Clless 9s

ttWH

T-UI; mON .AtTI

.nN .nNTI-L.nV P.wTOI

Chess Stars

www.chess-stars.com
Current Theory and Practice Series

The Modern Reti. An Anti-Slav Repertoire

Translation and editing by Semko Semkov

Cover design by Kaloj an Nachev

Copyright 2012 by Alexander Delchev

Printed in Bulgaria
ISBN: 978-954-8782-87-6
Contents

Foreword 5
Introduction 7

Part 1. Anti-QGA
l.lLlf3 d5 2.c4 dxc4 11

Part 2. Reversed Benoni

l.lLlf3 d5 2.c4 d4 25

Part 3. Anti-Slav and Anti-Chebanenko

l.lLlf3 d5 2.c4 c6 3.e3! rare 3d moves, ....if5, ....ig4, ...a6 set-ups 53

Part 4. Anti-Meran I
l.lLlf3 d5 2.c4 c6 3.e3! lLlf6 4.lLlc3 e6 83

Part 5. Anti-Meran II
4.lLlc3 e6 5.b3 lLlbd7 6.c2 .id6 7..ib2 with Ei:gl 95

Part 6. Anti-Meran III

4.lLlc3 e6 5.b3 lLlbd7 6.c2 .id6 7..ib2 0-0 8..ie2 115

Part 7. Anti-Queen's Gambit I

l.lLlf3 d5 2.c4 e6 3.g3 rare sytems; 3...dxc4 4.a4+ 147

Part 8. Anti-Queen's Gambit II

l.lLlf3 d5 2.c4 e6 3.g3 lLlf6 4..ig2 .ie7 5.0-0 0-0 6.b3 171

Index of Variations 211

Bibliography

Books

The English Opening, volume 2, M. Marin, Quality Chess 2010

Mastering the Chess Openings, volume 4, John Watson, Gambit 2010
Starting Out: The Reti, Neil McDonald, Gloucester Publishers 2010
The Dynamic Reti, Nigel Davies, Everyman Chess 2004

Electronic/Periodicals
Mega Database, Chess Base
Chess Informant, Sahovsky Informator
New in Chess Yearbook, Interchess
Chess Today

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

4
Foreword
In 2010, ex-world champion Anto
aneta Stefanova invited me to assist
her in preparing for the forthcom
ing FIDE Gran Prix series of tour
naments. My concrete task looked
relatively easy to me - I had to im
prove her "service", that is, the ef
fectiveness of her play with White.
Without sufficient familiarity
with top level women's chess and,
more importantly, with Antoane
ta's playing style and psychology,
I rushed to acquaint her with the
latest opening developments in the
most popular lines. My analyses
were detailed and promised a solid
edge in various topical variations.
However, this approach failed
altogether. We were ruthlessly pun
ished twice and she was generally
lacking in confidence in the most
principled openings. Although her
result was not a catastrophe (3-4th
place), for the next tournament we
decided to make a U-turn and em
brace another approach. This time,
we would be aiming to throw the
opponents out of their home prepa
ration and made them use their
own brains in unfamiliar positions.
We switched from move-by-move
memorisation to plan-oriented
thinking. This tactic proved to be
especially unpleasant for the Chi
nese girls, who were deprived of
their opening preparation and had
to fight on our ground. The result
was excellent - a victory in the last
game would have brought first prize.
Stefanova-Hou Vifan
FIDE GP Ulaanbaatar 2010
l.lUf3 lUf6 2.c4 e6 3.g3 dS 4.ig2
ie7 S.O-O 0-0 6.b3 b6 7.ib2 cS
S.e3 ib7 9.lUc3 lUc6 1O.cxdS lUxdS
l1.lUxdS 'lWxdS 12.d4 lUaS 13.lUh4
(13.dxcS!;!;, Part S) 13...'lWdS 14.dxcS
hh4 lS.gxh4 'lWxh4 16.hb7 lUxb7
17.c6 lUcS lS.'lWd4 'lWh6 19J''1adl EiacS
20.b4 lUa4 21.ial eS 22.'lWxeS 'lWg6+
23.@hl 'lWxc6+ 24.f3 f6 2S.'lWg3 Eif7
26.Eigl lUc3 27.Eid2 'lWc7
2S.'lWg2 (2S.'lWg4+ -) 2S...
lUbS 29.e4 (29.hf6!+ -) 29...lUa3
30.'lWg4 lUc4 31.Eid3 lUeS 32.heS
S
 Foreword
fxeS 33J'!gd1 l'!cf8 34.l'!d7 'lWc6
3S.l'!xf7 l'!xf7 36.l'!d8+ l'!f8 37.l'!xf8+
mxf8 38.'lWfS+ and a draw was
signed 20 moves later.
The Chinese super-GM who
eventually won the tournament and
the Grand Prix and went on to be
come a World Champion, was vis
ibly nervous during the opening.
She had to find a series of accurate
moves over the board, spent a lot of
effort and, not surprisingly, com
mitted some serious mistakes in the
middlegame, being short of time.
One final blow on move 28 or 29,
and Stefanova would have won the
tournament.
This experience has convinced
me that in the computer era one
might achieve better practical re
sults with a flexible opening stra
tegy, based on understanding of the
middlegames plans.
One year later I have decided to
present my analyses to the reader.
Many of them deal with positions
which are blank spots in opening
theory. In my opinion, the most in
teresting section is about the bayo
net attack, where Slav fans face a
head-on assault on their king after
l.tDf3 dS 2.c4 c6 3.e3 tDf6 4.tDc3 e6
S.'lWc2 tDbd7 6.b3 .id6 7..ib2 0-0
8.l'!gl
6
Instead of the usual struggle for
gradual equalisation in the Classical
Slav, Black has to solve urgent, very
concrete problems - how to avoid
a debacle in the next 10-15 moves.
Don't worry though, for less blood
thirsty players (of which I am one!),
I also cover the set-up with 8..ie2.
This leads us to the question:
For whom is this book written?
Club players have probably noticed
that their opponents as a rule are
well prepared against the central
openings l.e4/1.d4. If you are dis
appointed with your results, or just
tired of endlessly studying the lat
est analyses in the most explored
variations, you'll find here a viable
repertoire versus 1...dS. You might
also use my suggestions as surprise
weapons.
Note, however, that my work
has nothing in common with the
SOS-type articles. It was meant for
a top-level professional and this
repertoire is designed to serve for
many years. It is based on com
plex positions without early pawn
clashes in the centre. This shifts the
focus towards middlegame plans
and reduces the impact of home
brewed novelties. At the same time,
it is no less ambitious than the Si
cilian, or the Griinfeld, which were
the subject of my previous books.
Most of the material is fresh and is
not covered anywhere else.
Alexander Delchev
February 2012
Introduction
In 1923 Richard Reti introduced an
amazing new set-up:
Reti-Fischer
Vienna 1923
l.liJf3 liJf6 2.c4 e6 3.g3 dS 4.g2
c6 S.b3 liJbd7 6.b2 e7 7.0-0 0-0
8.d3 b6 9.liJbd2 b7 lOJk1 l'!c8
1l.l'!c2 cS 12.'!Wa1 d6
13.cxdS exdS 14.liJh4 l'!e8
1S.liJfS f8 16.liJc4 '!Wc7 17.liJce3
b8 18.xf6 liJxf6 19.1iJh6+ gxh6
20.xf6, with a big advantage.
Reti was a universal player who
had been successfully opening with
both l.d4 and l.e4, but in this game
he placed his faith in a flank stra
tegy. After the double fianchetto, he
continued his attack with pieces to
achieve a winning position in only
19 moves. At this stage he had no
pawns beyond the third rank!
This strategy brought him a no
table success in that year. He beat
Rubinstein and Tartakower, but his
biggest triumph was yet to come. In
the New York super-tournament of
1924, Reti used his system to crush
the reigning World Champion Ca
pablanca (his first loss in 8 years!),
the ex-Champion Lasker and the fu
ture Champion Alekhine. After that,
no one would dispute the name of
the new-born opening system.
Reti was the first to notice that
the mainstream openings were
running short of fresh ideas. Lasker
and Capablanca also thought that
chess would suffer a "draw death",
but they were too strong to worry
about that. Reti was the thinker
who pushed forward chess under
standing.
To be sure, his ideas were swift
ly put under the microscope. Black
has since discovered some solid set
ups which has taken the sting out of
his system to some extent. How
ever, White has been struggling
7
 Introduction
lately to achieve even the slightest
advantage in the Slav jMeran and
the Queen's Gambit. These open
ings have been explored in detail up
to move 30. This has brought about
the now frequent phenomenon of
some grandmasters' games consist
ing entirely of home preparation.
The players sit behind the board
and finish the game without having
made a single move of their own.
This approach requires most of all
a photographic memory. 90% of
the time for preparation goes into
opening analysis and tracking the
latest trends in the theory. Many
youngsters are willing to pay the
price in their pursuit of quick re
sults. The flip-side is that the flood
of information has made the latest
novelties available to all. Nowadays
even a very weak opponent might
catch you unawares and crush you
without a fight. That's why many
good players begin to revise their
approach. They seek flexibility and
surprise. That accounts for the
growing popularity of flank open
ings as the English and the Reti.
Frankly, I believe that it will be
more and more difficult and unre
warding to play the most principled
openings.
The modern Reti has signifi
cantly evolved in the last few years.
It has transformed into an English
Reti hybrid which is often used as
a tricky move order aimed at side
stepping certain systems. Instead
of employing a double fianchetto
versus any and every black set-up,
8
White often transposes to other
openings. For instance: 1.lLlf3 lLlf6
2.c4 e6 3.lLlc3 - Anti-Nimzo, or
1.lLlf3 lLlf6 2.c4 g6 3.lLlc3 - Anti
Griinfeld.
In this book I consider only the
pure Reti schemes which arise af
ter 1.lLlf3 dS 2.c4. Note that Sicilian
fans might prefer 1.c4 and turn to
2.lLlf3 in the event of 1...e6 or 1...
c6. Thus they would avoid 1.lLlf3 dS
2.c4 d4, which is by far the sharpest
Black's response.
The main section of my book is
the Anti-Slav set-up:
1. lLlf3 dS 2.c4 c6 3.e3!.
It borrows from Reti the idea
of delaying d2-d4. White does not
rule out this move permanently, he
merely waits for the best moment
for it. In my opinion, this approach
has no drawbacks. The pluses are
obvious:
We avoid the Slav because our
move order discourages an early
....tg4 or ....tfS owing to the possi
bility of \Wb3.
We avoid the main lines of the
Chebanenko System.
 Introduction

We rule out the Noteboom and which will be good if Black advan
other "triangle" variations. ces his b-pawn to h4.
My proposed repertoire would
We can always transpose to
have been vulnerable to move or
some mainline Meran, but we
der tricks had not I devoted two
should do so only rarely, when the
sections to l.ttJf3 d5 2.c4 e6. Now
resulting position is known to be in
3.d4 c6 would have been awkward
White's favour. My repertoire hard
so I consider 3.g3, with the mai
ly requires any knowledge of the
branches 3...dxc4 4.\'tla4+! and 3...
Meran. I examine instead a kingside
ttJf6 4.i.g2 i.e7 5.0-0 0-0 6.b3!?
pawn storm with 1"lgl and g4, and, as
an alternative, quiet development
with i.e2 and kingside castling. In
both plans, the delay in playing d4
enables dangerous tactical blows,
based on the latent power of the b2-
bishop. In the latter set-up, White
commonly controls the centre with
f2-f4, leaving the long dark diago
nal open.

The delay in playing d4 is also
very useful versus the Queen's
Gambit Accepted: l.ttJf3 d5 2.c4
dxc4 3.e3! ttJf6 4.i.xc4 e6 5.0-0 a6
6.\'tle2! c5 7.1"ldl!
I'm not against the Catalan, but
it has been heavily explored lately
and the character of the game is
more suitable for Kramnik than for
a club player.

In the diagram position Black

can choose to keep the tension in
the centre, when we should switch
to d4, or he can take on c4, e.g. 6...
c5 7.i.b2 ttJc6 8.e3 b6 9.ttJc3 dxc4
1O.bxc4 i.b7 11.\'tle2

We could have transposed to the

QGA with 7.d4, having sidestepped
the Romanishin Variation with
...i.g4, but the rook move contains
more venom. The main point is that
it keeps open the option of d2-d3!,

9
 Introduction
This variation is still inade
quately covered in opening books.
Mihail Marin advocates in his The
English Opening, volume 2 a plan
with d2-d4, but it is ineffective, for
many reasons.
I investigate the classic plan of
a kingside pawn storm with f2-f4j
g2-g4, also keeping h2-h4-hS-h6 in
mind. It leads to strategically un
balanced positions without forcing
variations, where the cost of every
move is higher for Black because
his king is in danger.
Of course, I also examine the re
versed Benoni set-up l.ttJf3 dS 2.c4
d4. This part is slightly out of step
with the rest of the book. I consider
three different options for White.
10
The emphasis, however, is on the
extremely sharp gambit 3.b4 f6
4.e3 eS S.cS as 6..ib5+!? c6 7..ic4.
I also analyse in detail 3.e3 ttJc6!
4.b4. I'm afraid that general consi
derations and plans would be use
less here. White aims to open up
the centre at any cost and tactics
should prevail over strategy.
Finally, I should like to stress
that most of the book is based on
my own original analyses. I have
found my sources to be mostly
unsatisfactory for the aims of this
book, so I had to develop my own
theory in many lines. You'll find
very few overlaps with other pub
lications. That should give you an
edge over your opponents.
 Part 1

Anti-Queen's Gambit Accepted

1 .f3 d5 2.c4 dxc4

11
Part 1
Main Ideas
1.f3 d5 2 .c4 dxc4 3.e3!
White's set-up is rather deceiv
ing. It looks like play will transpose
sooner or later to the QGA main
lines, but this is true only with an
important addition - we can choose
to enter only the main lines that are
better for White! One example is
the Steinitz system:
3 ... c5 4 .ixc4 f6 5. 0 - 0 e6
6.Wfe2! c6 7. gdl! i.e7 (7...a6
8.d4! b5 8.dxc5! ) 8.c3 0 - 0 9.d4
12
White gets a favourable IQP po
sition because 9...Wfc7 would give
him a serious lead in development
after 1O.dxc5! i.xc5 1l.a3!.
9 . . . cxd4 . 1O.xd4! xd4
1l.exd4 d5 12 .Wff3 xc3
13.bxc3 c7 14 .ib3
White is dominating all over the
board. He can develop his initiative
in the centre after 14...i.d6 15.c4!
i.xh2+ 16.<;!;>h1 .id6 17.c5 .ie7 18.i.f4
'Wc6 19.xc6 bxc6 20.i.a4 i.d7
21.gab1 gfd8 22.gb7 <;!;>f8 23.gxd7,
Gelfand-Adianto, Cap d'Agde 1998,
or attack on the kingside follow
ing 14... gb8 15.gd3! i.d6 16.'Wh5 h6
17.c4 b6 18.gh3.
Of course, most QGA fans do not
put their knight on c6, but prefer to
fianchetto their bishop first:
6 . . . a6 7. gdl! b5 8 . .ib3 .ib7
 1.tt:Jf3 d5 2.c4 dxc4

White as it leaves him with the bet

ter bishops. See game 2 Toma
shevsky-Ganguly, Moscow 2007,
where he finished off the opponent
with a kings ide attack.
Another version of this struc
ture arise after:
9...tt:Jbd7 1O.axb5 axb5 11.xa8
xa8 12.tt:Ja3! b4 13.tt:Jc4 ie7 14.d3!
9. a4!

The point of White's design is to

provoke ...b4 and then fix the back
ward pawn on c5 with d3. This is
the reason White has refrained
from d4 in favour of 7.d1. White
aims to reach this typical position:

9... b4 10.dS! ttlc6 1l.ttlbd2

'flYc7 12 .ttlc4 The open a-file could be only in
White's favour. Black can also try:
8 ttlc6 9.a4 b8 10.axb5 axb5
.

1l.ttlc3 c4 12 . .ic2 ttlb4 13.d3!

White's game is easier. His plan

is to build up a pawn barrier against
the b7-bishop with d3-e4, then de
velop the queen's bishop on f4. As The open files underline White's
a rule, trading the knights favours lead in development.

Points to remember:

White plays d1 before moving the d-pawn.

Then White meets ... tt:Jc6 with d4 and recaptures on d4 by knight.
If Black plays an early ...a6 and ...b5, White tries to provoke ...b4
and puts his central pawns to d3, e4.

13
Part 1

Step by Step

1..!lJf3 d5 2 .c4 dxc4 Black hopes to transpose to the

QGA with ...liJf6, ...e6. The move or
der is irrelevant for this set-up.

Alternatives are:

a) 3...b5?!
Black is unable to hold the gam
bit pawn s9 this move only wastes
time and create weaknesses on the
queenside: 4.a4 c6 5.axb5 cxb5 6.b3,
with an initiative.

b) 3...liJc6 4.i.xc4 e5

Instead of fortifying the d5-

pawn, QGA fans prefer an open cen
tre and easy piece development.
They might even argue that l.liJf3
has deprived White of the aggressive
option of l.d4 d5 2.c4 dxc4 3.e4!?
In fact, the delay of d4 brings
about positions of independent sig
nificance and enables many new
possibilities for White. Now we can follow up in the Si
cilian style with 5.'lWc2 i.d6 6.a3 liJf6
3.e3! 7.liJc3 0-0 8.b4, but:

3.liJa3 a6 4.liJxc4 b5 5.liJe3 i.b7 5.d4! is more unpleasant. Then

offers Black comfortable equality. 5...exd4 6.0-0! liJf6 7.exd4 only looks
like the QGA with 3.e3 e5, because
3 . . . c5 the knight is already committed

14
 l.liJf3 d5 2.c4 dxc4
to c6 at the expense of ...d6. Thus
Black cannot castle now and 7...e7
8.d5!t is gloom for him. Perhaps
that is why Ivanchuk and other play
ers prefer:
s...e4. Indeed, the position af
ter 6.liJfd2 \Wgs 7.g3 liJf6 8.liJc3 \Wg6
9.ltJd5 liJxd5 1O.xdsfs, Drozdovs
kij-Ivanchuk, Odessa 2007, is un
clear. However, White can hamper
Black's natural development by:
6.\Wb3!
Now 6...\wf6 7.liJes liJxes 8.dxes
f5 9.g4! \wf3 lOJglt would give
White the upper hand so Black is
forced to concessions - the knight
must go to the edge of the board. Af
ter:
6...liJh6 7.liJfd2 \Wgs (7...5 8.liJc3
.\kd7 9. ..Ike2) 8. ..Ikfl (This is only a
temporary retreat. The bishop will
return to an active position very
soon.) 8...fs 9.liJc3 liJb4 1O.liJbS,
White has the more active pieces.
c) 3...g4 4.xc4 e6
This is a clumsy attempt to
transpose to the Romanishin Vari
ation of the QGA. After s.\Wb3! ..Ikxf3
6.gxf3,
1.Il .tl.ll
jjj jjj
j
the thematic pawn sac 6 ... ltJd7?
does not work and Black must re
sort to the passive:
6.. :c8 7.d4 liJf6 8.liJc3 where
White has the bishop pair and a
strong centre. He could always
meet ...c7-c5 by d4-ds!, opening
play to his favour. For instance:
8...liJbd7?! 9.e4 a6 1O.e3 b5 (10...
c5 11.d5) 11.xe6!? fxe6 12.' xe6+
e7 13.e5. White's attack is worth
more than a piece here. More cle
ver is:
8...a6!, hoping for 9.e4?! bs
1O.e2 cs!, but 9. ..Ike2! anticipates
this idea (9...c5 10.d5). Play might
continue 9...liJbd7 1O.e4 c5 11.ds!
exds 12.liJxds liJxds 13.\Wxds \Wc7
14.e3 Eld8 (14...d6 15.0-0-0:1;)
15.0-0-0 ..Ike7. White can develop
his initiative with 16.4 0-0 17.e5
or 16.Elhgl g6 (16... 0-0 17. .\kh6 f6
18.\Wd6) 17.4 0-0 18.es.
d) 3... e6
I.Sokolov has been testing this
idea lately. Now 4.liJa3 c5 5.liJxc4
liJc6 6.b3 liJf6 7.b2 g6= has no bite
so White should go for the aggres
sive:
4.liJgs! ds s.e4!
1.Il 't!V .tI.ll
jjj jjjj
.t ttJ
j fJ,
fJ,/j /j /jfJ,/j
:ttJ'iV :
15
 Part 1

5....ic6 4 . .ixc4 f6 5 . 0 - 0 e6
5...e6 should be met by 6.exd5
(6.d4 looks interesting, but af
ter 6....ic6 7..ixc4 .ie7! 8.h4 h6
9.tt:lxe6 fxe6 1O.h5+ i>f8 11..ixe6
e8 12.f5+ tt:lf6 13.d5 .id7 14.e5
.ixe6 15.dxe6 'l'tfh5 16.xh5 tt:lxh5
17.g4 tt:lc6! Black gains the initia
tive.) 6...xg5 7.dxe6 e5+ 8.e2
tt:lc6 9.exf 7+ i>xf 7 1O.i>d1! xe2+
1l. .ixe2 tt:le5 12.b3! (12.tt:la3? .ixa3
13.bxa3 tt:lf6 14..ib2 tt:ld3 was in
Black's favour in Panchenko-Ro
zentalis, Lvov 1987.) 12...tt:ld3 (or
12... cxb3 13.f4 tt:lc6 14. .ic4+ i>g6 6.'l'tfe2!
15.f5+ i>xf5 16J!1+ i>e5 17.tt:lc3
tt:la5 18.axb3 tt:lxc4 19.bxc4 i>e6 Note this fine point. We are fol
20.tt:lb5 i>d7 21.1'!f7+ tt:le7 22..ia3 lowing the typical QGA scheme of
c6 23.tt:ld4 i>e8 24JU2) 13J1 tt:lf6 development, but Black cannot do
14.bxc4 tt:lxcl 15.i>xcl. the same. His set-up is connected
with...b5 which would offer White
6..ixc4 e6 7.d3! a lever for a queenside attack with
Black's bishop is extremely a2-a4. The delay of d4 enables new
clumsy at c6. It takes the best place possibilities which face Black with
for the Queen's knight and rules out some problems.
...c7-c5 which is an essential move
for controlling the centre. White Now Black chooses between:
will comfortably complete devel A. 6...tt:lc6 and B. 6...a6.
opment and push d3-d4 to domi
nate the board, for instance, 7...tt:lf6 6...tt:lbd7 has never been tested
8.tt:lc3 h6 9.tt:lf3t. Instead, in Gran in this position. The knight is pas
delius-I.Sokolov, Stockholm 2010, sive on d7 and that could be un
Black made another mistake: 7... derlined with 7.d4 a6 8.a4!, for in
.ie7?! 8.g4! tt:lf6 9.h3! d7 (9...h6 stance:
1O.tt:lxf7 i>xf 7 11.xe6++-) 10.0-0 8...c7 9.e4! cxd4 1O.e5 tt:lb6
tt:la6 when 1l.tt:ld2! would have as 1l..ib3 tt:lfd5 12.tt:lxd4t;
sured White of a big advantage. The
knight is heading for e5. Besides, 8...cxd4 9.exd4 .ie7 (9....id6
after 1l.a3 b5 12..ia2 tt:lc5 13.tt:lc3 1O.tt:lc3 0-0 1l.tt:le5t) 1O.tt:lc3 0-0
as 14..ie3 h5 15.Ei:fd1, Black's king 11.Ei:d1 tt:lb6 12..ib3 tt:lbd5 13.tt:le5
would also be very vulnerable. .id7 14.Ei:d3!.

16
 I.ltJf3 d5 2.c4 dxc4
A. 6 tOc6 7Jdl!
The point of White's strategy is
to be flexible and keep all the op-
tions open. Now 7...a6 will final-
ly see 8.d4! when 8...b5 would face
9.dxc5! - see line B.
7 !i.e7 8.tOc3 (or 8.d4 cxd4
9.tt:lc3!) 8 0 - 0 (8...a6 9.d4 b5
. .
1O.dxc5!) 9.d4
We have reached by transpo
sition the Steinitz Variation of the
QGA. It has been unpopular late
ly as it offers White a good version
of an IQP position. He has not lost
tempos on !i.d3xc4 like in the Nim
zo, nor on a4 (this comes in a set
with a hole on b4), like in one of the
main lines of the QGA. As a result,
White usually achieves to relocate
his queen to a more active place on
the kingside and gains a lasting in
itiative.
9 cxd4
After 9.. :c7 1O.dxc5! .ixc5
1l.a3! White is way ahead in deve-
lopment in this symmetrical posi
tion: 1l ...a6 12.b4 .id6 13..ib2 ltJe5
(13 ....id7 14.!!acl) 14.ltJxe5 .ixe5
15.f4! .id6 (15....ixc3 16..ixc3 b5
17..ixf6 gxf6 18 ..id3 .ib7 19.!!ac1
b6 20.@f2) 16.!!ac1 e7 17.c2
(or 17.d3 !!d8 18.ltJe4 ltJxe4
19.xe4, Schmittdiel-Aigner, Vel
den 2008) 17...b5 18.ltJe4 ltJxe4
19.xe4 .ib7 20. d4.
Let us examine now both op
tions:
AI. 1O.exd4; A2. 1O.ltJxd4!.
A1. 10.exd4
White hopes to use his knight
against the enemy king so he keeps
more pieces on the board. I sup
pose that Black should regroup for
defence, starting with:
10 . . tOd5!?
It is risky to divert the queen's
knight from the e5-square because
White will occupy it immediately:
a) 1O...ltJb4 1l.ltJe5 ltJbd5 12.!!d3!
lifting the rook to the kingside.
b) 1O...ltJa5!? 1I..id3 b6 12.e5!?
Avrukh's recommendation. It might
be insufficient for an advantage,
but the other continuations have no
bite. The queen is heading for g3.
12....id6 13.g5 g6! 14. g3. In my
opinion, Black can hold now with
14...lt:lh5! 15.g4 ltJg7 16..ih6 !!c8,
followed by ...It:lc6 or ...e7. Then
Black can defend with ...f6 or ...f5.
This idea is yet to be tested.
17

c) 1O...a6 1l.a3 tLlaS 12.a2 tLldS
13.tLleS:!;.
11.a3
11.b3 would not change Black's
plan with 11...f6.
11 . . .if6
Vacating e7 for the c6-knight.
12.V;Ye4
12.tLle4 removes a hit from dS
and allows Black to complete deve
lopment with 12...d7.
In Arutinian-Sprenger, Cap
pelle la Grande 2010, White failed
to achieve any advantage after
13.tLlxf6+ WExf6 14.WEe4 tLlce7 1S.d3
WEfS 16.WEh4 tLlg6 17.WEg3 (17.hfS
tLlxh4 18.xh7+ wins a pawn, but
does not pose any problems to
Black: 18.. .';t>xh7 19.tLlxh4 Elac8
20.d2 Elc2 21.Elabl f6.) 17...
tLlgf4 18.xfS tLle2+ 19.@hl tLlxg3+
20.hxg3, draw.
12 tLlce7
18
Part 1
Black's set-up is known from the
Semi-Tarrasch Defence. It is a bit
passive, but very solid. Black needs
only to develop his Queen's bishop
and then he can think about play
of his own. The game I.Sokolov
Legky, France 2004, went further
13.d3 tLlg6 (13... g6) 14.d2 b6
IS.WEg4 b7 16.tLle4 Elc8=.
A2. 10.tLlxd4!
Practical results seem to indi
cate that White's play after this re
capture is easier than after 1O.exd4.
The point is that the e2-queen now
gets free access to f3 or hS. This
continuation attracted attention af
ter the spectacular debacle Topa
lov-Kharlov, FIDE-Wch k.o. New
Delhi/Teheran 2000:
1O...tLlxd4 1l.exd4 WEe8 12.tLlbS
WEd8 13.tLlc3 WEe8 14.tLlbS WEd8
IS.f4 d7 16.tLlc7! Elc8
17.dS! Elxc7?
The only move was 17...exdSD
18.tLlxdS Ele8 19.WEd3 a4 20.b3
c6 21.tLlxe7+ WExe7 22.ElacU, with
a strong bishop pair.
18.d6 Elc6 19.dxe7 WExe7 20.b5!
Elb6 21.xd7 tLlxd7 22.WEd2! eS
 l.lLlf3 d5 2.c4 dxc4
23.ie3 lLlc5 24.b4 lLle4 25.'1Wc2
l'!xb4 26.a3+-, winning material.
10 lLlxd4
Symmetry does not help Black
due to his undeveloped bishop on
cB: 1O...'lWc7 1l.lLlxc6 'lWxc6 12.e4.
1l.exd4 lLld5 12 .'lWf3 lLlxc3
13.bxc3 'ifc7 14.ib3
The latest attempt of Black to
revive the variation. He prepares
...b6 or even ...b5. Instead, 14...id6
15.c4! ixh2+ 16.'i!7h1 id6 17.c5 ie7
1B.if4 'lWc6 19.'lWxc6 bxc6 20.ia4
id7 2U'lab1 E!fdB 22.E!b7 'i!7f8
23.E!xd7 was better for White in Gel
fand-Adianto, Cap d'Agde 199B.
15J;d3 ! .id6
15...b6 16.'lWh5 g6 17.'lWh6 id6
1B.E!h3 f6 19.id2 leaves the e6-
pawn very sensitive.
16.'lWh5 h6 17.c4 b6 18.l:!h3
White owns the initiative. De
spite the exchanges, his kingside
attack is looming, for instance: lB...
if4 19.'lWh4 ixc1 20.E!xc1 ib7 21.E!g3
E!fdB 22.E!d1 E!bcB 23.'lWg4 g6 24.ic2
'i!7hB 25.E!e1, when 25...'lWxc4? fails
to 26.'lWh4 'i!7g7 27.ixg6+-.
B. 6 . . . a6 7. l:!dl!
This idea of M.Marin from 1997
poses serious problems to Black.
White anticipates 7...b5 B ib3 ib7
when instead of the standard and
innocuous d4, he has the nasty sur
prise 9 a4!.
7. . . b5
7...lLlc6 8.d4 is in White's favour:
a) 8...b5 9.dxc5! 'lWc7 1O.id3
ixc5 11.a4! b4 (1l ...bxa4 12.E!xa4
lLlb4 13.ib5+:t) 12.lLlbd2 ib7
13.lLlb3 id6 14.id2 0-0 15.E!ac1
'lWe7 16.e4!, Kacheishvili-Krush,
Las Vegas 2009. White has a sig
nificant advantage due to his active
pieces and the weaknesses in the
enemy's pawn structure. After the
nearly forced 16...lLlg4 17.h3 lLlge5
18.lLlxe5 lLlxe5 19.ibl E!fcB, White
could have add also the bishop pair
advantage with 20.lLla5!.
b) 8...cxd4 9.exd4 ie7 1O.lLlc3
0-0 is a typically worse for Black
IQP position.
White can achieve a small plus
with 11.d5 or maintain the tension
with 1l.ig5.
19
 Part 1
See game 1 Wojtaszek-Buh
mann, Aix-les-Bains 2011.
8. .ib3 .ib7
B...ttJc6 9.a4 ):!bB 1O.axb5 axb5
Il.ttJc3 (1l.d4 c4 12.c2 ttJb4 13.ttJc3
ttJxc2 14.1!tfxc2 ttJd5 was unclear in
Campbell-B.Lalic, CCF 2011) 1l...c4
12.i.c2 ttJb4 13.d3!:! cxd3 14.i.xd3
ttJxd3 15.):!xd3 1!tfc7 16.e4 i.e7, Ras
mussen-Pina Vega, Havana 2011,
when 17.e5! ttJd7 IB.ttJe4 b7 19.):!c3
1!tfdB 20.ttJd6+ .ixd6 21.exd6 0-0
22.ttJe5 would have given White the
edge.
9.a4!
We see the cunning idea behind
7.):!dl. White has waited for ...b5 to
20
switch the focus of the game to the
queenside and force an important
positional concession from Black.
The big difference in comparison
with the Classical QGA is that 9...
ttJbd7 1O.axb5 axb5 1l.):!xaB 1!tfxaB
12.ttJa3! b4 will be met by 13.ttJc4
i.e7 14.d3! followed up by e3-e4. In
stead, Piket-Lautier, Amber-blind
fold, Monte Carlo 1995 continued
with:
12...i.xf3 13.gxf3 b4 14.ttJb5 1!tfbB
15.d4 i.e7 16.e4 0-0 17.e5 ttJd5 (17...
ttJh5 IB.f4 g6 19.d5 exd5 20.i.xd5
ttJb6 21.i.e4 ):!dB 22.):!xdB+ 1!tfxdB
23.ttJa7) IB.hd5 exd5 19.dxc5
ttJxe5 when 20.f4 ttJf3+ 21.1!tfxf3
1!tfxb5 22.):!xd5 wins a pawn.
9 . . . b4
Attempts to retain a flexible
queenside pawn structure face
Black with difficulties:
a) 9...1!tfb6
This was Black's choice in the
source game Marin-Kaeding, Sit
ges 1997.
White chose 10.d4, but the po
sition after 1O...ttJbd7 1l.ttJc3 b4
(1l...c4 12.i.c2 b4 13.a5 Wic7) 12.a5
1!tfc7 13.ttJa4 is unclear. Even more
challenging is 1O...c4!? 1l.i.c2
ttJc6 12.axb5 axb5 13.):!xaB+ i.xaB
14.ttJc3 i.e7 15.e4 ttJb4 16.i.bl 0-0
17.e5 hf31B.1!tfxf3 ttJfd5.
I think that White should try to
use the fact that the pawn is still on
d2by:
10.axb5!? axb5 1l.):!xaB haB
12.i.c2!?
 l.LUf3 dS 2.c4 dxc4
Enhancing the threat of LUa3.
Of course, 12.LUa3 c4 13.c2 is also
possible, when 13... LUbd7 14.d3
ha3 lS.bxa3 c3 16.e4 0-0 should
be pleasant for White due to his
centre and the bishop pair.
12...e7 13.LUa3 b4 (12...c4
13.d3) 14.LUc4 Wic7 1S.b3. White has
the better pieces.
b) 9...c4?! 1O.c2 Wib6 1l.axbS
axbS 12J'!xa8 xa8 13.d3!. This
possibility assures White of the ad
vantage.
10.d3!
Our plan is to build up a pawn
barrier against the b7-bishop with
d3-e4, then develop our queen's
bishop to f4 and put a knight on c4.
Then we could play on both flanks
thanks to our domination in the
centre. Generally, we do not mind
trading the knights as we will re
main with the better bishops.
10 . . LUc6
Or 1O...e7 1l.LUbd2 0-0 12.LUc4
as 13.e4 Wic7 14.g3 LUc6 1S.f4.
1l.LUbd2 Wic7
After 1l...LUaS 12.c2 e7
13.e4 Wic7 we can swap one pair of
knights: 14.LUb3 LUxb3 lS.xb3 h6
16.LUd2 (or 16.h3 0-0 17.e3!;j;), with
good prospects.
12 .c4
White's play is easier. See game
2 Tomashevsky-Ganguly, Mos
cow 2007.
21
Part 1

Complete Games

1 . Wojtaszek - B u hma n n 1 2 . . .tZl b4

Aix-I es - B ai ns 24.03.20 1 1
1 . d4 d5 2.c4 c 6 3 .tZlf3 tZlf6 4.e3
e6 5 . .id3 dxc4 6 . .ixc4 c5 7 .0-0 a6
8.e2 tZlc6 9J d 1 cxd4 1 0 .exd4
.ie7 1 1 .tZlc3 0-0
This is a very risky approach.
Black gets full control of the dS
square, but he gives too much free
dom to White's pieces in the centre.
Alternatives are:
a) 12...bS 13. .txf6 (13. .tb3 ttJaS
14. .tc2 .tb7 lS. .txf6 .txf6 16. .te4t)
13... .txf6 14.dS bxc4 lS.dxc6 Wc7
16.ttJe4 Wxc6 17.ttJxf6+ gxf6 18.E1.ac1
.tb7 19.E1.xc4t.
b) 12... ttJdS 13..txe7 ttJcxe7
14.ttJeS bS lS..txdS ttJxdS 16.ttJe4
.tb7 17 .ttJcS Wen.

1 2 . .ig5 1 3 .tZl e5 b5 1 4 . .ib3 .ib7

Another typical method of play

ing against the IQP is to prevent
ttJb4 by 12.a3. Then Black answers
12...ttJdS 13.We4 (13..ta2!?) 13... ttJf6
14.Wc2 ttJdS lS..ta2 .tf6 16.ttJe4,
with somewhat better prospects.
The isolated pawn is always a
double-edged asset so most White
players prefer to trade it gain
ing a small edge with 12.dS!? exdS
13.ttJxdS .td7 ! [13...ttJxdS 14. .txdS
Wc7 (14...We8 lS.E1.el) lS.We4 .tf6 This move is not too difficult to
16.ttJgS .txgS 17. .txgS] 14. .te3 ttJxdS find, but it is a novelty! Previously
lS. .txdS We8t. White had played lS.a4 or lS.E1.ac1.

22
 1.<'2lf3 dS 2.c4 dxc4

15..Jxf7 16.WI'xe61!9fB

White's initiative proves to be

very strong. 16.. :t!e8 17.a3 <'2lc6
18.ixf6 .bf6 19.<'2le4 V9xe6 20.Jtxe6
also favours him.

17Je1 ttJc6 1B.xf6 .ixf6

19.ttJe4 ttJa5 20.ttJxf6+
21.Wfxf7+ Wl'xf7 22.xf7+
23J'ac1
gxf6
'i!7xf7
We shall meet the same pawn
structure in Part 7, but White's bish
op is on g2 there. On b3 it is more
exposed to attacks (with <'2laS) and it
looks very passive, but on the other
hand, it may easily become the hero
of a crushing kingside attack.

10...ttJc6 11.<'2lbd2 c7

I think that Black should have

White has a clear edge. The gene
ral rule is that without queens, a
rook+two pawns are stronger than
a bishop+knight when the rooks
have enough open files.
23....id5 24.E:c7+ c;t>g6 25.b3
lLlc6 26.E:d7 ttJb4 27.E:e3 E:cB
28.h4 E:c1+ 29.'i!7h2 c6? (29...
Elal 30Jg3+ fS 3U'xh7 xa2
32.hS xf2 33.h6) 30.E:g3+ c;t>f5
31.E:xh7 ttJd5 32.h5 E:c2 33.h6 E:xf2
34.E:hg7 1-0
taken c4 under control with 11...
CLlaS 12.ic2 ie7. Thus he would
be able to change the pawn struc
ture (which is not in his favour as
it is now): 13.e4 (after 13.b3 <'2ldS
14.ib2 <'2lc3 Is ..\hc3 bxc3 16.<'2lb1
0-0 17.<'2lxc3 c7 Black is a pawn
down, but his pieces have plenty
of good squares.) 13 ... <'2ld7 14.<'2lc4
<'2lxc4 IS.dxc4 V9c7 16.eS d8 17.f4
0-0 18.<'2lgS g6. Of course, White re
tains the initiative, but at least his
attack is not too easy, e.g. 19.h4 h6
20.<'2lf3 hS.

2. Tomashevsky-Ganguly
Moscow 16.02.2007
1.<'2lf3 d5 2.c4 dxc4 3.e3 ttJf6
4.ixc4 e6 5.0-0 a6 6.1!ge2 c5
7Jd1 b5 B..ib3 .ib7 9.a4 b4 10.d3
12.<'2lc4 ttJg4?!
Black's idea to exchange the
knights does not spare him from
a kingside attack. He could have
tried 12....lie7 13.e4 0-0 14.aS (14.
gS [gadS lS.\We3 <'2lhS; 14.h3 :8ad8

23
 Part 1

lS..ie3 ttld7 16.Ei:acl ttlb6 17.ttlxb6 1 B .f4 !

xb6 18.c2 ttlaS 19.ttld2 Ei:c8=)
14...ttld7 1S.i.a4 Ei:fd8 16..ie3:;!;. An excellent positional decision!
White seizes the initiative on the
kingside. Black's problem is that
without knights he lacks any coun
terplay.

1 B . . .Y;Yc7 1 9 .9f1 .if6 20 .f5 e5

2 1 ..ie3 Y;Y e7 22.h4 0-0

This position is still very com

plex.

1 3 .g 3 ! ttlg e5 1 4.c!lJfxe5 c!lJxe5

1 5.c!lJxe5 Y;Yxe5 1 6 .e4 gcB 1 7 . .ic4
.ie7

Black is totally helpless. The fi

nal stage of the game does not need
comments.

23.gf2 <.t> h B 24.gaf1 gfdB 25.gg2

gd4 26 ..ixd4 exd4 27 .Y;Y h 5 gfB
2B.g4 g6 29 .Y;Y h 6 g 5 30.hxg5 1 -0

24
 Part 2

Reversed Benoni
1 . f3 d 5 2 . c4 d4

25
Part 2

Main Ideas

l . tLlf3 d5 2 . c4 d4 The truth is somewhat different.

The variety of plans White has tried
is due to the fact that he has not any
clear path to an edge.
Of course, that would not make
me write off the whole system so
easily. Most of the chess giants
have played it successfully (recently
Kramnik has crushed Topalov, for
example). However, White should
abandon the deceptive idea that he
could choose some standard Re
versed Benoni set-up with g3, g2,
lLla3, lLlc2 and outplay the oppo
nent. Practice has proved that such
According to my database, this
slow plans gradually lead White to
move is 6 times less popular than
a cramped position. Let us never
2...c6 or 2 ...e6, but in my opinion
forget that Black has a space advan
it is the most unpleasant retort to
tage and once he has consolidated,
2.c4. I have spent many months on
his game may turn to be preferable.
analysis, only to confirm my con
White has an extra tempo and
viction that White should hardly
he should use it to destroy the en
be better in a position where the
emy centre at any cost.
opponent has a space advantage.
And make it fast!
When I ran out of fresh ideas I even
Lately White has been living up
began asking for help other highly
to that understanding so the analyt
respectful Chess Stars authors, but
ical focus has been shifting towards
they also thought that Black should
extremely sharp gambit lines. They
be OK here. are still nearly unexplored, but
White's future is there. I'm sure
Many authors have spent plenty we'll be witnessing more top level
of ink to describe the multitude of games in variations like 3.M f6 4.e3
plans White has in his possession. eS S.cS as or even 3.e3 lLlc6 4.b4?!.

26
 l.LUf3 dS 2.c4 d4

In the next pages I will acquaint

you with the course of my research.
I will spare you many well trodden
paths which I think to be irrelevant
for contemporary theory. I will pay
attention only to the critical varia
tions according to my analysis.

First of all, I rejected the popu

lar plan 3.g3 LUc6! 4.g2 eS 5.0-0
IiJf6 6.d3 in view of 6... aS! or even
6...d6!?, with comfortable deve
lopment. The other old line:
3.e3 LUc6! 4.exd4 LUxd4 S.LUxd4
xd4 6.d3 also seems without
prospects. For instance: 6... c6! (A
modern approach. Black restricts
White's knight and plans to retreat
to d6, followed by long castling. )
7.ltJc3 LUh6!
This is practically a blank spot
in the theory. There are quite a few
games, but most of them are ab
solutely irrelevant and full of mis
takes. That is not without reason
though. The position is very unbal
anced and it hides many traps. I
have developed brand new attack
ing plans which should be a nasty
surprise to your opponents.

Perhaps the best risk/reward ra

tio offers the variation:

3.b4 f6 ! (3... cS?! 4.e3 f6 S.bxcS

eS 6.exd4 exd4 7.d3:t:LUc6 8.e2 and
3... g6?! 4.g3 g7 S.d3 eS 6.g2 !1Je7
7.0-0 0-0 8.a4!? as 9.bS cS 8J:'1a2!
:ga7 1 1.e3 dxe3 12.fxe3:t: should not
be of great concern to White. ) 4.e3
Black plays next ...LUfS, ...eS, eS S.cS as
with active pieces. This has been
topical lately so I consider it in
detail. According to my analysis,
Black is holding in the complica
tions after 8.e2LUfS 9.0-0 eS lO.g4
1tJh4 1 1.e3 d6, but a future game
might introduce some new idea for
White.
On the fourth move, however,
White has the enterprising pawn
sac 4.b4?! dxe3 S.fxe3LUxb4 6.d4: 6.i.bS + !?

27
 Part 2
6.tt'lxe5 is a forced draw while
6..ic4, as played by Kramnik, is not
entirely sound.
6 . . . c6 7.i.c4
I have found only three examples
in my database. White is threaten
ing with 8.tt'lxe5 which will be possi
ble after 7...ax b4. Other possibilities
are 7...g4 and the all hot 7...tt'le7
which was introduced in Game
3 Berczes-Prusikin, Bundes-
28
liga 2012. I analyse this important
novelty in detail in the "Complete
Games" chapter. Look also at line
B3 in the "Step by Step" chapter. I
think that it is useless to give here
some "main" line since the next 20
moves require an utmost precision
and should be learned by heart. My
general opinion is that White has
full compensation for the mate
rial deficit and the main question
is would Black survive or not. It is
extremely difficult to find over the
board all the tactical subtleties that
often hold together his defence.
Even if the opponent spent hours
on home preparation, he might fall
victim of misleading "suggestions"
by the chess engines. The future
will shed more light on this tricky
variation, but its current state is ab
solutely playable, with fair practical
chances for White.
Part 2
Step by Step
1.J dS 2 . c4 d4
A. 3.g3; B. 3.e3; c. 3.b4.
The SOS variation 3.c5?! has no
bite, for instance: 3...c6 4.a4
id7 5.xd4 e5 6.xc6 xc6 7.c2
d4 8.e3 xc5 9.c3 f6 lO.a3 a6
11.b4 d6=, B.Lalic-Dubreuil, Cap
pelle la Grande 2011.
A. 3.g3 c6!
3...c5 or 3...g6 offer White more
chances to develop an initiative in
the reversed Modern Benoni.
4 . .i.g2 eS s.o-o f6 6.d3 as!
This is more accurate than 6...
1e7 7.b4 ttJxb4 (or 7....i.b4 8.ttJxe5
ttJxe5 9.a4+ ttJc6 1O.xc6+ bxc6
11.xb4;!;) 8.ttJxe5 0-0. With a
pawn majority in the centre, White
has more possibilities to fight for
the edge than in the main line. The
recent game Kramnik-Topalov,
20th Amber Blindfold Monaco
14.03.2011, went 9.ttJd2 d6 1O.a3
ttJa6 11.f4! c6 12.Eib1 e7 13.tDdf3
tDg4 14.tDxd4 tDxe5 15.fxe5 xe5
16.e3 and White's central pawns
keep Black at bay, leaving him with
out a clear plan.
6...d6!? discourages 7.b4 by
defending e5. After 7.e3 (7.g5 h6
8.hf6 xf6 9.ttJbd2 0-0 1O.ttJe4
d8 11.tDxd6 xd6 12.a4 1d7
l3.a3 xa3 is probably slightly
better for Black, Ni Hua-Edouard,
Biel 2011) 7...0-0 8.exd4, the game
is balanced, for instance: 8...tDxd4
9.ttJxd4 exd4 10.ttJd2 Eib8 11.Eie1
ig4 l2.ttJf3 draw, Ni Hua-Pono
mariov, FIDE World Cup, Khanty
Mansiysk 2011.
After the text, White should
undermine the opponet's centre
because slowplay with 7.tDa3 c5
would allow Black to consolidate
and keep his space advantage:
8.ttJd2 0-0 9.ttJc2 f5 lO.b3 Eie8
11.1b2 e4+, Agrest-Salgado Lopez,
Ourense 2009.
29
 Part 2

7.e3 dxe3! Now the main line is B1. 4.exd4,

but B2. 4.h4?! may pose more prob
Straightforward and simple. lems to Black. The gambit could be
7...ii.e7 8.exd4 exd4 9.4Ja3 0-0 quite dangerous.
1O.4Jb5;t gives White active pieces.

8.he3 .ie7 9.d4 Bl. 4.exd4 xd4 5.xd4

Wxd4 6.d3 c6!
9.4Jc3 0-0 does not change
much the character of the position. A modern approach. Black re
Sooner or later White will have to stricts White's knight and plans to
push d4. retreat the queen to d6 and castle
long.
9...exd4 10.c!lJ xd4 xd4
11..ixd4 7.c3 h6!

1 1.Wxd4 0-0 12.4Jc3 c6 is totally

equal. Black's only weakness at a5
is easy to defend.

11... 0-0 12.c3 c6 13.c5 ii.g4

14.Wa4 Wd7=.

B. 3.e3 c6!

3 ...c5 4.b4 f6 is considered in

line C.

30
 l.tLlf3 d5 2.c4 d4
This manoeuvre is a potential
killer of White's set-up. The knight
will control from f5 all the impor
tant dark squares in the centre and
will be discouraging a bishop's sor
tie on e3. Evidently, White cannot
stand it for long so he should con
template g2-g4. It is still impossi
ble, though.
8.h3 looks like a reasonable
preparation, but practice has seen
Black at least equalising after 8 ...
liJf5 9.g4 e5+!, for example:
1O.liJe4 tLld4 1 1.f4 a5+ 12.d2
d8, Naiditsch-Bauer, 201 1 Mul
house 2011. That's why I prefer a
quick development:
8.e2 f5 9.0-0
9.g4 tLlh4 lOJ'1g1 d6+ IS too
risky.
9 ...e5 1O.g4 h4 1l.e3 d6
A critical position. I think that
White should not be better. He is
ahead in development, but concrete
analysis shows that he cannot reap
dividends from that. At the same
time, his castling position is com
promised and may tell in future.
12.e4
12.d4 would have been fine if
Black took on d4 ( 12 ... exd4 13.xd4
xd4 14.hd4:t), but the flank at
tack 12...h5! pinpoints the sore
point of White's set-up - the shaky
position of his king. I investigated
in my analysis 13.gxh5, but Black
has sufficient counterplay. Instead,
the game D.Gurevich-W.Watson,
Saint John 1988, continued 13.f3
d7 14.d5 g6. Here 15.g5 would
have kept the kingside closed, but
most players would prefer to be
Black anyway.
12...c7 13.f4 e7 14.f5
White would be better if he
could fix the c6-pawn while main
taining his kingside stable with f5
and h3. However, he lacks a tempo.
It is tempting to put in 14.c5, but
apart from 14...f5!?, Black also has
14 ... h5 15.d4 exd4 16.xd4 xg4
17.xg4 hxg4 when the forced se
quence 18.tLld6+ xd6 19.cxd6
d7 20.hg7 0-0-0 21.xh8 l"1xh8
22.d4 l"1h6 23.g7 xd6 24.l"1fd1
tLlf3+ 25.Wfl lLld2+ 26.l"1xd2 xd2
27.xh6Ieads to a draw.
14...h5 15.h3
The correspondence game Bo
ger-Persson, ICCF 2007, went 15.c5
hxg4 16.xg4 d7 17.h3, with a
tense fight. However, computers
suggest 15 ... g6!?oo.
31
 Part 2
15 . hxg4 16.hxg4 c5
Apparently, Black is holding
here, despite the somewhat hang
ing state of the h4-knight. Here are
some illustrative variations :
17.el .id7 IB . .if2 .ic6 19.f6
.ixe4 2 0 . fxg7 :9:gB 2 1 . dxe4 ttJg6oo ;
17.:9:f2 .id7 IB.:9:h2 .ic6 19.el
.ixe4 2 0 .dxe4 VNc6 2 1 .:9:dl VNxe4
2 2 .i>f2 g6 23 .hl gxfS ! (23 . . . xhl
24.:9:dxhl gS 2S . .if3) 24.xe4 fxe4
2S.gS f6 2 6.gxf6 hf6 27 . .ig4=.
B2. 4.b4?! (risky ! ) 4 . dxe3
4 . . . .ig4 gives White the initiative
without any material consolation
after S.bSltJeS 6 . .ib2 cS 7.bxc6 .ixf3
B.gxf3 ttJxc6 9 .f4.
5.fxe3
This gambit is quite tricky. It fac
es Black with some difficult choices.
Another point in its favour is that it
is completely unexplored.
5 ttJxb4
32
S . . . eS ! ? has never been played,
but it may be a decent alternative.
The position after 6 . bS ttJ b4 7.ttJc3
.ifS B.e4 .ig4 is strategically unba
lanced and unclear.
6.d4
6.VNa4+ ttJc6 7.d4 has no bite:
7 . . . .id7 B.b3 eS 9 .dS ttJ b4, Kozul
A.Petrosian, Slovenia 1994.
I have analysed after 6.d4 the
aggressive B 2 1 . 6 . . . eS, and the solid
B 2 2 . 6 . . . cS ! ? and B 2 3 . 6 . . . e6.
B21. 6 . e5 7.a3 c!lJc6
The first game in this variation,
Takacs-Rubinstein, Merano 19 24,
saw 7 ... e4 ! ? B.c!lJfd2 ltJd3+ 9.hd3=.
White should be more aggressive:
B.ttJeS ! ltJc6 9 .a4 .id6 1 O . ltJxc6
VNh4 + 1 l.g3 ,hg3 + 1 2 . hxg3 xhl
13 .bS ! ltJf6 14.ltJeS+ c6 IS.ltJxc6
.id7 16.eS+ .ie6 17.VNd6 ! bxc6
IB.xc6+ i>e7 19 . a4 l'!.hbB 20 . .ia3+
i>dB 2 1.ltJc3 .id7 2 2 .d6, with a
strong attack.
8 .ie2 e4 9.c!lJfd2
.
 l . lDf3 dS 2 .c4 d4
9 ... f5
The e4-pawn is important. After
9 .. :gS 10.lDxe4 'lWxg2 11.f3 'lWg6
[11. ..'lWh3 1 2 . lDbc3 h4+ 13.lDg3
(13.lDf2 d7 14.c2 lDf6) 13 . . . d6
14J"la2 lDf6 15.0-0 h3 16 .g2 d7
17.e4 lDg4 1 8.f3 0-0 19.eS e7
20.lDdSt] 1 2 . lDbc3 e7 13.lDdS
ih4+ 14.'i!?d2 d8 IS .b2 lDge7
16.tLlf4 White obtained a strong
initiative in Tymrakiewicz-Allicock,
Sunningdale 2 0 07.
10 .ib2!
. hh2 + ! . At the same time, h e dis
Of course we should not let the
black knight to eS as in Omearat
Raetsky, Abu Dhabi 2 0 l O : l O .dS
tLleS ll.b2 h4+ 1 2 .g3 e7 13.lDc3
tLlf6't'.
10 g5
This has been played in the
stem game Eklund-Brynell, Swe
den 2 0 0 9 . The queen's sortie is
undoubtedly the most challenging
retort to White's gambit, and the
engines love it. Still, I suppose that
the restrained development :
10 . . . lDf6 ! ? is more unpleasant.
11.0-0
White should be careful to avoid
a blockade of his central pawn pair.
It is much better to sac another
pawn with c4-cS , d4-dS-d6, than to
allow a blockade, as in the variation
l 1 .dS? ! lDb8 ( l l . . .lDaS ! ?) 1 2 . lDb3
lDbd7 13.0-0 lDcS 14 .d4 b6. We
should move one of our pawns only
if we are able to secure the march of
its neighbour.
1 1 . . . d6
First line of the engines. Black
creates threats like 12 . . . hS, 13 . . .
courages ideas with g2 -g4.
Alternatives are :
a) 1 1 . . .b6. By preventing c4-cS,
Black threatens to put his bishop
to d6 in full comfort. We have no
choice but force play by 12 .dS! lDb8
( 1 2 . . . lDaS 13.lDc3 e7 14.lDcxe4 fxe4
lS.lDxe4 0-0 16.d6 cxd6 17.lDxf6+
xf6 18 .dS + ) 13.JthS+ g6 14.lDxe4,
with an attack.
b) 1l . . . Jte7 ! ? The same re
strained tactic started with lO . . .
lDf6 . We should d o something
quickly before Black has castled:
12.g4 ! ? fxg4 13.lDc3 0-0 14.lDcxe4.
The game is strategically unbal
anced and unclear. Play might con
tinue 14 . . . 'lWe8 IS .dS lDb8 16.Jtd4 b6
17.e1 lDa6 18.g3co.
12 .cS ! Jte7 13 .b3 WdS
Dragging the bishop to c4 under
the knight fork. 13 . . . d7 14.lDc3
lDaS 1S .a2 is awkward to Black.
14.c4 d7
33
 Part 2
Is . .ic3 !
I have also analysed Is:a2 tLlg4
16 . .it7+ @d8 17.gel .ih4 1 8 .g3 .igs
19.tLlfl gf8 2 0 . .ic4 ( 2 0 . .ihS tLlf6
2 1 ..ie2 tLlds+; 20 . .ib3 tLl as 2 1 .h3
tLlf6 2 2 .ds '!Nt7) 2 0 . . . tLlaS 2 1 .h3
tLlf6 2 2 .ds '!Nt7 2 3 . tLl bd2 tLld7 24.d6
tLlxc4 2s.dxc7+ @xc7 2 6 . tLl xc4 .ie7
27J;acl .ixcs where the pin along
the a2-g8 diagonal is clogging
White's pieces. The text is slower,
but White keeps the extremely im
portant light-squared bishop. His
superiority in the centre promises
him full compensation for the pawn:
Is . . . tLld8 16.a4 tLlg4 17.gel .igs
18.tLlfl tLle6 1 9 . tLl a3 a6 2 0 .gadl .ih4
2Ule2 0-0 22 . .ib4 gd8 2 3 . ds @h8
24.ged2.
11.0-0 '!Nxe3+ 12.l!?hl
34
A spectacular position. For only a
couple of pawns White has achieved
a big advantage in development and
a mobile pawn centre. My analysis
suggests that the position is very
dangerous for Black, despite the
"opinion" of the engines :
12...c!lJf6
12 . . . tLlxd4? fails to 13 . .ihs + g6
14.tLlxe4 ! .
13.d5 c!lJe7
13 . . . tLld4 is already possible,
but White retains the attack af
ter 14.tLlxe4 ! tLlxe2 Is.tLlxf6+ gxf6
16 .gel @t7 17.gxe2 '!Nf4 18.tLld2 .id6
19.tLlf3.
13 ... tLlaS is more complicated.
We continue with 14.tLlc3 .id7 (14 . . .
.id6 Is . tLl cxe4 ! ! + - fxe4 16 . .ihs +
l!?e7 17.gel '!Nd3 18.cS .ixcs I9 . .ixf6+
gxf6 2 0 .'!Ncl .ie3 2 1 . '!Nxc7+ .id7
2 2 . tLlxe4 '!Nxe4 23.d6+ @e6 24 . .if3
'!Nf4 2s.gadl --+) 15. tLl bs !
15 . . . 0-0-0
Is ... hbs does not solve Black's
problems in view of 16 . .ixf6 gxf6
17 . .ihs+ I!?d8 18.cxbs '!Ngs 19.94 ! !
fxg4 2 0 . .ixg4 gg8 2 1 .h3 . Black's
 l . ctJ f 3 d 5 2 .c 4 d4

king will hardly find a safe haven.
The opposite coloured bishops hin
der his defence. One of the many
entertaining variations is 2 1 . . .h5
22.E1f5 g6 23.E1xh5 e3 24.ctJf3 ctJc4
2S.d6! ctJxd6 2 6.d5 and Black is
would have been gloom for him. I
prefer to complete development
before targeting something in the
black camp.
16 lDf4 (16 ... a6 17.d4 id7
.

helpless. 1B.i'!adl)

16.id4 h6 17.ctJb3 ! hb5

18.lDxa5 .

,*J: 1. i.
.t..t..t. .t..t.
I,i) \\IV
tb1. f'::, .t.
f'::,$,.t.

jL f'::,t<:,
'iV M
White's attack is strong. Il
17Jxf4!
lustrative variations are: 1B . . . ieB
19.b3 b6 2 0 .c5 ; 1B . . . id7 19.ctJxb7!
17.ib5+ c6 1B.g3 cxb5 19.E1xf4
wxb7 2 0 .b3 + 'it>cB 2 1 .i'!ab1 ; lB . . .
h5 is unclear.
a6 19.i'!b 1 ctJxd5 ! ? (or 19 ... i'!d6
20.ixa7; 19 ... id6 2 0 .h3 ctJd7 2 1 .c5
17 Y;lfxf4 18.lDb5 e3 19 ..ixf6
e2 2 2 .c6! b6 23 .xe2 ctJbB

gxf6 20.d6 cxd6 21.cxd6 .ixd6

24.E1xf5; 19 . . . ie7 2 0 . E1xf5) 2 0 .cxd5
22.lDxd6+ @e7 23.lDc4 .ie6
E1xdS 2 1 . ixa6 xa6 2 2 .ctJxb7 c4
24.Y;lfel a5 25.lDbxa5@f726.Y;lfb4
23.xa7! ! i'!xd 1 24.i'!fxdl.

White's two pieces are stronger

14.lDb3 lDg6 15.c5
16.lLlc3!
In the stem game Eklund-Bry
nell, Sweden 2 0 0 9 , White won after
16.ib5+ 'it>dB 17.i'!a2 f4? 1 8 .ctJ1d2
fS 19 .ctJa5 i'!bB 2 0 .c6 b6 2 1 .ctJb7+
WcB 2 2 .ctJc4 ig4 23.ixf6 gxf6
24.d6 cxd6 25 .lDcxd6 + 'it>c7 26.b3
e5 27.E1d2 ie6 2B.a4 a5 29.ctJxa5
d6 3 0 .ctJc4 hc4 31.Wa7+ 1-0,
but 17 ... a6 ! 1B.ic4 ctJg4 19 .id4 h6
Y;lfg5
than the rook, because Black's king
has no shelter.
B22. 6 . c5!? 7.a3 lDa6
This variation is underesti
mated, probably because of the pas
sive position of Black's knight at a6.
However, I like Black's position. It
is quite solid so the game cannot be
decided in the opening. White will

35
 Part 2

have to regroup his forces in order in both variations, but only practice
to prepare for a further advance in can tell which is the best option.
the centre. I tried to investigate in
deep the possibilities of both sides. a) 9)lJbd2 (9 .b2 lLlf6) 9 .ig7 .

Black seems to hold, but only with 10.gb1.!Llf6 1l .id3 0-0 12.0-0

active counterplay by . . . e6.

S.d5

There is nothing to gain by

maintaining the tension in the cen
tre. Black only gets additional pos
sibilities as B.d3 g6 9 . lLlbd2 cxd4
1 O . exd4 lLlcS ! .
Black's general plan i s . . . g6,
. . .g7, . . . lLlh6-fSjg4, for instance :
B.'IWa4+ d7 9.\1*lb3 bB I0 .dS g6
1 l .lLlc3 g7 1 2 . bl lLlh6 13.d3 0 - 0
14.0-000;
B.lLlc3 g6 9.dS 1J.g7 1 O .d2 lLlf6 The easy part has ended. Now
11 .d3 0 - 0 1 2 . 0 - 0 lLlg4. Black must define his plan. He can
either stay passive, manoeuvring
S...g6 to improve his pieces, or attack
the centre. The first approach of
fers White a lasting initiative on the
kingside :
12 . . . lLlbB 13.\1*lel lLlbd7 14.\1*lh4
\1*lc7 IS.b2 eB (1S . . . h6 will be im
mediately attacked by 16.g4 ! gS
17.lLlxgS .!LleS IB.\1*lg3) 16.c2 !

Now White should decide where

to put the aI-rook. At first sight,
bl looks the better place. The rook
hampers the break . . . b7-bS and re
stricts the cB-bishop. On the other The bishop is heading for a
hand, from a2 it can arrive faster at more active position at a4. This is
f2 or e2. White has compensation essential in variations like 16 ...b6

36
 l.ttJf3 dS 2.c4 d4
17.e4 ttJhS 1B..bg7 <t!?xg7 19.ia4
e5 2 0.dxe6 :Bxe6 2 1.ttJgS ttJdf6
22.4:lxe6 + he6 23.ttJf3;l;.
16 ...aS denies ia4, but leaves
the black king without support.
After 17.e4 eS 1B.dxe6 :Bxe6 19.:Bbe1,
White's attack is strong.
The most reasonable answer
seems to be:
16...h6, to cover the gS square
and make room on h7 for the f6-
knight. White leads the attack with
17.g4 ! ttJ h 7 (17...gS 1B.ttJxgS hxgS
19.xgS--+) 1B.ttJe4 hb2 19.:Bxb2
IiJdf6 2 0.xh6 ttJxg4 2 1.h4 fS
22.4:lf2 --+.
Obviously, Black needs counter
play in the centre:
12 e6! 13.e4 c!lJh5
13 ...b6 14.eS ttJxdS (14 ...ttJg4
15.e2) 1S.cxdS exdS 16.:Be1 ib7
17.4:lf1 inverts the roles when Black
has to prove that his 3 pawns bal
ance the opponent's extra piece.
14.e5 (14.e1 b6! ?) 14 .. exd5
15.cxd5 xd5 16..ic4
16 ..c6
Black has only moves here. The
knight on a6 might easily become
"a collateral damage" of the attack
on 0. 16 ...dB 17.e2 ttJc7 saves it,
but after 1B.ttJe4 White is domi
nating on the dark squares.
17.c!lJg5 .ie6 18.b3
Perhaps White has nothing more
than a draw after 1B...heS 19.ttJxe6
fxe6 2 0.:Be1 ttJf4 ( 2 0 ...id4+ 2 1.@h 1
ttJg7 22.xb7) 2 1.xb7 ttJh3+
2 2.@h 1 ttJf2 + 23.@g1=.
b) 9 ..id3 .ig7 10J::!a2 c!lJf6
11.0-0 0-0 12.'el
Shifting the queen to the
kingside. 12.e4 bS (12 ...ig4
13.ttJbd2) 13J'!e1oo also needs analy
sis. After the text, the typical 12 ...
e6 13.e4 :BeB 14.h4 exdS 1S.exdS
ttJhS 16.:Be2 id7 17.:BxeB+ ixeB
18.f2 offers White compensation.
Computers propose:
12....ig4!? 13.c!lJbd2 e6 14.e4
exd5
Or 14 ... aS 1S.h3 ixf3 16.ttJxf3
xe1 17.:Bxe1 ttJd7 18.eS. The
bishop pair and the overwhelming
advantage in the centre overweigh
Black's extra pawn.
15.cxd5 ge8 16.f2
37
 Part 2
This position is double-edged.
White has excellent compensation
for the pawn, for exampl e :
16 . . . c 4 17 ..bc4 lLlxe4 l S . lLlxe4
E1xe4 19.1Lld2 ! .id4 2 0 .lLlxe4
.ixf2 + 2 1 .E1axf2 .if5 2 2 . d 6 ! (22 .
.ib2 E1cS 2 3 .lLlf6 + 'i!tfS 24.lLlxh7+
'i!tgS 25.lLlf6 + is a draw) 22 . . . lLlc5
23. lLlxc5 ;
16 . . . .id7 17 . .ib2 lLlg4 l S .'IWg3 b5
19.hg7 'i!txg7 2 0 .h3 lLlh6 2 1 ..ib1 c4
2 2 .lLld4 'lWb6 2 3 . 'lWc3 'i!tgS 2 4 .a4 lLlc7
25.'i!th1 a6 2 6 . 'lWe3.
B23. 6...e6 7.lOc3
I believe that the pieces should
be developed on their most active
places. Accordingly, the knight
should go to c3 (and not to d 2 ) , and
the bishop to d3 (but not to e 2 ) .
7... lOf6 8.a3 lOc6
Let us elaborate a little over this
position. Obviously, White should
advance his central pawns in or
der to repel the main defender of
Black's king - the f6-knight. How-
3S
ever, 9 . e4 will be strongly met by
9 . . . e5 ! , favourably returning the ex
tra pawn. On the other hand, 9 . .ie2,
hoping for e4 on the next turn, is
too passive. Black will complete
development with 9 . . . .ie7 1 0 . 0-0
0-0 leaving us wondering what to
attack. In Fressinet-Istratescu, Le
Port-Marly 2 0 1 2 , White chose :
9 . .id3 . Then the mundane 9 . . .
.ie7 is not so comfortable: 1 0 . 0-0
0-0 1 l . 'lWe2 lLla5 12 . .ib2 c5 13.E1ad1
cxd4 14.exd4 a6, Munoz Moreno
M ancebo Ibanez, Aragon 1997,
when 15.d5 ! exd5 16.lLlxd5 lLlxdS
17.cxd5 .id6 l S . 'lWc2 h6 19. 1LleS
would have given White a tangi
ble initiative. However, Istratescu
surprised his opponent with 9 . . .
g6 ! and i t turned out that White
could not find even a trace of an
attack against the perfectly pro
tected king: 1 0 . 0-0 .ig7 1l . .ib2 0-0
1 2 . 'lWe2 b6 13.E1ad1 .ib7 14.'i!th1 'lWe7
15.e4 E1adS 1 6.d5 exd5 17.cxd5 lLlaS
l S . E1fe1 when best would have been
lS . . . E1feS+ . White' s set-up simply
does not work against g6. In my
opinion, the fianchetto of the dark-
 l . ttJf3 dS 2 . c4 d4
squared bishop should be attacked
by h2-h4. Naturally, White should
firstly carry on e3-e4-eS. Thus
9.ie2 begins to look not so bad - it
produces fine results in the event of
9 ... g6 1O.e4 g7 (lO . . . eS is a mistake
now due to 1 l .dxeS ttJg4 12 .'&xdS+
I1;>xdS 13 .gS+ e7 14.f4) 1 1 .eS
ItJd7 12.h4! 0-0 13 .hS ttJe7 14.hxg6
hxg6 1S.h6-7.
Unfortunately, after 9.e 2 , Black
can opt for the other set-up - 9 . . .
ie7! . I had t o find a multipurpose
move which would enable e3-e4,
but should also be effective against
9 ... ,ie7. Thus a novelty has been
born :
9J!bl!
We prepare a very original rook
lift i3b1-bS-gS.
9 ... g6 weakens f6 thus allowing
1O.e4 ! g7
1O . . . eS is slightly better for White
after 11.dxeS '&xd 1 + 1 2 . mxd1 ttJg4
13.liJdS l:!bS 14.me1 g7 1S.ttJxc7+
\!;>fS 16.e2 ttJgxeS 17.f4.
H.eS ttJd7
12 .h4! 0-0 (12 .. .f6 13 .hS fxeS
14.,igSt) 13.hS-7.
9...e7 10 .id3 0-0 11.0-0
b6 12Jb5!?
White suddenly brings the heavy
artillery closer to the enemy king.
The ensuing complications require
practical tests, but it is clear that
White has sufficient compensation
for the pawn. Black's position is not
e nviable. Too many lines lead to a
debacle . The engines suggest:
12".i.a6
It is risky to remove the knight
from the centre with 12 . . . ttJaS (hop
ing for . . . cS) 13.l:!gS h6 14.l:!g3 d6
1S.ttJeS b7 16.e4 ttJeS:
:i 'ir :i *
iAi i i
i Ai i
tD
!'!:, !'!:, !'!:,
8 tD
.t 'iV
Our plan has triumphed. One
last precise prophylactic move and
Black should be unable to defend :
17.mh1! xeS (17 . . .f6 1S.g4
fxeS 19 .xe6+ mhS 2 0 . ,ixh6+-)
39
 Part 2

IB.dxeS Wfd4 19.ttJbS WfxeS 2 0 . .if4 This is the most fashionable

WfcS 2 1 ..ixh6. continuation. Its main lines are
very sharp and require exact knowl
13.l'g5 h6 14J:g3 .id6 15J:h3 edge of forced variations.

3 .. f6!

Let us investigate the alterna

tives :

a) 3 . . . g6? ! . Black spends a tem

po to develop his bishop to a pas
sive square. White should enjoy
some initiative after fianchettoing
his own bishop and opening the
centre with e3. Perhaps he should
firstly stabilise his space advantage
on the queenside: 4 .g3 .ig7 S.d3 eS
15..ic8 6 . .ig2 ttJe7 7.0-0 0-0

lS . . . .ib7 could be met the same

way: 16 . .ibl ! l'!eB (16 . . . ttJ aS ? 17.Wfc2
ttJxc4 IB.ttJgS + - ) 17.Wfd3 ttJ aS IB.e4
eS 19.ttJdS ..... .

16..ic2! e5 17.l'!h4 exd4

18.exd4 ge8 19.b5

White is clearly better, for in

stance : 19 ... a6 2 0 .ttJxd6 Wfxd6 B.a4 ! ?
2 1 .Wfd3 bS 22 . .ixh6.
This improves o n B.ttJbd2 as !
(Any delay of this break leads to a
difficult position as in Miroshni
C. 3.b4
chenko-Gasanov, Poltava 2 0 0B,
which went B . . . h6? ! 9 . ttJb3 gS 1 O . a4
ttJg6 l 1.bS ttJd7 1 2 . aS a6 13 . .ia3 l'!eB
14.ttJfd2 fS lS.Wfc2 9 .ttJb3 gS 1 O . a4
ttJg6 1 l.bS ttJd7 1 2 .aS a6 13 . .ia3
l'!eB 14.ttJfd2 fS lS.Wfc2.) 9 .bS cS
1 O .bxc6 ttJexc6 1 l . .ia3 ttJb4 1 2 .ttJel
ttJBa6 13.ttJc2 l'!eB 14.hb4 axb4
lS.a3 bxa3 16.l'!xa3 l'!bB=, Damlja
novic-Sedlak, Subotica 2 0 0B.

40
 l .lLlf3 dS 2 .c4 d4

I think that White cannot aspire 6.e3 if5 7.ib2.

to the initiative without breaking in
the centre with e2-e3 .
8 . . . aS 9 .bS cS 8 . !!a2 ! !!a7 1 l . e3
dxe3 12 .fxe3

12.ixe3 is also pleasant for

White.

12 . . . b6 13.lLlc3 !!d7 14.lLlel

This position has not been test-

ed yet.

The alternative on move 4 is risky:

4.bS cS

After 4 .. .f6? ! S.e3 eS (or 5 . . .

dxe3 6.fxe3 e S 7.lLlc3 ie6 8 .d4
14 .. .fS ib4 9 .id2 lLld7 1O .a3) 6.exd4 exd4
(6 . . . e4 7.1!tfe2 e7 8 .lLlh4 fS 9.g3)
If Black waited passively with 7.id3 icS 8 .e2 + White is better.
.f6, White would put a knight on
..

dS and prepare g4, h4. S.e3

IS.e4 f4 16.!!af2 (or 16 .ih3 ! ?

gd6 17.hc8 xc8 1 8 . !!g2;!;) 1 6 . . .
E & .t .t;li) !:
..

i i i ' i
gd6 17.lLldS gS . Here in Markus
Brkic, Banja Vrucica 2 0 0 9 , 1 8 . d4 ! !
i l!, i
would have come out of the blue.
l!, i
b) 3 . . . aS t::, tjj
l!, t::, t::, l!, t::,
Black aims to define the pawn tjj iV i. :
structure on the queens ide. That
would give him a free hand in the S . . .f6 !
centre where he has a space advan
S . . . g6 is known to be in White's
tage. He will try to put in both . . . cS
favour since the game Euwe-Ale
and . . . eS. Practice does not shed
khine, Amsterdam 1926, which
enough light on this variation. I
went 6.exd4 ig7 7.d3 cxd4 (7 . . .
would suggest to open the centre
ixd4? 8 . lLlxd4 xd4 9.1!tfc2 ! ,
quickly with :
Euwe) 8 .g3 lLld7 9.lLlbd2 ltJcS
4.1!tfa4 + ! id7 (4 . . . lLld7 or 4 . . . c6 1O .lLlb3 b6 11.lLlxcS 1!tfxcs 12 .ig2
should be met by S.ib 2 ) S .bS cS LDh6 13.0-0 0-0 14.a4 ! .

41
 Part 2
6.exd4
If White missed the moment for
this capture, Black would have the
option of recapturing by the eS
pawn.
After 6.g3 eS 7.d3 i.d6 8 .i.g2
0,e7 9.0-0 0-0, it is unclear what to
do with the queenside pieces.
6.id3 eS 7.c2 is dubious due to
7 . . . ie6 ! . Black completely ignores
the hit on h7 and he is even plan
ning to use the open h-file against
the enemy king: 8.!xh7 0,e7 9.0-0
c8. The good half of White's pieces
are stuck useless on the queenside.
6 . . . cxd4 7.i.d3 !
After 7.0,h4 0,h6 8.d3 0,f7 9 .g3
0,d7 1O .ig2 0,cS 1 1 . 0-0 gS 12 .0,f3
ifS Black is dominating in the cen
tre, Grigoryan-Javakhishvili, Aix
les-Bains 2011.
7 . . .eS 8 . 0-0 id6 ! 9 .0,h4 0,h6 ! ?
1O .hS+ 'kt>f8 , with unclear play.
c) 3 . . . cS . Choosing the Closed
Benko with colours reversed. Reti
believed that White should answer
4.e3 ! . I consider the position after
4 .. .f6 below, following the move or
der with 3 . . .f6. Alternatives are:
4.bxcS 0,c6 S.g3 eS 6.d3 !xcS
7.ig2 0,f6 (Black had a fine game
even after 7 . . . h6 8.0-0 0,f6 9.ia3
!xa3 10.0,xa3 0-0 11.0,d2 e7
12.0,c2 ifS 13 J'!bl, Kramnik-Tka
chiev, blitz Moscow 2009, when the
best move is 13 .. J''! ad8 to protect d4
in the event of 14.f4 e4.) 8.0-0 0-0
9 .ia3 .ba3 10 .0,xa3 1fS=;
42
4.ib2 . The bishop has no busi
ness here. It is better to trade it
via a3 . 4 .. .f6 S.d3 eS 6 . bxcS !xcS
7.0,bd2 0,c6 8 .g3 0,h6 = .
4.e3 e5!
This variation is a tough nut to
crack.
Note that 4 . . . dxe3 ? ! S .fxe3 eS
6.cS i.e6 7.ibS+ c6 8 .ia4 0,h6
9.0-0 (9.ib3 d7 1 0 . 0-0 0, a6) 9 . . .
i.e7 1 0 .ib3 (l0.d4 ! ? 0,d7 1 1.ib3)
is much better for White, for exam
ple: 10 . . . d7 11.0,c3 (or 1l.d4 0, a6
1 2 . bS cxbS 13.0,c3 !Xb3 14.xb3 b4
lS. 0, e4) 1l . . . 0,a6 ( 1 2 . a3 0, c7 13.d4
0-0-0, draw, Korotylev-Kramnik,
Moscow 2 0 07) 12 .d4 ! .
4 . . . cS is seldom seen. Then
S.exd4 cxd4 6.cS as ! ? 7.i.c4 axb4
8.b3 e6 9.0-0 !xcS has turned
well for Black in Mastrovasilis-Dao
Thien Hai, Khanty Mansiysk 2 0 10.
White should opt for:
S.bxcS eS 6.exd4 exd4
The most aggressive continu
ation here is 7.id3 , but it can be
counterattacked with 7 . . . i.xc5
8 . 0 - 0 0,e7 9.c2 0,bc6 1 0 . ia3 a5
11.!xcS xcS 12 J'!el
 l . lLlf3 ds 2 . c4 d4
12 . . . lLlb4 ! ? Now the endgame
after 13.a4+ .id7 14.b3 lLlxd3
1S.xd3 0-0-0 16.xd4 xd4
17.ttJxd4 lLlc6 looks fine for Black.
So we choose the quiet :
7.d3;!; lLlc6 B . .ie2
Or B.g3 ! ? hcS 9 . .ig2 e7+
1O.M1 .ie6 1 1 . lLlbd2 fidB 1 2 .fib1 d7
13.e2 lLlge7 14.lLle4 b6 1S.lLlxcS
bxcS 16 . .ia3 ; B.lLlfd 2 ! ?
B . . .hcS 9 . 0-0 lLlge7 1 O .lLlbd2
0-0 11. lLl b3
White has a very good version
of the reversed Benko Declined.
The e-file is open, the black pieces
are unstable. For instance, after
1l . . .id6 1 2 . fib 1 b6 ( 1 2 . . . .ic7 13.fie1
b6 14 . .ib2 lLlfs 1s .g3 as 1 6 . .ifl a4
17.ttJbd2;!;) 13.lLlbxd4 lLlxd4 14. lLlxd4
txh2+ 1s.mxh2 xd4 16 . .if3 fibB
17.l"le1, White's bishop pair is ruling
over the board . Instead, the game
Azmaiparashvili-Stefansson, Mos
cow 1994 went 1l . . . b6 1 2 .lLlxcS (or
12.l"lb1 as 1 3 . fie 1 d6 14 . .ifl lLlg6
1S.ttJxcs bxcs 16.lLld2 fs 17.g3 .id7
18.ig2;!;, Pirc-Kostic 1936) 1 2 . . .
bxcS 13.lLld2;!;.
S.eS
Black's set-up is based on the
fine point s.exd4? e4! and White
is already worse: 6.e2 (6.lLlh4
was killed in the game Reshet
kov-Bogdanovich, Odessa 2 0 1 0 ,
which went 6 . . . xd4 7.lLlc3 e3 !
8.hs+ g6 9 .fxe3 xc3 1 O .lLlxg6
xa1 1 1 . lLlxhB+ mdB 1 2 . ds+ lLld7
13.xgB xc1+ 14. mf2 cs ! 1s.bxcs
xd 2 + 16 . .ie2 mc7 17.fid1 c2
1B.fids lLles 19.h3 fs+ 0-1.) 6 . . .
e7 7.lLlg1 lLlc6 B.e3 lLlxb4 9.md1
lLlh6 1 0 .lLlc3, Summerscale-Lane,
Torquay 2 0 0 9 , when 1O . . . .id7 ! , in
tending 1 1 . xe4 .ia4 + , would have
assured Black of the edge.
The gambit idea S.cS is White's
main hope for an advantage. It is
coming into fashion and I would
like to add my two cents in its de
velopment. Black will have to show
brilliant defence in order to survive
the attack.
a) s.b3 is more restrained, but
it has no bite. Sakaev showed the
best set-up for Black:
S . . . cS 6.bxcs lLlc6 7.exd4 exd4
(7 . . . lLlxd4 B.lLlxd4 xd4 9 . .ib2
e4+ 1 O .e3 xe3+ 1 1 .fxe3 hcS
12 . .ie2 lLle7 13.d4 exd4 14.exd4
.ib4+ 1s.lLlc3;!;)
43
 Now White's dark-squared bish f3.) 13.f3 f7 14.fxe4 fxe4 IS.bS and
op has not any prospects. After: the e4-pawn should perish after g3 ,
192 .
8.1a3 f5! 9.1e2 lLlf6 10.0-0 1e7
1l.d3 lLld7 12.lLlbd2 lLlxcS 13.c2
(13.,txcS ,txcS 14.bS 1b4 ISJiabl
1c3 16.lLleS a6 17.a4 bS 18.lLlxc6
bxa4 19.1Llxd8 @xd8 =) 13 . . . 0-0
14.lLlb3 lLle6 (14 . . . b6 IS.lLlfd2 1b7
16.1f3 lLle6 17.aeU) IS.he7 xe7
16.fel 1d7 17.1f1 d6, the game is
balanced.
8 .1d3 hcS 9.0-0 lLlge7 10 .1a3
b6 was also equal in Tomashevsky 8 .lLlxc6 bxc6 9 . f3 fS 1 O . fxe4 fxe4
Sakaev, Serpukhov 2 007. 1l.lLlc3 lLlf6 1 2 .g3 e7 13.1g2 1e6
14.a4 1d7 1S.0-0.
b) S.a3 cS 6.exd4 cxd4 7.d3 as is
fine for Black.

c) S.bS is a computer move. Black

can chose between S . . . cS 6.1d3 1d6
7.c2 g6 8.exd4 exd4 9.0-0 lLle7=
and S ... lLlh6 6.i.b2 cS 7.1d3 if5
8.hfS lLlxfS 9. 0-0 i.e7 1O.c2 lLld6
1l.d3 0-0 12.lLlbd2 lLld7= .

d) S.1b2 could be answered by

S . . . cS= , transposing to 4 . . . cS. How
ever, Black has no reason to decline
the pawn sac. After S . . . dxe3 ! 6.fxe3
ixb4 7.a3 (7.cS lLlc6 8 .1c4 !xcS
9.0-0 lLlh6) 7 . . . 1e7 8:c2 lLlh6 Main branches are:
9.1d3 g6 10.lLlc3 c6 he has the bet Cl. 6.lLlxe5 = ; C 2 . 1c4; C3. 6 .1bS+ .
ter game.

5 . . . a5 ! Ct. 6.xe5, with a forced draw

S . . . d3? ! leaves Black overex
tended: 6:b3 ! e4 7.lLld4 lLlc6
Or 7 . . . aS 8.lLlc3 ! fS 9.lLle6 e7
1O .a4+ f7 1l.lLlxf8 @xf8 12 .1a3
lLlf6 (After 12 . . . axb4 13.xa8 lLla6
14.c6 b6 IS.g3 ! bxa3 16.i.g2 lLlf6
17.0-0 White breaks through with
after 6 . . . fxeS 7.h5+ d7 8 .f5+
@e8 9.h5+ , Van Wely-Kramnik,
Nice 2008.
C2. 6 . .ic4? ! axb4 !
This greedy move challenges
White to prove that his attack is

44
 1.'Df3 dS 2 .c4 d4
worth two pawns. The blitz game
Kramnik-Aronian, Moscow 2009
saw the weaker 6 . . . dxe3 7.fxe3 axb4
8.d4 'Dc6 9.0-0 'DaS 1O. 'DxeS 'Dxc4
11.lLlxc4 e6 12. 'Dbd2 'De7 13 .b2
d5 14.f3 bS 1S.cxb6 'Dxb6
16.lLlxb6 cxb6 17.c6+ d7 18 .dS
ie7 19.'Dc4 bS 2 0 . 'DeS+-.
6 ... 'Dh6 ? ! is also dubious: 7.0-0
i.e7 (7. . . d3 8.a4+ 'Dc6 9.'Dc3 fS
1O.bS 'Db4 1l.a3 'Dc2 12 .gb1 xcS
13.lLlh4 g4 14.h3 hS 1S.dS
\!{fS 16.c4 and the d3-pawn
should fall; 7 . . . axb4 8.exd4 exd4
9J'!e1+ e7 1O.b2 'Dc6 1l.a3 bxa3
12.lLlxa3+-) 8 .bS ! hcS (8 . . . g4
9.h 3 hS 1O .a3) 9.'Dxd4 ! xd4
1O.hS+ , winning.
7.exd4
Alternatively, 7.ib2 hcS and
now:
a) 8.'DxeS fxeS 9.hS+ 'it>d7
1O.exd4 is fighting for a draw with
10 ... ixd4 1l.hd4 exd4 12 .dS+
rlie7 13.f7+ 'it>d6 14.dS + . Black
could keep on playing by 1O . . . e7 ! ?
(but not 1O . . . 'Df6 11.e2 ! hd4
12 ..bd4 'Dc6 13 .ib2 ge8 14.0-0).
b) 8.'Dxd4 'De7 9.'De6 he6
1O.xe6 d3 11 .c1 d6 12.c4
'Dbc6 13.0-0 'Dd8 14.g4 dS
1S.c2, Chuchelov-Schenk, Muel
heim 2010, when 1S . . . d6 ! would
have put White's compensation to
the test.
7 exd4 8. 0-0
.
White admits the fact that he
has no concrete threats and aims
for a positional compensation.
8.b3 puts the - battery in the
scope of Black's queen's knight: 8 ...
'Dh6 9.0-0 ixcS lO.d3 (lOJ'!e1+ 'it>f8
11.a3 'Dc6 12 .b2 'DaS) 1O ... 'Dc6
1l.ge1+ 'it>f8 12.'Dbd2 (12 .hh6 gxh6
13.'Dbd2 'it>g7) 12 . . . 'DfS ! 13 .c2 gS
Apart from the two extra pawns,
Black also has a space advantage.
8 . . . .b:c5 9.d3
White must quickly create con
crete threats to avoid ending up
in positions like on the above dia
gram. However, in my analysis I
have found more than enough de
fensive resources for Black.
a) 9.ge1+ 'it>f8 (9 . . . 'De7!?) 1O.d3
(1O .c2 'De7 11.b3 b6! 12 .d3 b7
4S
 Part 2
13.li.Jbd2 .AdS+) 1O . . . li.Jc6 11.li.Jbd2 g6
12.li.Jb3 .ib6 13.h3 Wg7+ .
b) 9. li.Jh4 g6 1O.d3 li.Jc6 lU!el+
fS 12.li.Jd2 li.JaS! .
c) 9.c2 li.Je7 10 . .AbS+ c6 11.VNxcS
cxbS 12.li.Jxd4 1ltldS ! 13.xb4 li.Jbc6
14.li.Jxc6 li.Jxc6+.
9 ... li.Jc6 !
9 . . . li.Je7? ! 10.li.Jbd2 li.JdS lU!e l +
fS 12.li.Jh4 gives White a n attack.
An essential move. 1O . . . li.Jge7
1l.li.Je4 .Ab6 12 J!el li.JaS allows the
tactical hit 13.li.JfgS ! fxgS 14.VNhS+
d7 lS.,ixgS with an initiative, for
example: lS . . . li.Jxc4 16.dxc4 aS
17.cS Wc6 lS.cxb6 wxb6 19 .VNh4
eS 20 .li.Jd2;t.
1l.li.Jb3 li.Jxc4 12.dxc4 b6
13.lOfxd4 h:d4 14. li.Jxd4 li.Je7
15."Bh5+ g6 16."Bf3 tfxd4 17 .Ae3
ga3 18.h:d4 19.9xf3 m
46
White has landed in a murky
endgame.
C3. 6 .Ab5 + ! ?
I like this move. I n my opinion ,
it leads to very tangled positions
which are difficult for both sides.
Sharp players should feel in their
element here.
6 c6 !
The game Iturrizaga-Efimov,
Khanty-Mansiysk 2 0 1 0 , saw:
6 . . . li.Jc6? ! 7.exd4? e4 S.dS when
S . . . li.Jge7! 9.li.Jc3 exf3 1O.xf3 ax b4
1l.dxc6 bxc6 would have been
a big improvement over S . . . dS .
However, the novelty:
7.0-0! gives White the upper
hand:
7 . . . axb4 (7 . . . li.Jge7 S .bxaS xaS
9.a4 .ig4 1O .d3;t) S.exd4 exd4 (S . . .
e4 9.el; S . . . dS 9 . d3 exd4 10.el+
li.Jge7 1l.a3 b3 12 .li.Jbd2 xcS
13 .VNxb3 .Ad7 14 . .Ab2t) 9.el+ .Ae7
(9 . . . li.Jge7 1O .b3) 1O . .ib2 WfS 1l.a3
bxa3 12.li.Jxa3 hcS 13 .cl, with ex
cellent compensation.
7 . .ic4
 l.tLlf3 d5 2 .c4 d4

8 . . . e4

Or 8 . . . axb4? ! 9.exd4 hf3

1 O .xf3 xd4 1 1 . b3; 8 . . . lLlh6
9.exd4 xf3 1O.xf3 xd4 11 .b3
axb4 12 .b2 xc5 13.d4.

9.h3 hf3

9 . . . h5 1O .g4 exf3 l 1 .gxh5 axb4

1 2 . xf3 xc5 13 .b2 dxe3 14.d4 fa
vours White.

1 O .gxf3 axb4 11.b3

7 lLIe7! ?
1 1 .a3 hc5 1 2 . fxe4 e7 13 .b2
A very original move, aimed xe4 14.axb4 !!xal 15.xal id6
against lLlxe5. Alternatives are : 16.xd4 g6 + 17.'it>hl f5 18.'it>g2
is a draw while 1 l.fxe4 lLla6 ( 1 l . . . d3
a) 7 . . . g4
1 2 . a3) 1 2 .ib2 tLlxc5 is unclear, with
Played in Nikolaevsky-Savchen full compensation for White.
ko, Kiev 1995 when 8 . exd4? loses to
1 l . . . lLlh6 1 2 . exd4 xd4
B ... hf3 9 ."\1*'xf3 'I1;Vxd4 1 O . b3 xal
11.0 - 0 d4 1 2 .xg8 !!xg8 13.xg8
axb4- + .
8.b3 i s also dubious: 8 . . . tLlh6D
9.bxa5 [or 9.exd4 hf3 10.gxf3
(1O.xf3? xd4 11.c3 'I1;Ve4+ 1 2 .'it>fl
axb4 13.xb4 d4 14.b2 hc5)
10 ... xd4 1 1.b2 f4 1 2 . a3 tLlf5+]
9 ... (9 . . . hc5 1O.exd4 hd4
l1.tt:lxd4 xd4 12.lLlc3) 1O.gxf3
'fNc7! ! t 13.a4 !!a7.
Here 13.xb4 xc5 14.xb7
8.0-0 ! xc4 15.xa8 0-0 16.b7 d4
17. lLl c3 exf3 18.b3 gives White a
slightly better endgame with an ex
change for a pawn .
13.!!el leads to more tangled
positions with full compensation
for White since 13 . . . xal 14.ib2
xa2 15.!!xe4+ e7 16.!!xe7+
'it>xe7 17.e3 + 'it>d8 18.ixa2 !!xa2
19.d4+ tLld7 2 0 .xb4 !!a7 2 1 .lLla3t
is dubious for Black.

47

b) 7 . . . dxe3 8 . fxe3 axb4
Peralta-Cafolla, Port Erin 2 0 0 2 ,
saw 8 . . . .ig4 when simplest would
have been 9 . b3 tLlh6 1 O . bxaS E!:xaS
(l0 . . . tLld7 11.d4) 1 l . tLl c3 . Another
possible line is 8 . . . e4 9 .tLld4 axb4
when 1 O . a3 provides White with
good compensation. For example,
1O . . . tLlh6 1 1 . 0-0 .ig4 1 2 . c2 hcS
13.axb4 E!:xa1 14.bxcS .
9.d4. In this complex position
White has compensation for the
pawn .
c) 7 . . . axb4 8.tLlxeS tLl h 6 !
O f course, 8 . . . fxeS 9 .hS+
'i!?d7 1O .fS + 'i!?c7 1 1 .xeS+ 'i!?d7
12 . .ie6 + 'i!?e8 13 . .ixc8+ is not in
spiring for Black.
9 .tLld3 tLla6 1 O . hS+ g6 1 1 .e2
dxe3 1 2 . .ib2 .ie7
13.xe3 !
13.dxe3 tLlxcS 14.tLlxcS
1S.tLld2 e7 16.f3 is unclear.
13 . . . tLlfS 14.e4 tLlg7 (l4 . . . 'i!?f8
1S.ha6 E!:xa6 1 6 . 0 - 0 ) 1S.ha6 E!:xa6
16.tLlxb4 E!:a4 17.a3 0-0 18.c2 E!:aS
48
Part 2
19.d4 tLle6 2 0 .c4 'i!?g7 2 1 . 0 -0 E!:e8
2 2 .E!:eU. A rare case where White
has an extra pawn !
8.0-0! ?
8 .bS i s not in the spirit o f the
variation. Play would be unclear
after 8 . . . tLlfS 9 .c2 (9.e4 tLlh6
1 O .d3 tLlt7 1l.bxc6 tLlxc6oo) 9 . . . dxe3
1 0 .fxe3 .ixcS 1l . .id3 tLle7.
8 . . . axb4 9 . .ib2
This position is the cutting edge
of the theory in the 6 . .ibS + ! ? line.
I expect new interesting develop
ments very soon. The only definite
thing I dare say is that White has
full compensation for the pawn
and his game is easier as he is the
attacking side. It is possible, how
ever, that with a perfect defence
hcS
Black may be able to keep the bal
ance. See detailed commentaries
of Game 3 Berczes-Prusikin,
Bundesliga 2 0 1 2 in the "Complete
Games" chapter.
Pa rt 2
Complete Games
3. Berczes - P rusiki n
Bu ndes liga , Baden 22.0 1 .20 1 2
1 .f3 d 5 2 . c4 d 4 3 . b4 f6 4.e3
e5 5.c5 as 6 . .ib5+ c6 7 ..ic4 e7
8.0-0 axb4 9 . .ib2 ! (9 .b3 ttJdS)
9 . . . ttJa6?
After this thematic move, Black's
game is clearly worse. Critical is:
9 . . . dxe3 10.fxe3 ttJfS (lO . . . ttJdS
1l.ttJxeS) 11.e2 hcS 12 .d4 i.a7!
(In many variations Black's king es
capes through b6.)
After long exhaustive analyses,
I am still unsure what is White's
best way to proceed in this posi
tion. He has full compensation for
the pawns and his attack at many
points looks decisive, but somehow
the engines miraculously succeed
to hold Black's position together
with a series of only moves. There
are too many branches, however, to
reach a definite assessment. Let us
see what future practice will show.
Meanwhile, I think that over the
board White's game is much easi
er. He seems to always have some
kind of perpetual check in reserve,
should Black display a perfect de
fence. I have analysed three main
branches :
a) 13.a3 ? ! e4
13 . . . e7 14.axb4 i.e6 IS.i.xe6
xe6 16.dxeS fxeS 17J'!xa7 l'!xa7
18.e4 ttJh6 19.ttJxeS is unclear,
but only Black can be better: 19 . . .
b 6 2 0 . ttJd2 ttJt7 2 1 .ttJg4 0 - 0 ( 2 1 . . .
ttJ d 7 2 2 .hg7 l'!g8 23.ttJf6+ ttJxf6
24.hf6oo) 2 2 . ttJf6+ = .
14.ttJel ttJd6 lS.i.b3 ttJa6 16.ttJc2
e7 17.hS+ g6 (17 . . . 'it>f8 18.axb4oo)
18.e2 l'!f8 19.ttJxb4 ttJc7+.
b) 13.dxeS ttJd7!
49
 Part 2

13 . . . tDxe3 loses to 14.exf6 gxf6 An incredible position. Even

lS.'i!;>h1 Wie7 16.l"le1 'i!;>d8 17.tDbd2
'i!;>c7 18.tDf1 l"le8 19.tDxe3 .b:e3
2 0 .i.c1 + - .
14.tDbd2 fxeS !
14 . . . Wie7 lS.tDd4 tDxd4 16.exd4
tDxeS 17.l"lae1 favours White :
more incredible is that it may turn
to be equal with perfect defence
from Black!
a1) 17.tDxc8 + ? ! Wixc8 18.'i!;>h1 Wic7
19.f2 'i!;>d8 2 0 .l"lfe1 tDxc4 2 1.Wih4+
'i!;>c8 2 2 . Wixc4 bS 23.Wixb4 l"le8 24.a4
i.c5 2S.Wig4 l"lxa4 26.l"lxa4 bxa4
27.Wixa4 Wib7+'.
a2) 17.tDxeS tDxeS 18 . .b:e5
tDxc4+ 19.'i!;>h1 tDxd6 2 0 . l"lae1 i.e6
21.i.xg7 Wig8 2 2 .l"lf6 'i!;>d8 23 . .b:h8
'i!;>c7 24.l"lxe6 Wixh8 2S.l"lxd6 'i!;>xd6
26 .e7+ 'i!;>dS.

17 . . . i.g4 (or 17 . . . 'i!;>d8 18. 'i!;>h 1

tDg6 19 .Wid1 Wic7 2 0 .dS i.d7 ( 2 0 . . .
i.b8 2 1.g3) 2 1 . tD e4 'i!;>c8 2 2 .d6 aS
23.i.xf6 gxf6 24.tDxf6) 18.Wie4 Wid7
19.'i!;>h1 0-0-0 2 0 .dxe5 fS 21.e6 ! , with
an initiative. Here is an example:
2 1 . . .Wixd2 2 2 .Wib1 l"lhe8 23 .e7 l"ld7
24.i.c1 Wic3 2S.i.e6 l"lexe7 26 . .b:d7+ Apparently, White has no more
l"lxd7 27.l"le8+ l"ld8 2 8 . l"lxd8 + 'i!;>xd8 than a perpetual here.
29 .i.gS + t 'i!;>c7 30.a3 ! ? i.e2 31.i.f4+
a3) 17.tDf7 tDxc4+ 18.'i!;>h1 Wig8
'i!;>b6 32 .l"le1 i.d3 33.Wixb4+ xb4
19.tD3gS (19.tD3xes tDdxe5 2 0 . tDxe5
34.axb4 'i!;>bS 3S.l"le7 i.d4 36.l"lxb7+
tDe3 2 1.l"lf7+ Wixf7=) 19 . . . tDxb2
'i!;>c4 37.i.dM.
2 0 . l"lae1 i.b8 21.Wixb2 h6 2 2 . tDxe5
lS.tDe4 tDxe3 (lS . . . e7 16.tDfgS) tDxe5 23.l"lxe5+ .b:eS 24.WixeS+ i.e6
16.tDd6+ 'i!;>e7 2S.Wic7+ 'i!;>e8

50
 l.ltJf3 d5 2 .c4 d4
Again, a perpetual check is
looming after 26.xb7 hxg5 or
26.lLle4.
b) 13.lLlbd2 e7 (13 . . . ltJd7
14.lLlg5) 14.ltJe4 lLld7 15.dxe5 (15.
ib3 @dB 16.ltJfg5 fxg5 17Jxf5
exd4 IB.lLlxg5 d3 19.xd3 xe3 +
20.xe3 .be3+ 2 1 . @hl .bg5
22.gxg5 geB 23.gxg7 ge2+) 15 . . .
tiJxe3 16.ltJd6+ @dB 17.@hl
17 . . . @c7 (17 . . . ltJxfl IB.ltJf5 fB
19.9xfl @c7 2 0 .e6) IB. ltJd4 .bd4
19.1xd4 ltJxfl 20.ltJf5 dB 2 1.e6
i>b8 22 .gdl gaS 23.xfl (23.exd7
gxf5 24.dxcB+ is unclear because
Black's rooks are too active.) 23 . . .
tiJb6 24.e7 c7
25.ltJxg7 (After 25 . .bb6 xb6
26.lLld6 ge5 27.ltJf7 gxe7 2 B .ltJxhB
the knight cannot join the other
white forces.) 25 . . . xe7 26 . .bb6
ge5 27.f3 xg7 28.b3 g4
29.a4 gaS 30 .i.xa5 b5 31..bb5
cxb5 3 2 .xb5+ i.b7 33.gg1 gcB = .
1 0 .exd4 exd4 1 1 Je1
11.ltJxd4 ltJxc5 12 .h5+ ltJg6
13.gel+ @d7 14.a3t also retains the
initiative.
1 1 . . .tZlxc5
1 2 .tZlxd4
White has opened the centre
without significant material losses.
This is a sure sign that he is on top.
12 . .bd4 b5 13 ..bc5 bxc4 14.i.xb4
@f7 15.ltJc3 lLld5 16.i.xfB ltJxc3
17.cl gxfB 1B.dxc3 would have
been better for him, but the text
keeps more tension . Now 12 . . . lLla4
is insufficient due to 13.b3 lLlxb2
14.xb2 (14.i.f7+ @d7 15.lLle6
a5 16.xb2 ltJd5 17.lLlxfB+ gxf8
IB.i.xd5 cxd5 19.d3) 14 . . . b6
15.a3 @dB 16.i.e6, with a strong at
tack.
1 2 . . . 'i!Yd6 1 3 .a3! tZla4 1 4.'i!Yb3
tZlxb2 1 5.'i!Yxb2 d8 1 6.axb4 xa 1
17 .'i!Yxa 1 'i!Yxb4 1 8 . .1e6 c7
51
 Part 2
1 9 . tLlc 3 ?
Berczes has been playing per
fectly so far and he needed only one
more accurate move to finish off
the opponent : 19.ttJa3 ! (defending
the d4-knight ! ) 19 . . . 'i!fb8 2 0 .,ixc8
ttJxc8 2 1 . ttJxc6 + .
19 . . . b S 20 .Axc S ?
White was either i n time trouble
or he thought that everything was
52
winning. 2 0 .ttJe4 c5 2 1 .ttJb3 .lxe6
2 2 .ttJbxc5 .lc8 23J'!:bl Wfc4 24.d3!
was decisive.
20 . . . tvxd4! 2 1 . .ih3 f5 22 .tva5?
This mistake turns the tables.
2 2 .,ixf5 Wfa7 23 .Wfbl ttJxf5 24.Wfxf5
was still in his favour.
22 . . .g6 2 3 J: a 1 .ig7 24.g 3 tLlc8
25.Ag 2 tva7 26 .tvxa7+ tLlxa7
27 J:e1 tLlcS 2S.f1 E:dS 29 .e2
E:eS+ 30.d 1 E:xe 1 + 3 1 .xe1
Axc 3 32.dxc 3 c7 33.e2 d6
34.e3 tLl e7 35 ..if1 tLl d5+ 36.d4
c5+ 37.d3 tLlf6 3S .f3 b5 39 .d2
c4 40.e3 c5 4 1 .d2 tLld5
42 ..ie2 b4 43.cxb4+ tLlx b4 44.Ad 1
d4 45 . .ia4 g5 46 . .id7 f4 47.gxf4
gxf4 4S . .if5 h6 49 .Ah7 tLl d 5
50 . .ig 6 tLle 3 5 1 .e2 c 3 52 . .if7
c2 53 ..ig 6+ b2 54 . .id3 c 3 0-1
 Part 3

Anti-Slav
Anti-Chebanenko

53
Part 3
Main Ideas
1.c!DfJ d5 2.c4 c6 3.e3 !
Our aim i n the opening i s to
reach the following position :
Here I advocate a rare and very
aggressive approach, connected
with a kingside attack by g2-g4. I
will also examine a more restrained
development where White also de
lays d2-d4, but does not burn bridg
es behind him and castles short.
The problem is how to steer the
opponent into our desired set-up.
In this part I examine several sys
tems where Black avoids typical
Meran structures.
54
3 !Dd7 4.ttJc3 ttJgf6 5.c2 e5
hardly deserves much attention,
because White obtains a good cen
tre and the initiative after 6.cxd5
ttJxd5 7.d4 .td6 8 . .td2 ! ? , preparing
0-0-0.
By playing 2 ... c6, Black has
clearly displayed his intention to
lead out his bishop to f5 or g4. Our
move order is aimed exactly against
that. This thematic development
does not work here because of the
weak d5-pawn and the possibility of
White's knight to jump to d4 or e5
with tempo. For example :
3 J.f5 4.b3 b6 (4 . . . c7
.
5.cxd5 cxd5 6.ttJc3 e6 7.a4+ ttJc6
8.ttJd4 .tg6 9.b3;!;) 5.cxd5 xb3
6.axb3 cxd5 7.ttJc3
Black is losing a pawn.
3 J.g4 4.b3 c7 5.ttJe5 .te6
6.d4 ttJd7 7.ttJc3 dxc4 8.,ixc4 hc4
 l.ctJf3 d5 2 .c4 c6 3 . e3 !
9.llJxc4 ctJgf6
E - .t E
" 'if. " "
, .
ti:J 6
'iV tLJ ::,
::, jj ::'
g :
1O.a4 e6 1 1 .e4;t.
Perhaps you should remember
that we meet any bishop sortie with
b3 , without exchanging on
dS. This m ight be useful i n some
lines with . . . '!!1 c 7.
3 ...lLlf6 4.lLlc3 oig4 is an at
tempt to improve on the previous
line. 5.'!!1b 3 '!!1 b 6 6 . ctJe5 .ie6 7.d4
White has a space advantage on
the queenside. H e can develop his
initiative with ctJa4.
3 g6 4 . ctJ c3 .ig7 5.d4 ctJf6
6.e2 transposes to the Schlech
ter System which is too passive for
contemporary tastes.
3 e6 4.b3! f5 is another pas-
sive set-up, based on the Stonewall
System .
A t least Black can argue here
that he is exploiting White's 3 .e3.
I ndeed, the king's fianchetto is a
more popular weapon against the
Stonewall. H owever, I adhere to
Petrosian's view that the best ap
proach of White towards the Dutch
is to not hamper the opponent to
play it! Black's dark squares are
chronically weak and we should
underline that by trading dark
squared bishops.
5.e2 lLlf6 6.0-0 oid6 7..ia3!
Now 8 . . . e5? ! 8 .cxd5 cxd5
9 . .ib5 + ! ctJc6 1 O .'!!1 c 1 gives White
control over the dark squares, so
Black should find another way of
developing the c8-bishop. He can
try to manoeuvre it to h5:
55
 Part 3
7 . . . 0-0 B.d4 ltJe4 9.c1 ltJd7
1O ..bd6 ltJxd6 1l. ltJbd2 ltJf6 12 .ltJeS
d7
White has clear play on the
queenside:
13 .cS ltJf7 14.ltJxf7 l"!xf7 1S.ltJf3
eB 16.b4. See Game 4 Laznicka
Valeanu, Herceg Novi 2 0 0S.
More often Black fianchettoes
his bishop. Then we attack with b4 :
7 . . . ha3 B.ltJxa3 d6 9.'1Wc1 0-0
1O .b2 b6 1l .d4 b7 1 2 . l"!ac1 ltJbd7
13.ltJeS l"!acB 14.b4 e7
White has a small advantage af
ter 1S.ltJb1 ltJxeS 16.dxeS ltJ e4 17.f3
ltJgS 1B.cS.
Black can try to prevent b3-b4
by . . . as, but then the b6-pawn and
the whole queenside become weak
er. White should exploit that by
opening files on that wing:
S6
13.ltJa4! Now 13 . . . b7? ! stum
bles into 14.ltJxb6;t while 13 . . . cS
14.cxdS exdS 1S.l"!bc1 or 13 . . . e7
14.cxdS give White a lasting initia
tive on the queenside.
One of the most popular sys
tems nowadays are the set-ups with
an early . . . a6. The first critical posi
tion arises after:
3 . . Ajf6 (3 . . . a6 4.tOc3 tOf6)
4.tOc3 a6 5.Yfc2 !
White can also transpose to the
extensively tested main line of the
Chebanenko with S.d4, but I think
that he has good prospects for an
edge with the Reti approach.
I'll not pay much attention to
5 . . . .ig4 6.ltJeS e6, because Black's
set-up is obviously dubious. White
 l.tLlf3 dS 2 . c4 c6 3.e3 !

has a space advantage after 7.d4

bd7 8.i.e2 !

5 e6 is the second most popu

..

lar Black's choice (after S . . . g6), but

it looks like a concession. After all,
it does not fit in with Black's initial
plans. In my opinion, the bayonet
attack would be the most challeng
ing continuation:
6.b3 ! ? tLlbd7 (or 6 ... cS 7.d3 !
c6 8.a3, with fresh, interesting
play) 7.ib2 id6 8Jl:gl ! ? I examine
this plan in the next part.
5,..b5 is more consistent. 6.b3
e6
White is going to gain space
with cS and switch his attention
to the breakthrough e3-e4 : 8 . . . e7
9.0-0 0-0 1O .b2 b7 11 .cS or 8 . . .
id6 9.ib2 b7 1O .cS, transposing
to Part 6/ line Cl. See my game 12
Delchev-Brunello, Porto Carras
10.11.2 011.
5 g6 is the main line. We have
. . .

6 ... g6 7.cxdS cxdS 8 .b2 fS nothing better than :

9.id3 hd3 1O.xd3 e6 11.a4! de
monstrates one of the main ideas of 6.d4 .ig7 7 .td3 0 - 0 8. 0 - 0
.

our opening strategy:

This is the Schlechter System

White's pieces are like a com
pressed spring. Black often seizes
more space, but then he finds him
self overextended and underdevel
oped: 11 . . . b4 1 2 . tLle4 ! g7 13.tLlxf6+
hf6 14.hf6 xf6 1SJcl.
7.d4 tLlbd7 8 .ie2
with c2 a6 inserted.
8 . . . .tg4
The bishop would be too passive
on b7: 8 . . . bS 9.b3 tLlbd7 1O .h3 b7.
We apply the usual recipe against a
b7-bisop - we prevent . . . cS and pre-

57
 Part 3

pare e4: 1 l . cS ! as 1 2 . a3 eS 13.dxeS should delay the queenside advance

lLle8 14.e4 ! t . in favour of development.

9.lOe5 .te6 l O .e5! lObd7 a) 1l . . . 'Wxd7 12 . .id2 ! .ifS 13 .hfS

1l.lOxd7 gxfS 14.lLle2 It>h8 1S.lLlf4 1'!g8 16.f3
.ih6 17.lLld3 1'!g7 1 8 .1'!ae1 1'!ag8
19.1'!e2 'We6
This position has been reached
in Laznicka-Postny, Kolkata 2 0 09.
Perhaps best is 2 0 . 1'!fe1 in order
to use the open e-file after 20 .. .f4
2 1 . lt>h1 lLlhS 2 2 .exf4t.

b) 1 l . . . lLlxd7 12 . .id2 fS 13.b4 .if!

14.f4 !
Both sides have defined their
plans. Black will attack on the
kingside while we'll try to exploit
our space advantage on the op
posite wing. Note, however, that
White should not hurry to push
b4 ! . If Black exchanges the queens
through fS , we will have a more ef
fective plan - a rook lift to b3 ties
Black down with the defence of the
b7-pawn, then the breakthrough White should b e slightly bet
e3-e4 should gain some edge. See
ter after .id2-e1-g3 . See game 5
game 6 Postny-Himanshu,
New Delhi 2 0 1 1 . If Black keeps Koneru-Stefanova, Ulaanbaatar
the queens for an attack, again we 2010.

Points to remember:

White meets . . . .ifS or . . . .ig4 with 'Wb3 .

White meets ... g6 with d4.
If Black plays an early . . . a6 and . . . bS, White answers b3, completes
development and plays c4-cS followed up by e3-e4.
If Black plays an early . . . a6 and . . . g6, White answers d4, later -
c4-cS and does not hurry with b4.

S8
 Pa rt 3

Step by Step

1.f3 d5 2.c4 c6 3.e3 ! cxdS (S . . . ttJxdS 6.d4) 6.d4 e6 7.i.d3

a6. I suggest to keep the tension in
the centre.

I n this part I examine several

systems where Black avoids typical
Meran structures.

A. 3 . . . ttJd7; B. 3 . . . g6 ; C. 3 . . . i.fS ;
D. Set-up with . . . i.g4 ; E. 3 . . . e6 fol
The only sensible way to stick
lowed by .. .fS ; F. Set-up with an
to our repertoire since S.b3 ? ! eS !
early . . . a6.
6.cxdS cxdS 7.i.b2 a6 is unaccepta
The main variation 3 . . . ttJf6
ble for White. The queen move pre
4.c3 e6 is subject of the next parts.
vents this variation in view of S . . . eS
6.cxdS cxdS? 7.ttJbS.

A. 3 d7 5...e5 6.cxd5 xd5 7.d4

Black aims to occupy the centre Sicilian adepts might prefer

by . . . eS. This logical move is seldom 7.i.e2 i.d6 8 . 0-0 0-000 as in Fressi
seen in practice probably because net-Kasimdzhanov, Nancy 2 0 1 1 .
his position after cxdS is too pas
sive. At the same time, it is very 7....id6 8 ..id2!?
solid and I do not think that White
has much after 4 . ttJ c3 ttJgf6 S.cxdS Preparing 0-0-0.

S9
 Part 3
8 0 - 0 9.ttJe4 !J.c7 10. ttJeg5
.
g6 1l.e4 ttJ5f6
12.!J.c4 !
12.ttJxeS ttJxeS 13.dxeS ixeS
14.0-0-0 looks promising. Indeed,
14 . . . Yi'e7 1S.<it>bl (or IS.h3 c5 16.ttJf3
!J.d4 17.i.d3t) IS . . . cS (IS . . . !J.d7
16.i.e3 b6 17.h3 cS IB.ttJf3) 16.h4
llJg4 17.!J.c4 bS IB.!J.dS i.b7 19.ixb7
Yi'xb7 2 0 .hS ttJxf2 2 1.Yi'xcS;!; favours
White, but lS . . . Yi'c7 ! , taking f4 un
der control, is unclear. It is safer to
lead an attack with short castling.
12 . . . exd4 13. 0 - 0 ti'e7 (13 . . . cS
14J':iael) 14.gae1t
The pawn structure resembles
the variation with 3.e4 eS in the
QGA.
B. 3 . . . g6 4.ttJc3 !J.g7 5.d4 ttJf6
6.!J.e2 0 - 0 7. 0 - 0
Play has transposed to the
Schlechter System. I'll only mention
the two most popular continuations :
60
a) 7 ... a6 B.Yi'b3 e6 (B ... dxc4
9.ixc4 bS lO.i.e2 lDbd7 11.e4;!;)
9J3dl
The character of play is not
forced. White can also develop
with 9 .i.d2 llJbd7 (9 . . . lDe4 lO .i.e!
lDd6 11.cS lDfS 12 J:!dl lDd7 13.e4;!;)
lOJ3fdl.
9 ... lDbd7 10.a4 as 11.Yi'c2;!;.
b) 7 . . .dxc4 B.ixc4 !J.g4 9.h3
!J.xf3 10 .ti'xf3 lDbd7 lU3dl eS 12 .dS
e4 13.lDxe4 lDxe4 14.Yi'xe4 lDb6
lSJ3bl geB 16.Yi'd3 cxdS 17.!J.bS ge7
IB.Yi'b3 gcB 19.i.d2;!;, Turov-Ipatov,
Nakhchivan 2011.
C. 3 ... .if5
This thematic development does
not work here because of the weak
dS-pawn and the square d4 which
can be occupied with tempo by lDf3.
4.ti'b3 eb6
4 . . . Yi'c7 S.cxdS cxdS 6.ttJc3 e6 of
fers White two good options :
a) 7.llJbS Yi'b6 B.Yi'c3 ! ttJc6 9.lDd6+
 l.ttJf3 d5 2 .c4 c6 3.e3 !

ixd6 1O :xg7 ttJb4 1 l .ttJd4 r;!;e7 Kramnik-Eljanov, Moscow 2010,

12.'hh8 Wic7 13.ttJxf5+ exf5 14.Wic3 1l . . . Wib6, with an unclear game.)
IiJc2+ 15.r;!;d1 ttJxa1 16.Wixc7+ fixc7 4 . . . Wic7 5.ttJe5 fie6 6.d4 ttJd7 7.ttJc3
17.b3 ttJf6 18.f3 , with the bishop dxc4 8.fixc4 fixc4 9.ttJxc4 ttJgf6.
pair advantage. (White is also somewhat better af
ter 9 . . . b5 1O.ttJd2 ttJf6 11.ttJce4.)
b) 7.Wia4+ ttJc6 (7 . . . ttJd7 8.ttJb5
%Yb6 9.Wid4) 8.ttJd4 fig6 9 .b3 a6
1O.tLlxc6 bxc6 1 l.fia3 fixa3 12.Wixa3
liJe7 13.fie2 0-0 14.0-0.

5.cxd5 Wixb3 6.axb3 cxdS

7.c3

The game Flores-Peralta, Men

doza 2008, went 10.e4 b5 11.ttJe3
when 11.. .e5 would have blocked the
centre. I propose 1 O .a4 e6 1l .e4.

4.c3 J.g4 (4 . . . fif5 5.cxd5 !

cxd5 6.Wib3)

At first glance Black has solved

the problem of the Slav bishop, but
It turns out that Black is losing a couple of simple moves dissipate
a pawn :
the illusion :

7... e6 8. bS d8 9. xa7;t. S.Wib3 Wib6 6. e5

D. Set-up with . . . .ig4

3 . . . f6

It makes sense to take the sting

out of Wib3 . The immediate 3 . . . fig4
4.Wib3 is awkward for Black (4.cxd5
cxd5 5. Wib3 allows the defence 5 . . .
%Yc7 6.fib5+ ttJd7 7.ttJc3 fixf3 8.gxf3
e6 9.d4 ttJgf6 10 .e4 a6 1 l .fie2 ,

61
 Part 3

6 .. .ie6 14 .ib8 15. d2 e5 16. tLJb6

l'l:a7 17.3 h5 18.dxe5 he5

Alternatively: 19 .ic3:t.

a) 6 . . . i.f5 7:xb6 axb6 8.cxd5

LDxd5 9.LDxd5 cxd5 1O .i.b5+ LDd7 E. 3 . e6 4.b3 f5
l1.f4 g6 12 .d3 f6 13.LDf3 @f7 14.e4
dxe4 15.dxe4:t, Chatalbashev-Drab
ke, Antalya 2 0 04 .
b) 6 . . . e6 7.LDxg4 LDxg4. White
has a bishop pair and a spatial ad
vantage. Now, or on the next move,
he can play ;Vc2 , with a flexible
pawn formation. RaLGarcia-Gin
zburg, Buenos Aires 1995, went
8.i.e2 LDf6 9 .d4 LDbd7 10.0-0 i.e7
l1 .;Vc2 0-0 12 .b3 l'l:fc8 13 .i.b2:t.

7.d4
Now the other bishop got stuck
on f8 and should look for sideways.
7
. tLJbd7
7 . . . g6 8 .i.d3 i.g7 9 . 0-0 0-0 is a
rare line of the Schlechter system.
After 1O .;Va3 dxc4 1l.LDxc4 ;Vc7
12 .i.d2 l'l:e8, Basagic-Ivacic, Slove
nia 1996, White has more space and
good control of the centre.
8. tLJxd7 J.xd7 (8 . . . LDxd7 9.cxd5)
9.tLJa4 W1xb3 1 0 . axb3 a6 1l . .id3
e6 12 . .id2 .id6 13 . .ia5 e7 14.c5
In this version of the Stone
wall, White should trade dark
squared bishops and expand on the
queenside. Of course, he can play
d4 at any moment and transpose
to standard lines. Then he should
follow the set-up of the game Ilin
cic-Kiroski, Belgrade 1995, which
went 1.d4 d5 2 .c4 e6 3.LDc3 c6 4.e3
f5 5.LDf3 LDf6 6.i.d3 i.d6 7.0-0 0-0
8.b3 LDe4 9.LDe2 LDd7 1O.a4 b6
l1 .i.a3 i.xa3 1 2 . l'l:xa3 as 13.l'l:a2 i.b7
14.l'l:c2 l'l:c8 15.;Vb1 ;Ve7 16.l'l:fc1.

White has some space advan

tage and his outpost on b6 is cramp
ing the opponent's queenside.
The game Ponomariov-Ivanchuk,
Dagomys 2 0 1 0 , went :

62
 l . lLlf3 d5 2 .c4 c6 3 . e3 !

White has a slight pull here, but 1O .YlYc1 YlYc7 1l.hd6 YlYxd6
he lacks the thematic plan with h4. 12.YlYa3 xa3 13.xa3
I propose to rip benefits from hav
ing delayed d4 and c3 :

S . .ie2 f6 6. 0 - 0 i.d6 7.i.a3 !

Black loses control over the dark

squares, because 13 . . . lLlg4 14J''lacl
i.d7 15.h3 e4 16.hxg4 would cost
him a pawn.
El. 8 . . . e5? ! ; E 2 . 7 . . . i.xa3 ; E3. 7 . . .
b6; E4. 8 . . . lLle4. 13 e4 14.d4 i.d7 lS.i.xc6
..

bxc6 16.lUcl k8 17.1'kS;!;.

El. 8 eS? !
E2. 7 i.xa3 8. xa3 YlYd6

This is consistent, but Black is 9.cl 0 - 0 10 .YlYb2 b6 1l.d4

underdeveloped and his pieces are .ib7 12.acl bd7 13.eS ac8
hanging. 14.b4 YlYe7

8.cxdS cxdS 9 .ibS+ !

Taimanov-Persson, Stockholm
1994, saw 9.d4 i.xa3 1 O .lLlxa3 e4
11.ttJe5 0-0 12 Jcl a6 13 .YlYc2 lLlfd7
14.f4;!;, but the text is more straight
forward.

9 . . . . c6

9 ... i.d7 drops a pawn to 1O.i.xd7+

xd7 1l.i.xd6 YlYxd6 12 .YlYc2 while
9...'it>t7 10.d4;!; is also gloom. White has a small advantage,

63
 Part 3

but the centre is still unstable. In Petersburg 2 0 1 0 , went 13 . . . .ib7? !

the game Malaniuk-Tregubov, 14.tLlxb6;t. 13 . . . c5 is more precise,
Krasnodar 1997, White spent a but after 14.cxd5 exd5 15.Ei:bc1
tempo on 15.tLld3 and allowed ex White enjoys a lasting initiative on
changes after 15 . . . Ei:b8 16.tLlb1 Ei:fc8 the queenside.
17.tLld2 tLle4 18.tLlf3 c5. He should
have started redeployment of his
knight right away with : E4. 8 c!De4 9.cl c!Dd7
.

10.hd6 c!Dxd6 1l. c!Dbd2 c!Df6

15.c!Dbl c!Dxe5 16.dxe5 c!De4 12. c!De5 .id7
17.f3 c!Dg5 18.c5;t.

E3. 7 . . b6 8 . .bd6 xd6 9.d4

White transposes to a typical

Stonewall. Without dark-squared
bishops, this is a sensible decision ,
which should assure h i m o f some
edge. The RHi-approach 9 . tLl c3 0-0
1O.cxd5 exd5 1 1 .c2 is also possi
ble, because 1 1 . . . f4 (11 . . . .ie6 1 2 .d4)
would stumble into 1 2 . e4 ! .
This is the most passive, but
9 c!Dbd7 10. c!Dc3 0 - 0 11.c2
.
solid development of the light
c!De4 12.Ei:abl a5 squared bishop. Black avoids any
weaknesses and hopes to withstand
a long siege after:
13.c5 c!Dt7 14.c!Dxt7 gxt7
15.c!Df3 .ie8 16.b4

Game 4 Laznicka-Valeanu,
Herceg Novi 2 0 05 , is a model ex
ample how to exploit White's domi
nance by opening a second front on
the kingside.

F. Set-up with an early a6

Black has prevented b4, but
now 13.tLla4! pinpoints the weak 3 c!Df6 (3 a6 4.c!Dc3 c!Df6)
. .

ness on b6. Grigoryan-Yudin, St. 4. c!Dc3 a6 5 .c2

64
 1.tt'lf3 dS 2 .c4 c6 3.e3 !
S.d4 transposes to the exten
sively tested main line of the Che
banenko System. Then S . . . bS 6.b3 is
still in course of development, but I
find it difficult to recommend a line
against S . . . e6, followed by . . . c6-cS.
I play it successfully as Black.
Apart from S:c2 , White has
also tested S.b3. He was slightly
better in Morozevich-I. Sokolov,
Sarajevo 2 0 07, after S . . . .ig4 6 . .ie2
e6 7. .ib2 tt'lbd7 B.h3 hf3 9.hf3
ib4 1O :'c2 0-0 11.0-0 Wfe7 12 J''!: a d1
gacB 13.g3 tt'leS 14 . .ig2 E:fdB 1S.tt'le2
llJg6 16.f4 tt'leB 17.fS exfS 1B.tt'ld4:t,
but S . . . .ifS 6 . .ie2 h6 7 . .ia3 tt'lbd7
8.0-0 e6 9.MB tt'lxfB 1O.d4 tt'lBd7
equalised in Radjabov-Vallejo Pons,
Porto Carras 2 011.
The queen move aims to impede
the development of Black's bishop.
Black has tried here : Flo S . . . .ig4 ;
F2. S . . . e6; F3 . S . . . bS; F4. S . . . g6.
Fl. 5 . . . .ig4 6.e5 .ie6
6 ... e6 7.tt'lxg4 tt'lxg4 is solid, but
rather dull. White is assured of a
long-term small edge with .ie2 , d4,
b3, .ib2 .
7.d4
The pure Reti approach 7.b3 ! ?
(delaying d2-d4) i s also a n option :
7 . . . g6 B . .ib2 .ig7 9 . .ie2 0-0 10.0-0
.ifS (1O . . . tt'lbd7 1l.tt'la4:t) 11.d3 Wfd6
1Voa4 tt'lbd7 13.cS Wfc7 14.f4,
Janssen-Svetushkin, Kavala 200B.
7 . . bd7
8 . .ie2 !
Ivanchuk and Dreev have played
B.tt'ld3 to win their games after S . . .
tt'lb6? 9.cS tt'lbd7 1O.tt'lf4 .ig4 1l.h3
.ihS 12.tt'lxhS, Ivan chuk-Postny,
Kallithea 2009 or B . . . gS? 9.cxdS
tt'lxdS 1O .h4 g4 11.e4, Dreev-Pana
rin, playchess.com 2 004. However,
Black should have sacrificed a pawn
with an excellent compensation af
ter B . . . .ifS ! 9.cxdS cxdS 1O .Wfb3 e6
1l.Wfxb7 .id6 12 . .ie2 0-0 13.0-0 Wfe7
14.Wfb3 E:fcB.
Natural development of the
bishop to e2 is better. Now it is not
clear how Black should continue.
6S
 Part 3

8 . . . dxc4 9.lLlxc4;l; surrenders the

centre, but 8 . . . lLlxeS 9.dxeS lLld7
1O.cxdS cxdS 11.f4 is not a good
option either. As a whole, White
has the upper hand owing to the
awkward position of the e6-bishop.

F2. 5 . . . e6

I prefer to continue here in the This treatment of IQP positions

Reti style with b3, .id3, leaving the
pawn at d2. This development is
considered in line F2 2 . However,
switching to well tested lines with
6.d4 is also popular, and it is cer
tainly less committal.
F21. 6.d4; F22. 6.b3 ! ?
F21. 6.d4 c5
Alternatively:
a) 6 . . . bS 7.b3 lLlbd7 8 . .id3 .ib7
9. 0-0 .ie7 1O.cS is rather gloom
for Black. He cannot push 10 . . .
e S i n vew o f 11.dxeS lLlg4 1 2 .lLlxdS,
Khenkin-Kaeser, Osterburg 2006.
Therefore, he cannot prevent White
from playing e4, with a big space
advantage.
(with a4), has been popularized by
Topalov and now we see another
top player choosing it. Black's task
is not easy. Nakamura had previ
ously played this position as White
against Ni Hua, London 2 0 0 9 : 10 ...
.ie7 11.0-0 0-0 12.Ei:ac1 lLlb6 13 . .id3
h6 14.b3 lLlbdS lS.lLlxdS lLlxdS.
Now 16 . .ic2;l;, building a .i- bat
tery against the enemy king, would
have provoked . . . fS and the hole
on eS would give White an edge. It
is interesting to see Nakamura in
Black's shoes :
1O . . . lLlb6 1 1 . .id3 lLlbdS 1 2 . lLlxdS
lLlxdS 13.0-0 lLlb4 14 . .ixb4 .ixb4
lS . .ixh7! . So he had no improve
ment at all ! White was on top after
lS . . . g6 16.hg6 fxg6 17.xg6+ .
7.cxd5 exd5 8 . .ie2

b) 6 . . . lLlbd7 7 . .id2
Carlsen has crushed Nakamura
with this move in Monaco 2 0 1 1 .
Practice has seen Black gradually
equalising after 7.b3 .id6 8 . .ib2 0-0
9 . .ie2 eS 1O.cxdS cxdS 11.dxeS lLlxeS
1 2 . 0-0 .ig4 13.lLlxeS heS 14 ..ixg4
lLlxg4 lS.h3 lLlf6.
7 ... dxc4 8.a4 ! cS 9 . .ixc4 cxd4
1 O . exd4

66
 l.tLlf3 d5 2 .c4 c6 3.e3 !
8 c!Dc6
8 ... cxd4 ! is seldom seen, but
I think that it gives Black more
chances to equalise. Then 9.exd4
lLlc6 10.0-0 (or 10.tLle5 d6 ! ) 10 . . .
id6 ( 1 0 . . . e7 is passive. White
is slightly better after Il.tLle5 0-0
12.lLlxc6 bxc6 13.tLla4 c7 14.d3)
11.g5 e6 12 .d2 h6 is roughly
equal. Thus White should choose:
9.tLlxd4 d6 1O. tLlf5 0-0 1l.tLlxd6
xd6 1 2 . 0-0
12 ... tLlc6!
Naturally, the side with an iso
lated pawn should avoid exchang
es. Gustafsson-Handke, Germany
2 010, saw 12 . . . g4 13.E1dl he2
[13 ... tLlc6 14.h3 (14.tLlxd5 he2
15.tLlxf6+ xf6 16.xe2 E1ad8)
14 ... he2 15.tLlxe2 E1ac8 16.b3]
14.tLlxe2 tLlc6 15.d2 tLlg4 16.tLlg3t.
13.E1dl e5 = .
9.c!De5! c7 1 0 . c!Dxc6 xc6
1O . . . bxc6 stumbles into 1l.e4 ! !
cxd4 (1l . . .dxe4 12.tLlxe4 cxd4
13.tLlxf6+ gxf6 14.0-0) 1 2 .tLlxd5
lLlxd5 13.exd5 b7 (or 13 . . . b4+
14.d2 hd2 + 15.xd2 c5 16.0-0
0-0 17.b4 cxb4 18.xb4 c3 19 .c4
d7 20.E1ac1 b2 21.E1bl c3
2 2 . E1fc1 xc4 23.hc4t) 14.0-0 d6
15.f3 cxd5 16.E1el+ @f8 17.xc7
xc7 18.b3 d6 19 .b2t, Zhou Wei
qi-Ni Hua, Danzhou 2010.
11. 0 - 0
1l.dxc5 i s premature due to 11...
xc5 = . Note that 1l . . . xc5 12 .d2
d6 does not transpose to the main
line since White has more useful
moves than castling, namely 13.E1cl
0-0 14.b3t, Bu Xiangzhi-Ni Hua,
Guangzhou 2010.
1 1 ... d6
After 1l . . . cxd4 12.exd4 the sym
metry works in White's favour
because his pieces are better sta
tioned : 12 . . . d6 (l2 . . . b4 13 .f4
hc3 14.xc3 xc3 15.bxc3 d7
16.a4t leaves White with a back
ward pawn on c3, but the bishop
pair and the open b-file gave him
an edge in Mamedyarov-Wang Vue
Elista 2008.) 13.f3 h6 14.d2 0-0
15.E1acl e6 16.b3 d7 17.tLla4
tLle4 (17 . . . E1ae8) 18 .he4 dxe4,
67
 Part 3
Xiu-Deepan Chakkravarthy, Subic
Bay 2009. Now 19.1Llb6;!; wins the
exchange for a pawn.
12.dxc5 'lWxc5 13 .id2 0-0
14.acl
This position has arisen in Ma
medyarov-Grischuk, Nalchik 2 0 0 9 .
Apparently, Black did not want to
defend a somewhat passive set-up
with an isolated pawn after 14 . . . .ie6
1S.d3 so he decided to get rid of
it by 14 . . . d4 1S.lLle4 lLlxe4 16.'lWxe4
'lWeS 17.'lWxeS heS 18 .b3;!;. The
threat of .if3 hampers Black's de
velopment.
F2 2 . 6.b3 ! ? c5
6 . . . lLlbd7 7.d4 .id6 8 . .ie2 0-0
9 . 0-0 transposes to a well known
Anti-Meran pos ition, but 7 . .ib2
.id6 8 . g1 ! ? is more entertaining. I
examine this plan in the next part.
7 . .id3 ! (7.cxdS exdS= ) 7 . . . c6
8.a3
8 . g6
68
Black bolsters his castling posi-
tion. 8 . . . d4 9.lLle4 favours White, but:
8 . . . .ie7 deserves serious atten-
tion. After 9 . .ib2 0-0 I have ana
lysed 1O.lLle2 (lO.cxdS exdS 1l.lLla4
b6 12.hf6? .ixf6 13.hh7+ loses
material) 1O . . . b6 and now White
can organise a kingside attack using
the g- or the h-file :
a) 1l.cxdS exdS 1 2 . lLlg3 g6 13 .h4
.id6 14.hSt.
b) 1l.g4 eS (first line of the en
gines) 1 2 .lLlxeS lLlxeS 13 .heS lLlxg4
(13 . . . dxc4 14.bxc4 hg4 1SJg1 g6
16.b1 .if3 17.g3 .ib7 18 . .ixg6 ! !
i s crushing, for instance : 1 8 . . . fxg6
19.!%xg6+ It>h8 2 0 .!%h6 !%f7 2 1 .lLlf4
It>g8 2 2 . lt>e2) 14 . .ib2 .ih4 1S.lLlg3
dxc4 16.bxc4 g6 17.h3 lLlf6 18.!%gl
!%e8
White has a strong initiative.
His next step should be to open
the centre in order to activate the
second rook: 19.'lWc3 It>f8 2 0 . 0-0-0
.ib7 2 1..ic2 bS 2 2 .d4t.
9 . .ib2 .ig7 1 0 . e2 0-0
1l.cxd5 xd5
It would be interesting to test
1l . . . exdS 12.xcS ig4oo.
 l.lLlf3 dS 2 .c4 c6 3.e3!
12.c3 d8 13 .e4 d7
14.hg7 'it>xg7
This position has been reached
in the game Iturrizaga-Malakhatko,
Cannes 2 0 1 0 . White has the better
pieces. He could have fixed his ad
vantage with 1S.e2 hitting cS and
preparing d2-d4.
F3. S bS 6.b3
Black is trying to follow the Che
banenko approach, but in this set
up %Yc2 is more useful than d4. For
instance, the thematic:
6 ... g4?! faces 7.lLleS and the
bishop can not retreat to fS while
the unsuspecting 7 . . . hS? turned
out to be a blunder in Ivanchuk
Bu Xiangzhi, Sofia 2008: 8.cxbS
cxbS? ! 9.xbS+ ! axbS lO. lLlxbS and
the game was over.
In practice Black often tries to
exploit the early development of
White's queen by:
6 . . . g6 Here again, White is bet
ter off without d4: 7.cxdS cxdS
8 .b2 fS 9.d3 xd3 1O.xd3 e6
(the long diagonal is open so 10 . . .
g7? loses t o 1l.lLlxdS ! )
Black's queenside i s weakened
and White should immediately take
advantage of that by:
11.a4!
Instead, Bu Xiangzhi-Matlak
ov, Moscow 2010 saw 1Uk1 lLlbd7
1 2 . lLle2 d6 13Jk6 e7 14.xf6
xf6, when neither 1S.c2 a3
16.0-0 0-0 17.l"1b1 e7 18.d4 d6
19.1Llc1 lLlb8 = , nor 1S.0-0 e7 16.a4
bxa4 give a substantial edge.
Following 11.a4, 11.. .bxa4
1 2 .lZlxa4 bd7 13.0-0 d6 14.l"1fcl
is clearly in White's favour as well
as 1l . . . b4 12 .e4 g7 13.lLlxf6+
hf6 14.hf6 xf6 1SJ'kl.
69

6 . . . e6 7.d4
This move is not mandatory yet,
but sooner or later White will have
to play it. He can start with 7.b2,
though .
7 .. .!lJbd7 8.e2
White's general plan in this
position is to gain space on the
queenside with c4-c5 and prepare
e4. 8 .d3 b7 9.c5 also serves well
this goal and gives White an edge,
for example, 9 . . . g6 10.0-0 g7
11.e4;l; (It is possible to delay e3-e4
for a few moves: 11.a3 a5 12 .b2
0-0 13.{Je2 '?fic7 14.h3 . ) 11 . . . b4 12 .e5
bxc3 13.exf6 {Jxf6 14.'?fixc3 , with the
more active pieces.
8 . . .te7
8 . . . d6 9.b2 b7 1O .c5 trans
poses to Part 6/ line Cl. See my
game 12 Delchev-Brunello,
Porto Carras 1 0 . 1 1 . 2 0 1 1 .
9. 0 - 0 0 - 0 10 .b2 b7 11.c5
70
Part 3
White is not interested in keep
ing the tension in the centre any
more. For example, after 11.E:adl
'?fic7 White has not anything better
than 12 .c5, but then why has he put
the rook on d1?!
1l . . . b4
Or White will play b4 himself.
12 .!lJa4 a5 13.a3 .ta6 14.ha6
gxa6 15.axb4 axb4 16 .!lJel YlYb8
17 .!lJd3 ga5 18.f3 gb5 19.gaeU
White has the better prospects,
Ilincic-Popchev, Jagodina 1993.
F4. 5 . g6 6.d4
Perhaps I should explain why we
have to play d4 right now. Indeed,
White can pursue the Reti scenario
with 6.b3 g7 7.b2 0-0 8 .e2 , but
after 8 . . . E:e8 9.0-0 e5 all the fun is
for Black. We should aim for a more
active set-up, connected with i.d3.
6 .. .ig7
 1.ct:lf3 dS 2 .c4 c6 3.e3 !
Black plays this move almost
automatically, but the big ques
tion is what happens in the event
of 6 . . . ifS . The only chance to get
something out of the opening is to
exploit the temporary weakness of
the enemy queenside. We answer
with 7.b3 bS (7 . . . E1a7 is too pas
sive. White is somewhat better fol
lowing 8 .ie2 ig7 9.0-0 0-0 1O.a4 ! ,
for example, 1O . . . ct:le4 11. ct:lxe4 ixe4
12.ct:ld2 ifS 13.aS;l;) 8.cxdS cxdS
I would like very much to play
here 9.a4, but then 9 . . . b4 ! ! 1O .xb4
Itlc6 would earn the initiative.
The game Matveeva-Gunina,
Moscow 2009, went 9.ct:leS ct:lbd7?
1O.a4 b4 11. ct:lxdS ct:lxeS 12.ct:lxf6 +
exf6 13.dxeS fxeS 14.ic4 d7 (14 . . .
Wff6 ! i s stronger, but White retains
some pull after lS .id2 id6 16 .idS
e6 17.ixe6 xe6 18 .xe6+ fxe6
19.e4 E1c8 20. me2 me7 21.E1acl;!;)
and here lS.e4 ! would have given
White a terrific initiative : lS . . . ixe4
[or lS . . . ie6 16 .ixe6 xe6 17.xe6 +
fxe6 18.ie3 E1c8 19.md2 icS (19 . . .
e7 2 0 . E1hc1 m d 7 21.E1xc8 E1xc8
22 .md3) 2 0 .ixcS E1xcS 21.E1hc1 E1xc1
22 .E1xc1 md7 23.E1cS;l;] 16.0-0 ic6
17.igS ig7 18 .xb4.
However, 9.ct:leS is probably not
the best move. The engines pro
pose a paradoxical retort: 9 . . . e6 ! ?
1O .f3 hS! and I could not find a way
to break through Black's trenches
along the fifth rank.
In the position on the last dia
gram, White should revive the
threat of a4 by:
9.id2 ! ct:lc6 1O.ct:leS ct:laS 11.Wdl
ct:lc4 12 .a4 !
The pawn structure after
12. ct:lxc4 dxc4 (12 . . . bxc4 13.b3;l;)
looks to be in White's favour be
cause he can gradually put his cen
tral cluster in motion, for example,
13 .ie2 E1c8 14.if3 ie6 lS .c2 ig7
16.0-0 0-0 17.ct:le2 idS 18 .e4 ib7.
However, Black's pieces are well
stationed.
12 . . . ct:lxd2 13.xd2 b4 14.ct:la2 e6
lS.f3 b3 16.ct:lb4 b6
* .t
i i
i \l!V i l.il i
i tlJ .t
Cj !J
i
'it"
a w i. :
The a6-pawn may be protected,
but only at the expense of Black's
king. White deals a terrific blow :
17.ci:Jxa6 ! E1xa6 18.E1c1 ! ! and
there is suddenly no defence against
ibS + , e.g. 18 . . . E1xa4 19.E1c8+ me7
2 0 . E1b8 c7 21 .ibS+-.
71
 Part 3
7.J.d3 0 - 0 8. 0 - 0
This is the Schlechter System
with 1!9c2 a6 inserted. White's game
is slightly better because he has
more space on the queenside.
8 J.g4
.
The bishop would be too dull on
b7, for instance:
a) 8 . . . bS 9.b3 ttJbd7 1O.h3
1O . .ib2 .ib7 1UUd1 (ll.cS eS)
11..Jc8 12Jac1 1!9c7 13.a3, with a
small edge, is also possible, Akes
son-Hammer, Gausdal 2 00S.
1O ... .ib7 1l.cS as 12.a3 eS
13.dxeS ttJe8, Chernin-Murey, Co
penhagen 1986. In these structures
White can only gain an advantage
with the breakthrough e3-e4 and it
is quite strong here.
b) 8 . . . ttJbd7 9.b3
Undoubtedly, 9.cxdS cxdS
10 . .id2 b6 (10 . . . bS 11.fcU) 1l.fc1
retains some pull: 1l . . . .ib7 12 .ttJe2
c8, Maiorov-Flachet, Sautron
2008, 13.1!9a4 !;l;.
9 ... e8 1O . .ib2 dxc4 1l.bxc4
72
eS 12.a4 1!ge7 13 . .ia3 1!ge6 14.d5 ,
Lautier-Relange, Besancon 1999.
c) 8 ... .ie6 9.c5 ! .ig4 (9 ... ttJbd7
1O.ttJg5) 1O .ttJe5 ttJbd7 1l.h3 ! .ie6
12 .ttJf3 b6 13.cxb6 c5 14.b3;l;, Akes
son-Hellsten, Sweden 1998.
9.e5 .ie6 10 .c5!
The right plan ! Alternatives are:
a) 10 . .id2 gives Black the oppor
tunity to open the centre with :
1O . . . dxc4 !
1O . . . ttJbd7 allows White to re-
turn to the plan with c4-c5 after
1l.ttJxd7 ixd7 (1l . . . 1!9xd7 12.cxd5
cxdS 13.ttJa4 ac8 14.ttJcS 1!9c7
15.1!9b3 .ifS 16.ixf5 gxf5 17.acl)
12 .cS;l;.
1l.ttJxc4 b5 !
The breakthrough . . . e7-e5 as
sumes that Black will part with the
bishop pair: 1l . . . ttJbd7 12 .h4 ! ixc4
13.ixc4 e5 14.dxeS ttJxe5 15 . .ie2
1!9d6 16.ab1 ad8 17.fd1 ttJeg4
18.g3;l;.
12 .ttJeS cS 13.dxcS 1!9c7 14.ttJf3
1!9xcS 1S.ttJd4 .id7 16.h4
In A.Maric-Zhu Chen, Moscow
2001, White became even worse af
ter 16.a3? ! ttJc6 17.ttJb3 1!9d6 18.ad l
ttJeS+.
16 . . . 1!9b6 17.a4 ttJc6 18.ttJxc6
ixc6 19.axbS axbS 20.xa8 xa8
21.c1, M.Gurevich-Van Haastert ,
Belgium 2003. Here 2 1 . . . ttJg4 ! ?
22 .h3 ttJeS equalises.
 1.1Of3 d5 2 . c4 c6 3 .e3 !
b ) 1O .b3 has been played in a
fair amount of games:
10 . . . c5 ! 1l.lOe2 lOfd7!
I have tried once 1l ... cxd4
1 2 . exd4 lOc6 13.lOxc6 bxc6;!;, but
White is better here. Another
option is 1l . . . dxc4 12 .bxc4 lOfd7
13 . CiJxd7 ixd7 14Jbl lOc6 15.dxc5;!;.
12.lOxd7 1Oxd7 13 .i.b2 cxd4
14.exd4
14.lOxd4 dxc4 15.bxc4 tLlc5
16.CiJxe6 tLlxe6 17.ixg7 xg7 18 .i.e4
b8 does not promise White much.
I would rather take Black here.
14 . . . dxc4 15.bxc4 E1c8 16.tLlf4
(16.d5 tLlc5 = ) 16 . . . ixc4 17.ixc4
ttJb6 18 .ixf7+ E1xf7 and only Black
could aspire to the advantage here.
10 . lLlbd7 1l. lLlxd7
D41. 1l . . .'xd7; D42. 1l . . . lOxd7.
D41. 1l . . :f;!fxd7
Black aims to swap his light
squared bishop through f5. The
drawback of this plan is that Black's
pawn formation loses its flexibility.
That might be critical in an endgame
when Black pushes ... e5 and the f5-
pawn remains without support.
12 .td2 !
Occasionally, White plays 12 .b4.
This advance has no alternatives in
the event of 12 . . ..if5 13.ixf5 gxf5,
but if Black swaps queens by 13 . . .
\Wxf5, White will remain without
an effective plan. I show in the
commentaries of game 6 Postny
Himanshu, New Delhi 2011, that
without queens White has more
winning chances with the pawn on
b2. In the endgame he should hit
b7 by the rook lift E1al-c1-c3-b3 and
prepare e3-e4.
12 . . . .tfS 13.hf5 gxf5 (13 . . .
\Wxf5? ! ) 14.lLle2 h8 15.lLlf4 gg8
16.f3 .th6 17.lLld3 gg7 18.gae1
gag8 19.ge2 1!Ye6
73
 Part 3

This position has been reached It is difficult to break through

in Laznicka-Postny, Kolkata 2009.
White is slightly better, because
he has the plan with b4-bS in re
serve while the opponent can only
rely on tactical tricks. Perhaps
best is 20Jfe1 in order to use the
open e-file after 20 .. .f4 21.h1 lLlhS
2 2 . exf4;!;.
F42 . 11 xd7 12 .Ad2
after 14.lLle2 eS 1S.dxeS lLlxeS, al
though in Macieja-Wojtaszek, War
saw 2009, White got an edge after
16.ic3 h4?! (16 . . . e7) 17.id4
lLld7 18.c3 hd4 19.xd4 xd4
20.exd4 lLlf6 21.a4 lLle4 2 2 .ga2 ie8
23.f3 lLlgS 24.lLlf4 id7 2S.gfa1 e6
26.lLlxe6 he6 27.bS;!;.
With the text, White fixed her
space advantage and will enjoy a
lasting pressure.

This move order is not obliga

tory. White can also begin with
12 .b4 when 12 . . .fS should transpose
to the main line. Black can also try
to prepare . . . eS by 12 .. .f6, but this
move is theoretically unimportant
as White can force .. .fS with 13.f4
(13 . . . if7? ! 14.fS). Note that 13.e4
dxe4 14.lLlxe4 if7 as in Tomashevs
ky-Yevseev, Sochi 2006, is also
good for White following 1S.ib2 h6
16.lLld2 fS 17J'Uel.

12 . . . f5 (12 . . .f6 13.f4 if7 14.fS See game 5 Koneru-Stefano

e8 1S.e4) 13.b4 f7 14.f4! va, Ulaanbaatar 2010.

74
Part 3

Complete Games

4. Laznicka-Valea n u Laznicka makes a good strategic

Herceg Novi 1 S.09 .2005 decision :

1 .c4 eS 2 . f3 d5 3.e3 cS 4.b3 1 4. xf7 gxf7 1 5. f3 .ie8 1 S. b4

f5 5.ie2 fS S.O-O idS 7 .ia3 ge7 1 7 .e5 d 7 1 8 . xd7 .ixd7
0-0 8.d4 e4 9 :ilYc1 d7 1 0 .ixdS 19 .VNb2 VNc7 20 .f4 !
xdS 1 1 . bd2 fS 1 2 .e5 .id7
1 3 .c5 f7

White's play i s s o clear and con

sistent that it does not need any
comments.
The position looks too closed
and Black obviously relies on the 20 . . . bS 2 1 .gfc 1 gee8 22 .VNc3
well known fact that one weakness geb8 23.a4 as 24.gab 1 b5 25.ga1
should not lead to a loss. This logic .ic8 2S.ga3 .ib7 27.gca 1 VNc8
is still correct indeed. White can 28.g1 a2 c,t>f8 29 .VNa 1 c,t>g 8 30.h3
not win only by attacking c6 or an gS 3 1 .g4
other queens ide pawn. His decisive
breakthrough should happen on
the opposite wing. That explains
his following move. Generally, the
si de with a space advantage should
avoid exchanges, but in the con
crete example, the black knights
would be dangerous if White tried
to open the kingside. So the young

75
 Part 3
The second stage of the game
begins. The question now is could
White infiltrate his queen or king
into the enemy camp.
31 . . . 'it> h S 32.'it>g2 'it>g S 33.'it>g3
'it>f7 34.h4 'it>f6 3S . .tf3 'it>f7 36.hS
Wg S 37.hxg6+ hxg 6 3S.axbS axbS
39J3a7 gxa7 40.gxa7
40 . . .gS
This move only helps White,
but Black cannot stay passive any
more. After 40 . . . mg7 41.hl hB
42.xhB + mxhB 43.mh4 mg7
44.mg5 he is in fact in a zugzwang:
44 . . . mf7 45.gxf5 gxf5 (45 . . . exf5
46 . .bd5 + ! ) 46 . .th5 + me7 47 . .tg6
md7 4B .mf6.
Following 40 . . . g5, White wins
with 41.hl gxf4+ 42.mxf4 fxg4
43 . .bg4 g7 44.mf3 ! (clearing the
h2-bB diagonal) 44 . . . me7 45.h4+
(or even 45. h2 UB+ 46. me2 xg4+
47.md2+-) 45 . . . Wf6+ 46.xf6+
mxf6 47.mf4 me7 4B.me5 md7
49J!a2 ghB 50.gf2 gh7 51.gf7+ .
4 1 .Wa6?! WeS 42.Wb6? !
gxf4+? 43.exf4 fxg4 44 . .txg4 'it>f6
4S.WaS gaS 46.gxaS .txaS 47 .Wa 1
.tb7 4S .We 1 Wd7 49 .WeS+ 'it>f7
76
SO.fS exfS S1 . .txfS WdS S2 .We6+
'it>fS S3.Wh6+ 'it>f7 S4 ..te6+ 'it>e7
SS.WgS+ 1 -0
S. Koneru-Stefanova
U laan baatar, 201 0
1 .d4 dS 2 .e4 e6 3.tLle3 tLlf6 4.e3
a6 S.We2 g6 6 . .td3 .tg 7 7. tLlf3 0-0
S.O-O .tg4 9.tLleS .te6 1 0 .eS tLl bd7
1 1 .tLlxd7 tLlxd7 12 ..td2 fS 1 3. b4
.tf7 1 4.f4
An important typical position
with a closed centre. In the previ
ous game we have observed the
same pawn formation, but without
dark-squared bishops. In principle,
this exchange should favour White.
On the other hand, here Black's
pawn chain on the queenside is
weakened and she cannot keep it
close by . . . b5. To take stock, White's
game is easier. She will push b5 and
open files on the left wing. So Black
should hurry to create counterplay
or at least distract some white piec
es from the queenside.
1 4 . . . e6
This move looks ugly, but it is
the only way to achieve the break-
 1.4Jf3 dS 2.c4 c6 3.e3 !
through . . . g6-gS. 14 . . . 4Jf6 lS.e1
4Je4 16.h4 would have allowed
White to activate her passive bi
shop.
1 5.a4?1
This advance was not going to
run away. Besides, White had a
more urgent task. He should decide
how to meet . . . gS. The most con
sistent way is lS.!e1 gS 16.g3, but
then 16 . . . gxf4 17.4 eS 1B.dxeS
ig6 19.e6 4JeS 2 0 .4Je2 (or 2 0 J''1 a d1
4Jxd3 2U'1xd3 eB 2 2 .4Je2 xe6
23.4Jd4 e4 24.d2) 20 . . . 4Jxd3
21.xd3 ixa1 2 2 . xa1 eB 23.4Jd4
f6 24.f1 xe6 2S.4Jxe6 xe6
leads to a position where White's
better bishop is hardly enough for a
win. The further course of the game
suggests that White should make a
useful prophylactic move:
lS.@hl ! Now lS . . . gS fails to
16.fxgS xgS 17.e4, so Black should
prepare it by lS . . . 4Jf6 when we ac
tivate the knight - 16.4Ja4. (16.!e1
gS 17.4Je2 4Je4 1B.ixe4 dxe4 19.b1
gxf4 2 0 .4Jxf4 is also possible.) 16 . . .
4Je4 ( 1 6 . . . 4Jd7 i s too passive and
gives White a comfortable edge af
ter 17.e1 gS 1B.ig3 g4 19.4Jc3 hS
20.a4 h4 2 1.!e1 f6 2 2 .b1 h6
23 .bS axbS 24.axbS 4Jf6 2S.bxc6
bxc6 26.4Ja4;!;.) 17.4Jb6 bB 1B.e1
gS 19 .ixe4 dxe4 20.4Jc4.
1 5. . .g 5 1
l S . . . 4Jf6 does not make much
sense: 16.bS axbS (16 . . . aS only
makes a target after 17.bxc6 bxc6
1B J''1fb 1 c7 19.b3 fbB 20.ab1
lLle4 21.4Jxe4 fxe4 22 . .ie2) 17.axbS
4Je4 1B.4Jxe4 fxe4 19.i.e2. The
black pieces are very passive.
1 6.fxg5
Of course, White should not
open the g-file with 16.g3 gxf4
17.gxf4 hS 1B.bS f6? She could
also try 16.4Je2, intending to re
capture on f4 by knight. The prob
lem with this set-up is that Black
can exchange the knight for her
bishop and reach a drawish posi
tion, for example: 16 . . . gxf4 17.4Jxf4
e7 18.ab1 (lB.bS stumbles into
1B . . . axbS 19.axbS xa1 20.xa1 eS)
1B.f3 4Jf6 19.!c3 (against . . . eS)
19 . . . 4Je4 2 0 .ixe4 fxe4 21.g3 @hB
2 2 .e2 h6 23.f1 ixf4 = .
I t i s difficult for White t o avoid
this scenario. We clearly see that
she has missed the moment for ma
noeuvring the bishop to g3 via el.
1 6 .. .'xg5 1 7. g;, h 1
I t is late for 17.4Je2 e7 18.e1
(lB.@h1 4Jf6 19.!e1 4Jg4), because
Black can take over the initiative
with 1B . . . eS ! 19.4Jg3 (19 . .ixfS .ig6
77

2 0 .hg6 xfl+ 21.\t>xfl hxg6 2 2 .i.f2
f8 23. \t>gl exd4 24.ltlxd4 e8 is ex
cellent for Black.) 19 . . . i.g6 2 0 .ltlxfS
hfS 2 1.hfS exd4 2 2 .,ixh7+ \t>h8
23.i.fS d3 . I'd say that this posi
tion is unclear, e.g. 24.dl hal
2S.xal+ \t>g8 2 6.f3 xfS 27.xfS
f8 28.xf8+ xf8 29.d4 fSoo.
1 7 . . . .ih6 1 8 Jif3 ctt h 8
1 9 Jg 3
19.1tle2 g8 2 0 . g3 e7 21.h3
ltlf6 ! is equal. White would not be
able to avoid the exchange of her
knight.
The rook is more useful on gl,
but 2 0 .ltle2 g8 2 Uh3 g7 2 2 .gl
ltlf6 23.ltlf4 i.xf4 24.exf4 .illS is also
fine for Black.
The knight is shaky on g3, but
at least it covers the g-file and gives
White a respite, so she can pro
ceed with her opening plan on the
queenside. Play is double-edged,
but dynamically balanced.
78
Part 3
22 .. J:!g 8 23.bS axbS 24.axbS
trg 7 2S. bxc6 bxc6 26.tra4 ! tOf6
27 .trxc6 tOg4
A critical position. Black is
threatening with . . . ltlxh2. A good
defence is 28.c7 ltlxe3 (28 . . . ,ixe3
29 . .b:e3 lLlxe3 30.f4 lLlg4 31.f3;!;)
29.f3 xd4 30.xt7 xd3
31.,ixe3 he3 32 .f6+ g7 33.c6
c4 ! 34.xe6 xg3 3S.e8+ g8
36.xe3 xc6 37.xfS b7, with
a draw. However, White makes a
natural move:
28.gf3 tOxe3??
A horrible blunder. 28...
lLlxh2 ! was the logical continua
tion of Black's attack. The game
should finish peacefully then after
29.\t>xh2 i.hS ! 30.xe6 (or 30.f2
xg3+ 31.\t>gl .ixe3 32 .he3 xe3
33.d6 ! , with a compensation for
the pawn.) 30 . . . hf3 31.eS ! ! xeS
32.dxeS e4 33.he4 fxe4 34.lLlfS
and White even has some initiative
thanks to her fantastic knight.
29 J:txe3 f4 30.gxe6 .ixe6
3 1 .trxe6 fxg 3 32.trxh6 ga8
33.hxg3 ga 1 + 34.ctt h2 g d1 3S.J.e2
1 -0
 l.liJf3 d5 2 .c4 c6 3.e3 !
6. Postny-H ima ns h u
N e w Delhi 201 1
1 .d4 dS 2.c4 c6 3 . c3 f6 4.e3
a6 S.'lWc2 g6 6 . .id 3 .ig7 7 . f3
0-0 8.0-0 .ig4 9 . eS .ie6 1 0 .cS
bd7 1 1 . xd7 'lWxd7 12 . .td2 .ifS
1 3 . .txfS
1 3 . . . 'lWxfS
Black does not achieve good re
sults in this endgame, but in fac
this recapture is thoroughly con
sistent. After 13 . . . gxf5, White has
a clear plan with b2-b4-b5, while
Black's pressure along the g-file is
not too effective. The fine point be
hind the queen swap is that in the
endgame the pawn advance on the
queenside does not give White any
benefits. This game illustrates why.
On the other hand, I will show that
without queens White receives oth
er active options as the e4 break
through. All in all, Black is playing
for two results only and should de
fend with patience and precision.
1 4.'lWxfS gxfS 1 S.f3 d7 1 6.e2
e6 17 J:ac1
Black's only counterplay is ... e5
and White should be constantly cal
culating the consequences of this
break. For instance, 17.liJf4 would
have let in 17 . . . e5 18.liJh5 f4.
17 . . J:Uc8 ( 1 7 . . . e5 18.ii.c3) 1 8.
gc2
18.liJf4 would have stumbled
into 18 . . . eS so White decides firstly
to consolidate his centre against a
possible breakthrough and then im
prove his minor pieces. However,
I'm not entirely sure that the knight
would stand on d3 better than on
e2. By all means, White should
not take on eS if Black pushed it.
Instead, he should bolster d4 hop
ing to win the fS-pawn. In that sce
nario, the knight is best stationed
exactly on e2 from where it can go
to g3. The d2-bishop, on the con
trary, does need a better diagonal.
It could be shifted to g3, but:
18 . .taS ! ? was also a good option.
Let us examine the two possible ap
proaches of Black now - the passive
waiting and the active . . . eS:
79
 Part 3
a) 18 . . .'i!lf8 19Jc3 ! ?
We'll see later i n the game that
the plan with b4-bS is not enough
for an advantage because it opens
the a-file for the black rook. It is
much better to lift a rook to b3 and
bind the opponent with the defence
of the b7-pawn.
19 . . . gab8 2 0 .@f2 @e7 21.gb3
.if6 2 2 .gg1 ! ?
The mobility of White's heavy
pieces makes the difference. Black's
rooks are tied down to the b7-pawn
and would not be able to compete
on the kingside. White should be
looking to open play by g4 or e4. It
is good to drag the h-pawn to h4.
Then g3 would open a file and/or
produce a distant passer.
22 . . . hS 23.lLlf4 h4 24.lLld3 .ig7
2S.g3 hxg3+ 26.gxg3 (26.hxg3)
26 . . . .if6 27.gh3 eS 28.dxeS lLlxeS
80
29.lLlxeS .ixeS 30.f4 .if6 31.@f3t.
White will put his rook on h6 an
push h2-h4-hS to force the enemy
into a zugzwang .
b) 18 . . . eS 19.9fdl ge8 2 0 .@f2
.ih6 21.gc3t. The threat of gb3 is
very unpleasant.
1 8 . . . f8 1 9 .1Llf4
19 . .taS was still possible as 19 . . .
e S 20.gdl .ih6 21.gc3 i s good for
White.
19 . . . eS 20.lLle2 f4 was possible,
but White retains some pull with
2 1.gdl.
20.c!ild3 c!ilf8 2 1 . .ie 1 c!il g6
22 . .ig 3 h 5
A critical position. In my opini
on, White's only sensible plan here
is connected with e4. He should
have tried:
23.gc3 @d7 24.gb3 ga7 2S.gel
The idea of doubling rooks on
the b-file will be countered by ... eS:
2S.gc1 f6 26.gb6 eS. Another option
is 2S.@f2 f6 26 . .td6 if8 27 . .txfB
gxf8 28 .gel eS 29 .e4 fxe4 30 .fxe4
 l.lLlf3 dS 2.c4 c6 3 .e3 !
dxe4 31.l:l:xe4 It>e6 3 2 . lt>e3 l:l:f7= . As
we see, Black's plan with . . .f6 and
. . . eS is viable and White should be
precise with his move order. On the
contrary, the idea of . . . as is dubious
since the pawn will be doomed af
ter 2S.l:l:e1 as 26.l:l:a3 l:l:e8 27.l:l:a4 f6
28.lLlc1 eS 29.lLlb3;!;.
2S . . .f6 26 . .id6
Black cannot prevent e3-e4.
For instance, 26 . . . eS? would stum
ble into 27.e4. The only question is
which are the best settings for the
central break. Play might continue:
a) 26 ... .if8 27.,ixf8 l:l:xf8 28 .e4
dxe4 29.fxe4 fxe4 30.l:l:xe4;!;. Both
White's rooks are more active than
its black counterparts.
b) 26 . . . l:l:e8 27.g3
I have also checked the idea of
manoeuvring the knight to b6, but
Black's king arrives to c7 just in
time to help its rook to flee from
the cage : 27.lLlc1 .ih6 28.lLle2 .if8
29.M8 l:l:xf8 30.lLlc3 It>c7 31.e4 f4
32.exdS exdS= .
27. . ..ih6 28. lt>f2 lLlh8 (28 . . . eS
29.dxeS fxeS 30 .e4 fxe4 3 1.fxe4
d4 32 .l:l:f1 .if8 33.lt>e2) 29.e4 fxe4
30.fxe4 dxe4 31.l:l:xe4;!;.
2 3. b4?! 'it>d7 24.a4 h4 2S . .id6
.if8 26 ..ixf8 13:xf8 27.13:b1 f6 28.bS
axbS 29.axbS 13:fc8
White has fulfilled his plan, but
now the ultrapassive rook at a8
suddenly doubled its power. Black
has only to push eS to completely
balance the game.
Passive play usually costs dearly
in chess. Black had the obvious 3 1 . . .
e S with good counterplay, but per
haps Himanshu could not evalu
ate the position after 32.dxeS fxeS
33.fxeS <tt> e 6 34.lLlf4+ lLlxf4 3S.exf4
l:l:c3, which of course, cannot be
worse for him.
32.'it>f2 13:a8? 33.13:db2 e7
34.13:b3 g6 3S.'it>f3 h8??
Black does not realise that he
must seek to trade knight at any
cost. The reason for that is the pos
sibility of White to produce a pas
ser on the h-file. It is very strong In
a knight endgame while in a rook
endgame it would be nearly irrel
evant. So he should have stayed
81
 Part 3
active with 3S . . . eS 36.dxeS fxeS
37.fxeS me6 3S.lt'lf4+ It'lxf4 39.exf4
!'laS. After the text, Black can not
avoid an exchange of the rooks.
36.g 1 b2 c!Ll f7 37. bxc6+ bxc6
38.gb8 gxb8 39.gxb8 c!Lld8 40 . c!Ll b4
gc8 4 1 .gxc8 cbxc8 42 .g3 hxg 3
43.cbxg3+- cbd7 44.cbh4 rtJe7
45.cb h 5 cbf7 46.cbh6 e5 47.h4
exd4 48.exd4 c!Ll e6 49 .c!Ll xc6 c!Llxf4
50 . c!Ll b4 c!Ll g 2 5 1 . h 5 (Sl.c6 was
an easy win.) 51 . . . f4 52.c6? (S2.
It'l xdS+-) 52 ... c!Lle3 53.c7 c!Llf5+
S2
54.cbh7 c!Ll d 6 55.c!Lld3?
SS.lt'lxdS still preserved practi
cal chances after SS . . . f3 S6.lt'le3 f2
S7.mh6 lt'lcS SS. lt'lfl lt'ld6 S9.dS lt'lfS+
60.mh7 lt'ld6 61.lt'lg3 lt'lcS 62.mhS fS
63.mh7 It'ld6 64.lt'lxfS It'lxfS 6S.cSW
flW 66.We6+ mfS 67.d6 lt'lxd6:t,
55 . . . f3 56.c!Llf2 c!Ll c8 57.c!Llg4 c!Lld6
58.c!Ll h6+ cbe6 59.c!Llg4 cbf7 60 . h 6
c!Ll c8 61 .c!Llf2 c!Lle7 62 . c!Ll d 3 c!Ll c8
63. c!Llf2 c!Lle7 64. c!Ll h 3 c!Ll c8 65. c!Llf2.
D raw.
 Part 4

Anti-M era n I
1 . f3 d 5 2 . c4 c6 3.e3 f6 4 . c3 e6

83
Pa rt 4

Main Ideas

In the previous part, I have dealt

with Black's various ways to avoid
the Meran set-up. Now I will exa
mine positions where Black plays
4 . . . e6, but then he tries to trick us
out of our repertoire by exploiting
our particular move order.

1.f3 d5 2.c4 c6 3.e3 f6

4.c3 e6 5.b3 ! ? Now 7 . . . eS (7 . . . dxc4 8.ixc4)
8.cxdS cxdS 9.tt'lbS tt'lc6 1O.tt'lxd6
xd6 does not equalise completely,
because White's queen is not on c2
and instead he has made a useful
developing move : 1l.d4 e4 12 .tt'leS
ie6 13.tt'lxc6 xc6 14J:kl d6
15.0-0. Accordingly, Black prefers
in the overwhelming majority of
games 7 . . . tt'lbd7 when 8 . 0-0 (or
8.c2) , followed up by c2 , will
An important finesse. I f we take us to the main line of Part 6.
choose S.'c2, I have no idea what
to recommend against the rare se 5 . . . bd7 6.c2
quence S . . . id6 ! ? 6.b3 O-O ! 7.ib2
eS ! 8.cxdS cxdS 9.tt'lbS c6 1O.tt'lxd6
'i;Yxd6.
Of course, we can switch over to
the normal Meran with S.d4. After
all, we have won already a small
battle having sidestepped the Slav
and the Chebanenko with . . . a6.
However, I have different plans for
White.
Following S.b3 id6 6.ib2 0-0 I consider in this part 6 . . . i.e7
White should continue with 7.i.e2 ! ' and 6 . . .b6 while the main line 6 . . .

84
 l.lLlf3 dS 2.c4 c6 3.e3 lLlf6 4.lLlc3 e6

id6 is examined in the next two The d-pawrl. should remain on d2

parts.
A. 6 .te7!?
This humble development of the
bishop, connected with a fianchetto
of the queen's bishop, is under
rated. In my opinion, White cannot
at least for a while. Note that If
Black starts with 6 ... b6, then White
obtains an alternative - he can fa
vourably alter the pawn structure
by 7.cxdS because 7 . . . cxdS does not
work - see Line B .
7 .tb2 0 - 0 S .te2 b6

prove any advantage by conven

tional means.
Avrukh's whole repertoire
against the Slav is based on the as
sumption that the following posi
tion (reached after the move order
l.d4 dS 2 . c4 c6 3.lLlf3 lLlf6 4.e3 e6
S.b3 lLlbd7 6 . .tb2 b6 7 . .td3 .tb7
8.0-0 i.e7 9.lLlc3 0-0 lO.'!Wc2) is
better for White:
9.ggl ! ? .tb7 10 .g4 lLleS
H.g5 e5 12.d4 e4 13.lLle5 .td6
14. 0 - 0 - 0 !

I prefer White here. He has a

clear plan on the kingside with gS,
hS, gS-g6 (perhaps after .th3 first)
and he also stands better in the cen
tre. All the black pieces are passive.

Here he examines only lO . . . '!Wc7 B. 6 b6 7.cxd5

11.lLleS and 10 .. J3c8 lU3adl '!Wc7
12.lLleS. However, it is well known A daring alternative is 7 . .tb2
that Black should prevent lLleS by i.b7 8 .i.e2 i.d6 (8 . . . .te7 9.g4 ! ? or
1O . . . h6! and then put the queen's 9J3g1) 9.g4 ! ?
rook on d8. A fresh example is Aira
petian-Dreev, Rus. ch. 2011, which
went 11J3adl '!Wc7 12 .h3 l::1 ad8 13.lLleS
dxc4 14.bxc4 lLlxeS lS.dxeS lLld7
16.f4 lLlcS 17 . .te2 a6 18.i.f3 bSi.

I advocate a completely different

approach. A sterner test of Black's
set-up is the bayonet attack g2-g4.

8S
 Part 4

The text gives White clear play See more about White's play in
without any risks. the annotated game 7 Delchev
Collutiis, Plovdiv 2010. I also con
7 exd5 8.d4 .id6 9 . .id3 .ib7
sider the same pawn structure in
1 0 . 0 - 0 0 - 0 1l .ib2
. Part 6, line A3. The only difference
is the white bishop staying on the
passive square e2, but that matters
only if Black pushes . . . cS-c5. If he
adopts a waiting strategy, White ex
ecutes the same manoeuvre - .if1-
g2 , the rooks go to el and dl.

Points to remember:

Play S.b3 before c2 .

Always attack . . . e6-eS with cxdS followed up by tbbS.
The queenside fianchetto . . . b6, regardless of Black's move order,
can always be attacked with l:'!gl, g4-gS.
A good solid approach against . . . b6 is cxdS, followed by short
castling and the manoeuvre .ie2-fi-g2

86
Part 4
Step by Step
1. 1tJf3 d5 2.c4 c6 3.e3 1tJf6 4.1tJc3
e6 5.b3 ! ?
The move order with:
5.'lWc2 seems less precise. Then
S liJbd7 6.b3 transposes to the
. . .
main line.
Black has tried 5 . . . c5 in only 2
games. After 6.cxd5 exd5, we can
lead our bishop to e2 or b5 and play
against the isolated d5-pawn :
a) 7.d4 a6 8.i.e2 liJc6 9.0-0 i.e6
lOJ!d1 cxd4 11.liJxd4 liJxd4 12J!xd4
ic5 13JldU.
b) 7.i.b5+ liJc6
White is interested in trading
the bishops with 7 . . . i.d7 8.ixd7+
xd7 (8 . . . liJbxd7 9.'lWb3) 9.d4 liJc6
10 . 0-0 c4 11.e4;!;, with a lead in de
velopment.
8.d4 cxd4 9.liJxd4 i.d7 10.0-0
i.d6
Or 1O . . . liJxd4 11.exd4 !J..e 7 12 Je1
0-0 13.ixd7 (13.'lWb3;!;) 13 .. .'xd7
14.i.g5 Elfe8 15.'lWb3 Elad8 16.Ele5;!;;
1O . . . Elc8 11.liJxc6 bxc6 12 .i.a6;!;.
11.liJf5 ixf5 12.'lWxf5 0-0 13.Eldl.
However, at this moment I
have not a decent recommendation
against:
5 . . . i.d6 ! 6.b3 (It is not late to
return to the common lines with
6.d4 ! ) 6 . . . 0-0 ! 7.i.b2 e5 ! 8.cxd5
cxd5 9.liJb5 liJc6 1O.liJxd6 'lWxd6
This position is about equal :
11.d4 liJb4 12 .'lWd1 e4 13.liJe5 liJe8
14.a3 f6= , Grandelius-Al Sayed,
Bosna Open 2011;
11.Elc1 i.g4 12 .'lWc5 'lWe6 13.'lWb5
Elab8=, Jakubiec-Jaracz, Czech Rep.
2011. So the queen on c2 is not use
ful against . . . i.d6, . . . 0-0 and . . . e5.
87
 Part 4
5 tObd7
s . . . .id6 6 . .ib2 0-0 is a very im
portant deviation.
Note that 6 ... es 7.cxds cxds
8 . lLlbs lLlc6 9.lLlxd6+ xd6 10 . .ibs
e4 11.lLld4 0-0 is in White's favour:
12 . .ie2 [Hracek advocates 12 .c1 ! ?
1 2 . . . lLlb4 (12 . . . lLlxd4 13.,ixd4 EidB
14.b2;!;) 13.a3 lLla6 14 . .ie2;!;] 12 . . .
lLlxd4 13.i.xd4 .if5 14.0-0 lLld7
IsJ:kl (ls.f3 exf3 16.Eixf3 .ig6
17.Eif4t) Is . . . lLles, Dautov-Mueller,
Lippstadt 2 0 0 0 and now 16.f4 exf3
17 . .bf3 lLlxf3+ IB.xf3 .ie4 19.g4
g6 20.xg6 hxg6 2 1.Eic7 b6 2 2 .h4
f6 23J:!fcl;!; is rather pleasant.
White should answer 6 . . . 0-0
with 7 . .ie2 ! and now:
a) 7 . . . dxc4 B.,ixc4 !
After B.bxc4 es 9.d4 exd4 ! (9 . . .
lLlbd7 1O. 0-0 EieB 1 1.c2 e7 leaves
White more chances to fight for the
edge, for instance, 12.Eiael or 12 .a4
e4 13.lLld2 lLlfB 14.cs .ic7 Is . .ia3
lLlg6 16.lLlc4 .ig4 17.lLld6 ixd6
IB.cxd6 d7 19.,ixg4 xg4 2 0.f3
exf3 21.Eixf3 Eie6 2 2 .Eiafl;!;, Ehlvest
Shankar, Kolkata 1999) 10.exd4
.ig4 ! , Black solves the problem of
his queen's bishop development -
BB
11.0-0 lLlbd7 12 .h3 .ths 13.Eiel (13.
lLlh4 ,ixe2 14.lLlxe2 EieB Is.lLlfs .ifB
16.lLleg3 = , Tomashevsky-Motylev,
Sochi 2007.) 13 . . . EieB 14.c2 = .
I n two recent games White has
tried in vain to refrain from 9.d4 :
9.d3 EieB 10. 0-000 .tg4, Aronian
Gelfand, Leon 2010;
9.c2 .ig4 10.lLle4 lLlxe4 11.xe4
fS 12 .d3 ixf3 13.gxf3 e7oo, Vitiu
gov-V.Popov, Rus. ch. 2011.
B ... bs 9 . .ie2
In the rapid game Aronian-Gel
fand, Leon 2010, White retreated
the bishop to d3, but after 9 . . . .ib7
he hurried up with 10.lLle4 lLlxe4
11.,ixe4 when 11 .. .fs ! 12 . .id3 lLld7
turned out to be unclear. Instead,
he had to play 10J:kl ! and follow
ing 10 . . . lLlbd7 l1.lLle4 lLlxe4 12 .,ixe4
EicB 13.V;Ye2 fs retreat to bl, with a
slight pull. For instance: 14 . .ibl
e7 1s.0-0 es? ! 16.e4 f4 17.d4.
9 ... b4 (or 9 ... lLlbd7 10.lLlgs .ib7
11.c2) 10.lLla4 lLlbd7 11.Eicl .ib7
12 .d4 V;Ye7
This position has been reached
in Zvjaginsev-Khenkin, Herceg No
vi 2001. White should be somewhat
better after 13.lLld2 EiadB (13 . . . es
 l.lLlf3 dS 2 .c4 c6 3 .e3 lLlf6 4.lLlc3 e6

14 .lLlc4 fJ.c7 1S.lLlxeS ! lLlxeS 16.lLlcS 12 . . . h4+ 13.g3 e7 14.fxe4

ic8 17.dxeS; 13 . . . cS 14.dxcS lLlxcS
1S.lLlxcS .ixc5 16 . .ixf6 gxf6 17 . .if3)
14. c2 eS 1S.dxeS .ixeS 16 ..ixeS
lLlxeS 17.cS gd6 18. 0-0.
b) 7 ... eS 8.cxdS cxdS 9.lLlbS e4
Halkias-Rapport, Vrnjacka Ba
nja 2010, saw 9 . . . . lLlc6 10.lLlxd6
xd6 1l.d4 e4 12 .lLleS i.e6 when
13.lLlxc6 xc6 14.gc1 d6 15.0-0
gfc8 16.h3 is slightly better for
White owing to his bishop pair. Af
ter the exchange of rooks along the
c-file, Black's queenside will be vul
nerable as both of his minor pieces
are busy on the other wing.
1O .lLleS (1O.lLld4 also deserves
attention, planning to undermine
the centre later.) 1O . . . lLle8 1l.lLlxd6
lLlxd6 12.f3
dxe4 15.0-0 cud7. Here the simple
16.lLlxd7 fJ.xd7 17.i.a3 would have
given White some pull.
As a whole, this variation is a
blank spot in theory.
Another rare alternative is 5 . . .
b6 6.fJ.b2 .ib7. In Epishin-Jaracz,
Bratto 2 0 07, White has introduced
here an original set-up based on the
contra-fianchetto : 7.g3 cubd7 8 . .ig2
i.e7 9 . 0-0 0-0 1O .d4 c7 1l.gc1
gad8 12.cxdS exdS 13.cuh4 b8
14.cufS.
In a later blitz game, Kram
nik-Grachev, Wch. blitz Moscow
2010, Black put his bishop on d6
and was close to equalising: 8 . . .
i.d6 9.d4 0 - 0 10.0-0 h 6 1l.gc1 'fie7
12.cud2 gfd8 13.e4 dxe4 14.cudxe4
i.a3 1S . .ixa3 xa3 16.e2, but
16 . . . aS? ! gave White a lever on the
queenside: 17.gfd1 e7 18.cuxf6+
cuxf6 19.1LJa4.

The plan with g2-g4, as in the

main line, is more enterprising.

6.ti'c2

White has fulfilled his plan to

obtain a complex, nonstandard po
sition. His dark-squared bishop has
great potential. Look at the vari
ation 12 .. .f6 13.lLlg4 .ixg4 14.fxg4
cuc6 15.0-0 d4 16.b4 d3 17.b3+
'it>h8 18.i.d1 e7 19.9c1 gac8 and
now 2 0 .gS ! fxgS 2 1.i.g4t unleashes
the bishop pair. The game Val.Po
pov-Inarkiev, Russia 2 0 1 1 went:

89
 Part 4
I will examine here A. 6 . . . .te7
and B. 6 . . . b6 while 6 . . . .id6 is con
sidered in the next two parts.
A. 6 . . . .ie7 7 .tb2
7.d4 0-0 B . .td3 b6 9 . 0-0 .tb7
10 . .tb2 (10 J!dl 'ffc 7 11 . .tb2 h6) 10 . . .
h6! 11.!!adl (1l.cxd5 cxd5= ) 1l . . . 'ffc 7
is fairly level.
7 . . . 0 - 0 S .te2
B.!!gl e5 9.cxd5 cxd5 10 .tUb5 is
not with tempo so White would be
deprived of this resource.
S . . .b6
The modern treatment of such
positions is based on a pawn storm
on the kingside. However, the sac
rifice 9.g4 is dubious: 9 . . . tUxg4
10 .!!gl tUgf6 11. 0-0-0 .tb7. So we
should support the pawn by the
king's rook.
90
9.0-0 .tb7 W.!!acl !!cB ll.'ffb l
c5 12 .cxd5 tUxd5 13.tUxd5 hd5
14.d4 leads to a symmetrical posi
tion where Black should be able to
neutralise White's slight lead in de
velopment.
9 . . . .tb7 10 .g4 eS H.g5 e5
12.d4 e4 13.e5 .td6 14. 0 - 0 - 0 !
It is better to keep the centre
closed and shift the focus of the
game to the kingside. 14.f4 exf3
15.xf3 'ffe 7 16.e4 dxe4 17.'ffxe4
tUc7 1B.xe7 he7 19.0-0-0 !!feB is
only equal.
14 . . . xe5 15.dxe5 Axe5
16.q"bl!
Again, 16.tUxe4 hb2 + 17.xb2
regains the pawn, but gives Black
counterplay - 17 . . . 'ffe 7 IB.tUf6+
<;!;>hB 19.tUxeB !!axeB 20.cxd5 cxd5
21.<;!;>bloo.
16 . . . 'ffb S 17.h4 c7 lS.h5
geS 19 . .tg4t
I prefer White here. He has a
clear plan on the kingside with g5-
 1.4:lf3 dS 2 .c4 c6 3.e3 4:lf6 4.4:lc3 e6
g6 (maybe after i.h3 first) and he
can even exert pressure on the c
file by 1=!cl. All the black pieces are
passive.
B. 6 . . . b6 7.cxd5
I think that White should take
the opportunity to alter the pawn
structure in his favour since Black
cannot recapture by the c-pawn
(7 . . . cxdS? 8.lDbS).
If you like adventures, however,
you might prefer 7.i.b2 i.b7 8 .i.e2
i.d6 (8 . . . i.e7 9.g4 ! ? or 9.1=!g1) 9.g4 ! ?
A daring attack, entirely i n the
spirit of our book. The game Sal
vador-Danielian, Cento 2011, con
tinued 9 . . . h6 1O.1=!g1 'fffe 7 11.h4 lDeS
12 .gS? ! (12 .lDxeS ! .beS 13 .gS hxgS
14.hxgS lDd7 1S.f4 i.d6 16.cxdS cxdS
17.lDbS would have been more ac
curate.) 12 . . .hxgS 13.hxgS lDxf3 +
14.1xf3 lDd7 1S.cxdS lDeS? (15 . . .
cxdS !oo) 16.ie2 cxdS 17.f4 lDc6
18.lDbS 1=!c8 19.ixg7 ! + - .
In the diagram position, Black
could refrain from wakening his
kingside by 9 . . . 0-0, but castling
under the fire of all White's pieces
is also a risky decision. White can
try, for instance, 1O.gS lDe8 11.h4
eS 12 .i.d3 ! ? g6 (12 . . . e4 should be
losing to 13.lDxe4 dxe4 14.1xe4 g6
1S.hS+-.) 13.hS lDcS 14.i.e2 lDe6
1S.0-0-0 'fffe 7 16.lDh4--+. The battle
is entirely on Black's territory.
The text is more restrained and
it gives White clear play without
any risks.
7 exd5 8.d4 .td6 9 .td3 .tb7
. .
10. 0 - 0 0 - 0 1l .tb2
.
I prefer this position for White.
He can manoeuvre at ease, constant
ly threatening with e3-e4 or b3-b4-
bS. Black's only active plan is con
nected with . . . c6-c5, but then White
would have typical play against the
hanging central pawns. Look at the
annotations to game 7 Delchev
Collutiis , Plovdiv 2010.
91
Pa rt 4
Complete Games
7 . Delchev-Collutiis
E u ro Club C u p , P lovd iv 201 0
1 .tLlf3 tLlf6 2 .c4 c6 3 . tLl c3 d5 4.e3
e6 5.YlYc2 tLl bd 7 6.b3 b6 7 .cxd5
I think that the only sensible al
ternative to this exchange is 7 . .ie2
i.b7 8 . .ib2 .id6 9.g4 ! ?
7 . . . exd5 8.d4 iod6 9 .iod3 .ib7
1 0 .0-0 0-0 1 1 .iob2 ge8 1 2 .gfe 1 !
This pawn structure defines two
main points of conflict - White is
constantly threatening with e3-e4
while Black may seek counterplay
with . . . c6-c5. In practice, Black
avoids to commit himself too ear
ly with this breakthrough since it
would give the opponent clear play
against the hanging pawns. He
would like to shift firstly the queen
92
from d8 and the text is aimed ex
actly against 12 .. .'e7? ! - 13.e4 ! .
1 2 . . . gc8 1 3.gad 1
This rook should stay away from
c1 to avoid exchanges. 13 J'!ac1 c5
14.e2 lL'le4 15.i.a6 ha6 16.xa6
lL'ldf6 17.dxc5 i'!xc5 18 .h3 d7
19 .d3 i'!ec8 20.lL'le2 i'!xc1 21.i'!xc1
i'!xc1+ 2 2 . lL'lxc1 was only equal in
Portisch-Huebner, Manila 1990.
13.e4 does not bring any dividends
at this point: 13 ... dxe4 14.lL'lxe4 lL'lxe4
15.he4 lL'lf6 16 . .if5 i'!c7 17.i'!xe8+
xe8 18.lL'le5 cst, Polugaevsky
Karpov, Roquebrune 1992 .
1 3 . . .YlYc7
Black decides to evacuate his
queen to b8. In Atalik-Buono, Brat
to 2 0 05, this square was taken by
the bishop:
13 . . . .ib8 14 . .if5 ! c7 15.g3 i'!cd8,
but this set-up made effective
White's other active plan - 16.b4 !
a6 17.a4 b5 18.a5 d6 19 ..b3.
White has firmly blocked the left
flank and he can finally concentrate
on the centre : 19 . . . .ic7 2 0 . i'!a1 i'!e7
21.lL'ld2 h5 2 2 .e4 ! .
 l.ttJf3 dS 2.c4 c6 3.e3 ttJf6 4.ttJc3 e6
Let us examine now the typical
position with hanging pawns:
13 ... cS 14.dxcS bxcS. [The isola
tor after 14 . . . ttJxcS is somewhat dull
for Black, e.g. lS . .ifS ttJe6 16.\1;Vd3
(16.\1;Vbl .ib4) 16 . . . .ib8 17.EkU]
lS.i.fS ! Not only pinning the d7-
knight, but also bolstering the de
fence of White's castling position. It
seems at first that it is even better
to include ttJbS first: lS.ttJbS i.b8
16 . .ifS and win the exchange after
16 . . . a6 17 . .ixf6 ttJxf6 18 .hc8 \1;Vxc8,
but deeper analysis shows that
White would be unable to escape a
draw by perpetual after 19.ttJc3 (or
19.ttJa3) 19 . . . d4 ! = . With a knight on
c3, this idea is insufficient for equal
ity: lS . . . d4 (lS . . . g6 16.hd7 \1;Vxd7
17.ttJxdS ! ! hdS 18.\1;Vc3 l'!e6 19.e4
i.c6 2 0 .eS) 16.exd4 i.xf3 17.l'!xe8+
\1;Vxe8 18.gxf3 cxd4 19.1'!xd4 hh2 +
20.cbfl .ieS 21.l'!d3;!;. The pin i s still
a cause of concern to Black.
1 4.Wb1 gedS 1 5. g 3 !
A very important redeployment
of the light -squared bishop. From
g2 it will restrain any ideas with . . . c5
an d support e3-e4 at the same time.
1 5 . . . ge7 1 6 . .tf1 a6 1 7 . .tg2 WbS
1 S .We 1
The charm of this position is
that White can be manoeuvring for
many moves without risking to spoil
it significantly. The opponent's task
is much more difficult. In the dia
gram position I could have pushed
18.e4 dxe4 19.ttJxe4 ttJxe4 20.l'!xe4
l'!xe4 21.\1;Vxe4 ttJf6 ( 2 1 . . .c5 22 .d5)
2 2 .\1;Vc2 ttJdS (22 . . . c5 23.ttJgS)
23.ttJe5, with some initiative, but
I decided to play on accumulation
of small advantages and wait for an
opportunity to deal a tactical blow.
My next step is to trade my passive
bishop and launch a minority attack
on the queenside. I'm not afraid of
counterplay with 18 . . . c5 19.dxc5
bxcS because my pieces are well
set up exactly against this break
through. I could continue with
2 0 .ttJh4 d4 2 1.ttJa4 dxe3 2 2 . l'!xe3
l'!xe3 23.\1;Vxe3 hg2 24.cbxg2;!;.
1 S . . . geeS 1 9 . .ta3 geS 20 ..txd6
Wxd6 2 1 .ge2 ge7
Without dark-squared bishops,
the hanging pawns lose most of
their dynamic potential. Moreover,
the c5-pawn is rather weak: 2 1 . . .c5
2 2 . dxc5 bxcS 23.ttJel ttJb6 24.\1;Va3i.
93
 Part 4
22 J:e2 geeS 2 3. tLl e2 tLle4
24.tLlf4 as 25.tLld3 f6
I'm beginning to rip dividends
from my manoeuvres. Black's last
move weakens the dS-g8 diagonal
and the kingside as a whole. It also
takes away the f6-square from the
knight.
26.tLld2 tLl g 5 27 .Wb2 .la6 2S.tLlf4
Black should avoid exchanges of
knights as they are his only hope
for counterplay. For instance, 28 . . .
tLl e 4 2 9 J:!dc1 !ic8 30.tLlxe4 dxe4
3 1..ih3 g6 32 .a3 fS 33.b4 sets up
d4-dS and White's advantage be
comes tangible.
94
29.!ide1 tLlbS 30.!ie3 g5
It is becoming difficult for
Black to wait any more. After 30 . . .
g 6 31.c2 @g7 32 .h4 ttld8, every
thing is ready for 33.e4, e.g. 33 . . .
ttle6 34.ttlxe6+ l:%xe6 3S.exdS !ie1+
36.!ixe1 !ixe1+ 37.@h2 !ie2 38 .c1
cxdS 39.!ic7+;!;.
31 . tLl d 3 f5 32 .We2 !ie6?!
Finally a chance to bust the en
emy's centre has shown up ! 32 .. .f4
33.exf4 hd3 34.xd3 gxf4 3S.ttlf3
was still playable although the e-file
would be of no real use to Black.
Now the game ends in a few
moves:
33.tLle5! tLlxe5 34.dxe5 WdS
(or 34 . . .xeS 3S.ttlf3 f6 36.ttld4
!id6 37.ttlxc6) 35.f3 US 36.tLld4
gxe5 37.gxe6! f4 (37 ... ttlxc6
38.ttlxc6 c7 39.b2 ! !ife8 40.ttlb4
b7 41.ttlxa6 xa6 4 2 .f4 !iSe6
43.hdS is a mate in 16, according
to the computer.)
1 -0
 Pa rt S

Anti - M era n I I
1 . tLlf3 d S 2 . c4 c6 3 .e3 tLlf6 4 . tLlc3 e6 S . b 3 tLl b d 7
6 .9c2 id 6 7 .ib2

95
Pa rt 5
Main Ideas
1.f3 dS 2.c4 c6 3 .e3 f6 4.c3
e6 S.b3 bd7 6.Yc2 .td6 7 . .tb2
This part is the core of my book.
It is mostly based on my own re
search on the bayonet attack with
g1 followed up by g4.
The move 8.g1 ! ? (after 7 . . . 0-0),
has been introduced in 2002 and
has not accumulated enough prac
tice yet. It scores very well accord
ing to my database, achieving 67% .
Naturally, draws are seldom seen
in such strategically unbalanced
games.
I would like to stress strongly
that this sharp plan is not my weap
on of choice or at least it is not my
only repertoire. At the recent Euro
pean team championship I chose
the reliable and flexible ie2 , con
nected with short castling. It is a
perfect option if you want to avoid
risk. I analyse it extensively in part
6. Perhaps it is best to vary both
96
plans depending on the opponent
and the tournament situation.
If you do opt for g1, though,
you must play to the utmost of your
tactical abilities. The next 10 moves
are likely to decide the game. This is
even more valid for your opponents
though. Black often tries to follow
recipes which are familiar from the
calmer variations, and get crushed
right in the opening. I hope that my
book will help you contribute your
own share of nice miniatures.
Let's divide Black's defensive
plans into two groups : with an early
castling and without it. Another
major distinction is whether Black
plays . . . a6, or pushes . . . e5 without
this preparation. That leads to dif
ferent pawn structures.
1. Black does not castle and re
stricts g5 by . . . h6:
7" .a6 8J;gl ! ? eS (or 8 . . .'e7
9.g4 h6 1O .ie2 ! followed by h4)
9.g4 h6 (9 . . . e4 1O .g5). Now we
should not hurry to open the h-file
to the enemy rook. Instead, we can
complete development and strike
in the centre :
10. 0 - 0 - 0 ! e4 11. d4 eS
12.h3
 4.lLlc3 e6 S.b3 lLlbd7 6.c2 d6 7.b2
12 b5 (12 . . . 0-0 13.d3 ! ) 13.d3 !
exd3 14.hd3 bxc4 15 .te2 ! 0 - 0
16.f4.
2. Black plays . . . eS without . . . a6.
7 0 - 0 8.gg1 e5 9.cxd5
lLlxd5 (9 . . . cxdS 10 .lLlbS b8 11.c1)
1 0 . lLle4 .tc7 1l.g4! ge8 12 .g5
lLlf8 13.h4 .tfS
White owns the initiative. After
14.a3 .ig6, one way or another, he
will play h4-hS, even if he would
have to trade light-squared bish
ops. Then he should consider both
g5-g6 or hS-h6. See the detailed
commentaries of game 8 Wojta
szek-Krysztofiak., Leba 2004.
3. Black plays ... eS after ... a6.
7 0 - 0 8.gg1 a6
This move has not been tried
yet. It looks very consistent to me
so I have spent a lot of time on it.
9.g4 b5 10 .g5 e8 1l.h4! e5
12.h5 bxc4 13.bxc4 gb8 !
I n this extremely sharp position
White should take on d5 - not to
win a pawn, but to point two more
pieces at the black king.
14.cxd5 ! cxd5 15.lLlxd5 .tb7
16 .tc4!
I'm afraid, you should memorise
the rest of my analysis from "Step
by Step", line B. Both sides have to
follow a narrow path which reaches
a position where White retains the
initiative.
4. Black does not commit him
self with pawn moves in the centre.
7 0 - 0 8.gg1 'fJe7 9.g4 ! ?
When Black avoids action in
the centre, White can also spend a
tempo on development by 9.i.e2 ! ?
97
 Part 5

See game 9 Halkias-Haznedar tings and move orders. This is due

oglu, Antalya 2001. to the fact that . . . b6 does not ad
dress in any way the critical square
e5 and the fight for the centre. That
gives White a free hand on the
kingside.
Of course, we can also exchange
on d5 and enjoy a small, but lasting
edge as shown in other parts of the
book.

Black has delayed . . . e6-e5, but

he may be never capable of achiev
ing it anymore as White can suc
cessfully contest the dark squares:
Typica l Tactica l Motifs
a) 9 ... ia3 ? ! 1O .g5 tLle8 1l.!xa3
'l'Nxa3 12 .h4 f5 13.gxf6 tLldxf6 14.tLle5
This part deals with very unbal
'l'Nd6 15.f4, Polak-Petr, Ostrava
anced positions where tactical vigi
2010.
lance is often decisive. The follow
b) 9 . . . tLle5? ! 10.tLlxe5 !xe5 1l.g5 ing attacking guide should help you
tLld7 12.f4. sniff out possible blows from afar.

c) 9 ... tLle8 ! ? 1O.h4 b6 12.tLlg5 g6

If Black does not bar the main
13.f4 f6 14.tLlf3 e5
diagonal by . . . e5, he often becomes
victim of a i-'I'N battery:

Analysis

White's threats are more sub

stantial after 15.dxe5.

5. Black plays . . . b6.

1.h6 ! g6 2.cxd5 cxd5 3.tLlxd5 !
exd5 4.'l'Nc3+-.
The bayonet attack is effective
against the fianchetto in all its set-

98
 4.ltJc3 e6 5.b3 ltJbd7 6.'I1;!fc2 .id6 7 . .ib2
Nogueiras-Kirk
Dos Hermanas 2003
17.b4 ! ltJa6 18.ltJe4+-.
Sometimes disaster reaches
Black on both neighbour diagonals:
Petkov- Friedrich
Cesenatico 201 0
14 . .id3 ! e4 (or 14 . . . g6 15.hxg6
fxg6 16 . .ixg6 l'!xf3 17.hh7+ f8
18 . .ifS) 1S.ltJxe4 ! dxe4 16 . .ixe4+ - .
Chadaev-Potapov
Russian ch. 201 1
White continued with 12 .h5
b5? and got a big advantage after
13 . .id3 ! e4 14.ltJxe4 dxe4 1S . .ixe4.
However, Black could have pre
vented the debacle with 12 . . . e4. So
the best setting of White's idea is
12 . .id3 ! e4 13.ltJxe4 dxe4 14 . .ixe4
g6 1S.h5 ! .
Another interesting position
arises if in the last variation Black
chooses 12 . . . g6. White follows
up with 13.h5 ltJc5 14 . .ie2 lLle6
15.0-0-0 'I1;!fe7 16.lLlh4.
White will throw into the fire f2-
f4 to crack enemy's defence.
If a white knight can reach f6
(from dS or, more often, from e4),
we should always be on the alert to
spot possible tactics based on the
opening of the g-file :
l.lLlf6+ ! gxf6 2.g6 ! ! , with an at
tack.
99
 Part S

Here is another example :

We see the pawn structure

14.lLle4 ! ! (14.bxc4 fS ! is unclear)
14 . . . cxb3 IS.axb3. White has a
strong initiative. For instance: 15 . . .
c7 16.lLlf6+ lLl dxf6 17.gxf6 lLlxf6
1 8.1/;Yc3 id8 19J3a4 !
that arises after an early . . . eS. The
threats of lLlf6+ or gS-g6 restrain
Black's counterplay. White can fur
ther increase his striking power by
bringing the queen to the kingside
after <i!ffl, 1/;Ydl-f3.

Points to remember:

If Black plays . . . eS without . . . a6, we take on dS and install a

powerful knight on e4.
I f Black plays . . . e S after . . . a 6 , w e d o not take o n d S and proceed
with our attack on the kingside with 13gl.
If Black restricts gS by . . . h6, we complete development, castle long
and then play in the centre or follow up with h4.

100
Part 5
Step by Step
1.f3 dS 2.c4 c6 3.e3 f6 4.c3
e6 S.b3 bd7 6.c2 .id6 7 . .ib2
7 0-0
Black can try to take the sting
out of the bayonet attack g2-g4
by delaying castling, but his king
would be even more vulnerable in
the centre:
a) 7 . . . a6 8J!g1 ! ?
The calmer alternative 8.d4 0-0
9 .!e2 is examined in Part 6jline C.
White has a tiny edge in this set-up
without any risks.
8 . . . e5 (8 . . . 0-0 9.g4 b5 transpos
es to the main line) 9.g4
It is amazing to watch how poor
ly Black performs in practice versus
this plan.
The game Hauchard-Relange,
Nantes 1993, continued 9 . . . d4? !
1O.exd4 exd4 11.lLlxd4 e5 12 .!e2
,ixg4 13.f4, when Blacks is close to
losing as 13 . . . !a3 14.lLla4 costs a
piece.
9 . . . e4 looks more consistent.
Perhaps White should choose 1O .g5
exf3 1l.gxf6 lLlxf6 12.l'!xgn because
1O .lLld4 lLle5 1l.g5 lLlfg4 12 .cxd5 c5
13.xe4 cxd4 (13 .. .f5? 14.lLlxfS 0-0
15.lLlh6+ gxh6 16.f4+-) 14.exd4 0-0
15.dxe5 lLlxe5 16.0-0-0 l'!e8;; is not
too clear.
The engines like :
9 . . . h6. This effectively discour
ages further advancing with 1O .h4
because opening the h-file would
tip the balance in Black's favour:
10 ... e4 1l.lLld4 (1l.g5 hxg5 12 .hxg5
exf3 13.gxf6 lLlxf6 14.l'!xg7 .te6co)
1l . . . lLle5 12 .g5 hxg5 13.hxg5 lLlfg4
14.cxd5 cxd5 15.lLlxd5 l'!h2 16.lLlc3
lLld3+ 17.i.xd3 exd3 18.xd3 lLlxf2
101
 Part 5
19 .c2 lLlh3 2 0 .l'Ui lLlxg5=t. How
ever, 9 . . . h6 weakens the kingside
and makes future castling risky for
Black. White can complete develop
ment with :
1 O . 0-0-0 ! e4 1l.lLld4 lLle5 12 .h3
and hit in the centre, for example :
12 . . . b5 (12 . . . 0-0 13.d3) 13.d3 !
exd3 14.hd3 bxc4 15.i.e2 ! 0-0 16.f4
lLlg6 17.'j;>b1. Everything is ready for
a decisive attack.
b) 7 . . . e7 is a flexible move that
might transpose to the quite line of
Part 6 after B.i.e2 0-0 9 . 0-0. There
are some lines of independent sig
nificance, where White combines
B . .!e2 with 9J::i g 1. See game 9
Halkias-Haznedaroglu, Antalya
2 001. White can also reveal his ag
gressive plans immediately:
BJ::ig 1 a6 9.g4 h6 10 .i.e2
Black has not castled yet so it is
good to make a developing move
instead of h4.
1O . . . b5
The thematic 1O . . . e5 1l.h4 dxc4
12 .bxc4 g6 13.g5 hxg5 14.hxg5 lLlh5
15.lLle4 was much better for White
in Simonian-Semenova, Voronezh
102
2010 . The text, however, weakens
the c6 square.
1l.h4 i.b7
1l . . . b4 12.lLla4 !'igB 13.g5 hxg5
14.hxg5 lLle4 15.cxd5 cxd5 16.c6
!'ibB 17.i.xa6 cost Black a pawn in
Dobrov-Menadue, Caerleon 2007.
12 .g5 hxg5 13.hxg5 lLlgB.
The game Bogner-Ris, Neu
hausen 2007, continued here with
14.0-0-0;1;, but this might not be the
best option. If Black opens the b
file, White should be able to contest
it with !'ib1. Therefore,
14.d4 ! looks preferable:
White has a space advantage
and the better pieces :
14 . . . e5 15.cxd5 cxd5 16.dxe5
lLlxe5 17.d2 !'idB 1B. 0-0-0 or
14 . . .bxc4 15.bxc4 !'ibB 16.!'ib1 I!tfB
17.lLla4 eB 1B.lLlc5 (or 1B.c5 .!c7
19. 1!td2 lLle7 20.!'ihU) 1B . . . icB
19.cxd5 cxd5 20.lLlxa6.
The above examples show that
White develops a strong initiative
very quickly. Perhaps Black should
seriously think about preventing
g2-g4 by 9.!'ig1 h5 10 .h3 h4 ! ?
Now let u s return to 7 . . . 0-0.
 4.lUc3 e6 5.b3 lUbd7 6.Wc2 d6 7.b2
White is now facing a fateful
choice : B.E!:gl ! ? or the more conven
tional B.i.e2, which is the subject of
Part 6.
8.gg1 ! ?
The inclusion o f B.i.e2 E!:eB 9.E!:gl
is probably in Black's favour. After
9 . . . e5 W.cxd5 lUxd5 ll.lUe4 lU7f6
(the rook defends the e5-pawn ! )
1 2 .lUxd6 Wxd6 13.a3 i.d7 Black
stands solidly in the centre.
After B.E!:gl, main branches are :
A. B . . . We7; B. B . . . a6; C. B . . . e5
A minor alternative is B . . . lUe5
9 .lUxe5 i.xe5 W.g4 ! (10 .f4 i.c7
1 1 . 0-0-0 e5oo) W . . . i.xh2 ll.E!:g2
i.c7 12 .g5 lUeB 13 .i.d3 g6 14.0-0-0.
White only needs to put in E!:h1 to
target all his pieces at the oppo
nent's king.
A. 8 We7 9.g4 ! ?
I have examined the position
after 9.i.e2 ! ? under the move order
7 . . . We7 8 .i.e2 0-0 9.E!:gl earlier in
this chapter.
9 . . . lUe8 ! ?
Houdini and Rybka like this
move. I have seen in practice only:
a) 9 . . . i.a3? ! . The exchange of
this bishop is positionally unsound
as now Black's dark squares are
too weak: W .g5 lUeB ll.ha3 Wxa3
12 .h4 f5 13.gxf6 lUdxf6 14.lUe5 Wd6
15.f4 d4 16.lUe4 lUxe4 17.Wxe4 lUf6
1B .Wg2, Polak-Petr, Ostrava 2010.
b) 9 . . . lUe5? ! 10.lUxe5 he5 ll.g5
lUd7 12.f4 i.d6 13 .i.d3 g6 14.h4 lUc5
15.i.e2 dxc4 16.i.xc4 b5 17.b4 lUa6
1B.lUe4+-, Nogueiras-Kirk, Dos
Hermanas 2 003.
c) 9 ... E!:dB W .g5 lUeB ll.h4 a6
12 .h5 b5, Edouard-Guliyev, Livry
Gargan 2 009, when 13 .i.d3 g6
(13 . . . lUfB 14.h6 g6 15.cxd5 cxd5
16.lUxd5) 14.hxg6 hxg6 15.cxd5
cxd5 16.i.xg6 fxg6 17.lUxd5 would
have been crushing.
10 .h4
At first I analysed 10.0-0-0
lUc7 ll.d4 b6 12 .d3 g6 13.@b1,
but Black generates considerable
counterplay on the queenside : 13 . . .
a 5 14.h4 (14.e4 dxc4 15.bxc4 e5;
14.cxd5oo) 14 ... lUa6 15.i.e2 i.b7
16.h5 a4 17.lUxa4 lUb4 18 .Wd2 dxc4
19.i.xc4 b5 20.lUc5oo. Perhaps it is
safer to leave the king at e1. In the
ensuing complications I was in
trigued by the position after:
103
 Part S
I O . . . tOe7 11.d4
It is time to pay attention to the
centre. It is unclear how to proceed
after 11.liJe2 eS 12 .liJg3 g6 13.hS bS.
1l . . . b6 12. tOg5 g6 13.f4 f6
14. tOf3 e5
Despite the seemingly danger
ous attack of Black, White's threats
are more substantial. Here are
some crazy variations :
15.dxe5 .ie5 (1S . . . fxeS 16. 0-0-0
exf4 17.cxdS and 17 . . :xe3+ fails to
18.'tt> b1+-) 16.exd5 exd5 17. lOb5
tOxb5
17 . . . liJe6 18.''d2 .ib7 19.0-0-0
fxeS is also possible, but now White
takes over the initiative with 20.fS
.ib4 (20 . . . ltJg7 21.ltib1) 21.'lWh2 a6
2 2 .fxe6 'lWxe6 23.ltJgS 'lWe7 24.ltib1
axbS 2S . .ixbSt.
18 . .ixb5 .ib7
Or 18 . . . .ixe3 19 . .ic6 !'gb8
2 0 . .ixdS+ ltih8 21. ltifl fxeS 22 .fxeS
104
.ib7 23 . .ixb7 !'gxb7 24.'lWe4 !'gbb8
2S.ltig2 ! .ixg1 26.!'gxgl.
19 . .ixd7!
A solid positional solution.
19 .'lWe2 fxeS 20 . .ixd7 exf4 21 ..ibS
a6 22 . .id3 wins a piece, but com
pletely abandons the initiative to
the enemy: 22 . . . !'gae8 23.!'gg2 (23.
!'gfl 'lWxe3 24.ltid1 .ic8 2S . .ic1 'lWxe2+
2 6 . .ixe2 .ixg4 27 . .ixa6 .ih3 = ;
23 .hS 'lWxe3 24.ltid1 'lWxe2+ 2 S . .ixe2
.ixg1 26.ltJxg1 d41Xl 27 . .ic4+ ltig7
28 . .ixd4+ ltih6 29 . .ixb6) 23 . . . .ib4+
24. ltif1 fxe3 2S.ltig1 !'gf4.
19 . . . 'lWxd7 20 . .id4 .ixd4
21.exd4 e7 22 .'lWd2 fxe5
23.tOxe5 xh4+ 24.\tldl gae8
25.gel
The eS-knight is ruling over the
board, for instance, 2S . . . .ia6 26.fS .
B. 8 . . . a6
This move has not been tried in
practice yet, but it is very consis
tent. Black is aiming to open the
 4.tDc3 e6 s.b3 tDbd7 6.'!Wc2 .id6 7 . .ib2
queenside with . . . b7-bsxc4, so long
castling becomes quite problematic
for White. His next two moves are
clear:
9.g4 b5 10 .g5 tDe8,
but now he must make a major
decision.
Black is ready to play . . . es, but
Il .d4 bxc4 12 .bxc4 E:bS ! is not too
inspiring. The threat of . . .dxc4 fol
lowed by . . . cS is unpleasant with
White's king stuck in the centre !
Therefore, White should probably
go for a head-on attack with :
1l.h4! e5
Every tempo is precious. 11 . . .
1b7 12 .hS e S 13.1d3 ! e 4 14.tDxe4
smashes Black's defence.
12 .h5
Now 12 .id3 fails to 12 . . . fS !
13. hfs E:xfS 14.V!1xfS tDcs.
12 . . . bxc4 13.bxc4 gb8 !
Black has n o time for 1 3 . . . tDc7
due to 14.tDa4, thretening to win
the es-pawn after c4-c5. Following
14 . . . e4 Is.h6 ! g6 16.V!1c3 f6 17.gxf6
E:xf6 1S.cs .ifS I9.tDeS, White fulfils
all his opening aims.
14.cxd5 !
White has to take the pawn be
cause 14J''!b l tDc7 1S.1d3 ! ? fS ! (15 . . .
e 4 16.tDxe4 dxe4 17.he4+ - ; I S . . . g6
16.cxds cxds 17.hxg6 fxg6 IS.hg6
E:xf3 19.hh7+ 'ktfS 2 0 .V!1g6) 16.ixfS
E:xfs I7.V!1xfS tDcS lS.V!1c2 e4 19 .tDh4
tDd3+ 2 0 .'kte2 as is awful for him.
The immediate 14 . .id3 gives
Black the opportunity to show the
idea of his previous move : 14 . . .fS ! ?
(or 1 4 . . . e 4 ls.tDxe4 dxe4 16.he4 fS
17.hc6 tDcS lS .idS+ 1e6 19 .he6+
tDxe6 2 0 .h6 g6 2 1.V!1c3 E:xb2 ! )
IS.cxds e 4 16.tDxe4 fxe4 17.he4
E:xb2 ! IS.V!1xb2 tDcS, with an initia
tive.
14 cxdS 15.tDxdS ib7 16.ic4!
. .
It looks like White's pieces are
hanging, but the force (meaning
105
 Part S
tactics ! ) is with him. Black must
play very carefully to avoid the nu
merous tactical mines which could
explode at any moment. For in
stance:
16 . . J3cB looks normal, but then
follows 17.ltJf6 + ! gxf6 (17 . . . ltJ dxf6
1B.gxf6 xf6 19.1tJgS+-) 1B.g6 ! !
E1xc4 (1B . . . ltJg7 19.1tJh4) 19.xc4
hxg6 2 0 .hxg6 ltJg7 2 1.ltJh4! and
White's initiative is undisputable.
The material is about balanced,
but the safer king's position and
piece activity tip the scales in
White's favour:
a) 2 1 . . .ltJb6 2 2 .g4 i.cB 23.e4
eB 24.a4 ! . While Black is tied
down with the defence of his king,
White can improve on the oppo
site wing. 24 . . . aS (24 . . . i.d7 2S.aS
ltJcB 26.gxt7+ E1xt7 27.ltJfS hiS
2B.xfSt) 2S.i.c3 i.e6 26.i.xaS ltJdS
27.gxt7+ xt7 2B.E1b1 E1aB 29.h1 ! !
fB 30.ltJg6+ eB 31.E1bSt.
b) 2 1 . . .c7 2 2 .gxt7+ E1xt7
23.xc7 .ixc7 24.E1c1 i.bB 2S.f3
h7 26.E1h1 gB 27.i.a3t.
16 . . . ltJc7 is not any better due to
17.ltJf6+ gxf6 18 .fS ! ! hB 19 .i.d3
e4 20.gxf6 E1gB 21.E1g7.
106
Perhaps the most accurate an
swer is:
16 b6 17. f6 + ! gxf6
18.gxf6+ h8 19. xe5 xc4
After 19 . . . i.xeS 2 0 . .ixeS ltJd6,
White should probably opt for the
slightly better endgame 2 1.i.xd6
xf6 2 2 .c3 xc3 23.dxc3 ltJxc4
24 ..ixbB E1xbB 2S.E1b1 ltJeS 26.E1g3
ltJd3+ 27.e2 ltJcS 2B.E1gg1 since
21.E1g7, amazingly, is not winning
- 2 1 . . .i.e4 2 2 .i.d3 E1cB 23 .b1 E1eB
24 ..ixd6 xf6 2S.h6 (2S.E1xt7 xt7
26.i.xe4 f6 27.i.b4 E1xe4 2 B.xe4
xa1+ 29.e2=) 2S . . . i.g6= .
20.1::! g7 xg7 21.fxg7+ xg7
22. c6+ xb2 23.xd8
A critical position. I'm sure that
computers should be able to defend
it, but over the board Black's task
is much more difficult. His king is
exposed and the raving queen is
able to create tons of threats. Gen
erally, White aims to trade a pair of
rooks in order to hamper the coor
dination of the enemy pieces. Look
at this variation :
 4.tUc3 e6 5.b3 tUbd7 6.Wfc2 .id6 7 . .ib2
23 .. J'!fxd8 24.Wfxb2+ f6 25.Wfc2
.ie7 (25 .. J'!bc8? 26.h6+ 'i!7xh6
2 7.Wfb2) 26J'!b1 .id5 27J'!xb8 l'!xb8
28.Wff5 .ha2 29.h6+ 'i!7h8 30 .Wfd7
.if8 31.Wfc6 hh6 32 .Wfxf6+ .ig7
33.Wfxa6 .ig8 34.f4. Black must
keep his pieces together in order
to avoid double attacks, but that
allows White to start rolling his
pawns.
23 .bS
24.l'!bl l'!fxdS
25.l'!xb2 l'!bcS 26.ti'f5 h6 27.l'!c2
l'!bS 2S.!!k4
White has more chances to win
than Black to draw here.
C. S e5 9.cxd5 tOxd5
White will get an invasion
square on c7 after:
9 . . . cxd5 lO.lLlb5
The thematic 10.g4 also de
serves attention, for example, 10 . . .
tUb6 1l.g5 lLle8 12.lLlb5 e4
13.lLlxd6 (13.tOfd4 .ie7 14.h4
.td7) 13 . . .Wfxd6 14.lLld4 (14.lLle5
.tf5 15Jk1 l'!c8 16.Wfd1 l'!xc1 17.Wfxc1
lLlc7 18.Wfc3 lLle6+) 14 . . . .td7 15.f4
l'!c8 16.Wfd1, with active pieces. The
knight move aims to gain the bi
shop pair advantage.
10 . . . .tb8
The assessment of lO . . . Wfe7 de
pends on the sharp variation ll.l'!cl.
a) 1l ... e4 12.lLlfd4 hh2 13.l'!h1
i.e5 14.f4 (14.i.a3) 14 ... exf3 15.gxf.3.
b) 1l . . . d4 1Vt:Jxd6 Wfxd6 13.exd4
exd4 14.Wfc7 Wfe6+ 15.'i!7dl. Despite
the awkward placement of White's
king in the centre, the rest of his
pieces are so active that Black can
not avoid simplification. For exam
ple : 15 . . . lLle4 16.hd4 lLldf6 17.i.c4
Wff5 18.Wfe5.
c) 1l . . . lLle8 12.lLlxd6 Wfxd6 13 .g4
f6 14.g5 (or even 14.lLlh4 lLlb6
15.c5;!;) 14 . . . e4 15.gxf6 lLldxf6
16.lLle5 .if5 17.Wfc5;!;.
1l.l'!c1 e4 12.tOfd4 .hh2 (12 . . .
lLle5 13.tOc7 lLld3+ 14.hd3 exd3
15.Wfc5 .hc7 16.xc7 i.d7 17.f3;!;)
13.l'!hl. White has an overwhel
ming advantage for only one pawn.
The game Lushenkov-Demianjuk,
Minsk 2008, went on 13 . . . i.e5
14.f4 exf3 15.gxf3 a6 16.tOc7 hc7
17.xc7;!;.
107
 Part 5

In the event of 10 .. :e7, White

should not hurry to eat the passive
bishop on d6. After l1. liJxd6 xd6
12 . a3 e8 13.d3 c5 14 . .ie2 b6 15.g4
i.b7 16 .h4 as 17.g5 a4 18 .h5 b5
19 .96 fxg6 2 0 .hxg6 h6, Bocharov
Belozerov, Tomsk 2009, his play
on the kingside reached an im
passe. Perhaps he should prevent
. . . V-f5 by 1l.g4 ! ?
This idea has been realized in
practice with the insertion of 1l.a3
i.c7 1 2 .g4. The game Iturrizaga
Sharapov, Dubai 2 0 1 0 went 12 . . .
liJ7f6 13.liJxf6+ xf6 1 4 . .id3 g6
15.i.e4 e7 16.g5t
Undoubtedly, White owns the ini
tiative, so I can understand Black's
impatient next move, which turned
out to be a blunder:
16 .. .f5? 17.gxf6 liJxf6 18.,ixg6
hxg6 19.Y;Yxg6+ h8 2 0 . liJxe5 h7
2 1.liJV+ 1-0.
1l.g4
This is a model position for the
bayonet attack. It is practically un
tested as I found only 2 games in
my database.
White will probably leave his
king in the centre. His next moves
are g4-g5, h2-h4-h5. Then he
should reevaluate the situation de
pending on the opponent's moves.
Black's play is not so clear. 1l ... liJb4
12.c4 as does not change much
as after 13.a3 liJb6 14.c3 liJ4d5
15.c2 , Black will probably have to
play 15 . . . e8 anyway. Then 16 . .id3
g6 17.g5, followed by h2-h4-h5 is
pleasant for White. Thus Black's
most natural move remains :
1l ge8 12.g5 liJf8 13.h4 .ifS
White has some initiative. See
the detailed annotations of game
8 Wojtaszek-Krysztofiak, Leba
2004.

108
Pa rt 5
Complete Games
S . Wojtaszek-Krysztofiak
POL-ch U 1 S Leba 2004
1 .c4 e6 2 . c!LJf3 d5 3.b3 c!LJf6
4 .ib2 c6 5.e3 c!LJ bd7 6 .YlYc2 .td6
.
7 . c!LJc3 0-0 SJg 1 e5 9 .cxd5 c!LJxd5
1 0.c!LJe4 .tc7 1 1 .g4 geS
Black's knight cannot land on
the d3-square : 1l ... ttJb4 12 .c4 as
13.a3 ttJb6 14.YlYc3 ttJ4dS IS.c2 .
1 2 . g 5 c!LJfS 1 3 .h 4 .tf5
1 4. a 3 ! 1
The pawn structure i s deceptive.
One might argue that White plays
the Sicilian with colours reversed,
but the character of position is in
fact completely different. White
will try to keep as low as possible
on the queenside in order to avoid
weaknesses. He will push a3 to res-
trict the dS-knight, but b4 is out
of his plans because the left flank
could shelter his king after an even
tual castling. An essential point in
White's attack is h4-hS and it seems
sensible to push it without any de
lay:
14.hS. Then mundane continua
tions like 14 . . . e7 1S.a3 gadS would
favour White - 16 . .id3 hS 17.ttJh4
.icS IS.0-0-0.
The only way to refute 14.hS
would be an attack in the centre by:
14 . . . ttJe6 ! IS.gel (threatening
16.ttJf6+ ) IS . . . ttJd4
White has two possible cap
tures, but Black retains counterplay
either way:
a) 16.ttJxd4 exd4 17.ttJf6+ ttJxf6
18.xfS dxe3 19.fxe3 ttJe4 (19 . . .
ttJxhS? 2 0 .ic4 gfS 21.g6 ! +-) 2 0 .gc2
( 2 0 .gg2 ttJg3 21.f3 ttJxfl 2 2 . gf2
109
 Part S
d7 23.c.i> xf1 ge6 24.gcS gaeS
2S.gfS gSe7oo) 20 . . . tt:lg3 2 1.f3
tt:lxf1 2 2 .gxf1 d7 23.h6 .ieSoo.
b) 16 . .ixd4 exd4 17.tt:lf6+ tt:lxf6
lS.xfS tt:lxhS (lS . . . tt:l dS 19 .96 fxg6
2 0 .hxg6 dxe3 2 1.fxe3 h6 22 . .ic4t)
19 ..ic4 g6 20.xf7+ hS 2 1..ie2 gfS
2 2 .c4 .tb6 23.gg4 d7 24.ttJxd4
ttJf4 2S.ttJc2 ttJxe2 2 6.xe2oo.
Wojtaszek prepares to castle and
that is a reasonable choice. Now
14 . . . ttJe6 would not be dangerous in
view of lS.0-0-0. Black would not
be able to hold the blockade on hS
anyway.
1 4 . . . .ig6 1 5.d3 f5
This looks a bit impatient. I do
believe that Black cannot survive
with a passive defence against the
beast on e4, and sooner or later
he would be forced to move the f
pawn. However, he should have
waited for a better moment, for in
stance: lS . . . .ihS 16 . .ie2
a) 16 . . . ttJg6 17.ghl aS 1S.ttJg3 .ig4
19.hS. The blockade is removed and
White easily redeploys his pieces
for the attack: 19 . . . ttJfS 20.ttJe4 ttJe6
2 1.d1 e7 2 2 .tt:lh4 .ixe2 23.ttJfS
110
dS 24.xe2 d7 2S.tt:lh4 gadS
26.tt:lf3 e7 27.gdgl.
b) 16 . . . aS 17.tt:lh2 .ixe2 lS.xe2
tt:ld7 19.9g3 e7 2 0 .hS e6 2 1 .tt:lf3
fS 2 2 .gagl.
c) 16 . . . e7 17.tt:lfd2 .ixe2
lS.xe2 gadS 19.hS as 20.gg3 hS
21.gag1
Now Black can wait with :
c1) 2 1 . . .ttJe6 2 2 .ttJf3 ttJfS (Again,
22 .. .fS does not fully equalise. After
23.gxf6 gxf6 24.e1 ggS 2S.ttJh4
gxg3 26.gxg3 ggS 27.gxgS+ xgS
28.e2 , White is still better, even
without rooks.) Perhaps White can
try to shift his queen to the kingside
with 23.f1, e2 , ttJh4.
C2) 2 1 . . .e6 2 2 .ttJf3 d7 23.f1
(23.g6 ! ? fxg6 24.hxg6 h6 2S.hlt)
23 ... b6 24.e2 cS 2S.g6 fxg6
26.hxg6 ttJxg6 (26 . . . h6 27.ttJfgS)
27.ttJfgS ttJfS 2S .hS g6 29 .h6
g7 30.h4.
The early breakthrough in the
game made weaknesses and gave
White a free hand on the kingside:
1 6.gxf6 ttJxf6 17 .0-0-0 xe4
1 8.dxe4 f6 1 9 . ttJ g 5 ad8
 4.ttJc3 e6 S.b3 ttJbd7 6.Wc2 .id6 7 . .ib2
19 . . . h6 is not too helpful due to
2 0 .f4.
20 . .ic4+ <bhB
2 1 JxdB
This exchange would have been
justified only if it were winning by
force. White's rooks are very active
and he should have preserved both
of them by 2 U'!:dfl! bS 22 . .ie2 ttJe6
23.f4.
21 .. JxdB 22 .f4 V!Je7?
Loses by force. From f6 the
queen defends the critical g7-
square. I could not find anything
decisive after 22 . . . h6 23.ttJf3 lLld7
2 4.ie2 ib6 (24 . . .ih7 2S.@b1 '<Mfe7)
2S.fxeS he3+ 26.@b1 We6 27.ic4
eS 2SJle1 if4 29.e6 lLlf6 30 .'<Mfg2
ih7 (The endgame following 30 . . .
b S 31.ixf6 gxf6 32 .Wg4 id6 33.l"lgl
bxc4 34.'<Mfxg6 '<Mfxg6 3S.l"lxg6 cxb3
36.@b2;!; is slightly better, e.g.
36 ... l"leS 37.l"lxf6 @g7 3S .l"lf7+ @g6
3 9 . l"lfS l"lxe6 40 .hS+ @g7 41.eS)
31. lLld4 ieS 32 .lLlfS hfS 33.exfS
ixb2 34.'<Mfxb2;!;. White stands nice
ly, but would he be able to make
progress?
23.V!Jg2 (23.hS ! hhS 24.V!Jh2 .ieS
2S.Wg3+-) 23 . . . b5 24.lLlf7+ <bgB
25.f5 bxc4 26.fxg6 hxg6 27.xdB
V!JxdB 2B .V!Jc2 cxb3 29 .V!Jxb3+
<bh7 30.h5 V!Jh4? 3 1 . hxg6+ xg6
32 .V!Jd 1 1 -0
9. H a l kias-Haznedaroglu
Anta lya 2001
1 . f3 f6 2 .c4 c6 3 . c3 d5
4.e3 e6 5.b3 .id6 6 ..ib2 bd7
7 .V!Jc2 0-0 B ..ie2 V!Je7 9J:g 1
It is hardly the best idea to com
bine ie2 with l"lgl. In many varia
tions the bishop would be much
more dangerous on d3 . We have
also seen examples where Black
plays . . . eS-e4, . . . lLleS. Then White
breaks through with d3 when ie2
would be a loss of tempo. On the
other hand, White might argue that
. .. '<Mfe7 is not too challenging either,
compared to . . . a6, followed up by
. . . bS or . . . eS. As a whole, 9.l"lgl is
playable and faces Black with con
crete problems. On the contrary,
the sac 9.g4 is dubious although
in Wojtaszek-Cichocki, Dzwirzyno
2 0 04, White had some compen
sation after 9 . . . lLlxg4 1O.l"lgl lLlgeS
11.lLlxeS lLlxeS 12 .d4 lLlg6 13.0-0-0
111
 Part S

dxc4 14.hc4 eS lS.dxeS ixeS 16.f4
.if6 17Jdel ge8 18.lt>bl. I prefer
to use the g-pawn for repelling
the knight from f6 rather than for
opening the g-file.
9 . . . e5
Black has also tested 9 ... .ia3,
but this exchange is only good if
White had already taken on dS. Af
ter 1O .,ixa3 Wlxa3 11.g4 gd8 12 .gS
ttJe8 13.h4 a6 14.hS bS lS . .id3 ttJf8
16.ttJeS !b7 17.ttJe2 Wld6 18.f4,

After a year, Haznedaroglu im

proved against Karadeniz, Ankara
2 0 0 2 , with 9 . . . . dxc4.
Perhaps he was afraid of 9 . . . eS
10.cxdS ttJxdS 11.ttJe4. Indeed, 11 . . .
.ic7? ! 12 .g4 would b e similar t o our
familiar game Wojtaszek-Kryszto
fiak, but trading bishops with 11 . . .
.ia3 ! 12 .g4 hb2 13:xb2 ge8+ White dominated on the dark
would have been fine for him. squares in Martinez-Fenollar, To
1O.bxc4 eS 11.g4 ttJcS 12 .gS ttJe8 tana 2011.
13.h4 as 14.hS a4
1 0 .g4 ttJ b6 1 1 .g5 ttJe8

Here White chose lS.ttJe4 and

had a good game after lS . . . ttJxe4, but
lS . . . .ifS 16.d3 ttJxe4 17.dxe4 ib4+
18.ttJd2 hd2 + would have been
awkward. Instead, lS . .ia3 looks
good enough, for instance, lS . . . ttJe6
16.ixd6 ttJxd6 17.g6 hxg6 18.hxg6
ttJcS 19.9xf7+ Wlxf7 2 0 .Wlg6 !t. We
should never forget that endgames
are usually favourable for White in
the bayonet attack.
1 2.h4
White had a good alterna
tive here - 12.cxdS ttJxdS 13.ttJe4
.ifS 14.a3, similarly to Wojtaszek
Krysztofiak. I suppose, it would
even be the best option as 12 .h4 e4!
13.ttJd4 a6 would have been a total
mess. The threat of . . . cS forces the
answer 14.a4 cS lS.aS cxd4 16.exd4.

112
 4.lLlc3 e6 5.b3 lLlbd7 6.c2 d6 7 . .tb2
Now the greedy 16 . . . lLld7?
17.lLlxd5 e6 18. 0-0-0 would give
White a tremendous initiative, but
16 . . . dxc4 ! 17.axb6 cxb3 18.'?4'xe4
ie6 leads to an "unclear" (mean
ing unclear to me ! ) position. In the
game, Black erred and gave the op
ponent control of the centre.
1 2 . . .,le6? 1 1 3 .c5 ,lxc5 1 4.c!tJxe5
c!tJd7 1 5.c!tJxd7 hd7 1 6..id3 g6 1 7 .h5
.ia3 1 S.ha3 Y!Yxa3 1 9.c!tJe2 c!tJg7
20.c!tJg3
The exchange of the bishops
weakened the dark squares, but it
reduced White's attacking poten
tial. Therefore, White should prob
ably change his strategy and play
on a long-term positional advan
tage with 2 0.h6 lLlfS 2 1 .,txf5 hfS
2 2 .'?4'c3 f6 23.lLlg3 ig4 24.lLle4;!;.
This variation suggests that Black
needs his knight and should keep
it by 20 . . . lLle6 21.'?4'c3 Vlfe7 2 2 .f4 f5
with double-edged play.
20 . . JaeS
Black misses the chance to trade
his bishop: 20 . . . i.g4 2 1.i.e2 he2
2 2 . l!ixe2 gae8 23.f4 lLle6 24.l!if2oo,
and becomes clearly worse.
2 1 .f4 Y!Yd6 22.0-0-0 c5
23.<bb 1 ? 1
This i s overly prophylactic. It
seems that Halkias was very reluc
tant to define the pawn structure on
the kingside, but it was the best line
for him : 23.h6 ! lLle6 24.'?4'b2.
23 . . . b5
24.Y!Yb2?
24.h6 lLle6 25.'?4'b2 c4 2 6 .f5 was
an easy win. The text allows 24 . . .
lLlxh5 ! 25.lLlxh5 gxh5 26.gh1 ig4,
with a defendable position. Instead,
Black chose the thematic, but losing
breakthrough .. .f5. That was what
Halkias was waiting for so long:
24 . . . f5? 25.gxf6 Y!Yxf6 26. hxg6
hxg6 27 .Y!Yc2 c4 2S .,lxg6 cxb3
29.axb3 .1g4 30 .,lxeS .ixd 1
3 1 .Y!Yxd 1 1 -0
113
114
 Pa rt 6

Anti - M era n I I I
1 . tLlf3 d S 2 .c4 c6 3 . e 3 tLlf6 4 . tLlc3 e6 S . b 3 tLl b d 7
6.VNc2 J.d6 7 .J.b2 0 - 0 B .J.e2

115
Pa rt 6

Main Ideas

In this part, I examine an alterna paths. Let us examine Black's main

tive approach based on short cas options :
tling. Then White plays d4 versus
set-ups with . . . a6 and often delays
1. Black takes on c4 after B . . .l:'!eB
this move in the other systems.
or B .. :e7.
l.o f3 dS 2.c4 c6 3.e3 tLlf6 The immediate B . . . dxc4 is sel
4.tLlc3 e6 S.b3 tLlbd7 6.Wc2 .id6 dom seen. White can organize an
7 . .ib2 0 - 0 8 .i.e2 ! ?

attack leaving his king in the cen

tre: 9 .hc4 bS 10 . .te2 .tb7 11.tLlgS
a6 1 2 .h4 ! ? cS 13 ..tf3 hf3 14.gxf3 .
Black's defence is not trivial, be
cause he has not any levers to use
for counterplay.

B .. JeB or B .. :'e7 are more cun

ning continuations as White has
to define his plans. It is a bit late
for the bayonet attack, so we have
nothing better than castling.

B .. :e7 9.0-0 dxc4 1O.hc4 bS

B . .te2 i s a flexible development, 1l . .td3 ! ? (1l . .te2) 1l . . . .tb7
where White wants to exploit the
absence of d2-d4. Play might also
transpose to familiar lines from the
Anti-Meran, where White is slightly
better, but the real clash is deferred
for the middlegame. In my opinion,
White has enough pure Reti-style
alternatives, so I do not see any
need to resort to the well trodden

116
 4.llJc3 e6 S.b3 llJbd7 6.\1;Yc2 ii.d6 7.ii.b2 0-0 B . .ie2 ! ?
A typical set-up. White can put a
knight on e4, obtaining good pros
pects for a kingside attack:
12 .llJgS h6 13.llJge4 ii.c7 14.f4.
Here is a similar example: B . . .
E1eB dxc4 1O.,bc4 b S 11 . .ie2 (There
is nothing wrong with 11 . .td3, but
I have no practical examples to
show.) 11 . . . .tb7 12 .llJgS E1cB, Petkov
Massanet, Palma 2009.
The game went 13.f4 ! ? , also in
teresting is 13.llJce4 llJxe4 14.llJxe4
ii.bB 1S.f4 fS 16.llJg3 g6 17.b4.
2 . Black enters Sicilian struc
tures with colours reversed.
B .. JeB 9 . 0-0 eS 10.cxdS llJxdS
1l.llJe4 .ic7 1 2 ,lLlg3 !
White's general plan i s t o push
d4, possibly after some preparation
as in the game Grischuk-Kramnik,
Wijk aan Zee 2011:
12 . . . llJfB 13.a3 as 14.d4 llJg6
1S . .id3 ! exd4 16.llJxd4 .tg4 (16 . . .
.ieS 17J'aeU) 1 7. .tfS llJf6
White has the upper hand.
3 . Black fianchettoes his bishop.
He should do it by following the
move order with B . . . geB 9 . 0-0 b6,
because B ... b6 could be attacked
with 9.gg1 ! ? .ib7 1O .g4 eS 11.gS llJeB
1 2 . 0-0-0 llJc7 13.b1 llJe6 14.h4
\1;Ye7
1S.d4 ! e4 16.llJeS;!;. White is
clearly ahead with his offensive.
B . . . \1;Ye7 9.0-0 b6?! is also a mis
take due to 1O.cxdS exdS 11.llJd4 ! .
117
 Part 6

After 8 . . . E!e8 9.0-0 b6, I suggest
to define the pawn structure with
10.cxdS exdS 1l.d4 i.b7 12.E!fdl Vlfe7
13 .g3 !
Black has now two options: 9 ...
bS or 9 . . . eS. Any delay would be un
reasonable since White will favour
ably open the centre with e4: 9 . . .
Vlfe7 1O.0-0 b S 1l.e4 e S 12.cxdS cxdS
13.dxeS llJxeS 14.llJxdS.

a) 9 . . . bS 10. 0-0 ib7

It is important to know the vari

ation 1O . . . bxc4 11.bxc4 dxc4

White sets the manoeuvre

ie2-fl-g2 which would give him
a lasting edge due to the possible
breakthrough e3-e4 - see Part 4,
game 7 Delchev-Collutiis, Plov
div 201O.
Black's only counterplay is 13 ...
cS, but 14.llJbS ib8 1S.ifl a6 16.llJc3 12 .llJa4! cS? 13.dxcS llJxcS
id6 17.ig2 is pleasant for White. 14.llJxcS ixcS IS.llJgS+-.

1l.cS ic7 13.b4 eS 13.dxeS llJxeS

4. Black plays . . . a6 14.llJxeS ixeS

He can do it on move 8 or after

8 . . . Vlfe7 9 . 0-0. Against this particu
lar set-up, we should not wait with
d4 any longer.

8 . . . a6 9 .d4

IS.f4 !

A thematic way t o ensure the

supremacy of our bishop along
the main diagonal. I give three ex
amples on that theme in the "Com
mented Games" chapter. See for

118
 4.lLlc3 e6 5.b3 lLlbd7 6.c2 d6 7.b2 0-0 B . .te2 ! ?

in stance, my game 1 2 Delchev White players love this posi

Brunello, Porto Carras 10.11.2011.
b) 9 . . . e5 10.cxd5 cxd5 11.dxe5
lLlxe5 12.0-0
tion because they have clear play
against the isolator. In fact, it is dif
ficult to extract something tangible
with a standard manoeuvring. That
is why I chose to analyse in depth
two games where White assumes
some risks by launching a direct at
tack with f4.
See game 13 Delchev-Van
Wely, Melilla, 2 1 . 11 . 2 0 1 1 and
game 11 Ionov-Yevseev, St. Pe
tersburg 2011.

Points to remember:

If Black takes on c4, we recapture by the bishop and leave the

pawn on d2.
We play d4 early in the opening after B . . J3eB 9.0-0 b6 10.cxd5
exd5 11.d4 and against . . . a6.
We exchange on d5 after ... b6 as above, or after B ... e5.
In the Sicilian structure after B ... EieB 9.0-0 e5 10.cxd5 lLlxd5
1l.lLle4 i.c7, we play 12.lLlg3 ! and aim for d4.

119
Pa rt 6
Step by Step
l.c!Of3 d5 2.c4 c6 3.e3 c!Of6 4. c!Oc3
e6 5.b3 c!Obd7 6.c2 .id6 7 .ib2
0 - 0 S .ie2 ! ?
9 . . . e5 10.cxd5 c!Oxd5 11.tZle4 gives
A. 8 . . . Ei:e8; B. 8 . . . e7; C. 8 . . . a6
D. 8 . . . dxc4, E. 8 . . . b6.
8 . . . e5 9.cxd5 tZlxd5 10.c!Oe4 ic7
lLtZlg3 transposes to lines A or B.
A. S . . .geS 9. 0 - 0
I think that the early ie2 does
not fit in well to the bayonet attack,
because it may prove to be a loss
of tempo. Still, some strong play
ers test this plan in rapid games.
Zvjaginsev-Maletin, ch-RUS rapid
2 011, saw the risky sacrifice 9.g4
tZlxg4 10.Ei:gl tZlgf6 11.tZlg5 tZlf8 12.0-
0-0 e5 13.f4 d4 14.tZlce4 .if5 15.tZlxf7
120
he4 16.tZlh6+ with a perpetual.
9J"!:gl is more "solid", but then
Black extra options compared to
the line without ie2 . The e5-pawn
is already protected so 11 . . . tZl7f6 is
possible. After 12.tZlxd6 xd6 13.a3
.id7 White's game is even a little
worse.
AI. 9 . . . dxc4 ; A2 . 9 . . . e5; A3 . 9 . . .
b6.
9 . . .e7 transposes to 8 . . . e7.
Al. 9 dxc4 10.ixc4 !
10.bxc4 e5 11.d4 (further tricky
play with 11.Ei:adl is dubious due to
11 . . . e4 ! ) brings about a well known
variation from the Anti-Meran,
where White has committed his
dark-squared bishop to b2 too early.
 4.lLlc3 e6 S.b3 lLlbd7 6.c2 i.d6 7.i.b2 0-0 8 .i.e2 ! ?
According to Dreev (The Meran
& Anti-Meran Variations, Chess
Stars 2011), Black has fair chances
after the flexible 11 . . . e7, keeping
open the option of . . . e4, or 11 . . . exd4
12.exd4 lLlf8 .
lO b5
10 . . . lLlb6 has been played in se
veral games. Perhaps the best retort
is 11.i.d3 eS 12 .lLle4 lLlxe4 13.,txe4
g6 14.d4;!; exd4 IS.l'!ad1 although
11.d4 lLlxc4 12 .bxc4, as in Benko
Mednis, Pasadena 1978, is also in
White's favour. Black cannot free
his play with . . . eS, while the other
break, 12 . . . cS 13.lLlbS cxd4 14.hd4
eS IS.lLlxd6 xd6 16.heS l'!xeS
17.l'!adl, is also horrible for him.
1l . .ie2 .ib7 12. g5 gc8
This position has been reached
in Petkov-Massanet, Palma 2 009.
Now 13.lLlce4 lLlxe4 14.xe4 i.b8
1S.f4 fS 16.lLlg3 g6 17.h4 is prefera
ble for White, but maintaining ten
sion with 13.f4 ! ? was not bad either.
After 13 . . . a6 14.a4 e7 1S ..ihS g6
16 ..if3 eS? ! 17.fS, White was clearly
better.
A2. 9 e5 lO.cxd5 xd5
1O . . . cxdS is typically bad if White
has 11.lLlbS. The game Izoria-Perez
Rodriguez, Dos Hermanas Inter
net 2 006, went 11 . . . i.b8 12 .l'!ac1 d4,
when best would have been 13.gS
h6? 14.lLlxf7 'it>xf7 1S . .ic4+ 'it>f8
16.i.a3 + .
1l.e4 .ic7
We see a typical Sicilian pawn
structure with colours reversed.
However, Black's pieces are very
passive and misplaced. His only
active plan is connected with . . . f5
and a kingside attack, but White's
last move underlines the futility of
such hopes. White, on the contrary,
has two promising plans. The one is
to play on the queenside, but Black
should be able to oppose it by . . . as,
WJe7. The other one is to push d4,
121
 Part 6

possibly after some preparation

as in the game Polugaevsky-Piket,
Aruba 1994: 1 2 J!adl lLlf8. Here
13.d4 exd4 14.lLlxd4 is premature
due to 14 . . . .ixh2+ IS.lt>xh2 h4+
16.lt>gl xe4 17.xe4 xe4 18 . .tf3
e7 and Black can hold this position,
but White could insert 13.lLlg3 lLlg6
14.a3 as when IS.d4 ! ;!; is already
fine. Polugaevsky opted for slow
manoeuvring with 13.d3 e7 14.a3
as IS.fel .tfS 16.lLlg3 ig6 17.c4
lLlb6 18 .g4 d7 19 .1Llh4, with a
slight pull. The text seems like an I have been following the game
improved version of this game. See Grischuk-Kramnik, Wijk aan Zee
also game 10 D.Gurevich-Vigo 2011. White has the upper hand.
rito, USA 1998 for more explana Nothing can stop him from rolling
tions. the f- and e-pawns. Perhaps the best
way to do it is 18 .h3 hfS 19.1LldxfS
12 . . . tLlfS ieS 2 0 .f4 hb2 21.xb2;!;, although
18.f3 ihS 19.ael, as played, was
The attempt to build up a Maro not bad either.
czy bind by 12 . . . lLlb4 13 .b1 cS 14.a3
lLldS IS.c1 b6, as in Gritsak-Golod,
Bie1 2009, is dubious. White can at A3. 9 . . b6
tack it by 16.b4 ! . The same break is
possible as a real sacrifice after 15 . . . Now that the bayonet attack is
h 6 , e.g. 16.b4 ! cxb4 17.axb4 lLlxb4 no longer an option, this fianchetto
18 .d4 exd4 19.1Llxd4, with excellent is quite viable. Perhaps it is even the
compensation. Of course, instead most solid retort to White's set-up.
of retreating to 13 .bl, it would be
even more consistent to activate the 10.cxd5
queen on c4.
I like this exchange since the
13.a3 a5 ! symmetry would be bad for Black
due to 1O . . . cxdS 11.lLlbS. White has
13 . . . W'e7 14.b4 g6 IS.acl gives another reasonable option though.
White clear play on the c-file. The position after 10 .d4 ib7
11.adl e7 12.fel ad8 has been
14.d4 tLlg6 15 . .id3 ! exd4 topical lately. It seems that 13.id3
16. tLlxd4 J.g4 (16 . . . ieS 17.ael;!;) eS 14.cxdS cxdS IS.dxeS lLlxeS is
17 .ifS tLlf6
only equal so White has been test-

122
 4.ttJc3 e6 S.b3 ttJbd7 6:c2 .td6 7 . .tb2 0-0 8 . .ie2 ! ?

ing 1 3 . .tfl eS 14.dxeS ttJxeS IS.ttJxeS lithea 2008, should be met by

\WxeS 16.g3 hS IS.xe4;j;.

13.g3

White's bishops are more ac

tive than their black counterparts.
Now 17.cxdS, creating an isolani,
deserves attention, but 17.ig2 dxc4
18.bxc4 ic8 19.ttJe2, as in Hracek
Schulz, Germany 2010, is also A critical position. White needs
pleasant for White. another two moves on ie2-fl-g2 so
maybe Black should attempt to get
10 exd5 U.d4 .tb7 12.gfdl some counterplay here. The most
obvious try is:
I have played the same pawn 13 . . . cS 14.ttJbS .ib8 IS.ifl a6
structure (see Part 4, game 7 Del 16.ttJc3 id6 17.ig2
chev-Collutiis, Plovdiv 2010) and
I think it is very comfortable for
White. Basically, he plays without
any risks. If he completed devel
opment and redeployed the light
squared bishop to g2, his game
would be better. It is good to begin
this manoeuvre right away in order
to avoid counterplay in the cen
tre. The game M. Gurevich-Dreev,
Cap d'Agde 2002 saw 12 J''! a dl e7 Now 17 . . . c4 is dubious due to
13 .:i!fel :i!ac8 14.fS (14 ..ifl cS 18.bxc4 dxc4 19.e4, but the hang
lS.bl c41') 14 . . . e6 = . ing pawns in the event of 17 . . . :i!ac8
18.dxcS bxcS 19.ttJh4 e6 2 0 .d2;j;
1 2 . . e7 would also be a cause of constant
concern to Black.
12 . . ttJe4 13.ttJxe4 dxe4 14.ttJeS
. 13 . . . :i!ac8 is more clever. White
Wic7, Skembris-Katzourakis, Kal- follows with 14.:i!acl, when 14 . . .

123
 Part 6

iDe4 (or 14 . . . cS lS.i.f1) lS.iDxe4 B1. 9 , . .b6? !

dxe4 16.iDd2 cS 17.dxcS (17.dS .ixdS
18.iDc4 hc4 19 . .ixc4 iDf6 20 . .ibS This move is almost always du
l':!ed8 21..ixf6 Wxf6 2 2 .'?ffxe4 = ) bious against the Reti set-up with
17 . . . .ixcS 18.iDc4 iDf6 19.Wb1 We6 out d4 because White gains the
2 0 .iDeS i.dS 2 1.i.bS l':!ed8 22 . .ia6 bishop pair advantage with :
gives him the better pieces.
In the game Bischoff-Boudre, 10.cxd5 cxd5 (lO . . . exdS
Bad Zwesten 2 0 0 2 , Black opted for 1l.iDd4) 1l.b5;j;
a passive stand :

13 gad8 14.i.f1 f8 15.i.g2

g6 16.d2 i.b8 17.geU

White has more space for ma

noeuvring while Black's task is not
so trivial. White's main threat is e3-
e4, but he has another possible plan
with b3-b4. He can also follow my
set-up against Collutiis - l':!ad1, Wc1,
ia3 , trading the bishops.

1l,. .i.a6 !
B. 8 , . .'?ffe 7 9. 0 - 0
1l . . . ib7 is too passive. The game
Moskalenko-Alsina Leal, Barcelona
2008, went 12 .iDxd6 '?ffx d6 13.l':!ac1
l':!ac8 14.'?ffb 1 a6 lS.Wa1 l':!fd8 16.h3
h6 17.lLleS iDe8 18.lLlxd7 Wxd7
19.i.d4 Wd6 2 0 .Wb2 l':!xc1 21.l':!xc1
l':!c8 2 2 . l':!xc8 .ixc8 23 .Wc3 .ib7
24.a4 f6 2S.f4.

12.fd4!

White is also slightly better after:

I covered the risky plan with
9.l':!gl ! ? in the previous chapter.
Now Black chooses between :
B1. 9 . . . b6; B 2 . 9 . . . eS; B3. 9 . . . l':!e8 ;
B4. 9 . . . a6; BS. 9 . . . dxc4.
a) 12.l':!fc1 l':!fc8 13 .Wd1 obbS
14 . .ixbS ia3 ! lS.l':!ab1 (lS.ha3 !
Wxa3 16.lLld4;!;) lS . . . .ixb2 16.l':!xb2
Wb4 (16 . . . l':!xc1! 17.Wxc1 WcS
18 .Wxc5 bxcS 19 .d4 cxd4 2 0 .iDxd4
l':!c8 21.f3 'it>f8 2 2 .l':!c2 l':!xc2 23.lLlxc2

124
 4.ttJc3 e6 S.b3 ttJbd7 6 .'!Wc2 .id6 7 . .ib2 0-0 B . .ie2 ! ?

rJ;; e 7= ) 1 7. .ie2 ttJe4 1B.ttJd4 g6

19.1'!bc2 'lWfB 2 0 . .ibS;l;, Roiz-Le
Roux, Caleta 2011.
b) 12.l'!ac1 l'!fcB 13.'lWbl l'!cS (or
13 . . . .bb5 14 . .bbS .ia3 15 . .ba3
'lWxa3 16.ttJd4;l;) 14.a4 .bbS 1S.,hb5
a6 16 . .ie2;l;.

12 . . . hb5

In Shulman-Grover, Reykjavik
2010, Black had to part with both
bishops after 12 . . . l'!fc8? ! 13.ttJc6!
fB 14.ltlxd6 'lWxd6 1S.,ha6 l'!xc6
16.'lWd1 l'!c7 17.l'!c1 l'!d8 18 .'lWe2.
13.xb5 gfc8 14.'lWdl Ae5
15.d4 Ab8 16.gcU.
Black's queenside is weak.
I have analysed in line A the
game Grischuk-Kramnik which
reached the same position, only
with . . . l'!eB instead of . . . 'lWe7. This
difference allowed Black to rede
ploy his d7-knight via f8-g6. In the
diagram position he can follow the
same routine:

12 . ge8
H2. 9 e5 10.cxd5 xd5
The alternative 12 . . . ltl7f6 allows
.

White to take over the e5-square

10 . . . cxd5 1l.ltlb5 gives White a
with 13.d4 (He can also maintain
strong initiative: 1l . . . .ib4 12 .a3 (12.
the tension with 13.a3 ! ? e4 14.ltld4
C2Jg5 ! ? e4 13.l'!fc1) 12 . . ..b5 13.l'!ac1
g6 15.f3 exf3 16 . .ixf3 ltlg4 17.ltlde2
(or 13.l'!fc1 e4 14.ltlfd4 ltle5 15.a4
or even 13.d3.) 13 . . . e4 (13 . . . ltlb4
id7 16.Aa3 Ab4 17 . .bb4 xb4
14.'lWc1 exd4 15.i.xd4 ; 13 . . . exd4
IB .'lWc3;l;) 13 . . . e4 14.ltlfd4 ltle5 15.f4;l;.
14 . .bd4 .ib6 15 . .ib2;l;) 14.ltle5.

11.e4 Ac7 12.g3

This move is not obligatory at

all. White could leave the knight
i n the centre and play a very good
version of the Sicilian with 12.d3 ! ?
Then 12 . . .f5 13.ltlg3 (13 . . .f4 14.exf4)
would give White chances of coun
te rattacking by e4.

125
 Part 6

The game Davies-Wessman, cxd5 12.lLlb5 e 4 13.lLlxd6 Wixd6

Katrineholm 1995, went 14 . . . g6? !
15.i.c4 lLlb6, when 16.a4 ! (threat
ening i.a3) 16 . . . lLlxc4 17.bxc4 c5
18.a5 a6 19.h3 would have been in
White's favour. For instance: 19 . . .
cxd4 2 0 . exd4 e 3 2 1..b3 o r 1 9 . . J'3e8
2 0 .f4 exf3 2 U '3xf3 .ie6 2 2 J'3af1 t
cxd4 23.exd4 he5 24.dxe5 Wic5+
25JUf2 lLld7 2 6.Wid2 ! ! , with a deci
sive attack.
Stronger is 14 .. J'3e8 ! , intending
to capture on e5. Perhaps White
should opt for a tiny edge after
15J'3fd1 lLld7 (15 . . . a5 16.a3) 16.Wixe4
(16 . .ic4 lLlxe5 17.dxe5 i.xe5 18 .i.xd5
cxd5 19 J'3xd5 i.xg3 2 0 .hxg3 f6
2 Uk5 i.d7 should be level.) 16 . . .
lLl5f6 17.Wif4 lLld5 18.lLlf5 Wif6 19.Wie4
lLlxe5 2 0 .lLlg3 l'3e7 21.dxe5 i.xe5
2 2 .he5 l'3xe5 23 .Wid4 l'3e8 24.1'3ac1
Wixd4 25.1'3xd4 i.e6 26.lLle4;\;.
13.d3 (aimed against . . . e4)
13 . . . tLlfS 14.a3 tLlg6 15.gfel Ad7
16.h3
14.lLle5;\;) 12.ll:lxd5 cxd5 13.dxe5
lLlxe5 14.1'3fdU.
1l.,ixc4 b5
a) 1l . . . e5? 12.lLlg5 l'3f8 13.f4 ! (13.
lLlce4 is also good.) 13 ... exf4 14.exf4
h6 (14 . . . lLlb6 does not help either
due to 15.1'3ae1 Wic7 16.lLlce4 lLlxe4
17.Wixe4 g6 18.f5 ! ) 15.1'3ae1 Wid8,
Polak-Alaimo, Siracusa 2 0 07. Here
simplest would have been 16.lLle6
fxe6 17.1'3xe6, winning a pawn.
b) 1l . . . h6? ! addresses the weak
ness of fl, but exposes g6. White
can exploit it by 12 .1'3ad1 e5 13.lLlh4!
lLlf8 At last all the light squares
are covered, but now the central
dark squares are hanging: 14.lLle4
lLlxe4 (14 . . . i.c7 15.dxe5 lLlg4 16.lLlf3
i.f5 17.Wic1 he4 drops the queen
after 18 .i.a3+-.) 15.Wixe4 exd4
16.Wixd4+- i.e5 17.Wixe5 Wixe5
18.i.xe5 1'3xe5 19.1'3d8.

There is a long manoeuvring

ahead.

B3. 9 . . . ge8

Black practically forces White to

define his future plans about the d
pawn.

10 .d4 dxc4

Alternatively: 12 . .id3
a) 1O . . . b6 transposes to line A3 .
12 .i.e2 is more passive although
b) 1O . . . e5 1l.cxd5 lLlxd5 (11 . . . things are not too clear. I have

126
 4.liJc3 e6 S.b3 liJbd7 6.%Vc2 .id6 7 . .tb2 0-0 S . .te2 ! ?
analysed 1 2 . . .b7 13.liJgS ! ? gacS
( 13 . . . h6 14.liJge4 .tc7 1S.a3 aS 16.gfcl;
13 . . . eS 14.liJce4 liJxe4 1S.liJxe4 exd4
16.liJxd6 %Vxd6 17.gad1 cS 1S.hbS
a6 19 .hd7 %Vxd7 2 0 .f3 %VbS 21.exd4
ge2 2 2 .%Vc3) 14.liJce4 cS ! ? C14 . . . .tbS
IS.liJcS liJxcS 16.dxcS eS I7.b4;!;) and
suddenly Black obtains consider
able counterplay, e.g. 1S.hbS liJxe4
16.liJxe4 hh2 + 17.<;t>xh2 %vh4+
IS.<;t>g1 he4 19 .%Ve2 cxd4 ! 2 0 . gfc1
( 2 0.gac1 hg2 =) 20 . . . liJeS.
12 . .tb7 13. g5
White has not much of a choice.
13.liJe4 ttJxe4 14.he4 fS ! 1S . .td3
cS ! offers Black the initiative.
13 .e4 eS 14.ttJe2 is unclear, but
Black's pieces are harmoniously
placed.
13 h6
13 . . . hh2 + 14.<;t>xh2 ttJg4+ loses
to IS.<;t>g3 %VxgS 16.f4.
14.ge4 IJ.c7
Black has two possible plans
here. One of them is to defend the
bS-pawn by . . . a6 and push . . . c6-cS.
Or he can open the centre by . . . eS,
.
leaving . . . cS for an appro ate mo
ment. It seems that 1S.a3 eS 6.gfe1
as 17.gac1 exd4 1S.exd4 %VfS i fair
ly balanced: 19.ttJxf6+ ttJxf6 20 ttJe4
ttJxe4 21.gxe4 gdS so White s ould
firstly activate his rooks :
Now 16 . . . eS is not good ow
ing to 17.ttJxf6+ ttJxf6 1S ..1fS exd4
19.9xd4. The game Sasikiran
Fressinet, Haute Evry 200S, went
16 . . . a6 17.ttJe2 ttJxe4 1S.he4 ttJf6
19 . .tf3 .tb6 when instead of 2 0 .e4
cS= , White could have preserved a
tiny edge with 2 0 .%Vb1 cS 2 1.hb7
%Vxb7 2 2 . dxcS hcS 23.hf6 gxf6
24.ttJg3 lJ.e7 2S.ttJhS gxc1 26.%VxcU
due to his safer king.
B4. 9 a6 10 .d4 b5
1O . . . eS 1l.dxeS ttJxeS 12.cxdS
cxdS leads to a typical IQP position
with Black's queen already com
mitted to e7. White can follow the
plan of the model game 11 Ionov
Yevseev, for instance, 13.gad1 1J.e6
14.ttJxeS heS 1S.f4 1J.c7 16 . .1f3 %VcS
17.%vf2 gadS 1S.ttJa4 %Ve7 19 .1J.d4,
D.Gurevich-B.Kogan, Estes Park
19S6.
11.e4 e5
1l ... b4 12 .eS bxc3 13.exd6 is
clearly better for White.
127
 Part 6

12.cxd5 f5 15.ltlg3 (15.ltld6 ! ? ixd6 16.cxd6

White has many paths to a
significant advantage, for exam
ple: 12 .exd5 cxd5 13.c5 ii.c7 14.c6
exd4 15.cxd7 dxc3 16.dxc8 l:!fxc8
17.ii.cl, with an extra piece.
12 . . .cxd5 13.dxe5 xe5
14.xd5
I do not understand what Black
was expecting from this line in Gel
fand-Smeets, Nice 2010.
xd6 17.f4t) 15 . . . g6 16.ii.c4+ I!;>g7
17.f4 ltlxc5 18.d4 exd4 19.exd4 ltle4
2 0 .xe4 fxe4 21.d5+, Miltner-Pel
letier, Germany 2 007. Black's king
will feel uncomfortable at h6.
b) 12 ... ic7 13.a4 (Anastasian
T.L.Petrosian, Yerevan 2006, saw
13.l:!adl f8 14.xf6+ xf6 15.f4
h6? ! 16.f5, but it was stronger to
keep the queen in the centre with
15 . . .'e7.) 13 . . . ltlxe4 14.ltlxe4 ltl c5
15.ltlxc5 xc5

B5. 9 . . . dxc4 10 .hc4!?

This plan, based on the idea to

keep the pawn on d2 and control
the centre with f4, is completely
new in this particular position.
Another way of deviating from
the big theory is:
16.ii.d3 ! . Piece play above all!
10.bxc4 ! ? e5! Il.ltlg5 ! (instead of The d-pawn can wait. 16 . . . h6 17.ii.a3
transposing to the Anti-Meran with a5 18.!Ubl ii.d8 19.ii.b4 c7 2 0 .a5
1l.d4) ii.e6 21.c5 a6 2 2 .ii.c3t. The e5 and
b7-pawns are weak. Later White
1l . . . c5
can finally open the centre with d2-
Another developing move is : d4.
1l . . . l:!e8 12.ltlge4
12 .d4 exd4 13.exd4 ltle6 14.ltlxe6
ixe6 15.ii.d3

a) 12 . . . xe4 13.xe4 ic7 14.c5

128
 4.1',c3 06 5.b3 1ilbd7 6."IIc 2 .t.d6 ..tb2 0-0 B . .t.02 ! ?
The mobile pawn pair i n the
centre gives White an initiative. I
have analysed here:
a) IS . . . ,ixh2+? 16.xh2 lLlg4+
17. g3 hS lS.f4 E:feS 19.E:hl.
b) IS . . ..ic7 16.E:fel E:adS 17.lLle4
tDxe4 IS.,ixe4 d6 (IS . . . h6 19.dS)
19 .,ixh7+ hS 2 0 .g3.
c) IS . . ..ig4 16.E:ael c7 17.h3
1hS IS.cS h2 + 19.hl f4 2 0 .dS
cxdS 2 1.,bh7+ hS 2 2 .lLlbS.
c) IS . . . E:feS 16.E:fel c7 17.h3
1d7 IS.lLle4 lLlxe4 19.he4 h6
2 0 .E:adU.
10 b5
Or lO . . . eS 1l.lLlgS!
a) ll ... bS 12 .d3 h6 13.lLlge4
ibS 14.lLlxf6+ lLlxf6 IS.lLle4:t.
b) 1l . . . lLlb6 12 ..id3 h6 13.lLlge4
ltJxe4 14.lLlxe4 !c7 IS.a3 fS (15 . . .
!::1 d S 16.f4 fS 17.lLlg3 e 4 IS.!e2
ltJdS 19 ..ic4 f7 2 0 .lLle2) 16.lLlcS
e4 17.!e2, with typical for the
Tai manov (with colours reversed)
pressure on the queenside.
c) 1 1 . . .h6 12.lLlge4 fi.c7 13.a4!
(threatening with .ia3) 13 ... E:eS
J
14.aS! (A more straightforward
plan is 14.a3 dS IS.b4.) 14 . . . a6
IS.f3 ! :t.
1l . .id3 ! ?
1l.!e2 !b7 12 .lLlgS h 6 13.lLlge4
1i.c7 14.f4 is also a worthy idea.
1l . . . .ib7
Perhaps Black should prevent
lLlgS with 1 1 . . .h6. White can then
choose 12 .lLld4 .ib7 13.f4 a6 14.lLle4
c5 IS.lLlxd6 xd6 16.lLlf3 E:fdS
17.E:adl or 12 .a4 b4 13.lLle4 lLlxe4
14.he4 !b7 1S.aS a6, with complex
play.
12. g5 h6 13.ge4 .ic7 14.f4
e5 15.g3 !
White has good prospects for a
kingside attack.
c. 8 . . a6 9.d4!
.
White has lost a tempo on .ie2
so the bayonet attack is not attrac-
129
 Part 6
tive anymore. Black is the first to
start an attack. The game Fressi
net-Spoelman, Wattenscheid 2011,
seems to confirm such an assess
ment: 9Jgl bS 1O .g4 bxc4 1l.bxc4
b8 12 .gS tile4 13.tild4 V!:Jc7 14.tilxe4
dxe4 IS.c3 cS+.
Cl. 9 . . . bS; C2. 9 . . . eS.
Cl. 9 . b5 10. 0 - 0 j,b7
Black does not succeed in elimi
nating the centre by 10 ...bxc4 ll.bxc4
dxc4, intending 12 .hc4 c5 due to:
12.tila4 ! . Now 12 . . . cS? 13.dxcS
tilxcS loses to 14. tilxcS hcS
IS.tilgS+-, Spirin-Collutiis, Aman
tea 2011, but 12 . . . b7 13 .tilcS ! ?
leads t o similar consequences.
H.e5
ll.adl is a bit too mun
dane. White will have to close the
queenside anyway. For example :
1l . . . e8 (1l . . . bxc4 12 .bxc4 cS
13.cxdS exdS 14.dxcS tilxcS IS.tild4
g6 16.tila4 c8 17.tilxcS xcS I8.V!:Jb3
V!:Jc7 19 .a3;!;) 12 .cS (or 12 .a3 bxc4
13 .bxc4 cS =) 12 . . . c7 13 .b4 eS!
130
Or 13 ... aS? ! 14.a3 V!:Je7 (14 . . . eS
IS.tilxbS cxbS 16.c6 hc6 17.dxeS
tilxeS 18.tilxeS heS 19 .heS xeS
20.V!:Jxc6;!;) IS.fel eS 16.tilxbS cxbS
17.c6 ; 13 V!:Je7 14.a4 eS IS.axbS
.
axbS 16.tilxbS cxbS 17.c6;!;.
14.e4 tilxe4 (14 . . .dxe4 IS.tilxe4
tilxe4 16.dxeS tilexcS 17.bxc5 V!:Je7
18.tilgS g6 19.e6 fxe6 2 0 .tile4; 14 . . .
exd4 IS.tilxd4 tilxe4 16.tilxe4 xe4
17.d3 h4 18 .g3 hS 19.fel)
IS.tilxe4 dxe4 16.tilxeS tilf6 17.f3 e3 !
18.fel tildS, with counterplay.
12 . . j,e7 13.b4
White has a considerable space
advantage. The inclusion of 13 . . .
as? ! 14.a3 i s not recommendable,
because 14 . . . eS could be punished
by 14.tilxbS e4 IS. tileS cxbS 16.c6
c8 17.cxd7 hd7 18 .bxaS haS
19 .V!:Jb3 b8 2 0 .c3 c7 21.fc1,
with total domination.
13 . e5 13.dxe5
13.e4 dxe4 14.tilxe4 tilxe4
IS.V!:Jxe4 tilf6 16.V!:Jc2 e4 17.tileS e8
is rather unclear.
 4.ttic3 e6 S.b3 ttibGr6.WI'c2 .id6 7 . .ib2 0-0 8.i.e2 ! ?
1 3 ttixe5
. 14.xe5 he5
15.f4 i.e7
White has the better pieces. See
my game 12 Delehev-Brunello,
Porto Carras 10.11.2011.
C2. 9 e5 10.exd5
. exd5
11.dxe5 xe5 12. 0 - 0
A very important position for
the i.e2 set-up. It can arise after dif
ferent move orders and White can
not avoid it, but he has nothing to
complain about. The isolator on dS
gives him an easy game with plenty
of potential for manoeuvring.
12 i.e6
12 . . ..ig4 13.ttixeS heS 14.hg4
ttixg4 lS.h3 ttif6 exchanges more
pieces, but Black is playing for two
results only.
See game 13 Delehev-Van
Wely, Melilla, 2 1 . 1 1 . 2 011.
12 . . . i.e6 is more complex, but
White retains a pleasant game af
ter 13.ttixeS heS 14.f4. See game
11 Ionov Yevseev, St. Petersburg
-
2011.
D. 8 . . dxe4
This exchange is seldom seen at
this point, but in fact it is more logi
cal here than after 8 .. .1''l e8 9.0-0,
because the rook move would be
superfluous in the event of 9 . . . dxc4
10 .hc4.
The only way for White to exploit
the text is to organise a kingside at
tack leaving the king in the centre.
9.he4
9.bxc4 eS 10. 0-0 1'!e8 11.d4 (11.
ttigS tticS=) is an extensively tes
ted position from the Anti-Meran,
131
 Part 6
where the bishop is not at its best
place on b2. Cornette-Arutinian,
Heraklio 2 0 0 2 , saw 1O .g4? ! tLlxg4
1l.tLle4 i.c7 12 .E:g1 fs 13.tLlegs, when
instead of 13 . . . h6? 14.E:xg4 ! , Black
had 13 . . . tLldf6 14.h3 tLlxf2 ! 1s.'i!?xf2
h6.
9 . . . b5 10 . .te2
1O .i.d3 i.b7 1l.tLlgs h6 12.tLlge4
1J.e7 13.0-0 is a solid alternative.
White can try to extract some value
from his better minor pieces, e.g.
13 ... b4 14.tLlxf6+ i.xf6 1s.tLla4 E:c8
16.i.h7+ 'i!?h8 17.1J.e4.
10 . . ..tb7 11.g5 a6 12.h4!?
An amazing move, typical for the
original style of Mihail Suba. 12 .a4
would be the standard positional
option.
12 . . . c5 13 . .tf3 hf3 14.gxf3
Black's defence is not trivial, be
cause he has not any obvious ways
to generate counterplay.
132
The game Suba-Cosma, Bucha
rest 1996, continued 14 . . . i.eS 1s.f4
.ixc3 16.hc3 E:e8, when instead of
17. 0-0-0, White would have been
clearly better following 17.'i!?e2 ! es
18.E:ag1--+. The Reti bishop is X-ray
ing the kingside, both white rooks
are active.
E. 8 . . . b6 9.E:gl !
I think that the bayonet attack
is White's best weapon versus the
fianchetto set-up. In my database it
scores 8-1! If Black parries the first
wave of the attack, things might get
complicated for White due to his
king in the centre. However, the
second players have not shown a
good defensive method so far.
9.g4 is dubious: 9 . . . tLlxg4
In Chadaev-Potapov, Russia
2011 was 9 . . . ib7 1O .gs tLle8 ll.h4 es
12 .hs bs? 13.id3 e4 14.llJxe4 dxe4
1s.,be4 fs, when 16.gxf6 llJdxf6
17.h6 llJxe4 18.hxg7 wins.
 4.lLJc3 e6 5.b3 lLJbd7 6.'?9c2 d6 7 . .ib2 0-0 B . .ie2 ! ?
1OJ%g1 lLJde5 !
Practice has only seen 10 . . .
f5?! 11.h3 [Perhaps the best move
o rder is 11.cxd5 exd5 (11 ... cxd5
12 .lLJb5) 12 .h3] 11 . . . lLJge5 (or 11 ...
lLJgf6 12.lLJd4t '?geB 13.cxd5 cxd5
14. lLJcb5 .ie5 15.lLJc7 hc7 16.'?9xc7
gbB 17.'?9d6 gfl 1B.lLJxe6 and White
won in Andreikin-S.Kaplan, Plov
div 2 00B) 12 .lLJd4 lLJc5 [12 . . . lLJbB is
too passive. White has an attack af
ter 13.cxd5 exd5 14.0-0-0 f4 15.lLJf3
e7 (15 ... i.f5 16.e4) 16.exf4 gxf4
17. lLJxe5 he5 1B . .id3--+] 13.0-0-0
This position was reached in
Aronian-Gelfand, Leon 2010. White
should have opted for 13.cxd5 exd5
(13 . . . cxd5 14.lLJcb5 lLJg6 15.h4)
1 4.f4 '?9h4+ 15.'tt> d1 lLJg6 16.b4 lLJe6
17 .lLJxc6 .ib7 1B .b5 mhB 19.'?9b3,
with a messy game.
11.lLJxe5 lLJxe5
It is much safer to have the f
pawn at fl, instead of f5. Now it is
unclear how to lead the attack after
1 2 . 0-0-0 lLJg6 (12 . . . dxc4 13.lLJe4--+).
I have also analysed:
12.cxd5 cxd5 !
12 . . . exd5 gives more chances
in view of the pawn march in the
centre : 13.0-0-0 lLJg6 14.f4 geB (or
14 . . . b5 15.f5 lLJe5 16.d4 lLJd7 17.e4
b4 1B.lLJa4oo) 15.f5 lLJe5 16.d4 lLJd7
17.e4 dxe4 1B.d5, although it is only
a draw following 1B . . . lLJe5 19.dxc6
lLJd3+ 2 0.hd3 exd3 21.lLJd5 dxc2
2 2 J %xg7+ mfB 23J''! gB + , with a per
petual.
13.lLJb5 .ib7 14.lLJxd6 '?9xd6 15.f4
lLJg6 16.h4 gacB 17.'?9bl. Now Black
opens up his bishop and takes over
the initiative with 17 . . . d4! 1B.h5
lLJh4 19 .hd4 f6t.
I should also note that the the
matic exchange 9.cxd5 is not too
promising, because after 9 . . . cxd5
(9 . . . exd5 is risky. In Mikhalevski
Misiano, Biel 2010, White owned
the initiative: 1O.lLJd4 i.b7 11.lLJf5
.ibB 12 .d4 geB 13.g4 lLJfB 14.h4--+.)
10.lLJb5, Black has 1O ... i.c5, cover
ing the c-file. Then 1l.a3 a6 12.lLJbd4
!b7 13.b4 .id6 is balanced.
9 .ib7
This is the most natural deve
lopment, but we'll see that White
gets a strong attack in this line. I
admire the fine sense of danger of
Dreev who chose twice in 2 0 0 2 :
133
 Part 6

9 . . :e7 1O.g4 .b3 14.hc4 .ib7 (14 . . . .ia6 15Jg4 ! ? )

l S . .id3 g 6 16.h5 e7 (16 . . . c5
This exchange is positionally du
17 . .ie4 he4 18.xe4) 17.gg3 lLlg7
bious, because Black remains with
lS.hxg6 fxg6 19 ..ie4 gacS 2 0 . 0-0-0
permanently weak dark squares,
but it does take the sting out of
White's onslaught. White should
switch to strategic play with a slight
plus for him.
1l.gS lLleS I2 .ha3 xa3 13.h4

The open h-file and the sensitive

square f6 define White's edge.
It is indicative that no one has
followed in Dreev's shoes in later
games.

13 . . . dxc4 10.g4 e5
Dreev improves on his game
Practice has also seen:
against Bischoff in Hastings 2000,
which saw 13 ... lLld6 14 . .id3 g6 1S.h5 a) 1O ... a6? ! is too slow. Petkov
lLlfS 16.cxdS cxd5 when 17 . .ixfS ! Friedrich, Cesenatico 2010, went
exfS 1S.lLlbS would have been clear 1l.gS tDeS 12 .h4 eS 13 .hS lLlc7
ly better for White. It would be also 14 . .id3 ! and the battle is over:
interesting to exchange immediate
ly on dS : 14.cxdS cxdS IS.lLld4 .ib7
16.lLlcbS lLlxbS 17.lLlxbS e7 18.c7
.ia6 19.d6 xd6 2 0 .lLlxd6 lLlcS
21..ixa6 lLlxa6 2 2 . @e2;!;.
The text fixes a target on c4 and
keeps the centre open. In Zvja
ginsev-Dreev, Poikovsky 2 0 0 2 ,
White recaptured 14.bxc4 lLld6
IS.hS .ia6 16.lLle4 lLlxe4 17.xe4
and was only slightly better. The 14 . . . e4 (or 14 . . . g6 IS.hxg6
exchange of the knights relieved fxg6 16.hg6 gxf3 17.hh7+ @fS
Black's defence. I think that White IS .if5) Here best is IS.lLlxe4 dxe4
has the better prospects after: 16 . .ixe4+-.

134
 4.ttJc3 e6 S.b3 ttJbd7 6.1lNc2 .id6 7 . .ib2 0-0 8 . .ie2 ! ?
b ) 1 O . . . cS 1l.gS ! . Aimed against
1 1 . . .ttJe4 which would stumble into
1 2 .cxdS exdS 13.ttJxe4 dxe4 14.1lNc3 !
f6 lS.gxf6 E!xf6 16.ttJgS .ieS, Rot
stein-Schmitz, Lido Estensi 2 0 0 2 .
Here 17.d4 exd3 18.1lNxd3 .ixb2
19.E!dl+- is winning. Stefanova
Sebag, Ulaanbaatar 2010, saw 1l.h4
Eic8 ? ! 1 2 . 0-0-0 ttJe4? 13.ttJxe4 dxe4
14.ttJgS and Black was pawnless,
but 1l . . . ttJe4! would have taken over
the initiative: 12.ttJxe4 dxe4 13.ttJgS
.ieS 14 . .ixeS ttJxeS IS.ttJxe4 1lNxh4
16.0-0-0 fS.
1l ... ttJe8 12 .h4 1lNe7
12 . . . dxc4 should be met by
13 .ttJe4 ! cxb3 14.axb3 .ic7 IS.hS as
in the main line.
13.hS dxc4
White intends to join the second
rook in the attack after castling.
Black should do something quick
ly in order to escape this scenario
and it seems reasonable to acti
vate the bishop. 13 .. .fS only opens
files on the kingside : 14.gxf6 ttJdxf6
1S. 0-0-0;l;.
1 4.ttJe4 ! !
There is no time to bother with
pawns. 14.bxc4 f5 ! is unclear.
(but not 14 . . . .ic7 IS.ttJe4 eS 16.ttJh4
fS 17.ttJc3 e4 18 .h6 g6 19.f4)
14 . . . cxb3 lS.axb3
White has a strong initiative.
For instance :
IS . . . .ic7 16.ttJf6+ ttJdxf6 17.gxf6
ttJxf6 18.1lNc3 .id8 19.E!a4 ! or
IS .. .fS 16.gxf6 ttJdxf6 17.ttJxf6+
ttJxf6 18.h6 g6 19.E!a4 ! (against i.e4)
19 . . . .ic6 20.ttJh4 .ie8 2 l..id3 + - .
The g6-square will be busted soon .
H.gS ttJe8 12. 0 - 0 - 0 ttJc7
13.h4
I'm following the game Socko
Ejsmont, Ustron 2006, but 13.'it>bl,
aimed against ... bS, is the better
move order.
13 . ttJe6 (13 . . . bS ! ) 14.'it>bl 1lNe7
This position cannot be cracked
without a break in the centre. Al
though IS.hS is possible, I think
that the best approach is lS.d4 ! e4
16.ttJeS;I;.
White is clearly ahead with his
offensive.
13S
Pa rt 6
Complete Games
1 0 . D . G u revich-Vigorito
USA 1 998
1 .f3 f6 2.c4 c6 3.e3 d S 4.c3
e6 S.b3 bd7 6.c2 .id6 7 . .ib2
0-0 8 . .ie2 e7 9 . h 3 eS 1 0 .cxdS
xdS 1 1 . e4 .ic7 1 2 . g 3
In the opening, White has played
9.h3, which I do not recommend.
This move might prove superflu
ous and by all means it is less useful
than 9.0-0. I chose to comment this
game in an attempt to shed some
light on the Sicilian pawn struc
ture we see on the diagram. There
is very little practical experience in
it and both sides usually struggle to
find a clear plan.
This type of the Sicilian with col
ours reversed can arise with Yffe 7 or
e8. In both cases, it makes sense
to retreat the centralised knight to
g3. That not only hampers .. .5, but
it also threatens with d2-d4. In one
game White preferred 12 .g4, but
136
this move fits in with g1 and not
with the passive h3.
1 2 . . . b4 1 3.b1 cS
In similar positions, Black usu
ally sends this knight to f8-g6 (after
. . . e8). White's loss of tempo on h3,
however, allows Black to display ac
tivity first and organise . . . e5-e4.
1 4.a3 dS 1 S.c2 as
Black should not miss the op
portunity to push 15 . . . e4 ! . After
16.d4 f5 (in the event of 16 . . . hg3
17.fxg3 Yffd 6 18.0-0 hh3 19.9xh3
Yffxg3 + Black has only a draw.)
17.b4 d3+ (17 .. ,4 18.bxc5 fxg3
19.f3;1;) 18 .hd3 exd3 19.Yffxd3 f4
White is on the defensive. His
king will never be safe on the
kingside because of the lever on
h3. e.g. 2 0 .ttJge2 fxe3 2 1 . dxe3 .id7
2 2 .0-0 ae8.
 4.liJc3 e6 S.b3 liJbd7 6.'<!;Vc2 d6 7 . .ib2 0-0 B . .ie2 ! ?

1 S J:c1 b S 1 7 .0-0 24.b4 axb4 2S.axb4 tLld7 2S.f4

White provocatively allows . . . e4,
but apparently Vigorito trusted him
without calculation. It was interest
ing to try 17.c4. Then 17 . . . e4 would
simply drop a pawn to 18.tLlxe4 and
17 . . JeB IB .d3 would take control
over e4 while leaving the bishop
active : IB . . . tLla6 19.0-0 bS 20 . .ixdS
cxdS 2 1.'<!;Vc6 E1bB 2 2 .'<!;VxdS b7
23.'<!;VxbS;!;.
1 7 . . ..id7
17 ... e4 ! IB .tLld4 '<!;Vd6 19.f4 '<!;Vg6
2 0 . lt>h2 .id7 would have been ba
lanced. Now White gets some ini
tiative.
1 S . .ic4! :gacS
gS 27 .fxgS ftlxeS 2S . .ixeS fxeS
29.:ge1 ftld3 30 .:gxeS ftlxb4 3 1 .:gc4
ftldS 32.:gexcS :gxcS 33.:gxcS :gdS
The golden endgame rule to put
the rook behind the passer should
have saved Black: 33 . . . E1bB 34.E1d6
tLle7 3S.E1d7 It>fB 36.tLle2 bS 37.tLld4
b4 3B.tLlb3;!;.
34.ftlfS <tIf7 3S.:ghS <tIgS
3S.ftldS ftlc3 37. ftlfS ftldS 3S.:gcS
<tIf7 39.ftl dS+ <tIe7 40 . ftl b S <tIf7
41 .<tIf2 <tIg7 42.<tIf3 :geS 43 . ftl d S
:ge3+ 44.<tIg4 :gb3 4S.ftlfS+ <tIf7
4S.<tIhS :gc3 47.:ghS <tIgS 4S.:gdS
ftlf4+ 49 .<tIhS ftlxg2 SO.h4 :gcS
S1 .:gdS+ f7 S2.:gd7+ <tieS S3.:gd2
:gxfS S4.:gxg2 :gf7 SS.:gb2 :gb7
SS.:gbS <tIdS S7.hS <tics SS.:gfS bS
S9.gS hxgS SO.hxgS 1 -0

1 1 . lonov-Yevseev
St. Petersburg 2 0 1 1

1 .c4 cS 2.ftlf3 dS 3.e3 ftlfS

4.ftlc3 eS S.b3 ftl bd7 S.Wc2
.id 6 7 . .ib2 a6 S.d4 0-0 9 . .ie2
eS 1 0 .cxdS cxdS 1 1 .dxeS ftlxeS
1 9 .ftlxeS 1 2 .0-0 .ie6

White also have the more active

piec es following 19.e4 tLlf4 2 0 .d4
exd4 2 1..ixd4 tLlce6 2 2 .tLlfS lMleB
2 3 . .ie3 bS 24.id3;!;. The text simpli
fies to a better ending which Black
should be able to hold.

1 9 . . . .ixeS 20 . .ixeS WxeS 2 1 .d4

ftl xe3 22.dxeS ftlxc2 23.:gxc2 .ieS

137
 Part 6

1 3. c!Ll xe5 ie6? ! 18.lLla4 ixb2 19.1Llxb2 a5

2 0 .lLla4 b4 2 1.d4;!;, but the bru
I think that mundane continu tal 17 . . . ixc3 would have left White
ations do not promise White any with a nice, but dead draw position
advantage : with opposite coloured bishops.
a) 13 J'ad1 c8 (13 . . . lLlxf3 + 17 ... ib8 18.xc8 ixc8 19.id4
14.ixf3 Yff c 7 15.h3 ih2 + 16.'i!;>h1 e8 = , Portisch-Kasparov, Dubai
ie5 17.d2 ac8 18.lLle2 ixb2 1986. White stands nicely, but he is
19.Yffx b2 c2 2 0 .Yff a 1 is slightly bet not threatening anything really.
ter for White.) 14.b1 a5 15.a1
fd8 16.lLld4 id7. White's pieces 1 3 . . . ixe5 1 4.f4 ib8 1 5.if3
stand solidly, but they are hamper
ing each other. Perhaps crucial for the assess
ment of this variation is 15.ad1
b) 13.ac1 c8 14.b1 lLlfg4
e8 (The e3-pawn is irrelevant.
15.lLlxe5 !xeS!
Black gets crushed after 15 . . . ia7
15 ... lLlxe5 16.cd1 h4 cost Black 16.f5 ic8 17.'i!;>h1 ixe3? 18.lLlxd5
a pawn in Halkias-Dominguez, li lLlxd5 19.ic4.) 16.'i!;>hU
nares 2 0 0 2 : 17.f4 (17.g3) 17 . . . lLlg4
18.ixg4 xg4 19.h3 h4 2 0 .f5 id7
2 1 .lLlxd5.
16.ixg4 (The game Kovalyov
Haslinger, Benidorm 2008, finished
in a draw after 16.h3 ih2 + 17.'i!;>h1
ib8 18.ixg4 ixg4 19.f4 owing to
the perpetual check following 19 . . .
h4 2 0 .lLlxd5 ixh3.) 1 6 . . . ixg4

White has the more active pieces.

He is intending if3 with pressure
on d5. Ionov speeds up play, but
eventually all the complications
should chill down to a rook end
game with a tiny edge for White.

1 5 . . .tyb6!

17.lLla4 The queen hurries to help its

king via e3. Or 15 . . . ia7 16.f5 ic8
White should preserve his 17.'i!;>hU.
knight. Wang Rui-Chandler, Kuala
Lumpur 2007, saw instead 17.d3 1 6.i>h 1

138
 4.lLlc3 e6 5.b3 lLlbd7 6.c2 .id6 7 . .tb2 0-0 S . .te2 ! ?

The computer suggests a n ex

tremely weird variation which
cannot attract any human, I'm
sure, because of the multiple pins :
16.f2 !a7 17.l::1 ael d4 lS.lLldl dxe3
19.1Llxe3 .id7 2 0 JWg3 !b5 2 Uf2
l"1feS 2 2 .i.e5 l"1adS 23.@hl g6oo.

1 6 .. .'ilYxe3 1 7 .fS .id7 1 8.ttlxdS

ttlxdS 19 . .ixd S YlYh6 2 0 . g 3 ! .ixg 3
29.l"1af1 l"1d6 30.l"1f3 h6 3l.a4 as
32.l"1fS cS 33.l"1xg7+ @hS 34.l"1xcS
l"1xf6 3S.l"1g3 l"1aS 36. @g2.

21 J11 g 1 .ixfS 22 .YlYg2 Ud8

23 . .ixb7 gab8 24.if3 .if4 2S.ixg7
YlYg6

The first critical moment o f the

game.
I think that White should play
2l.g2 !h4 2 2 .!c1 ! ?
Instead, White can "win" the
queen in a much better version
than in the game (because the h-file
is closed), but I doubt that he could
aspire to a win after 2 2 .l:U3 !f6
23.l"1h3 (23.l"1g1 .bb2 24.xb2 @hS
25.l"1h3 f6=) 23 . . .xh3 24.xh3
ixb2oo. His king is too vulnerable
under the fire of the long-range
black pieces.
22 . . . f6 23.l"1g1 !c6 24 . .bc6
xc6 2S.xc6 bxc6 2 6.!h6 !f6
27 .ixg7 !Kg7 2 S.f6 l"1adS
White can squeeze out a small
advantage, using the fact that
Bl ack's fS-rook is tied with the de
fe nce of the f7-pawn :
26.YlYf1 ? !
As they say, White does not
sense the danger. 2 6.!c3 xg2+
27.l"1xg2 + i.g6 2S .i.e2 @fS 29.l"1f1
!d2 would have been a draw end
ing. He is greedy to snatch the
queen, probably in order to feel the
moral winner of the opening battle,
but that is not a pragmatic decision.
The rest of the game is not too
interesting for us. White's defence
is not trivial and his blunder on
move 31 could be expected.
26 ... cit?xg7 27.gxg6+ hxg6

139
 Part 6
28 .Wc4 .ieS 29 J e 1 .if6 30 .Wxa6
d2 3 1 .e2? d 1 + 32.<it?g2 h8
33.Wb7 .id4 34.g3 g S 0-1
1 2 . Delchev-Brunello
Porto C arras E u . team c h . 201 1
1 .f3 dS 2.c4 c6 3.e3 f6
4 . c3 e6 S.Wc2 bd7 6.b3 .id6
7 . .ib2 a6
Perhaps it is better to conceal
this plan until White has devel
oped the bishop to e2, for example,
7 . . . 0-0 8.ie2 a6.
8.d4
8J01g1 ! ? is much sharper, but I
did not want to unnecessarily stir
the emotions of my team mates so
early in the match.
8 . . . b S 9 . .ie2
With this move order, I could
have found a more active place
for my bishop: 9.c5 ! ic7 1O .id3 !
'f!e7 11.e2 ! 0-0 1 2 . 0-0 e5 13.dxe5
xe5 14.he5 he5 15.xe5 'f!xe5
16.d4;t.
9 . . . .ib7
140
1 O.cS .ic7 1 1 .b4 0-0 1 2 .0-0 eS
Black has nothing to gain from
waiting:
12 . . Je8 13J01ad1 e5 14.e4 ! ;
12 . . . 'f!e7 13JUe1! e5
Or 13 . . . a5 14.a3 axb4 15.axb4
l:!xa1 16.ha1 e5 17.dxe5 xe5
18.d4 c4 (18 . . . g6 19.e4 dxe4
2 0 .xe4 d5 21.'f!b2) 19.e4;t.
14.dxe5
14.e4 is not too clear because of
14 . . . dxe4 15.xe4 xe4 16.'f!xe4 f5 !
17.'f!h4 (or 17.'f!e3 17 . . . e4 18 . .id1
f6) 17 . . .'f!xh4 18.xh4 .id8 .
1 4 . . . xe5 15.e4 l:!fe8 16.d4
dxe4 17.xe4 xe4 18.'f!xe4 as
19 . .id3 g6 2 0 .'f!xe7 l:!xe7 21.f5;t.
1 3.xeS xeS 1 4.dxeS .ixeS
1 S.f4 .ic7
1 6.e4
This move seemed very strong
to me during the game, but the
postmortem showed that, albeit
clearly worse, Black has chances to
hold his position. Perhaps I should
have kept the tension by 16.a4 !
(or 16.d1! 'f!e7 17.a4 ! ) 16 . . . 'f!e7
17.d1! (This move never crossed
my mind.) 17 . . . e4 18 . .id4 l:!fe8
 4.lLlc3 e6 S.b3 lLlbd7 6.Wc2 d6 7.b2 0-0 8 .e2 ! ?
( l 8 . . .a S 19.1Llc3 ! lLlxc3 20.Wxc3
axb4 2 1.Wxb4 .taS 2 2 .Wb2 bxa4
2 3.hg7 Wxe3+ 24.'.thl) 19.1Llc3:;!;.
I n this position all my long range
pieces are active and I could set up
a kingside attack.
1 6 . . . dxe4 1 7.tLlxe4 tLl xe4
Of course, Black should seek ex
changes. After 17 . . . lLldS? ! 18.E:ad1 !
the queen has not a comfortable
square: 18 . . . We7 19 . .td3 lLlxb4
(19 . . . hf4 20.E:de1 Wc7 2 1.g3 .th6
2 2 .lLld6 f6 23.!xh7+ '.th8 24.VNg6)
2 0 .Wc3 fS 2 1 .tlJd6 (White is also
better after 21.tlJg3 tlJdS 2 2 . E:de1
t7 23.Wd4 .tc8 24.hfS) 2 1 . . .tlJdS
2 2 .E:de1 VNd7 23.Wd4 hd6 24.cxd6
l:l:f6 2S.E:eS. White has a tremen
dous initiative, for example: 2S . . .
xd6 26 . .txfS E:af8 27 . .te6+ '.th8
28.fS .tc8 29.hdS cxdS 30.E:f3 Wc7
31..ta3
1 8 .Wxe4 ge8 1 9.Wc2
19.Wf3 VNd2 20 . .teS !xeS 2 1.fxeS
d4+ 2 2 . '.th1 WdS 23.E:ad1 Wxf3
2 4.hf3 E:e7 2S.E:d6 g6 is level.
A critical moment of the game.
Black denies me the c3-square, but
he misses the opportunity to trade
a pair of rooks by 19 . . . aS ! 20 . .td3
Wh4 2 1.g3 WhS 2 2 . a3 axb4 23.axb4
E:xa1 24.E:xa1 .tc8:;!;.
White certainly stands nicely,
but in the game I failed to break
through in a similar position.
Forced variations lead to only a
small edge after:
a) 2S ..te4 We6 2 6.hh7+ '.th8
27.Wc3 .teS 28.VNxeS WxeS 2 9 .!xeS
'.txh7:t. White has a clear extra
pawn indeed, but in positions with
opposite coloured bishops it is
more important to have a strategic
rather than a material advantage.
b) 2S . .tfS '.tf8 ! (after 2S . . . .txfS
26.WxfS Wg6 27.VNxg6 hxg6 28.E:a6
E:e6 29.E:a8+ '.th7 30.E:c8, Black
is close to a zugzwang : 30 . . . E:e2
31..tc3 E:e7 32.'.tf2 f6 33 . .td2 E:d7
34 . .te3 E:e7 3S.'.tf3 E:d7 36.E:e8.)
2 6.hc8 E:xc8 27.VNfS E:d8 2 8 . .tc3
Wg6 2 9.Wxg6 hxg6 30.E:a6 E:d3
31..ta1 l:l:b3 32.E:xc6 E:b1 + = .
I t may sound incredible, but
I was already short on time and
141
 Part 6
23.h2 e2 24Jb1 l3e8+) 20 . . .
tt:lxd5
Now 21.l3xd5 xa2 2 2 .hg7
does not win in view of 22 . . . l3g3 ! ' it
is easy to miss such a move! 23.l3f2
Wxb3. White has sufficient com
pensation for the pawn, but I could
not find anything decisive.
Therefore, in the diagram posi
tion it is probably better to take a
small edge in the endgame after
21.Wxd5 Wxd5 2 2 . l3xd5 !i.c7 23.l3d7
i.b6 24.h2 l3e2 25.l3xb7 i.c5
2 6.i.e5 l3xa2 27.l3d1 !i.f8 28.!i.b8;!;.
20Jxd5! b5!
I was understanding that I
should simply keep the extra pawn
with 21.l3e5 b6 (21.. .l3ed8 2 2 .tt:le4
tt:lxe4 23.Wxe4 d2 24.!i.d4 Wxa2
25.l3h5-+) 2 2 .h2, but the tempta-
144
tion to face the opponent with con
crete problems in his time trouble
was too strong to resist.
21 . . . tt:lxd5?
So he lived up to my expecta
tions and erred! The arrogant rook
is a challenging sight indeed, but
2 1 . . .l3e6 22 .l3g3 l3ae8 23.l3e5 ! (23.
l3d3 g6 24.h2 l3c8 25.tt:ld5 tt:le4oo)
23 . . . g6 24.tt:ld5 We1+ 25.h2
tt:lxd5 (25 . . . Wxg3+ 26.xg3 gxf5
27.tt:lxf6+) 26.l3xe6 l3xe6 (26 . . . fxe6
27.l3xg6+) 27.xd5 l3xe3= or 21 . . .
i.a3 2 2 .ha3 tt:lxd5 23.tt:lxd5 Wxa3
24.l3g3 g6 25.tt:lf6+ h8 26.Wg5
g7= were keeping the balance.
A horrible move after which my
advantage slips away.
23.l3g3 ! should be winning.
Black has two defences:
a) 23 . . . l3d8 24.h2 ! l3dd6 (24 . . .
l3xd5 25.xd5 Wxa2 2 6.!i.xg7 ! +
l3g6 27.!i.d4) 25.tt:lf6+ l3xf6 26.i.xf6
ge6 (26 . . . Wxa2? 27.gxg7+ !i.xg7
28.c8+) 27.Wg5 Wxa2 28.f5 gxf6
29.Wxf6 Wxb3 30.Wxa6+-.
b) 23 . . . gg6 24J"!xg6 hxg6
 4.llJc3 e6 S .b3 llJbd7 6.'\1;Ifc2 .id6 7 . .ib2 0-0 B . .ie2 ! ?

25.llJf6+ ! ! @hB 2 6.'\1;Ifd5 ! .

I missed this double attack on
th e rook and the f7-pawn. White has
full domination : 26 . . .:!J:a7 27 . .id4 ! !
'Bc7 (27 . . . gxf6 2B . .ixa7 @g7 2 9 . a4
is a pawn up) 2B.llJd7 @gB 29.llJxfB
'Bc1+ 30.@h2 @xfB 31.'\1;IfaB+ @e7
32 ..ib2 (32 . .ixg7? ! Yffc 7= 33.'\1;Ifxa6
Wc6) 32 . . .l'k7 (32 . . .l:!d1 33 ..ixg7
Wc7 34.'\1;Ifxa6) 33 . .ixg7, with very
good chances to win. 28 .fS ge8 29 ..1d4 'ilYc7 30.gf1
23 .. J::l xf6 24 . .1xf6 ge8 2S . .1eS gd8 3 1 .'ilYe4 gc8 32 .'ilYg4 'ilYd7
(2S ..id4 '\1;Ifxa2 2 6.'\1;Ifd7=) 2S . . . 'ilYxa2 33.'ilYg3? (33 .h4 ! @hB 34.hS f6
26.Yff d 7 ge6 27.'ilYdS 'ilYc2 ! 3S.h6 ! t) 33 . . . g;, h8 34. h4 f6 3 S . h S
h6 36.'ilYf3 as 37.gd 1 'ilYf7 38.e4
Here my opponent offered me a ge8 ! 39 . .1b2 a4 40. bxa4 bxa4
draw which I should have accepted. 41 .gdS 'ilYb7 42 . .1c3 'ilYb1 + 43.gd 1
The position is objectively level and .1cS+ 44.g;,h2? 'ilYxe4 4S.'ilYxe4
we both were in time trouble. gxe4 46.gd8+ g;, h7 0-1

145
146
 Part 7

Anti-Queen's G a m bit I
1 .f3 d 5 2 . c4 e6 3 . 9 3

147
Pa rt 7

Main Ideas

In Part 7 and 8 I examine the king's White takes over the initiative:
fianchetto against 2 . . . e6: a) 5 . . . b5 6.d3 ! ? cxd3 7.ttle5 ib7
8.a4 b4.
1.<:Jf3 d5 2.c4 e6 3.g3

9.a5 ! ? (threatening a6 ! ) 9 . . . ttle7

White can also play 3.d4 if he 1O.ttlxd3 ttla6 11.\!;lIa4 ttld5 12 J'd1
wanted to avoid the Slav and the \!;lIc8 13.e4, Pantsulaia-Bareev, Beer
Chebanenko, but would be happy sheba 2005.
with the possible Meran after 3 . . .
b) 5 ... ttlf6 6.a4 ttla6 7.ttla3 ha3
c 6 . Another option i s 3.e3 ttlf6 4.b3
8 .bxa3 ttlc5 9Jbl.
c5 which is the Semi-Tarrasch.
After 3 .g3, Black occasionally
The most challenging approach
tries to play in the spirit of the Tri
of Black is:
angle System:
3 . . . c6 4.ig2 (4.b3 !?) dxc4 5.0-0 3 dxc4 4.%7a4+

148
 l.llJf3 dS 2.c4 e6 3.g3
A. 4 . . . i.d7; B. 4 . . . c6; C. 4 . . . llJd7.
A. 4 ... ,td7 5.Yfxc4 c5 6.ftle5 ! ?
White gains the bishop pair ad
vantage in an open position with a
flexible pawn structure. The game
Aronian-Kramnik, Khanty Mansi
ysk 2010, went:
6 . . . lik6 7.ltJxd7 '!Wxd7 8 .1g2 ltJf6
9 . 0-0 i.e7 1O .'!Wa4 ltJd4 11.'!Wxd7+
'it>xd7 12 .ltJc3 gab8 13.gdl ghd8
14.b3 'it>e8 IS.i.b2 ltJdS
White should seek asymmetric
positions so 16.ltJxdS exdS 17.e3
liJe6 18.gael bS 19.4 looks consis
tent.
B . 4 ... c6 5.Yfxc4 b5 6.Yfc2
tb7 7.,tg2
White's main plan here is to put
pressure on the b- and c-pawns.
Here is an example of his stra
tegy: 7 . . . cS 8.a4 a6 9.axbS axbS
1O.gxa8 has 11.0-0 ltJf6 12 .Yfb3 !
b4 (12 . . . i.c6 13.ltJeS) 13.d3;!;.
White will later restrict the ene
my's pieces with e2-e4 as in Game
14 Radjabov-Negi, Khanty-Man
siysk 2011.
7 . . . ltJf6 8. 0 - 0 ltJbd7 9.ltJc3
a6!
9 ... !e7 offers White an addi
tional options as 1O.d4 ! ?
10 .a4! ,te7 11.d3! 0 - 0 12.e4
White's set-up is universal and
does not depend much on Black's
move order.
149
 Part 7

I propose here the retreat

His plan is to develop his bish
12 .ttJdl ! ? ttJc5 13.c2;!;, when 13 . . .
op on e3, safeguard it with h3, and
.idS 14.ttJe3 .ib3 15.bl;!; keeps the
bring the rook to dl. Then he will be
material balance.
well prepared for further advancing
in the centre or on the kingside.
Black's only counterplay is con
nected with . . . cS-c4.

12 gc8 13.h3 ! ?

A simple, but effective novelty.

Black is lacking a good place for his
queen. 13 . . .'c7 faces 14 ..if4 e5? !
15.ttJxe5 ttJxe5 16.d4 .id6 17.dxe5
.ixe5 18 . .ixe5 xe5 19.f4 c5+
2 0 . f2;!; while on b6 the queen will
also be under the bishop's X-ray:
13 . . . b6 14 . .ie3 c5 15.axb5 axb5
16.b4 ! .
c. 4 . ttJd7 5.i.g2 ! f6
Black often tries to develop his
bishop before White has put .ig2 :
5 . . . a6 6.xc4 b5.
However, this gives White ob
j ects to attack on the queenside.
Svidler-Kramnik, Moscow 2011,
The plot of the next few moves
is turning around the fate of Black's
light-squared bishop. It is not easy
to fianchetto it because of the threat
of ttJd4. Black can try . . . a6, b5 or
. . . b6. We know that the first set-up
has the drawback of offering targets
on b5 and c5. In practice, Black pre
fers the more restrained . . . b6. Then
we should open the centre by d4.
Let us consider a few typical lines.
a) 6 a6 7.c2! c5 8.c3 tfc7
.

went:
Or 8 .. .l:!b8 9.d4 .id6 1O.dxc5
7.b3 .ib7 8 . 0-0 gf6 9.ttJc3 ttJxc5 11.0-0; 8 . . . ga7 9.d4 b6
.ie7 1O .d3 0-0 11.a4 b4 10.0-0 .ib7 11.e4 ! ? (a natural move)

150
 1.tt'lf3 dS 2 .c4 e6 3 .g3

1l cxd4 1 2 .tt'lxd4 W1a8 13.f3;1;, fol

. . . 1971. His plan is to increase it by
lowed by l:!fdl. pushing eS or b2-b4-bS.

9. 0 - 0 ! b6 10 .d4 .tb7 1l .tf4

b) 6 c5 7. 0 - 0 b6 8 .d4 .tb7

id6 12.hd6 ti'xd6 13.!Udl 0 - 0 9.:gdl cxd4 10 .ig5 .ie7 1l. xd4

1 4.e4 cxd4 15.xd4;!; hg2 12.'it1xg2

White had a space advantage The c6-square and the pawn on

in Kortschnoj-F.Olafsson, Moscow a7 are weak.

lSI
Pa rt 7
Step by Step
1.f3 d5 2.c4 e6 3.g3
3.e3 is less effective than it was
against 3 . . . c6, because Black can
play . . . c5 in one step, thus getting a
good version of the Semi-Tarrasch.
I do not have fresh ideas for White
after 3 . . . f6 4.b3 c5 5.!b2 c6
6.cxd5 exd5 7.c3 .id6 S . .ie2 0-0
9. 0-0 geS 1O.d4
152
1O ... cxd4! 11.xd4 ttJxd4 12 .xd4
.ie5 13.d2 .ig4 14.,ixg4 ttJxg4
15.h3 f6 16.gfdl gcS 17.gac1 gc6 = .
The attempt t o get without b3
is countered the same way - Black
drags the white knight to d4: 4.c3
c5 5.cxd5 exd5 6.d4 ttJc6 7 . .ie2 cxd4
S.xd4 .id6= .
3.b3 i s inaccurate because Black
obtains additional possibilities in
comparison to the main line. For
instance, after 3 . . . c5 4.!b2 ttJc6
5.g3, he can play 5 . . . d4 6 . .ig2 e5
7.d3 .id6.
The king's fianchetto is more
enterprising. It may lead to Cata
lan set-ups, where the delay of d4
restricts Black's options. In fact,
White often does not play d4 at all.
The Catalan has been extremely
popular lately and you can find tons
of publications on it. In this book, I
will examine only lines which are in
the spirit of the Reti.
A. 3 . . . c6; B. 3 . . . g6; C. 3 . . . dxc4.
3 . . . f6 4.ig2 .ie7 5.0-0 0- 0
6.b3 is the subject of Part S.
3 . . . c5 4.cxd5 exd5 5.d4 leads
to the Tarrasch Defence although
 1.ttJf3 d5 2.c4 e6 3.g3

White can also follow the scheme Now Marin's method does not
with b3. work: 1O . .ig5 f6 1l . .id2 .ie6 ! when
12.1'xb7? b8 13 .Wla6 .ia5 14.ttJb5
A. 3 . c6 4 .tg2
.
.ic8 favours Black.

Al. 4 . . . dxc4 5. 0 - 0

5.a4 is also possible, but then

White should reckon with the vari
ation 5 . . . ttJa6 6.ttJa3 Wld5 ! ? , e.g. 7.e4
Wle6 8 .Wle2 ha3.

5 ttJf6

5 . . . b5 is consistent, but is it
worth to be suffering the next 20-
30 moves for only a pawn ? ! Apart
A1. 4 . . . dxc4; A2 . 4 . . . ttJf6. from the thematic 6.a4, White has
6.d3 ! ? cxd3 7.ttJe5 .ib7 8.a4 b4.
4 ... .id6 should be met by 5.d4 !
for two reasons :
1. If White is afraid of the pas
sive, old-fashioned Stonewall that
arises after 5 .. .f5, he should not
play close openings at all.
2 . He has not a better alterna
tive, because 5.0-0 concedes the
centre to 5 . . . e5. Marin advocates
6.cxd5 cxd5 7.ttJc3 ttJe7 8 .b3 .ic7 White has two good options
9.d3, but now he considers only
here:
9 . . . 0-0 10 . .ig5 ttJbc6. Instead, 9 . . .
a) 9.ttJxd3 ttJf6 10.ttJd2 !e7
tUbc6 ! i s fine for Black: 1l.ttJc4 0-0 12 .Wlc2 .
b) 9.a5 ! ? (threatening a6 ! ) 9 . . .
ttJe7 10.ttJxd3 ttJa6 1l.Wla4 ttJd5
12 .dl Wlc8 13.e4, Pantsulaia-Ba
reev, Beersheba 2 005.
In both variations White has an
excellent game.

153
 Part 7
6.a4 a6 7. a3 ha3 8.bxa3
c5 9.gbl
White will soon regain the pawn.
In practice, Black fails to neutralise
successfully his initiative :
9 . . . 0-0 1O.'I1*lc2 lLldS 1l.'I1*lxc4 'I1*laS
12 .d4 lLlxa4 13 .id2 lLldb6 14.'I1*ld3
'I1*lhS 1S.'I1*lc2, Stocek-S.Movsesian,
Plzen 2 0 0 1 ;
9 . . . 'I1*laS 10J':l:b4 c 3 1l.dxc3 0-0
12 .'I1*lc2 lLldS 13.e4 lLle7 14.eS 'I1*lc7
1S.h4 lLldS 16.E!e1, Hillarp Pers
son-Dreev, Helsingor 2009.
A2 . 4 ... f6 5. 0 - 0
My intention is to play d4 and
reach the Closed Catalan. Still, S.b3
ie7 6.0-0 0-0 is popular and exten
sively tested in practice. One pos
sible plan is 7.ib2, d3, lLlbd2 , e4,
with complex play. Here is one ex
ample : 7.ib2 lLlbd7 8 .d3 bS 9.lLlbd2
as 10.a3 ib7 1l.'I1*lc2 a4 12 .b4 bxc4
13.dxc4 cS 14.cxdS hdS 1S.bS;!;,
Macieja-Gelashvili, Kallithea 2008.
154
5 . . . bd7 6.d4
6 . . . .id6
After 6 . . . dxc4 7.a4 White re
gains the pawn in a favourable po
sition : 7 . . . ie7 (or 7 . . . aS 8.'I1*lc2 lLlb6
9.lLlbd2 lLlfdS 10.lLlxc4 lLlb4 1l.'I1*lb3
lLlxc4 12 .'I1*lxc4 ie7) 8.lLla3.
6 ... ie7 leads to the main line
of the Closed Catalan. White has
the initiative following 7.'I1*lc2 0-0
8.lLlbd2 b6 9.e4 or 7.b3 0-0 8.lLlc3
b6 9.lLld2.
White breaks through the centre
with e4 and obtains a spatial advan
tage. Practice has seen further:
a) 8 . . . E!e8 9.e4 dxe4 10.lLldxe4
lLlxe4 1l.lLlxe4 ie7 12.if4 lLlf6
13 .cS;!;, Ivanisevic-Ascic, Zupanja
2007.
b) 8 . . .ic7 9.e4 dxc4 1O.lLlxc4 eS
1l.dS lLlb6 12 .b3 lLlxc4 13.bxc4 id6
14.E!bU, Miroshnichenko-Gavrilov,
Moscow 2008.
 l.liJf3 dS 2 .c4 e6 3.g3
c) 8 . . . i.b4 9.a3 (9 .e4 eS is un
c lear , but 9.'<!9b3 as lOJ':idl, Lautier
B ologan, Enghien les Bains 2001, is
a fair alternative.) 9 . . . i.xc3 1O.bxc3
b6 11.cxdS cxdS 12 .c4 i.b7 13.a4;!;,
L. Filip-Olarasu, Albena 2011.
B. 3 g6 4 .ig2 .ig7 5. 0 - 0
liJe7 6.d4 lObc6
Marin considers in his The Eng
lish Opening, volume 2 only 7.cxdS
exdS 8.lDc3, but I see no reason to
open the black light-squared bishop.
7 0 - 0 8.e3 !
.
We'll exchange on d S only after
Bl ack weakens his queenside:
8 . . . b6
8 . . . dxc4 9 .'<!9a4 eS favours White
after 1O.lDxeS liJxeS 11.dxeS c6 (11 . . .
,beS 1 2 J':idU) 12.f4 ! . His pawn
clu ster in the centre is much more
dangerous than the opponent's
pawn majority on the queenside.
The game Pridorozhni-Bezgodov,
Khanty Mansiysk 2009, went 12 . . .
i.e6 13.e4 '<!9b6+ 14.hl 1':iad8
IS.liJdl f5 (or IS . . . 1':id3 16.liJf2 1':id7
17.g4) 16 . .ie3.
8 ... aS does not make much sense
either in view of 9.b3 a4 10 .i.a3;!;.
9.cxd5 ! exd5 10.a3 a5 1l.b3
White has a clear plan on the
queenside, but his secret and most
effective weapon is the break
through e3-e4. Then Black will re
gret his fifth move. Here are two
practical examples :
Dorfman-Vaisser, Moscow 1989:
1l . . . h6 12 .lDel i.e6 13.lDd3 gS
14 . .ib2 '<!9d7 1S.1':ic1 1':iac8 16.1':iel lDg6
17.b4 axb4 18.axb4 lDce7 19.bS.
155
 Part 7

The d5-pawn is weak, Black's Cl. 4 ... i.d7; C2. 4 ... c6; C3 . 4 . . .
pieces lack coordination ; ltJd7.

Bischoff-Khenkin, Recklinghau
sen 1996: 1l . . . i.a6 12J%e1 d6 Ct. 4 . . ..id7 5.1Yxc4
13 .i.b2 h6 14.c2 E:acB 15.e4 i.b7
16.E:adl.

White is ruling in the centre.

Cll. 5 . . . .ic6 ; C12. 5 . . . c5.

C. 3 . . . dxc4 4.1Ya4 + !
Cll. 5 . . . .ic6 6 . .ig2
Marin advocates 5.ltJa3 in
his repertoire trilogy, but I can
not agree with him. After 5 . . . ,ha3
6.bxa3, White practically plays
without a queenside. Therefore, he
should rely only on a kingside at
tack to win the game. Accordingly,
if Black is not superambitious, he
can quickly finish development and
castle. That would assure him of a
comfortable equality. For exam
ple, the game Dominguez-Meier,
Lubbock 2011, went on with 6 . . . b5
7.ltJe5 ltJd5 B.d3 (Marin's recipe)
B . . . cxd3 9.xd3 0-0 10.xb5 ia6
11.1Yb2 (or 11.b3 c5) 1l . . . c5 1 2 . 0-0 Black tries occasionally 6 . . . .id5
ltJ d7 13.ltJxd7 xd7 14.c2 l'!acB= . when 7.1Ya4+ d7 8.dl! ltJc6
Practical results confirm my opin 9.ltJc3 ltJf6 10.0-0 !J.e7 ll.ltJxd5 exd5
ion as 4.a4+ scores much better 12 .d3 0-0 13.id2 l'!feB 14.1Ya4 i.d6
than 4.ltJa3 . 15.l'!fel ltJe5 16.xd7 ltJexd7 17.b4

156
 1.tt'lf3 d5 2.c4 e6 3.g3
is slightly better for White. I think
that 7.'1Wc2 tt'lc6 B.'a4 ! or 7.'d3 !
may be even better, for example :
7 . . tt'ld7 (7 . . . tt'lc6 B.tt'lc3 tt'lf6 9.0-0
. 13.tt'lxc6 bxc6 14.d3 - the d3 pawn
ie7 1O.tt'lxd5 exd5 11.b3 bB
1 2 .d3 d7 13 . .td2 0-0 14.acl;!;)
B .O-O tt'lgf6 9.tt'lc3 .tc6 1O .c2;!;.
White will seize the centre with his
pawns while Black's counterplay
with . . . cS would be hindered by the
bishop on c6.
7. 0 - 0 bd7 8.c3 J.e7 9.e4!
White restricts the c6-bishop
and vacates the e2-square for the
queen. Then he will be able to oc
cupy the centre by d2-d4.
9 0-0
9 . . . tt'lcS stumbles into 10.d4!
ll'l cxe4 1l.tt'le5 .tdS 12.tt'lxdS exdS
13.bS+ .
1 0 .gdl!
10 .e2 is a mistake because
White loses control over the d4-
squ are after 1O . . . tt'lc5 1l.tt'leS d4 ! .
The rook move is aimed against this
as you can see from the variation
1O . . . bB 11.e2 tt'lc5 12.tt'le5 d4
is protected and Black is unable to
prevent d3-d4, e.g. 14 . . . eS 15 . .te3
b4 16.d4 exd4 17.hd4 fdB
1B.a3. That's why in Schmidt
Gdanski, Gdansk 1994, Black chose
10 . . . bB 1l.e2 b5, but now 12 .e5 !
tt'ld5 13.tt'ld4 .taB (13 . . . tt'lxc3 14.dxc3
hg2 15.'k!?xg2) 14.tt'ldxb5 gave
White a clear edge.
C12. S cS 6.J.g2
6.tt'leS ! ? is more simple and it
may be the better choice. The game
Aronian-Kramnik, Khanty Man
siysk 2010, went 6 . . . c6 7.tt'lxd7
xd7 B . .tg2 tt'lf6 9. 0-0 ie7 1O.a4
tt'ld4 11.xd7+ 'k!?xd7 12.tt'lc3 abB
13.d1 hdB 14.b3 'k!?eB 1S.ib2 tt'ldS
White has the bishop pair ad
vantage in an open flexible pawn
formation. In general, he should
seek asymmetric positions so
16.tt'lxdS exd5 17.e3 tt'le6 1B.ac1 bS
19.f4 100ks consistent.
6 . . . c6
157
 Part 7
6 . . . .ic6 7.0-0 tOd7 B.tOc3 tOgf6
9.e4 bS 10.tOxbS he4 11.d4 .idS,
as in Stellwagen-Sargissian, Por
to Carras 2011, should be met by
12 .e2 .ie7 13.tOc3 .ic6 (13 . . . 0-0
14.tOxdS tOxdS IS.dxcS hcS 16.b3)
14.dxcS hcS IS . .if4.
7. 0 - 0
White cannot save a tempo on
0-0 with 7.tOc3 due to 7 .. .:1kB ! . Note
that 7 . . . tOf6 is possibly inaccurate.
White has some initiative following
B .d4 bS 9 .d3 cxd4 10.tOxd4 tOeS
11 .dl b4 12.tOe4 tOxe4 13 .he4
l'kB 14.0-0 .ie7 1S . .if4 tOg6 16 . .ie3;!;.
7 . . . tOf6 B.d4 transposes to a
well known variation of the Cata
lan. Practical results suggest that
White retains a small edge after B . . .
cB 9 . dxcS o r B . . . cxd4 9.tOxd4 1'!cB
10.tOc3 tOxd4 11.xd4 .icS 12 .h4
0-0 13.hb7 bB 14 . .if3 1'!b4 1S.gS
.id4 16.d2 c7 17.tOdl, Kramnik
Naiditsch, Dortmund 2010, or B . . .
bS 9 .d3 1'!cB 1O.dxcS.
ISB
The text is more critical for the
assessment of this line. Now B.d4
bS 9.d3 cxd4 1O.tOxd4 tOxd4
11.xd4 tOf6 12 . .igS (12 .xa7 .icS
13.b7 tOdS 14.hdS c7 is a draw ;
12 .tOd2 .icS 13 .d3 0-0 14.tOe4
.ic6 IS.tOxf6+ xf6 16.hc6 xc6
17.xbS fcB) 12 . . . .icS 13 .h4
0-0 14.tOc3 h6 is about equal. The
plan with d4 is also insufficient af
ter B.tOc3 tOf6 9.d4 bS 1O.d3 cxd4
11.tOxd4 tOxd4 12 .xd4 .icS 13.d3
b4 14.tOe4 tOxe4 IS.he4 b6 16.a4
bxa3 17.bxa3 .ibS= . Perhaps White
can try to put pressure on cS with:
S.tOc3 f6 9.d3 a6
10.b3 ! ?
The game Velickovic-Sveshni
kov, Murska Sobota 2007, saw
10.ttJe4 bS 11.tOxf6+ xf6, when
12 .b3 ttJd4 13.ttJxd4 cxd4 14.a4
(14 . .id2 .id6 IS.1'!fc1 0-0) 14 . . . .id6
IS.axbS hbS 16 . .id2 0-0 17.1'!fc1 is
close to equal.
10 . . . .1e7 11 . .1b2 b5 12.'fU4
As long as c4 is under control,
White's game is easier.
 1.lLlf3 d5 2 . c4 e6 3.g3
C2. 4 c6 5.tbc4 b5 6.'9c2
ib7 7 .ig2
We have reached the same pawn
structure as we already examined in
Part 1. The fianchetto of the light
squared bishop does not change
White's general plan. He will attack
the b5-pawn, trying to drag it to b4.
That would concede the c4-square
and would assure White of a last
ing pressure in the centre. Here is
an example of White's strategy:
7 . c5 8.a4 a6 9.axb5 axb5 1O.gxa8
. .
,ba8 11.0-0 lLlf6 (11 . . . lLlc6 12 .liJc3
ttJb4 13.dl) 12 .b3 ! b4 (12 . . . ic6
13. liJe5) 13.d3;!:: .
7 . . lLlf6 8. 0 - 0 lLlbd7 9.c3
Following the main Reti strategy
- to be flexible and keep all options
open. Depending on Black's moves,
we could open the centre by d4, or
pl ay on restriction by d3-e4.
Occasionally, White inserts 9.a4
first, hoping to provoke . . . b4? ! .
The correct answer i s 9 . . . a6. Black
may try to deviate with 9 . . . gc8, but
it al lows 1O.d4 ! ? Besides, 1O.liJc3
should probably still transpose to
the main line, or even 10 .d3 fie7
11.liJc3 because 11 . . . b4? ! 12 .liJbl! c5
13.lLlbd2 0-0, as in Chuchelov-Post
ny, Belgium 2008, would be pleas
ant for White after 14.a5;!:: .
Note that 9.d4? ! c5 gives Black
nice counterplay in the centre.
9 a6!
Black prepares ... c5. Alterna
tives are :
a) 9 . . . i.e7. This mundane deve
lopment is inaccurate due to :
1O.d4!
Game 14 Radjabov-Negi,
Khanty-Mansiysk 2011, saw 1O.e4 ! ?
b 4 11.liJa4 c 5 12 .d3 gc8 13.b3 liJb6
14.liJb2 and White fulfilled the
positional aim of the opening - the
critical square c4 is in his full con
trol. Perhaps Black should have
tried the pawn sac 12 . . . c4 13.xc4
a5 14.c2 (14.b3 gc8 15.d4
liJxe4 16.i.b2 i.f6 17.e3 liJ c3) 14 . . .
gc8 15.dl 0-0, with some (albeit
insufficient) compensation.
10 ... a6 11.gdl
The thematic 11.liJe4 is also
good: 11 . . . liJxe4 12 .xe4 liJf6 (or
12 .. gb8 13 g4 ! 0-0 14 i.h6 i.f6 15
i.f4 gc8 16 e4 ! ) 13 .c2 b6 14.liJd2 !
(14.liJe5 ! ?) 14 . . . 0-0 15.liJb3;!:: , Vau
lin-Rabiega, Katowice 1993.
11 . . . c5 (11 . . . 0-0 12 .lLle5;!:: ) 12 .dxc5
bc5 . 13. liJe5 bg2 14. @xg2 ga7
15.f3 c8 16.lLlxd7 gxd7 17. gxd7
xd7 18.i.g5 i.e7.
159
 Part 7
Black is several tempos behind
in development and it is not easy
for him to find a stable place for
his queen. The game Gharamian
Libiszewski, Chartres 2 005, went
on 19.hf6 gxf6 2 0 .llJe4 0-0 2U'!d1
\Wc8, when best would have been
2 2 J'kU. Even better is 19J':id1! Vlic7
(19 . . . Vlic6 2 0 Jd2 0-0 2 l.llJd5;!;)
2 0 Jk 1 ! having in mind to meet
20 . . . 0-0 by 21.llJd5;!;.
The bottom line is: if Black does
not play . . . a6 early, White seizes the
centre with d4, because the thema
tic break . . . c5 does not work.
b) 9 .. .l'k8. This move discourag
es 1O.d4 owing to 10 . . . c5 ! , but 10.d3 !
should transpose t o the main line
after 1O . . . a6 1l.a4 i.e7 while 10 . . .
i.e7 1l.a4 b4? ! 12 .llJb1 c 5 13.llJd2
0-0 14.a5! was better for White in
Choma-Fier, Balneario Camboriu
2005.
10.a4!
It is necessary to keep an eye on
b5 or Black will favourably open his
b7-bishop with . . . c5.
lO .ie7
160
10 . . J'k8 could transpose to the
main line after 1l.d3, but White has
an even better option:
1U'!d1!
The same insidious rook move
as in the main line of Part 1. White
prepares d4 which was not good
right away due to 1l.d4 b4 12.llJe4
c5.
Let us examine Black's options
here:
a) 1l . . . Vlib6 12 .d4 c5 13.axb5 axb5
Black's two moves with major
pieces neglected the development
of his kingside so the inevitable
opening of the centre spells trouble
for him.
14.!e3 !
White had a tiny edge af
ter 14.Vlib3 cxd4 15.llJxd4 hg2
16.mxg2 b4 17.llJa4 Vlib7+ 18.Vlif3
Vlixf3+ 19.1lJxf3, Stanojoski-Lupu
lescu, Dresden 2008, but the text
practically wins a pawn.
14 . . . .ie7
14 . . . llJg4 further delays develop-
ment. White is able to choose be-
 l,lt:\f3 dS 2 . c4 e6 3.g3
tween lS.d2 lilxe3 16.xe3;!; and
lS.dxc5 hcS (lS . . . lilxe3 16.cxb6 !
lilxc2 17J':la7 gb8 18.gxd7 '>!lxd7
19 .1ileS+ '>!lc8 2 0 .lilxf7) 16.d3 0-0
17.hc5 xc5 (17 . . . lilxcS 18.d4)
18.d4, with a very active rook af
ter 18 . . . ixf3 19.exf3 lilgf6 20.xcS
ElXCS 21.gan.
lS.b3 !
Attacking the bS-pawn be
fore Black has castled. Now 15 . . .
c6 16.lileS! c 4 17.dS ! o r 16 . . . hg2
17.'>!lxg2 b7 18.f3 lildS 19.1ilxdS
exdS 2 0.gaS ! are bad for him.
The best option is to give up a
pawn with a dubious compensa
tion:
lS . . . lildS
White's pieces are better mobi
lised. He can eliminate to a bet
ter endgame with 16.xbS xbS
17.lilxbS lilxe3 18 .fxe3 0-0 19.Ela7
gb8 20.lile1 hg2 21.gxd7 .idS
22.gxe7 gxbS 23.dxcS gxb2 24.'>!lf2
gc8 2S.gan gbS 26.e4! he4 27.gc7
gbb8 2 8 .gdd7 or keep more tension
with 16.dxc5 hc5 17 . .bc5 gxcS
18. lilxdS .txdS 19.d3 lilf6 2 0.d4
0-0 21.b4;!;, e.g. 21 . . . gc6 2 2 .lileS
'<&xd4 23.gxd4 gc2 24.e4.
b) 11 ... M 12 .lile4 lilxe4 13 .xe4
c7 14.lile1 ! lilf6 lS.c4 c5 16.b3
hg2 17.'>!lxg2 b7+ 18.f3 .ie7
19.1ild3;!;. Black failed to organise
any counterplay and remained with
his backward pawn on cs.
c) 11 . . . e7 12 .d4! 0-0
An alternative is 12 ... b4 13.lile4
cS 14.lilxf6+ lilxf6 lS.dxcS aS
16 . .ie3;!;
Such positions are commonly
unpleasant for Black. His queenside
pawns are vulnerable, and he needs
to spend two tempos on castling
and making a luft. Now 16 . . . .bcS
17.hcS gxcS 18.d3 0-0 19.1ild4
would be gloom for him.
Berczes-Venkatesh, Dresden
2011, saw 16 . . . 0-0 17.lild2 hg2
18. '>!lxg2 lilg4 when best would
have been 19.1ilb3 lilxe3+ 2 0.fxe3
c7 21.gac1 c6+ 2 2 .'>!lf2 xa4
23.c4. Instead, 19.e4 lilxe3+
2 0.xe3 xc5 21.xc5 l:!xc5 2 2 .gac1
gd8 23.lilb3 gxd1 24.gxd1 gc7 (24 . . .
gc2 2S.gd2 gc4;!;) 2S.Elc1 also re
tained some advantage which Ber
czes eventually went on to convert.
13.lileS! (anticipating . . . cS)
161
 Part 7

cxd4 17.ttJbxd4 E:bB, Svidler-Gel

fand, Moscow 2011.

13 . . .'b6
13 .. .'c7 14.i.f4 i.d6 1S.axbS
axbS 16.E:a7, Andreikin-Grachev,
Moscow 2 0 1 0 , is obviously unac
ceptable, but 13 . . . ttJxeS ! ? 14.dxeS
ttJdS 1S.ttJe4 cS (1S .. .'c7 16.i.gS)
16.axbS axbS 17.ttJc3 b4 1B.ttJe4 c4
19.ttJd6 hd6 2 0 .exd6 fS 21.i.f4 E:O
2 2 .i.eS is not so clear.
After 13 .. .'b6, best is 14.ttJxd7!
ttJxd7 1S.i.e3 ttJf6 (1S . . . cS 16.dS)
16.ttJe4, with an edge. Instead, Giri
Menezes, Vienna 2011, saw 14.i.gS
E:feB? ! 1S.ttJxd7 ttJxd7 16.he7 E:xe7
17.b4. Black could have tried 14 . . .
ttJxeS ! ? with drawing chances i n the
endgame arising after 1S.dxeS ttJg4
16.he7 xf2 + 17.h1 cS 1B.ttJe4
xg2+ 19. xg2 ttJe3+ 20.f2
ttJxc2 2 1 .ttJxcS ttJxa1 2 2 .ttJxb7 E:feB
23.ttJd6 ttJb3 24.ttJxcB E:xcB 2S.E:d6
bxa4 26.E:xa6 h6 27.E:xa4;!;.
Similar positions arise in some
lines of the QGA. White usually
cannot make any progress from this
point.
b) 12 . .igS cS ! 13.axbS (13.ttJeS
hg2 14.xg2 cxd4 1S.ttJc6 eB
16.ttJe4 E:cB 17.axbS axbS 1B.E:fc1
hB 19.E:a7 ttJeS 20.E:xe7 xc6
21.hf6 xc2 22.E:xc2 gxf6 23.E:xcB
E:xcB= ) 13 . . . axbS 14.dxcS hcS
1S.E:xaB xaB 16.ttJxbS E:cB. We
see the same scenario as in the pre
vious line.
It seems that this method of
equalising against d4 works well if
Black has not lost a tempo on . . . E:cB.
11 0 - 0 12.e4

11.d3 !

I n this order o f moves, 11.d4 0-0

allows Black to solve the opening
problems by an interesting pawn
sacrifice:

a) 12.E:d1 cS ! 13.axbS axbS

14.E:xaB xaB 1S.ttJxbS i.e4 16.d2

162
 1.lt:lf3 dS 2 . c4 e6 3.g3

White's plan is to develop his 14 . . . gacS lS.axbS axbS 16.b4 !

bishop on e3, safeguarding it with transposes to the main line.
h3, and bring the rook to dl. Then
13.h3 ! ?
he will be well prepared for further
advancing in the centre or on the
I propose this novelty because
kingside.
13 . .ie3, as played in Ivanchuk
Black's only activity is connect
Grischuk, Khanty-Mansiysk 2011,
ed with . . . cS, so his most natural
allows 13 . . .lilg4 14 . .if4, when the
and flexible continuation is:
engines suggest 14 . . . b4 lS.lile2 gS
16 . .id2 cst.
Practice has also seen :
a) 12 ... cS turned out to be over 13.gd1 ygb6
optimistic in Shimanov-Goganov,
The b6-square is undoubtedly
St. Petersburg 2011: 13.axbS axbS
the best place for the queen. The
14.gxaS ygxaS 1S.lilxbS c4 16.dxc4
game Shimanov-Naiditsch, Stock
ixe4 17.yge2 lileS 1S.lilxeS ixg2
holm 2011, featured 13 . . . ygc7 14.h3
19.9d1 .icS 20 . .ie3 yge4, when
gfdS lS ..ie3 b4 16.lilb1 c5 17.lilbd2
2 1 .ygd3 ! ygxeS 2 2 .ixcS ygxcS
23.Wxg2 would have given White
the upper hand.

b) 12 . . . ygb6 defines Black's plans

one move too early. This could al
low White to save a tempo on gd1
after 13 ..ie3 cS 14.h3

It seems that White has achieved

White is slightly better because
the pawn couple on bS-cS will be
vulnerable after b2-b4 :
14 . . . gfcS lS.gfc1 ! .ic6 (lS . . . ygc6
is not a safer place for the queen
due to lild2-b3-aS) 16.b4;
all he wanted from the opening. He
needs only to put lilc4 to be thor
oughly happy, so the next move
looks consistent: 17 . . . c4 ! ?
This thematic sac often enables
Black to miraculously disentangle
his pieces. The stem game went on
with 1S.lilxc4?! lilxe4 19.dxe4 ygxc4
20.ygxc4 gxc4= . Critical is, how
ever:
1S.ygxc4 when the endgame af
ter 1S . . . YGxc4 19. lilxc4 lilxe4 2 0.lilb6
lilxb6 would be better for White.

163
 Part 7

1B . . . bB ! ? deserves attention, but White's pieces are more stable,

still 19.b3 lLlcS 2 0 .hcS (20. the bS-pawn is hanging. For exam
xb4 E:xd3 21.E:ac1 lLlfxe4 2 2 . lLlxe4 ple:
E:xd1+ 23.E:xd1 he4 24.xbB E:xbB 16 . . . eS 17.b1 i.c6 1B.E:c1;
2S.lLleS f6 26.lLld7 lLlxd7 27.E:xd7 16 ... c6 17.b1! cxb4 1B.lLld4
hg2 2 B.'it>xg2 b4= ) 20 . . . ,ixcS xc3 19.E:c1, winning the queen.
2 1 .lLlc4 a7 2 2 .c2 d6 23.'it>h2 In both variations we needed
leaves White with a healthy extra the king's rook on c1 but not on dl.
pawn, e.g. 23 . . . cS 24.E:d2 .
After 13 . . . b6 14.i.e3 cS 15.
C3. 4 ... d7 5 . .tg2 !
lS.axbS axbS 16.h3 E:fdB, White
chose in Almasi-Robson, Lub
bock 2010, 17.E:dc1 (17.b4 ! ) which
strongly suggests that White's rook
is not that useful on dl. On the
other hand, h3 looks indispensable
anyway.

13 . b6

13 . . . c7 hardly deserves atten

tion since White has 14.f4 when
14 . . . eS? ! fails to lS.lLlxeS lLlxeS
16.d4 .td6 17.dxeS heS 1B .heS
xeS 19.f4 cS+ 20.f2. This move order is more accu
rate than S.xc4 which would al
14 . .te3 c5 15.axb5 axb5 low Black to fianchetto the bishop
16.b4! with S . . . b 6 ! ? 6 . .tg2 .tb7. Compared
to the main line, White has not a le
ver on the queenside which the bS
pawn provides.

S . . . a6 6.xc4 bS is also topical.

Perhaps White should retreat here
to b3 in order to pressurise the bS
pawn :
7.b3 b7 B.O-O lLlgf6 9.lLlc3
i.e7 10 .d3 0-0 1l.a4 b4

164
 l.ttJf3 dS 2.c4 e6 3.g3

The fine point of my idea is

11.. .ttJcS 12.1!Nc2 b4 transposes to
Svidler-Kramnik while:
11 . . J%bB 12.axbS axbS 13.i.d2 b4
14.ttJd1 cS 1S.ttJe3 ttJb6 16.ttJeS hg2
17. 'it>xg2 Wfc7 1B.ttJ3c4 allows White
to achieve his strategic aim.
that 13 . . . .idS 14.ttJe3 .ib3 1S.WfbU
keeps the material balance since
lS . . . ttJxa4 (1S . . . ha4 16.d4; 1S . . . ttJdS
16.d4 ttJd7 17.Wfd3 ttJxe3 1B.i.xe3
.idS 19.9fcl as 20.ttJd2) loses to
16.ttJd2 . Another option is 13 . . . eS
as in Svidler-Kramnik, but then we
can snatch the pawn: 14.ttJxeS i.xg2
1S.'it>xg2 WfdS+ 16.ttJf3 ttJb3 17.gb1
ttJd4 1B.Wfc4 Wfxc4 19.dxc4.
6.ti'xc4

12.ttJdl ! ?
The very interesting game
Svidler-Kramnik, Moscow 25.11.
2 011, saw 12 .ttJb1 ttJcS 13.Wfc2 eS ! ?
(or 1 3 . . . .idS 14.ttJbd2 b3 1S.Wfc3 gbB
16.ttJc4) 14 . .ie3 (14.ttJxeS? loses a
piece to 14 . . . i.xg2 1S.'it>xg2 WfdS+
16.ttJf3 ttJb3) 14 . . . e4 1S.hcS hcS
16.dxe4 l/Jxe4

C31. 6 . . . a6; C32 . 6 . . . cS.

C31. 6" .a6

This move could turn out to be

a waste of time. White can open the
centre to use his significant lead in
development.
Svidler suggests here 17.e3 ! with
a b etter pawn structure. He chose
7.Wc2 !
in stead 17.ttJfd2, but the bold 17 . . .
txf2 + ! 1B.gxf2 ttJxf2 19.i.xb7 ttJg4
Discouraging 7 . . . bS due to
gave Black sufficient counterplay.
B.ttJeS, and hindering the fianchetto
12 . . . ttJcS 13.Wfc2 of Black's bishop.

16S
 Part 7

7 . c5 8.ti)c3 'ilYc7 13.ixd4 1J.e7 14.a4;!;.

Alternatives are :
a) 8 . . J:b8 9 .d4 .id6 (9 . . . cxd4
1O.xd4 .ic5 11.b3 .id6 1 2 . 0-0
0-0 13J:dl Wfc7 14 . .ie3;!;) 10.dxc5
xc5 11.0-0.
b) 8 . . Ja7 (freeing a8 for the
queen) 9.d4
9 . 0-0 b5 10 .d3 also deserves at
tention.
b2) ll.e4 ! ? This is more natural.
ll . . . cxd4 12.xd4 Wfa8.
Here in Hertneck-Schlosser,
Austria 2 005, was 13J:el 1J.e7! [13 .. .
1J.b4 14.a3 hc3 15.Wfxc3 0-0 (15 . . .
xe4 16.he4 ixe4 17J:xe4) 16.f3
i:c8 17.Wfe3;!;, Skembris-Malakhat
ko, Porto Mannu 2 009.] when
White found nothing better than
14.f3. Obviously, 13.f3;!;, followed by
i:fdl, would have been better.

9 . . . b6 1 0 . 0-0 .ib7
9. 0 - 0 !

This is more cunning than 9.d4,

because then Black can play the ac
tive 9 . . . b5 1O . .if4 Wfa7 ll.d5 xd5
12.xd5 exd5 as in Szabo-Keres,
Winnipeg 1967, when 13.a4 leads
to a messy game. The text dis
courages 9 . . . b5 in view of 1O.a4 b4
11.dl followed by b3, b2, d3. We
already know this structure from
White's game is easier. Exchang
line C2.
es only increase his advantage be
cause the possible invasion squares
9 b6 10.d4 .ib7 1l .if4 .id6
c6 and d6 lose there defenders. On
. .

12.hd6 Wfxd6 13.gfdl 0 - 0

the other hand, White should also
14.e4 cxd4 15.ti)xd4;!;
fight for tempos so ll.dxc5 would
not be precise. He should take on
c5 after . . . .id6 or make Black play
. . . cxd4 which would activate the f3-
knight. Possible continuations are :

bl) ll . .ie3 Wfa8 12J:fdl! cxd4

Further fight for tempo with

12 . . . h6 is risky, because White has
a more useful move: 13J:acl .ie7
(13 . . . g4 14.if4 cxd4 15.a4 e5
16 . .ih3+-) 14.dxc5 ixeS 15.ixc5
xc5 16.b4 cd7 17.e4 0-0 18.d4;!;.

166
 l.ttJf3 dS 2 .c4 e6 3.g3
White has reached a typical po
sition with an initiative for him. He
has more space and his plan is to
increase this advantage by pushing
eS or b2-b4-bS. Follow the model
of Kortschnoj-F.Olafsson, Moscow
1971:
15 'ffc7
. (1S .. .'I!fie7 16.'1!fie2)
16.'ffe 2 ga7 17.gac1 Yfb8 18.a4
.ia8 19.b4! gc8 20.b5 (claiming
the outpost on c6) 20 gac7 21.e5
. 10 .tg5 .ie7 1l.ttJxd4 hg2
hg2
2 1 . . . :B:xc3 loses to 2 2 . :B:xc3 :B:xc3
23. ttJc6 and the weakness of the last
rank is decisive.
22.'i!.>xg2 Wb7+ (22 . . . ttJe8
23.ttJc6 a8 24.e4 ttJcS 2S.Wb4)
23.ttJc6 (23.f3) 23 ttJe8 .
Compare the two diagrams !
Without having committed any
serious mistake, Black is totally
dominated. Now 24.ttJa2, heading
for b4, would have crowned Kor
tschnoj 's strategy.
C32. 6 . . c5 7. 0 - 0
7.b3 ! ? is much more popu
lar, but 7 . . . :B:b8 allows Black to
fianchetto his bishop.
7.d3 .id6 100ks equal. The text is
more simple and consistent. White
does not spend tempos on finesses,
but he stakes on better development.
7 b6 8.d4 .tb7 9.gd1 cxd4
12.'i!.>xg2
White is obviously better, Aro
nian-Gelfand, Nice 2010. The c6-
square and the a7-pawn are weak.
12 . . . :B:c8 would not be of much
help due to 13.a4 c7 14.ttJc3 0-0
1S.ttJdbS b7+ 16.f3 a6 17.ttJd6;!; so
Gelfand kept his rook on a8 :
12 . . . c8 and went on to gradual
ly equalise after 13.ttJd2 ? ! h6. White
should send his queen's knight
to another route : 13.Wxc8+ :B:xc8
14.ttJc3 ! 0-0 1S.ttJcbS a6 16.ttJa7 :B:c7
17.:B:acU, underlining the weakness
of c6.
167
Pa rt 7
Complete Games
1 4. Radjabov-Neg i
Kha nty- M a nsiysk 3 1 .08.201 1
This game is a fine example of
how to exploit the weakness of the
c4-square in the typical pawn struc
ture with a backward black pawn on
c5.
1 . f3 f6 2 . c4 e6 3.g3 d5
4 . .ig2 dxc4 5.a4+ c6 6.xc4 b5
7 .c2 .ib7 8.0-0 bd7 9.c3 .ie7
1 0 .e4
Black's last move does not pre
pare . . . c6-c5 so White can answer
1O.d4 ! ? Radjabov has another
set-up in mind. He restricts the b7-
bishop with e4-d3 and will try bind
the opponent with the defence of
the c-pawn.
1 0 . . . b4?
168
This is an instructive positional
mistake. Black can accept this pawn
structure only if he had a concrete
variation in mind. It is possible
that Black will have to play it later,
but only after White had put in a4.
Even then, the concession of the
c4-square should be a major flaw of
Black's position, but at least White
would not be able to open the a-file.
1O . . . a6 also seems inaccurate
in view of 1l.d4 c5 12 .d5 exd5
13.xd5 xd5 14.exd5 0-0 15.d6
.if6 16.lDg5;!;. Perhaps Black should
have castled. Then 1l.d4 would
be no longer scary due to 1l . . . b4
12.lDa4 c5.
1 1 . a4 c5 1 2 .d3
1 2 .. Jc8?
After this reckless move, Black's
 l.tLlf3 d5 2 .c4 e6 3.g3

queenside gets into a positional 1 7.a3 as 1 8 J'!fc1 Ud8 1 9.YlYe2

vice. I would prefer to throw in a
pawn, but keep my pieces active.
For instance : 12 . . . c4 13.xc4 a5
14.c2 l'3cB. It is true that White
retains the extra pawn, e.g. 15.d1
0-0 16.a3 bxa3 17.b3 l'3fdB 1B.ha3
ha3 19.1'3xa3 b4 2 0 .aU tLleB
2 1.l'3a2 tLld6, but it is usually very
difficult to convert the b-pawn in
such positions.
1 3. b 3 b 6 1 4. b2 0-0 1 5 . .lf4
h 5 1 6 . .le3
19.axb4 axb4 2 0 . l'3a7 bB
21.l'3cal was quite good, but White
wants to provoke another weak
ness. In many variations he will
have d4, e.g. 19 . . . l'3aB 2 0.hc5 hc5
21.d4 so Black's response looks
forced.
1 9 . . . e5 20 . .lh 3 ga8 2 1 . YlY c2 .lc8
22 . .lxc8 gdxc8 23. c4

White's strategy has triumphed.

The opponent has not a single ac
1 6 . . . YlYc7 tive piece.

Black looks completely at a loss. 23 . . . d 7 24.YlYa2 gab8 25.YlYb2

In my opinion, he should seek coun f6
terplay at any cost so 16 .. .f5 is in
dispensable. White would have an
edge after 17.a3 ! f6 1B.axb4 (lB.
tLle5 ! ? J.d6 19.tLlbc4 i.bB 20.l'3fdU)
1B ... fxe4 19.tLlg5, but at least both
armies would be in a direct conflict,
so White would be prone to mis
takes.
Play might continue with 19 . . .
lLlfd5 2 0 . tLlxe6 tLlxe3 21.fxe3 cxb4
2 2 .tLlc4 l'3xf1+ 23 .hf1 d7 24.l'3xa7 26 . .ld2 ? !
( 24.tLlf4 exd3 25.xd3 l'3aBt) 24 . . .
xe6 25Jxb7 tLlxc4 26.dxc4 i.c5 Radjabov was obviously feasting
27. d2t. his eyes on the position and did not

169
 Part 7

want to alter it. This often slips a still owns the initiative, though,
great deal of the advantage. His mi thanks to his active rook. There
nor pieces are already perfect. He fore, Negi should aim to trade it
only needs to invade the opponent's and 33 . . . l"!a8 34.l"!ca1 ltJd4 served
camp with a heavy piece. Stayed his well this goal. However, he misses
pawn on a4, his edge would have his chance and his position deterio
been only little and he should have rates again.
sought to open a second front in the
centre or on the kingside. In the ac
tual position, however, he had the
obvious 26.axb4 axb4 27.l"!a6 l"!a8
28 .\I9a2, accelerating the pace of
the game.

26 . . . g 6 27 . e 1 fS 2S.e3
YlYdS 29.d5 e6 30.axb4 axb4
3 1 .l"!a7 i.d6 32.e2 f5 33.ee3
33 . . . d41! 34Jea 1 f41 35.e4
Things have changed. Black al f6 36J 1 a6 ge6 37.gxe6 xe6
ready have a gorgeous knight which 3S.ga6 geS 39.xf6+ YlYxf6 40 .J.xf4
counterbalances the d5-one. White YlYe6 41 . .ie3 .ifS 42 . a 5 1 -0

170
 Pa rt S

Anti-Queen's Gam bit I I

1 . tDf3 d 5 2 .c4 e6 3 . g 3 tDf6 4 . .ig 2 i.e7 5.0-0 0-0 6 . b3

171
Pa rt S
Main Ideas
I.IOf3 d5 2.c4 e6 3.g3 1Of6 4 . .ig2
.ie7 5. 0 - 0 0 - 0 6.b3
I had a very difficult choice to
make between the Catalan with
6.d4 and the Reti-style 6.b3. In
practice, players' preferences are
divided almost equally between the
two options. In my opinion, Black is
very close to equality in the Catalan,
with accurate play. The focus in this
opening is on the endgame so one
needs good technique and subtle
positional understanding.
Eventually, I chose to advocate
6.b3. It is more flexible and allows
different move orders. At the same
time, the strategic ideas are clear
and easy to play even by beginners.
A repertoire with 6.b3 can also
be a bonus against other popular
openings. Here are two examples :
172
l.1Of3 IOf6 2.c4 e6 3.g3 b6 4.ig2
ib7 5.0-0 .ie7 6.b3 0-0 7.ib2 c5
8 .e3 d5 9.lOc3 ;
l.lDf3 c 5 2.c4 lDf6 3.g3 b6 4 . .ig2
.ib7 5. 0-0 e6 6.b3 ie7 7.ib2 0-0
8.e3 d5 9.lDc3 .
You see that this set-up could b e
used t o avoid the Hedgehog and the
main lines of the Queen's Indian.
Black has two main approaches
against 6.b3. One of them is to de
fine the pawn structure by taking
on c4 - set-up A. The other one is to
maintain the tension in the centre
and leave White to exchange on d5.
I'll consider it as set-up B . In both
cases Black fianchettoes his light
squared bishop.
Much less often Black chooses
the reversed Modern Benoni :
6 . . . c5 7. ib2 lDc6 8.e3 d4 9.exd4
cxd4 lOJe1!
 l.lLlf3 dS 2.c4 e6 3.g3 lLlf6 4.i.g2 i.e7 S. O-O 0-0 6.b3
White should not rely on the fact
that he has one or two extra tempos
compared to the Modern Benoni.
He has less space in the centre. If
Black consolidates and carries on
. , .e6-eS, the i.b2 may turn into a
really ugly piece. Look at the game
Obukhov-Kolomensky, Orsk 2 0 0 0 :
10 . . .!!e8 1l.a3 ? ! a s 12 .d3 i.cS
13.lLleS lLlxeS 14.!!xeS 'lWd6 ! 1S.'lWe1
i.d7 16.Eie2 eS 17.lLld2 i.fS 18.lLle4
lLlxe4 19.he4 he4 20.Eixe4 fS
21.Eie2 b6
This is the worst positional
scenario into which White could
be entrapped. It took him only 2 0
opening moves t o get into a murky
cramped position. And he has not
committed any apparent mistake !
The moral of this example is that
White should get rid of his restric
ted bishop at the first opportunity:
10 . . . Eie8 1l.d3 i.cS 12 .i.a3 !
I believe that only this plan
leaves White in the battle for the
opening advantage. See game 15
Sherbakov-Grigoriants, Mos
cow 1999 for more details.
It is also a good idea to trade a
pair of knights. This is especial
ly true when . . . eS looms as after
1O .. .f6. Then best is 1l.lLleS! lLlxeS
12.EixeS f6 13.Eie1 eS
Again: 14.i.a3 ! . Then we quickly
roll our queenside pawns, starting
with c4-cS.
A. 6 c5 7.i.b2 c6 8 .e3 b6
9.c3 dxc4 10 .bxc4 i.b7 n:ee2
This variation is still inade
quately covered in opening books.
It is true that Mihail Marin has
spent tons of ink on it in his The
English Opening, Volume 2, but in
173
 Part 8

fact his work has hardly advanced

theory any further. We'll see in the
"Step by Step" chapter that the plan
he advocates in the main line does
not work.

Marin bases his repertoire on

the idea of playing gfdl followed by
d2-d4 against any Black's set-up.
However, I see at least four varia
tions where this plan does not give
any advantage. Thus the question
how to play this position as White
remains open.
I will investigate the old classic
plan of a kingside pawn storm with
f2-f4/g2-g4, having also h2-h4-
hS-h6 in mind. It leads to strategi
cally unbalanced positions without
forced variations, where the cost of
every move is higher for Black be
cause his king is in danger.
14 . . . ge8 IS.a3 lLlc6 16.dS ! ? exdS
17.lLlxdS lLlxdS, A.Petrosian-G.
Kuzmin, Baku 1977. Here, 18.gxdS !
c7 19.gel would have bound Black
up and down.
b) 1l . . . d7 12 .gfdl ! gfd8 !
12 . . . gad8 is bad due to 13 .d4
cxd4 14.exd4 lLlaS lS.lLleS c8 16.dS
13.d4 cxd4 14.exd4

Black has tried nearly all legal

moves in the diagram position. His
only real counterplay is based on
the breakthrough . . . b6-bS (with, or
without the preparatory . . . a6) . Thus
the most consistent answer should
be 1l . . . a6. Amazingly, this continu
ation is relatively rare. More often
Black chooses developing moves. As a rule, White is slightly better
Two of them - 1l ... lLlb4 and 11. . . with hanging pawns if his knights
d7, allow a simple and straight are active, as in the diagram posi
forward approach from White. He tion, where d4-dS is always on the
does play d4, followed by d4-dS (if agenda. Shift the knight from c3 to
possible ! ) and takes over the initia d2 with the manoeuvre lLlc3-bl-d2
tive : (if the c4-pawn required extra pro
tection), and the game would be
a) 1l . . . lLlb4 12 .d4 cxd4 13.exd4 balanced if not even more pleasant
gc8 14.gfdl for Black.

174
 l.tDf3 d5 2 .c4 e6 3.g3 tDf6 4.i.g2 i.e7 5.0-0 0-0 6.b3
Unfortunately, the early d2-d4
hardly brings White an edge after
1l . . . a6, 1l . . JkS or 1l . . . c7. Instead,
White has another plan which is
well tested, but it is still hiding a lot
of fine points. He remains passive
in the centre and launches a pawn
storm on the kingside with the help
of the f- and g- (or g- and h-) pawns.
This set-up has a slower version
with l:l:abl + l:l:fdl , and a sharper vari
ant with l:l:adl (or l:l:abl)+l:l:fl. The
former is universal and it works
against most Black's defences. The
latter is more risky, but it is tacti
cally justified in the event of passive
inaccurate play from the opponent.
Note that engine's evalua
tions may be misleading!
White's attack usually devel
ops slowly and decisive collisions
happen after 15 or more moves of
redeployment and manoeuvring.
Thus a depth of 22-25 half-moves
is typically insufficient for the best
engines to assess correctly the full
potential of White's assault.
Let us now investigate :
1. 1l" .a6 12.gfdl ! ? c7
13.gabl! gab8 14 .tal gfd8 15.d3
The first stage of our plan is
complete. We have firm control of
b5, eventually we can reinforce it
with a2-a4. Any troubles along the
d-file are also ruled out for near
est future. In short, White has won
time for regrouping and launching
the pawn storm on the flank. His
next step will be tDel, followed up
by f2-f4 or g2-g4.
Black's task is to wait for us to
loosen the grip and counterattack
with . . . b6-b5 or even . . . V-fS. If he
rushed on the 14th or 15th move to
prepare . . . b6-b5 with . . . tDa7, then
d4! would have taken the initiative.
So he makes another useful move :
15" .J.a8 16.tDel and only now
16". tDa7
The continuation of the waiting
strategy with 16 . . . tDeS 17.f4 is con
sidered in game 19 Zaichik-Tiv
iakov, Moscow 1994.
17.,txa8 gxa8 18 .g4!
White's design is to push g4-g5,
g4, tDe4, h2-h4-h5-h6.
Note that White's set-up is
possible against practically
any move order of Black.
175
 Part S
2 . 1l c8?!
This move is a pure waste of
time. It does not prepare . . . b6-bS
so White can ignore this threat and
opt for a more aggressive set-up
than in the previous line :
Leaving the other rook on flo It
will enhance the effect of f2-f4-fS
and the rook lift gfl-f3-h3 is also a
valuable attacking resource. White
often begins with IVtJel, but the
text is more accurate. In this set
up, the best stand of Black's queen
is d7 from where it is eyeing all four
critical squares : bS, fS, d3 and d2 .
However, 12.gadl d7 would be
dubious due to 13 .d4 cxd4 14.exd4
.ib4 IS.dS ttJaS 16.ttJeS;!;.
An example of White's further
play is the following variation: 12 . . .
c7 13.ttJel a6 14.f4 ttJa7 1S.a4 hg2
16.ttJxg2 d7 17.g4 ttJe8 18 .fS
You should also see game
17 Kharlov-Kosyrev, Samara
20.06.2000 for 12 . . . ttJaS 13.ttJelo
Finally, 1l . ti'c7 is a particular
176
case. Play may transpose to line
1 after 12J'!fdl, or take a sharper
course following 12.ttJelo In the lat
ter case, White plays similarly to
line 2, but he is a tempo down. His
attack is still dangerous, though.
I was often hesitant during my
analyses how to recapture on g2
when Black exchanges the bishops
himself. Perhaps there is no gene
ral recipe, but I suggest to take by
knight - ttJxg2, when our rook is
on fl. We may send it later to f4-hS.
In the set-up with gfdl, our attack
will often feature g4-gS instead of
f4-fS. Then we'll need our knight on
another route - f3-eS-g4(f7)-f6/h6,
so we should prefer xg2 (or gfl
f2xg2).
I have avoided more details on
purpose, because play is not forced
and both sides have too many de
cent possibilities. It would be im
possible to analyse them branch
by branch. It is better to remember
the main plans and build up a good
understanding of the position. I
hope the next lines will arm you ad
equately for practical battles.
Attacking G u ide
I'll try now t o present the most typi
cal patterns of White's kingside at
tack. I begin with a few examples
where White's rook remains on
flo This enables tactical solutions
along the f-file :
 l.lLlf3 d5 2.c4 e6 3 .g3 lLlf6 4 . .ig2 i.e7 5.0-0 0-0 6.b3

Analysis The pressure on e6 forces Black

to concede the d5-square as 20 . . .
Wd7 i s bad i n view o f 2 1 .lLlce4 or
2 1.fxe6 fxe6 2 2 .d4.
2 0 ... exfS 2 U xf5 lLlbd7 2 2 . lLld5
E1e8 23.E1dfl E1c6 24.e4 i.d8 25.lLlxf7
Wxfl 2 6.Wh5+ Wg8 27.lLlxf6+ 1-0 .

Analysis

2 1 . lLlh5 b5 2 2 .lLle4 bxc4

23 J''!xf8 + ! .ixf8 24Jfl+- b7
25.f2 We7 26.lLlxc5 cxd3 27.lLlxe6.

Sherbakov-Log inov
St. Petersburg 1 998

19.f6 ! ! gxf6 (19 ... i.xf6 20.E1xf6 !

gxf6 21.lLlf4 E1fd8 2 2 . lLlh5 lLle8
23.lLle4) 2 0 .lLlf4 b5 2 1 . axb5 axb5
2 2 .d3 b4 23.lLle4 lLlxe4 24.dxe4
Wb7 25.lLlh5 e5 26.g5 + - .

Kharlov-Kosyrev
20.g5 ! ! hg5 2 U '!xf8+ Wxf8 Samara 2000
2 2.lLle4 i.h6 23.a5 (or 23J'!f1+
c;t>g8 24.g4+- We7 25.lLlf6+ lLlxf6
2 6 .E1xf6 g6 27.E1xe6) 23 . . . bxa5
24.lLlxc5 1-0.

M a rkus-Peru novic
Su botica 2008

White has simply followed the typi

cal attacking scheme and went on
to obtain a promising position de
spite Black's ingenious resistance :
21.E1h3 b5 2 2 . axb5 axb5 23.cxb5 c4

177
 Part 8

24.d4 lLld6 25.g5 g6 26.lLlf3 lLlxb5 White often resorts to the e

27.lLle5 'gc7. Here, best would have
been 28.lLlxt7 Wxt7 29.'gxh7+ Wg8
30 .'gh6 .if8 31.'gxg6+ .ig7 32.'gxe6.
White's attack is much slower
after 'gfdl. We needs additional
resources and that is usually the h
pawn. The following diagram is a
model position with an advantage
for White:
Analysis
pawn in order to repel the enemy
pieces. This commonly happens
when Black has sent his c6-knight
away from d4, for example to a7 or
b4:
Zaich ik-Tiviakov
Moscow 1 994

Ribli suggests here 27.f6

and again, Black is significantly
cramped.
1.h6. On the left wing White is
balancing while on thE: kingside he
enjoys a serious spatial advantage.
Black's pieces are on the defensive.
B. 6 . . . b6 7 ..ib2 .ib7 8 .e3
Note that this plan is no less ef
fective without queens :

A.Sokolov-Thesing
Berl i n 1 993

Now Black has two major pos

sibilities : to develop the knight on
c6 or d7.

29.g6 fxg6 30.lLlfg5 ixg5 when a) 8 . . . c5 9. lLlc3 tLlc6 10.cxd5

31.hxg6 ! ! was winning beautifully. tLlxd5 1l.tLlxd5 %\'xdS 12.d4

178
 l.lLlf3 dS 2 .c4 e6 3.g3 lLlf6 4 . .ig2 e7 S.O-O 0-0 6.b3
Black opts for this variation
mainly in order to kill all the action
(and trade as many pieces as pos
sible), and dry up the position prac
tically by force. It is a tough nut to
crack, indeed. We should be glad to
obtain even the slightest edge in the
endgame after:
12 ... lLlM 13.lLlh4 %1fd7 14.dxcS
%1fxd1 1SJUxd1 ixg2
16.xg2 hcS 17.a3
The good news is that White
still has some lead in development
while Black is playing for two re
sults only.
There is one long variation that
requires memorisation:
12 . . . l'!ad8 13.lLleS %1fd6 14.dxcS
%1fxcS 1S.lLld7
1S . . . . %1ffS (1S . . . %1fgS 16.h4 %1fh6
17.l'!cl) 16.e4 %1fgS 17.h4 %1fh6
18 .c1 gS 19.hxgS %1fg7 2 0 .eS ! h8 !
2 1.if4 .ia3 2 2 .b4 ! ixb4
This was Rath-G.Flear, Esbjerg
1982. Marin points out here to
23.l'!c1 .icS 24.l'!c3, intending l'!d3,
and White is on top.
b) 8 tLlbd7 9.tLlc3 (9.%1fe2 as ! ?
i s unclear to me) 9 tLle4
The waiting strategy 9 . . . cS
1O .%1fe2 l'!c8 (1O . . . lLle4 11.l'!fd 1 ! )
1l.l'!ac1 %1fc7 leads t o the following
position :
I like Malakhov's idea 12.tLlh4 ! ?
dxc4, transposing t o the structures
from set-up A.
10.'tfe2 ! .if6 11. cxd5 tLlxc3
12.hc3
179
 Part 8

12 . . .i.xdS 13.d4 i.e4 14.ac1 fie7

lS.ltJd2 ! ixg2 16.'ktixg2 cS 17.ltJe4
cxd4 18.exd4 ac8 19.fibS;j;

Black has delayed . . . cS and he

may be unable to achieve it at all.
That makes the whole complex of
light squares on the queenside,
and especially c6, rather weak, for White is menacing i.b4. All his
instance : 12 . . . hc3 13.dxc3 hdS pieces are active, in a sharp con
14.ad1 'itfe7 1S.ltJd4 or: trast to the opponent's army.

Points to remember:

In the reversed Modern Benoni, trade bishops with 1a3.

If Black takes on c4, our main plan is ltJe1 and a pawn storm on
the kingside. Only 1l . . .'itfd7 and ll . . . ltJb4 call for d4.
If Black plays . . . ltJbd7 and waits, we prepare f2-f4 by ltJh4.
If Black plays . . . ltJc6 and waits, we open the centre by exchanging
on dS and later on cS, hoping to use our lead in development.

180
Pa rt 8
Step by Step
1.c!lf3 d5 2.c4 e6 3.g3 f6 4 .tg2
ie7 5. 0 - 0 0 - 0 6.b3
Of course, the Catalan, which
arises after 6.d4, is an excellent
alternative. Moreover, White has
sidestepped quite a few complex
branches, for instance, the check
from b4.
After 6.b3, I will focus on:
A. 6 . . . c5 ; B. 6 . . . b6
Minor alternatives are :
a) 6 . . . d4
White can transpose to the vari
ations I consider on move 8 with:
7.e3 c5
7 . . . lLlc6 8.exd4 lLlxd4 9.!b2
lLlxf3 + (Black cannot hold the d4-
square anyway - 9 . . . !c5 1O.lLla3
followed by lLlc2 .) 1O.Wxf3 1'!b8
11.We2 b6 12 .lLlc3 .ib7 13.1'!ad1 ,ixg2
14.xg2 offers White some space
advantage. Botvinnik-Stahlberg,
Amsterdam 1954, went 14 . . . c6
15.lLle4 1'!b7 16.lLlg5 lLld7 17.lLlf3 !f6
18.d4;!;.
8.exd4 cxd4 9.!b2.
Besidses, 9.d3 lLlc6 10 .1'!e1 lLld7!
1l.lLla3 is also playable and leads to
typical Modern Benoni positions
with reversed colours.
b) 6 . . . a5 7.lLlc3 d4
After 7 . . . b6 8 .!b2 !b7, White
will play d4 at once or after 9.e3
lLla6. It looks like a side-line Cata
lan, where Black's plan is not too
clear while White will exchange on
d5 and will try to put pressure along
the c-file.
7 . . . lLlbd7 8 .!b2 lLlb6 9 .d3 a4
10.Wc2 a3 1l . .ic1 d4 12. lLle4 turned
well for White in Altykenov-Po
livanov, Alushta 2009. Of course,
8.d4 is also a sound choice.
181
 Part B

A. 6 c5 7 .i.b2 c6
a) 1O ... lLleB, aImIng for . . .f6,
... eS, or b) 10 ... e8 followed up by a
If Black wants to play the re bishop's retreat to cS or f8 .
versed Modern Benoni, he com
A minor alternative is 1O . . . d3.
monly prefers to do it on the next
This move only opens the diagonal
move. There is no objective reason
to White's worst piece and invites
for this because after 7 . . . d4, White
1l.lLlc3 . Then the only consistent
has not anything better than B.e3
continuation is Tal's choice 1l . . . lLld7
c6. In practice, he also tries B.b4,
[or 1l . . . lLlb4 12 .c1 e8 (12 . . . lLlc2
but then B . . . aS should equalise.
13.f1) 13.lLleS lLld7 14.lLlxd7 xd7
1S.hS;!;] 12 .a3 as when 13.lLlbS lLlcS
8.e3 b6
14.lLleS gave White the more active
pieces in Andersson-Langeweg, Am
B . . . dxc4 9.bxc4 d3 does not
sterdam 1973 .
make much sense because trading
queens does not solve Black's prob
a) 1 O lLleB
lems with development: 1O .b3
. .

dB ll.d1 a6 12 .xd3 xd3 13.lLlc3 This is too passive. White should

dB 14.lLla4 bB 1S.lLleS;!;. answer:
1l.lLleS !
B . . . d4 9.exd4 cxd4 10 .e1 ! leads
to the reversed Modern Benoni : White has less space s o he
should exchange a couple of minor
pieces to ensure more breathing
space and activate the g2-bishop. It
is also important to free the way to
the f-pawn in order to enable f2-f4.
The inclusion of 1l.a3 as is po
sitionally wrong in my opinion, be
cause it deprives White of the pos
sibility of trading its dark-squared
bishop: 12.lLleS lLlxeS 13.xeS f6
It is pointless to count the tem 14.e1 eS 1S.d3 lLlc7 16.lLld2
pos here. !b2 may be considered
a step in the wrong direction, but
the hit on d4 prevents the thematic
Benoni redeployment lLlf6-d7-cS. It
is more important that Black can
not achieve . . . e6-eS, which is the
cornerstone of any active plan for
him. He should work hard to enable
it with:

1B2
 1.tt'lf3 dS 2.c4 e6 3.g3 tt'lf6 4 . .tg2 fie7 5.0-0 0-0 6 .b3

White's bishop on b2 is un
employed. The imminent break
through f2-f4 will increase its im
pact on the centre, but not enough
for a serious advantage. Play may
continue 16 . . . lDe6 17 . .tdS hB 1B.f4
lDcS 19.lDe4, with mutual chances.
1l . . . lDxeS 12.gxeS f6 13.ge1 eS
13 . . . lDc7 ! ? is a way to avoid the
sharp variation with f2-f4. White This crazy endgame occurred
should probably transpose to the in Kosten-Luther, Austria 2009.
main line with 14.i.a3. Wahls The only thing I can say is that it is
Chandler, Germany 2 0 0 2 , saw a totally unclear to me. Only a very
similar plan : 14.d3 gbB 1S.i.a3 deep computer analysis can shed
i.xa3 16.lDxa3 eS 17.vtId2 vtId6 some light on it, but I prefer to play
1B.lDc2 hB 19.b4 b6 20.f4;!;. chess and not to spend my time on
memorising long variation, where
even a considerably weaker oppo
nent might beat me thanks to a bet
ter computer assistance. This game
went further 22 . . . gdB 23.ge7 lDe6
24.lDc3 gB 2S.lDdS fB 26 . .ihS
(26.gxe6 i.xe6 27.lDc7 .ifS 2B.lDxaB
gxaB 29.i.xb7 gdB 30.i.dS)
26 . . . lDgS 27.gae1 i.d7 2B .d4 .teB?
(2B . . . gacB ! ) 29.,ixeB gxeB 30.gxb7
14.i.a3 !
gxe1 31.xe1 lDf3+ 3 2 . f2 lDxd4
I recommend this exchange 33.e3 lDfS+ 34.f4 lDd6 3S.gd7
in most branches of the reversed lDeB 36.cS gcB 37.b4 as 3B.a3 axb4
Modern Benoni. In my opinion, it is 39.axb4 hS 40.fS gaB 41.g6 1-0.
principally wrong to play with bad
pieces in one's camp. Tarrasch's 14 ... i.xa3 1S.lDxa3 lDc7 16.gc1
formula was : one bad piece equals gbB 17.cS
a bad game. To be fair, White has
also a tactical solution of that prob
lem. It cannot be a main repertoire
si nce Black can easily sidestep it
with 13 . . . lDc7, but at least it is fun :
14.f4 exf4 1S.vtIf3 fxg3 16.vtIdS+
hB 17.hxg3 vtIxdS 1B.i.xdS i.d6
19 . bd4 lDc7 2 0.i.f3 i.xg3 2 1.i.f2
.

1xf2 + 2 2 . 'it>xf2 .

1B3
 Part S
White's pawn majority on the
queenside makes his game prefer
able. The blockading attempt 17 . . .
b S should b e met b y lS.cxb6 axb6
19.f4 exf4 20.f3, with an initiative.
b) 1O .. J%eS 11.d3 .tcS 12 . .ta3 !
I believe that only this plan
leaves White in the battle for the
opening advantage. Marin's recom
mendation 1 2 .lbeS tLlxeS 13JxeS
is insufficient due to 13 .. .'d6 ! '
See game 15 Sherbakov-Grigo
riants, Moscow 1999 for more de
tails.
Now Black has to define the fu-
ture character of the game.
lS4
AI. 9 . . . .ib7; A2 . 9 . . . dxc4 ; A3 . 9 . . .
.ta6.
Al. 9 . . . .ib7 10.cxd5 xd5
10 . . . . exdS 11.d4 offers White a
very easy game. One recent exam
ple is Vitiugov-Frolyanov, Russian
rapid ch. 2011: 11 .. J%eS (11 . . . .ta6
12.tLleS) 12 Jk1 .tfS 13.tLle2 (the pro
phylactic move 13.ge1 is more po
pular) 13 . . . tLle4 14.tLlf4 .ia6 lS.ge1
c4 16.bxc4 hc4 17.a3 gcS lS.tLleS
tLlxeS 19.dxeS bS 2 0.g4;!;.
An interesting alternative is:
12 ... exdS 13.d4 as ! ?
O r 12 . . . .ta6 13.ge1 c 4 14.bxc4
hc4 lS.tLld2 bS 16.tLlxc4 bxc4
17.a4 tLlb4 1S . .tc3 tLld3 19.gebU.
13.dxc5
13 .d2 ! ? is more restrained, but
White's plan, based on the weak
ness of the dS-pawn, is very clear
and easy to implement: 13 . . . a4
14.gfd1 axb3 lS.axb3 gxa1 16.ha1
aS 17 ..tc3 gdS lS.b2 .tfS 19.tLle1
cxd4 2 0.hd4 tLlxd4 2 I .xd4 aS
2 2 .tLld3 bS 23.b2 !d6 24.tLlf4
.txf4 2S.gxf4;!;, Hickl-A.Sokolov,
Switzerland 2007.
13 . . . bxcS 14.tLle1 !
14.tLlh4 hh4 lS.gxh4 xh4
16.xdS tLldS 17.eS f6 1S.g3 was
only equal in Reinderman-Irwanto ,
Jakarta 2011: lS . . . xg3 19.hxg3
hg2 20.'it>xg2 tLle6 2I..ia3 gtbS = .
 l.lLlf3 dS 2.c4 e6 3.g3 lLlf6 4.i.g2 i.e7 S.O-O 0-0 6.b3
14 . . . .ia6 1S.lLld3
White retains some initiative,
but he must be ready to sacrifice the
exchange after lS .. .1!c8 16.1!e1 c4
17.lLlf4 .ib4 18.lLlxd5 he1 19.xe1
cxb3 20.axb3 l!e8 21.d1, with a
great compensation.
12.d4
Black has tried many continua
tions here, but I'll stop my attention
to : All . 12 . . . l!ad8; Al2 . 12 . . . lLla5 ;
A13. 12 . . . lLlh4.
Minor alternatives are:
a) 12 . . . l!fd8 13.lLle5 d6 14.f3
(or 14.h5) lLlxeS lS.xb7 lLld3
16 . .ia3 l!ab8 17.xa7, Pigusov
Dzuban, Riga 1988.
b) 12 ... d8 13.lLleS (I would not
have mentioned this move since
13.dxcS .ixcS 14.lLle5 lLlxeS lS ..ixb7
l!b8 16 . .ie4:t is obviously in White's
favour, had not Kramnik chosen it
versus Radjabov in Kazan 2011.)
13 . . . lLlxeS 14.i.xb7 l!b8 lS . .ig2 cxd4
16.exd4 lLld7 17.dS exdS 18.xd5
.if6 19.1!ad1 .ixb2 2 0 .xdn. If
you had Kramnik's technique, you
would have fair chances to win this
endgame.
All. 12 gad8
This is an invitation to the long
forced variation :
13.e5 d6
13 . . . xg2+ promises Black only
a difficult struggle for the draw.
14.dxc5 xc5 15.d7
15 . . . . 'efS (lS . . . g5 16.h4 h6
17J:k1) 16.e4 'eg5 17.h4 'eh6
18 . .lct g5 19.hxg5 g7 20.e5
c;t}h8 ! 21 .lf4 .la3 2 2 .b4 ! .lxb4
18S

This was Rath-G.Flear, Esbjerg
1982 . Marin points out here to
23J!el .tcS 24.l'k3, intending gd3,
and White is on top.
A12 . 12 . . . a5 13.dxc5
The most principled answer, but
White has other good alternatives
as well: 13.h4 ! ? or 13.gel ! ? gac8
14 . .ia3 ! f5 (14 . . . gfd8?! IS.dxc5 ixc5
16.ixc5 c5 17.xdS gcxdS 18.d4
gSd7 19.b4) IS.dxc5 ixc5 16.ixc5
c5 17.ttJd4 eS 18.ixb7 ttJxb7
19.b4 el 20.xel ttJd6 21.gdU.
13 . . . xd1
13 . . . ixc5 14.e2 also favours
White : 14 . . . gfd8 IS.gfdl hS
(IS . . .fS 16 . .ic3 ttJc6 17.b4 ttJxb4
18.ixb4 ixb4 19.ttJh4) 16.h3 f6 !
17 . .ic3 ttJc6 18.c4 f7 19 .b4 .tf8
2 0.bS;!; (Marin).
14Jfxd1 hc5 15 .lc3 ! ?
15 Uc8
16. e5 hg2
17. Wxg2 f6 (17 . . . ttJc6 ! ?) 18 .ha5
bxa5 19. c4;!; (Marin) .
186
Part 8
A13. 12 . . b4
A tough nut to crack! Black has
considerable chances to make a
draw in this line. Compared to the
previous line, the knight attacks the
a2-pawn and that restricts White's
choice. Most often White answers
with :
13.h4
13.ttJel has similar ideas : 13 . . .
d7 14.dxcS xdl IS.gxdl ixg2
16. Wxg2 , with a tiny edge for White:
16 . . . ttJxa2 17.cxb6 axb6 18 .gd7 .tf6
19.ixf6 gxf6 20.ttJf3 or 16 . . . ttJdS
17.e4 ttJf6 18.Wf3.
13 d7
14.dxc5 xd1
15.gfxd1 .bg2
I tried to make 16.ttJxg2 work,
but then 16 . . . ixcs 17.a3 c2 !
18.gael ttJxa3 19.gxcS bxcS 2 0 .ixa3
gtb8 is unclear. In such endgames
a rook plus a distant passer is in no
way weaker than a .t+ ttJ .
 l.ltJf3 dS 2.c4 e6 3.g3 ltJf6 4 . .ig2 ie7 s.o-o 0-0 6.b3
16 . .ixc5
16 . . . bxcS ! ? is an interesting at
tempt to restrain White on the
queenside. The hit on a2 is very
annoying. In Ree-Brandenburg,
Netherlands 2011, White tried the
most logical move 17J'!d7, but the
possibility of 17 . . . ,ixh4 1B.gxh4
ElfdB= rescues Black. White should
play positionally: 17.ltJf3 ElfdB (17 . . .
a s 1B.a4;!;) 1B.ElxdB+ ElxdB 1 9 . .ic3
ltJc6 (19 . . . aS 2 0.a3 ltJc6 21.a4;!;)
2 0 . 'i!lf1 f6 2 1.'i!le2 'i!lt7 2 2 .ltJd2, with
a slightly better pawn structure.
17.a3 c6
I suspect that the position after
17 . . ltJc2 1B.Elac1 lLlxa3 19.ElxcS bxcS
.
2 0 .,ixa3 might be a technical draw.
18.b4
White has some edge since lB . . .
ie7 could b e attacked by IB.ltJfS.
A2 . 9 . . dxc4 10 .bxc4 .ib7
11.e2
A2 1. 11 . . . ElcB ; A2 2 . 11 . . . a6; A23.
11 ... c7; A24. 11 ... d7; A2S. 11 ...
ltJb4
A21. 11 . . . gc8
This move is Black's most popu
lar answer in practice. It aims to
discourage White's active play in
the centre. However, it is a pure
waste of time against a flank attack
with f4.
a) 12 J:!acl l3c7 13.d4 is possible,
but insufficient for an edge due to
13 . . . cxd4 14.exd4 l3d7 1S.Elfdl l3eB !
This is meant to take the sting
out of d4-dS. White has to play this
position fast because once Black has
consolidated, the hanging pawns
may really begin to hang. So Ma
rin's recipe 16.lLlbS a6 17.ltJeS seems
adequate, but he stops his analysis
too early, evaluating White's chan
ces as higher. Unfortunately, forced
play after 17 . . . lLlxeS 1B.dxeS ,ixg2
19.exf6 ,ixf6 20 . .bf6 (20.l3xd7
xd7 21.'i!lxg2 axbS 2 2 .,ixf6 gxf6
23.cxbS dS+ 24.'i!lgl l3aB 2S.Elc4
1B7
 Part 8
hS 2 6.a4 gd8= ) 2 0 . . . gxd1+ 21.gxd1
xf6 2 2 .\!;>xg2 axbS 23.cxbS gd8
24.a4 g6 leads us to a draw:
2S.gxd8+ xd8 26.e4 b8
27.c6 \!;>f8 . Black's king arrives
just in time to stop the passer:
2 8.aS bxaS 29.b6 \!;>e7 30.b7 \!;>d8.
My further efforts failed to make
the plan with d4 work so I had to
turn my attention to the right wing.
The universal set-up from
line A23 is playable : 12.gfd1 c7
13.gab1 gfd8 14.d3 b8 1S.ttle1 ttlaS
16.f4 hg2 17.xg2 ttlc6 18.g4 ttle8
19.ttle4 ttld6 2 0.ttlg3 b7 21.ttlhS f6
(2L .if8 2 2 .ttlf3) 2 2 .gS.
White achieves excellent practi
cal results with :
b) 12 .ttle1 ! ? Then 12 . . . c7? !
13.f4 ttl as leads to this position :
188
The game Malakhov-Bokros,
Mainz 2009, went 14.fS?! .b:g2
lS.ttlxg2 d7! 16.fxe6 (16.g4 gfd8
17.gad1 d3=F) 16 . . . fxe6 17.ttlf4
gcd8 18.gad1 id6 19.d3 ttlc6
2 0 .g2 (20.d4 ! oo) 2 0 ... ieS+. This
example teaches us what we should
NOT do - to push fS too early
when we could not support it with
g4 due to weaknesses along the d
file. Instead, White can exploit the
particular move order of the oppo
nent by 14.ttlbS d7 lS.ttlxa7 gcd8
16.hb7 ttlxb7 17.ttlf3 ttle4 18.d4,
winning a pawn, but at the cost of
the initiative. It is more in the spirit
of the position to prepare fS with
14.gadl ! ? or:
14.g4 .b:g2 lS.ttlxg2 gcd8
16.gad1 (16.gS? ttle8 17.gf3 d7
18.gh3 d3 19.hS g6 2 0 .e2
f6=F) 16 . . .d7 17.d3 ttlc6. (Other
wise Black is risking to get crushed
with his knight at the edge of the
board.) 18.fS.
We have reached a typical po
sition with an initiative for White.
Possible continuations are:
18 . . . ttle8 19.ttle4;!; or 18 . . . exfS
19.9xf5 ttleS (19 . . . gfe8 20.ttle1 id6
2 1.g2 i.eS 2 2 .ttlf3 d6 23.e4;!; ttld7
24.ttlxeS ttldxeS 2S.ttldS) 2 0.e4. The
 l.ltJf3 ds 2.c4 e6 3.g3 ltJf6 4 . .ig2 !i.e7 5.0-0 0-0 6.b3
pawn majority in the centre gives
White a free hand on the kingside.
In this set-up, the black queen
has no business on c7. After 12 .ltJel,
best is probably 12 ... 'I1;Yd7 13.f4 E:fd8.
Here 14.d3 and 14J!adl are
probably of equal worth. White's
chances are preferable. His plan is
to push g4, fs. Be sure to look at the
detailed commentaries of game
16 Markus-Perunovic, Subotica
2008.
My main line 12.E:adl ! ? presents
a clever move order which discour
ages 12 . . . Wld7 due to 13.d4 cxd4
14.exd4 i.b4 ls.ds ltJaS 16.ltJes;t. At
the same time, it leaves the other
rook on the kingside. That could be
crucial for the success of the immi
nent march forward of the f-pawn
after ltJel as I have shown in line b.
12 ltJa5
Another option is 12 . . . Wlc7
13.ltJel and then:
a) 13 ... a6 14.f4 ltJa7 1s.a4
Is.fs ! ? seizes the initiative, but
I'm following the game Sherbakov-
Loginov, St. Petersburg 1998, which
followed the plan with a slower, but
better prepared attack:
Is . . . .ixg2 16.ltJxg2 Wld7 17.g4
ltJe8 18.fs
18 . . . !f6
18 . . . ltJd6 allows the bold sac-
rifice 19.f6 ! ! [White has another
good continuation: 19.1tJf4 bs (19 . . .
ltJc6 20.ltJhs g 6 2 1.ltJg3t) 20.axbs
axbs 21.cxbs ltJaxbs 2 Vt:lxbs ltJxbS
23.f6.] 19 . . . gxf6 (19 . . . .ixf6 20.E:xf6 !
gxf6 21.ttJf4 E:fd8 2 2 .ttJhs ttJe8
23.ttJe4) 2 0 .ttJf4 bs 21.axbs axbs
2 2 .d3 b4 23.ttJe4 ttJxe4 24.dxe4
Wlb7 2s.ttJhs es 26.gs+-.
19.fxe6 fxe6 20.gs ! ! .ixgs
21.E:xf8+ c.t>xf8 2 2 .ttJe4 6 23.as
(or 23.E:f1+ c.t>g8 24.Wlg4+- Wle7
2s.ttJf6+ ttJxf6 2 6 .E:xf6 g6 27.E:xe6)
23 . . .bxas 24.ttJxc5 and Black re
signed.
b) 13 . . . E:fd8 14.f4 a6 Is.fs (15.
g4 ! ?) IS . . . Wld6 16.d3 E:b8 17.!al
ttJes 18 ..txb7 E:xb7 19.ttJf3
Black remained coordinated in
the centre, but he has not created
any counterplay. The game might
continue:
189
 Part 8
19 . . . exf5 (or 19 . . . ttJxf3+20 .Wlxf3
l'!bd7 2 1.fxe6 Wlxe6 2 2 . e4;!;) 20.e4 ! !
fxe4 (20 . . . ttJxf3 + 2 1.Wlxf3 Wld7
2 2 . e5 ttJg4 23.ttJd5 ttJh6 24 . .ic3 b5
25 . .id2t) 21.dxe4 Wle6 2 2 . ttJxe5
Wlxe5 23.ttJd5 Wle6 24.ixf6, with an
overwhelming advantage.
c) 13 . . . ttJa5 14.f4 .bg2 15.ttJxg2
l'!cdB I6.g4
We know already that f4-f5
should be well prepared. For in
stance, 16.f5 Wld7 17.ttJf4 exf5
IB.ttJcd5 ttJxd5 19.ttJxd5 ttJc6 20.l'!f2
ttJb4 ! 2 1 .l'!xf5 is only equal.
16 . . . b7 17.d3 a6 IB.f5 (IB .g5
ttJeB 19.1'!f3 g6 20.l'!h3 Wld7oo) IB . . .
ttJ e B 19.ttJf4 Wlc6
This position occurred in Supa
tashvili-Mkrtchian, Batumi 2003.
White has achieved everything
190
he was aiming to. He should have
switched now to concrete calcula
tions :
2 0 .fxe6 ! fxe6 21.ttJh5--+ when
2 1 . . .b5 2 2 .ttJe4 bxc4 loses to
23.l'!xfB + ! ixfB 24.l'!f1+- Wlb7
25.Wlf2 Wle7 26.ttJxc5 cxd3 27.ttJxe6.
13.d3 ! ? is a consistent alterna
tive. I analyse it in game 18 A.So
kolov-Thesing, Berlin 1993.
The text frees the lane to the f
pawn. See for more details game
17 Kharlov-Kosyrev, Samara
2 0 . 06.2000
A22. 11 a6
.
12.gfdl!
The threat of . . .b5 practically
rules out the plan with l'!adl. This
does not mean we renounce the
idea of a kingside pawn storm. We
are simply delaying it until the b5-
square is under firm control.
 l.lLlf3 dS 2.c4 e6 3.g3 lLlf6 4 . .ig2 .ie7 S . O-O 0-0 6.b3
Marin advocates 12 .d4 cxd4
13.exd4 c8 14.lLld1 hoping for 14 . . .
lLlaS? ! 1S.lLle3 Wic7 16.ac1 fd8
17.dS. This manoeuvre has two
drawbacks, though. Firstly, I do not
like much White's position after 14 . . .
bS ! ? More importantly, Black can
deny altogether the knight's transfer
to e3 by not exchanging on d4:
12 . . . lLlaS ! ? when White has not
anything better than 13.dxc5 bxcS
14.e4 lLld7oo 15.eS lLlb6. In the game
Brunello-Antonsen, Odense 2011,
White went on to lose a pawn after
16.ad1 Wic7 17.lLle4? he4 18.xe4
lLlaxc4't.
12 . Wc7 13.gabl! gab8 14 . .ial
The slogan of our opening cam
paign is prophylaxis ! We keep the
opponent at bay and we are still
hoping for d2-d4 at an opportuni
ty. For example, the thematic 14 . . .
lLla7 would stumble into 15.d4 ! bS
(1S . . . cxd4 16 .exd4 b5 17.dS Wixc4
18 .Wixc4 bxc4 19 .d6 was gloom
for Black in Todorovic-Cvetkovic,
Belgrade 2 007) 16.dS ! . In practice,
Black usually opts for the flexible :
1 4 . . . gfd8 15.d3 .la8
1S . . . lLla7 is still inaccurate due
to the same 16.d4 cxd4 17.exd4
The next moves will unleash the
power of the hanging pawns or, to
be precise, of the active long-range
pieces behind them. Practice has
seen further:
a) 17 . . . bS 18.dS! Wixc4 (18 . . .
exd5 19.1Llxd5 lLlxd5 2 0 . cxd5 xd5
2 Ul:e1 ! gave White many threats
in Kortschnoj-Kir.Georgiev, Lu
gano 1986.) 19.Wixc4 bxc4 20.dxe6
lLlbS (20 . . . fxe6 21.xd8+ hd8
2 2 .lLlgS lLlb5 23.a4 hg2 24.It>xg2
lLld7 25.lLlxe6 lLlxc3 26.E1xb8 lLlxb8
27.hc3;!;) 21.exf7+ 'i!?f8 2 2 . E1xd8+
E1xd8 23.lLle5, Espig-Gruenberg,
Leipzig 1980.
b) 17 ... E1bc8 18.Wib2 ! ! (setting up
an ambush ! ) 18 . . .b5 (18 . . . lLld7 19.c5
bS 2 0 .a4) 19.d5 Wixc4 20.dxe6 Wixe6
21.lLlgS Wib6 2 2 .lLld5 ! .
16.lLlel
This retreat is commonly con
nected with f2-f4, but it is also in
dispensable for my idea of advan
cing the g-pawn.
191
 Part 8
16 . . . a7
16 . . . llJe8 17.f4 was in game 19
Zaichik-TIviakov, Moscow 1994.
17.haS gxaS
Common continuation here
is 18.f4 l3ab8 19.a4, with double
edged play, for example, 19 . . . c6
(19 .. .'c6 2 0 .llJf3) 2 0.g4 llJb4 (20 . . .
llJ e 8 2 1.llJf3 llJd6 2 2J1 d 2 .if8 23.g5
llJf5 24.llJe4 Wb7 25.llJg3) 2 1.g5 llJe8
2 2 .llJf3 ( 2 2 . llJe4 llJd6 ! ) 2 2 ... llJd6
23 .llJe5 f6 24.gxf6 hf6.
These variations show that one
tempo could be decisive for the suc
cess of White's attack so I decided
to save a4 and began analysing
19.94 b5 2 0 .g5 llJe8 21.llJe4
192
Our plan is to open the main
diagonal with h2-h4-h5-h6 and
build the battery i.a1-Wb2. How
ever, Black can discharge part of
the tension with 2 1 . . .llJd6 22 . .ie5
llJc6 when 23.llJxc5 does not win
a pawn due to 23 . . . Wa7 24.ixd6
hd6 25.llJe4 i.xf4 = . This fine point
made me think that f2-f4 may be
not obligatory at all. Let us check:
IS.g4!
This natural move is a novelty.
Indeed, why should White weak
en his kingside with f4 when his
main plan is to push the h-pawn? !
Black cannot prevent g 5 by 18 . . .
llJ e 8 19.1lJe4 llJd6, because 2 0 .llJxd6
hd6 21. Wb2 is a little unexpected
double hit.
IS . . . gabS 19.95 llJeS 20:g4
d6
2 0 . . . b5 does not distract White
from the right wing: 21.llJe4 bxc4
2 2 .dxc4 l3xb1 23.l3xb1 llJc6 24.h4t.
21.h4 b5 22.h5
 l.ltJf3 d5 2.c4 e6 3.g3 ltJf6 4.i.g2 i.e7 5.0-0 0-0 6.b3
White has an obvious initiative.
The game may continue with 2 2 . . .
ltJc6 23.ltJf3 bxc4 24.gxbS '!WxbS
25J1b1 '!Wc7 26.dxc4 .ifS 27.h6.
A23. 1l ti'c7
A flexible move. It effectively
takes the sting out of the plan with
d2-d4 because 12.gac1 a6 13.d4
could be answered with 13 . . . ltJaS.
a) 12.ltJbS looks like a loss of
tempo since after 12 . . .'!WcS 13.gfd1
a6 14.ltJc3 ltJaS the queen is not
worse on cS than on c7:
lS.gab1 gbS 16 . .ia1 i.c6 17.e4
'!Wc7
lS.ltJeS is not dangerous due to
lS . . . ltJd7 19.1tJxc6 ltJxc6 2 0 .f4 ltJd4
21.'!Wfl ltJc2 2 2 .e5 b5f! .
The game Wang Yue-Hou Yifan
Danzhou 2011, saw:
lS.ge1 ltJd7 19.e5, when 19 . . .
gfdS would have kept the tension
in view of the threat of . . . b6-bS.
b) 12.ltJel ! ? is a risky, but en
terprising plan. Compared to the
position after 11.. .gcS, Black is
practically a tempo up. He can de
velop the rook on dS or keep it on
the queenside to support the break
through . . . b6-bS. Look at game 2 0
Pigusov-Zontakh, Sochi 2007.
Apart from 12.ltJe1 ! ? , White has
also the same plan as in line B 2 :
12.gfdl a6
Play has transposed to line B 2 .
A24. 1l ti'd7
This would be the ideal place
for the queen - it controls e6 and
f5 thus discouraging White's plan
with f2-f4-fS. Moreover, it puts
pressure on the d-file and supports
the breakthrough . . . b6-bS. How
ever, all this is in conditional mood
since tactics works in White's fa
vour. Black stumbles into difficul
ties after:
12 ... gadS is plain bad. Vega
Holm-Magem Badals, Cala Gal
dana 1999, went on with 13.d4 cxd4
14.exd4 ltJaS 15.ltJe5 '!WcS 16.d5 gfeS
193
 Part B

(16 . . . exdS 17.lLJxdS lLJxdS 1B.cxdS) .icS 29.lLJa6 .id4 30.ga2 gfB 31.'i!fg2
17 . .ih3 .icS, when 1B.lLJbS ! a6 (lB . . . gf7 32 .gd6.
bB 19.1LJxf7) 19.1LJd4 ! (Marin)
would have given him an edge.
After the text, things are far from
clear:

13.d4 cxd4 14.exd4 .if8 !

The e6-pawn is vulnerable.

A resilient defence, which Marin
does not consider. 14 . . . lLJaS 1S.dS ! is
fine for White, beyond doubt.
In the diagram position, lS.dS
has no venom due to lS . . . exdS
16.lLJxdS lLJxdS 17.cxdS e7! so I
took my time to analyse lS.lLJeS
eB 16.lLJbS gdcB 17 ..ia3 ha3
White is ready to bring a second
hit upon it with Ek6, lLJc7. The d4-
bishop is also hanging so it should
leave its active stand in the centre.
Here is an illustration of how jumpy
the white knight could be: 32 . . . .if6
33.gc6 ge7 34.lLJc7 'i!ff7 3S.aS bxaS
36.gxaS ge2 37.lLJbS 'i!fg6 3B.ga4
gb7 39.lLJd6 gbb2 40.lLJe4 gh6
41.gcB.
A25. 1l b4 12.d4 cxd4
. .

1B.lLJxa3 lLJaS and lS.gael e7 (lS . . . 13.exd4 gc8 14.!Udl

lLJaS 16.dS) 16.lLJbS a 6 1 7. .ia3 lLJb4
1B.lLJc3 gacB 19.9b1 as 2 0 .lLJbS.
Although White has some space
advantage, play remains complex.
Perhaps White should take a small
edge in the endgame following:
lS.a4 lLJaS 16.lLJeS eB 17.hb7
lLJxb7 1B.lLJbS gdcB 19.f3 lLJaS
2 0 .dS lLJxc4 2 1 .lLJxc4 gxc4 2 2 .hf6
gxf6 23.dxe6 (23.xf6 a6 ! ) 23 . . .
fxe6 24.xf6 g6 2S.f3 e4
26.xe4 gxe4 27.lLJc7 gcB 2B .gd7

194
 l.lLlf3 dS 2.c4 e6 3.g3 lLlf6 4 . .ig2 .ie7 S.O-O 0-0 6.b3
14 . ge8
Black has to do something about
the threat of d4-dS. Kharitonov
Dvoirys, Hoogeveen 2000, saw
instead 14 . . . c7 1S.a3 lLla6 16.dS
exdS, when the forced variation
17.lLlbS xc4 18.xe7 xbS 19.hf6
Elc7 2 0.d6 gxf6 21.lLld4 d7
2 2 .xf6 e7 23.f4 gives White a
nice attacking position.
15.a3 lLlc6 16.d5 ! ?
17.lLlxd5 xd5
I have been following the game
A.Petrosian-G. Kuzmin, Baku 1977.
Here, 18.ElxdS ! c7 19.Ele1 would
have bound Black up and down.
The only sensible answer 19 . . . lLlaS
2 0 .EleS lLlxc4 21.Elxe7 Elxe7 2 2 .xe7
xe7 23.Elxe7 hf3 24.hf3 lLlxb2
2S.Elxa7 g6 26 . .idS Elf8 27.Elb7 lLla4
28 . .ic4 ! leads to an endgame where
Black has no useful moves : 28 . . . 'it>g7
29.h4 h6 30.'it>g2 'it>g8 31. 'it>f3 'it>g7
32.'it>e3, going to the queenside.
A3. 9 .ia6
This development of the bishop
is very tricky. Black wants to deny
us active play in the centre. One
mundane move as 1O .d3? ! (Ma
rin's recommendation), and White
should forget about any opening
advantage. Black will get time for
bringing the rooks in the centre and
his pawns would be even more ac
tive than ours.
Kramnik's approach against
Giri, Monaco 2011, makes more
exd5 sense:
1O .e2 d7 1l.Elfd1?! Elac8
12 .d4 cxd4 13.exd4 Elfd8 14.lLleS
e8 1S.f3 .ib7 16.f4 = .
Chiburdanidze-Kalevic, Belgrade
1992, introduced the missing ele
ment: 1O .e2 d7 1l.Elfd1?! Elac8
12 .lLlbS ! Elfd8 13.d4 cxd4 14.exd4
e8 1S.Elac1 .ib7 16.lLleS a6 17.lLlxc6
Elxc6 18.lLlc3. However, this is still
about equal.
Our only chance is an energetic
assault in the centre with d4, lLleS,
lLlbS. In this scenario, Elfd1 is un
necessary and only gives Black's
queen time to retreat comfortably
to e8. So we should play concrete
chess :
10 .e2 d7 11.d4!
This is my two cents in the theo
ry of this variation. I was amazed to
find that this move was a novelty.
Now Black faces some problems:
1l cxd4 12.exd4 gfd8
12 . . . Elac8 13.lLleS is awkward
as the queen has not a safe haven:
19S
 Part S
13 . . .'c7 14.tilbS ! hbS lS.cxbS
tilxeS 16.dxeS tild7 17J"!ac1 b7
lS.a3 ! 13xcl 19.13xc1 13cS 2 0 .e3 13c7
2 1.b4;!; cS
2 2 .13e 1 ! 13c2 23 . .id4. White can
easily control the invasion squares
along the c-file with his bishop pair.
Black will have to passively wait for
White's attack with f4-fS.
13.e5 %YeS 14.b5 !
The point of my idea. Now 14 . . .
hbS (Or 1 4. . . 13acS lS.tilxc6 13xc6
16.tilxa7 13xc4 17.bxc4 hc4 18.d2
a4 19.13fc1 xa7 2 0.a4;!;, lS.13ac1 !
may be even better.) lS.cxbS tilxeS
16.dxeS tild7 17.a3;!;, followed by b4,
gives White full control of the board.
Later he will continue with f4-fS.
196
B. 6" .b6 7 .ib2 .ib7 S.e3
bd7
S . . . cS 9.tilc3 transposes to line A
9.c3 !
9.e2 i s more popular because
it makes 9 . . . tile4 pointless. Indeed,
after 9 . . . cS 1O.cxdS tilxdS 11.13fd1 !
if6 12 .d4 cxd4 13.tilxd4, White
would have a small edge.
However, I have no idea how to
meet 9 . . . aS ! . The game Tikkanen
Grandelius, Vasteras 2011, went
10 .d3 a4 1l.tilbd2 cS 12.13ae1 axb3
13.axb3 13a2 = 14.tilb1? dxc4 lS.bxc4
.ixf3=F. 9.tilc3 is more natural, but
9 . . . a4 ! 1l.tilxa4 dxc4 12.13fc1 bS !
13.tilc3 13bS 14.bxc4 bxc4 gives
Black active play, for example:
lS.13cb1 hf3 16.xf3 tileS 17.d1
tild3 lS.c2 tilb4 19.d1 tild3 = .
After 9 .tilc3, m y main line
branches to:
B1. 9 . . . tile4; B 2 . 9 . . . cS.
9 . . . aS is not too clear here, be
cause 10 .d3 puts the question to
Black's last move.
 l.lLlf3 d5 2 .c4 e6 3.g3 lLlf6 4 . .ig2 .ie7 5.0-0 0-0 6.b3

Bl. 9 . . . c!tle4 10.tve2 !

Marin's repertoire i s based on:

10.c!tle2, but I can not see even a
trace of an edge. Let us check fur
ther:
1O . . . .if6 11.'lWc2
11 . . . dxc4 12 .bxc4 c5 (12 . . . lLld6)
13.lLlf4 lLld6 14.tvb3 cxd4 15.exd4
j':1cB 16.j':1fc1 tvc7 was better for
Black in Dominguez-Lysyj , Khanty
Mansiysk 2011. 12 . . hd5
11 ... hb2 12.tvxb2 'lWf6 13.d4 The symmetry after 12 . . . hc3
dxc4 14.bxc4 lLld6 1S.lLld2 hg2 13.dxc3 hd5 14.j':1ad1 'lWe7 15.lLld4
16. <;!;>xg2 cS 17.a4 is a bit more pleasant for White
since the active knight on d4 has
access to c6. This should ensure
control over the d-file. On the other
hand, 13 . . . exd5 14.j':1ad1 'lWe7 1S.lLld4
lLlf6 16.c4 c5 17.lLlf5 'lWe5 1B .g4 gives
White an initiative on the kingside.

13.d4 i.e4 14.ac1 tve7

White was slightly better in Diz

Marin ends his analysis here
with the assessment "", but this
is obviously an example of wishful
thinking. 17 . . . cxd4 1B.exd4 eS 19.aS
lLlxc4 2 0 .lLlxc4 'lWc6+ 2 1 . <;!;>g1 'lWxc4
is a dead draw.
10 i.f6 1l.cxd5 c!tlxc3
dar-Markus, Sibenik 2007, due to the
weak light squares on the queenside.
This could be underlined with:
15.c!tld2 ! hg2 16. <;!;>xg2 c5
17.c!tle4 cxd4 18.exd4 ac8
19.'lWb5

No one wished to try 11 . . . exdS

as 12 .d4 leads to a typical pawn
structure where White maintains
lasting pressure along the c-file.

12 .hc3

197
 Part 8

White is menacing .ib4. All his Or 11 ... .if6 12.cxd5 exd5 13.d4
pieces are active. e7 14.ac1 fd8 15.dxc5 ltJdxc5
16.c!ild4;!;.
12 .hc3 .if6 13.cxd5 hd5 14.d4
B2. 9 c5 9.c!ilc3 c!ilbd7 cxd4 15.ltJxd4 hg2 16.<;t>xg2 c8
17.acI.
9 . . . ltJc6 10.cxd5 is considered in
line AI.

1 0 .ti'e2 k8

Alternatives are :
a) 10 . . . ti'c7 1U!ac1 c8 (11 . . .
ad8 12.fd1) transposes t o the
main line. Perhaps White does
not even need ac1 and should try
White is slightly better and
11.ltJh4 immediately.
went on to win in Dizdar-Naum
kin, Voskresensk 1990. Black's
b) 1O . . . ltJe4 11.fd1 !
queenside pawns and the c6-square
11.cxd5 ltJxc3 12 .hc3 hd5 are sensitive.
13.e4 .ib7 14.fd1 ti'c7 15.ac1 ad8
16.d4 b8 17.d5 exd5 18.exd5 .if6 lUac1
19.1tJh4 is an important line.
11 .ltJh4 at once also deserves
consideration.

1l ti'c7

Black used to struggle in such

positions because White's knight
proved to be powerful on f5. How
ever, the game Poettinger-Maren
tini, Triesen 2011, showed a simple
defence: 19 . . . g6 ! 2 0.d2 d6+.
11 . . . ltJxc3 12.c!ilh4!?

198
 l.lLlf3 d5 2.c4 e6 3 .g3 lLlf6 4 . .ig2 .ie7 5.0-0 0-0 6.b3
In my opinion, this brain
child of Malakhov is the only real
way for White to set problems to
Black. 12.cxd5 lLlxd5 13.lLlxd5 hd5
14.e4 .ib7 15.d4 '\1;lfb8 16J'fd1 cxd4
17.lLlxd4 gxc1 18 .hc1 gd8 19 . .if4
lLle5 2 0 .b4 is totally equal. Marin
claims an advantage for White, but
after 20 . . . '\1;lfd6 21.gd2 '\1;lfc7 2 2 .b5
.if6 23.lLlc6 hc6 24.gxd8+ '\1;lfxd8
25.,he5 it is high time to sign a
draw.
12 . . . dxc4
If Black waits too long with this
exchange, we could anticipate it
by taking on d5 and c5. My game
Delchev-Cornilius, Benasque 2011,
reached this position:
16 . . . '\1;lfa8 17.lLlfS .if8 , when the
active knight on fS enabled the
combination 18.lLlxd5 ! lLlxd5 (or
18 . . . hd5 19.9xd5 ! lLlxd5 20.lLlh6+
@h8 21.'\1;lfh5 lLl7f6 2 2 .hf6 lLlxf6
23 .'\1;lffS '\1;lfb8 24.lLlxf7+ @g8 25.e6)
19.1Llh6+ @h8 20.'\1;lfh5 lLl7f6 2 1..ixf6
gxf6 ( 2 1 . . .lLlxf6 2 2 .'\1;lfxt7 gxh6
23.hb7) 2 2 . .ie4 1-0.
13 .bxc4 .b:g2 14.xg2
White has obtained a good ver
sion of the pawn structure which I
have considered in line A. The rook
is useless on c8, but that is balanced
by the equally misplaced c1-rook.
The d7-knight should loose time to
reach its best square c6. In Mala
khov-Potkin, Moscow 2010, Black
tried to save such a redeployment
with:
14 .'llY c6
.
I see here at least two good ways
to expand on the kingside : 15.e4
gcd8 16.e5 ttJe8 17.ttJe4 f6 18.exf6
ttJexf6 19.ttJg5 e5 2 0 .he5 or the
standard 15.d3, followed by f4,g4,
f5.
The stem game went instead
15.f4 ttJe4 when 16.ttJxe4 '\1;lfxe4
17.d3 '\1;lfc6 18.e4 gcd8 19.1Lle3 .if6
20 .e5 .ie7 2 1.f5 would have also
seized the initiative.
199
 Part 8
Pa rt S
Complete Games
1 5. Sherbakov-G rigoriants
M oscow 1 999
1 .c4 e6 2.<f3 f6 3.g3 d5 4.ig2
ie7 5.0-0 0-0 6.b3 c5 7.ib2 c6
S.e3 d4 9.exd4 cxd4 1 0J:e1 ic5
This is the best place for the
bishop. It not only protects the d4-
pawn, but also bars the fifth rank
to White's rook so it cannot reach
bS. However, Black's move order is
probably not accurate. He should
play 1O .. J'l:e8 first in order to avoid
the exchanges after 11.eS ! xeS
12J!xeS vgd6 13J'l:xcS vgxcS 14.i.a3.
It is generally beneficial to White to
trade these two minor pieces. In the
game, White still has to work hard
to achieve them.
1 1 .d 3 13eS
1 2 .ia3 !
200
White often plays here 12.llJ eS
(or 12 .a3 as 13.eS). After 12 . . .
lLlxeS 13.13xeS vgc7? ! 14.f4 i.d6
1S.hd4 ! gives White compensation
for the exchange although the posi
tion remains very complex:
1S . . . heS 16.heS vge7 17.lLlc3
lLld7 18 .d4 lLlxeS 19.fxeS 13f8
2 0 .lLlbS f6 2 1.vge2 gb8 2 2 . lLld6.
The same sacrifice is possible with
the insertion of 12.a3 as and then
13.lLleS lLlxeS 14.gxeS vgc7? ! 1S.f4
i.d6 16.hd4 heS 17.heS vge7
18.lLlc3 lLld7, but here best is 19 .id4
(instead of 19 .d4 lLlxeS 2 0 .fxeS
gf8 21.lLlbS f6 22 .lLld6 vgc7 23 .vge2
vgb6oo) 19 . . . gb8 2 0 .vgg4 fS 2 1.vge2
b6 2 2 . lLlbS i.b7 23.hb7 gxb7
24.i.c3 lLlf6 2S.ge1, with pressure.
These problems of Black have
a simple solution, though : instead
of 13 . . . vgc7, Black should play 13 . . .
 l.ttlf3 d5 2.c4 e6 3 .g3 ttlf6 4.i.g2 i.e7 5.0-0 0-0 6 .b3
Wld6 ! , achieving comfortable equal
ity. For example: 14.f4 bB 15.a3
(or 15.ttld2 i.d7 16.Wle2 i.c6 17.f1
hg2 IB.Wlxg2 i.b4 19.ttle4 ttlxe4
2 0 . xe4=) 15 . . . a5 16.b4 axb4
17.axb4 hb4 IB.ttla3 .id7 19.ttlb5
hb5 2 0 . xb5 b6.
1 2 . . ..ixa3
Black has also tried:
a) 12 ... .id6 13.hd6 Wlxd6 14.a3
a5 15.ttlbd2 h6 16.Wlc2 e5 17.cS
White has a clear plan to ad
vance his queenside pawns. Dam
ljanovic-Nikcevic, Igalo 1994, saw
further 17 . . .WlfB IB.ttle4 ttlxe4
19.xe4 .if5 20.eel .ih7 2 1 .ttld2
'it>hB 2 2 .abl f5 23.b4 axb4 24.axb4
a3 25.ttlc4.
b) 12 . . . ttld7 13.ttlfd2 ! ha3
14.ttlxa3 ttlc5 15.ttle4 ttlxe4 16.he4
f5 17 . .ig2 e5 IB.Wld2
White is far ahead with his
queenside plan and went on to win
in Obukhov-Kolomensky, Nizh
nij Tagil 2007: IB . . . Wle7 19.ttlc2 a5
2 0 .f4 Wld6 2 1.fxe5 xe5 2 2 . xe5
Wlxe5 23.el 'IMlf6 24.Wlf4 h6 25.Wlc7
'it>h7 26.eB Wlg5 27 . .id5 ttlb4
2B . .igB+ 'it>hB 29.ttlel .ie6 30.xaB
Wle3+ 31.'it>fl hgB 32 .'IMlf7 1-0.
1 3.xa3 eS 1 4.c2 as 1 S.a3
.ifS 1 6. h4 .ig4 1 7 .trd2 h6
A critical position. White has
good prospects on the left wing,
but he should think about prophy
lactics, too. IB .h3 ! .id7 19.b4 axb4
20.axb4 xal 21.xal would have
been the right course to take since
IB.b4, as played in the game, misses
IB . . . axb4 19.axb4 xal 2 0 . xal e4 !
21.dxe4 d3 2 2 . ttle3 .ie2 23.f3 g6
with double-edged play.
1 8 . b4 gS 1 9 . bS b8 2 0 . h 3 .ic8
2 1 . f3 trd6 22.h4 ( 2 2 .ttlxg5 ! )
22 . . . g4 23. fxd4 (23.ttlxe5 ! +-)
23 ... exd4 24.trxh6 :1xe1 + 2SJxe 1
bd7 26.trg S+ q"h7
Now 27.ttlxd4 wins at once, but
quite amazingly, Sherbakov some
how fails to convert his advantage.
201
 Part S
27 . .ie4+ xe4 2S J::xe4
Wg6 29 .WdS b6 30.xd4 e5
3 1 J:U4?? (3UeS ! +-) xd3
32J:U6 .ie6 33.WxaS Wxf6 34.We4+
Wg6 35.xe6 fxe6 36.Wb7+
Wg7 37.We4+ Wg6 3S.Wb7+ Wg7
39.We4+ Wg6 YZ-YZ
1 6. M a rkus-Peru novie
S u botiea 02.09 .200S
1 . f3 f6 2 .e4 e6 3.g3 d 5 4.b3
.ie7 5 ..ig2 0-0 6.0-0 e5 7 . .ib2
e6 S.e3 b6 9 . e3 dxe4 1 0 . bxe4
.ib7 1 1 .Yfe2 geS
Black anticipates the plan with
d2-d4, but against the other dan
gerous plan - f2-f4, his rook is even
worse than on its initial square.
From cS it does not support the
only possible counterplay based on
. . . a6, . . . b6-bS.
202
1 2 . e 1 ! ? (12J%adl ! ?) 12 . . .Wd7
1 3.f4 gfdS 1 4.d3
14J!d1 is a valuable alterna
tive. In some variations White may
need the additional control over bS
which e2 provides.
The game Panchenko-Kucera,
Pardubice 1992 went on with:
14 . . . ttJeS. This is a thematic ma
noeuvre. From d6 the knight will be
eyeing fS and supporting the break
through . . . bS.
1S.g4
The flexible 1S.d3 also deserves
attention. It may save g4, as in
the variation 1S . . . ttJb4 (1S . . . ttJd6
16.ttJf3) 16.ttJf3 ttJd6 17.ttJeS c7
1S.e4 ttJc6 19.ttJxc6 ixc6 2 0 .fSt.
1S . . . ttJaS
1S . . . if6 looks safer: 16.d3 (16.
gS i.xc3 17.i.xc3 ttJd6oo) 16 ... ttJd6
17.ttJf3.
16.d3 i.xg2 17.xg2 a6
White chose here 1S.a4 .if6
19J!b1 gbS 2 0 .gS i.xc3 2 1.i.xc3
xa4 when 2 2 .gf3 d7 23.gh3
would have given him a nice attack
ing position. However, he could
 1.lt:lf3 d5 2.c4 e6 3 .g3 lt:lf6 4 . .ig2 i.e7 5.0-0 0-0 6.b3

have saved the prophylaxis in fa

vour of the straightforward IB.lt:lf3
It:ld6 19.f5 b5 20.e4 ! , for instance :
2 0 . . . bxc4 21.e5 It:leB 2 2 .f6 i.fB
23.fxg7 hg7 24.lt:lg5.

1 4 . . . b4 (or 15 . . . lt:ld6 16.lt:lf3)

1 5J: d 1

White has good control of the

1 5 . . a6
. 25.f5 c6 26.lt:le4;!; h6 (26 ... cd7
It is curious to investigate :
15 . . . hg2 .
The d7 suggests that White's
knight would be more active on f3
rather than on g2 so we recapture:
16.xg2 a6 17.a3 lt:lc6 IB.lt:lf3
IB.g4 looks premature due to
IB ... b5 ! 19.95 It:leB 20.cxb5 axb5
2 1 .lt:lxb5 bB 2 2 . a4 lt:ld6.
IB . . . b7 19.a4 !
centre and a clear plan on the king
side with e3-e4-e5, It:le4, f4-fS. The
mere fact that Houdini 2 at depth
22 gives as a first line 20 . . . lt:lg4
21.h3 It:lf6 reveals the truth - Black
has no counterplay. White can fol
low 2 2 .g4 i.fB 23.e4 c7 24.e5 It:leB
27.fxe6 fxe6 2B.lt:lfg5+-) 27.d2
when 27 . . . xa4 loses to 2B.fxe6
fxe6 29.lt:lfg5 ! bxg5 30.f3+-. Let
us examine now a more sensible
retort:
20 . . . lt:leB 2 1.g4 It:ld6 2 2 .e4 d7
which enables the manoeuvre
. . . It:ld6-b7-a5-c6.
23.f5 It:lb7 24.g5

Again, White has n o reason to

give the opponent counterplay. Af
ter 19.94 b5 2 0 .fS bxc4 2 1.g5 It:ld5
2 2 . lt:lxd5 xd5 23.f6 .id6 24.dxc4
xdl 25.xdl dB the game is ba
lanced.
Now 24 . . . exf5 25.exf5 It:ld6
19 . . . lt:lb4 2 0 .e2
26.lt:le5 or 24 . . . eB 25.f6 i.fB
It is funny to watch the engines 2 6.fxg7 hg7 27.e5 are clearly
preferring Black here. gloom so Black follows up his plan:

2 03
 Part S

24 . . . ttJaS 2S.g6! fxg6 2 6.fxg6

hxg6 27.g2 gS 2S .ttJe2 ttJxc4

20 . . . exf5

Perunovic could have conced

The battle has suddenly heat
ed a clear advantage after 20 . . . eS
ed. Both sides have fulfilled their
2 1.ttJf3 ttJc6 2 2 .g2 , but he prefers
plans and it is time for calculations.
to open the centre.
29.i.xg7 is probably not winning
although it is a close call for Black.
2 1 Jlxf5 til bd7?1 2 2 . til d 5 1
The kill is a little unexpected:
29.dxc4 ! ! xdl 30.g4 ! d6 Game over - 2 2 . . . ttJxdS 23 .hS.
31..ieS d7 32 .ttJf4+-, threatening
ttJg6, hS. 22 .. J'eS 23Jdf1 13e6 24.e4 .idS

1 6.tilf3 VNe7 1 7 .a3

25.tilxf7 i>xf7 26.VNh5+ i>gS

27.tilxf6+ 1 -0
It is time to take stock. White has
regrouped successfully, his pieces
are ready for a kingside assault. On 1 7. Kharlov-Kosyrev
the contrary, Black's counterplay is Samara 20.06.2000
nowhere to be seen. No wonder the
game did not last long. 1 .e4 e6 2.g3 d5 3.J.g2 tilf6 4.tilf3
J.e7 5.0-0 0-0 6.b3 b6 7 . .ib2 J.b7
1 7 . . . tile6 1 S . til g 5 til bS 1 9 . .ixb7 S.e3 e5 9 .VNe2 tile6 1 0.tile3 dxe4
VNxb7 20.f5 1 1 . bxe4 geS 1 2 .gad 1

2 04
 l.lLlf3 d5 2.c4 e6 3.g3 lLlf6 4.i.g2 !le7 5.0-0 0-0 6.b3

pleasant. That is why Kosyrev tries

to distract his opponent from the
right wing, at least for a while:

1 2. . .a5 1 3.e1

This i s the common continu

ation of course, but I think that
White should persist with the flex 21 . . . b51? 22.axb5 axb5 23.cxb5
ible approach by 13.d3 ! ? as in the c4 24.d4 d6 25.g5 g 6 26.f3
next commented game. xb5 27.e5 gc7

1 3 . . . .bg 2 1 4.cbxg2 gc7 1 5.d3

BaS+ 1 6.cbg1 gdS 1 7.f4 a6

2S.Bf2 ? 1

Here White missed the hit

1 S .g4 2B.lDxt7! ! xt7 2 9 .l::( x h7+ gB
White chooses to attack on the 30.gh6 .ifB 31.gxg6+ ig7 32.gxe6.
h-file. In my opinion, 1B.f5 ! ? was a
perfectly sound alternative. 2S . . . xc3 29 . .i.xc3 gd5

1 S . . . Bb7 1 9 .a4 eS 20.gf3 29 ... lLlc6 allows the same tactics

gcd7 2 1 .g h3 - 30.lDxt7 ! .

It becomes evident that Black is 30.e4 g b 5 3 1 .Be 1

behind with his counterplay. If he
tries to support the breakthrough Kharlov is hesitant again. After
. . . b6-b5 with 21. .. lDd6, then 2 2 .g5 3l.f5 ! exf5 3 2 .exf5 hg5 33 .fxg6
g6 23.'<lMg4 would be extremely un- hxg6 34.d5 ! gxd5 35.ge1+-,

205
 Part 8
all the files and diagonals are
gaping wide open for the invading
white army.
31 . . . b6 32.<;i;>h 1 (32.@g2 !
E1b2 + 33.@hl ! + -) 32 . . Jb7 33.d5
:gb1 34.dxe6
It was better to keep the d-pawn:
34.@g2 ! + -.
34 . . . fxe6 35.'it>g2 b3 36.:gxb1
c2+ 37 .d2
37 . . .xe4+?! 3S.:gf3 xb 1
39.d4?
39.ltJg4 ! retained the attack.
This move would have been inef
fective had Black taken on bl two
moves earlier due to 39 . . . Wxe4+ .
After White's mistake, Black has
plenty of counterplay to keep the
balance.
206
39 . . . ltJ b 3 40 .xc4 ltJc5 41 . lLl d 3
c2+ 42.lLlf2 :g d 7 43.:ge3 E1 d 2
44.f1 :gd5 45.b5 rJif7 46 .c6
:gd3 47 .f3 :gxe3 4S .xe3 lLl e4
49 .xe4 xc3 50.lLlg4 .id6
5 1 .b7+ iJ.c7 52.lLle5+ rJig7 53. lLlf3
c4 54.lLld4 xd4 55.xc7+ 'it>gS
56.e5 xe5 57 .fxe5 rJif7 5S.rJif3
rJie7 59.'it>e4 %-%
1 S . A.Sokolov-Thes i ng
Berl i n 1 993
1 .c4 e6 2.lLlf3 d5 3.e3 lLlf6 4.b3
c5 5.iJ.b2 lLlc6 6.g3 iJ.e7 7.iJ.g2 0-0
S .O-O b6 9 . lLl c3 dxc4 1 0 . bxc4 .ib7
1 1 .e2 :gcS 1 2 .:gad1 lLla5 1 3.d3
The idea of this move is to see
where Black's queen is going. For
example, after 13 . . . Wd7, White
could gain a tempo by 14.ltJe5 Wc7
15.f4, or even change plans by opt
ing for 14.e4 E1fd8 15.e5 ltJe8 16.ltJe4.
1 3 . . . lLld7
This game is interesting mostly
because it is a rare example where
White gladly trades queens and
still retains an attack. Another par
ticularity is the pawn on d3 which
 l.ltJf3 d5 2 .c4 e6 3 .g3 ltJf6 4.i.g2 i.e7 5.0-0 0-0 6.b3
allows ltJd2 instead of the common
ltJel.
1 4. lLl d 2 .ixg2 1 5.xg2 Vll c 7
1 S .f4 as 1 7 .a4
I prefer 17.f5 ! with an attack
in the event of 17 . . . b5 18.fxe6 fxe6
19 .'I1*rg4.
1 7 . . . Vll c S+ 1 S .g 1 gfeS 1 9 .94
lLl b7
20.Vllf3 ! ?
A very original decision ! The
knight on d2 does not support well
the attack with f4-f5 so White em
braces "a slow ball" strategy. His
only weakness on d3 will then be
defended by the king so White is
not running any risk.
20 . . . Vllxf3 2 1 . lLlxf3 gcdS 22.f2
lLl a 5 2 3.e2 lLl cs
Sokolov has everything under
control and now he can launch a
pawn storm on the kingside.
24.g5 lLl b4 25.h4?!
White should have prevented
25 .. .f5 26.gxf6 gxf6 by 25.1':1g1 !
25 . . . lLl bS? 2S.lLle4 lLl d 7 27Jg 1
fS 2S.h5 gaS
Black apparently hoped that
without queens White's attack
would not be decisive, but he was
wrong.
29.gS fxgS 30.lLlfg 5 .ixg5
31 .lLlxg5
White misses the chance to win
beautifully with 31.hxg6 ! ! + - i.f6
(31. . .hxg6 32.1':1xg5) 32 .gxh7 'tt>f7
33.ltJd6+ 'tt> e7 34.ltJxe8.
3 1 . . .'e7 (31 . . . e5 ! ) 32.lLlxh7
gxh5 33.gxg7+ dS 34. lLl g 5 1':1gS
35Jh7 e5 3S. lLl f7+ eS 37.lLlg5+
dS 3S.fxe5+ lLl xe5 39.lLle4+
eS 40 . .ixe5 xe5 41 .1':1e7+ f5
42JU1 + gS 43JfS mate ! 1 -0
1 9 . Zaic h i k-Tiviakov
Moscow 1 994
1 .lLlf3 lLlfS 2. c4 bS 3.g3 eS 4 . .ig2
.ib7 5.0-0 1e7 S.b3 0-0 7 . .ib2 c5
S.lLlc3 d5 9.e3 dxc4 1 0 . bxc4 lLl cS
1 1 .d3 Vllc 7 1 2 .Vlle 2 as 1 3 Jab1
gabS 14 . .ia 1 gfdS 1 5.gfd 1 .iaS
1 S. lLl e 1 lLleS
2 07
 Part 8

30.gxh6 lLlb3 31.h7+ Wh8 32.1!xb3

cxb3 33.lLlf4 e8 34.g4-+,

The knight is heading for d6.

From there, it not only supports
. . . b6-b5, but restrains the c3-
knight. but humans tend to avoid them
with the defending side.
1 7.f4 d 6 1 8 .a4
23.e5 .te7
Vokac-A.Ivanov, Prague 2010
saw 18.lLle4 lLlxe4 19 . .be4 b5 when
2 0 .g4 ! g6 2 1.e2 would have
provoked an important weakness
of the long diagonal. Black should
probably plug it with 2 1 . . .f6, but the
threat of g4-g5 would remain after
2 2 .g4 lLlb4 23 . .ba8 1!xa8 24.a3 lLlc6
25.cxb5 axb5 26.1!xb5.

18 ... b4 1 9 .,ba8 gxa8 20.e4 24.g4

e8 2 1 .gd2
White makes a useful move us
ing the fact that the opponent has
not serious counterplay.
21 . . .tvc6 22.f3 .tf6
The engine Houdini suggests
the dubiously looking manoeuvre
22 . . . lLld6 23.g4 lLlb7 24.f5 lLla5, ob
viously heading for c6. "He" might
be able to save his skin in variations
of type of 25.e5 .if8 26.lLlg5 h6
27.lLlh3 b5 28 .fxe6 fxe6 29.g5 bxc4
Game 18 A.Sokolov-Thesing
has taught us that endgames in
this structure offer White an abso
lutely safe position with a clear plan
on the kingside: 24.e4 ! ? 1!ac8
25.xc6 1!xc6 26.1!bdl h5 27.Wg2
g6 2 8 .h3 lLlg7 29.g4;t;.
24 .. Ja7 25.f5 gad7 26.gbd 1
.tf8
A critical moment. White should
have played here 27.f6 ! with an
edge. Instead, he wastes a tempo

208
 1.1Of3 d5 2.c4 e6 3.g3 1Of6 4.i.g2 i.e7 5.0-0 0-0 6.b3
and Tiviakov gets just enough coun
terplay to maintain the balance :
27 . .1b2 ? ! c7 28.f6 b5! 29.axb5
axb5 30.fxg7 .1xg 7 3 1 . e4 bxc4
32.dxc4 gxd2 33.gxd2 gxd2
34.fxd2 Wa4 35.f3 Wa2 36.<bf1
Wb1 + 37.e1 ca6 38.c3 Wg6
39.e4 h5 40.gxh5 Wf5+ 4 1 .Wf3
c6 42.f2 ab8 43 .Wxf5 %-%
20. Pigusoy-Zontakh
Soc h i 06.05.2007
1 . f3 f6 2.c4 b6 3.g3 .1b7
4 ..1g2 e6 5.0-0 .1e7 6.b3 0-0
7 . .1b2 c5 8.c3 d5 9.e3 dxc4
1 0 . bxc4 c6 1 1 .We2 Wc7 1 2 . e 1
(I recommend 12 JUd1)
1 2 . . . a6
It seems reasonable to organise
counterplay with . . . b5.
12 . . .l''! ad8 13.f4 lOb4 is also pos
sible, but White has some initiative
after 14.Eif2 ! ? (14 . .ixb7 xb7 15.a3
lOc6 16.g4 lOe8 ! 17Jb1 1Od6 18 .i.a1
f5 turns the tables in Black's fa
vour.) 14 . . . .ixg2 15J'!xg2
Ribli suggests here to reduce
White's attacking potential with
15 . . . b7! 16.g4 lOe4 17.a3 lOxc3
18 . .ixc3 lOc6, with a defendable po
sition.
Instead, the game Sorokin
Slipak, Villa Martelli 1997, conti
nued with 15 . . . d7 16.g4 1Oe8 17.g5
lOd6 18.1Of3 lOf5 19.1Oe5 b7, when
2 0.a3 ! lOc6 21.lOg4 Eid7 2 2 .lOe4
Eifd8 23.i.c3 a6 24.l''! b 1 would have
been gloom for Black because of the
lack of useful moves. For instance,
24 . . . lOa7 Ioses to 25.lOef6+ .
1 3.f4 a 7 1 4.a4
14.g4 turned well for White af
ter 14 . . . Eiab8 15.g5 1Oe8 16.lOe4 1Oc6
17.Eif3-+ f6 18.Eih3, Galliamova-Bot
sari, Innsbruck 1987, but critical is
14 . . . b5 ! 15.g5 1Oe8 16.lOe4 1Od6. e.g.
17.i.e5 lOc6 18.lOxc5 lOxe5 19.1Oxb7
lOxb7 2 0.fxe5 bxc4oo.
1 4 . . . .1xg2 1 5. Wxg2
209
 Part 8
1 5 . . . 'llY c 6? !
This move may be principled
because it brings another hit on the
critical square b5, but it cuts off the
a7-knight from the kingside. 15 . . .
:gad 8 ! 16.g4 lD e 8 was preferable :
17.lDe4 lDc6 (17 . . . lDd6 18.lDxd6
ixd6 19.lDf3 lDc6oo) 18.g5 lDd6oo.
1 6.f3 :!Ud8 1 7 .'llYe 2
I see no need for this retreat.
17.g4 :gab8 18.5 gives White nice
attacking prospects.
210
Now Black succeeds to trade a
couple of minor pieces.
1 7 . . . e4 1 8 .e5 xc3 1 9 .J.xc3
'llY c 7 20.g4 c6 2 1 . g 5 xe5
22 .J.xe5 J.d6 23.J.c3 e5 24.f5
Despite the exchanges, White
has a clear positional advantage.
24 . . . e4 25.'llY h 5 'llYc 6 26. h 1 'llYe 8
27 JU2 ga7 28.gg 1 g6 29.'llY h 6 J.f8
30.'llY h 4 J.g7 31 .J.xg7 xg7 32.gg4
g8 33.'llY h 6 gxf5? 34.gh4 'llYe 6
35.'llYx h7+ f8 36.gh6 'llYe 5 37.gf6
'llYa 1 + 38.g2 'llYd 1 39.'llYxf5 1 -0
Index of Variations

Part 1. Anti-QGA
Vlf3 d5 2.c4 dxc4 3.e3
3 . . . cS (3 . . . bS 14; 3 . . . lt'lc6 14; 3 . . . g4 15; 3 . . . e6 15) 4 . .hc4 It'lf6 5.0-0 e6
6.e2 a6 7J''l d l! 19
6 ... lt'lc6 7.!'idl e7 (7 ... a6 8 .d4 19) 8.lt'lc3 0-0 9.d4 cxd4 (9 . . :c7 17)
10 .exd4 17
1O.lt'lxd4 18

Part 2. Reversed Benoni

1.c f3 d5 2.c4 d4
3.g3 lt'lc6 29
3.e3 lt'lc6 (3 . . . cS 42) 4.exd4 3 0
4 . b4 dxe3 ( 4 . . . ig4 32) S.fxe3 It'lxb4 6.d4 e S 32
6 . . . cS 34
6 . . . e6 38
3.b4 f6 (3 . . . g6 4 0 ; 3 . . . aS 41 ; 3 . . . cS 42) 4.e3 eS (4 . . . dxe3 42; 4 . . . cS 42) S.cS
S . . . aS 6.lt'lxeS 44
6.c4 44
6.bS+ 46

Part 3. Anti-Slav; Anti-Chebanenko

Vilf3 d5 2.c4 c6 3.e3
3 . . . lt'ld7 59
3 . . . g6 6 0
3 . . . fS 6 0
3 . . . lt'lf6 ( 3 . . . ig4 61) 4.lt'lc3 ig4 61
3 . . . e6 4.b3 fS 63
3 . . . lt'lf6 4.lt'lc3 a6 S.c2 65 (S.b3 65)
S . . . ig4 65
S . . . e6 6.d4 66
6.b3 68
S . . . bS 69
S . . . g6 70

Part 4. Anti-meran I
1.lt'lf3 d5 2.c4 c6 3.e3 It'lf6 4.lt'lc3 e6
S.b3 (S .c2 87) It'lbd7 (S . . . d6 88; S . . . b6 89) 6.c2 e7 9 0
6 . . . b6 91

211
Part 5. Anti-meran II
1.c!ilf3 d5 2 .c4 c6 3.e3 c!ilf6 4.c!ilc3 e6 5.b3 c!ilbd7 6.YlYc2 i.d6 7 .ib2

7 . . . 0-0 (7 . . . a6 1 0 1 ; 7 . . . YlYe7 1 0 2) BJ'%gl ! ? (B .i.e2 - Part 6) YlYe7 1 0 3

B . . . a6 1 0 4
B . . . eS 1 0 7

Part 6 . Anti-meran III

1.c!ilf3 d5 2.c4 c6 3.e3 c!ilf6 4.c!ilc3 e6 5.b3 c!ilbd7 6.YlYc2 .id6 7 ..ib2
0 - 0 8.i.e2
B .. JeB 9 . 0-0 dxc4 12 0
9 . . . eS 121
9 . . . b6 122
B . . . YlYe7 9 . 0-0 b6 125
9 . . . eS 125
9 . . . geB 126
9 . . . a6 127
9 . . . dxc4 128
8 . . . a6 9.d4 bS 13 0
9 . . . eS 131
8 . . . dxc4 131
B . . . b6 132

Part 7. Anti-Queen's Gambit I

1.c!ilf3 d5 2.c4 e6 3.g3
3 . . . c6 4.i.g2 dxc4 153 (4 . . . i.d6 153)
4 . . . tiJf6 154
3 . . . g6 155
3 . . . dxc4 4.a4+ (4.tiJa3 156) 4 . . . i.d7 S.xc4 i.c6 156
S . . . cS 157
4 . . . c6 159
4 . . . tiJd7 S.i.g2 tiJf6 6.xc4 a6 165
6 . . . cS 167

Part 8. Anti-Queen's Gambit II

1.c!ilf3 d5 2.c4 e6 3.g3 c!ilf6 4 .ig2 i.e7 5. 0 - 0 0 - 0 6.b3

6 . . . cS 7.i.b2 ttlc6 8.e3 b6 9.ttlc3 i.b7 10.cxdS ttlxdS 1l.ttlxdS xdS 12 .d4
gad8 185; 12 . . . ttlaS 186; 12 . . . ttlh4 186
9 . . . dxc4 10 .bxc4 i.b7 11.e2 gc8 187;
1l . . . a6 1 9 0 ; 1l . . . c7 193; 1l . . . d7 193; l1.. .ttlh4 194
9 . . . i.a6 195
6 . . . b6 7.i.b2 i.b7 B.e3 ttlbd7 9.ttlc3 (9.e2 as 196) 9 . . . ttle4 197
9 . . . cS 199

212

missed to get a clear advantage af
ter 21.f3 ! Then 21. . .'lWxf4 loses to
2 2 .g3 'lWxb4 (or 2 2 . . .'c4 23.'lWxc4
bxc4 24Jid7 cB 2S . .td4 e6
26 . .thS+ -) 23.d4 'lWaS 24.h4 g6
2S ..tdS cxdS 26.'lWf2 .
Black can double his rooks on
the e-file with 2 1 . . .aeB, but 22 .g3
'lWh3 (22 .. .'h6 23 .fS ! 3e7 24.f6
gxf6 2S.d4 looks crushing.)
23 .'f2 3e7 24.fe1 f6 2S ..tc3 'lWfS
26.xe7 xe7 27.a3.
Compared to the game, here my
f3-bishop dominates its poor coun
terpart on b7. Therefore, I can even
swap rooks, retaining the edge.
Play might continue 27 . . . hS
2B.e1 xe1+ 29 .'lWxe1 mf7 30 .'lWe2
g6 31..te4 'lWg4 32 .'lWxg4 hxg4 33.fS
gS 34.mf2 and the g4-pawn is
about to perish - 34 . . . me7 3S.me3
mf7 (3S . . . .tcB 36.hc6 hfS 37 . .tb7
.td7 3B .ha6 fS 39 . .tb7; 3S . . . .teS
36 . .td4 ! ) 36 . .td3 .tcB 37.me4.
21 . . J'xd3 22 .tYxd3 gd8
2 2 ... eB ! was more accurate -
23.g3 'lWe7 24 . .tf3 as 2S.'lWc3 'lWe3 +
26.'xe3 xe3 27.mf2 d3 2B.me2
dB 29.a3 f6 30 ..tc3 axb4 31.axb4
mf7;!;.
142
Part 6
23 .'c3 YlYh6 24.gf3 (24 . .tg4
.tcB 2S . .txcB xcB 26.g3 eB 27.e1
xe1+ 2B .'lWxe1 mfB = )
24 . . . f6 2S.g3?!
Only 2S.'lWb3+ mfB (2S . . . dS
26.g3 mfB 27.d3 'lWg6 2B.xdS
cxdS 29.mf2 'lWfS 30 ..tf3 hS;!;)
26.eS ! heS 27.fxeS .tcB 2B.exf6
gxf6 29.h3 keeps some edge.
Now Black's bishop enters play
and I cannot make any progress.
2s . . .Ac8 26.ge3 '.!?f8 27 .if3 id7
28.gd3 ie8 29.gxd8 .ixd8 30.YlYd3
Ac7 3 1 .h4 YlYg6 32 .YlYxg6 hxg6
33.'.!?f2 as 34.a3 axb4 3S.axb4 id7
36.'.!?e3 rtJf7 37.g4 Ab8 38.gS ic7
39 . .ic3 .id8 40 .ie4 ie7 41 .rtJf2
id8 D raw.
1 3. Delchev-Va n Wely
M e l i l l a , 2 1 . 1 1 .201 1
1 .f3 dS 2.c4 c6 3.e3 f6 4.c3
e6 S.b3 bd7 6.YlYc2 id6 7.,ib2
0-0 8 .ie2 a6 9.d4 eS 1 0 .cxdS
cxdS 1 1 .dxeS xeS 1 2 .0-0 .ig4
1 3. xeS ixeS 1 4.,ixg4 xg4
1 S. h 3 f6
1 6.YlYfS
 4.lt'lc3 e6 S.b3 lt'lbd7 6.VBc2 J.d6 7.J.b2 0-0 S . .te2 ! ?
I already had this position last
year in the same team tournament
against Mchedlishvili, but this game
was not published in any database
so my opponent did not know it. The
idea of the queen move is to leave
the open c-file with tempo. Howev
er, the queen would probably stand
better on d3. For instance :
16.E!:ac1 ! E!:cS 17.E!:fd1 VBaS where
Black is playing for two results only.
The game J ovanic-Sedlak, Varazdin
20 04, went lS.VBd2 E!:fdS 19.1t'le2
xd2 20.E!:xd2 E!:xc1 + 2 1.hcU.
l S .VBd3 was stronger.
Now lS . . . hc3 19 .hc3 VBxa2
2 0.hf6 is clearly out of question,
so Black should probably concede
a small edge after lS . . . E!:fdS 19.a3
l'k6 2 0 .b4 VBc7 2 1.bS E!:c4 2 2 .bxa6
bxa6 23.lt'le4 E!:xc1 24.lt'lxf6+ hf6
2 5. E!:xc1 VBb6 26.hf6 VBxf6 27.VBc3
Wd6 2S.a4 h6 29.aS;!;. It is not a fun
to defend this ending.
16 Je8 1 7 .gac1 Wla5 ? 1
..
17 . . .E!:cS was better. Then lS.lt'la4
1xb 2 19.1t'lxb2 is rather dull so
Whit e has to follow up with lS.f4
.1xc 3 19.hc3 It'le4 (19 . . . E!:xe3??
2 0 .ixf6) 2 0 .id4 E!:c6 21.E!:xc6 bxc6
2 2 . E!:c1, with a complex game. The
text is sharper, but the ensuing
complications favour White.
1 8 .f41 .id6 1 9 Jtcd 1 1
19.E!:fd1 is dubious due to 19 . . .
'\1;lfb6 ! 2 0 .lt'lxdS It'lxdS 21.E!:xdS
VBxe3+ 2 2 .'it>h1 '\1;lfxf4.
19.1t'la4 It'le4 2 0 .id4 looks at
tractive as the dS-pawn is weak.
Then mundane continuations as
20 . . .ifS are in White's favour:
21.E!:fd1 bS 2 2 .lt'lc3 g6 23.lt'lxe4 ! (23.
'\1;lfxdS E!:adS 24.lt'lxe4 E!:xdS 2S.lt'lf6+
'it>hS 26.lt'lxeS+ E!:xd4 27.exd4 ih6
2S.E!:cS hf4) 23 . . . gxfS 24.lt'lf6+
'it>hS 2S.lt'lxeS+ f6 26.lt'lc7 E!:bS
27.E!:c2 . Most black pawns are weak.
However, the paradoxical 20 . . .
E!:adS ! maintains the balance. The
fine point of this move is seen in
the variation 2 1.!b6 '\1;lfd2 2 2 .E!:d1
'\1;lfe2 23.E!:de1 '\1;lfd2 24.hdS E!:xdS
and suddenly White has not a good
defence against . . . bS.
1 9 . . . .if8?1
This is outright bad. My oppo
nent was already in time trouble
and he definitely missed my next
move. He should have played :
19 . . . E!:xe3 ! 2 0 .lt'lxdS ( 2 0 . E!:xdS?
It'lxdS 21. lt'lxdS '\1;lfcS 2 2 . lt'lxe3 '\1;lfxe3+
143