 008.
b) 10 . .ig5 �e7 1l.Eld1 h6 12 . .ixf6+
(12 .�f4 lt.Je4) 12 . . . gxf6+, Landenber­
gue-Matlakov, Rhodes 2 013.
lO • . . tLle4!
Sveshnikov wrote in 2010: 'White
has not found an advantage yet'. In
2 0 14, that is all the more true. The
latest top level game is Predojevic­
Carlsen, Lillehammer 2 013. It went:
�d7
Or 11.Eld1 + .id7
ll . . . �e7 12.lt.Jbd4 lt.Jxd4 13.lt.Jxd4
Eld8 14 . .id3 lt.Jf6 15.lt.Jb3 �d6= .
12 . .ie3 (12.lt.Jbd4 lt.Jxd4 13.lt.Jxd4
e5) 12 . . . .ixe3 13.fxe3 �e7 [13 . . . a6
14.lt.Jbd4 �e7 15 . .id3 li.Jf6 (15 . . . lt.Jd6=)
16.0-0 E!hd8 17.e4 lt.Jg4 18.Elfe1 f6= ,
Sveshnikov-Balogh, Warsaw 2005]
14.�d3 lt.JcS (14 ... lt.Jf6 15.0-0 lt.Jg4
16.Elfe1 lt.Jce5 17.�fl .ixb5 18 . .ixb5
Elhd8= , Schellmann-Kempinski,
Dresden 2 0 08) 15.�c2 lt.J e5 = .
ll . . . tLlxg5 12 .hg5 + f6 13 . .ie3
he3 14.fxe3 �e7 15.0-0-0 tLle5
16 .ie2 .id7
•
Predojevic has ben following
in the footsteps of Baklan (against
Sutovsky, Plovdiv 2 003). In both
games White could not hold the bal­
ance!
295
 Part 9
B. 6.ie3 cxd4 7.cxd4 ib4 + !
8.c!i:Jc3 0-0
9.id3
a) 9.ie2 is passive - 9 . . . �aS !
10.�b3 b6 11.ctJeS (11.0-0 ia6
12 .ha6 ttJxa6 13.a3 hc3 14.bxc3
gac8 1S.c4 ttJb8 is similar) ll . . . ia6
12 .ha6 hc3+ 13.bxc3 �xa6. The
hanging pawns are not dangerous
because Black has traded two minor
pieces. Thus White's chances for an
attack are minimal. Radjabov-Iv­
anchuk, Amber-rapid Nice 2009,
went further 14.c4 gcs (14 ... �b7
1S.O-O ttJc6=) 1S.O-O ttJc6 16.gfd 1 = .
b ) 9.a3? ! hc3 + 10 .bxc3 �aS
11.�c2 (11.�b3 b6 12.ttJeS ia6
13.�b4 �xb4 14.axb4 hf1 1s.gxfl
gcs 16.@d2 ttJc6) ll . . . b6 12.id2 !
(12 .id3 ia6) 12 . . . ib7 is more tan­
gled, but White cannot complete
development without concessions :
13.id3 .bf3 14.gxf3 ttJc6 1S. O-O
eS; 13.c4 �hS 14.ie2 �g6 ! 1S.�xg6
hxg6 16.0-0 ttJc6 = .
9 ••. b6 10.0-0
White can prevent . . . ia6 with
10 .�a4, but the queen will be mis-
296
placed on the left flanc. After 10 . . .
hc3 + ll.bxc3 ib7 1 2 . 0-0 ttJbd7
13.gfb1 gfc8 14.�c2 �as 1S.id2
hf3 16.gxf3, Black obtained an at­
tack. His knight pair completes well
the queen - 16 . . . �hS 17.@g2 ttJdS
18.ia6 gc7 19.c4 ttJ7f6 2 0.a4 ttJe7,
Vlassov-Sutovsky, playchess.com
2003.
10 i.xc3 11.bxc3 ia6 12.ixa6
•••
lll x a6
It is easy to play as Black here.
White's pawn structure is really
"hanging". He should squeeze some
dynamic resources out of his piec­
es while we have the obligatory ma­
noeuvre . . . ttJa6-b8-c6 and pressure
on the c-pawn. Although the en­
gines assess the position as rough­
ly equal, practical experience is in
Black's favour. Here are two exam­
ples :
13.�e2 ttJb8 14.c4 �e4 1S.ttJeS
ttJbd7 (1S . . . ttJc6=) 16.f3 �b7 17.ttJxd7
ttJxd7 18.a4 gfc8 19.aS h6 2 0 J'Uc1
�c6 2 1.�d2 �d6 2 2 .if4 eS= , Ste­
vic-Grachev, Biel 2 0 1 1 ;
13 .�a4 ttJ b 8 14.c4 �d7 1S.�xd7
ttJbxd7 16.a4 gac8 17.ttJd2 eS 18.aS
exd4 19 .hd4 gfe8 2 0 . axb6 axb6
 Alapin: 3 . c3 dS
21.E1fel = , Stevie-Wojtaszek, Porto
Carras 2 01 1 .
C. 6 . .id3 J.e7! 7. 0-0 0-0 !
Note that our best move or­
der here is slightly different from
the other sixth moves. White de­
velops his pieces more aggressive­
ly so we must castle quickly. We are
not afraid of the endgame after 8.c4
�d7! 9.dxcS E1d8 10 .J.c2 (lO.�eS
�d4 ll.�e2 �bd7 1 2 .�xd7 J.xd7
13 .J.e3 �eS 14.�c3 J.xcS= ) 10 . . .
�xd1 11.E1xdl E1xdl+ 12 .J.xdl �bd7!
13.c6 (13 .b4 aS) 13 . . .bxc6 14.�c3
J.b7 1S.b3 cS= . Potapov-Hammer,
Aix-les-Bains 2011, went on 16.J.f4
�e4 17.�xe4 J.xe4 18.J.e2 f6 19.E1dl
J.c6.
8 .J.e3 cxd4 9.cxd4 b6!
10.�c3 �d8 !
Another difference ! We are
planning . . . �dS so we should not
present a tempo on �e4 by retreat­
ing to d6.
lU�cl
ll.�e2 planning E1adl, J.cl is
more popular, but Black gets com­
fortable play against the isolated
pawn - 11 . . . J.b7 12.E1adl �dS 13.J.cl
�d7 14.a3 (14 . .ibl � 7f6 1S.�e4 aS
16.�eS J.a6 17.�xf6+ �xf6 18.J.d3
J.xd3 19.E1xd3 E1c8+, Pap-Zubarev,
Paleochora 2010) 14 . . . E1c8 1S.�e4
h6 16.�g3 E1e8 17.E1fel .if8 18.h4
J.d6 19.J.a6 .ixa6 2 0 .�xa6 J.xg3
2 1.fxg3 �c7 2 2 .g4 �c4+, Kalezic­
Malakhov, Budva 2009 .
ll.�eS J.b7 12 .�e2 is a modifi­
cation of the above line - 12 . . . �c6
13.E1adl �b4 14.J.c4 �bdS 1S.E1fel
E1c8 = .
ll J.b7 12 .ibl �bd7 13.�e2
.•• •
Or 13 .�d3 J.xf3 14.gxf3 �b8.
13 ... �d5 14.J.d2 gcs 15.gfel
We have reached an IQP position.
The firm control of dS and comfort­
able development assure Black of an
easy game. Sooner or later we'll have
to define the pawn structure with
. . . �xc3. Perhaps we should not de­
lay this exchange - 1S . . . �xc3 ! ? . The
only game in my database saw in-
297
 Part 9
stead 15 . . J%e8 16.!1cd1 (16.'119 d 3 ll:lf8
17.ll:le5 f6 18.ll:lf3 �d7=) 16 . . . ll:lxc3 ! ?
17.bxc3 .idS ! 18.�d3 (18 . .id3 �c7)
18 . . . g6 19.�e3 b5
The transformation from an iso­
lated pawn to a hanging duo c3-
d4 is in Black's favour. The latter
structure is more passive and easy
to attack. 20.�f4 �c7 21.�g4 ll:lf6
(Black should seek to trade dark­
squared bishops in order to win the
c3-pawn. Perhaps more energet­
ic is 2 1. . . .id6 2 2 .h4 h5 23 .�h3 i.f4)
22 .�h3 .ixf3 23.�xf3 ll:ld5, Malak­
hov-Zhou, Rapid Ningbo 2 0 1 0 .
White cannot organise a dangerous
attack without a knight.
D. 6 .ie2 ll:lc6 7.0-0 cxd4 (7 . . .
•
fi.e7 8.c4 �dB 9.dxc5 ! gives White a
pull) 8.cxd4 .ie7 9.ll:lc3 ti'd6
298
a) 10.fi.g5 is a typical develop­
ment when White has a bishop on
d3. 10 . . . 0-0 11.!1c1 (ll . .ixf6 .ixf6
1 2 .ll:le4 ti'f4 13.ll:lxf6+ ti'xf6 14.!1cl
!1d8 15.!1c4 .td7=) ll . . . ll:ld5 12.ll:le4
(12 . .ixe7 ll:lcxe7 13 . .id3 b6 14.ll:le5
fi.b7 15.ll:le4 ti'd8 16.!1e1 ll:lf4 17.fi.f1
ll:lf5 18.�g4 ll:lg6 = ) 12 . . . �d8 13 . .ixe7
ll:lcxe7 14.ll:le5 b6=.
b) 10 . .ie3 0-0 11.!1c1 !1d8
White's bishops are not impres­
sive. That give us time to complete
development with . . . .td7-e8 or . . . b6,
. .. fi.b7:
b1) 12 .a3 b6 13.ll:lb5 (13 .ti'a4 fi.b7
14.!1fd1 !1ac8 15.ll:le5 �b8 16.fi.f4
�a8 17.fi.f1 ll:lxe5 18.dxe5 ll:ld5
19.ll:lxd5 .ixd5=) 13 . . . ti'd7 14.ll:le5
ll:lxe5 15.dxe5 ll:ld5 16.fi.d2 fi.b7= .
b2) 12.�d2 ll:ld5 ! 13.ll:le4 (13.
ll:lxd5 �xd5 14.fi.c4 ti'a5 15.�e2
.id7 16.!1fd1 .ie8 17.a3 !1ac8 18 . .ia2
fi.f6 = ; 13.!1fd1 fi.d7 14.ll:le4 'l!tfb8
15.ll:lc5 .te8 16.a3 b6 17.ll:le4 ll:la5
18.�c2 !1c8 19.�d3 !1xc1 2 0.!1xc1
b5) 13 . . . ti'b4 14.�c2 , Benjamin-Ye­
rmolinsky, Parsippany 1996, 14 . . .
�b6 ! threatening . . . e5 ! t .
 Alapin: 3.c3 dS

10 ••• �d8 ll.J.f4 axbS 1S.i.e4 has been extensive­

ly tested. It retains more pieces in­
1Llt'le5 is an older move which deed, but the hole on eS after 15 . . .
has drifted out of fashion. Black tt:lb4 16.'1We2 fS 17.i.b1 tLldS 18 .-tes
can be only glad to see its knight .id7 19.i.c2 .ic6 2 0.i.b3 �d7 is not
swapped because exchanges de­ to everyone's taste.
crease White's attacking poten­
tial and the pawn on c6 would bol­ 15.tt:lc7 (15.4Jc3 i.g4) 15 ••• gbs
ster the blockade of the isolated
pawn. 11 . . . 0-0 12 .i.f3 (1 2 _l[)xc6 bxc6
13.tLlc3 E!b8 14.tLla4 tLldS 15.�c2
i.d6 16.g3 �f6 17.E!d1 eS 18.dxe5
heSt) 12 . . . \Wb6 13.tLlxc6 (13.a4 a6
14.tLlxc6 bxc6 15.tLla3 i.b7 16.tLlc4
�c7 17.i.e3 tLld5 18.E!c1 !!feB 19.'1We2
cS= ) 13 . . .bxc6 14.tLlc3 i.a6 15.E!e1
!!adS 16.\Wa4 E!xd4 17.\Wxc6 �xc6
18.hc6 !!eSt.

ll .•• tt:ld5 12 .ig3 0-0 !

•

In my younger years, I played

against E.Sveshnikov 12 . . . a6 13.4Jc3
0-0, but there is no urgent need to
flap the knight away - yet !
13.J.c4
13.E!c1 �b6 14.\Wd2 (14 . .ic4 i.d7
15.E!e1 E!fd8 16.tLlc3 4Jf6 17.4Ja4
\WaS 18.tLlc5 �b6) 14 . . . E!d8 1S.E!fd1
i.d7 16.4Jc3 (16.4Jd6 hd6 17.hd6
tt:lxd4) 16 . . . E!ac8 = ;
13.4Je5 fS ! ? 14.h3 f4 with an ini­
tiative.
13 a6 ! 14.hd5 (14.4Jc3 4Jxc3
•••
1S.bxc3 bS 16.i.b3 i.b7) 14 exd5 ••.
The c7-knight can also be saved
with 16.�b3 i.d6 17.\WxdS hg3
18 .\Wxd8 !!xd8 19.hxg3 4Jxd4 and
16.E!cl i.g4 17.h3 hf3 18.\Wxf3 i.d6
19.tt:lxd5 hg3 2 0.fxg3 4Jxd4 = .
Sveshnikov writes that Black is
bound to struggle after the text, but
the position is a dead draw, with op­
posite-coloured bishops :
16 .•• J.f5 ! 17.tt:lxa6
17.E!c1 4Jb4 18.a3 i.d6 19.�b3
i.c2 2 0 .�e3 hc7 21.axb4 heS
22 . .txes E!c8 23 .�d2 .tfs.

17 gc8 18. tt:lxc6 gxc6 19.tt:lc5

.••

Entering a drawish line. 14 . . . hc5 20.dxc5 gxc5 21.J.e5 = .

299
300
 Part 10

King's Indian Reversed

l.e4 c 5 2 ./t)f3 e6 3.d3 .!ilc6 4.g3

The idea of 3 .d3 is to meet 3 . . . d5 by 4.4Jbd2. This move order allows White
to build a King's Indian set-up. In contrast, 3.g3 d5 forces White to take on
d5 and play against an isolated pawn. I consider this trendy line in Part 11.

301
Pa rt 10. Ki ng's I n d i a n Reversed

Main Ideas

The typical reasoning behind l.e4
c5 2 . �f3 e6 3.d3 �c6 4.g3 is:
"I'm playing the King's Indian De­
fence where the pawn should stay
on e4 instead of e3. So I'll have prac­
tically two extra tempi". It is com­
pletely wrong, because we'll not
comply and will choose a Sicilian
set-up ! If White stubbornly persists
in delaying d4, he could easily end
up with an inferior position right in
the opening.
Black's general aim is to get hold
of the centre by . . . e6-e5 and proceed
further with . . . f7-f5. At the same
time, he has not discarded yet plans
with . . . d7-d5 which is keeping the
enemy in haze. White can expand
on the queenside by a3-b4 or gain
space in the centre by pushing d4 :

A. 8.l:�el (threatening e5) 8 . . .

e 5 9.a3 d 6 10.b4 a6 !

4 .•• �ge7 5 . .ig2 g6 6.0-0

Black commonly meets h2-h4 by

. . . h6. This insertion is in Black's fa­
vour.

6 .•. .tg7 7.c3 0-0

302
 King's Indian Reversed

Do not allow b4-b5 ! It will be dif­ cxd4 in order t o keep c 3 unavailable

ficult to win such a position as White to White's knight.
can blockade the kingside, too. For
instance, when we push . . . f7-f5-f4, 9 .exd5
he will defend with h3, g4, lt:lh2.
Or 9.e5 �b6 10.lt:la3 (lO.dxcS
n. lt:lbd2 h6 12.gb1 J.e6 ! ? �c7 ! ) 10 . . . cxd4 ll.cxd4 f6 12 .Eiel
J.d7, see game 52 Bologan-Ivan­
Black has completed develop­ chuk, Moscow 1996.
ment and stands well in the centre.
9 ... lt:lxd5 10.dxc5 %Ya5
U.tl:lbd2 �xc5 12.tl:le4
B. 8.d4 d5 ! ?

12 ...�e7 13.c4 tl:lb6 14.c5 tl:ld5

The point here is to delay 8 . . . 15.gel �c7 16.gbl b6 = .

303
Pa rt 10. Ki ng's I n d i a n Reversed
Step by Step
l.e4 c5 2 . �f3 e6 3.d3 �c6 4.g3
4 .•• �ge7
White's insipid play also al­
lows early flank activity like 4 . . . b5
S.i.g2 i.b7, but such treatment is
not in the spirit of this book. Com­
mon sense and practical experience
show that central strategy brings
better results, moreover, it is easier
to implement on the board.
4 . . . d5 looks consistent. This set­
up has numerous adherents, but I
do not approve of it. In my opin­
ion, White's play is too easy and
straightforward after S.lt:lbd2 , fol­
lowed by i.g2 , 0-0, E:el, eS, lt:lfl,
i.f4, h4, lt:lh2 (e3)-g4. It is safer to
304
restrict enemy's options in the cen­
tre and on the kingside by fianchet­
toing the dark-squared bishop.
4 . . . g6 is a reliable move order
which commonly transposes to our
main line. It allows two variations
of independent significance though :
5.d4 cxd4 6.lt:lxd4 i.g7! 7.lt:lb5 dS
8.exd5 exdS 9.'\WxdS Vfie7+ 10 .i.e2
i.g4 with full compensation for the
pawn ;
S.i.gS '\Wc7 6.i.g2 i.g7 7.c3 lt:lge7
8 . 0-0 0-0 9.d4 (9 .'\Wd2 dS 1 0 J:%el
f6 ll.exdS lt:lxdS 12 .i.h6 E:d8 with
a good control of the centre) 9 . . .
cxd4 10.cxd4 d S ll.eS h6 1 2 .i.f6
hf6 13.exf6 lt:lfS 14.g4 lt:ld6 15.'\Wcl
'\Wd8 = .
5.i.g2
White may prevent 5 . . . g6 with
5.b3? ! , but on b2 the bishop stands
worse than on cl - S . . . dS 6 .'1We2
(6.i.b2 d4 7.lt:lbd2 eS) 6 . . . g6 7.i.b2
d4 when 8.e5 i.g7 9.lt:lbd2 0-0 10.a4
loses a pawn to 10 . . . '\WdS ll.lt:lc4
lt:lxeS.
5 ••• g6 6.0-0
 King's Indian Reversed
6 .d4 cxd4 7.ll:\xd4 i.g7 8.ll:\xc6
bxc6 9 .�d6 is dubious owing to 9 . . .
fS !
According to Dvoretsky, it is still
early for 6 .c3 i.g7 7.d4 cxd4 8.cxd4
due to 8 . . . �b6:;:: .
6 .c3 i.g7 7.i.e3 ! ? is an inter­
esting move order. Perhaps Black
should accept the challenge by play­
ing 7 . . . b6. Black is threatening . . . d7-
d5, so White must follow up with
8.d4 cxd4 (8 . . . d5 ! ? ) 9.ll:\xd4. This
position is still unclear:
Black has considerable counter­
play after 9 . . . i.b7 see game53
- play with less risks, you better
Grigore-Itkis, Bucharest 2 0 0 0 .
6.h4 i s commonly met b y 6 . . . h 6
which i s a useful move anyway.
6 • • . ll:\ge7
7.c3
7J:�el d6 (or 7. . . e5! ?) 8.c3 eS!
commits White's rook to el too ear­
ly. Basically it is better placed on fl.
7.ll:\bd2 0-0 8 .ll:\h4? ! dS 9.f4 can
quickly turn the tables in Black's
favour because the only threat of
White, g4, is easily parried by . . . fS,
while the queenside activity of Black
( . . . b6, . . . as, . . . i.a6, . . J:�a7 or . . . gbs,
. . . bS) should gradually prevail.
7 • • • 0-0
Black's general aim is to get hold
of the centre by . . . e6-e5 and proceed
further with . . .f7-f5 or . . . d6-d5. That
would be possible if White delayed
d3-d4, which is not dangerous, but
significantly changes the character
of play. Black is unable to prevent
the central break-through, but it is
up to him to decide what type of po­
sition to get after it.
If you are looking for clear
choose:
7 . . . e5 ! ? . Then 8.i.e3 d6 9.d4
exd4 10.cxd4 i.g4 reduces the ten­
sion to a roughly equal position :
ll.dxcS (ll.dS hf3 12.hf3
ll:\d4:;:: , Filipovic-Markus, Zadar
305
 Part 10

2003) ll ... dxcS 12.l!Jbd2 (1Vt:'lc3
ltld4 ! ? = , e.g. 13.�a4 ltlec6 14.lt:'lxd4
cxd4 15.e5 dxe3 16.�g4 exf2 +
17J3xf2 0-0=) 12 . . . b6 13 .Eib1 0-0
14.a3 ltld4 ! ? 1S . .id4 (1S.b4? ltlbS+)
15 . . .cd4+ Koskela-Zhukova, Istan­
bul 2003.
7 ... 0-0 is flexible, but allows
White to seize and hold the centre.
The resulting positions are double­
edged, with plenty of chances for
the better player. White's main con­
tinuations are :

In his book How to Beat the Si­

cilian Defence, G. Jones actual­
ly shows how White can fight for
equality in this line. He suggests :
8.a3 (instead of 8 . .ie3) 8 . . . d6
9.b4 a6 10 . .ie3 b6 ll.ltlbd2 0-0
12 .Eib1 h6 13 .bxc5

A. 8.Eie1; B. 8.d4

He has also tried:

I reached this position in one
of my games. Black pieces are well
placed, the centre is in my control,
but White can probably maintain
the balance after 13 . . . dxc5 (Jones
considers only 13 . . . bxc5 ? ! ) 14.ltlc4
Eib8 (perhaps I should have tried
the sharper 14 . . . b5 ! ? 1S.ltlb2 �d6
16.ltld2 ltlaS 17.f4 exf4 18.gxf4 .ib7
19.e5 �c7 2 0 . ltl e4oo) 15.a4 and the
game Reyes Larena-Delchev, Beni­
dorm 2007 ended in a draw after
1S . . . .ie6 16.�c2 �c7 17.Eife1 Eifd8
18.ltlfd2 bS 19.axb5 axbS 2 0 . ltlb2
ltlaS 21.c4 bxc4 2 2 .ltlbxc4 ltlxc4
23.ltlxc4 .if8 24.Eixb8 Eixb8 25.'�cl
.ixc4 26.�xc4 ltlc6 27.Eicl. · veloped in Black's favour after 11 . . .
a) 8 . .ie3 b6 9.d4 is another ver­
sion of the plan of line B. White
wins a tempo on .ie3 , but on the
other hand, the bishop is not well
placed there. After . . . ltlfS, it should
move elsewhere because Black
would solve all his problems if he
managed to trade it. Then . . . b7-b6
could turn to be just a present from
White's side. 9 . . . d5 10 .e5 lt:'lf5 11 . .ig5
f6+.
b) 8.ltla3 eS 9 .lt:'lc4 d6! 10 .a4 h6
ll . .id2 is extremely passive. In the
game Sakaev-Rublevsky, St. Pe­
tersburg 2 0 0 1 , Black had the ini­
tiative after ll .. .fS 1 2 .exf5 .ixfS
13.'�e2 �d7 14.ltlh4 .ig4 1S.f3 .ie6
16.f4 exf4 17.gxf4 dS. Moutousis­
Rogozenko, Debrecen 1992, also de­

306
 King's Indian Reversed
.ie6 12 .b4 �d7 13.a5 f5 14.�e2 cxb4
15.exf5 .hf5+.
c) 8.ttlbd2 is similar to 8.ttla3.
Black has many good possibilities.
He can choose the flexible 8 . . . d6,
planning . . . b5, . . . b4 and .. .f5 . My
game Pace-Delchev, Lugano 2 007,
went 9.�e1 e5 10.a4 h6 1U3b1 a6
12 .b4 cxb4 13.cxb4 b5 . My pieces
are more active.
The only reason behind this
move is to provoke e5, which is on
Black's agenda anyway! I consider
it in detail because it is surprising­
ly popular.
8 . . . e5
9.a3
9.tt:la3 d6 10 . .ie3 b6 11.'?9d2 has
been tested at highest level in Bru­
zon-Topalov, Wijk aan Zee 2 0 05.
White's idea is to trade dark­
squared bishops and open the cen­
tre with d4. Topalov decided to
prevent it with ll . . . .ig4 12 .h3 .h£:3
13 . .h£:3 �d7 14 . .ig2 f5 15.exf5 gxf5
with unclear position. All Black's
pieces are well placed and co-ordi­
nated. The more conservative ll .. .f5
yields good results, though.
9.tt:lbd2 d6 often transposes to
the other lines. In the game Udovcic­
T. Petrosian, Zagreb 1970, White at­
tempted to hinder plans with . . .f5 by
lO.tt:lfl h6 11 .ttle3 .ie6 12 .tt:lh4, when
12 . . . d5 ! 13.'?9f3 d4+ allowed Black to
gain space in the centre.
9 . . . d6
Another good plan is to prevent
b4 by 9 . . . a5. White answers 10.a4
d6 ll.tt:la3 h6 and Black is ready for
d5 or f5, for example, 12.ttlb5 d5= .
10.b4 a6 ! ?
10 . . . b6 allows 1l.b5 tt:la5 12.c4
with a closed position which would
be difficult to break. In the game
Grischuk-Najer, Moscow 2 014,
White chose the double-edged
ll.tt:lbd2 .ie6 12.ttlg5 .id7 13 .tt:lc4 h6
14.tt:lf3 .ie6 15 . .id2 b5 16.tt:le3 when
16 . . . �b8 would have equalised.
ll. �bd2 h6
We can try a little provocation ­
ll . . . .ie6 ! . Then 12.tt:lg5 .ic8 13.tt:lc4
b5 14.tt:le3 h6 15.ttlf3 .ie6 would be
fine - 16.ttld2 a5 . However, White
can transpose to the main line with
12J3b1 h6 ! ? .
307
 Part 10

Mladenov, Schwaebisch Gmuend

2 014;
10. lLia3 cxd4 11.cxd4 f6 1 2 . l"l:e1
!d7 13.l"l:b1 (13.exf6 hf6) 13 ... l"l:ac8
14.if4 fxeS 1S.dxeS l"l:f4 ! ? 16.gxf4
\!;lfb4 17.�b3 l"l:f8�, see game 52
Bologan-Ivanchuk, Moscow 1996.

9 ..ft:lxd5
• 10.dxc5 �as
lt.c�:lbd2

After this novelty, I d o not see l l .c4 runs into ll . . . lLidb4 12 .\We2
any sensible plan for White . 13 .i.b 2 l"l:d8 13 .i.e3 tt:l d3 14 . lLi c3 tt:lb2 !
(preparing d 4 ) 13 . . . b 5 14.tt:lb3 \Wb6
was pleasant for Black in Torre­
Gheorghiu, Manila 1973. 13.lLib3 c4
gives a good version of an open Si­
cilian while 13.\Wc2 l"l:b8 (or 13 . . . b6)
again passes the move to White.

B. 8.d4 d5 !

Practical experience has seen

Black struggling after 8 . . . cxd4
9.cxd4 dS 10.eS f6 ll.l"l:e1. White's
knight finds a good stand on c3, par­
rying enemy counterplay on the c­
file. The text is aimed against that
possibility. My suggestion from The Safest Si­
9.exd5 cilian, 2008 - 12 . . . �aS is still valid :
a) 13.�e2 \!;lfc7 14.l"l:d1 (14.c4
9 .eS \!;lfb6 hinders the normal de­
lLif6 1S.lLic3 eS 16.igS ie6=) 14 . . . b6
velopment of the enemy queenside.
1S.l"l:bl (1S.a4 ib7) 1S . . .h6 16.c4 h6
White has tried without success
17.b3 l"l:ad8 leads to double-edged
three moves here :
play (18.ib2 eS).
10.dxcS \!;lfc7! 11.if4 llJxeS
In 2 0 14, I can add that 13 ... �c7
12.lLixeS !xeS 13.heS �xeS= ;
could be saved - 13 . . . b6 ! ? 14.c4 ia6
10 .b3 id7 1l.ia3 cxd4 12.cxd4
1S.id2 �a4 16.b3 �a3 17.icl �e7
�Uc8 13.icS (13.lLic3 lLib4 14.\!;lfd2 aS
18.ib2 eS 19.l"l:adl l"l:ad8oo.
1S.l"l:fc1 if8 16 ..if1 lLifSt, Abergel­
Kveinys, Cappelle la Grande 2003) b) 13.igS h6 14.id2 \!;lfc7 1S.c4
13 ... \!;lfd8 14.l"l:e1 lLifS, Bravo Lutz-Sv. llJde7 16.llJc3 a6! 17.l"l:el l"l:d8 18 .�cl

308
 King's Indian Reversed
'it;h7= , Zhang Zhong-Rublevsky,
Poikovsky 2 0 04.
G. Jones offers as an improve­
ment 17.�cl 'it;h7 18 .�c2 gds
19.gadl .id7 2 0 .�e4 'and White can
return to attacking on the kingside'.
Jones' statement must be a joke
since Black has more pawns on the
kingside and only he can attack
there. After 2 0 . . . .ie8 (ensuring the
d4-square), White's alleged attack
might be launched (and stopped)
with 21. �h4 ltJfS, or 2 1 .h4 'it;g8 ( 21. . .
gac8 ! ? wins a pawn - 2 2 .h5 ltJaS
23 .hxg6+ ttJxg6+).
Black has active play in this line.
In many variations, the queen re­
turns home so it may be a better
idea to put it there at once. The
game might continue :
13.c4 ltJb6 14.c5 ltJd5 15.gel
'ec7 16.gbl b6 = .
309
Pa rt 10. Ki ng's I n d i a n Reversed
Com plete G a m es
51. Kindermann-Volokitin
Bundesliga 2003
l.e4 c5 2.c!i�f3 e6 3.d3 lDc6
4.g3 tDge7 5.J.g2 g6 6.h4 h6 7.c3
.ig7
8.0-0
The combination of 6.h4 with
8.0-0 looks strange. White only
weakens the g4-square and his
castling position. I do not see what
he achieves in return. Black com­
monly plays . . . h6 anyway. 8 . .ie3 d6!
9.�d2 is a more consistent attempt
to draw benefits from 6.h4, but 9 . . .
e5 10.lLla3 .ie6 leads White t o a n im­
passe. 11. 0-0-0 is rather risky (apart
from losing the a2-pawn). 11.d4
exd4 12.cxd4 dS+ is also d ubious so
310
White can castle short 11.0-0, when
ll . . .fSt earns Black the initiative,
despite the uncastled king.
8 . • . e5
8 . . . 0-0 might be even better.
Then 9.!!el (an attempt to improve
on Harikrishna-Navara, Dagomys
2008 which saw 9.d4 dS ! ) 9 . . . e5
10 .J.e3 b6 ll.�d2 @h7 12 .d4 exd4
13.cxd4 dS ! was slightly better for
Black in Maze-Navara, Caleta 2 0 14.
9.a3
Or 9.�e3 d6 10 .a3 ( 10 .�d2 J.g4
ll.lLla3 hS 12 .lLlc2 0-0 13.d4 exd4
14.cxd4 dS lS.exdS lLlxdS 16 . .ig5
�d6 draw, Minasian-M.Gurevich,
New Delhi/Teheran 2 0 0 0 ) 10 . . . 0-0
ll.b4 b6 12 .b5 lLla5 13.c4 .ie6 14.lLlc3
fSoo, Minasian-M. Gurevich, New
Delhi/Theran 2 0 0 0 .
9 0 - 0 10 .b4 b6 ll.b5 lD a5
. • •
12.c4 d6 13.lDh2 !
The break . . . t7-f5 is looming so
White tries to organise his king-
 King's Indian Reversed
side's defence. The march of the
black f-pawn should be prevented
at all costs.
13 • . • �e6 14.h5
14 • • • f5?
Every coach warns his pupils
against hasty decisions and pre­
mature attacks. Black needed only
one tempo to finish his preparation
for a direct action. After 14 . . . �b8 ! ?
15.ltlc3 fS 16.hxg6 f4�, his position
would have been very promising.
Instead, he sacrifices the exchange.
The problem is that White's castling
position cannot be seized by an as­
sault on the weakened light squares.
Black needs open files and rooks to
use them.
15.exf5 gxf5 16 .ixa8 ti'xa8
• f3? 32.�d2 ti'd8 33.J.xa5 bxa5
17.�b2 d5? !
Black is nervous because the op­
ponent is already threatening f4.
However, this move only deprives
Volokitin of future counterplay
against the c4-pawn (and weakens
eS) , so he should have kept it for
reserve. 17 . . .'�e8 18.f4 .if7 19.ltld2
.ihS 2 0 .'1MfeU is in White's favour,
too. Perhaps 17 .. .f4 18.g4, counting
on the closed character of the posi­
tion, is the best practical chance, but
it is difficult to admit a mistake . . .
18.l;el d4 19.ltld2 J.-fl 2 0 .g4!±
�h8
White launched a counter-at­
tack and his chances are clearly bet­
ter. Now 21.'1Mff3 ! ? 'IM/e8 22 .'1Mfg3 �g8
23.�e5 would have beeen promising.
The rest of the game is entertaining,
but irrelevant to the opening.
21.gxf5 ti'e8 22.ti'g4 J.h.S
23.ti'h3 l;f5+! 24.c!Oe4 �g6 25.lt:lg3
g-fl 26-l!�adl c!Of5 27.lt:le4 ti'e7
28.�cl? lt:lh4:ii 29.f4 J.h5 (Or
29 . . . exf4 30.ltlg4 lt:lxc4 ! 31.dxc4
he4 32.�xe4 'IM/xe4 33 .'1Mfh4 'IM/g6
34.�hl f3+.) 30.l;d2 exf4 3U�f2
34.ltlc5± ti'g5 35.�hl ltlf5 36.ltle4
ti'h4 37.ti'xh4 lt:lxh4 38 .c5 J.g6
39.c6 J.e5 40.lt:lg4 .if4 41.l;c2
ltlg2 42.l;xg2 fxg2 + 43.�xg2
gg7 44.�f3 gn 45.�g2 Sudden­
ly White starts repeating the moves.
45.ltlef6 .ie3 46.@g3 is probably
winning, without risks at that.
4s l;g7 46.�f3 gn 47.�g2
.••
Draw.
311
 Part 10

16.gxf4 �b4 17.Vb3 lf!f8i

52. Bologan-lvanchuk
Moscow 1996

l.e4 c5 2.tl)f3 e6 3.d3 c!i)c6

4.g3 g6 5.i.g2 i.g7 6.0-0 c!i)ge7
7.c3 0-0 8.d4 d5 ! 9.e5 'M>6
10.c!i)a3 (10.dxc5 is met by 10 . . .�c7)
10 cxd4 ll.cxd4
••.

White is already on the defen­

sive. He should have avoided dou­
bled pawns by 18J"!e3. From now on
Black is in total control and does not
leave any chance to the opponent.

ll ••• f6 18 Vxb3 19.axb3 lf!xf4+ and

• • •

Black went on to win 50 moves lat­

lvanchuk is persistently rasp­ er 0-1
ing at the enemy centre. If he chose
a flank strategy like ll . . . ltJb4, . . . a5,
. . . .id7, he would risk to see his ini­ 53. Grigore-ltkis
tiative evaporate at some point. Bucharest 2000
Then White's supremacy on the
kingside might prevail. lvanchuk l.e4 e6 2.d3 c5 3.g3 c!i)c6 4 .ig2
•

prefers to fight for every inch of the g6 5.c3 .ig7 6 .ie3 b6 7.c!i)f3 ltJge7
•

centre. The mere fact that Black is 8.d4 cxd4 9. c!i)xd4 .ib7
able to choose between two tempt­
ing plans shows that he has solved
the opening problems.

12.lf!el .id7 13.lf!bl lf!ac8

14 . .if4 fxe5 15.dxe5 lf!xf4!?

15 ... ltJb4 16.�d2 a5 (16 ... ltJxa2?

17J"!a1 E!f4 18.gxf4;t) was a good op­
tion too, but Ivanchuk grasps the
chance of annihilating the enemy After 9 . . . .ia6 10 .�a4, the bishop
central pawns. should retreat to b7 anyway.

312
 King's Indian Reversed

10.f4 14.e5 tt:\f5 = ) 13 .he5 heS 14.fxe5

tt:\c6. The arising position is tan­
White could try to hinder the gled and double-edged. 15.�d6 :1!e8
opponent's development with 10.0- 16.tt:\a3 �a6 17.:1!f2 tt:\xe5oo.
0 0-0 11.tt:\xc6 hc6 12 .ygd6. Then
12 .. J!e8 13.tt:ld2 �b8 14.tt:\c4t main­ 1 2 .c!b a3 d6 13.ygd2 �xd4
tains White's spatial advantage.
In Nijboer-Sakaev, Elista 1998,
Black came to the plan with t7-f5,
but started with 12 . . . �e8 . It may
turn to be a waste of time since af­
ter 13.tt:\d2 fS 14.£3 Sakaev found
nothing better than repelling the
queen by 14 . . . tt:\c8. I suppose that
12 . . . tt:\c8 13 .ygd2 fS? should be bet­
ter. If White's leaved the d2-square
free for his knight with 13 .�d3 , then 14.cxd4?
13 . . . d5 equalises. 14.hd4 eS 15.fxe5 would have
been equal.
10 .•• 0-0 11. 0-0 �c8
14 £5+ 15.e5 .b:g2 16.ygxg2
•••

Black is playing by "general con­ �d5 17.J.f2 dxe5 18.dxe5 ygd7

siderations". That is not a priori 19.�acl �a4 20 .ygf3 g5 2 1.fxg5
bad, but he had a more concrete ap­ he5 2 2 .yge2 yge4 23 .�xe4 fxe4
proach like 11.. .tt:\xd4 ! ? 12 .hd4 eS 24.�c4 J.g7 25.a4 �b4 26.b3
(or perhaps 12 . . . hd4 13.ygxd4 dS � d5 27.�e3 0-1

313
314
 Pa rt 1 1

Ra re Lines on M ove 3

l.e4 c5 2 .ll:\ f3 e6

A. 3 .b4
B. 3.b3
C. 3.c4
D. 3.ll:\c3
E. 3.�e2
F. 3.g3

315
Pa rt 1 1 . Rare Li nes on Move 3

Main Ideas

In this chapter, I'll try to arm you A. 3.b4 cxb4

against different "rare" lines which
are not that rare at lower level. I'm
sure that any devoted Sicilian play­
er is only too happy to see White
deviating from the main lines. Still,
some variations could be tricky and
quite embarrassing if you do not
know how to face them. Another
problem when studying them is that
most of the games in the database
are between low rated players and
only tend to obscure things instead
of providing a clue. I think that it is
unproductive to spend much time
on rare lines, so I'll choose for you
some essential information which
should be enough. Have in mind
that in most variations Black has
other good options (unlike in the
main lines where many moves are
critical ! ) . If you are used to another
line and like it, stay with it.
I examine six different systems
in one chapter, but the only thing
that they have in common is that
Black is fine without too much ef­
fort. Otherwise they should be
treated separately.
If you feel unhappy to defend
with an extra pawn for some initia­
tive, you should look at game 57
S.Williams-Delchev, London 2 0 13
which featured 3 . . . b6 ! ? . I recom­
mend you to grab the gift, because
this version of the Sicilian Gam­
bit is not too dangerous for Black.
White has committed his knight to
f3 and that deprives him of two im­
portant attacking plans. The one
is linked with the exchange sacri­
fice :ga3 which is pointless here. he
knight is barring the third rank, it
is pointless. The knight on f3 is im­
peding also his queen in its longing
for the kingside and especially the
square g4.

316
 Rare Lines on Move 3
4.a3 d5 5.e5 c!Oc6 6.axb4
hb4 7.c3 i.e7 8.d4 i.d7 9.i.d3
�k8 10.0-0 a6
You must watch out here for c3-
c4, for instance, meet ll.lt:lbd2 by
ll . . . lt:laS ! .
B . 3 .b3
White hopes to gain a spatial
advantage by pushing eS. Anoth­
er idea is to swap the light-squared
bishop via bS and play a hedge­
hog structure after c4. In that event
White's position would be very sol­
id. I propose to deprive the oppo­
nent of both positional "threats". At
least he will feel unhappy . . .
3 • • • a6 ! ? 4.i.b2 c!Oc6 5.c4 ! ?
5.d4 cxd4 6.lt:lxd4 stumbles into
6 .. .'�f6 ! 7.c3 .icS pinning the d4-
knight.
5 . . . 'i'c7 6.g3 c!Of6 7.e5 (7.lt:lc3
l"!b8 ! ? 8.i.g2 b5= ) 7 . . . c!Og4 8 .'i'e2
c!Oh6 9 . c!O a3 J.e7 10. c!Oc2 0-0
n.J.g2
ll . . . b5! 12.0-0 bxc4 13.Ybc4
a5 = .
C . 3.c4 c!Oc6
White's "secret" thought is to
enter hedgehog structures without
having to lose tempi on manoeuvres
like lt:lf3-d4-b5-a3 as considered in
Part 7. His "little" problem is that if
he does not push d4 at once, he'll
never be able to achieve it, because
we'll close the centre with 4 . . . e5 !
And if he does play 4.d4, the pres­
sure on the e4-pawn will force him
317
 Part 11

to fetch another black pawn toward game 56 Leventic-Delchev, Zadar

the centre : 4.d4 cd4 5.llJd4 llJf6 2 004.
6.llJc3 �b4 7.llJc6 bc6 8.�d3 eS ! fol­
lowed up by . . . .ixc3 and . . . d6.
D. 3.lLlc3
4.llJc3 e5 !
This continuation is often used
This move challenges the fun­ by White to sidestep rare move or­
damental principle of quick devel­ ders, for instance 3.d4 cxd4 4.llJxd4
opment in the opening. It is all the 'l't!b6.
more provoking when played by the
second player who practically lags 3 • • • lLlc6 4 .ib5
•

two tempi behind the enemy. Yet

most good players prefer it ! There White's last move initiates a
are two reasons for that: strange hybrid between different
1. 4 . . . llJf6 5.�e 2 ! gives White systems, which is amazingly popu­
some initiative. lar lately.
2. 4 . . . e5 is not dubious at all ! In
positions with closed centre the only 4 • • • llJ d4 !
reasonable plan is a flank break­
trough and White's knight on f3 is
an obstacle before the f-pawn. That
allows Black to fight for the initia­
tive with an early . . .t7-f5. The aris­
ing positions are strategic and the
better player usually outplays the
opponent.

5.d3 d6
5.0-0
Or 5.�c4 llJf6 ! 6.0-0 a6 ! 7.a4
'l'tlc7. Black is following the nor­
mal Sicilian ways of development.
When White plays d3, he retreats
the knight to c6 with good play.

5 • • • a6 6 .id3 llJc6 ! ?
•

A funny position. Both sides have

Black pushes . . . fS and gets good lost tempi aiming mostly at hinder­
counterplay. He can develop the ing enemy's plans. It seems to me
dark-squared bishop to e7 or g7. See that White is outwitted. He cannot

318
 Rare Lines on Move 3

even transpose to the main open Si­
cilian lines because 7.�e2 could be
challenged again by 7 . . . ll::l d 4 ! ? .
an early stage of development and
there is not written theory so Black
is at his own. The only serious way
to exploit White's tricky move or­
der is:

The flank attack i s perfect­

ly grounded here. The rook move
prepared �fl and d4, but deprived
the f3-knight of any retreat square.
Black comfortably finishes develop­
ment after 8 .g3 d6 9 .b3 �g7 1 0 . .ib2
h6 ll.�fl ll:Jge7= , Souleidis-Bou­
sios, Halkidiki 2 0 0 2 .

3 d5 ! ? 4.exd5 (4.d3 dxe4)

E. 3 .'ee2 � c 6 4.c3 d5 ! · 5.d3
•••

4 exd5 5.d4 (S.i.g2 'ee7+ ! ) 5

••• • • •

�f6 6.g3 fi.e7 7.fi.g2 0-0 8 .0-0 �f6 ! 6.fi.g2 fi.e7 7.0-0 0-0 8 .�c3
b5 = .
8.i.e3 cxd4 (8 . . . .tg4 ? ! 9.dxc5)
9.ll::l xd4 i.g4 !

8 . . . �c6 9 .fi.e3

The c3-pawn provides a good

target to attack.

F. 3.g3 ! ?
9 fi.g4! 10 .h3 cxd4 ll.hd4
•••

This i s a very trendy line. I also fi.e6 12 .ge1 gc8 = , Chadaev-Ivan­

choose it occasionally. It is still at chuk, Olginka 2 0 1 1 .

319
Pa rt 1 1 . Rare Li nes on Move 3

Step by Step

l.e4 c5 2.�f3 e6 What can White really count on

by giving up a pawn? He does take
the opening initiative and he redu­
ces Black's control of the centre.
Chess Stars recently published
the 550 pages-thick Soloviov's in­
vestigations on the Sicilian Gam­
bit, The Modern Anti-Sicilian (via
the move order 2 . a3 ! ?) . The author
summarised his opinion for the
readers of this book in the following
way: 'My feeling is that White has
A. 3 .b4 compensation for the pawn, but not
B. 3.b3 more. For engines, Black has a nice
C. 3.c4 position. Over the board, it is the
D. 3 . ./Llc3 opposite. Perhaps it is a matter of
E. 3.�e2 taste. Defenders like Black, attack­
F. 3.g3 ers prefer White. ' A few months ago,
when I faced 3.b4 for the first time
A. 3.b4 cb4 in my practice, I decided to ques­
tion White's strategy by 3 . . . b6 ! ? .
See my annotations t o game 57
S.Williams-Delchev, London 2013.
After the game, I took my time to
make my homework and eventually
came up with the following sugges­
tion for more principled readers :

4.a3

4.d4 (4.�b2 ./Llf6) gives Black

additional possibilities as 4 . . . ./Llf6

320
 Rare Lines on Move 3
5.i.d3 dS 6.e5 lDe4. The knight hin­
ders White's development, for ex­
ample, 7.0-0 lDc6 8.a3 (RlDbd2
lDc3 9.'Wel i.d7) 8 .. .f5 9.exf6 �xf6
10.i.e3 i.d7 ll.axb4 i.xb4 12 .c3
lDxc3 13.lDxc3 i.xc3 14J!bl i.b4
15.lDe5 aS+, Zvjaginsev-Neverov,
St. Petersburg 2 0 1 0 .
G M Zvjaginzev i s the only top
level GM who plays the gambit in
classical time controls on a regular
basis. He is some sort of the last of
the Mohicans.
4 d5 5.e5 tbc6 6.axb4 .ixb4
•••
7.c3 .ie7 8.d4 .id7 9.i.d3 �c8
10.0-0 a6
Black should be able to reach
this position against any White's
move order.
We'll wait for an opportunity to
develop our knight to h6. The cen­
tre is closed so we can leave our king
there for a while and attack the c3-
pawn.
U.tbbd2
a) 11.lDfd2 lDh6 1 2 . lDb3 lDfS
13.i.a3 i.xa3 14J'!xa3 l'!c7 15.�e2
lDa7! 16.lDc5 i.bS I do not see
-
enough compensation after the ex­
change of the light-squared bishops.
b) 1l . .ia3 lDa7 12.i.xe7 lDxe'T+.
c) ll.lDel lDa7 12.f4 lDh6 13.g4
lDbS 14.f5 i.gS and it is unclear how
to develop the attack. White should
not forget that his king is also in
danger. Look at the funny variation
15.f6 gxf6 16.i.xg5 fxg5 17.'Wf3 tbxc3
18 .�h3 'Wb6 19.lDf3
19 ... lDf5 ! 2 0 .gxf5 tbe4 2 l.�h6 g4
2 2 .'Wg7 l'!f8 23.�xg4 �b2 24.l'!a5
l'!cl 25.�g2 �xg2+ 26.@xg2 l'!g8+
27. @h3 l'!xfl 28.i.xfl exfS with a
mating net despite the trade of
queens.
u . . . tba5!
It is essential to prevent c3-c4 !
which would follow after ll . . . lDa7.
12 .�c2
Or 12 .lDb3 lDc4 13.l'!el h5 14.lDfd2
i.b5 15.i.c2 lDb6 16.l'!e3 'Wc'T+.
12 • . • 'Wc7 13.J.b2
13.i.xh7? 'Wxc3 14.'Wbl 'Wb4
15.i.b2 tbh6 16.i.c2 0-0 was hop-
321
 Part 11
less for White in Zvjaginsev-Rod­
shtein, Moscow 2012.
1 3 �h6
•.• 14.gfcl .ib5
15.hb5+ axb5 16 .ia3 ha3 •
17.gxa3 0-0 18.Wb2 �c4 19. �xc4
bxc4 20.gcal
It will be difficult to convert the
extra pawn due to the activity of
White's heavy pieces.
B. 3.b3
3 .•. a6 ! ?
Alternatively, Black might ex­
plore 3 . . . �c6 4 . .ib2 lt:lf6.
I used to play before 3 . . .b6, but
then 4.d4 cxd4 5.lt:lxd4 '1Wf6 is quite
risky. The text move is more useful.
4 . .ib2 �c6
White must define his plans.
Main continuations are :
Bl. 5.d4? ! ; B 2. 5.c4 !
Other minor options:
a) 5.'1We2 ! ? d6 6.g3 lt:Jf6 7 . .ig2
ie7 8.0-0 0-0 9.d4 cxd4 10.�xd4
id7 with a pleasant open Sicilian.
322
b) 5 . .ie2 d5 ! ? Black enters a fa­
vourable position with an isolated
pawn. Another decent solution is
5 . . . lt:Jf6 6.e5 lt:Jd5 7.0-0 lt:Jf4 ! 8 J:%el
d5 9.exd6 (9.-ifl g5 ! ?--+) 9 . . . .ixd6
lO . .ifl 0-0 = . 6.exd5 exd5 7.d4 lt:Jf6
8 . 0-0 .ie7 9 . dxc5 .ixc5= . Black's
plan is . . . 0-0, .. J%e8 , . . . .ig4, . . . lt:Je4.
He should only avoid exchanges.
c) 5.g3? ! (if White wanted to
fianchetto the bishop, 3.g3 ! would
have been the better choice) 5 . . .
d5 ! 6.exd5 exd5 7 . .ig2 (7.d4 lt:Jf6
8 . .ig2 '1We7+) 7 .. :�e7+ 8 .'1We2 �xe2+
9. c;t>xe2 .if5+.
Bl. 5.d4? ! cxd4 6. �xd4 �f6 !
By this original move, Black pins
the d4-knight along the main diag­
onal. Soon the bishop will crucify it
on the other diagonal, f2-c5, forcing
c3. Evidently, White's strategy is a
failure.
7.c3 .ic5 8 .ie2 �ge7
•
This is a good and safe develop­
ing move. However, if you feel like
grabbing a pawn, you can follow in
my footsteps: 8 . . . '1Wg6 ! ? 9.0-0 �xe4
10.lt:ld2 �f4 ll.g3 �c7 12.lt:lxc6 dxc6
 Rare Lines on Move 3

13.tt:le4 !i.e7 14.c4 tt:lf6 15.tt:lxf6 !i.f6 Alternatively: 7.tt:\c3 l'%b8 ! ?

16.!i.xf6 gxf6, Rivas Pastor-Delchev, 8.!i.g2 b5= ; 7.Y!le2 e5= ( 7. . . d6 ! ?) .
Benasque 2 0 05. However, White
has probably enough long-term 7. . . �g4 8.Y!le2 �h6
compensation. I encountered later
9.!i.f3?! tt:�es 1o. o-o tt:lf6 1l.tt:ld2 o-m=, I'm following the game Ro­
Paljusaj-Delchev, Sibenik 2 0 07. manov-Svidler, Sochi 2 0 1 2 , where
Black had a comfortable game. An­
9. 0-0 0-0 other option is 8 . . . !i.e7 9.!i.g2 0-0
10.0-0 f6.
The game is balanced, but White
should still make a couple of accu­ 9 . � a3 J.e7 10.�c2 0-0 ll . .ig2
rate moves. For instance, 10.tt:la3 ? !
i s not enough, owing t o 1 0 . . . d5
ll.exd5 exd5+.

B2. 5.c4 ! ? Y!lc7 (discouraging

6 .d4? in view of 6 . . . cxd4 7.tt:\xd4
Y!le5)

ll . . . b5 ! 12.0-0 bxc4 13.�xc4

a5 = .

C. 3.c4 �c6

6.g3

6.!i.e2 tt:lf6 7.tt:lc3 tt:ld4 8.d3 is too

modest. Perhaps simplest would be
8 . . . tt:\xe2 9.�xe2 d6= .
6.tt:lc3 tt:lf6 7.g3 (7.d4 cxd4
8.tt:\xd4 !i.c5) 7 . . J%b8 ! ? 8 .!i.g2 b5
9 .d3 !i.e7 1 0 . 0-0 0-0 1 U%e1 d6= .

6 . . . �f6 7.e5 Cl. 4.d4; C2. 4.tt:\c3

323
 Part 11
Cl. 4.d4 cxd4 5.�xd4 �f6
6.<tlc3
6.<tlxc6 dxc6 ! ? ( 6 . . . bxc6=) is good
for Black
6 •.. .tb4 7.�xc6
Black can answer 7.f3? ! or
7.�d3? ! by 7 . . . 0-0, while 7.tLlb5
could be faced by 7 . . . a6 ! ? 8.ttJd6
cJ;; e7 9.�f4 e5 10.ttJf5 'i!ff8 11.�g5 d5
12.cxd5 !f5=, Lilov-Drenchev, Sun­
ny Beach 2007.
7 •.• bxc6
It is widely believed that 7 . . . dxc6
8.�xd8+ 'i!ld8 is equal, but in my
opinion 9.e5 ! ? gives White a lasting
initiative on the kingside. It could
be developed with h4 and l'!h3 .
8 .id3
•
Or 8.e5? ! ttJe4 9.�g4 ttJxc3 10.a3,
when 10 . . . !f8 ! ? is the safest way to
get an edge : ll.bxc3 �a5 12.�g3 g6
13.!f4 !g7+, Alburt-Sunye Neto,
Wijk aan Zee 1980.
8 ••• e5! 9. 0-0 0-0
324
9 . . . hc3 is overoptimistic. White
can seize the initiative with 1 0.bxc3
d6 11.�a4 �c7 12 .c5 d5 13.exd5
ttJxd5 14.c4 ttJe7 15.�b2 0-0 16.l'!ae1
f6 17.�c2 h6 18 .f4t.
10 .ie3
•
Here is the best place for the
bishop. After 10 .!g5 h6 ll.!h4
!c5 ! Black seizes the g1-a7 diago­
nal. In the game Christiansen-Rib­
li, Deutschland 1992 White tried to
attack firstly on the queenside with
12.a3? ! a5 13.'Wd2 d6 14.'i!lh1 !e6,
and then on the other wing -15.f4? !
exf4 16.l'!xf4 g5 17.l'!afl, but 17 . . . !d4 !
18.l'!4f3 ttJg4 ! 19.!g3 ttJe5 2 0.!e5
!e5+ proved to be in Black's favour.
This example shows that 1 0 .!d2
could turn out to be a waste of time.
Black answers .. !c5, . . . d6, . . . !e6.
10 • . • hc3
In his game against Kram­
nik, Kramnik-Kasparov, New York
1994, Kasparov played 10 . . . d6? !
11.tLla4 ! . In his comments he wrote
that Black should double the oppo­
nent's pawns to achieve equality.
ll.bxc3 d6 12.h3 !e6
 Rare Lines on Move 3

13.f4cc

Game 54 Chandler-Emms,
Hastings 2 0 0 0 (see "Complete
Games") confirmed Kasparov's
evaluation. Black had everything
under control.

C2. 4.lilc3
I did not like 9 . 0 - 0 ! h4 (9 . . . .ih6
1 0.d3) lO.ltJhl .ih6 ll.d3 d6 12.f4
(12 . .ixh6 lt:\xh6 13.f4 .id7) 12 . . . exf4
13 . .ixf4 hf4 14.Elxf4 �gS lS.Wffl
.ie6 16.W/f2 0-0-0 17.Elfl h3 18.g3
�eS. White's pawn structure is
more flexible.

4 ... eS is a solid option which

4 ••. e5 ! ?
Any move which hinders d4
should be a fair alternative. Still,
4 . . . W/b6, played by Swiercz, Sakaev,
Movsesian, looks a bit too extrava­
gant to me.
4 ... lt:\d4 ! ? · is more logical as it
solves the problem altogether! Its
main advantage is that the arising
positions are nearly unexplored.
S.lt:\xd4
If White delays this exchange,
we'll get the possibility to save . . . e6-
eS - S.d3 lt:\e7 6.4Jxd4 cxd4 7.lt:\e2
lt:\c6 8.g3 bS ! ? with an initiative,
Dolzhikova-Muzychuk, Kharkov
2012.
brings excellent practical results.
Black is two tempi behind in a sym­
metric position, but he can push
.. .fS without any preparation.
White can develop his bishop to
e2 or g2 . The same is valid for Black.
White's play on the queenside
can be slowed down with . . . aS. Look
at the game Naiditsch-Perunovic,
Paks 2014:
S.a3 aS 6 ..ie2 d6 7.d3 .ie7 8.4JdS
fS 9.exfS .hf5 10.0-0 lt:\f6 ll.lt:\h4
.ie6 1 2 .4Jxe7 W/xe7 13.f4 0-0 14.fxeS
dxeS

S . . . cxd4 6 . lt:\ e 2 e S 7.lt:\g3 g6

8 . .ie2 hS
32S
 Part 11

Black's game is already slightly 4 . . . l!Jge7 is also frequently seen,

better. but it gives White a free hand in the
S.d3 d6 6 . .ie2 .ie7 7.0-0 f5 centre.
8.exf5 h£5 9 . .ie3 l!Jf6 lO.l!JgS 0-0
ll..if3 W'd7 12.l!Jge4 l'!ab8 13.l!Jxf6+
hf6 14 . .ie4 (Pridorozhni-Jakoven­
ko, Aix-les-Bains 2011) 14 . . . .ig4
1S ..if3 .ifS 16 . .ie4 �h8 = , Zhigalko­
Roiz, Olginka 2011 (16 . . . .ig4 17 . .if3
draw, Solak-Nikolov, Albena 2012).

White can also fianchetto his

bishop - S.d3 d6 6.g3 .ie7 7 . .ig2 fS,
see game 56 Leven tic- Delchev, Za­
dar 2004.
5.0-0

After S ..ic4 l!Jf6 ! 6.0-0 (6.e5

D. 3.lbc3 l!Jg4 7.l!Jxd4 l!JeS ! +) 6 ... a6 ! (beware
the trap 6 . . . d5? 7.exd5 exdS 8.l!Jxd4
This continuation is often used cxd4 9.l!Jxd5 ! l!JxdS lO .�hS--+)
by White to sidestep rare move or­ White's bishop is vulnerable on
ders of the opponent, for instance c4, for instance 7.d3 W'c7 ! ? 8 .l!Jxd4
3.d4 cxd4 4.l!Jxd4 W'b6. Nothing un­ cxd4 9.l!Je2 .id6 10 .h3 l!Jxe4 ! = . The
usual so far, we play: last move explains why Black has
been delaying b7-b5, which is also
3 .tbc6
••
good, by the way. Some players pre­
fer 7.a4 instead of 7.d3.
If your repertoire includes the
Kan, you should choose 3 . . . a6 !
when 4.d4 would lead to the most
pleasant version of the Kan - with
l!Jc3 - while 4.g3 bS is double-edged
and in any case not worse for Black.

4 . .!b5

White's last move leads to an

amazingly popular lately hybrid be­ Now 7 ... l!Je4? ! is dubious in view
tween different systems. of 8.l!Jxe4 dS 9 .d3 dxc4 10.dxc4 .ie7
11 . .ie3 l!Jxf3 12 .�xf3±, Carlsson­
Jansson, Stockholm 2 0 04.

326
 Rare Lines on Move 3

7 . . . d5 8.exd5 exdS 9 . .ta2 .tg4

10 .'�e1+ .te7 11.tt'lxd4 cxd4 12 .tt'le2
d3 ! ? gives Black full compensation
for the pawn after 13.cxd3 0-0 14.d4
l':l:e8 15.f3 .tfs 16 . .tb1 hb1 17.l':'i:xb1
Wffc 7 18.Wid1 .td6 19 .g3 l':l:e6�.
Perhaps the simplest solution
has been demonstrated by Kiril
Georgiev against Tissir, Gibraltar
20 05: 7 . . . Wffc 7!? 8.h3 (8.d3 tt'lc6 ! ? is
similar) 8 . . . .te7 9 .d3 tt'lc6 ! ? White
defended his bishop and the knight
retreats back. Dl. 7 . .te2 ; D 2. 7.l':'i:e1

Black has many good plans after Dl. 7 .ie2

•

5.a4. I propose 5 . . . a6 6 . .tc4 (6 . .td3

tt'l c6) 6 .. .'�f6 7.tt'lxd4 cxd4 8.tt'le2
·
Wig6. Black's play is clear and easy.
9 .tt'lg3 hS 10.Wfff3 tt'lf6 ll.h4 (ll .d3
.tb4) ll . . . .td6 12 .d3 tt'lg4 13 . .tf4.
This position occuerred in Drago­
jlovic-Papp Zoltan, Szeged 1998.
Black could have gained the edge
with 13 . . . .tb4 !+ .
The most consistent response
to S . .td3 is 5 . . . tt'lxf3 ! ? (S . . . tt'le7
6.tt'lxd4 cxd4 7.tt'le2 tt'lc6= is a wor­
thy alternative) 6 .Wffxf3 a6 7.0-0
(or 7.b3 .td6 8 ..tb2 .teS= ) 7 . . .'�c7
8 . l':'i:e1 .td 6 ! 9.g3 tt'lf6 10 . .tf1 .teS=
ll . .tg2 bS 1 2 . a4? ! b4+, Iordaches­
cu-Rublevsky, FIDE-Wch k.o. Trip­
oli 2 0 04. Black's play in this game
was clear and straightforward.
In mathematics, you just write :
'7 . . ."Wc7 (or 7 . . . tt'lf6) 8.d4, transpos­
ing to our main line Taimanov' and
move on to other pending prob­
lems.
You should also consider 7 . . .
bS ! ? 8.d4 cxd4 9.tt'lxd4 .tb7 10.tt'lxc6
hc6, Andreikin-Caruana, Biel
2010.
Finally, 7 . . . tt'ld4 ! ? i s a capital
way of preventing d4. Perhaps it is
enough for equality, but you risk to
be summoned by the chief arbiter
for a booze control. . .
D2. 7.gel g5 !

5 . . . a6 6 . .id3 ti)c6 ! ?

A funny position. Both sides

have lost tempi aiming mostly at
hindering the opponent's plans. It
seems to me that White has been
outwitted. He cannot even trans­
pose to the main open Sicilian lines :

327
 Part 11
The flank assault is perfectly
grounded here. The rook move to el
prepared .tfl and d4, but deprived
the f3-knight of retreat squares. In
practice White has tried :
8.�fl .tg7 9.b3 g4 lO.tt:lel h5
n . .tb2 d5 (it is better to keep the
centre closed - ll . . . d6 ! ) 12 .exd5
exd5 13.tt:la4 hb2 14.tt:lxb2 tt:lge7oo
Safarli-Brunello, Khanty-Mansiysk
2010.
8.h3 h5 9.ttlh2 .td6 10 .b3 (10.
ie2? g4 11.g3 h4- + , Fruteau-Skrip­
chenko, Chartres 2005, 10 .g3? �c7)
10 . . . g4 11 . .tb2 .te5 ! ;
8.g3 d6 9.b3 .tg7 10 . .tb2 , Soulei­
dis-Bousios, Halkidiki 2 0 0 2 , 10 . . . g4
ll.tt:lh4 h5 ;
B.@hl - see game 55 Mark
Tseitlin-Avrukh, Tel Aviv 2002.
E. 3. '9e2 � c 6 4.c3
White's queen does not fit in to
the g3-design: 4.g3 tt:lge7 5 . .tg2 g6
6.0-0 .tg7 7.c3 0-0 8.�dl e5 = .
4 . . . d5 ! 5.d3 �f6 6.g3 ie7
7.i.g2 0-0 8.0-0 b5 = .
The c3-pawn provides a good
328
target to attack. The Skripchenko­
Koneru, blitz Moscow 2 0 1 0 , went
on 9 .e5 �d7 10 .h4 b4 n . .tf4 bxc3
12 .bxc3 .ta6 13.tt:lbd2 �a5 14.�fcl
�ab8 15.ttlfl, when best would have
been 15 . . . �a3 with a clear advan­
tage.
F. 3.g3 ! ?
White avoids the risks o f the
open Sicilian and all the theoretical
overhead it involves. The play takes
a manoeuvring course without tac­
tical complications. This allows the
strategically-oriented players to im­
pose their style which would be im­
possible after the sharp 3.d4 . Some­
times the surprise effect can bring
quick dividends. Look at my game
Delchev-Bagaturov, Edirne 2 0 1 3 :
3 . . . tt:lf6 ? ! 4.e5 tt:ld5 5 . .tg2 d6 6.0-0
tt:lc6 7.exd6 hd6 8.d4 0-0 9.dxc5
.b:c5 1 0 .'?9e2 V!fc7 ll.c4 tt:lde7 12 . .tf4
.td6 13 .hd6 V!fxd6 14.�dl V!ic7
15.ttlc3 a6 16.�d2 .td7 17.tt:lg5 �adS
18.�adl h6 19.tt:lge4 e5. I made all
the most natural moves and already
my advantage is decisive - 2 0 . �xd7!
�xd7 2 1 .tt:lf6+ gxf6 2 2 .�g4+ was
winning.
 Rare Lines on Move 3

Black can prevent eS with 3 . . . b) S.d3 d6 6.0-0 i.e7

lt:lc6 4.i.g2 lt:lf6 which I analyse as
line F1, or he can attempt to exploit
White's tricky move order with 3 . . .
d S which i s line F2. The choice de­
pends entirely on your taste. Line F1
is closer to the spirit of the Sicilian
while line F2 offers active piece play.

Fl. 3 . . . �c6 4 .lg2 �f6

.

Both sides have an enormous

5.'tYe2
a) s.�c3 indicates that White
intends to push d4, having side­
stepped the Kan and a number
of other lines. The principled an­
swer is S . . . d.S 6.exdS exdS 7.d4 cxd4
8.�xd4 i.g4 ! . I got this position
against Malakhov. He chose 9.lt:lxc6
bxc6 10 .'\Wd4 i.e7 11.0-0 0-0 12.lt:la4
when best is 12 . . . E:e8 with an excel­
lent position.
Similar is 9.'\Wd3 i.cS 10 .i.e3
'tYb6 ll.lt:lxc6 bxc6 12.i.xcS �xeS
13.0-0 0-0.
The latest attempt to improve
White's play was 9.f3 i.d7 10 .i.e3
i.e7 11.0-0 0-0 12.E:e1 E:e8 13.lt:lb3,
Safarli-Tregubov, Loo 2 013, when
13 . . . i.b4 deserves attention.
choice here. For instance:
7.c4 0-0 8 .lt:lc3 a6 9.d4 cxd4
10.�xd4 i.d7 followed up by . . . bS;
7.E:el 0-0 8.c3 i.d7 9.�bd2 '\Wc7!
(aimed against e4-eS) 10 .a3 E:ac8
ll.d4 cxd4 12.cxd4 eS 13.dS lt:lb8
14.lt:lf1 aS = , Van der Weide-Kemp­
inski, Groningen 1996;
7.lt:lgS h6 8.lt:lh3 eS 9.f4 i.g4
10 .�d2 �d4 11.lt:lf2 i.f3 .
5 • • . e5
Fighting for every inch of the
centre. Black can also allow e4-eS af­
ter s . . . d6 6.0-0 i.e7 7.c3 (7.E:d1 0-0
8 .d4 cxd4 9 .lt:lxd4 lt:lxd4 10.E:xd4
i.d7 ll.c4 a6 12.lt:lc3 bS�) 7 . . . 0-0
8.d4 (8.E:d1 dS ! 9 .eS lt:ld7 10.d4 aS
ll.i.e3 b6�) 8 . . . dS 9.eS lt:ld7

329
 Part 11
This is the favourite set-up for
Black of my club teammate Gan­
guly. He is Anand's second and one
of the world's most erudite open­
ing experts. He always puts his
queenside pawns on b6 and a5 :
10.j,e3 a5 ll.a4 b6 12Jk1 j,a6
13.�d1, E.Hansen-Ganguly, Lin­
ares 2013, 13 . . . b5 ! 14.axb5 hb5
15.lLlbd2 l'!b8 ;
10.h4 b6 1l . .ig5 (1Lj,f4 a5 12.l'!e1
j,a6 13.�d1 l'!c8 14.j,fl hf1 15.Wxfl
b5 16.lLla3 �b6 17.h5 b4 18.lLlc2
bxc3 19.bxc3 �b2 2 0.h6 g6 2 l .�d2
cxd4 2 2 .lLlcxd4 �xd2 23.hd2
lLlc5+, Vea-Ganguly, Caleta 2 0 14)
ll . . . a5 12.l'!c1 j,a6 13.�e3 l'!c8 ! (13 . . .
b5 14.lLlbd2 �b6 15.he7 lLlxe7
16.dxc5=, Nitin-Ganguly, Kanpur
2014) 14.lLlbd2 l'!c7 (14 . . . l'!e8) 15 . .ifl
hf1 16.lLlxf1 b5+.
6.0-0 Ae7
White has not shown a convinc­
ing plan here. The most natural try
is 7.c3 0-0 8.d4, but 8 . . . cxd4 9.cxd4
exd4 10.e5 ltJeS ll.l'!d1 �b6 is good
for Black. Therefore, White should
develop in the King's Indian man­
ner:
330
7. lLla3 0-0 8 . c3 d6 9.d3 ges
lO.lLlc4 .if8 ll.lile3 h6
I'm following the game Nai­
ditsch-E.H ansen, Reykjavik 2 0 14.
I'm not sure about Black's last
move. ll . . . l'!b8 planning . . . b7-b5-b4
is worth considering
t 2 . lLlh4 g6 13. Whl J.g7 14.Ad2
gbs = ts.gael
Obviously White's plan is f4 so it
would be good before moving the b­
pawn to defend the c6-knight with
15 . . . j,d7! . Then 16.f4 would drop a
pawn after 16 . . . exf4 17.gxf4 lLlxe4.
F2. 3 • • • d5 ! ?
The point i s that White cannot
enter King's Indian structure with
4.d3 since 4 . . . dxe4 5.dxe4 �xd1 +
6.Wxd1 lLlf6 can be only in Black's
favour.
4.exd5 exd5 5.d4 (5.j,g2
�e7+ ! ) 5 lLlf6 !
• • •
I t i s better t o castle quickly than
display activity with 5 . . . lLlc6 6.j,g2
 Rare Lines on Move 3

.ig4 7. 0-0 .ie7 in view of 8.dxc5 ! . 9.dxc5 hc5 i s equal ( 9 . . .d 4 is

a dubious attempt to play in the
6 .tg2 .te7
• Tarrasch style - 10.lL!a4 i.g4 11.h3
i.f5 12 .a3) 10.i.g5 d4 11.lL!e4 i.e7
6 .. .'�e7+ is a mistake. In Rabi­ 1 2 .lL!xf6+ hf6 13.i.xf6 1!;Yxf6 14.lL!e1
neau-Delchev, Fourmies 2 0 13, was i.f5 15.lLld3 lL!e5 16.lL!xe5 \Wxe5 = ,
7.'�e 2 ? ! \Wxe 2+ 8.'i!?xe2 lL!c6 and I Balogh-So, Saint Louis 2012.
was already better.
7.lL!e5 cxd4 8.0-0 \Wxe5 9J:�e1
lL!e4 10 . .if4 'I!;Yf6 11.f3 .ie7 is also
pleasant. White should answer:
7 . .ie3 ! lL!g4 8.0-0 · lL!xe3 9 .fxe3
lL!c6 1 0 .c4 ! and Black's centre is
crumbling down.

7.0-0 0-0 8 . lL! c3

8 . .ie3 cxd4 (8 . . . i.g4 ? ! 9.dxc5)

9.lL!xd4 i.g4 ! (9 . . . lL!c6 10.h3 ! was
9 . . .tg4 ! 10 .h3 cxd4 ll.hd4
slightly better for White in Adams­
.

.te6 12.l!el gcs =

Kryvoruchko, Tromso 2013, since
Black's bishop could not find an ac­
12 . . . h6 13.lL!e2 i.d6 14.hf6
tive stand) 10 .1!;Yd3 'I!;Yd7 11.c4 (11.
\Wxf6 15.lL!ed4 i.c5 16.c3 l!ad8
lL!c3 lL!c6 12 J'!fe1 l!fe8=) 11 ... lL!c6
was also balanced in Sadvakasov­
1 2 .lL!xc6 bxc6 13.lLlc3 = . If White at­
Khalifman, Liepaja 2 0 0 1 .
tempts to trick us with the move or­
der - 8.h3, we'll not take on d4 -
13.�e2 i. f5 14.c3 g e s 1 5 .te3•

8 ... lL!c6 ! 9.i.e3 \Wb6 10.lLlc3 \Wxb2

.ie4 16.�ed4 \Wd7 17.�d2 .tg6
11.lL!a4 'I!;Ya3 12.lL!xc5 b6+.
8.dxc5 hcs 9.i.g5 lL!c6 1 0 .lLlc3 Ivanchuk follows the rule to
i.e6= 11.lLld2 ? ! h6. keep more pieces when playing
with an isolated pawn, but 17 . . .
8 • . • lL!c6 9 .ie3
•
hg2 was also good. The game went
on 18.lL!xc6 (18 .\Wa4 a6) 18 . . . bxc6
9.i.g5 h6 10.i.xf6 i.xf6 11.dxc5 19.1!;Ya4 i.d6 with mutual chances,
hc3 12 .bxc3 1!;Ya5 = ; Chadaev-Ivanchuk, Olginka 2011.

331
 Pa rt 1 1 . Rare Lines on Move 3

Com plete Gam es

13 . . . exf4 14.hf4 tfb6 15.�hl

54. Chand ler-Emms
Vlfc5
Hastings 2000

l.e4 c5 2.�f3 e6 3.d4 cxd4

4. �xd4 �c6 5.c4 �f6 6.�c3 J.b4
7.�xc6 bxc6 ! 8.J.d3 e5 9.0-0 0-0
10.J.e3 hc3 (10 . . . d6? ! ll.tLla4 ! t)
ll.bxc3 d6= 12.h3 J.e6

White failed to achieve substan­

tial compensation for the split pawn
structure. Still, the bishop pair helps
him to hold on.

13.f4 Black improves his rooks before

If White seeks initiative on the
queenside, the pawn on c4 could fall
quickly, as in the game Van Weer­
sel-Nill, Chisinau 2 0 05, 13.l'%bl
Wfc7 14.Wfd2 (14.Wfa4 ll:ld7) 14 ... ll:ld7
15.l'%fdl l':1fd8 16.l'%b4? c5 17.l'%b2 ll:lb6
18.Wfe2 f6+.
Or 13.Wfa4 Wfc7 14.f4 exf4 15.J.xf4
tLld7 16J"i:fdl cS+, Doell-Kovalev,
Dresden 2010.
eating the enemy pawns. 17 . . . J.xc4 is
about equal : 18 .i.e3 WfhS (18 . . . Wfa5?
19.l'%xf6 gxf6 2 0 .Wfg4 'it>h8 21 .i.d4)
19.i.d4� ll:le8 2 0 .J.xa7 dS 2 1.i.b6 = .
18.tfd2 ffc4?!
Black has achieved perfect co­
ordination between his pieces. He
could have displayed that by open­
ing up the centre : 18 . . . d5 ! Instead,

332
 Rare Lines on Move 3
he inexplicably decides to sacrifice
the exchange accepting an inferior
position.
19 .ig5 c!Lle4
• 20 .ie4
• �e4
2 1 .id8 gd8 =
•
The a7-pawn is weak, but appar­
ently Black should be able to hold
the enemy to a draw.
2 2 . gd3 d5 2:tge1 �h4
24.gd4 �e7 25.�d3 �a3 2 6.�c2
gbs 27.ga4 �c5 28.�d2 V!/e7?
29.�d4 gb6 30.�e5?
White misses his chance to -reach
a clear edge by 30 .V!/b6! ab6 31.gas
V!Jf8 3 2 . gfs <i>f8 33.a4±. Now Black
gradually becomes better.
30 h6 31.ge3 'it>h7 32.gg3 f6
.••
33.�e3 V!/d7 34.V!Jc5 .if5 35.V!/a5
gb7 36.gb4 .ie4 37 .gxb7 V!/xb7
38.�d8 c5 39 .V!/d6 c4 40.'it>h2
a6 41.a3 a5 42.<i>hl a4 43.'it>h2
�f7 44.ge3 f5 45.V!Je5 V!Jf8 46.ge2
V!Jxa3 47.�e4 fxe4 48.�f5+ 'it>h8
49.V!/c8+ 'it>h7 50.V!Jf5+ g6 51.�f7+
'it>h8 52.�f6+ 'it>h7 53.�f7+
'it>h8 54.V!Jf6+ 'it>g8 55.�xg6+
'it>f8 56.�xh6+ 'it>e7 57.V!/g5+
'it>e6 58.�g6+ 'it>d7 59.�f5+ 'it>c7
60.V!Jf7+ 'it>c8 61.V!Ie8+ 'it>c7 Draw.
55. Mark Tseitlin-Avrukh
Tel Aviv 2002
l.e4 c5 2 . c!Llf3 e6 3.c!Llc3 c!Llc6
4.i.b5 c!Lld4 5.0-0
5.tt:lxd4? cxd4 6.tt:le2 V!Jg5 is
sometimes seen in blitz . . .
5 a6 6 . .id3 c!Llc6 ! ? 7.gel g5 !
• • •
8.'it>hl
8 • . • .ig7
Avrukh should have tried to de­
velop his initiative by 8 . . . g4 ! ? 9.c!Llgl
h5 lO . .ifl (10.f4 .ih6 ! ) , when 10 . . .
.id6 ! best fits i n the spirit o f the
Taimanov. Black's design is certain­
ly not to mate the opponent, but to
cramp him and prevent the freeing
d2 -d4
9 .ifl h6 10.d3 c!Llge7 ll.c!Lle2
•
d5? !
Black is inconsistent. Firstly
he refrained from g4, now he sud­
denly opens up the centre. ll . . . tt:lg6
1 2 .tt:lg3 0-0 13.tt:lh5 .ih8 = looks
preferable to me.
12 .c3 dxe4 13.dxe4 V!/xdl
14.gxdl b6 15.c!Llg3 ggs
15 . . . tt:lg6! 16.tt:lh5 0-0 would be
roughly equal.
16.c!Lld2
Black's awkward break-trough
in the centre only lost his tempo-
333
 Part 11

rary initiative and activated the en­

56. Leventic-Delchev
emy pieces. Now 16.lLJh5 i.h8 17.h4±
Zadar 2004
would have underlined this fact, but
White sticks to routine manoeu­
l.e4 c5 2 . lLJf3 e6 3.c4 lLJc6
vring and again falls under pres­
4.lLJc3 e5 5.d3 d6 6.g3 .ie7 7 .ig2 •

sure. This time he gets into trouble

f5
because of his weak light squares.

16 lLJe5
.•• 17. lLJc4 lLJxc4
18 .hc4 .ib7 19.�gl? ! .ie5
20.f3 lLJc8 21.lDh5 �e7 22 .ie3 •

�d6 23.-ifl .ic6 24.gd2 .ib5+

25.gadl ga7

Black is behind in development

indeed, but the closed centre saves
him from trouble. Practice has
shown that White is unable to re­
fute Black's design.

8.exf5 hf5 9. 0-0 �d7 10. lDd5

26.b3? lLJf6 11. lDg5 lLJxd5 12 .hd5 hg5
Avrukh suggests 26.i.d3 as the
only move (26 . . . lLJc4 27.i.c4 i.c4 12 . . . h6 13.lLJt7? fails to 13 . . . gf8+,
28.b3 i.bS 29.a4 i.c6 30.a5±), but but 13.lLJe4 is probably equal .
26 . . . i.d3 27.gd3 gd7+ still leaves
Black with the better game because 13.hg5 .�b4 14 .ie4
• 0-0
of the ridiculous knight at the edge. 15 . .ie3 a5? !
26 .ifl 27.�fl f5 ! (winning
••.

a pawn) 28.ef5 lLJxf5 29.gd3 We all know the AB C book, but

hh2 30 .if2 g4 31.f4 ®f7 32 .g3
• somehow tend to forget about it
�g6 33.�g2 �xh5 34.�xh2 gc7
35.gds gxd8 36.gxd8 gc6 37.a4
a5 38.gbs �g6 39.gg8 + �f7
4o.gbs gd6 4t.gb7+ �g6 42.�g2
hs 43.gbs �f7 44.gb7+ �f6
45.gbs gd2 46.®fl gd3 47.gxb6
�xg3+ 48.�g2 h4 49.hc5 gd2+
so.�gt �f5 st.gbs h3 s2.ghs
gdl+ 53.�f2 h2 0-1
in the heat of the battle. 15 . . . .ixe4 !
16.dxe4 �e6+ would have fixed the
pawns in my favour because my
knight would be more mobile. Per­
haps subconsciously I wanted to
keep more pieces on the board.
. 16.a3 (16.f4 ! ?) 16 . . . lLJc6
17 .id5+ �h8 18.f4
•

334
 Rare Lines on Move 3
18 . . . .ie6
It would have been better to
open lines : 18 . . . exf4 ! 19.hf4 l!Jd4+.
19 .'ea4 exf4 20 .bf4 �U5 •
21 .ie4 gf7 2 2 .gael .ih3 23.gf2
•
ges 24 .ie3 •
24.�dl ! wou ld have kept the bal­
ance because 24 . . . .ig4 (24 . . J�ef8 = ;
24 . . J'l:ff8 =) fails to 25 . .ixc6 ! gxel
26.�xel �xc6 27.�e4 !±.
24 gxf2 25 . .ixf2 Y9f7 26.�dl
• . •
l!Je5 27 .id5 Wf6 28 .d4? (28.
•
�d2+) 28 gf8- + 29 .ie3 �g4
••• •
30 .if4 g5 31.i.g2 gxf4 32 .ixh3
• •
fxg3 0-1
57. S.Williams-Delchev
London, 12.12.2013
l.e4 c5 2.�f3 e6 3 .b4
My first reaction to this move
was that I felt a bit scandalised and
my ego was ofended. Then I came to
my senses and pondered rationally
over the situation. The truth was
that I was not preapred against the
Sicilian Gambit. So finally I replied:
3 • • • b6 ! ?
This move was proposed by Da­
vid Bronstein in his book David
Against Goliath. It may not be the
most principled retort, but it is an
effective antidot to White's idea.
Black keeps control of the centre.
My wife, WGM Djingarova, played
3 . . . b6 back in 2 005 so I can say that
I was still in my preparation !
4.bxc5 bxc5 5.�a3
White had tried 12 ( ! ) difer­
ent moves, but the text seems to
be a novelty. The knight is head­
ing for c4-d6. Muzychuk-Djingaro­
va, Nova Gorica 2005, saw the most
logical S.l!Jc3 l!Jc6 6.gbl .ie7 7.i.c4
l!Jf6 8.0-0 0-0 9.gel d6 10 .d4 cxd4
ll.l!Jxd4 l!Je5 12 .i.fl i.d7+ and Black
can be satisfied with his position.
She achieved a Scheveningen for­
mation where white has weakened
considerably his queenside.
5 • • • i.b7! 6.e5 �e7 7.h4?!
I could not understand this
move. Instead of developing quick­
ly in this semi-open position, White
loses a tempo and weakens his
335
 Part 11
kingside. The natural continuation
is 7J�bl �d5 (7 . . . �e4 ! ? 8.d3 hf3)
8.�c4 a6 (preventing ltlb5) 9.0-0
ltlbc6.
7 •.• �f5 8 .td3
• 13 .te4 gc8 14 .bc6 dxc6
8 •.• .td5
I didnt want to close the diago­
nal to the bishop although 8 . . . ltlc6
9J:'!bl l'!b8 10.ltlc4 �a6 ! + would have
been fine. Another forcing contin­
uation is 8 . . . �d4 (removing a de­
fender of the e5-pawn and prolong­
ing the scope of my bishop) 9.ltlxd4
cxd4 10.\1;!fg4 \1;!fa5 ll.ltlc4 \1;!fd5.
9.c4
White creates a gaping hole on
d4 and restricts his own pieces ltla3
and �fl. Perhaps 9 .�5 maintained
the balance.
9 ••• hf3 10.\1;!fxf3 �c6
n . .tb2?!
336
It was simpler to trade piec­
es with ll.hf5 ltld4 12.\1;!fb7 l'!b8
13.�xa7 exf5 14.ltlb5 = .
l l . . . �fd4 1 2 .bd4 • cxd4
• •
White's set-up is a total mess. He
has not any co-ordination between
pawns and pieces. The a3-knight is
exceptionally useless. With my next
moves I convert my strategic advan­
tage into a pawn.
15.�c2 d3 16. �e3 �d4 17.0-0
\1;!fxe5 18.gfel
Or 18.ltlg4 \1;!fd4 19.l'!ael �d6
2 0.l'!e4 \1;!fb2 2 1.\1;!fxd3 l'!d8 2 2 .�e2
0-0+.
18 ••• .td6 19. �g4? !
This is a step in a wrong direc­
tion. It was better to try 19.g3 0-0
2 0 .ltlc2 WaS 2 1.�xd3 l'!fd8+ - the
pawns are equal, but all my pieces
are active.
19 \1;!fd4 20.c5 hc5 2 U : !acl
. . •
�d6 22.ge4 tlli2 23.gcel 0-0
(Game over) 24.�xd3 gfd8 25.\1;!fc4
�xd2 26.h5 �d5 27.�e2 f5
28.�e6 fxg4 29.�xg4 i.f8 30.h6
gc7 31.g6e5 �xa2 32.�g5 gbs
33.ges gxe8 34.gxe8 �al+ 0-1
Index of Branches

Taimanov l.e4 c5 2 )t)f3 e6 3.d4 cxd4 4.liJxd4 c!lJc6

5. tt:\xc6 278
5.c4 tt:\f6 324
5.tt:lb5 d6 6 ..if4 e5 7 . .ie3 a6 8.tt:\5c3 tt:lf6 9 . .ic4 249
9 ..ig5 250
9. tt:ld2 251
6.c4 tt:lf6 7.tt:l lc3 (7.tt:\5c3 252; 7.tt:ld4 253) 7 . . . a6 8.tt:la3 b6 (8 . . .
.ie7 253) 9 . .ie2 .ib7 10.0-0 .ie7 (10 . . . tt:l b 8 254) ll . .ie3 0 - 0 12.f4 255
12.�b3 257
12.f3 26 0
1 2 Jk1 262
5.tLlc3 �c7 (6.tt:ldb5 284; 6 .tt:\xc6 281)
6.�d3 283
6 . .ie2 a6 7.0-0 tt:lf6 8.a3 17
8.'it>hl tt:Jxd4 9.�xd4 .ic5 10 .�d3 b5 ll.f4 18
ll . .ig5 20
8 . .ie3 .ib4 9.tt:la4 .ie7 10 .c4 23
1 0 .tt:\xc6 25
6 . .ie3 a6 7. .id3 tt:lf6 8 . 0-0 tt:Jxd4 9 . .ixd4 .ic5 10 . .ie2 84
10 . .ixf6 85
10 . .ixc5 86
8 . . . tt:le5 9.tt:lf3 tt:Jeg4 89 (9 . . . d6 89; 9 . . . tt:Jfg4 89)
9.h3 .ic5 10.tt:la4 91
10 .�e2 93
10.'it>h1 94
7.�d2 tt:lf6 8.0-0-0 (8 .f3 ; 8.f4 53) 8 . . . .ie7 9.f3 (9.f4 54) 9 . . . b5
10 .g4 tt:Jxd4 11.�xd4 61
ll . .ixd4 .ib7 12 .g5 58 (12.�f2 57)
12 .'it>b1 6 0
12 . .id3 61

6.f4 a 6 7.tt:Jf3 128

7 . .ie3 129
7 . .ie2 133
7.tt:lxc6 �xc6 8 . .id3 b5 9.�e2 .ic5 137
9 . . . .ib7 139

337
6.g3 a6 7.i.g2 d6 8.0-0 i.d7 9.ll'lde2 163
9 .ll'lb3 164
9.i.e3 167
9.l'!e1 168
9.a4 169
9.ll'lxc6 i.xc6 10.l'!e1 170
10.a4 172
7 . . . h5 8 .h4 174 (8.i.g5 174)
B.h3 175
8.0-0 176

Kan l.e4 c5 2.ll'lf3 e6 3.d4 cxd4 4.ll'lxd4 a6

5.i.e2 29
5.c4 ll'lf6 6.ll'lc3 i.b4 7.i.d3 2 0 6
7.�d3 ll'lc6 8 .ll'lxc6 dxc6 9.�xd8+ 2 0 9
9 .e5 2 0 9
7.�c2 21 0
7.�f3 212
5.i.d3 g6 6.ll'lc3 214
6.c4 216
5 . . . ll'lf6 6.0-0 Vffc 7 7.Vffe 2 d6 8.c4 g6 9.ll'lc3 i.g7 10 .l'!dl 222
10.ll'lf3 223
10.i.e3 224
5.�c3 Y!lc7
6.i.e3 ll'lf6 7.i.d3 i.b4 65 (7 . . . i.c5 ; 7 . . . b5 65)
6.i.e2 b5 7.0-0 i.b7 8.l'!el ll'lc6 9.ll'lxc6 dxc6 10 .e5 l'!d8 ll.i.d3 c5 1 2 .Vffh 5 33
emg
1 2 .Vffg4 35
6.i.d3 ll'lf6 (6 . . . ll'lc6 98) 7.Vff e 2 99
7.0-0 d6 8.a4 1 0 1
B.Vffe 2 1 03
B.f4 1 05
7.f4 107
6.f4 b5 7.i.d3 i.b7 B.�f3 143 (8 .�e2 153)
6.g3 i.b4 7.ll'lde2 179
7.i.d2 ll'lf6 8 .i.g2 ll'lc6 9.ll'lb3 0-0 181
9 . . . i.e7 182

338
Anti-Sicilian Systems
l.e4 c5 2 . �f3 e6
3.c3 d5 4.exd5 �xd5 5.d4 lt:lf6 6.lt:la3 293
6 . .ie3 296
6 . .id3 297
6 . .ie2 298
3.d3 lt:lc6 4.g3 .it:lge7 5 . .ig2 g6 6.0-0 .ig7 7.c3 0-0 8.l:!e1 3 0 7
8.d4 3 0 8
3.b4 32 0
3 .b3 a6 4 . .ib2 lt:lc6 5.d4 322
5.c4 323
3.c4 lt:lc6 4.d4 324
4.lt:lc3 325
3 .lt:lc3 lt:lc6 4 . .ib5 lt:ld4 5.0-0 a6 6 . .id3 lt:lc6 7 . .ie2 327
7.l:!e1 327
3.�e2 328
3.g3 lt:lc6 329
3 . . . d5 33 0

339
 Fo rward
Chess

Most Chess Stars books are also available in the interactive electronic for­
mat ForwardChess. It is a free application which presents the books as they
look in print. Furthermore, it also provides a board which displays the text
moves or your own analysis. Read more information about Forwardchess
at:
http : I I chess-stars.com/Forwardchess.html

Here is the link to the iOS version:

https :/ /itunes.apple.com/us/app/forwardchess/id5430 05909?mt=8

This is the Android version:

https : I /play .google.com/ store/ apps/ details?id=com.forwardchess

You can also use Forwardchess on Windows desktops and notebooks

through the Android emulation from http : / / www . bluestacks.com/

340