Vous êtes sur la page 1sur 178




Gloucester Publishen pic www.everymanchess.com
First published in 2003 by Gloucester Publishers pic (formerly Everyman Publishers
pic), Gloucester Mansions, 140A Shaftesbury Avenue, London WC2H 8HD

Copyright© 2003 Jacob Aagaard

The right of Jacob Aagaard to be identified as the author of this work has been as­
serted in accordance with the Copyrights, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, electrostatic, magnetic
tape, photocopying, recording or otherwise, without prior permission of the publisher.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN 1 85744 325 X

Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.

All other sales enquiries should be directed to Everyman Chess, Gloucester Pub­
lishers pic, Gloucester Mansions, 140A Shaftesbury Avenue, London WC2H 8HD
tel: 020 7539 7600 fax: 020 7379 4060
email: info@everymanchess.com
website: www.everymanchess.com

Everyman is the regi• 'f Random House Inc. and is used in this
work under licensr 'nc.

EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Chief advisor: Garry Kasparov
Commissioning editor: Byron Jacobs

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Production by Navigator Guides.
Printed and bound in Great Britain by Biddies Ltd.

Acknowledgements 4

Bibliography 5

Introduction 7

1 Simple Truths 10

2 Primary Concepts 23

3 Defining Weaknesses 35

4 Squares - And How Pieces Exploit Them 45

5 Analysing Your Own Gap1es 55

6 Positional Sacrifices 62

7 Positional Exercises 66

8 Solutions to Exercises 85

In January 2002 I started an exercise program on the Internet. The main idea was
first and foremost to write a book with a lot of really good positional exercises, as I
felt that this would be something at which I would be good. My main idea by run­

ning it as a training option for sale on the Internet was to have the material tested
before publishing it in book form. This turned out to be a good choice, as this book
has included only about 60% of the exercises from the e-mail program. And that is
30-40% from the first few months and close to 90% from the program's last few
I would in that connection like to thank Brian Kristensen, Jan Hondenbrink,
Kjell Arne Brekke, Dan Gross and Richard Dumbolin for their very positive par­
ticipation in the program. I would especially like to thank Ivo Timmermans, who
took the teachings to heart and allowed me to use some of his games as examples in
this book.
Danny Kristiansen also has my gratitude for some last minute proof reading.
Besides the following exercises I have also included a number of minor chapters
with some basic methods of positional evaluation. Some of these chapters were
originally published as articles in the Swedish magazine Schacknytt, but they were
always intended for this book and written in English. I would like to thank Ari
Ziegler for his comments, suggestions and his constant encouragement, as well as
Robert Splingberg for his excellent translations and pleasant working relationship.
In this way I was able to get maximum response to my material.
I would like to thank both my editor Byron Jacobs and Dan Addelman for creat­
ing a positive work environment. Finally I would like to thank Erwin for his worth­
less advice and good friendship.

Jacob Aagaard,
Copenhagen, Stockholm, Glasgow, January 2003.


The following books were in my mind during the writing of this book, providing
me with fresh material, ideas, references or friction. I consider them all worthy of

Attack and Defence, Mark Dvoretsky and Artur Yusupov (Batsford 1998)
Canyou be a Positional Che.rs Geniu.r?, Angus Dunnington (Everyman 2002)
Canyou be a Tactical Ches.r Geniu.r?, James Plaskett (Everyman 2002)
Fischer- His Approach to Chess, Elie Agur (Cadogan 1996)
In.rlnlctive Modern Che.r.r Ma.rtetpiece.r, Igor Stohl (Gambit 2001)
Positional Play, Mark Dvoretsky and Artur Yusupov (Batsford 1996)
Po.ritional Sacrifices, Neil McDonald (Cadogan 1995)
Practical Che.r.r P!Jchology, Amatzia Avni (Batsford 2002)
Rtassm Your Ches.r Workbook, Jeremy Silman (Silman James Press 2000)
Rta.s.rm Your Che.r.r, Jeremy Silman Geremy Silman Press 1994)
School ofChe.r.r Excellence 1-3, Mark Dvoretsky (Edition Olms 2001-2002)
Secrets ofChess Intuition, Alexander Beliavsky and Adrian Mikhailchischin (Gambit 2002)
Secrets ofModem Che.rs Strategy, John Watson (Gambit 1999)
Techniquefor the Tournament Player, Mark Dvoretsky and Artur Yusupov (Batsford)
The Cntical Moment, Iosif Dorfman (Game Mind 2002)
The Method in Chess, Iosif Dorfman (Game Mind 2002)
The Middlegame 1 and 2, Euwe and Kramer (Hays 1994 - Originally published 1964)
The Road to Chess Improvement, Alexander Yermolinsky (Gambit 2000)

Ex ce lling at Positional Chess

The Seven Dead!J Chess Sins, Jonathan Rowson (Gambit 2000)

Think Uke a Grandmaster, Alexander Kotov (Bats ford 1994 - New edition)
Trainingfor the Tournament Player, Mark Dvoretsky and Artur Yusupov (Batsford)
Understanding the Sacrifice, Angus Dunnington (Everyman 2002)

Periodicals and Databases

Chessbase Megabase 2002
Chess Informant
The Week in Chess


When I wrote Excelling at Chess about a different book. In Excelling at Chess I

year and a half ago I was sure that no­ came with fundamental arguments and
body would want to read it. It turns out · a number of different chapters focussed
I was wrong. The book is (thus far) around thinking like a human instead of
clearly my best-selling work. I also suf­ thinking like a computer, something I
fered from other forms of insecurity. I will discuss again below. In this book I
had some ideas that I thought to be have focussed on the method of ab­
correct, but my beliefs in myself were stract positional thinking. Most of the
limited and I had yet to test these ideas ideas are borrowed from Excelling at
with substantial material. In other Chess, but here they are explained and
words, I was afraid of having adjusted used in a practical framework. My main
the results to the ideas, instead of hav­ aim has been to show the method in
ing drawn the ideas from the empirical practice - not that I suggest an algo­
material. This was one of the main cata­ rithm for solving positional exercises at
lysts in starting the positional exercises the board, but because I suggest that
program. I wanted to pick a wide variety the development of intuition and the
of examples of a positional or tactical general ability to play good positional
nature and expose them to critical chess can be learned by solving exer­
study. Over the more than a year that cises in the right way. Of course there
the program ran I became sure that I are many ways to study positional chess,
was right in my ideas - even more so and solving exercises is only one of
than I had believed. I could see it in my them. My claim is that it is a smart
own games and in the games of my stu­ method as you get used to thinking
dents. positionally. The exercises in this book
This book is a product of 'post­ should be enough for you to advance
Excelling' thinking. In many ways it is a from struggling in the dark to making
remake, and in many ways it is a quite strong positional evaluations at the
Excelling at PositionBI Ch ess

board. twt:tve. Tal and Capablanca w��e well

But this is not just a workbook, of known for this. They did not thor­ ·

course. The chapters lead the way and oughly examine the positions but played
the exercises are the path. In the exer­ on feeling and imagination. Tal's sacri­
cises I have discussed issues that I felt fices were often incorrect and players
required the most attention, being not like Polugaevsky and Korchnoi made it
the only considerations in positional a habit to find a flaw in Tal's ideas
chess, but nevertheless central - and · through deep calculation. Alekhine
not particularly well described in other found that Capablanca was a better
sources. There are issues about which I chess player than he, and also found
could have written independent chap­ that when Capablanca was in his ele­
ters - prophylactic thinking is one such ment, the endgame, he became lazy and
example, but I feel that this is a rather stopped working. Consequently Alek­
complex concept and players ready to hine ended up beating Capablanca in
deal with this are also ready for the the technical phase. Capablanca relied
books by Dvoretsky/Yusupov (Posi­ too much on intuition while Alekhine,
tional P� and Trainingfor the Tournament who had no world class intuition,
Player in particular). And why repeat worked hard at the board and solved
what Mark has written there? Addition­ the problems with the aid of calculation
ally I could have included a chapter on and logic.
'missing bishops', a concept I have a The perfect chess player would have
great affection for, but I felt that it was an adjustable balance between all these
too marginal compared with the rest of three facets of his talent. In complex
the book. tactical positions he would use some
There are three initial ways to deal intuition and some logic to deduct the
with a chess position. candidates moves, after which there
The first is, simply, calculation. If I would follow calculation. In technical
do this, then he does that. Then I an­ positions he would take into account
swer thus and he will be shaking like a natural technical considerations and
leaf in fear. adhere to general guidelines, adjusted
The second is intuition: Okay, what with short lines of blunder checks. This
do we have here? I don't know, I think is what is normally called positional
I will just play this. Finally there is ab­ chess. A good example is the following
stract thinking: It seems like I should game.
develop - where does this piece belong?
Hmm, maybe my knight will get stuck Fischer-lbrahimoglu
on the edge of the board and not take Siegen 1970
part in the game for some time. Perhaps Caro-Kann Defence
I should exchange rooks...
Of course all three ways of thinking 1 e4 c6 2 d3 d5 3 .!Dd2 g6 4 o!Dgf3
exist all the time in our minds. But to i.g7 5 g3 lbf6 6 .i.g2 0-0 7 0-0
what degree? Some players are very in- i.g4 8 h3 i.xf3 9 'l'xf3 lLlbd7 1 0

Introdu c t ion

1i'e2 dxe4 1 1 dxe4 1i'c7 1 2 a4 .l:l.ad8 Black has only one reasonable move,
and it is prevented. Simultaneously
White progresses with his own slow
improvement of his position.
1 3 . . . b6 1 4 i.e3 c5 1 5 a5 e5
Black is putting all his pawns on dark
squares, inevitably causing the light
squares to be severely weakened. Now
White's worst placed piece is no longer
the bishop on e3, rather the knight.
Therefore the manoeuvre aimed at dS is
very logical.
1 6 ll::ld2 ll::le8 1 7 axb6 axb6 1 8 ll::lb 1 !
This is a position that is difficult to 1i'b7 1 9 ll::lc3 lbc7
calculate. A general plan must be Black has organised a modest defence
formed and executed. To do so one has on the light squares and is about to gain
to take all kinds of elements into con­ counterplay of sorts with ...b6-b5. A
sideration. For a strong player like quick comparison of pieces reveals that
Fischer, finding the key to the position the knight is the least valuable of
was probably rather quick. But for the White's minor pieces (the bishop on g2
less experienced, juniors, club players has enormous potential from c4, dS and
and hopeless IMs like myself, a posi­ h3) and that the knigh t on c7 is the
tional analysis will be of great help. most valuable for Black. Therefore an
Here a comparison of pieces as well as a exchange is not illogical. At the same
search for ideal squares will explain time the grip on the light squares and
Fischer's treatment. It turns out that the the prevention of ... b6-b5 are Issues.
cl-bishop belongs on e3, that f2-f4 will Therefore Fischer played:
not generate an attack but rather create 20 ll::lb5!!
weaknesses in the white camp; White's From here on it is pure power play.
knight would like to get to d6 at some White's knight might have been better
point- if Black plays ...e7-e5. It appears than its counterpart on c7 but the supe­
that there is little scope indeed for Black riority of the remaining pieces is enor­
to better his position. Only the knights mous, and that is what counts on the
can be improved and they need to use - scoreboard.
cS as a trampoline on their way to e6 20 . . .1i'c6 21 lbxc7 1i'xc7 22 1i'b5
and d4 (the ideal square). As White's :as 23 c3 .l:l.xa1 24 .l:l.xa1 .l:l.bB 25
knight is not ideal on c4, and as any l:ta6 i.f8 26 i.f1 �g7 27 1i'a4 .l:l.b7
rook to dl would serve only to invite 28 ib5 ll::lb8 29 .l:l.aB .i.d6 30 1i'd1
the rook to leave f8, the following or­ lllc6 31 1i'd2 h5 32 .ih6+ �h7 33
ganisation of White's forces is logical: i.g5 l:tb8 34 .:Xb8 lbxb8 35 i.f6
1 3ltJb3! lllc6 36 1i'd5 tba7 37 i.e& 'i>g8 38
This is, by the way, prophylaxis. i.xf7 + 1i'xf7 39 11fxd6 1 -0

Ex c e lling at Positional Chess

It is clear that calculation could not Let us take an example from John's
have brought you any success in this book.
game, yet calculation is what many play­
ers would have used to try to solve
these problems. To learn to use the
right tools at the right moment is an
important part of excelling at chess. In
this book the main part of the exercises
are quite positional. We often find our­
selves in situations where we need to
solve positional questions, but where
calculation plays an important part; we
set positional goals and use tactics to
implement them, or alternatively we
have to look out for tactics in one way White to play
or another.
In modem day chess at the top a Here Yusupov played 8 lllh 3. Wat­
player's mood tends to be aggressive, son writes: 'Don't put your knights on
leading to a search for more compli­ the rim! Well, knights are living on the
cated types of positions. In fact chess edge these days, as we shall see in chap­
changed considerably when Kasparov ter 5. But the case before us is really
arrived on the scene, and again when simple. Neither side is about to make
computers began to 'comment' on the any dramatic pawn-breaks, so there is
elite players' performances in analysis plenty of time to manoeuvre pieces to
rooms and bedrooms allover the world. their best posts. In the case before us,
Kasparov introduced the initiative as a that would involve the knight going to
much more important part of positional d3 via £2; where would it go from e2?
chess than was previously understood, As McDonald points out, iLif2-d3 could
and the computers made us pay more be followed by lL!d2-c4 and .i.e 1-d2-c3
attention to the benefits of concrete with a three-way attack on the forward
analysis. Some commentators, like John e-pawn.'
Watson, have made the 'misassumption' Now what is wrong with this? Most
that this has made the lessons of yester­ of it is nothing but correct. It is clear
day to some extent irrelevant. His no­ abstract thinking, and very sound. The
tion of rule independence seems to be a problem is this thing about knights on
little flawed. In his book Secrets t?[Modern the rim. In his chapter 5, where the
Chess Strategy he tries to argue that, in knights live on the rim, they only do so
the old days, the top players followed as long as there is a concrete advantage.
some rules in a rigid way, whilst today When the advantage disappears the
they use calculation and the magic of knights race towards the centre. The
intuition, which is a result of heavy ex­ same goes for this example. The knight
posure. in no way lives on the rim it is going


towards the centre. I am sure that Tar­ Modern Chess Strate!) as a great piece of
rasch, who was not an idiot, would have work which does treat the enormous
no problems with this. He was one of evolution there has been in positional
the greatest chess players of his age, understanding since 1935. It would be
coming across as dogmatic. There is a strange if the period from 1876-1935
story about a man who had put his rook had greater leaps in understanding than
behind a passed pawn, as prescribed by 1935-2003. It would be strange if some
Tarrasch. His friends had then laughed of the observations made by the old
at him. He mailed the position to Tar­ masters were not mistaken. John pro­
rasch and asked him whether or not he vided an excellent analysis of many new
was right in following his advice. Tar­ concepts in positional chess, and has
rasch assured the man that the move been rightfully praised for it. But to
played was good, and that in the future claim that the paradigm of thinking has
he would indeed do well to follow his completely changed is going too far.
advice. Only, in the given position, l:le8 Still, this is only one conclusion in
checkmate was a stronger move! John's book. And if you make up your
I believe that John is mistaken in his own mind and take from John and from
view on Tarrasch and the others as Jacob what you find useful and relevant,
dogmatic people who did not think. I am sure that my two books and his
Evidence (their games) suggests other­ book will be able to teach you some­
wise... The above diagram is a clear-cut thing.
situation of a knight not living on the I mentioned earlier that the internet
edge but manoeuvring towards the cen­ program had given my pupils and I
tre in the most flexible way. If you un­ some tools that proved useful in over­
derstand the rule as not being allowed the-board play. Some of these can be
to put your knight on the edge of the seen in the different articles in the book,
board under any circumstances, then but I would like to give an example
you are truly dogmatic, as well as stupid. from my most recent game and from
And Tarrasch was not stupid. If, in­ three games from Ivo Timmermans'
stead, we choose to understand it in most recent tournament.
terms of knights generally having less
influence on the edge and greater power Hei-Aagaard
in the centre (which, according to �­ Danish league 2003
gaard the linguist, is actually the most
obvious semantic interpretation) then Carsten H0i is Denmark's latest
the above manoeuvre makes a lot of Grandmaster. Despite the fact that he
sense. could have received the title back in
We might have a different view on 1993, he was awarded the title only re­
how to treat the past, but we try to cently. The positions where I felt the
solve positional questions in a some­ usefulness of the line of thinking cur­
what similar fashion. rently under discussion began after 13
I have continuously praised Semts of moves:

Excelling at Positional Chess

ltJeg1, but his position does not make a

positive impression. Carsten eventually
came up with a strong move, defending
the f4-square in return for conceding
the initiative.
1 5 .tg3 �hB 16 0-0-0 itlb4!
Again unable to find something use­
ful, I make a slightly unusual move. I
did not approve of 16...ltJcS 17 b4l?
tl'lxe4 for the reasons given above (even
though it does make more sense here),
so I decided to simply harass his well
Black. to move placed queen. The knight still has cS as
an ideal square, but .tZ'lcS with tempo,

Here I used 25 minutes, trying to followed by ... a7-a5, is nicer. Therefore

choose between the plans ...tl'lf6xe4 and after
...tl'lc5xe4, in both cases to make way 17 ..b3
for my bishop on c8. I was emotionally I simply returned with
dissatisfied. I had the feeling that it had 1 7 . itla6!
. .

to be possible to play something less Carsten could also find nothing use­
forcing, as both white knights have no­ ful about having his queen on b3, so the
where else to go than e4. Why should I game ended in a draw after
help my opponent by opening the king­ 18 .-c2 itlb4 %-%
side? Finally I used the ideal squares
technique and came up with the follow­ Iva Timmermans is a 42 year-old
ing manoeuvre. doctor (single and handsome, girls) with
1 3 .../tlf& 1 4 _.c2 ltlh5! an Elo rating that has hovered around
The f4-square is the weakest point in 2200 for centuries (at least two). He
White's position. No pawn can ever followed the training program from
control it (without Black's help) and the start to finish, missing occasional weeks
f2-pawn makes it inaccessible in similar due to tournaments, a heavy workload
fashion. Additionally the f4-square is a or new episodes of Friends. In his last
key factor in the fight for the light tournament, the Rilton Cup in Stock­
squares in White's camp. Carsten now holm, he illustrated much of the bene­
saw lines like 15 0-0-0 itJf4 16 l:.g3 fits gleaned from following the pro­
lDxe2 17 tl'lxe2 'iibs. It quickly turned gram, as well as the usefulness of solv­
out that after 18 l:.dgt l:.£7 19 i.f6! this ing combinations for an hour every day
line would win for White, but it was during the month before the tourna­
equally easy to find 17 ...f4! 18 ltgg1? ment. The latter exercise is, in my opin­
'irhs when Black wins a piece. White ion, an absolute must for the ambitious
can apparently avoid losing material chess player. Not one hour per day, but
with 18 l:.a3 1Vhs 19 ltJ£3 ltJcS 20 a minimum of three to four sessions of


half an hour each week. Even I do it to teristics of the position are typical. The
keep sharp, and I no longer consider pawn strucrure favours White, who has
myself to be an ambitious player. I see it a defended passed pawn on dS. Black
as exercise for the head, which holds has dynamic chances because he has the
the chess player's muscles, just as arms pawn majority on the queenside; Black
and legs carry the muscles of other ath­ should be careful here as advancing the
letes. pawns can create weaknesses. The only
Anyway, here are a few positions open file is the a-file. White has the
from lvo's games, annotated by the bishop pair yet the position is static and
man himself: closed and the d6-square seems ideal for
the knight. Black plans the manoeuvre
Gleizerov-Timmermans ...tt:Je8-d6 but exchanging dark-squared
Stockholm 2002 bishops with ...�cS is also an excellent
Catalan idea, weakening the dark squares and
making f2-f4 less attractive. The bishop
1 d4 d5 2 c4 c6 3 liJf3 liJf6 4 llJc3 trade also brings Black closer to the
e6 5 g3 dxc4 6 ig2 b5 7 llJe5 llJd5 desired endgame of knight versus bad
8 a4 a6 9 axb5 cxb5 1 0 llJxb5 axb5 bishop (.ig2). The disadvantage of
1 1 l:txa8 ib7 1 2 l:ta7 ...lbe8-d6 is that it temporarily locks in
After the game Gleizerov was disap­ the rook on £8, and by the time Black is
pointed not to have played 1 2 lla1 with ready to activate the rook White con­
a near decisive advantage. trols the a-file, e.g. 21...t0e8 22 'it'd2
1 3. ..'iFb6 1 3 l:txb7 'iFxb7 14 0-0 .tcS 23 l:lal with advantage. It is there­
Jie7 15 e4liJf6 1 6 d5 'iFc7 1 7 llJg4 fore logical to start improving the posi­
. Later we found that 1 7 t0£3! would tion of the rook and only then the mi­
have given White a clear advantage. It is nor pieces, thus prompting 21 ...l:la8.
all about the light squares. Th.is gives White something on which
1 7 ...lDbd7 1 8 l:te1 0-0 1 9 if4 e5 to ponder- 22 'i'd2 and 22 'i'c2 can be
20 lt:lxf6+ llJxf6 21 ie3 met with 22 lbg4 or 22.. .:ta2.'
... .

21 :aa 22 if3 l!Je8 23 'W'c2 Jic5


24 b3 llJd6 25 bxc4 Jixe3 26 l:xe3

'W'xc4 27 'ii'xc4 Y.a-Y.a
The grandmaster offered a draw to
avoid a worse fate.

Stockholm 2002
Frtnch Defence

1 e4 e6 2 d4 d5 3 lDc3 ib4 4 e6
lt::le7 5 a3 .txc3+ 6 bxc3 c5 7 liJf3
lvo explains: "The positional charac- b6 8 a4 .i.a6 9 ixa6 lt:lxa6 10 0-0

Excelling at Positional Chess

h6 1 1 ...e2 liJb8 12 .ia3 ltJd7 13 c4 return, threatens .i.xcS. White has a

dxc4 14 dxc5 bxc5 15 •xc4 0-0 1 6 small edge after 20...�xeS 21 �xeS
l:tab1 'ilc7 1 7 l:tfe1 l:tab8 1 8 l:tb3 �xeS 22 'ifxcS 'ifxeS 23 :XeS.'
l:tfd8 20 ...'iia5 21 l:ta4 l:tb1 + 22 q;,h2
ltJb4 23 l:tb3 l:txb3 24 cxb3 ltJd5 25
h5 .:l.b8 26 .:l.e2 ...c3 27 .:l.c2 Wxc4
28 .:l.xc4 l:txb3 29 .ixc5 a5 30 .id6
ltJ5b6 31 l:tc7 l:tb2 32 l:ta7 l:txf2 33
:txa5 l:ta2 34 .ic5?
34 Aa7! 11xa4 3S :Xa4 �xa4 36 g4
would still have made a draw. Now
Black is better.
34 ... ltJxc5 35 l:txc5 l:txa4 36 �g3
lLld5 37 l:tc8+ Wh7 38 :Lb8 lLlf4 39
:Lb7 ltJxh5+ 0-1

In this posttlon Iva's next move Despite losing to the strong Russian
made his opponent look at him in sur­ grandmaster, Iva was very pleased with
prise. A low rated amateur is not sup­ his effort. His goal was not to score as
posed to make such deep moves. I have many points against these guys as pos­
to say here that the knowledge of this sible but to play as well as he could,
plan comes from previous knowledge enjoy playing and learn from stronger
of the opening, and not positional exer­ players. That this attitude will get him
cises alone. far in the long-term I have no doubts.
'After 18...l:.fd8 White faces the The following game, from the sixth
choice as to what side of the board on round, illustrates what he will be doing
which to play. Attempts to switch to his former equals in the not too dis­
pieces to the kingside are frustrated be­ tant future.
cause the eS-pawn has to be defended.
White has to ftnd a way to strengthen Johansson-Timmerman&
his position without disturbing the co­ Stockhohn 2003
ordination between the pieces, and Sicilian Defence
none of the pieces can be improved
right now.' 1 e4 c5 2 c3 ltJf6 3 e5 lLld5 4 d4
1 9 h4! cxd4 5 lLlf3 ltJc6 6 cxd4 d6 7 .ie2
'Squeezing Black on the kingside. g6 8 0-0 .ig7 9 exd6 ...xd6 1 0 lLlc3
White threatens h4-hS (gaining space) lLlxc3 1 1 bxc3 0-0 12 ltJd2?! l:td8
and prevents all tricks connected with 1 3 'ilb3 b6 '14 ..tf3 ..te6!
...�xeS and ...l:tdl.' Black has achieved the better open­
1 9 . ..ltJc6 20 l:tc31 ing. The white pawns are weak and
'A short move that puts pressure on dominated.
cS. White has prevented ...�aS and, in 1 5 •a4 .id5 1 6 .ia3 •c7 1 7 l:tac1 ?


This rook move creates problems for Later, when the initial excitement
White, although after 17 .:Z.fe1 he would about this move faded, we discussed
be worse. whether 20...li:la5, with a permanent
'Black has pressure against the pawns positional advantage, was a reasonable
on c3 and d4, and if the pawns remain alternative to this little combination
blocked for a while Black can play on (which seems to win a pawn, but no
the light squares c4 and d5. A solid more).
move could be 17...:tac8, completing 21 f4?
development. However, the solution to As so often happens White cracks
the problem should be easier to find if under the pressure of repeatedly being
we ask which pieces Black wants to ex­ faced with new and unforeseen prob­
change. In the most simplified version lems. The alternatives were as follows:
Black aims for a good knight versus bad Black wins after 21 l:te2 ltJ£3+!! 22
(dark-squared) bishop ending. Hence gxf3 l:t g5+ 23 �h1 'i'f4 24 'i'c6 .:Z.h5 25
Black wants to exchange light-squared ..xa8+ �g7, which leaves 21 .:Z.e1!
bishops as well as the remaining bishop 'i'xc3 22 .tb4 'ii'c6 23 'Wb3 (23 ..xc6
for White's knight. Any exchange of f!Llxc6 24 .i.xe7 l:e8 25 l:tde2 :Xe7 26
heavy pieces is also welcome, of course. .:Z.xe7 lilxe7 27 :Xe7 llxd4 28 �fl lla4
Here Black can force matters with a and Black wins) 23... li:lc4 24 .:Z.c2 :X d4
subtle move:' 25 l:t xe7 aS 26 �e1 lle8 27 .:Z.ce2 :Xe7
1 7....th6 28 :Xe7 .l:e4 29 l:xe4 'i' xe4 with a
"Threatening 1 8....i.xf3 ensures that clear advantage for Black.
the right pieces are removed from the 21 . . .�c4 22 i.c1
board. After the forced sequence 18 22 l::te2 .:Z.aS is the sad reality.
lied1 .i. xd2 19 � xdS l:txdS 20 :Xd2 22 .. .�xd2 23 .txd2 e6 24 g4 :ca
Black can chose between 20...ttk5 and 25 f5 gxf5 26 gxf5 �f5 27 lbf5
20...ltla5, gaining complete control over exf5 28 Wc2 Wc6 29 1i'd3 •e4 0-1
the light squares.'
18 lcd1 i.xd2 19 i.xd5 ltxd5: 20 This book is about abstract thinking,
l:lxd2 �e51? about understanding chess consciously.
Intuition might work for some people
who are exposed to a lot of chess posi­
tions, but I am of the belief that know­
ing why you should do what you should
do is just as important. And especially
for people who have reached a cenain
level and cannot seem to improve de­
spite working extensively with tactics
and openings, trying to get a deeper
understanding of the game might be the
way forward.

Simple Truths

In this chapter I want to talk about mis­ other pawn and found himself in a rook
takes that are often committed but so endgame, which was still winning. He
obvious that, when you realise you are concentrated well but overlooked some­
guilty of making them, you tend to un­ thing and the ending was drawn. After
derplay their importance. The first one is the game he would not hear of misplay­
connected to simple evaluation tech­ ing a winning position. He would have
nique. won had he simply made the correct
I have a friend who is close to IM choice on move 50!
level; he is a really talented player with Well, I often see people drift from
several norms and an ELO rating that is comfortable winning positions into dif­
so close that two good games one day ficult winning positions when, suddenly,
will make him an IM. Nevertheless he is winning requires considerable effort.
violating one of the simplest rules of Chess becomes hard again.
chess so often that even I noticed this There are very few people in the
as a weakness in his play. This is best world who talk about bad positions we
illustrated with an example. know we can draw in the way Kramnik
In one game he had an ending with did after his match with Kasparov.
three extra pawns, a lead in develop­ Most of us know that we are susceptible
ment and a clearly superior pawn struc­ to making errors. We misplay our win­
ture. From then on he played rather ning and drawn positions again and
carelessly. In order to exchange a few again. So in order to make it easy for
pieces he gave up a pawn, as well as the ourselves we try to play accurately, even
momentum of direct attack on the king. when we are three pawns up. w� do not
Later he just waited, resulting in his wait for the-opponent to resign, but try
opponent gaining some activity for his to make him do so. Such an approach
pieces and, suddenly, a few threats. To needs extra effort, but we get used to
protect himself my friend gave up an- winning, rather than watching the op-

Simple Truths

ponent losing. And sometimes we get a concentrated only once and saw a two­
full point quickly because our opponent move line. But now it is time to wrap
realises that we are not going to gradu­ up the full point and put it in the bag.
ally let the victory slip away and, conse­ Such a position might very well win by
quently, he tries something desperate. itself, but my lead in development
The main mistake my friend commit­ should be enough to tear White's posi­
ted was to decrease his level of concen­ tion apart. Now I chose to force him to
tration. Although he has sufficient pres­ castle queenside because I had seen a
ence ·of mind not to blunder, he still more or less forcing 'I.Vin.
made enough inaccuracies for his posi­ 22 . ..:tae81 23 0·0·0
tion to go from easily winning, to win­
ning and to drawn.
The following game is an antithesis
to my friend's performance, where I
concentrate right until to the end, mak­
ing the most of my ability to concen­
trate (although the game can be said to
be over straight out of the opening).

G reger-Aagaard
Danish League 2001
Sicilian Defence
23 lth3 e3 24 'ii'xd3 ex£2+ 25 �d2
1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4 4 Wxb2+ 26 l!Jc2 l:lel, when Black wins
tbxd4 tbf6 5 tbc3 e5 6 lLldb6 d6 7 everything, was the tactical basis of my
.tg5 a6 8 tba3 b5 9 .txf6 gxf6 1 0 line. But what now? How to proceed?
lLld5 f5 1 1 .id3 .te6 1 2 •h5 .tg7 23 .. J:tb8!
1 3 c3 0-0!? 23...l:.a8 would probably have been
A sharp sideline but healthy enough. the choice of many, as the threat ....Uxa3
1 4 exf5 .ixd5 1 5 f6 e4 1 6 fxg7 would force White into 24 'ife3, which
:tea loses the game after the exchange of
Here White has a choice between queens. But I did not see any reason for
ii.e2 and .ic2, but he completely over­ my opponent to make it to an endgame.
looked my intended response to 17 Not that I had any doubts whether or
ii.xbS and lost a piece. not I would win it, but I did not see it as
17 .txb5?? l:le5! 1 8 ...h6 axb5 1 9 the strongest option available for me.
lbxb5 .ic4! 20 tba3 You can compare it to choosing be­
20 lZkl4 l!Jxd4 21 cxd4 l:tgS! and tween winning a pa-wn and a piece.
...WaS+ will decide the game on the 24 l:lh3 l:ta5! 25 l:txd3 exd3 26 tbc4
next move. 'ii'b3!
20 . . . .td3 21 h4 ..b6 22 'ii'd2 26...�xa2, of course, wins immedi­
So far it has been easy for Black. I ately, and then after 27 'l'xd3 comes

Excelling at Positional Chess

27...'ii'b3! with a direct win, as can be exchanges - as we learn to do when we

seen in the game. But I was focused and grow up - but instead found that the
saw another way to decide the game. task becomes more difficult. The twn­
27 •xd3 l:txa2 28 l:e1 �51 0-1 ing point for me came in the following
position in 1995 (after 42 lL!dS).
Obviously not a Beauty Prize game,
although I am quite proud of it. I played Jaksland-Aagaard
two very nice moves (23...llb8! and Denmark 1995
26 ...'ii'b3!) and kept maximum concen­
tration all through the game. However,
it is not a game worth publishing any­
where - White just blnndered a piece, as
a pupil of mine pointed out (it should
be said that this pupil does have a ten­
dency to draw his winning positions...).
I once saw a game being played in
Gennany, Movsesian being White
against Korchnoi in the final round of
an open event. Movsesian had made
7'12/8, while Korchnoi was on only 6Vz
points. The advantage of the fitst move Here I considered playing 42 �xd5

saw Movsesian obtain a small advan­ (as, I believe, many people would). The
tage, and then he systematically made reasoning is simple: if the exchange is
poor exchanges, leading to an endgame possible I have less pieces on the board
which he knew was drawn and which he and my pawn will connt more. How­
drew. I am always impressed with this ever, this is far from the truth. The real­
kind of confidence and control, but I ity is that the bishop on c6 is so much
still find it foolish in 99% of cases. better than the knight on dS that to
Recently a friend of mine had a win­ trade these pieces would be terrible.
ning knight(s) ending with a clear pawn White would also gain a passed pawn.
up. Her king had a fast lane to the cen­ To me this game was a kind of a break­
tre and she had fewer weaknesses than through as I realised these things and
her opponent. But she used a lot of afterwards saw them manifest them­
time trying to calculate a variation until selves quite clearly with an easy win:
the end; she thought she had succeeded 42 .. Jbb3 43 l:g7+ �f3 44 l:h7
but, somewhere in one of her sub-lines, l:tb1 + 45 'ii>h2 l:b2+ 46 'ii>g1 l:tg2+
she had overlooked something, and the 47 �f1 l:tc2 48 l:th3+ �g4 49 l:c3
resulting pawn ending was then losing ..i.xa4 50 l:.xc2 ..i.xc2 51 c5 �g3 52
instead of winning. I have seen this mis­ lt:lb4 ..i.a4 53 'ii>g 1 ..i.b5 0-1
take being made many times, including I did not really realise what kind of
in my own games. I have often tried to rule was working here, but I do now. It is
'simplify' the technical task through the same that counts in all the examples

Simple Truths

mentioned above. You see, normally we a modest lead, which they do not like,
are talking about chess positions as win­ and other 'level' positions which they do
ning, clearly better, slightly better, equal, like. Once you appreciate that you
unclear, slightly worse and so on. But should play the kinds of positions you
there are no real definitions of winning .:... like, then you have already made a great
because what is a winning position? Is it improvement of your understanding of
one you w1n 100% of the time, as my practical chess.
first Coach Henrik Mortensen tried to It is my claim that this kind of informa­
convince me, or is it a position where tor evalllation in over the board analysis is
you can prove a win as I read it in In­ one of the reasons why some people
fonnator? Well, if I was interested in calculate too much. When there are

opening theory as an independent disci­ forced variations you need to calculate to

pline, and not as something that helps the end (funnily enough, most calculating
me in my practical efforts, I would per­ pk!Jers have a tendency to neglect this),
haps still have the same definition as but when you are considering positional
Informator. But I don't I believe Henrik factors there seems to be a general
is correct. You should consider a posi­ agreement among leading instructors
tion to be winning only when you have (Yermolinsky and Silman are the first to
no doubt whatsoever that you are going spring to mind) that you perform a bitm­
to win it In the first example of this sec­ der-check but no actual analysis. Silman
tion my friend went from a winning po­ has his own ideas about how these posi­
sition to great winning chances to draw tions should be treated. He calls this the
because he did not realise one simple Silman thinking techniq11e and uses some­
thing. thing he calls Jantary positions. This has
The choice ofmoves sho111d not be made on an some resemblance with what I would call
exact verdict of the finalposition,
bllt on whether the search for the ideal sq11are for the
or notyo11r position has improved or worsened. pieces. The only difference is that I look
This might seem obvious to the point at the individual piece and try to keep
of being naive, but for many playeril this things as simple as possible. Silman has
'theory' does not find a way into their no problems working with three pieces
practice. And for my friends above this for each player, but how about six or
would have saved them the embarrass­ seven? Personally, I would get confused
ment and pain of throwing away easily trying to juggle all these pieces in my
winning positions. head at the same time. Consequently I
The mistake has its origin in forgetting prefer looking at each individual piece.
that chess is a game in which we should But the basic idea is good. Yermolinsky
use practical measures to assist us in is more of a self-taught player and there­
making our decisions - not theoretical fore has no greater ideas or advice con­
measures such as clear advantage, slight cerning thinking methods. But if I could
edge or winning positions etc. I think offer one simple piece of advice it must
most players remember the day they real­ be this - ask yourself at the end of a line:
ised that there are positions which offer Am I making progress? Is my task easier

Excelling at Positional Chess

or more difficult after my planned move? After 16......c5 White has a tactical
Tills advice is exactly what Svidler advantage from the exchange on d6: 17
must have been following during the lLlxe7+! lLlxe7 18 ixe6 fxe6 (note that
following (very instruCtive) game. this is highly different from 18...:Xe6
after 16...Afe8 - see next note) 19
Svidler-A.Sokolov W'xd6! W'xc2+ 20 �a1 and now both e6
Elista 1994 and e7 are hanging, when 20...tLlc6 21
Sicilian Defence 1i'xe6+ �h8 22 ltJgS h6 23 tLlf7+ �h7
24 'it£5+ �g8 25 lLlxh6+ �h8
1 e4 c5 2 lDt3 e6 3 lDc3 lDc6 4 d4 (2S...gxh6 26 'ii'g6+ �h8 27 lld7 tLle7
cxd4 5 lD xd4 d6 6 f4 lbt6 7 .ie3 28 'ii'xh6+! �g8 29 WgS+ �h8 30 :Xe7
e5 8 c!bt3 lDg4 9 'ii'd2 tbxe3 10 wins) 26lLlf7+ �g8 27 'i'g4! wins.
'ii'xe3 exf4 1 1 'ii'xf4 .ie6 12 0-0-0 1 7 i.b3!
.te7 1 3 �d5 0-0 1 4 �b1 lies 1 5 From a pure positional point of view
.1e2 'ii'a5 this is the most pleasant move to play.
Here White has a structural advan­ By guarding the king it prepares for the
tage due to the control over the dS­ attack on the weakness on d6. 17
square but his pieces are still not ideally lLlxe7+?! is too greedy. After 17...llxe7
placed. His bishop needs to find a bet­ 18 .ixe6 llxe6 Black is already freed
ter square and it is not obvious yet how somewhat from all his troubles, which
to activate the h1-rook. should alarm White (he has not made as
much progress as Black!) that perhaps
no pawn was worth this. And then after
19 llxd6? 'i'c7 20 .z:r.hd1 (20 eS tLlxeS!
with the idea of 21 .z:r.xe6 ti'xc2+ 22
�a1 00! and Black wins) 20...tLld4!!
Black wins material.
1 7 . . ....c5
With the logical idea of ...liJaS-c4
(xb3) to fight for the control of dS.
18 Ad3!
The piece that needs to be activated
before the direct assault is the rook and,
1 6 i.c4! as the weakness White is attacking is the
The most obvious improvement of d6-pawn, the rook belongs on the d-ftle.
the position. The control over dS is Another good feature about the text is
strengthened and so is the king's posi­ that it prevents ...ltlaS (prophylaxis)
tion. The idea is not to exchange on e6, while improving the pieces. This is what
but to guard the king from b3 and identifies a great move.
eliminate the pressure on the diagonal, 1 8 . . . b5
as can be seen from the next move. 18 ..tLlas 19 llc3! wins for White.

1 6 . . .l:fe8 1 9 l:l.hd1

Simple Truths

19 l:c3?! makes no sense now. It is are some transcendental rules in chess, as

better to play with all the pieces. if I resist the validity of variations as
1 9 . . .ltJa5 20 lt.Jxe7+ proof. This is, of course, absurd, and I
Now, fully developed and organised, will not waste more time with it, other
White can cash in. The rule is that you than to say it is not true.
should develop fully before taking ad­ In the Svidler game I am unsure if he
vantage of permanent weaknesses. - saw the 20...lDd4! combination at the
20 . . .l:txe7 21 l:txd6! ltJxb3 22 l:tdS+ end of 17 �e7?!, but I am pretty confi­
.=.as 23 AxeS+ AxeS 24 axb3 h6 25 dent that he felt Black was getting too
l:.d4 :ca 26 'i'd2 'itrh7 27 b4? much freedom too soon, and that it was
27 h3! is necessary. thus a logical to be sceptical about 17
27 . . .'iic7 28 b3 :taB! 29 l:td61 a5 30 CiJxe7 altogether. Any player with com­
bxa5 l:txa5 31 e5 •a7?! 32 •d4 mon sense, regardless of whether he be­
•as 33 •d3+! g6 34 •d4 l:.a3 35 lieves there are no truths in chess, should
l:td8 'iia 5 36 l:teS? test his assumptions afterwards by ana­
36 llb8. lysing the position. As Esben Lund told
36 . . . b4! 37 ltJd2 'W'a6? me: Wben I think I am being clever I am being
37...'ifb5. real!J strpid. The one who has no doubts
38 lZ:lc41 l:ta2 39 l:.b8 h5 40 l:txb4 about his view of the world is in danger
i.f5 41 ltJe3 1 -0 of being more wrong than the one who
The note to move 17 is very impor­ believes in himself but is always open to
tant The weakness is not running away; the possibility that he might be wrong.
it is a static feature and White therefore We are never too smart to learn more...
uses his space advantage and freedom to Personally, when a truth becomes so
manoeuvre to improve bis position before complicated that I do not feel in com­
beginning the attack on the weakness. plete control over all aspects of it, I get
This, of course, reflects another, very suspicious. I remember the Nobel Prize
obvious, simple truth: Your attack will winner in Mathematics said that he felt
have greater strength if you increase the he really understood a theory or concept
m.unber of pieces you include. only if he could grasp it in one unifying
In the Svidler game White organised idea. This is the reasoning behind the
first and only then attacked. Conse­ next chapter's discussion of Primary Con­
quently Black did not gain any real com­ cepts as a possible way to penetrate many
pensation for the pawn as White did not positions. But for now I will just wish
lose anythlng important in terms of ac­ you luck with improving your positions,
tivity when he finally decided to try to your chances and your results.
win the pawn. Some might argue that
there were tactics defending the pawn, Explanation of Terms
and that was why White did not take it !'!formant evaluatiotr. A theoreti-
These might include those who want to cal/ scientific evaluation of a poSltton
understand my ideas about conceptual mixed up with some ptactical considera­
thinking and the fact that I claim there tions. Very useful for organising opening

Excelling a t Positional Ch ess

theory in ECOs and for cross-lingual In this position White blew it with
annotations, but not a great help for de­ 1 9 exf5?. An abstract notion would
cision making at the board. · look at the enormous lead in develop­
Calculating pkfyerr. Players who tum on ment White possesses and the need for
Fritz at once without first having more an immediate attack; if not the static
abstract thoughts about the position. features such as the ruined pawn struc­
Bliinder-check Just checking if you ture will begin to count. The primary
missed some threats before you play the concept here is the exploitation of the
move you find natural, based on posi­ lead in development to get the rooks
tional evaluation. connected on the seventh rank. As this
Silman Thinking TechniqMe: 4 five-step would win the game immediately it
method that is meant to organise your takes priority over all other considera­
thoughts so you pay attention to what is tions in · the position. Therefore 19
relevant Probably good · as a training .icS!! is the right move, preventing
method but, most likely, also unsuitable ... l:te7. After this Black has no defence.
for peak performance at the board.
Fantasy positions: When you move the Further Reading
pieces around in your head to set up
some kind of preferable situation, with The &ad to Chess Improvement
the hope you will at some point be able (Alexander Yermolinsky)
to create something similar. A wonderful book built around Yer­
Ideal Sq11are: Most pieces on the board molinsky's own games. The book pre­
have squares where they - in the given sents the notion of Trends and illustrates
pawn formation - would be best placed very well how a 2350 player managed to
Primary Concept: A single unifying idea climb to the top board of the U.S. team
whose implementation would govern the simply by analysing his own games.
fate of a position. Example:
Reassessyour Chess
Lund-Hajnal (Jeremy Silman)
Budapest 2002 I am normally unhappy with books that
preach fixed methods of thinking but,
despite the Silman Thinking Techniq11e,
this is a wonderful book which I would
recommend to everyone with an Elo
rating under 1 BOO. The book succeeds
in delivering the absolute basics of
chess, the basics of which all of my pu­
pils - and, at times, even players at in­
ternational level - have too limited a
knowledge. I like all books that verbal­
ise what I asswne I already know, so I
can check it out...

Primary Concepts

Mark Dvoretsky, Jeremy Silman and I books, Reassessyour Chess and The Reassess
have a lot in common. We have all writ­ your Chess Workbook. These books are
ten books about how chess should be especially well designed for players under
studied and played - although, arguably, 2100, but my pupils above this level -
on different levels. We also have some­ and me, too - have allfound some ideas
thing else in common - the belief that all therein useful.
positions have some kind of governing But let us return to the question of
idea which is more important than all how to find the most important charac­
other ideas in the position. Dvoretsky teristic in a position. In a tournament
writes about it in his book Attack and game the way I normally try to do this is
Defence, Jeremy Silman works with similar. via candidate moves and some calcula­
ideas in How trJ Reassess your Chess and I tion. After this I know much more about
write about it in my book Excellisg at the position than I did in the beginning,
Chess. and then I know what is important to
Dvoretsky does not mention anything achieve.
about how you have to find this govern­ In training situations I use some very
ing idea, only that it is good to use it naive methods based on cognitive psy­
when you do. For top players this is suf­ chology. The key idea is that pattern rec­
ficient, but for lesser souls without a ognition is easy if the pattern is present
natural super-talent for chess Silman and in the short-term memory. Some ex­
I have different ideas to penetrate posi­ periments in the US have shown that
tions. patterns with no other relations than
Silman works with his own system of structural are easily transferred from one
imbalances. It is very useful as far as I area to another, strengthening the
can see, and I recommend anyone inter­ thought process and enhancing the abil­
ested in ways of thinking other than ity to solve complex problems. What this
blind calculation to read his two main means in terms of solving a positional
Ex c elling at Positio nal Chess

situation is that by first identifying con­ opment.

cepts and ideal squares for the pieces in a Nevertheless there are two things
given position we can bring this to the that I would normally consider doing
forefront of our mind. Then when we here - comparing pieces and finding
finally calculate we will do so with an ideal squares. In this situation, in order
unexpected level of accuracy and speed. to save space, I will just find the ideal
Of course we will calculate slighdy less, squares for Black: the aS-rook is good
but most oversights are performed in the right now, the knight might go to f4 but
first one or two moves in a given line, for now it is well placed, the bishop
and this is where we should improve our could hardly be better but might at
calculation. some point go to c6, the £8-rook should
probably be on d8 and the queen on b6.
Enough talk - let us look at a position. White has no good squares easily
available for his pieces. The queen can­
Borgo-Acs not improve and nor can the knight (on
Charleville 2000 f5 it would be quite lonely), but the
Sicilian Defence bishop might be better on gS, although
this takes time. Meanwhile the queen's
1 e4 c5 2 lbf3 d6 3 d4 cxd4 4 rook is simply miserable. Only the
lbxd4 lbf6 5 f3 e5 6 i.b5+ lbbd7 7 king's rook and the king are easy to im­
lbf5 a6 8 i.a4 d5 9 exd5 b5 1 0 prove (casding is coming).
i.b3 lbb6 1 1 ll:le3 i.c5 1 2 "W'd3 0·0 About exchanges. White should seri­
1 3 ll:lc3 i.b7 14 .i.d2 i.d4 1 5 lbf5 ously consider exchanging on � dS, for
a5 1 6 a4 i.xc3 1 7 bxc3 bxa4 1 8 with pawns on both sides of the board
i.xa4 i.xd5 1 9 lbe3 ll:lxa4 20 l:txa4 the bishop versus knight situation
would be beneficial for him. The ex­
change of queens might also be in
White's favour because it is more prob­
lematic for White to find a good home
for the queen than is the case for Black.
So as Black we now know how to place
the pieces and we know what we · want
to prevent. If we look at it from a static
point of view the correct move is
20 . .'ib6.
. The reason is simple - this is
how we want to place our pieces. This
also prevents lDxdS for at least a few
In this position it does not take a moves, homing in on the gl -a7 diago­
long time to work out that Black has the nal. The only problem is the hanging
advantage. The aS-pawn is a potential piece. However, there is a rule called the
danger, the doubled pawns are obvious 90% rule, which states that in 90% of all
weaknesses and Black leads in devel- situations the move which is correct for po.ri-

Primary Con cepts

tiona/ reasons works out tacticai!J. Here it P Nielsen Timman


makes us curious to see if the move we Sigeman & Co 2002

would like to play can be played. It Slav Defence
turns out that it works just fine for
Black. 1 d4 d6 2 c4 c6 3 lt:lf3 lt:lf6 4 lt:lc3
20 . . .'ifb6! 21 0-0 dxc4 5 a4 .i.f5 6 �e5 liJbd7 7 llJxc4
After the double exchange on d5 lbb6 8 lbe5 a5 9 g3
there is a simple check on b l , winning
the exchange and the game. Now Black
easily exploited his advantages to win
the game.
21 . . .l:tfd8 22 We2 .i.e& 23 l:th4 a4
24 c4 :d4 25 :h3 .i.d7 26 :g3
lt:lh5 27 l:tg5 �t4 28 Wt2 t6 29 :g3
:c8 30 c3 l:td3 31 'it'h 1 lt:lh5 32
lbd5 •xf2 33 :xt2 lbxg3+ 34 hxg3
Axc4 0-1

Perhaps it is not possible in these

limited pages to do full justice to this In this position Black has two main
idea. Although it has been fully explain­ concerns. 1) White is about to play i.g2
ed the transformation from an ideal to and e2-e4, and this might be annoying.
practical use is difficult. This is probably 2) Black needs to complete develop­
why Dvoretsky is more interested in ment. Timman, a truly creative player,
building up his pupils' intuition than in does not pay sufficient attention to
finding algorithms that work specifically these points, while Heine, one of the
for the club player. So the torch has top players of tomorrow (I hope), ex­
been handed to the rest of us. For the ploits Timman's carelessness with a
time being I am satisfied with being able combination of rapid development and
to make fire. Perhaps in the future I will simple threats.
attempt to create electric light. .. 9 . . .lbfd7?
One of the primary concepts in a Perhaps Timman had seen that there
position most often seen is develop­ was an earlier game with 1 0 f'bd3 here.
ment, which (of course) occurs in all However, the best continuation is 9. . .e6
games. But remember that it is not 1 0 i.g2 i.b4 with a balanced game in
meant in the sense that only develop­ Gurevich-Gulko, Salt Lake City 1 999.
ment is important in a position, rather 1 0 lbxd7!
something along the lines of: if you Gaining time.
ignore the need for development (or 1 0 . . . -.xd7 1 1 e4 i.g4?
improving your worst piece) you will I do not like this move at all. At the
suffer as a consequence. The following moment White has no problem weaken­
game illustrates how this can happen. ing his kingside slightly thanks to his

Excelling a t Po sitio n a l Chess

lead in development. 11 ...�g6 1 2 �e3 .txd4 2 1 llxd4 •f2 22 Aed1 •xh2

is still better for White (the knight is 23 •xb7 1 -0
exposed on b6). It is remarkable that a world class
1 2 f3 i.h3? player like Timman can end up in such
This is just bad. Now White identifies trouble by neglecting development and
a weakness on b7 (b6) and at the same failing to identify a chief weakness. This
time finishes his development. White is the danger of intuition and calculation
has a clear advantage after 1 2...J.h5 1 3 alone.
�e3 because after 1 3. . .e 6 there is 14 g4
J.g6 15 dS!, opening up the position In the Chapter How intuition develops in
with tempo, exploiting his lead in de­ his book Attack and Difence Dvoretsky
velopment. describes a training method he calls in­
1 3 i.xh3 'ifxh3 tuition training (starting on p. 67). The
idea is simple - you have about five
simple positions, slightly different in
nature, and you have to 'solve' them in
fifteen minutes. This, of course, helps
develop a number of different abilities
but, most importandy, it provokes intui­
tion in a way that can be compared to
muscle development in weight training.
I am a keen supporter of this combina­
tion of solving exercises and having a
good discussion about the solutions -
hence this book.
1 4 'ifb31 When I started chess coaching as
Developing the queenside with gain more than just a hobby I immediately
of time. understood that I needed some tools to
1 4. . . lla6 1 5 i.e3 ...g2? help explain decisions based purely on
Black continues to ignore both his positional considerations to players who
weaknesses and his poor development. are not blessed with natural, strong in­

1 5...'i!fc8 1 6 dS tbd7 17 llct is clearly tuition. But wherever I looked for such
better for White but there is still a game tools I found only outdated ideas. Of
to be played. Now White wins. course a good place to start is a
1 6 0-0-0 'i'xf3 1 7 l:the1 g6 Nimzowitsch classic, but the nature of
It was too late to save the game as positional chess seems to me to be far
the following line indicates: 1 7 ... e6 1 8 more complicated than the great father
dS! �b4 19 dxe6 0-0 20 e7 lle8 2 1 of the Danish chess tradition had envi­
�xb6 :Xb6 22 lld8 and Black loses. sioned. Nonetheless his main work, My
1 8 d5 System, is obligatory reading for anyone
Black is finished. hoping to nail down those key ·chess
1 8 . . .i.g 7 1 9 i.xb6 0-0 20 .td4 concepts. The book was published in

Primary C o n c ep ts

1 925, and since then there have been of a position. I have taken an example
other major works. In the 1950s a Rus­ from a recent book that I rather like, Can
sian named Lipnitsky published a book you be a Positional Chess Genius? by Angus
which can be translated as Problems of Dunnington.
Modem Chess Theory, where various as­
pects of chess 'rules' are dealt with. This
book has, unfortunately, never been
translated, but a Russian friend of mine
explained the content to me and it
seems that the book can be seen as a
deeper version of my own book Excel­
ling at Chess. In the 1 990s we had the
Mark Ovoretsky books. So far there are
about 1 0 of them, but I must honestly
say that I have lost count. Most impor­
tant of these are Positional P�, Training
for the Tournament P�er and a recent White to move
book entitled School of Chess ExceJJente 3,
Strategic P�. The others are, of course, Let us try to compare the pieces and
also great, but these three deal more from this make some deductions. We
with positional play. There are also · should compare pieces that are likely to
some works by Euwe and Kotov that be exchanged, so here the e2-knight and
are worth studying. c7-bishop are comparable, as are White's
But neither Dvoretsky's books nor bishop and its opposite number on e6
any of the others mentioned here gave · (as Black has no intentions of taking on
me the tools I needed to explain how very c4), leaving (by means of elimination) the
simple solutions were found, with the knights on c4 and f6. This might not be
exception of the principle of the worst what happens in the game, but that is of
placed piece (see page 31), which is ob­ less importance because the main idea
viously a useful tool in positional consid­ behind comparing pieces is to get a good
erations. Eventually I came up with sense of both the position and of what
some ideas that could be used to explain exchanges are likely to favour who. I
manoeuvring as something based on always do this from the top but often
more than individual solutions to indi­ you will find it useful to reduce the proc­
vidual examples. ess to some important pieces and pay
The main notions are those of primary less attention to others. In this case,
concepts (see chapter 1), comparing though, we consider the whole army.
pieces and ideal squares. Kings first - principally Black has a
safer king position as he has committed
Comparing Pieces no pawns and thus not created weak­
This is a simple exercise which can at nesses. The difference is minimal but is
times help to get a better understanding present nevertheless and should there-

Ex c elling at Po s i tio n a l Chess

fore be included in the comparison. tion Black is already threatening . . ...i.dS!,

The queen is well placed on c2, attack­ which would be the choice after a move
ing a potential weakness on f5, and there like 1 3 lbf4, e.g. 1 3... .i.xf4 1 4 gxf4 .i.dS!,
is no risk of being disturbed. The other when Black has fully equalised. So a
queen still has to find the right square, so natural conclusion will be to try to pre­
I prefer White somewhat. vent ....i.e6-d5.
White's queen's rook seems to have We also learned that Whlte has easier
easier access to a good open file, while play and that he should probably try to
the potential queenside minority attack play for an advantage in one way or an­
adding to its influence. This makes the other. This is apparent from the number
rook superior to the one on aS. of preferable pieces. Normally this is
White's other rook also has more something you get a better feeling for
possibilities than its opposite number. after having performed this little exer­
I prefer White's light-squared bishop, cise. Even thoughyou can carryo11t an overview
especially in view of the respective pawn of aU the pieces at a glance, yo11 will be able kJ
formations. For me it is obvious that increase the amo11nt ofpossibilities and concepts
Black favours an exchange of these two yo11 can see ifyo11 look at parts of the hoard
bishops due to the resulting weakened individllai!J. This example is typical. So is
light squares around the white king (the the next, where we come to finding ideal
exchange also trades a 'bad' bishop for squares.
White's wonderful bishop).
I believe White's knight is a little bet­ Ideal Squares
ter than the dark-squared bishop as the The subject of ideal squares is (again) not
latter has no active possibilities of its an exact science, rather it concerns how
own but can only hope for the exchange we get a better feel for the position.
- White, on the other hand, can choose From the point of view of a coach dis­
when (and if). However, there is poten­ cussing a position with a pupil, this is an
tial for the bishop to become strong, so excellent indicator of positional under­
it is not a clear choice. standing - or the absence of it.
Finally, neither the c4-knight nor I mentioned it briefly in Excelling at
Black's knight are too well placed, so I Chess, where I called it the Christmas
prefer neither. exercise. The key idea is that you ask
Now, what did we learn from that? your piece: 'Where do you want to go my
First we learned a little about who has little friend? What would you like for
the most room for improvement in the Christmas?' Remember that this does
position, and we also learned something not refer to what you would like to do
about what kind of improvement that overall in the position - that would be a
might be. The most obvious exchange to different exercise, which can also be use­
seek for Black is that of his bad bishop ful. But I like to cut everything down to
for White's good bishop, especially as small pieces before I perform a full
this would create weaknesses around the analysis.
white king. Actually in the diagram posi- White's king is fine and the queen is

Prim ary Co n c ep ts

ideal, safe behind its own pawn chain the knight can hop into e4, but to call
and hitting the £5-pawn. White's rook these squares ideal would be misleading.
would be better on c1 but, in the case of Nevertheless, things change, and these
a minority attack, might be perfect on pieces are not that badly placed.
bl, or even al. The king's rook looks So what can we do with this? Well, we
best placed on dl but it is not so easy to can easily identify White's worst placed
decided just yet. The bishop is perfect on piece as that which is furthest from its
g2 and the king's knight belongs on f4. ideal square. Therefore a possible plan
And it is as simple as that. But what could be to redirect the knight to cS in
about the knight on c4? Imagine remov­ some way. Here it makes little sense to
ing this piece and being able to put it go via c 1 and b3 as White would then be
somewhere else on the board: susceptible to any kind of opening up of
the position. So lbe5-d3-c5 seems to be
the right direction, and this also elimi­
nates Black's main idea of ....idS in
more than one way. Incidentally, it was
what White ended up playing in the

Budapest 1 995
Trompovs)g Attack

1 d4 lbf6 2 i.g5 d5 3 i.xf6 exf6 4

Where would you place e3 c6 5 lt:Jd2 .i.d6 6 g3 0-0 7 .i.g2 f5
a White knight? 8 tLle2 liJd7 9 o-o lt:Jf6 1 0 c4 dxc4
1 1 ll»lc4 i.c7 12 'i'c2 .te6
Personally, I would prefer �o place it This is the initial diagram position.
on cS. From there it disturbs the b�hop 1 3 l!Je5!
on e6 and attacks the weakest sp�t in The knight is on its way to cS, from
Black's camp, the b7-pawn. For Black it where it can exen pressure on b7. At the
goes like this: The queen would be per­ same time the text gives the other knight
fect on dS after the exchange of bishops. the option of going to f4 without being
The a8-r09k belongs on d8, the other removed by the enemy bishop. Other
rook on e8, the e6-bishop on dS and the moves prove to be insufficient, e.g. 1 3 b3
other bishop and the knight have no g6! and White will have to live with
obvious good squares - an important ... .i.dS, or 13 tLlf4?! i.xf4 14 gxf4 .idS!
observation. However, one should re­ and Black is no worse.
member that these last two are not par­ 1 3 . . .lbd5?
ticularly poorly placed where they are, This move makes little sense as the
and they still have some decent squares knight has no business on d5 and the
at their disposal. The bishop has d6 and square is now unavailable to the bishop.

E x c elling a t Po sitional Chess

Angus Dunnington gives some analysis 17 b4 a6 1 8 a4 ..td6 19 b5 axb5 20

and some comments in his book but, axb5 .:Xa1 2 1 .:Xa1 'ilc7 22 bxc6
unfortunately, he does not reach the bxc6 23 'ila4 l0d7 24 l086 .ixa6 25
depths of this position. I have taken the 'ilxa6 lObS 26 'ilc4 h5 27 l084 h4
liberty of analysing some alternatives. 28 .!Oc5 hxg3 29 hxg3 J:tc8 30 Ab1
Not very pleasant is 1 3 ... i.xe5 14 'ile7 31 .l:b7 'ilea 32 e4 i..xc5 33
dxeS liJd.S (or 1 4...lt:ld7 1 5 W'c31 'it'c7 16 dxc5 fxe4 34 i..xe4 l0d7 35 :&7
f4 when White has a clear positional �5 36 'ilc3 .:tdB 37 l:.c7 9e6 38
edge - Black can quickly find himself in �g2 .l:td7 39 J:tc8+ �h7 40 'it'a1 lld1
trouble, e.g. 16...£6?? 17 t!Jd4 1:ae8 1 8 41 'irxd1 9xc8 42 'ilh5+ 1 -0
l2Jxe6 :Xe6 1 9 i.dS! and White wins) 15
lbd4 g6 1 6 e4 fxe4 17 tbxe6 fxe6 1 8 One Move from the Ideal Square
'it'xe4, and although Black's position is I have often noticed that a piece is best
solid there seems little by way of future improved to the point where it is just
counterplay. If the queenside pawns start one move away from its ideal square.
to move they will only become weak, Only when our pieces occupy this posi­
and the knight is sitting pretty with no­ tion are they ready to be transformed
where attractive to go. into their perfect state. If you look at
Simply bad is 1 3...lbd7? when White the diagram at the beginning of this
has the brilliant computer-like 14 l2Jxc611 chapter you will see that the e2-knight
bxc6 15 i.xc6 tl)b6 (after 1 5..1tb8 16 dS and both rooks are all waiting, one
White regains his piece with interest) 1 6 move away from their ideal squares.
i.xa8 tbxa8 1 7 l:tfd1 and Black is quite Here is another example:
But more prudent than all this is a
simple move like 1 3 ...'i'e71?, bringing the
pieces into play. After 14 lbd3 (White
cannot play 14 lbf4 due to 1 4...he5 1 5
dxeS tbg4 1 6 'it'c3 i.c81 an d Black will
win a pawn for which White's compen­
sation is nothing special) 14...g6 1 5 ltJcS
i.d6 and Black is only slighdy worse and
can hope for a successful defence.
14 .!0d3! g6 1 5 llX:5 i..c87
The bishop should not be down here.
If White wants to exchange it Black White to move
should not be disappointed. Try to com­
pare the pieces; try to look at ideal In this famous position, from the
squares. Better was 1S.. .llb8. game Lasker-Capablanca, St. Petersburg
1 6 t003 .!0f6 1914, White's knight on e6 is already
How obvious it seems now that Black ideal, but let us take a look at the rest of
did not play accurately. The game ended: the pieces. The :d1 wants to reach a7,

Prim a r y Co n c ep ts

The other h7 and the king is better stood. That is why looking through un­
moving away from the same diagonal as annotated games from a database or
the bishop - in the game Lasker chose playing through the complete works of
the g3-square for the king, and that Averbakh will not seriously improve
seems to be a wise choice. The other your chess. Instead we should work
knight needs to find itself a good with annotated games and - even better
square, and in the game this leads to e4- - discuss positions with stronger players
e5 followed by iDc3-e4, from where it who know something about teaching.
rules the world. But let us take a look at Analysing your own games in depth and
how the game went. trying to understand the reasons for the
31 hxg5 hxg5 32 J:h3! mistakes you commit is, obviously, also
The rook is on its way to its ideal · a part of this. It is my hope these simple
square, h7, simultaneously vacating g3 tools will prove helpful for you in this
for the king. But now after task.
32 .:d7 33 �g3 �e8

White continues to improve his Improving your Worst Placed Piece

pieces slowly with A student of mine informed me that in
34 .:l.dh 1 i.b7? ! the recent book by Grandmasters Alex­
Then comes the final breakthrough. ander Beliavsky and Adrian Mikhal­
35 e5 ! 1 chishin, Secrets of ChessIntuition, there is a
The . pawn is o f little importance. chapter entitled Improving yo11r Worst
Control over the dark squares means Placed Piece. I immediately rushed off to
everything. buy the book since this is something I
35 . . . dxe5 36 .!Lle4 .!Lld5 37 .!Ll6c5! - have been preaching to all my students
Now Black is lost. Notice how the for as long as I can remember. How­
rooks have not rushed off to their ideal ever, the book was a major disappoint­
squares since they could do little on ment, although the idea is not. They
their own. After 34 ... �b7 Black was lost ascribe the concept to GM Makogonov,
in a sea of forks but it is still worth not­ but I have a feeling that players have
ing that White had prepared his pieces, been aware of it all the way back to the
one by one, to almost their optimum beginning of the last century. In fact I
before finally slotting them into ideal cannot recall where I got the idea from
squares all at once, so to speak. White but I think I actually pieced it together
now won easily. by myself, too. So I am sure thousands
37 . . .i.c8 38 .!Dxd7 .i.xd7 39 J:h7 of other people who think at the chess­
lU8 40 J:a1 �d8 41 .:taB+ .ic8 42 board have also done so.
.!Llc5 1 -0 But enough talk. Let us jump to the
ftrst example. It is not squeaky clean but
· It is my firm belief that intuition does does demonstrate the practical use of
not develop as random pattern recogni­ the idea in a tense situation. The posi­
tion, but rather the recognition of pat­ tion is taken from my first ever victory
terns previously investigated and under- over a 2600-player.

Exc elling at Positional Chess

term o r as less valuable. This is what a

computer does. Here is an example:

B russels 1 99 1

Black to move

In this position Black has no imme­

diate way to improve. Action on both
wings seems currently to be unjustified,
although Black is well organised. I must Black t o move
admit that I was completely lost in
terms of fl.nding a plan here. All I could This position is actually rather simple,
see was that Jonny might attack my e6- . yet at the same time highly complex.
pawn in some lines and that the rook on Black has some advantage but it is
a8 was not really contributing to my mainly based on time. If White had two
game. So for that reason I played the moves ('it>£2 and l:l.hel) he would have
innocent-looking 24 l:tae81 ? How the
... no problems whatsoever. Therefore for
game continued is not really relevant for Black it is a matter of whether he wants
this chapter, but I can say that I did not to play for an advantage with an attack.
at one moment regret the move, which The answer is, not surprisingly: Yes. So
is the kind of move one could imagine what is to be done? The key rule in at­
Petrosian and Karpov making. Instead tacking chess is that all pieces should
of having to worry about tricks later I join in the offensive. And here the least
protect my only unprotected piece in likely piece to play a part in the attack at
the position while slightly improving my the moment is the a8-rook. Thus the
worst placed piece. It is not great logic , correct plan is to get this rook into ac­
but good and healthy practical chess. tion in some way.
Let me try to be semi-philosophical 18 . . b5!

about the subject. We could look at the 1 8. .. tt:Jxg3+!? 1 9 hxg3 bS! is another
pieces in terms of points - as all of us way to play the same idea. It might give
did in our younger days - but count White more breathing space but still it
only those that are taking part in the seems dangerous.
action. We regard those with only po­ 1 9 'i'xb5
tential as worth nothing in the short- Here comes the hard evidence for the

Prim a r y C o n c e p t s

supremacy of the ... b6-b5 thrust. 1 9 placement (on the second rank) and
1i> £2 i s the standard untangling move, how to implement the subsequent plan.
but here Black wins time: 1 9 .. 1te6 20 This kind of treatment is presented by
...a3 ltlg4+!! 21 fxg4 ...f6+ 22 �gl Mark Dvoretsky in Attack and Deftnce,
ltlxg3 23 hxg3 l:tae8 and the final piece Jeremy Silman in How to &assess yo11r
enters the attack with deadly effect. Chess and myself in Excelling at Chess. We
After 1 9 ltlxbS l:e6 20 'it'a3 ltlxg3+ all explain it in a different way and have
21 hxg3 .:.bs 22 ltlc3 (22 'i!fd3 1i'a5) different methods of reaching this con­
22...ltlg4! Black has a very strong attack. clusion but, in essence, we agree.
1 9 .teS!? is the reason why ...ltlxg3 But back to the worst placed piece.
can be considered instead of 1 8...b5. What I disliked about the Beliav­
Then 19 ... b4 20 ltlbS l:.e6 21 ...a4 g4 sky /Mikhalchishin book is that it is
gives Black good attacking chances. mainly just a collection of simple exam­
1 9 ...�xg3+ 20 hxg3 l:b8 21 'i'd3 ples. There are few ideas in the book
l:txb2 and it feels like even fewer thoughtful
Black has a clear advantage and went annotations. In Chapter 1 2 (p. 1 03-107)
on to win as follows: they fonnulate the theory concerning
22 l:td2 'i'b6 23 g4 'Wb81 24 �d 1 this idea as follows:
'Wg3 25 tbf2 Ab6 26 lth3 l:tbe6 27 The late Makagonov (one of Kasparov's
Ad1 'i'c7 28 'i'd2 l:te3 29 l:tc1 ? I first trainers) was a strong positional p�er
'Wf4 3 0 ltd1 g 6 3 1 �g1 �g7 32 and formulated some usefol general principles.
'ifc1 l:te2?? (32...ltld7!, heading for c4, The mostfamous of these is that, in balanced
is the quicker route to victory) 33 'i'xf4 positions, when neither side has any direct
gxf4 34 tbd3 g5 35 a4 l:ta2 36 g3 th"ats or concrete plan, it is necessary either to
fxg3 37 tbc5? g2 38 l:tg3 l:tee2 39 nlocate your IJIOrst plaad piece to its best
�b3 l:teb2 40 tbc 1 l:txa4 41 /Od3 squan, or to exchange it off. (p. 1 03).
l:tba2 42 Axg2 Axg2+ 43 �xg2 This description would perhaps fit
l:txd4 44 �2 �d7 45 �e3- l:ta4 46 well with my game, but the Short game
l:tc1 tbb6 47 l:tc7 tbc4+ 48 �e2 does not really fall into the description
l:ta2+ 49 �e1 l:ta3 50 �e2 l:ta2+ of a balanced position with no threats,
51 �e1 aS 52 f4 gxf4 53 tbxf4 tbe3 unless you choose to see this as a very
54 l:ta7 a4 55 g5 a3 56 g6 �g2+ superficial evaluation, of course. The
57 tbxg2 l:xg2 58 l:txa3 fxg6 59 main point in the Short game is that the
l:ta6 �h6 60 l:td6 l:tg5 61 �2 �h5 absence of the hl -rook from the pro­
62 �3 �h4 63 l:a6 �h3 64 �f2 ceedings allows for Black to establish a
l:tg4 65 l:ta3+ �h2 0-1 kind of power play for some time, af­
In this example the rooks on a8 and fording him the possibility of obtaining
h1 had something in common in that a decisive advantage. Actually a devd­
they both lacked scope and therefore opment ·advantage is best understood
had little value. Black proved an advan­ with the help of a situation taken from
tage by improving his worst placed ice hockey where, for a limited time, a
piece and, in doing so, found the ideal team can have more players on the ice

Ex celling at Positio nal Ch ess

than the opposition, a situation that great). It takes some thinking and a free
must be exploited to the maximum. mind to find the ideal square - or best
The final, most clear-cut, example of square, as B & A call it. But I prefer to
improving your worst piece is taken use the Christmas exercise - where
from the Beliavsky/Mikhalchishin book. would this bishop be placed if the wish
were granted? The answer: b6!
lvanov-Benjamin 1 4 . . . .i.d81 1 5 ••2 c6 1 6 l:.d1 i.c7
Jacksonville 1990 1 7 h4 ....7 1 8 g3 �g7 1 9 .!i:lf3 a4
20 h5 ..1a5 21 l:.c1 'ird7 22 Afd1
l:.ae8 23 ¢'g2 f5 24 exf5 .:l.xf5 25
.!Lle4 .!Llxe4 26 'irxe4 l:.ef8 27 l:.d3
l:xh5 28 .!i:lh4 l:.xh4 29 gxh4 .:l.f4 30
'ire2 'iff5 31 c5 cxd5 32 cxd6 i.b6
33 l:.f1 e4 34 .:l.g3 d4 35 'ird2 e3
36 'ire1 'ird5+ 37 �h3 'ire6+ 38
�g2 'ird5+ 39 f3 'ifxd6 0-1 --

After completing this chapter I was

made aware of the following quote
from Dvoretsky's book, Positional P�
Black to move In positions of strategic manoeuvring (where
time is not significant) seek the worst placed
Here you could argue that the worst piece. Activating that piece is often ·the most
placed piece is the fB-rook as it has no reliable Wt!J ofimprovingyourposition.
scope and even no moves. This would,
of course, be partly true, but as the pri­ Explanation of Terms
mary way for the rook to gain scope is Ideal Square (the Best square or the
by . .. f7-f5, it is not too bad. Moreover, Christmas square): The square from
how do we decide which piece is the where a defined piece can have the
worst placed? I have found that a good greatest possible influence. Not defined
way to do this is simply by trying to find by the possibility of getting there, but
the ideal squares for all the pieces. In more formulated as an ambition.
this case the only piece with which we Undefended pieces: The cause of so
would have problems is the bishop, much misery in the chess world today!
which has no scope (as we cannot hope Pour p�: A term from ice hockey,
that White is kind enough to open up where a player is sent off the ice for two
with f2-f4). Therefore we have to find a minutes _(due to a penalty). The remain­
way for this piece to contribute to the ing six players against five is then the
action. Bishops tend to be better off powerp� situation. In chess this can be
being some way from the centre in or­ applied to a position where a player has
der not to be disturbed (of course an one or more pieces on the board not
untroubled bishop in the centre is taking pan in the actual struggle.

Defining Weaknesses

All positional chess is in some respect pawns. The three weaknesses are d4, c2
related to the existence of weaknesses in and b2, Black's pawn being very weak.
either your or the opponent's position. Due to the placement of the minor
However, it is surprising how many play­ pieces Black is, in fact, already losing.
ers know little about defining weaknesses The c2-pawn is not so weak thanks to
and how to relate �o them. In this chap­ the protection it receives from the
ter I will provide some examples of what bishop. Note that it is an important de­
a weakness is, and some ground rules tail that this bishop is very well placed
about how weaknesses can be defined. where it stands. In principle the b2-
Take a look at the following position: pawn is a little bit weak as it is without
any defence. But here a main rule
Timoshenko-Chernov comes into play - A 111(Jaknm is onfy weak
Bucharest 1 993 if it can be exploited/ attacked Here the
weakness of b2 is not so important be­
cause none of Black's pieces is currently
able to attack it. So both c2 and b2
could be weak according to theoretical
definitions but, in practice, they are not.
Only d4 is weak, and mainly so because
so many white pieces are ready to attack
As we shall see in the game Black
also has some problems with his king­
side in view of the two bishops pointing
in that direction. White cannot justify an
This is a posicion where the defmi­ attack on the king at the moment but all
tion of weaknesses mainly concerns weaknesses will count at one time or

Ex c e lling s t Po sitio n s / C h e s s

other. In fact White undermined his Here White could have exploited the
opponent's weaknesses. weakness with 20 f3 .i.e6 21 �h1 �h8
1 6 J.e5 ltJxe5 1 7 l:xe5 (jje7? 22 .ifl and the d4-pawn is lost.
After this move White has an easy to 20 . . .i.e6
way to simultaneously exploit both the After 20 ... a6 21 .:.e7 :.e8 Black tries
weakness of Black's king and the vul­ to profit from White's ostensibly weak
nerable d4-pawn with the aid of a dou­ back rank, but calculation proves that
ble threat. Of course only one target can after 22 l:lxe8 l:txe8 23 llxe8 'i'xe8 24
be addressed. lbxd4 'i'e1+ 25 i.fl lLle4 26 'i'e3 Black
1 8 'i'e41 l:fd8 19 'Wh7 + �f8 20 is a pawn down in the endgame without
'W'h8+ (jjg8 21 :ae1 compensation.
White's attack is irresistible. After 21 Ad 1 ..tg4 22 l:.de1 ?
21...a6 (or any other waiting move) Again 22 f3 would be good, winning
White has the following decisive attack: the d-pawn. For some reason White
22 i.h7 �e7 23 Le6+! fxe6 24 'i'xg7+ chose not to weaken the diagonal down
etc. Instead Black tries to prevent this, to his king, but exact calculation, as well
intending to defend the knight on g8 as the absence of a dark-squared bishop
with ... f7-f6. With this in mind White in his opponent's arsenal, should have
reacts with a little combination. convinced him to take the pawn. Now
21 . . . i.d5 22 l:.xd5! 'W'xd5 23 (jjc7 the game went on for another 1 34
'W'd7 24 .i.h7 1 -0 moves, with winning chances for both
In a more recent game Black de­
fended slightly better: This example mainly concerned weak
pawns. Indeed technical positions often
Bromann-Raetsky relate chiefly to the weakness of pawns.
Denmark 2002 But we also saw the weakness of the
back rank (a check was annoying in one
1 7 ...(jjf6 1 8 'W'f4 Afd8 1 9 l:d1 i.g4 line) and the weakness of the king
{'i'h7+ decided the first game).
One thing is clear from all of this:
Weaknesses are aiWt!JS defined according to the
pieces that remain on the board. Nimzo­
witsch pointed out many years ago that
the domination of an open flle has little
importance if all pieces have been ex­
In the f;llowing example Black seizes
control of the centre and eliminates one
of his own weaknesses by accepting
doubled pawns. This is an excellent
20 l:l.de1?\ example of how tactics and positional

D e fining We a k n e s s es

goals can unite m a great display of positional ruin.

chess. 1 3 . . . b3 1 4 i.d1 ltJc5 1 5 i.xf6 exf6!
After this slightly WlUSual recapture e4
Videki-HIIIarp Persson is a weakness, and we have a good rea­
Budapest 1 996 son why black did not give a check on
1 6 i.e2 .l:.e8 1 7 i.xc4 lbxe4 1 8
lbxe4 lbe4+ 1 9 i.e2 'iie8 20 li:ld4
f5 21 l0c6 i.h6 22 'ifd1 a5 23 �f1
Taking over the initiative and the two
bishops. White seems to be lost here.
24 'ifxe2 .ta6 25 c4 'ifxe2+ 26
'iti>xe2 .txc4+ 27 'it'd1 �8 28 l0d4
.txd5 29 f3 J.e3 30 lbb5 .l:.c8 31
.:e1 f4 32 lLla3 .ib7 33 .l:.xe3 fxe3
34 �e2 i.a6 + 35 �xe3 :te8+ 36
In this position the weakest spots in <ltd2 lle2+ 37 �c3 l:xg2 38 ltJb5
'White's camp are the pawns on b2 and lbh2 39 .l:.d1 <lte7 0-1
c3, the latter being under direct pressure
from the b4-pawn. It is important to The next example has some similari­
note that the fall of the c3-pawn would ties - which should be quite obvious.
lead to further losses for 'White.
1 2 . . . c41 Feher-Titov
Black needs to act quickly in view of Budapest 1 990
the temporary nature of the weakness.
Given the opportunity, White's next
move would have been 13 c4!, keeping
the queenside pawn structure intact.
This would be followed by b2-b3 (after
moving the rook away from the long
diagonal), leaving the rest if the battle to
be carried out on the kingside.
1 3 .tc2?
1 3 i.xc4 bxc3 14 bxc3 (1 4 '6'xc3
tt:he4) 14 .. 1lc8 can be very awkward for
White because 15 i.d3 liJxdS! sees Black
win at least a pawn. But White can tty a Here Black is fully justified in recap­
:nove like 15 J.b3!? with the idea of turing with the pawn - opening the e­
1 5 . ..ltJxd5 16 J.xdS! and compensation ffie, getting a square for the c8-bishop
for the queen. Overall White had to go with the break ... f6-f5 and, as a result,
for something like this, for he now faces simultaneously eliminating White's cen-

Exc e lling a t Positional Ch e s s

tre. Again the weakness of White's change not only does the dS-pawn re­
queenside is worsened by the introduc­ strict the bishop but there are weak­
tion of the bishops. At the same time d3 nesses on the queenside that did not
is exposed as a potentially weak square exist a move earlier.
in the white camp. Notice that this 1 4 ...j.xd5! 1 5 exd5 tbb4 1 6 c3?
square is weak only because of the well­ 16 l::tc t was a lot better.
placed knight on cS. Black won the 1 6 . . . /0d3! 1 7 •xd3 'i'xb3 1 8 'i'b5?!
game easily: White has been careless for some
1 5 exf6! 1 6 0-0 f5 1 7 b4?
•.. moves and now he accepts a weak pawn
Black was, of course, a lot better on bS which will either be lost or, as in
anyway, but after this move he is just the game, the cause of much misery.
winning. Instead an admittedly unattractive move
1 7 . . . �xe4 1 8 'i'b3 lilc3 1 9 .l:lbe1 a4 such as 1 8 l:fb1, intending 1 9 l1a3, was
20 ...c4 j.d7 21 ...f4 .l:lc8 22 .i.d3 probably necessary. But this just shows
...f6 23 �c4 ..i.b5 24 �xd6 ..txd3 how poor White's prospects have be­
25 �xeS :ctxc8 0-1 come. Nevertheless, the solid nature of
the position might still give White a
Weaknesses obviously have a major chance to save the game.
importance in the middlegame, and this 1 8 . . Ji'xb5 1 9 axb5 b6 20 .l:la4
is indeed a book about the middlegame The rook is very much misplaced out
but, usually, their creation is in the mid­ here, although it is difficult to find an­
dlegame and their exploitation is in the other move. Now comes a very nice
endgame. One of these endgames is reply, the like of which is seen in many
that of opposite coloured bishops, positions where there is nothing much
which people know remarkably little going on.
about despite the fact that there is not 20 .. . .1:lc51
so much to know. Opposite coloured
bishop endings are all about weak­
nesses, passed pawns and domination,
as will be well illustrated by Black in the
next example.

Donoso Velasco-5ilva
Santiago 1994
Sidlian Defence

1 e4 c5 2 lilf3 li:)c6 3 d4 cxd4 4

li:)xd4 g6 5 lilc3 .ig7 6 lilb3 �f6 7
j.e2 0-0 8 0-0 d6 9 i..g 5 i..e 6 1 0 Black forces White to push the c­
Wh1 l:tc8 1 1 f4 l:te8 1 2 i..f3 aS 1 3 pawn, which creates new weaknesses
a4 Wb6 1 4 tbd5?! that could prove problematic. Beca��se
I do not like this move. After the ex- these weaknesses will persist for a very long

D e fining Weak n e s s e s

time th� are called permanent (or static) through the game. The immediate ad­
weaknesses. It is often the case that one vantage is the opening of the e-file.
player willinvest time in order to induce 24 :.xeS+ 'it>xeB 25 g3
permanent weaknesses in the enemy In a fabulous article entitled From the
camp, as Black does here with 20 .. ..:cs. Simple to the Complex in the book Tech­
21 c4 l:tc7 nique for the Tournament P�e:r, Mark
Here Black could also have played Dvoretsky explains that the side trying
21 ...lbd7 to prevent the exchange on f6 to win an opposite coloured bishop
and transfer the knight to cS. White ending should try to place his pawns on
would be in a bad way after 22 b3 llc7 the opposite colour of his bishop, while
23 lie 1 .!Des 24 lla3 �f8 when his the side defending should try to place
pieces are clearly not very well placed his pawns on the same colour as his
and his pawns are more susceptible to bishop (p.89). The reason for this is
attack than Black's, and therefore weak. that, effectively, the board is divided
22 :.e1 �fB into two 'camps' of equal size: light and
White elects to enter an opposite dark squares. To win the game you need
coloured bishop ending which, unfor­ to take over some enemy territory, and
tunately, is close to losing. to advance a pawn you need to cross
23 .i.xf6?1 just as many light as dark squares. A
common defensive set-up in endings
with opposite coloured bishops is that
of a fortress. The idea is simple - the
bishop and a limited number of pawns
can be enough to gain full control of
over half of the board and, in doing so,
prevent the remaining pawns from ad­
vancing. To win you have to avoid this
situation and create passed pawns. The
only way to do this is to fight the
In this example 25 f5 l:te7! 26 l':al
23 b3 lDd7 24 l:ta3 lDcS would have l:te3 would give Black a winning posi­
transposed to the note to Black's 2 1 st tion thanks to his well-placed rook. But
move. after 25 g4! White would probably be
23 exf6!!
... better off than in the game. The double
This recapture is probably rather ob­ thrust of the pawn is logical since it ad­
vious if we retrace our steps and look at dresses Black's next, after which Black
the previous examples, but I am sure has a lasting positional advantage on the
that most players would have played kingside.
23 . . ..i.xf6 on auto-pilot. The different
. 25 . ..f5
advantages of the text will become ap­ The weaknesses in Black's kingside
parent, one by one, as we progress pawn structure are f1 and h 7, but there

Ex c elling at Positional Chess

is obviously no way for White to attack well as the passed pawn on the other
these, so Black can manoeuvre freely. flank combine to be decisive. The funny
The same cannot be said for White, thing is that if there were a black pawn
whose Achilles Heel is the h2-pawn. on b6 the game would probably still be
Note that Black's pawn is far more use­ drawn because White would defend his
ful on f1 than e7. weakness on g3 while infiltration by the
26 b3 l:e7 27 b4? king on the queenside is not possible.
After this I have no faith White's po­ Black does have a plan involving .. f1-

sition. Black gets a passed pawn and a f6, ... g6-g5 and ... h5-h4 but whether this
passage for the king to penetrate the is enough to win the game is not clear. I
queenside. As we shall see it would would guess not. The position would
probably be better to just lose the pawn. not count as that of two weaknesses as
White could activate the rook with 27 the immobility of the king is also a
:a21 :e 1+ 28 �g2 :b 1 29 .l:.e2+ �d7 weakness.
30 :e3 i.d4 31 .:td3 i.cS when Black As it is the position counts as one of
clearly has the advantage but still has to two weaknesses - more of this below.
ftnd a winning plan. 36 <l.>d3 i.g 1
27 l:te1 + 28 �g2 i.c3 29 bxa5
..• 36 ... J.f2! was even stronger. But
bxa6 30 .:ta2 White lets Black create a passed pawn
30 cS dxcS 31 .:tc4 J.d4 offers White on the kingside anyway, so there was no
no hope. reason to force him to do it.
30 . . .�d7 31 �f2 l:a1 1 32 l:xa1 37 g4 fxg4 38 hxg4 h4
i.xa1 33 i.d1 i.d4+ 34 �e2 rj;;c 7 Two distant passed pawns are
34...J.g1 35 �fl Axh2 36 'iPg2 is ob­ enough to win.
viously too soon, but the white king can­ 39 f5 g5 40 Wc3 �b6 41 <l.>b3 <l.>c5
not effort to step further away. 42 <l.>a4 'it>b6 43 i.f3 i.d41
35 h3 h5 Before infiltrating with the king Black
improves his bishop to an ideal square.
44 i.h1 i.f6 45 i.g2 ..tdB 46 i.h1
�c5 47 Wb3 Wd4 48 ..tg2 ..tb6 49
i.h1 f6 50 i.g2 <l.>e3 0-1

Creation of Weaknesses
An important part of positional play is
to create permanent weaknesses in the
opponent's camp. The idea is that in a
position with many weaknesses the de­
fending side will be so busy defending
that the protection of the accwnulated
Although level in terms of material weaknesses will eventually prove impos­
the situation is beyond salvation for sible to carry out. This is why we often
White. The weakness of the kingside as see strong players not trying to win

De fining Weaknesses

weak pawns immediately but instead have 22... a6!, blockading the pawns,
allowing them to survive until they can whereafter b4 is a potential weakness.
be picked up without having to make And 22 bS would then be met with
concessions. 22... cxb5 23 axbS i.b6 with a playable
The following examples are typical. position.
22 .tel '*g6 23 .1d4!
Karpov-Lautier Typical Karpov. Before he carries out
Dortmund 1 995 his own plan he makes a slight im­
provement to his position. Here he ex­
ploits the enemy queen's lack of breath­
ing space to improve his pieces and ex­
change Black's only defender of the
queenside, thus facilitating the execu­
tion of the minority attack.
23 . . . .txd4 24 ti:lxd4 .td6 26 e4
.te6 26 tLlf3 '*14 27 b6!
Now it is time. After this break Black
is doomed to finding himself with a
weak pawn on the queenside.
27 . . . axb6 28 axb5 lbe6
Nothing seems to be going on in this Probably Black should have changed
position. Now Karpov starts what is a the course of the game here with
standard plan in these kind of Queen's 28...i.xh3!? 29 eS .ixg2 30 �xg2 lt:Jxe5
Gambit positions - the minority attack. 31 ttlxeS 'i'xeS with compensation for
The iclea is really simple. By advancing the sacrificed piece. I have a feeling that
his two pawns on the queenside White white is still slighdy better, but I think
aims to create a weakness among that the came is quickly spiralling out of
Black's pawn majority. In this way control for Black, and immediate action
White can create a target on which to was needed.
focus his offensive operations, while 29 �xe5 ...xe5 30 bxc6 bxc6 31
Black has problems creating anything ltac1
on the kingside, where White is well The minority attack has been exe­
represented. cuted to perfection. Black is left with a
20 b4! �g6 21 a4 weak c-pawn, to which his forces willbe
White is now threatening to create a tied in defensive duties. Karpov makes
weakness with a4-a5-a6, which will un­ the most of his advantage thanks to
dennine the support of the c6-pawn. excellent technique.
21 . . .a6 31 ...l:d6 32 '*c3 tl\d7 33 .te21
21 ....i.c7!? is a possible improvement. A forcing line, leading to problems
The idea is that White wants to see what for Black's bishop.
Black is doing before he selects a pawn 33 l:xd1 + 34 .txd1 '*xc3 36
. . ..

advance. After 22 aS Black would then l:xc3 J:.c8 36 f4 f6 37 �e21

Ex c elling a t Pos itio nal Chess

Improving the worst placed piece. I chose to exchange queens and go

Black is probably lost already. into an endgame, predicting the follow­
37 . . . c5 38 it:\d4 i.f7 39 i.g4 :tea ing pawn formation:
40 it.'lb5 lt'!b6 41 l0d6 lba4 42 :ta3
:as 43 e5 i.d5 44 e6 1 -0
Black resigned in view of 44. .r�>f8 45 .

lte3! :a7 46 l2Jc8 etc.

From one of my own games:

Copenhagen 1997

Here White will be able to create a

second weakness in the enemy camp
when he wants, for after g3-g4 he will
either exchange on hS to leave a weak
pawn there or he will himself make an
outside passed pawn, which, indeed,
does count as a weakness. Actually I
think the rule would be easier to under­
In this position Black has just pushed stand if it was called the rule of two advan­
with 29 ... h5?!, intending to generate tagu. Here it is of great importance that
counterplay on the kingside. This move the bishops remaining on the board
is poor for two reasons. First there is 30 operate on the light squares. If they
'i'd2!? with the idea of 31 'i'gS, illustrat­ were dark-squared the h4-pawn would
ing how abandoning the gS-square has be weak.
led to the weakening of the g7-square, The idea of the rule of two weak­
thereby endangering the king. This is a nesses is that, in order to win in the
weakness, of course, only in view of the endgame your opponent needs to have
well placed knight on f5, and probably two weaknesses. I have said that this
not the worst downside to 29 ... h5. should be called two advantages be­
A bigger problem is that in the long cause having a passed pawn or an active
run White will be able to create a passed king also counts. In fact everything that
pawn on the kingside. Consequently I could be important seems to count.
played a move that, after the game, my This makes the rule a bit fuzzy, but it
opponent claimed was a mistake. A still has great practical use. Of course
move he did not understand. However, one weakness is occasionally enough,
as we shall see, my evaluation of the but in the majority of cases it is the right
position was probably correct. strategy to induce a second weakness in

D e fining We B k n e s s e s

the opponent's camp before setting as I did not believe that this would in
about exploiting the first. any way be sufficient to counter the
30 'ird4! 'irxd4 threats on the kingside and in the cen­
Black has no good squares for the tre.
queen so the coming endgame is practi­
cally inevitable.
3 1 tbxd4 .ta4 32 l:.c 1 �f8 33 f3
�7 34 �2 lilc5 35 .tf1 .td7 36
l:lxe8+ l:txe8 37 l:ld 1 lba4 38 l:td2
g6 39 .te2 tDb6 40 h4 l:le7 41 .td3
�g7 42 tbc2 l:te8 43 lbe3 .tc6 44
.te2 Wf6 45 l:ld4 <iPe6 46 .td1 lidS
47 .tb3 l:.d6

50 .. .a4 5 1 .tc2 a3!?

Otherwise the a-pawn might also be
weak, sooner or later.
52 bxa3 .:!.dB 53 �b3 :as 54 h6
l:.xa3 55 h6
With the threat of 56 g5! etc.
55 . . . g5

White has now positioned his forces

in the best possible way. The dS-pawn
is under considerable pressure and
Black must constantly watch out for the
break with c3-c4. All of this makes it
the appropriate time to create a second
weakness, this time on the kingside.
48 g4! hxg4 49 fxg4 a5
Black is tired of waiting and seeks ac­
tive counterplay. In practice this makes
a real difference, although it does not 56 l:td 1 ?
change the evaluation of the position. Not throwing away all of the advan­
To alter the direction of play is essential tage but missing an easy win. After 56
for Black. l:.b4! lbd7 · 57 c4! Black has no good
60 1i>g317 moves left and will soon find himself in
I allow Black to generate counterplay a completely hopeless situation.
on the queenside at the cost of a pawn, 56 .. . l:.a8 57 :n l:.hS 58 tDts .i.d7

E x c elling a t Posi tional Chess

59 �c2 �e5 60 :th 1 ? example of how these two kinds of ad­

Another grim mistake. After 60 .:e1+ vantages contradict each other.
\tf6 61 lt:lg7! i.e6 62 ctJhS+ �e7 63
Itb ! ctJd7 64 h7, a s given by Donev; Boe·Aagaard
W'hite still ha s a large advantage due to Denmark 1 992
his passed pawn. I feel that Black does Scotch Game
have genuine chances ro save the game
here but this line was still the only win­ 1 e4 e5 2 ltJf3 4Jc6 3 d4 exd4 4
ning try. I would have played it had T ttlxd4 .tc5 5 �e3 1i'f6 6 c3 0ge7 7
seen 61 lt:Jg7!, obviously. .i.c4 b6 8 0-0 i.b7 9 li:Jb5 0-0·0 1 0
60 . .. .ixf5! 61 .ixf5 �a4 62 h 7 i.xc5 bxc5 1 1 lbd2 a6 1 2 �a3 d51?
�xc3 63 �3 % - % After this move Black's pawn struc­
Koneru-loseliani ture on the queenside is in ruins bur the
India 2002 lead in devdopment is also important.
1 3 exd5 tZ:Ixd5 1 4 'iff3 �e51 1 5
'ifxf6 gxf6

In this position White played the very

stron g 1 6 1Lf5! and claimed an advan­
tage. Humpy Koneru writes the follow­ Black has six 0) isolated pawns, two
ing in New In Chesr. 'Although all Black's sets of doubled pawns and hardly any
queenside pawns are on light squares, dark squares yet he might even be better
Black's light-squared bishop is useful to here. TI1e open ftles have no long-term
protect the potential weakness on c6.' value, neither has the threat of lt:Jf4,

This simple point is the logic behind but at the moment these factors fully
India's strongest. female player. She is compensare for the weaknesses In fact

indeed right, and won the game in com­ this is quite common - one player has
fortable style. the long-term advantages and the other
the shon-term advantages. The game
Permanent versus Temporary now ended in a draw:
Advantages 1 6 .li.xd5 .ixd5 1 7 f3 i.xa2 1 8 tZ:Ie4
This heading could be the theme of a ie6 1 9 tt:lxc5 Itd2 20 ltf2 :hd8 21
full book! Here I v.-ill just give a brief .=.xd2 l:txd2 22 lDbS ih3 % ·%

Squares - And How
Pieces Exploit Them

In this chap ter I will feature a few basic ing itl This makes it more vulnerable
truths about positional chess - aspects ilian the rest of the pieces, and for this
of the game in which I have seen rather reason the knight usually needs a fixed
capable players fail from time to time. I sttongpoint somewhere in the centre,
d o not intend to provide a compl e te from where it can exercise maximum
picrure as I expect that a number of pressure on the opponent's position.
readers understand most of this already, This kind of square is ideal if .it is right
but as the points in this chapter are es­ in front of one of ilie opp onen t' s pawns
sen tial and rather uncontroversial - in this way the knight can use ilie
truths, I feel that we all should know opponent's pawn as a shield from rooks
them. So please read on when you feel I and queens. A good example of this is
am insulting your intelligence. the following:

Knights and Squares Mohr�Volokitin

The knight is the weakest of the minor Portoroz 2001
pieces for a specific reason - it has no
long-range possibili ties. As control of
the centre is a main part of positional
ches s it has been said that 'knights on
tl1e rim are dim' because only when a

knight is near to the centre can it con­

trol s quares in this busy area. Another
characteristic of the knight that makes it
weaker than other pieces is its silly walk
(this is, of course, also its strength).
Thanks to the \Vay it moves around it
can never take a piece that is threaten-

Ex c e lling a t Pos i tio n a l Chess

In this position Black played: 1 7 lba4 .:tea 1 8 l:l.ac1 c4 1 9 lbb6

1 3 . . .lbe51? :tc 7 20 1i'b4 f5 21 lbxc4 i.h6 22
This move has both good and bad lbe3 .:txc1 23 :txc1 fxe4 24 i.g4
points. The minuses are that the long �h8 25 .te6 1rh4 26 g3 'itd8 27
diagonal for the bishop gets closed and l:.e1 .:tf3 28 i.g4 .l:f8 29 i.e6 .:tf3
White gets a protected passed pawn on 30 i.g4 .l:lf8 � -1h
dS. On the plus side White has to give The d6-square was a very good post
up his two bishops (or lose a tempo), for the knight in this game, but it could
and the newly vacated d6-square be­ have been viewed as a little passive as
comes an ideal outpost for a knight, as well. Take a look at the following ex­
can be seen in the next diagram. Which ample.
of these pluses and minuses is most
important in this position is not obvi­ Wells-Aagaard
ous, even after seeing the rest of the Copenhagen 1996
game. A quieter move was 13 ...'ife7. Queen's Indian Defence
1 4 .txe5 dxe5 1 5 a5!
A strong positional move, designed 1 d4 lbf6 2 ll:lf3 e6 3 c4 b6 4 a3
not so much to prevent ... b7-b5 (which J.b7 5 ll:lc3 d5 6 .tg5 dxc4 7 a4
could be difficult to achieve) but rather J.e7 8 1i'c2 lbc6 9 ltd 1 ll:la5 10
to isolate the c-pawn from the rest of llle 5 0-0 1 1 lbxc4 ll:lxc4 12 i.xc4
the queenside pawns, and thereby make c5 1 3 d5 exd5 1 4 exd5
it weak.
1 5 . . .lbe8 1 6 1i'd2 lbd6

This was, at the time, a critical line in

the Queen's Indian. Later White found
Here the black knight is as good as it more aggressive ways to play the posi­
gets. From d6 it exerts pressure on e4 tion, although I have a feeling that there
and c4, as well as being able to help is general equilibrium.
with the breaks ...b7-b5 and . . f7-f5. The
. In this position Black could play
game is probably somewhat balanced, along the lines of the previous example
with White, if anyone, a little better due with 14 ...llJe8?1, which I believed was
to the weakness of the c-pawn. the · right move in the post-mortem

Squares - A n d Ho w Pie ces Exploit Them

analysis, but after 15 �e3 lL!d6 16 .i.d.3 1 4 . . . i.d6? 1 5 i.e2!

h6 17 0-0 I did not like Black's position. A true grandmaster move from my
The problem is that none of the pieces grandmaster opponent. Now the pin
is sufficiently active. The knight might and the strong d-pawn become the
look pretty on d6 but, compared with most important aspects of the position.
the previous example, influence over e4 And, importantly, the bishop finds a
and f5 has little importance, while the better square than c4.
absence of pawns on a6 and e5 makes it 1 5 0-0? �xh2+!, winning a pawn for
easier for White to remove the knight Black, was my simplistic threat.
from d6 (and thereby improve the value 1 5 . . . .*.e5 1 6 0-0 a6 1 7 lL!e4 .llc8 1 8
of the passed pawn - or at least tie d6 :a7 1 9 lL!xc5 ..txb2 20 lLle4 :tea
more black pieces down to the blockade 21 'ii'xb2 :xe4 22 .*.e3 lL!d7 23 i.f3
of the pawn). All in all not a promising :e5 24 i.d4 :b5 25 'ii'e2 lL!c5 26
position. But what else? i.e& :as 27 :te1 lL!e6 28 'ii'e3
14 ...lL!g4 is a suggestion of Fri� but 'ii'xd6 29 i.xb6 'ii'xc6 30 :dB+ lLifB
after 15 i.xe7 'i'xe7+ 16 �2 (16 'ti'e2 3 1 'ii'e8 'W'xe8 32 l:texe8 h6
'l'h4 and. Whit� has problems complet­ Here something funny happened. My
ing development) 16 .. .'ii'e5 White has opponent had only ten seconds remain­
17 1i'e4! with an endgame advantage ing with which to reach move 40 and,
due to the strong passed pawn. Perhaps trying to make his move, he dropped
1 6 . 'ii'h4!? is better, but I still prefer
. . his rook into my lap. When I tried to
White. hand it back to him, he was already
I believe the right move must be standing next to me, ready to search for
1 4.)t�h5!, e.g. 1 5 .ie3 �d6 16 fLle4 (16 the piece. After the game he explained
0-0? 1i'h4 demonstrates the main prob­ that the fair play I had displayed was
lem of having unprotected pieces - they uncommon in Open tournaments on
tend to fall off the board...) 16...lL!f6 17 the European circuit.
lLlxf6+ (17 fLlxd6 li'xd6 18 0-0 fLlxd5 33 :xt8+ �h7 34 i.xa5 1 -0
offers White no compensation)
17 ...'i'xf6 18 0-0 :lad8 19 :lfe1 i.e5 Finally, thinking about knights, here
and Black seems very close to being is a little exercise on the theme. What
equal. Alternatively after 15 .i.xe7 'i'xe7 should Black play after the following
16 'i'e2 'iltb4 White has continued moves?
problems with development, and in the
event of 1 6 tbe2 Ld8 Black can always Figerstrem-Aagaard
return the knight to quite an attractive Sweden 2002
square on f6, from where it puts pres­ undon System
sure on the d-pawn.
The solution to the problem is that 1 d4 lL!f6 2 lLlf3 g& 3 i.f4 i.g7 4 e3
the knight is best placed on f6, and in d& 5 ..tc4 lL!c& 6 o-o o-o 7 h3 i.d7
some situations on f4. In the game I 8 lL!bd2 lL!a5 9 i.e2 c5 1 0 lbb3
played anti-positionally. cxd4 1 1 exd4

Ex c e lling a t Positional Chess

b e able to use the dS-square for both

the knight and the bishop (as happened
in the game). A funny note is that after
1 2 'iVd2 lt.Jc4 1 3 .i.xc4 bxc4 14 lt.JaS
Black can exploit the ciS-square to the
maximum with 14... c3!, winning imme­
One of my juniors suggested
1 1 ...lt.Jd5 12 .i.h2 bS with the reasoning
that on t l ...bS White has the possibility
of 12 dS!?. I agree that it is possible, but
the opening of the long diagonal for the
Black to move g7-bishop makes up for the loss of the
ciS-square. And the pawn on dS also
The solution is, of course, to take the seems to be a little weak. Again - pluses
light squares in the centre. The knight and minuses. I prefer l l...bS as White
on f6 could be better placed because on needs to take care of the unprotected
f6 it hampers the bishop's control over bishop on f4 and possibly lose a tempo.
the centre, while not having enough Alexander liked t t ...lt.JdS first Posi­
influence of its own. The best square is tional chess is about evaluating pros and
obviously dS. cons - we don't always come to the
Black's other knight would love to same conclusions and we don't always
come to c4, of course. Therefore some agree that one is right and another
of my juniors chose 1 1 ...:cs, which is wrong.
quite a logical move. But after 12 lLlxa5 The game continued 1 2 'Oxa5 'ii'xa5
..xaS 13 c4 the position seems to be a 1 3 l:e 1 with a complex struggle ahead.
little better for White. The right move
must be: Bishops - The Archers of Chess
11 . . . b5! While knights have to get close to their
victims in order to hit them with a
deadly blow, bishops have true long­
range power. And because they go di­
agonally across the board they can be
compared to archers in a medieval bat­
de, shooting down their enemies while
simultaneously staying out of harm's
way. Bishops are known as inferior to
knights if their are pawns on only one
flank, or if the position is closed. This is
because their long-range abilities are
then worthless. However, in open posi­
Black has a pleasant position and will tions with pawns on both flanks, bish-

Squares - A n d Ho w Pie ces Exploit Them

ops are generally superior to knights c4 and with the text Fischer makes cer­
due to their ability to quickly shift their tain that the knight cannot threaten the
attention from one side of the board to bishop. The pawn might be on a light
the other, and even sometimes influ­ square but the most important factor is
ence both sides simultaneously. that the rest of the pawns are on dark
In the following position the white squares.
bishop is very strong. 33 . . .llJe7 34 h3 lllc6 35 h4 h5
It was very uncomfortable to let
Fischer·Taimanov White play h4-h5 and g3-g4-g5, but
Vancouver 1 97 1 now White forces Black to put the king­
side pawns on light squares. I believe
that in both cases Black would have lost
the game.
36 l:d3+ �c7 37 l:d5! f5
37...:b8 38 l:tgS and White wins.
Black has no way to avoid weakening
his position.
38 lld2 l:f6 39 l:e2 �d7 40 l:e3 g6
Sooner or later this would have to be
41 i.b5 :d& 42 'i!te2 �d8?
This is considered to be the losing
Taimanov wrote about thi s endgame: mistake but I doubt that Black would
'In what followed I felt like Dr. Watson have been able to withstand the pres­
who could only play along and watch sure forever. Even theoretically.
the resourcefulness and imagination of 43 .l:td3!
the great Sherlock Holmes.' Fischer executes the winning plan.
25 i.f1 ! The bishop dominates the knight.
Forcing Black to play ...a6-a,5. Of 43 . . .�c7 44 :xd6 �xd6 45 �d3
course White would like Black to pqt llle7 46 .tea �d5 47 i.f7+
his queenside pawns on light squares so The bishop has the ability to lose a
the bishop would be able to attack move, something the knight cannot do.
them. Here White uses this to penetrate with
25 . . . a5 26 i.c4 :ta 27 �g2 �d6 his king.
28 �3 llJd7 29 l:e3 lt:lb8 30 :d3+ 47 . . .�d6 48 �c4 'itc6 49 i.e8+
�c7 31 c3 lllc6 32 :e3 'it>d6 �b7 50 �b5 lllc8 51 .i.e&+ �c7 52
If Black tries to do something about i.d5 &iJe7 53 i.f7 �b7 54 i.b3
his pawns with 32 ... a4 he will just have Fischer knows the old Russian trick
an additional weakness on a4 after 33 of showing who is the master by repeat­
a3! etc. ing the position a few times.
33 a41 54 . . .�87 55 i.d1 �b7 56 .i.f3+
The bishop really likes the outpost on rilc7 57 �86 lllc8 58 .i.d5 lt'Je7 59

Excelling a t Positional Chess

.ic4 ll:lc6 60 .if7 t£Je7 61 .tea rooks and queens in a way that makes
Black is in zugzwang. Now he has to them seek shelter behind pawns and/ or
win a piece. Unfortunately for him, it other lesser beings. Often the heavy
loses the game. pieces stand with their backs against the
61 . . .�d8 62 ..txg6 lt:Jxg6 63 �xb6 wall and show a little muscle while they
�d7 64 �xeS lt:Je7 let the young and restless fight the bat­
This is another situation where a tle for the centre. When did you last see
bishop would have been more useful as a combination where only the major
the knight can block a few pawns but pieces remained on the board? Tactical
does so in a very inflexible fashion. properties belong to some extent to the
65 b4 axb4 66 cxb4 lt:Jc8 67 aS minor pieces, while the major pieces
lt:Jd& 68 b5 llJe4+ 69 �b6 �cs 70 come to life only when plenty of blood
�c6 �bS 71 b6 1 -0 has already been spilt on the board.
However, there is one situation
When I refer to knights as cavalry, where heavy pieces come into their be­
bishops as archers, rooks as cannons ing in the middlegame. This is when you
and queens as wizards it is for a reason. have a rook against two pieces, or a
By understanding the pieces in this re­ queen against three pieces, and the minor
spect I find it easier to do them justice. pieces are paorfy co-ordinated.
For children in particular these images Tal in particular was great at using
are useful. the heavy pieces against minor pieces.
The following position is probably the
Heavy Pieces: most extreme case of positional com­
Big Powers - Big Responsibilities pensation with a heavy piece.
Knights and bishops are nonnally the
most important pieces in the middle­ Tai-Panno
game. The reason is simple: rooks and Portoroz 1958
queens are more susceptible to incom­
ing fire than minor pieces. Of course
there are such things as rook sacrifices,
exchange sacrifices, queen sacrifices and
so on. But there are also blunders and
trapped pieces (pins, forks and other
disasters). Heavy pieces are chivalrous
yet fragile species.
The fewer pieces there are an the board the
mare sq11ares wiU become available for the
heai!J pieces an� canseq11entfy, their strength
increasesJar every exchange.
This is the theory, at least. With White is about to regain some mate­
plenty of pieces on the board, knights rial, but hardly enough. And on top of
and bishops are capable of hassling the all that, Black effectively has a passed

Squares - A n d Ho w Pie ces Exploit Them

pawn on the c-file. But there are other had he not known the games of Tal?
factors which are significant. Black's We will never know, but it is clear that
king is in trouble, the a6-pawn is about it was easier for Kasparov to know
to fall and White will have a passed about this material imbalance after
pawn, too. However, without his su­ studying the games of Tal.
preme understanding of the capacities
of the heavy pieces Tal would not have Van Wely-Kasparov
gone for this endgame, and the world Tilburg 1 997
would have been robbed of a master­
piece. White managed to win after:
30 'ii'xf8+ �g5 31 bxc4 bxc4 32 g3
.ta4 33 h4+ ;;.g4 34 �h2 .tf5 35
...f6 h6 36 ...e5 l:le4 37 ...g7+ �3
38 'l'c3+ �e3 39 �g1 .tg4 40 fxe3
h5 41 ••1 l:lxe3?
Tal writes the following in what
Murray Chandler and others have called
the best chess book ever: 'Fatigued by
the foregoing struggle, Panno makes a
mistake. 4L.:e6 would have drawn
quickly since 42 e4 gets nowhere after White has played a Sicilian reversed
42... c3. Now White has real winning and has made the mistake of not devel­
chances.' The book is, of course, The oping his pieces properly. Of course
IJ.je and Games ofMikhail Tal. Personally Black has not made it · easy for him ei­
I prefer the endgame university by Mar� ther, and now Kasparov takes over the
Dvoretsky, but this is a very recent initiative with a classical piece sacrifice
work (to be published in English at ap­ that is not so difficult to calculate.
proximately the same time as this book). 1 5 o&l41 1 6 exd4 ,

42 Wf1 + �e4 43 'ii'xc4+ �3 44 16 "it'cl b6 does not improve White's

•n + �a4 45 ft'xa6 <t>d4 46 ...d6+ situation so the sacrifice has to be ac­
�c4 47 a4 l:te1 + 48 �2 l:te2+ 49 cepted.
�1 l:la2 50 "ifa6+ 'iti>d4 51 a5 c4 1 6 . . . exd4+ 1 7 .te2
52 ...b6+ �d5 53 a6 :ta1 + 54 �2 1 7 �dl? loses to 1 7 ... b6 and 17 �d2
c3 55 a7 c2 56 'l'b3+ �d6 57 b6 makes little difference as ... .ih6+ is
'iid 3+ 1 ·0 coming.
1 7 . . ..txg2 1 8 0-0-0
The following example of a rook Thanks to . .i.£3 there were no alter­
. .

dominating two minor pieces could natives.

have been played by Tal himself, but 18 . .. .txh1 1 9 :txh1
was played by the greatest player of our This position was easy to foresee but
time instead. The question is - would to evaluate it is something else. With the
he have been able to play in this fashion next sequence of moves Kasparov en-

Excelling a t Positional Chess

sures that his slightly better co­ dominate the minor pieces it will even­
ordination is maintained and that the tually be outnumbered.
lead in development is kept all the way 26 .id1 a5! 27 ..txd4 a4 28 .txg7
into the endgame. �xg7
1 9. . .'i'd5! 20 l:.e1 The exchange could seem to favour
20 :gt ? 1be2! etc. White as two versus one should be bet­
20 . . .l:e5! ter than thtee versus two, but the ex­
The queen's rook is the worst placed change of bishops increases the- number
piece, and soon it will be participating in of squares to where the rook can travel
the game. White continues to retreat, safely, and also marks the pawns on a3
being poorly co-ordinated. and h2 as weaknesses. An important
21 ..b3 point here is that minor pieces have
Also possible is 21 ltlb3, when after great difficulty playing against a rook
21 ...ltae8 (21...lte7!?, as suggested by and a passed pawn on the edge of the
Winants, is perhaps a safer path to an board. It is no surprise that Kasparov
advantage) 22 .ixd4 ltxe2 23 :Xe2 uses this important feature of the mate­
:Xe2 24 'iixe2 'i'xb3 25 .txg7 r:/;xg7 rial imbalance to win the game.
Black has a much better ending in view 29 llld2 .l:.e5 + 30 ..te2 b5
of the extra pawn and White's weak­
nesses. However, White has some draw­
ing chances after 26 We7+ 'ii'f7 27
'i'd8!, when at least Black cannot domi­
21 . . . -.xb3 22 lllxb3 .l:.ae8! 23 �d1
23 ltlxd4 l:.5e7 24 �d2 ltd7 and all
the insufficiently protected white pieces
can no longer be protected.
23 .. . .l:.xb5 24 .if3 l:txe1 + 25 �xe1

Materially speaking the situation is

roughly equal. Rook and pawn is
probably not quite enough against two
minor pieces, while rook and two
pawns might be slightly better. Here
there is no doubt. The white knight
cannot find a stronghold anywhere, the
bishop cannot find scope and the white
pawns are divided and leaderless. The
dark forces have won the battle for
Domination. If the rook cannot 3 1 �d 1 Ad5 32 �c2 g5 33 .tf3

Squares - A n d Ho w Pie ces Exploit Them

l:d6 34 h3 �g6 35 lL!b1 h5 36 llJc3 position is how many pieces there are
g4 37 .i.g2 �f6 on the attacking side compared with the
Black's king is going to the centre. defending side. Black will find it very
The text allows White to exchange the difficult bringing his pieces out of the
h-pawns and free his bishop &om its two comers because he has to simulta­
obligation, but with the entry point on neously look out for his Icing's safety.
h2 available for the rook Black is well Even a materialistic chess program such
compensated. as Fritz evaluates this as winning for
38 hxg4 hxg4 39 d4 �g5 40 ;rd3 White.
:h6 41 �82 f4 42 J.e4 :h31 But compare this to the following
Domination once again. Black is position.
winning on both flanks. Hartvig-Raetsky
43 �d2 lr.h2 44 �e1 g3 45 fxg3 Tastrup 2002
fxg3 46 �1 l:f2+ 47 �g 1 b4! 48
axb4 a3 49 d5 �4! 50 i.g6 cxd5
51 lbxd5+ �g5 0-1
52 i.b1 ltb2 53 ltlc3 %:txb1+! 54
ltlxb 1 a2 etc.

The King and the Right to Castle

To many players the loss of the right to
castle is itself a reason to worry. If you
find yourself in a situation such as the
one below (taken from an opening trap)
this is easy to understand:
In this position White quickly played:
1 9 J.h5 + ?
Taking away Black's right to castle.
But before rushing to do this he should
have asked himself if this was to his
advantage. In fact it transpires that
Black wants to put his bishop on a6 and
his rook on b8, and this is achieved one
move faster after the check. Compared
to the previous example White's pieces
are not about to storm the enemy king.
And the king would probably have gone
Here Black is in serious trouble, al- to e 7 anyway as this square is safer than
1ftough there are currently no direct g8. All in all, White should have played
threats against the king. But they will 1 9 i.f.3 with a slightly inferior position.
come! The central pawns offer little Now the game ended quickly.
help as the key factor in this kind of 1 9 . . .'�e7 20 J.f3 a5 21 l:b1 .i.a6

Ex c e lling a t Po s i tional Chess

22 �e2 l::thb8 23 'ii'c2 amazement upon 3 ... e5 (after 1 d4 tlJf6

After 23 :d 1 :Xc4 24 bxc4 �xc4+ 2 c4 d6 3 tlJc3), believing that Black has
25 �el l:xbl Black also wins. lost his head. But after 4 dxe5 dxe5 5
23 . . ..txc4+ I 24 bxc4 l::tb2 0-1 'ifxd8+ �xd8 Black does not have
Black wins the queen. After the game problems with the king.
he was not unhappy with the extra
tempo his opponent gave him to carry
out his plan.

The 'superstition' of many players re­

garding the right to castle is well illus­
trated by the following example.

Black's king stands on the d-file but

after ... c7-c6 and .. .r/;c7 it will be per­
fectly safe. That does not mean that
giving up the right to castle just to get
into an endgame is necessarily justified.
It m eans that when you have the right
to remove your opponent's chance of
In this well known position, opening castling, or he is threatening to remove
theory considers 4 tlJ£3 to be the best yours, that you should try to find out
way to fight for an advantage. I know who really benefits from this, rather
many club players who would look with than taking matters for granted.

Analysing Your Own Games

It is well known - and has been for a can analyse our games. They have been
century - that one of the key ways of arranged according to the amount of
improving is to analyse our own games. work involved. Thus you might want to
For this there are some obvious rea­ go as far as Level 4 and no further, or
sons. First we get to check our intuitive until Level 6, and no more. Please note
decisions against more concrete evi­ that they are not in the order in which
dence. Secondly, we might get a better they should be conducted. I simply
insight into our general thought proc­ want everyone, even the laziest of the
esses. Thirdly, we will probably learn a lazy, to be able to benefit from these
lot about the openings we play, one way pages.
or another. But most importantly, if we
analyse our games correctly, we will be 1 1 Write down three new things you
able to see recurring mistakes. And that have learned from tl'le game
is obviously an imponant step in_ the Well, how long can this take? After a
direction of correcting them. while it actually becomes more difficult
In this chapter I will try to offer as you will eventually run out of new
some good advice as to how this can be · t:rungs to write. However, I am sure that
conducted effectively. It is my feeling seeing each game also as a stepping­
that even greats like Yusupov and Yer­ stone to new knowledge will benefit
molinsky, when they talk about analys­ your overall performance. (I thank
ing your own games as a way to im­ Coach for this idea.).
prove, fail to offer much advice about
how to get the best out of your hard 21 Always write down the time
work. Perhaps it is me, but I feel that spent during the game
more precise guidelines could be drawn, This is a well known idea and should be
and I have tried to do so. followed strictly. Quite simply, when
Below are eight levels upon which we you write down the moves you also

Ex c elling a t Po sitional Chess

write your time, or the time of your op­ Smolkov-Aagaard

ponent. I have found in my work with Voronezh 2002
pupils that this will always reveal where Sicilian Defence
mistakes are quite commonly placed
during the course of a game. With one 1 e4 c5 2 ttlf3 d6 3 d4 cxd4 4
pupil, for example, his obvious mistakes �xd4 �f6 5 �c3 g6 6 �e3 .i.g7 7
tended to be a result of playing a move t3 o-o a •d2 �e:s 9 o-o-o �d4 1 o
instantly, or pondering over it for 30-40 .txd4 .te6 1 1 �b1 9c7 1 2 h4 :tea
minutes without seeing anything at all. 1 3 h5 �xh5??
After realising this he started using his Please do not try this at home - or
time better and went up 100 Elo points anywhere else. As far as I can see from
- from 2200 to 2300 - very quickly. my database I am really the only survi­
From then on real improvements were vor of this mistake. White's continua­
unfortunately needed in his chess, but tion is pretty straightforward, but I still
they came, and now he is regularly play­ failed to see it. In my defence I would
ing for IM-norms and pushing his rat­ say that a very beautiful Russian blonde
ing in the direction of 2400. was watching me as J . was thinking...
I am certain that if you start writing Oh, how easy it is to lose concentration!
down the time after each move you will 1 4 .txg7 �xg7 1 5 g4 �f6 1 6
notice something you did not know, or 1i'h6+ �g8
perhaps you will notice something you
did know, but had done nothing about.

3 1 Check the opening theory

This is easy. If you have a database on
your computer you check some critical
positions and some strong GM games
that are played along the same lines as
your game. The same goes for using
books, of course - just look at what
others have played. I know a guy who
lost the same game twice in two years -
all the moves were identical! For obvi­ How to win? Take a minute and see
ous reasons I will keep his name to my­ it to the end.
self. Here is a recent game of my own ih 1 7 e5! dxe5 18 · gs �h5 1 9 :Xh5
which I survived a terrible mistake that gxh5 20 .td3 e4 21 .txe4?
no one else has ever survived at such a 21 ltJxe4! has been played in so��
high level. I learned a lot from looking games (all gimes), after �hich I - h�d
through the games on my computer intended 21 .'ii'f4, knowing that it lost,

database and in my Dragon book - even though I had not seen the follow­
mainly that I did not understand the ing straightforward combination - 22
system and should play something else! �f6+ exf6 23 i.xh7+ �h8 24 .1f5+t

A n a l y s in g Your O wn G a m e s

�g8 25 'irh7+ 'it£8 26 'ifh8+ 'ite7 27 in his S•hool of Chess Ex•ellence 2 - TadicaJ
gxf6 mate. Plq. :.4. •hess p�er depends on his own dis­
2 1 . . ..-e5! coveries to no less degree than those of others.
At least covering some squares and They are closer to him, and they are
thereby staying alive Gust). For some better remembered, since at some point
reason I always seem to play wonder­ they have already been deeply consid­
fully in 'decided' positions. ered and sensed by him. Your own
22 ..txh7+ �hS 23 ..te4+ �gS 24 games should be carefully preserved,
i.xb7? and sometimes again examined.'
A grave mistake caused by miscalcu­ Mark is right here, as he tends to be.
lation. 24 �7+ �£8 25 j\_xb7 would The thing is that we learn much less
still give White a virtually winning posi­ from being given conclusions than we
tion. do from finding them ourselves. This is
24. ..:dSI why it is so valuable to analyse your
Now White no longer has an advan­ own games. The idea of always writing
tage. down on the same day the moments in
25 .1xa8 :xd1 + 26 li:lxd1 'i'd4! games that you felt were critical is
A nice little trick. probably rather obvious, but I will take
27 b3 a few seconds of your time to explain it
27 ll'lc3 'ili'g1+. anyway. Later, when you get around to
27 . . . .-xd1 + 28 'if.?b2 'i'd4+ 29 �b1 analysing your game thoroughly (in my
'i'd1 + %-% case I have recently started analysing a
game played eight years ago!) you will
Incidentally, games such as this - find this record of your emotions dur­
whether they are lost, won or drawn - ing the game highly beneficial. It will
can be a good motivation for getting also suit you as a motivational tool.
started on some interesting studying. When you write down your feelings and
And almost always you will find some­ wtanswered questions from a tense
thing new that you did not consider game they will hang around in your
either during the game or the post­ memory until you address them.
mortem. Here the 22 ll'lf6+ reply to
2 1 ...'ili'f4 (see note to White's 21 st) even 5) Analyse the game yourself. Only
went undiscovered when I went when finished should you refer to Fdtz
tluough the game with my opponent. These days most people analyse their
We analysed only 22 l:.gl and decided games by turning on Fritz in their hotel
that it won for White. room/flat/house, and relax with a re­
freshing drink. This is what we call ana­
4) Write down the critical moments of lysing our games[ It is no wonder that
the game, the things you saw during modern chess is increasingly oriented to
th8 game and what you think went openings, and playing styles are becom­
wrong. Do this the same evening. ing more tactical The importance of
Mark Dvoretsky writes the followmg deeper positional understanding is fading

Ex c e lling a t Pos itio n a l C h e s s

into the background because everything choose not to follow it - and always re­
seems to be evaluated in terms of mate gret it. Our intuition is a powerful tool
or material rather than squares and plans and when we feel something we should
and so on. This is mainly valid for ama­ be alert to this priceless information,
teurs and not for the very top players, especially when we play or analyse. If we
who clearly understand chess very well in have any sort of uncomfortable feeling
all its facets. about a conclusion then it is time to go
Computers can be a useful tool to us deeper into that position. A common
when we work on one of our biggest reason for 'ignoring' your intuition is
weaknesses, such as complex tactics, but believing in others before yourself. I
cannot help with strengths such as logic, have often had bad feelings in the open­
structures, long-term planning, intuition ing, playing some line simply because it
etc. So when it comes to analysing our has been evaluated as equal in a book or
games in detail (something that is diffi­ magazine. But this was always wrong.
cult to find the time to do during tour­ Later analysis showed me time and time
naments) we should start by devoting again that there is no better guide in life
considerable time to it What we are do­ than your own, well-founded intuition. -
ing is not only searching for the truth, to It is also important to append a ques­
find some justification for our sour or tion mark to all conclusions you feel sure
happy feelings about a game and its re­ about, as this is often where you can im­
sult, but also investigating how we think prove on your deeper understanding of
and how our ability to make correct deci­ things. For although your intuition is a
sions at the board is affected Addition­ strong and magical tool, it is not flawless
ally, of course, we develop our ability to - which is probably the first thing your
think further just by thinking. This is intuition has told you about the lines
where we do not want to use Fritz or above!
similar programs as they should not It is generally considered a good idea
think for us at home. Lance Annstrong to express your conclusions in words as
does not prepare for the Tour de France in well as in variations when you have ana­
his car, and nor should we prepare our­ lysed your games and come to conclu­
selves for tournaments by letting the sions. For this reason I often like to ana­
computer think for us. lyse my games with others (it is not a
When you do analyse your games try good idea to talk to mysel£1). I believe in
to devote plenty of time to questioning both ways of thinking - abstract, long­
your decisions. I have a saying that you term concept oriented thinking and con­
should believe in yourself and believe crete, provable thinking. A move should
yourself. The first part concerns not set­ fall into both camps in order to really
ting limits for our abilities, while the sec­ impress.
ond part means we should be aware of When you have finished analysing
our inner voice. I have personally experi­ your games and feel there is little further
enced, again and again, that I have a feel­ progress to be made alone, it is time to
ing in some positions or situations and I go through your games with a computer.

A n al ysing Your O wn G a m e s

The computer will then find the tactics two weaknesses and none of the activity
you were unable to find yourself lbis will I should have.
provide you one of two deeply rewarding Analysis showed that 24 a4 lLlc5 25
experiences - either you will be sup­ .:tact £5 26 .:XeS! dxcS 27 'i'c3 would
ported in your own analysis by someone give \Vhite a lot of compensation for the
who sees everything, or you will see now exchange and possibly genuine winning
doors of wisdom open in front of your chances. No computer could have told
eyes, doors that were previously closed me that
to you. The thing is that to be able to
understand what these doors really hide
you first need to have done all the work
yourself. You cannot truly realise what is
new for you if you do not test the limits
of what you know.

Here are some examples from my

games in the summer of 2002, when my
trainer and I fotmd great improvements.

Copenhagen 2002 In this position I was Black against a
very talented 1M from Argentina,
Rueben Felgaer, who made his final
norm in this tournament (still Copenha­
gen Open 2002). We both saw the fol­
lowing line: 13 .i.xf6 gxf6 14 ...xf6 l:l.g8
15 g3 .i.g7 1 6 'ii'gS 'iit£8 1 7 'ii'ct h5 with
a complex struggle in which Black might
very well be better. What neither of us
considered was that after 14 ltk3! l:l.g8
1 5 lbe4 Black is close to being position­
ally lost. It is quite funny that a young
German IM, who is clearly stronger than
Here I played 24 'i'xe4?, believing that myself, had actually intended to play 14
I could create the activity necessary to 'i'xf6 against me in a game some months
make a draw after the exchange of a pair later. Of course I deviated from this line
of rooks. I was sadly wrong, though the due to 14 lbc3!.
position might hold some drawing A computer could have made neither
chances after 24....l:lxb5 25 l:l.acl .:tc5 26 of these discoveries. I would like to say
:X�S 'i'xc5. Here I played 27 h3? fol­ that I found them myself, analysing my
lowed by 28 a4? and the rook manoeuvre games for hours and hours, but even dtat
...llffi-a8-a5 made it obvious that I had was not enough. Mark Dvoretsky found

Ex celling a t Po s i tio n a l Ch e s s

them in an afternoon, going through my for them in the future. I have found
games guite rapidly. that after starting to use this way of ana­
Fri� however, did find one big hole lysing my own games I have improved
in my analysis in the game with Felgaer. both my tactical ability (considerably)
After the moves and my positional understanding. I ad­
1 3 lLlc3 i.c5?! (13...i.e7 with an even vise others to find their own methods,
game was better) 14 Wh4 liJd7! ? 1 5 of course, but it should pay to have
i.g3?! I was doing all right. But better these considerations in mind when ap­
had been 1 5 i.xg7! l:tg8 1 6 'i'xh7 proaching the task of analysing your
lLlf6 and now the discovery I should games.
have found myself, but which I saved for
Fritz: 61 Check for structural assistance in
ChessBIIsB to gain additional insight
When we analyse our games we occa­
sionally ftnd that the principles of some
types of positions completely evade us.
This is when we have little idea of what
candidate moves to look for and of
which plans are important, or when we
thought we had a good position and ·
then suddenly nothing seems to work.
On these occasions it is very useful to
use the functions of Ches.rBase or similar
programs to search on structure. This is
1 7 :xe6 + ! 1 usually pawn structures perhaps com­
1 7 �6? i.xf2+! (17 ...i. f8 1 8 'i'xf6 bined with ECO codes. Often a gre.?-t
i.xg7 1 9 'ii'gS 'iPf8 20 'fibS! serves only revelation will come when we look
White's interests) 1 8 �h1 ltlg4 1 9 'i'gS through games between really strong
i.xe1 20 :Z.xel ltlf2+ 21 �gl �6 22 players in positions we do not under­
�ft :ds 23 l:te2 (23 'ii'h4 :td2) stand.
23...�d1 and Black is better.
1 7 . . . 'i'xe6 1 8 .ixf6 .i.xf2+ 1 9 �h1 71 Toumament reports and
'i'g4 20 'ii'e4+ Wxe4 21 lLlxe4 J.b6 Diagnosis of weaknesses
22 :d 1 :g4 23 tLld6+ �d7 24 h3 After having analysed all my games
24 ltlxb7+ Wc6 25 lbd8+ i.xd8 26 from a tournament there is something
.i.xd8 :c4 27 c3 .:b8 is drawish. that I generally enjoy doing I make a

24 . . .llg6 25 lLlc4+ 'iilc7 26 Af 1 and list of all my mistakes from my games,

White has some advantage. By seeing and describe them. Often I find that I
tactics such as these after first being make one or two recurring mistakes,
completely blind to them I do not only and that these are not completely in the
acguire more information about my shape and form that I believed. A tacti­
games but I also increase my awareness cal mistake is not just a tactical mistake.

A n al ysing Your O wn G a m e s

Often, as with the examples in this Tournament r-eports: A report based on

chapter, there is a common theme that deep analysis of your latest tournament.
embraces all the mistakes. In this sec­ Not just the games themselves, but also
tion it is a lack of imagination. All the on the tendencies in your game. De­
things I overlooked were because I did signed to show the way forward.
not look at them at all not incorrect

evaluation or miscalculation, but pure Proposed Further Reading

After making this list and finding The &ad to Chess Improvement
your most serious weaknesses it is natu­ (Yermolinsky)
ral to continue with:
Arl of Chess Anafysis
8) Training based on tournament Qan Tinunan)
reports In this book Timman analyses games to
If you know where you lose points a level that can inspire you to see what
there is nothing as logical as building a depths chess analysis can reach.
training program based on eliminating
these weaknesses. For every weakness Instructive Modern Chers Masterpieces
there is a remedy, and it is never the (Igor Stohl)
same. I hope you will find yours. This book is really impressive. The
games are deeply analysed and very well
Explanation of Terms annotated, although it might be consid ­

Class pkgers: Auto-thinking individu­ ered heavy going from an average club
als. player's point of view.

Positional Sacrifices

Ifyou can 't beat them, confuse them - Es­ dynamic tradition. For these players the
ben Lund. initiative is more important than mate­
We all like to make sacrifices that rial. Thus far there seems to be no supe­
lead to checkmate - not only because rior playing style.
they look nice, but also because we win To me there is a clear difference be­
the game. We become aesthetically ex­ tween sacrifices aimed at generating an

cited when we are able to sacrifice ma­ initiative and those designed to gain
terial for an attack, sacrifices that cannot lasting positional (often structural) ad­
be calculated but which later turn out to vantages. Unfortunately it seems that
be correct, or at least very dangerous. chess literature fails to take this into
These tend to be the ones that find their account. In the following example the
way into magazines and books. sacrifice is what I would call positional:
Some players don't feel comfortable
sacrificing material and others don't feel L.B.Hansen-Loginov
comfortable accepting 'free' pieces - it Stara Zagora 1989
depends a lot on taste. Some players
belong to the classical school, as exem­
plified by the likes of Capablanca, Smys­
lov, Petrosian, Fischer, Karpov and
Kramnik. To a certain extent these have
a universal style and sacrifice material
from time to time. However, their main
emphasis is on control, technique and
structure. Then there are players like
Alekhine, Tal, Shabalov, Kasparov, Shi­
rov, Grischuk, Morozevich and others,
who belong to what could be called the

Positio n a l S a c rific e s

Here White sacrificed with 28 :c&! examples is the following: Material isjust
.i.xc6 29 :xc6 and held a lasting ini­ another positional jattor. It is sometimes
tiative on the light squares. The reason important, sometimes not. What good
why I would call this a positional sacri­ is it to have an extra bishop and pawn if
fice is that it is not just a matter of time it is the classic 'wrong' bishop and
and of getting to the king. The fact that rook's pawn and the defending king
the light-squared bishop is removed occupies the relevant comer? In such a
from Black's camp is very significant position the fact that the comer is oc­
and, in the long-term, this slows down cupied is clearly a positional factor, and
his activities on the kingside . Meanwhile it clearly equals White's extra material.
White prepares a slow invasion on the The same principle goes for a number
light squares. of other situations. A simple illustration
Another example, which is very fa­ from one of my own games starts in the
mous, is the following. following position.

Lutz-Karpov Aagaard-Olsen
Dortmund 1993 Copenhagen 1999

In this position Black earned a long­ I feel that this position is slighdy bet­
term advantage with 21 ...hf4! 22 ter for White, but that nothing is yet
gxf4 i.xf4+ where the pawn, the decided An important factor will be
domination of the dark squares and the how the minor pieces come to function.
somewhat weakened bishop on d3 But in the game Black decides to grab a
proved important. Actually Black pawn at the cost of significant time.
showed that it was not a matter of mat­ White chooses to use the time to re­
ing ideas when he followed up with 23 group his pieces, giving him a lasting
�b1 'it'e51 24 •xe5 .i.xe5 with at positional advantage. I think a move
least, equal chances in the endgame like 14...J:c8!?, intending 15 �d2 .ih6,
(which he went on to win due to supe­ provoking weaknesses, or perhaps even
rior technique). the immediate 14... .1h61?, should offer
What is characteristic in all of these Black reasonable chances of fighting for

Ex c e lling a t Po sitio n a l Ch e s s

equality. In the game White i s winning Korchnoi played this and four similar
within a few moves. games in the period of 1 998-2000 with­
1 4 . . . Wd7 1 5 ..d3 ._a4 1 6 lbd2 out conceding even a draw. These
..-xa2 1 7 lbc4 'W'a6 1 8 l:ta 1 ..b5 1 9 excellent results are mainly down to his
l:ta5 'i'd7 20 l:ta6 f5 2 1 exf5 'W'xf5 great playing strength. Alexander Khal­
22 Wxf5 gxf5 23 l:tfa 1 l:tfe8 24 � 1 ifman has a score of 3'/z/6 wi th Black in
.if8 2 5 f3 l:te7 26 l:te1 l:tc7 2 7 l:te3 related positions.
l:td8 28 b3 l:tcd7 29 l:te1 .tg7 30
lbe3 f4 31 lbf5 i.e5 32 g3 l:lc7 33
l:te4 fxg3 34 hxg3 i.f& 35 lbxd6
l:te7 36 l:txe7 i.xe7 37 l2:\f5 �8 38
l:txa7 .i.d6 39 f4 .tb8 40 l:tb7 i.d&
41 �e2 1 -0

As I indicated above, a positional sacri­

fice is C'ompensated in stT'IIctural advantages, a
tfynamic sacnjice is compensated in gain if
tempo and thf'fafs of mate or material gain.
Of course there are mainly borderline
cases, but when you have to decide But the main thing is that Black need
whether or not to sacrifice it is impor­ not prove his compensation immedi­
tant that you understand what kind of ately since it is of a positional nature
sacrifice you are contemplating. If it is a and therefore long-term. Whether it is
positional sacrifice, then you will have enough is another question.
compensation independent of time, In the following position the com­
being able to improve your position pensation is of immediate character.
slowly. On the other hand; a sacrifice
based on attack or other combinational Nezhmetdinov-Mikenas
properties requires you to act accord­ Kazan 1 948
ingly. A good example of positional
compensation is the Benko Gambit,
where situations such as the following
are common.

see following diagram

1bis is a theoretical posmon taken

from the game Korchnoi-Adorjan,
Germany 1 998. Black has positional
compensation because of the two half­
open files and the strong bishop. How­
ever, this compensation can be debated,
and White does have a full pawn more. This position is from the 1 1 th match

Po s itio n a l Sa crifices

game between Rashid N ezhmetdinov, White has a strong attack.

one of the greatest attacking players, 1 7 . . .�84 1 8 ii'xc5 .l:lfB
and Mikenas, played in Kazan in 1948. Nezhmetdinov was not one to pro­
In this position White has compensa­ vide lengthy analysis of his games; he
tion for the piece mainly because of the offered various ideas. Here he mentions
exposed black king which, if it were to 19 f3+ as a threat, which can be seen in
crawl to safety on h7 or g8, would se­ the line 18 ... ttJc6 19 f3+ �xf3 20 'i'c4!
verely reduce White's compensation to and White has a winning attack. One
inadequate. Therefore White needs to line is 20 .. .'1Wh4+ 21 'itdl lit>e4 22 dS+
act on his compensation as soon as pos­ �xeS 23 .:et+ �f6 24 gS+ etc.
sible. But Black committed an error... 1 9 0-0 �3 20 h3 b6 21 'i'c3+ �e4
15 . . .'i'd8? 22 'i'c4! 1 -0
Nezhmetdinov gives 1 5. ..ttJc6 16 d4+
�hS 1 7 'i'xcS 'ii'e7 with chances for Using sacrifices to unbalance the
both sides, bui: I find it hard to believe game in order to improve the prospect
that White should be okay here. An of generating winning chances is proba­
endgame would be better for Black due bly one of our favourite ways to burn
to the opposite coloured bishops and bridges before we cross them. It is one
the way White's pawn structure works of the things that makes chess both dif­
against the cl-bishop while helping ficult and interesting. In modern chess,
Black's minor pieces. In my opinion sacrifices designed to create an initiative
White needs to accelerate development are very common, and top players have
if he is to demonstrate compensation in a rather pragmatic attitude to being an
a position like this - but he simply can­ exchange up. In a training session with
not. After 1 6 0-0 Black would have the Mark Dvoretsky, where we analysed my
very nice move 16 .. .'ifr'h5!!, giving him games, it turned out that I had continu­
the better game. The idea behind the ally missed strong exchange sacrifices
move is to side-step all the possible that Mark had spotted relatively easily.
dangerous checks. Note that since To be a strong positional player it is
White has used up all the minor pieces necessary 'to have a freer relationship
capable of operating on the light with material matters than is the case
squares the king is as safe on hS as it with most average (rated) players. Mate­
would be elsewhere. This is an instruc­ rial is mere!J anotherpositionalfattor. It is as
tive example of a sacrifice in which the simple as that.
requirement is an immediate attack, but If you would like to delve deeper into
where the immediate attack is not a positional sacrifices I can warmly rec­
possibility, and the sacrifice (in this case ommend two books on the subject,
starting with 9 .i.xf7+) should be held namely McDonald's Positional Sacrijkes
r�sponsible for the sad state of White's and Dunnington's Understanding the Sacri­
posmon. fice. Both books are filled with interest­
1 6 d4+ �5 1 7 g4+ ! ing subjects, and both authors have my
Now the situation has changed. deepest respect.

Positional Exercises

These positional exercises are all taken not yet have found the time.
from the email training program I ran In all the exercises I have a distinct
during 2002. I would suggest that you idea regarding the best move. I have
take 15-20 minutes to solve each exer­ invested considerable time in these po­
c1se. Set a chess clock if that makes you sitions and I have checked them with an
feel more comfortable. The idea behind average of 5-10 pupils and participants
the exercises is not to guess the best in my e-mail program. In 95% of the
move but to find it and, subsequently, cases my own investigations and those
find the underlying plan. Solving all of of my pupils and participants in the
these exercises and comparing your so­ program have validated the decision and
lutions with my proposed solutions will annotations of the world class players
give you a lesson in positional chess that who played these positions. Compared
is more valuable than any other I would to most other positionally oriented
be able to propose. The solutions tend workbooks, this has given me an oppor­
to be, not surprisingly, a reflection of tunity to understand which exercises
the previous chapters. In order to fully were working and whlch were not. I
understand everything below it might hope you willhave enjoyable and educa­
be a good idea to read these, should you tional hours with these exercises.

Positio n a l Exercis e s

Exercise 1 : White to move Exercise 4: Black to move

Exercise 2: Black to move Exercise 5: White to move

Exercise 3: Black to move Exercise 6: White to move

Excelling a t Positio n a l Ch e s s

Exercise 7 : White to move Exercise 1 0: White to move

Exercise 8: Black to move Exercise 1 1 : White to move

Exercise 9: White to move Exercise 1 2: White to move

Positio n a l Exercis e s

Exercise 1 3 : White t o move Exercise 1 6: Black to move

Exercise 14: Black to move Exercise 1 7 : White to move

Exercise 1 5: White to move Exercise 1 8 : White to move

Ex c e lling a t Po s itio n a l C h e s s

Exercise 1 9: White to move Exercise 22: Black to move

Exercise 20: White to move Exercise 23: White to move

Exercise 2 1 : Black to move Exercise 24: Black to move

Positio n a l Exercis e s

Exercise 25 : Black to move Exercise 28: White to move

Exercise 26: White to move Exercise 29: Black to move

Exercise 27: White to move Exercise 30: White to move

Exc elling a t Positio n a l C h e s s

Exercise 31 : Black to move Exercise 34: Black to move

Exercise 32: White to move Exercise 35: White to move

Exercise 33 : Black to move Exercise 36: White to move

Positio n a l Exercis e s

Exercise 3 7 : Black to move Exercise 40: White to move

Exercise 38: White to move Exercise 41 : Black to move

Exercise 39: Black to move Exercise 42: White to move

Ex celling a t Positio n a l C h e s s

Exercise 43: White to move Exercise 46 : White to move

Exercise 44: White to move Exercise 47: Black to move

Exercise 46: Black to move Exercise 48: Black to move

Positio n a l Exercis e s

Exercise 49: White to move Exercise 52: White to move

Exercise 50: White to move Exercise 53: White to move

Exercise 51 : White to move Exercise 54: White to move

Ex c e llin g a t Positio n a l Chess

Exercise 55: White to move Exercise 58: White to move

Exercise 56: White to move Exercise 59: White to move

Exercise 57 : White to move Exercise 60: Black to move

Positio n a l Exercis e s

Exercise 61 : White to move Exercise 64: White to move

Exercise 62: White to move Exercise 65: White to move

Exercise 63: White to move Exercise 66: White to move

Ex c elling a t Po si tional Chess

Exercise 67: White to move Exercise 70: White to move

Exercise 68: Black to move Exercise 71 : White to move

Exercise 69: White to move Exercise 72: Black to move

Positio n a l Exercis e s

Exercise 73: White to .move Exercise 76: White to move

Exercise 74: Black to move Exercise 77: White to move

Exercise 75: White to move Exercise 78: White to move

Ex c e lling a t Po sitio n a l Chess

Exercise 79: Black to move Exercise 82: Black to move

Exercise 80 : White to move Exercise 83: White to move

Exercise 8 1 : Black to. move Exercise 84: White to move

Positio n a l Exercis e s

Exercise 8 5 : White t o move Exercise 88: White to move

Exercise 86: Black to move Exercise 89: White to move

Exercise 87: Black to move Exercise 90: Black to move

Ex c e lling a t Positio n a l Chess

Exercise 91 : Black to move Exercise 94: White to move

Exercise 92: Black to move Exercise 95: White to move

Exercise 93: White to move Exercise 96: Black to move

Positio n a l Exercis e s

Exercise 9 7 : White to move Exercise 1 00 : White to move

Exercise 98; White to move Exercise 1 0 1 : White to move

Exercise 99: White to move Exercise 1 02: White to move

Ex celling a t Positional Chess

Exercise 1 03 : Black to move Exercise 1 06 : White to m ove

Exercise 1 04: White to move Exercise 1 07: White to move

Exercise 1 05: White to move Exercise 1 08 : White to move

Solutions to Exercises

Exercise 1 : White to move inject some energy into his pieces. But
Andersson-Vaganian Black also has to respect the weakness
Skellefui 1 989 on d6. His only way to protect it is by
exploiting his control over d4. There­
fore White should attack dovm the d­
ft.le with 11c 1-c2-d2.
1 5 lc2!
Threatening 1 6 lbb5! etc.
15 ... 86 1 6 l:td2 tbd4
And now comes the exchange sacri­
fice, which guarantees White a slight
edge. Remember that the e6-knight -

now on d4 - and the bishop are the

best black piecest
1 7 hd4 cxd 4 1 8 l:Lxd4! J.xd4 1 9
W'hite is fully developed. As the rook 'i'xd4
on fl might support f2-f4 it is good for White is slightly better and is already
now. It is also hard to say where the threatening, for example, �d5 and
queen is best placed, and in fact it is .tg4. Therefore Black must activate his
quite fine where it is right now. Only pieces and surrender d6.
the rook on c1 needs to be improved, 1 9 ...'i'a7! 20 'i'xd6 fladB 21 lt)d5
and b2 needs a little protection. Black, \tlg7?!
on the other hand, is in a bit of trouble. Here Black could have improved his
Of his pieces, only the e6-knight and play with 2L'i'c5! 22 'i'xc5 (22 'i'g3
the bishop are really well placed, al­ 'i'd4! gives Black good counterplay)
though Black is considering . a7-a6 and. . 22 .'tJxcS 23 f3 and White has the

.. b7-b5 in the near future in order to

. advantage in the endgame with his two

Exc elling a t Posi tio n a l Chess

pawns and strong knight on dS. 14 . . .dxe4 1 5 dxe4 ll:lbd7 1 6 ll:ld5

22 rtd1 ! 1 6 lbh4!? i.xe4 17 lbhfS with some
White is clearly better now and went compensation was suggested by
on to win. Ehlvest.
22 . . ."i'c5 23 •g3 a5 24 "i'c3+ f6 1 6 . ..ll:lc5 1 7 llntf6+
25 ll:lf4 "i'e7 26 •xa5 •xe4 27 .:td4 17 ll:lb6? lbxb3 1 8 .z:r.xd8 %hxd8 wins
"ike7 28 1i'c7 �h8 29 .i.f1 'ire5 30 for Black while after 1 7 .ta2 i.d6 the

l:xd7 l:txd7 31 •xd7 "i'xf4 32 'irxb7 problems on the d-file are solved. Now
l:b8 33 "i't7 •de 34 c5 •ts 35 it becomes clear that Black is better
"ikxf8+ .:txf8 36 b4 :as 37 b5 l:xa2 placed.
38 .i.c4 l:a1 + 39 �h2 9/;g7 40 c6 1 7 . . .Wxt6 1 8 .ic2 •cs 1 9 l:te1
1 -0 Here White could have forced a
repetition with 19 %1d5! 'ii'f6 20 J:td 1 ,
Exercise 2: Black to move but now Black takes over.
Ehlvest-Anand 1 9 .. . l:ad8 20 .i.d2 "i'e6 2 1 l:ab1
Riga 1 995 "ikc4 22 b4 ll:ld3 23 .i.b3 "ikc6 24
.i.d5 l:xd5 25 exd5 -.xd5 26 l:ed 1
e4 27 ll:le 1 l:e6 28 .i.e3 .i.d6 29
.ic5 .i.xc5 30 bxc5 l:tg6 31 'ii'e3
'W'h5 32 c4 bxc4 33 l:d2 .i.e& 34
llc2 .td5 35 h3 �h7 36 �h1 f5 37
f4 exf3 38 ll:lxt3 .l:.e6 39 'W'd4 .i.xt3
40 gxf3 'Wxh3+ 41 l:h2 'W'xf3+ 42
l:g2 l:g6 0-1

Exercise 3: Black to move

Novgorod 1997
Black is fully developed. His queen's
rook could be better placed but it is not
so important right now, unlike the c6-
knight, which is without prospects and
obstructing the bishop on b7. Actually
this is a Christmas exercise (or ideal
square exercise, if you like) . Where
would this knight be best placed? The
correct answer is cS.
13 . . .ll:lb8! 1 4 .!De3
1 4 ll:lg3 cS!? gives Black a chance to
build up a powerful centre. Anyway, it
seems obvious that the knight is well The key problem for Black in this
placed on e3. position is the 'hanging' bishop on a6

So l u tions to Ex ercis e s

after the key push ... b5-b4. After Black's 17 ...lbxe4 1 8 Wc2
next the queen's rook does suffer a little 1 8 lL!c3? ltb8 1 9 'i'c2 :Xb21 is sim­
but, potentially, it is only one move ple.
from being activated, so it is not so 1 8 . . .lbdf6
critical. It is more important that the Black is well placed and White has a
rest of the camp is working together. If few co-ordination problems, but per­
White gets time to develop freely he will haps White will still be able to hold bal­
be more naturally placed (mainly due to ance if he respects this. In the game he
the a6-bishop). failed to do so and was soon tom apart.
1 4 . . .1i'c8! 1 9 g4?1
14 ... l:c8 1 5 ..ie3 'i'c7 1 6 ..ift 'i'b7 19 lL!g5! to challenge the strong cen­
17 ..if4, as played in Van Wely-Karnsky, tralised knight seems to be better.
Amsterdam Donner 1 996, seems easier 1 9 . . .'1'd7 20 g5?! �h5 21 i.h2 f5!
for White, and Black has to consider the Black's advantage is now obvious.
fate of the d-pawn. 22 lbc3 .labS 23 llab1 i.xc3!
1 4...Wb6 1 5 ..ie3 illustrates beyond Maintaining the powerful knight.
any doubt why the queen cannot go to 24 bxc3 l:txb 1 25 l:txb 1 i.c4 26
b6 - the c-pawn is pinned. lbd2 l0xd2 27 'W'xd2 f4! 28 l:te1
1 5 i.f4 l:te51? 29 l:te4? .l:txe4 30 .txe4
This is the test of Kasparov's new 1i'xh3 3 1 .tg2 'ii'g4 32 We1 lbg7!
move. Possibly better is 1 5 ..i fl ! b4 1 6 33 f3 'ii'xg5 34 'ii'b1 �f5 35 Wb8+
lbb5 '5'b 8 1 7 'i'a4!? ..ib7 1 8 .tf4 ..if8 �g7 36 •xa7+ �h6 37 'ii'f7 i.f1 1
1 9 axb4 lbxe4 20 lbxd6 .i.xd6 21 'ifxd7 38 �xf1 l0e3+ 39 'it>e1 'trh4+ ! 40
.txf4 22 gxf4 'i'xf4 with a highly com­ �e2 'l'xh2 41 �d3 �f5 0-1
plicated position. Later 1 5 .tn was
played in several games, with the con­ Exercise 4: Black to move
clusion (thus far) being that chances are Miles-Korchnoi
even. Lugano 1 989
1 5 . b4 1 6 �4 b3!

This is the key idea. Black jurrips at

the opportunity to establish a strong
knight on e4 and open b-file simultane­
ously - and free of charge! If you did
not ftnd this move, you did not fully
solve the exercise.
1 7 1i'xb3
1 7 'fib 1 is possible, but take a look at
the rook in the comer, the queen and
the knight on a4, and then look at
Black's pieces. It is clear that although
White does not lose the e-pawn right This exercise is really quite simple.
now, something bad is going to happen. The light-squared bishop is not too im-

Exc elling a t Positio n a l Chess

pressive and has little scope for im­ 55 i.g6 .tc1 + 56 �e2 .i.g5 57 .tf5
provement, while White's other bishop Ad4 58 i.g6 �b4 59 .tts .:.d3 60
is superior. Therefore the following ex­ �e1 l:d4
change sacrifice, creating a formidable Easier was 60 . . l:ta3! 61 l:lxa3 �xa3

centre, is quite logical. 62 Le4 �xa4 63 <li'dl <ltb3 64 .ic2+

21 . . .ll:lxe4! 22 'if'xc2 l:xc2 23 ..td3 �c3 65 i..a4 �b4 66 .ic2 a4 etc.
l:l.xb2 24 l:txb2 ll:lc5 61 �e2 e3 62 �f3 l:tf4+ 63 �g2
Black has compensation and is per­ .:td4 64 �3 �b3 65 l:ta1 Af4+ 66
haps even a little better already. Either �g3 llf2 67 i.e6+ �c3 68 ::tc1 +
way, the position is more difficult to �d4 69 �h3 e2 70 <ii>g3 :n 0·1
play for White in practice. Neither his
bishop nor his rooks have any good Exercise 5: White to move
SCfUares. Korchnoi-Timman
25 �c2 Ac8 Brussels 1991
25 . . f5!?
. is possible, but this might
give White a sense of purpose as the
pawn chain can be attacked with f2-f3
and g2-g4, thus revitalising the bishop.
Therefore Black is doing the logical
thing - flrst improve your strongest
piece! This is the endgame and these
things count!
26 l:te1 ll:ld7 27 a4 �8 28 g3 f5 29
f3 �7 30 �g2
30 g4!? was more optimistic, but
White still seems to have too little po­
tential activity. This is really a difficult exercise.
30 . . ..:.c4 31 lla2 ll:lf6 32 l:td1 �e7 There is no clear-cut way of deciding
33 lld3 �d7 34 l:td1 l:tc5 35 �b3 between the various possibilities. White
llc3 36 i..c 2 i.c5 37 l:l.d3 llc4 38 should consider whether the two bish­
d6 e4 39 fxe4 fxe4 40 i.b3 l:tb4 41 ops are enough for the pawn deficit,
flc3 ¢'xd6 4 2 h4?! �e5 and this appears to be the case if we
42 .. . .id4! was stronger. compare pieces and SCfuares. The d6-
43 i.f7 l:tb1 44 i.xg6? pawn is soft and will most likely fall
44 �h3! tLidS 45 l%c4 keeps the dis­ anyway (lLie4 is a possibility). The
advantage to a minimum. choice is now how White should choose
44 t0d5
... to maintain the bishops.
Now Black is much better. 1 7 .txd6-?
45 l:cc2 �e3+ 46 �h3 �xc2 47 This affords Black easy development
l:txc2 :tb4 48 l:ta2 i.d4 49 �g2 and restores his control over the light
i.b2 50 h5 �d4 51 �f2 �c3 52 SCfl,lares. Indeed White should be careful
�e3 �b3 53 i.f7+ �c2 54 g4 �c3 now. Alternatives:

So l u tions to Ex ercis e s

1 7 .ia3! seems best. Black is weak on suffers from tardy development.

a number of dark squares but it is only 1 9 . . . .id7 20 g3 l:tac8 21 .ih37
d6, e7 and d4 he is trying to control. lile5! 22 :hd1 :hd8 23 .ig2 :xeS
Now 1 7... e5! is practically forced as it is 24 .ixb7 a5 25 .ie4 lbg4? !
hard to see any other moves (after 25 .. £5! is preferable.

17... '1te7 1 8 .id3! Black cannot allow 26 l%6d2 f5 27 .ig2 lbf6 28 f4 l:b8
.i.xf5 and his king is awful on e7, e.g. 29 .cte 1 h6 30 l:td4 ltb4 31 lted 1 g5
1 8 ...lt:Jfd4 1 9 l:lhe 1 and it is evident that 32 a3 .ctxd4 33 ltxd4 lt:\g4 34 h4!
Black is in trouble - with tiJdS just one lbe3 35 if3 gxh4 36 gxh4 .ic6 37
of several ways for White to infiltrate - ..te2 lild5 38 '1;;d2 lbxc3 39 bxc3
although there is no forced way to an ..te4 40 :c4 :d5+ 41 �e3 'itd6 42
overwhelming advantage). After .ctc8 ltc5 43 .ia6 � - �
(1 7 ...eS!) 1 8 t2Je4 0-0 1 9 t2Jxd6 White
stands well. One option is to break Exercise 6: White to move
Black's centre with f2-f4, while also in­ Karpov-Portisch
teresting is 1 8 tiJdS!? (Korchnoi) with Tilburg 1 988
the idea . of t2Jb6, when Black's queen­
side is rather cramped.
1 7 .ib6!? is an interesting move with
some justification. The key idea is to
prevent Black from playing ... Ad8 to
protect the d6-pawn. Later White will
attack it with .ic7 and t2Je4. My main
objection to this move is that the knight
in many lines after i..a3 comes to b6
and dominates completely. It is not easy
to decide, but this is probably only the
second best move.
17 g4? does not work at all Black, White clearly has a space advantage
naturally, will not allow his bishop to be and prospects of an attack against the
locked in, so after 1 7... dxc5 1 8 gx£5 he enemy pawns. His best piece stands on
will play 1 8... e51, when White is strug­ c6, which can be neutralised by Black's
gling for equality. Look at the d4-square · knight but just not right now. 38 .ie2!?
- White lost control of it by exchanging is a fine move, but Black would never
his bishop. On 1 9 :gt both 1 9. g5!? .. play 38 ... a5?, losing what is left of his
and 19 .. .'itf7 20 t2Je4 lt:Jd4! appear to be light square control. Remember that
better for Black, although there are still your opponent will try to counter your
tactics to consider. plans, so 38... l:la8! is the only move.
1 7 . . .lbxd6 1 8 J:txd6 '1;;e 7 1 9 c5!? This position had already been on the
Trying for more than the position . board, in fact. Now after 39 g4 White
promises. After 19 :d2 the position is appears to have achieved a preferable
in balance. Notice that Black no longer version of the game, which also ga've

Ex c e lling a t Positio n a l Chess

him an advantage. However, concrete 41 'i'xb6?! 'i'fS and White is too far
calculation reveals that the black rook is away from the !dngside.
better placed on aS than on e8 if the 41 ....:h87!
world collapses. After 39 ...hxg4 40 41...'ii'c7, favouring White, was nec­
.i.xg4 lLk:S! it is good that a6 is pro­ essary. The rook is simply best placed
tected. Therefore Karpov chose the on e8. White's next move-would proba­
appropriate path when he decided on bly be 42 :tb3!, overprotecting e3 and
his next. preparing an advance on the kingside.
38 g4! 42 ...g4 ...e8 _

38 h4?! merely creates a weakness on 42...'i'c7 43 f4 and White is coming.

the dark squares. A possible variation is 43 :xb6 .:h4 44 •f3 .:xa4 45 :ba
38...'�g7! (preparing for g3-g4) 39 g4? ...d7?1
hxg4 40 J.xg4 .:th8 41 hS .!LieS! 42 A mistake. After 45...'ii'e4 46 'i'xe4
o!LlxeS J.xeS and Black suddenly has a :Xe4 47 :as .:ta4 48 o!Llb8 :as (48... a5
very active position in which there is no 49 .:ta6) 49 e4 White has some advan­
reason why he should lose. Note that 43 tage but it is stillhard to win.
hxg6?? 'i'c2! 44 :n (44 �f3 fS!) 46 .:as .th4 4 7 e4 ..tf6 48 •d3
44...'i'xg6 followed by ... f7-f5 is win­ 1ib7 49 .:ba •d7 50 •c2 .:a3 51
ning for Black. Ab3 :a 1 7
38 . . .hxg4 51 ...:Xb3 52 'ii'�b3 followed by 'W'a3
38... h4 has been suggested but after gives White very good winning chances.
39 'i'f4!, intending g4-g5, Black is in Now it is over, but I've thrown in an
trouble. extra exercise here.
39 .ixg4! �g7
Black cannot avoid the exchange of Exercise 7 : White to move
bishop for knight as 39 ...o!Llc5 40 'iixb6 Karpov-Portisch
'i'xb6 41 :Xb6 o!Llxa4 42 Axa6 gives Tilburg 1 988
White a clear advantage.
40 .txd7 •xd7

52 .:t3! 1ib7
52...J.e5 53 lLlxeS dxeS 54 'i'b2 l:ta4
41 Wi'f4! 55-'i'xeS+ �g8 56 Ab3 and White wins.

So l u tions to Ex ercis e s

53 J:txf6 'ii'b5 54 'ii'c3 'ii'f1 + 55 �g3 crucial part. White's only way to break
'ii'g 1 + 56 �h4 1 -0 open the queenside is through a4-a5, so
Black will need to play .. ..:bs at some
Exercise 8: Black to move point to protect b6. But then we need
Yusupov-Lutz to remember that ltle3-c2-a3-b5 will
Horgen 1994 most likely follow, so we need a safe
spot for the king. This is b8, hence
Black's next two moves.
1 9 . . . .:b81 20 lLlc2 .:b7 21 lLla3 �b8
.Black would always have to play this,
but by doing so right away he allows
himself the opportunity to decide on his
plan for counterplay with the exact
knowledge of where White is going, a
luxury he did not have on move 19.
22 lLlb5 l:td8
And Black has a solid position.
23 ..te2 'iihS !? 24 'ii'a4 a5! 25 ..i.xc8
Closed positions are often the most lLlxc8 26 l:l.b2 'ii'e8 27 'ifc2 'ifd7 28
difficult to play and, unfortunately, you ..te3 % - %
cannot always avoid them. Here Black
needs to organise his forces in the best Exercise 9: White to move
possible way. Yusupov-Rozentalls
The primary concern for Black in this Germany 1 99 5
position is to find a. safe- heaven for his
king. This is not so easy! Players with
prior knowledge of this type of position
can probably find some of the ideas, but
the initial concept is very difficult to
find. First, castling kingside seems very
dangerous. White will not find it t<:>o
difficult to attack the h6-g5 pawn duo in
view of the missing dark-squared
bishop, and Black has little chance of
creating active counterplay on the
queenside in time. Secondly, castling
long would send the king the right way, White is generally well placed here,
but at the same time leave the defending but so is Black. Both sides have com­
pieces on the kingside. So the right pleted development, but this is not the
move 1s ... middlegaineQ); and there is no possibil­
17 . . .�d8! 18 ..i.g4 'it>c7 19 l:.b1 ity of attacking the kings. Consequently
So far so good, but now comes the it is more precise to talk about the ttan-
Ex c elling a t Positio n a l Chess

sition from middlegame to endgame. placed.

Since no immediate action is any good 21 Jbc4 22 .l:.d7

White should simply improve some­ White has a slight advantage which,
thing, and that is the king. in the game, proved to be enough.
1 6 �1 1 22 .l:.c7 23 .l:.ad 1 1 f6 24 l:xc7

This was a new move at the time, ltJxc7 25 l:d6 l:b8 26 'it>e2 e5 27
impossible to find by mere calculation g3!? 'iti>f8 28 .l:.d7 lbe6 29 'it>e3
yet easy to find if we remember our axb4?! 30 cxb4 l:c8 31 'it>d3 .l:.c6
rules (without being slaves to them of 32 h4 h5 33 i..c3 'it>e8 34 .l:.b7 �d8
course, as we never know which rule is 35 l:b8+ 'it>e7 36 b5 l:d6+ 37 'it>c4
valid when until we investigate it). tL:Ic5 38 .ib4 l:d4+ 39 'it>c3 l:td3+
1 6 ltJd4 .idS! sees Black take control 40 'ittc2 l:d4 41 .l:.b7 + ! �e8 42
of c4 and equalise immediately (now the .ixc5 bxc5 43 l:xg7 l:b4 44 l:b7
b2-b.ishop is going to have problems 'it>f8 45 'it>c3 'it>g8 46 b6 �f8 47 g4
activating). hxg4 48 fxg4 'it>g8 49 h5 'it>h8 50
1 6 i.bS has similar drawbacks, since h6 1 -0
after 1 6 ... i.c6! 17 i.xc6 �c6 White
has problems with the light squares Exercise 1 0: White to move
once again. Then we have 1 8 bxaS Karpov-Portisch
(White is trying to force something, but Tilburg 1988
perhaps it is safer to force a draw with
1 8 c4!? :xc4 19 bxaS bxaS 20 i.xf6
lLlxf6 21 ZlxaS) 18 ...bxa5 19 lt:leS but
Black just plays the cool 1 9 .. .l:tb6! and,
if anyone is better, it is Black. After 20
i.cl He can remain cool or force a
draw with 20 ... l0xe5 21 :XaS :cs 22
l:txeS lLldS 23 .i.d2 ltbc6 24 l:te4 lLlxc3
25 i.xc3 :xc3, when White's winning
chances seem quite slim.
1 6 . . . .id5 1 7 l[)d2 lDe5
1 7 ...axb4 1 8 cxb4 .l:txa1 1 9 ..txa1 and
the bishop pair secures White a slight This is actually our Karpov-Portisch
advantage. game from exercises 6 and 7. Here
1 8 f3 lDe81? Black's only good thing is the c-pawn
Going to d6-c4. Note the variation and the potential force of the queenside
1 8...lLlc4? 1 9 lLlxc4 .i.xc4 20 .i.xc4 pawns, so Karpov chooses to eliminate
l:txc4 21 bxaS bxaS 22 %haS - this time them as · rus own advantages will never
it works! go away (the weakness of c6 and the
1 9 e4 .ic4 20 tL:Ixc4 l[)xc4 2 1 prospects of a queenside offensive).
.ixc4! 1 9 i..e4! !
21 .tel lLled6 and Black is well Simply taking control over d3, but

So l u tions to Ex ercis e s

that is far from the full st�ry. The . . 56 �h4 1 -0

bishop should not hope for e4 to be its
permanent resting square. Exercise 1 1 : White to move
1 9 lLla2 has been suggested. I wonder Korchnoi-Yusupov
what it is with these l:iJa2/lLla7 moves... Horgen 1 995
1 9...b5 and now a typical line is 20 l:iJb4
'ifaS! 21 t:£Jc6 t:£Jxc6 22 dxc6 1i'xa4 and
Black has good play. With lL!a2 White is
forcing play, wanting to reach c6 at once.
1 9 . . .l:teS
The weakness of the kingside is an
illusion. After 19 .. ."..c8 20 f4! t:£Jg6
(20...lLld7 2 1 ..t£5 and White wins
something very soon) 21 ..txg6 hxg6 22
lLle4 White wins material. The weakness
on e3 is not enough to compensate for
this as it is not clear right now how
Black is going to attack it. I originally had the feeling that this
20 i.c2! exercise was rather difficult but it seems
The key idea is to play b2-b3 and put that this is not the case. The important
the queenside under pressure. 20 f4? factor is that none of White's pieces
l:iJd7 21 'i'xc4 tlkS favours Black. other than the knight needs improving,
20 . . .ltcS 21 �4 i.e7 and as there is no reason to alter the
2l ...b5 22 b3 secures White a slight pawn structure this should be done
advantage. immediately.
22 b3 cxb3 23 i.xb3 1 4 �d 1 !
White is better. I think most of you Prefacing l:iJd 1 with 14 h3 is an alter­
can remember the finish of the game. native but as .....ltg4 does not appear to
23 . . ....d7 24 :xeS •xeS 25 .l:b1 be a very strong move there is no need
._f5 26 lbd2 :tbS 27 .td 1 •cs 2S for White to be side-tracked.
�b3 .tf6 29 lt:ld4 ...b7 30 "ii'c21 g6 1 4 . . ..tg4 1 5 e4
3 1 .te2 �g7 32 "ii'b3 llld 7 33 lllc6 White has a clear advantage. One of
l:taS 34 ._b4 ._c7 35 �g2 h5 36 h3 the keys to this position is that the g2-
�gs 37 .td1 :es 3S g4!? bishop is guaranteed to become a
Here we are - Home again! strong piece, particularly after Black's
3S . . . hxg4 39 .txg4! ¢'g7 40 .txd7 light-squared bishop has been ex­
'ifxd7 41 'i'f4 :hS 42 'i'g4 'i'eS 43 changed. Therefore the text serves to
lbb6 .l:h4 44 'irf3 .:txa4 45 l:bS gain space for White. His bishop is
'ird7?! 46 :as .th4 47 e4 .tf6 48 waiting on g2 but, as this is not a posi­
'ii'd3 'irb7 49 l:lb8 ...d7 50 ...c2 .l:a3 tion with mutual attacks, this is okay. It
51 ltb3 l:la 1 ? 52 l:f3! 'irb7 53 l:lxf6 has the potential to be enormous on c4
...b5 54 'ifc3 'i'f1 + 55 �g3 'i'g1 + or dS, for example.

Ex c elling a t Pos itio n a l Chess

1 5 . . . c6 Exercise 1 2: White t o move

1 5 ... .id7 1 6 tbe3 c6 1 7 dxc6 i.xc6 Kamsky-Kramnik
1 8 tt:\£5 'ii'e6 1 9 l:lfdt leaves White in Lucerne t 993
total control.
1 6 lLie3
Also possible, and probably better, is
16 dxc6! llfc8 (16 ... :ac8 17 'ii'd3! llfd8
18 cxb7 and White wins a pawn) 1 7
cxb7 (1 7 .d3 llxc6 offers Black
chances of resistance) 1 7 ... 1i'xb7 1 8
lUc3 .icS and, although Black has
gained some compensation for the
pawn, White should remain on top.
This is a good time to mention what
Alekhine said the best players will alw�s

trade in a material advantage for a positional White has a number of factors in his
advantage, as the latter is much easier to play. favour here but his pieces do not co­
This is the reasoning for Korchnoi's ordinate too well. Fortunately this is
move. easily taken care of. The most obvious
1 6 . . . cxd5 1 7 lilxg4 lilxg4 1 8 exd5 is the knight, which needs to fmd a bet­
l:.ae8 1 9 l:tfe1 tLlf6 20 'i'b3 ter square. On e6 the knight will com­
White has some advantage. pletely dominate Black (g7, f8 and d8)
20 . . .'i'd7 2 1 l:e2 .l:.e7 22 i.a5 llfe8 and keep the e-file closed, until that
23 i.c3 h5 very special moment when the knight
Black is back in business, but moves and Black suffers a serious inva­
23 ...'iff5 was better. sion. Therefore we should first improve
24 -.c4 •fs 25 ltce1 :ca 26 ._b3 the advanced roo� as it occupies the
l:tce8 27 i.b4 'i'd7 28 i.d2! i.c5!? transit square.
29 i.g5 'i'f5 30 h4 e4 31 'i'c2 ltc7 24 ltc6!
32 i.xf6 'i'xf6 33 ltxe4 llxe4 34 Here the rook is now perfectly placed
l:txe4 g6 35 ..,d2 l:te7 36 llc4 i.d6 and will be relatively undisturbed be­
37 �h2 l:te8?! 38 f41? �g7 39 i.f3 cause it is on a light square (where
�g8 40 �g2 'it>g7 41 'i'c3 lte7 42 White is in control). Already a6 is under
'i'c2 :ea 43 a3 �g8 44 1i'b3 l:e7 pressure, while g6 is also a potential
45 ltc8+ <;tJg7 46 -..b& lld7 47 i.e4 target, and in the meantime the e6-
i.e7 48 'i'xf6+ �xf6 49 �3 i.d8 square is under control.
50 �e2 l:tc 7 51 l:td8 �e 7 52 :as 24 ...fxe6
lld7 53 'ittf3 f5 54 ..id3 lld8 55 lta7 Black cannot accept a passed pawn in
ltd7 56 b4 i.bB 57 :as l:tdS 58 the heart of his position (a reasonable
i.c4 �d6 59 a4 Ac8!? 60 a5 llxc4 blockade is unlikely) and 24...1lc7 25
61 ltxb8 ltxb4 62 l:tg8 % -% l:txa6 is clear.
25 tllc 5 'iff7 26 tLlxe6 l:teB 27 'W'b3!

So l u tions to Ex ercis e s

Escaping the pin and staying in con­ There are two natural squares for this
trol over the light squares. An impor­ piece but neither is ideal. On gS the
tant effect of this move is that ... a6-a5, bishop is well placed but, because Black
which Black is aching to play, has now has no issues in the centre, ... t2Jd7 will
been ruled out. most likely come. The second option
27 . . . �h8 28 g3 1i'f5 29 :l.e2 for the bishop is b2, but b2-b3 slightly
Planning tLlxg7. damages White's pawn structure as it
29 . . .1lae7 30 l:xa6 weakens the dark squares. It is easy to
And the winner is ... imagine something with ... i.f6 being
30 ... i.f6 3 1 l:l.e3! �h7 32 :l.f3 'ii'e5 very annoying later. Gelfand discovered
33 lLJf4 'ii'b2 34 'ii'd3 :l.e4 35 �g2 that eS is the Christmas square. How­
Wd4 36 'ii'xb5 .:se7 1 -0 ever, I have my doubts about his move.
Additionally, we should never forget
Exercise 1 3: White to move how our opponent wants to place his
Gelfand-Ljubojevic forces. Therefore we quickly realise that
Linares 1 993 all his pieces have only very limited pos­
sibilities. The c8-bishop belongs on b7,
the queen's rook on d8 (c8) and the
queen on c7. We would, of course, be
happy to do something to disturb
Black's set-up, as well as to help our­
1 2 l:l.d 1 1?
Gelfand gives this an exclamation
mark, but I prefer 12 .lf4l cxd4 1 3
.:tad1 as a more natural set-up. Or even
the simple 1 3 ltfd1 if this is the desired
pattern. Moreover Gelfand's suggestion
White needs to develop the three re­ in the next note indicates that I am cor­
maining pieces. Black also needs to de­ rect.
velop three pieces. First we should try 1 2 . . . cxd4?!
to decide where the rooks belong. In 12 .. .'ifc7! is Gelfand's suggestion as a
the game Gelfand insists that the best more flexible try. Perhaps Ljubojevic
posts are al and d l , which seems had not anticipated White's next move.
slightly odd to me. I think the right 1 3 .i.f4!
squares are d 1 and e1 in view of the Now matters soon become difficult
open files. Or at least d1 and fl, with for Black, whose queen has no good
the potential to use el should there be squares.
time. Make up your own mind, the an­ 1 3 . . ..tc5?
swer is not carved in stone. But the big 13...J.b7 14 tLlxd4 'i'e8! was practi-
question for White is what should be cally forced according to Gelfand. We
done with the dark-squared bishop. see now the problems Black experiences

Ex c e lling a t Pos itio n a l Chess

with space after White took control issue in solving this regards the primary
over c7. (1 4...'ifc8 1 5 tlJfS is very un­ concept. As some of you might not
pleasant for Black). have read my book, or have it dear in
1 4 .ie5 i.b7 1 5 �xd4 �d7 1 6 your memory, then let me redefine pri­
�b3! mary concepts with this position as an
A key move. Black cannot do any­ example. Instead of calculating from the
thing about the eS-bishop in view of beginning (setting Fritz on to the task,
.i.xh7+ etc. for example) we should figure out what
1 6 . . ...g5 we want to do. We have the advantage
1 6 ... i.e7 17 .ixh7+ �xh7 18 ed3+ over a computer in that we know what
is a pawn worth taking. we want to calculate. Here the key thing
1 7 i.g3 .:adS 1 8 �xc5 �c5 1 9 for Black is to get the king to safety, so
i.c2 we should figure out how to do it. For­
White has a dear advantage due to tunately we can make it work directly!
his control over the dark squares. 1 8 . . . 0-0!
19 ... f51 7 20 f3 f4 2 1 i.f2 e5 22 b4! There are simply no alternatives to
�d7 23 c5 bxc5 24 bxc5 �h8 25 this move. I am sure that if you sug­
h4! + - Wh5 26 'i'b5 i.xf3 27 gxf3 gested any, it was with a feeling that it
'i'xf3 28 'ii'd3 ...g4+ 29 ¢>h2 e4 30 really did not work...
'i'xe4 �f6 31 'i'g2 _.e2 32 :lxd8 1 9 Wxd7
:xd8 33 :lg1 g6 34 i.e1 ! Wc4 35 Others fail to trouble Black. 19 l:tacl
:n :es 36 i.b3 'ii'xc5 37 'ili'b2 Wg7 'ii'd6 20 lbc4 'iVe7 is equal and 19
38 J.c3 f3 39 .txf6 + ¢>h6 40 i..xh7+ ..t>xh7 20 li'xd7 i..dS! gives
'ii'd2+ ¢>h 5 41 _.d5+ 'ii'xd5 42 Black full compensation for the pawn.
i.xd5 .:e2+ 43 �g3 1 -0 The bishop is very strong and the heavy
artillery is well placed. Note that
Exercise 1 4: Black to move 20... �d5! is a classical example of un­
Gelfand-Anand forcing play. Instead of forcing the issue
Biel 1 993 Black improves his position without
taking captures and threats into consid­
1 9 . . ..:fd8 20 i.xh7+!
This was White's surprise. If you saw
both this and the best reply, then you
should get out and play some tourna­
20 b4 'iVb6! 21 .e7 !txd2 22 ..i.xb7
'ifxb7 23 :Xa7 'ii'xe7 24 ltxe7 l:tb2
leads to a draw, while 20 l:tacl 'i'b4 21
ltc4 .xc4 is also fine for Black!
20 ... �xh7?
This position is razor sharp. The key Forced was 20 <itf8! 2 1 'ii'a4 l1xd2

Solu tions to Ex ercis e s

and Black's activity will be enough to that when one player has considerable
compensate for the pawn. All of room for improvement in his position
White's pieces could find better while the other has very little else to
squares ... improve, then the latter should consider
21 'ii'xf7 l:xd2?! immediate action. This is such a case.
21...�xg2! 22 �g2 llxd2 (22...'1i'g5+ And the weaknesses to be attacked are
23 �h1 l::txd2 24 l:tg1 helps White; Gel­ on the light squares, starting with e6.
fand provides a winning line, but we 1 7 :le2! �f6 1 8 l:de 1 .tea
have already seen enough) 23 '1i'xe6 1 8...i.g8? 1 9 lDbS and Black loses.
:b6 24 'ii'e4+ :lg6+ 25 �hl and White After 1S ... :de8 19 i.x£5 Black is also in
is a pawn up! bad shape.
22 l:a4! 'ii'g5 23 g31 1 9 ..tc41
White wins the queen! Brilliant. The pressure on the light
23 . . .e5 24 Ah4+ •xh4 25 gxh4 squares continues with a clear advantage
l:td6 26 h5 �e4 27 'ii'e7 l:bb6 28 as a result. Here White uses the weak­
'i'xe5 l:te6 29 'i'f4 1 -0 ness of dS to eliminate the pressure
against d4.
Exercise 1 5: White to move 1 9 . .. llJxc3
Gelfand-Adams 19 ... lDb6? 20 lDbS 'ii'd7 21 i.e6 is
Candidates 1 994 (4th game) the tactical justification of White's play.
1 9 ... lDf4 20 .i.xf4 'i'xf4 21 lDdS '1i'd6
22 lDxf6 'ii'xf6 23 dS is also very un­
pleasant for Black.
20 bxc3 h6 21 a4?1
Gelfand prefers 21 h4, probably with
the idea of h4-h5 and a continued attack
on the light squares, as well as the pos­
sibility of lD£3-gS!? with a lot of enter­
21 . . . b6 22 h4! lL'la5 23 .ta2 c5 24
lL'lg5! .ta6?!
24... 'ii'c7! 25 dxcS! bxcS 26 lbe6
White stands much better. None of i.xe6 27 i.xe6 i.xh4 28 i.x£5 with
his pieces can be said to be really inac­ advantage to White. But this was better
tive and he has no weaknesses. Black, than what follows.
on the other hand, is not very well co­ 25 l:te6 'i'd7 26 'ilfxf51 ! i.xg6
ordinated and has problems on the e­ 26 ... hxg5 27 hxgS.
file in particular. White's d1 -rook would 27 'it'g6! ..,7
be most useful on the e-file. The exer­ If instead 27.. 1tf6 then 28 l:te8+ llffi
cise has a lot in common with the theo­ 29 i.b t .
ries in Dorfman's new book The Method 28 'ii'xf7 l:xf7 29 hxg5 cxd4 30
in Chess. There he states quite logically cxd4

9 '1
Ex c e lling a t Positio n a l Ch ess

30 g6!? was also possible, with the trouble as his bishop is poor and the
idea of 30 ..l:tff8 3 1 J.xh6! gxh6 32 .:f.e7
. structure invites an attack. So the king
and a clear win. stays put in the centre, where it has
30 ....tc4 31 :leS+ l:.xeS 32 l:.xeS+ nothing to do.
�h7 33 .tb1 + g6 34 gxh6 tlJc6 35 The only chance was 1 3...bxc6! 14
i.e3 J:te7 36 J:cS .tdS 37 .td3 l0b4 lba4 Wa5+ 1 5 c3 c5 16 lbxcS .i.xc5 1 7
38 .te2 ie6 39 AdS l2Jd5 40 .tgS b4 which, thus far, you should b e able
J:td7 4 1 :es .tf7 42 J:tb8 .te6 43 to see. You should also know that this is
.tf3 l:!f7 44 J:d8 lf5 45 i.d2 ll)f6 one of those positions where you can
46 l:.a8 g5 47 l:txa7+ 'ii>xh6 48 J:ta6 bend your immediate forcing tenden­
ll)d7 49 a5 ::tb5 50 axb6 1 -0 cies. Open Pandora's box and disappear
in the mist... 17 ... J.xd4!! (1 7 ...J.xb4 1 8
Exercise 1 6: Black to move ax b4 is terrible - obviously) 1 8 bxaS
Gelfand-Adams J.xc3+ 1 9 �f2 0-0 20 .:tel J.xaS fol­
1 994 Candidates (6th game) lowed by . . £7-f6 and Black has good

counter-chances in a practical game, if -

not in the analysis.
1 4 0-0
The rest is just suffering for Black.
1 4 . . . g6 1 5 1Wd2 'flc7 1 6 l:.f3 b6 1 7
lbd1 ! h5 1 8 a4! i.c5 1 9 l:.c3 i.e 7
20 b4! 'i'b7 21 l:.b3 i.d7 22 lbe3
J:cs 23 a5?! .tbS! 24 axb6 a6! 25
:tc3 �d7?
25 ...ltxc3! improves.
26 J:tc5! :xeS 27 bxc5
The move 27 J.xcS! is more to the
This is one of these cases where you point.
need to make a really tough decision 27 . . ..:tea 28 c4 �xc4 29 J:tc 1 �as
between two kinds of positions - a 30 lbxc4 •d5 3 1 ltJaS .tc6 32 .ie3
hopeless position and something close '*e4 33 l2Jxc6 •xc6 34 h3?! 'i'bS
to Pandora's box. The real exercise is to 35 '9i'c2 •c6 36 �2 a5 37 l:.a 1
realise that, positionally, you cannot live :as 38 J:ta4 h4 39 'ifa2 •bs 40
with a static position, and therefore you •c4 •c6 41 '*a2 Wb5 42 c61 •xc6
will have to try something else. This of 43 J:txa5 l:.b8 44 l:.a7 'i'e4 45 •b3
course means that you will have to see �f8 46 J:td7 g5 47 :td4! ...,5 48
the line I am talking about, and have the J:c4! l:b7 49 J:tcS+ �g7 50 J:tc7
will to find the unforcing moves that gxf4 5 1 .td4! •e4 52 '*f3 1 -0
stamp your will on to the position.
Enough talk, here is the solution: Exercise 1 7: White to move
1 3 . . .i.xc6? Gelfand-Adams
After this mistake Black is in serious 1994 Candidates (8th game)

Solu tions to Ex ercis e s

better due to both the knight's superior­

ity and the structural advantages.
17 b3 'i'd3 is just plain bad and the
interesting 17 'i'd1 ! ? does not work out
well after 1 7 ...'i'xb2 1 8 'i'xd6 Wxc3 1 9
'i'dS+ (1 9 :acl 'Wd4 20 'i'dS+ l:t£7!)
1 9.. l:lf7 20 :tacl 'ifd4 21 l:txcS :tdB 22

'i'xd4 lhd4 and the game is on its way

to a draw. One way could be 23 f3 :d2
24 lt£2 l::d l+ 25 l:.£1 :ld2 etc.
1 7 . . . 'ii'c4 1 8 .l:e1 b5 1 9 'ii'g4!?
19 'ite2?! would be a positional mis­
This is a really interesting position take, as explained above.
and difficult to evaluate. Luckily we do 1 9 . . .1i'e6 20 'ii'e2 1tb8 21 a4 bxa47
not need to do so to find the best An improvement is 21 ...b4 22 aS!
move! But anyway, let us resort to one with an unclear struggle ahead. Now
of the old methods of evaluation. White comes out on top (22 lted1
Where should the pieces be placed? tl:laS!).
White, for sure, wants the queen's rook 22 '*'a61 .:b6 23 'W'xa4
on d 1 , while the queen could go to d7 White has some advantage.
or d2 and the bishop - who knows? 23 . . . 'i'b37 24 'i'xb3+ .:xb3 25
White needs to somehow break a hole J:ted 1 ! l:bb2? 26 .:xd6 l:tfb8 27 h3
in Black's centre, or forever live with a .:ab6 28 ltd5 a5 29 i.xe5 l:tb1 + 30
bad bishop. But in the position White is l:l.xb1 l:l.xb1 + 31 �h2 c4 32 J..d6 a4
under attack. 33 ltf5 h6 34 :ta+ �h7 35 .:as
1 7 ::tfb1 ! .:b3 36 e51 J:txc3 37 e6 1 ·0
The only move that helps White in
his quest to obtain his favourite piece Exercise 1 8 : White to move
placement. Now the idea is 'i'g5-d1 -d2 Gelfand-Kasparov
followed by £2-£3 and l:tb1 -d1 (maybe), Linares 1994
with i.g3-f2 and b2-b4 in some posi­
Less good is 17 :ab 1 'i'c4! and here
we actually do not need to know any
more, even though it is nice to assure
oneself that the rook endgame after 1 8
...d 1 :td8 1 9 'i'dS+ 'i'xd5 20 exd5 CiJe7
21 .i.h4 ltd7 22 J..xe7 l:xe7 is slightly
preferable for Black thanks to the f-ftle
and the prospects of an invasion on the
4th rank. Note that 21 c4 CiJ£5 is hardly
an alternative here, when Black stands

Ex c elling a t Positio n a l Chess

Here the primary concept is control 1 6 f3 i.e6 1 7 'ifd2 'fke7 1 8 lbb5 a6

over the dark squares in the centte. The 1 9 lbd4
light-squared bishops each cover theit
diagonal (b1-h7 and a2-g8) and do not
contest each other. But the fight for d4,
e5 and f4 is hard. The appropriate
course for White, then, is to bring his
pieces into this fight as quickly as possi­
ble. The logical way to do this is to start
with the knights.
1 3 li)ge2!
1 3 ltJh5!?, as played in Kasparov­
Yusupov, Linares 1993, is also a good
move, but the text is sttonger.
1 3 . . .l:lc8 White has a solid edge here. His
Black cannot v.rin the fight for the pieces are very well placed and Black
dark squares and instead begins to will continually have problems with the
complete his development. dS-pawn. White also has the possibility
13 ... d4 14 .i.xd6 'Wxd6 1 5 lbb5 fails of an advance on the kingside.
for Black and 13 ....i.xf4 14 lihf4 d4 15 1 9 ...t2Je5 20 b3 i.d7 2 1 g4
lDxe6 fxe6 1 6 exd4 tihd4 17 i.b 1 gives White has established a positional
White a position in which the bishop is advantage and cannot further improve
superior to the knight and where the e­ his position much more, so now it is
pawn is a slight weakness. Allin all not time for the attack. Black is probably
a pleasant position for Black. worse here but should. still be able tD
1 4 i.b 1 i.xf4 defend.
I dislike this move and in the game it 21 . . .lbg6 22 lbg2 lbea 23 .:xeS
also seems to tum out badly. Of course .txc8 24 Ac1 lL!d6?
these things are linked for me as I have 24...i.d7 was better.
seen what happens, and perhaps Kas­ 25 i.xg6!
parov had not. From a static point of Picking up a pawn.
view Black should consider preventing 25 . . . hxg6 26 lllf4 .l:le8 27 lL!xd5
tDc3-b5-d4 with 14... a6!? and simulta­ 1i'g6 28 lL!f4 1i'e5 29 lLIQ2 g5 30
neously retain the tension. 1i'b4 lbb5 31 lbxb5 axb5 32 'i'd4
1 5 li:lxf4 .ig4 1i'e7 33 'iic5 1i'f6 34 1i'xb5 J:ldB 35
White is seriously considering taking 'i'e2 i.e6 36 ,.f2 :td3 37 h4 gxh4
this bishop all the time. This move pro­ 38 lbxh4 g5 39 lL!g2 'ife5 40 'i'c2??
vokes a weakening of White's structure White has-missed some easy v.rins be­
but, as we shall see, this can also be cause of time ttouble and now_ fails
used positively, so one starts to wonder completely. Winning was 40 e4 'Wd6 41
if something like 15 ...a6!? was a better l:e1 %ld2 42 .:te2 .:tdl+ 43 �e1 i.d7 44
move. e5 'i'e7 45 'We3 and White will eventu-

1 00
Solu tions to Ex ercis e s

ally untangle. There are no ways to equalise, or

40 Wd6 41 lLle1 :xe3 42 'i'f2 'i'e5
..• even get near! 13 .. .llb8 14 lLld6 is good
43 ttlg2 .l:!.e2 44 .:I.e 1 . l:xf2 45 ltxe5 for White and 13 ... b3 14 lbd6 ltJdS 1 5
ltxf3 46 l0e3 i.d7 47 l:.xg5+ �f8 .id2! i.xd2 1 6 'i'xd2 lb7f6 1 7 lbc4
48 .!bf5 i.xf5 49 gxf5 Y.z - % .id7 1 8 a4! l:tb8 1 9 l:lfbt 'ii'c7 20 lbfeS
gives White the dark-squared control he
Exercise 1 9 : White to move needs to rule the centre and round up
Gelfand-Epishln the black b-pawn.
Dos Hermanas 1994 13... axb5? 1 4 axb4 seems to 'Win a
pawn and after 13 ... bxa3? 1 4 ltJd6! both
the a-pawns � presumably fall and
White will have the two bishops.
1 4 lDcl6 .i.xf3 1 5 Wxf3 J..c7 1 6
By gaining time White picks up the b­
pawn. Now, with the bishop pair and an
extra pawn, White is on the way to a
good result.
1 6 . . .We7
1 6 ...'i'c8 1 7 axb4 liJdS 1 8 l/JaS!
tLlxb4 19 .ie4 and White is doing very
Dorfman's aforementioned book has well, while 16 ....ixh2+ 1 7 �xh2 'i'c7+
one very interesting observation. He 18 �g1 'i'c3 19 'i'e2 'ifxa1 20 i.b2
noted that there often arises a situation 'i'a2 21 J..c4 b3 22 lla1 sees the queen
in which one player can improve his trapped.
pieces easier than the other, and the 1 7 axb4 e5 1 8 ..txa6 exd4 1 9 exd4
correct reaction to this, should you be Wxb4 20 i..e3 ltfb8 21 l:tab1 'i'e7
the one unable to improve your posi­ 22 i.b5 .:a3 23 J..c6 lLlb6 24 l:fc1
tion significantly, is to react ql}ickly. h5 25 h3 We6 26 .l:!.b5?
lbis position is such a case. But how? 26 lbc5 wins.
Well, White should see two things - the 26 . . .Wa2?
tactic which is really plain to see and the 26.. .lbc4! was better.
ideal square for the c3-knight on d6. 27 'i'd 1 ?I
This should make it possible to fmd the After this Epishin finds an inventive
right execution of the combinational defence - probably not enough to save
idea. the game but, in practice, these things
1 2 b41 cxb4 1 3 lLlxb5! are always difficult.
After 13 axb4 .ixb4 14 lbxbS the 27 ttlbd5!1 28 ltc2 ttlc3!! 29 ltxa2

bishop on b4 protects d6 - hence the ba2!! 30 ..-n lLlxb5 31 g4 lL'ld6?!

reversed order of moves. Without this 31...hxg4! 32 'i'xbS gxh3 33 �fl
point the exercise is not solved. J:at+ 34 �e2 h2 35 'l'c4 �e8 36 i..c t
1 3 i.b7
..• would also give White -extremely good

Ex c e lling a t Positio n a l Ch ess

winning chances. �f7 25 �xg7 + 'Ot>xg7 26 •d4+

32 ltJxd6 �xd6 33 g5 lbea 34 g6 �g8 27 'it>c 1 'Wh4 28 lLJc3 .!Zlg4 29
�f6 35 gxf7+ �fa 36 •c4 l:a7 37 'it>b2
-.e6 l:d8 38 i.b5 l:e7 39 ._f5 1 ·0 29 i.xc8! is preferable.
White won on time. 39...l:tf7 40 �c4 29 . . .�61 30 i.a6 l:tb6 3 1 lbb4 ltJe3
llc7 41 i..e6 offers White good winning 32 l:h 1 ._f2 33 �a2? lbg4? 34
chances according to Gelfand. -.xf2 lLJxf2 35 llhf1 d4+ 36 �a3
dxc3 37 llxf2 liaS 38 �a4 lbxb5 39
Exercise 20: White to move e3 �d4 0-1
St. Petersburg 1994 Exercise 21 : Black to move
Szeged 1994

White must develop his kingside In .

the event of ttlf.3 Black will play .. . f5.f4

and ... g6-g5, trying to generate an attack. White is threatening to play 1 1 f4!, 1 2
White can delay this substantially with i..xg7 and 1 3 ttl f3 with complete con­
the aid of a thrust of his own. trol of the centre. The knight has no­
1 2 f4! where to go from e5 so Black has to act
Black has no way to equalise. at once, using the only exposed piece
1 2 . . . exf4 White has - namely the bishop on d4.
Look at what the prevention of Note that White has good control over
Black's kingside expansion has done to most of the light squares with the
the e7-knightl pawns on dS and e4 working together
1 3 gxf4 'ii?h8 1 4 lbf3 �e6!? with the bishop on e2.
Trying to mix it, but without any suc­ 1 0. . . c51
cess. This pawl.! sacrifice makes it possible
1 5 lbd4 �g8 1 6 .txb7 d5 1 7 c5! for Black to use his slight lead in devel­
White has a very clear advantage. opment to solve his problems.
1 7 . . . l:b8 1 8 c6 ltJg4 1 9 ._b4 lbc8 1 1 ..txc5
20 lLJd1 ! 'Wh4+ 21 'it>d2 ..h3 22 No alternatives, of course.
�b2 l:te8 23 J:e 1 ! lLJxh2 24 lbc2 1 1 . ..lllec4

1 02
Solu tions to Ex ercis e s

1 1 ...�bc4? 12 'ii'b3 b6 1 3 �d4! was Exercise 22: Black to move

not the idea! Kacheishvili-5vidler
1 2 ..ixc4 Szeged 1994
This is forced as 1 2 il.xb6 'i'xb6! 1 3
�xc4 'ii'xb2 does not suit White's inter­
ests. Nor does 1 2 'ii'b3 �xb2! 1 3 :c1
(13 'ii'xb2? �a4 14 �xa4 ..txb2 1 5
.!Dxb2 'i'aS+! wins for Black)
1 3...i.xc3+! 1 4 'ifxc3 �2a4 when Black
retains his lead in development and
eliminates the pressure in the centre
without losing anything (obviously
Black is better here).
1 2 . . .lLlxc4 1 3 'i'b3 lL\xb21
White has no control over the light
squares and Black should therefore seek White is only just mainwrung a
to exploit the momentum. It is often blockade on the dark squares. Although
the case that when you have a lead in Black is a pawn down he chooses to
development and are entering tactics exchange the gueen to be able to use his
you should keep an eye open for any of advantage on the dark squares to the
the opponent's pieces that might be maximum. The thing I find important
hanging. Remember what John Nunn here is the strength of the g7-bishop, a
said: Loose pieces drop off. factor which is not permanent but
1 4 'W'xb2 'ifc7 rather dependent on the long diagonal
Here the bishop is struggling on cS. (and the fact that White has not yet
1 5 .ib4 played £2-f4 and e4-e5).
15 Wa3 b6 16 .ib4 aS 1 7 :c1 axb4! 20 .'ii'd6! 21 9xd6
. .

1 8 'ii'xb4 .ia6 and Black will soon have Also possible was 21 :ab1 'ifxb4 22
a winning attack. All his pieces are play­ :Xb4 :fc8! when Black is not disap­
ing, none of the White's are. . . pointed about trading off his worst
1 5 . . . a5 1 6 lt)ge2 axb4 1 7 1i'xb4 piece. Indeed he is happy at the coming
..ig4! invasion on the 2nd rank - 23 �xbS
Inducing a weakness on the dark i.xbS 24 l:txbS :Xa2 25 lLlgt i.h6!
squares. (when you have the initiative you
1 8 f3 ..id7 1 9 0-0 b5 should try to address certain problems
Black has good compensation for the with the time in mind; here the back
pawn. rank problems are solved at the same
20 'iirh 1 ? time as the bishop threatens to come to
, Better was 20 l:.abl! .:tfc8 2 1 lDxbS e3 to offer valuable support in an at­
i.xbS 22 'iVxbS l:.xa2 and the position tack) 26 f4 :cc2 and Black is better.
remains less clear. 21 . . . exd6 22 .l:tab1 .l:tfb8 23 l:tb4
And here is another exercise: The rook is a sad blockader, having

1 03
Exc elling a t Po sitio n a l Chess

none of the flexibility enjoyed by In this position we have a King's In­

knights, bishops and queens. dian Defence with the variation being
23 . . ..:ta3 24 :tc1 the Four pawns Attack. The difference
White is lost after 24 l:tb3 b4 2S between normal positions from that
l:txa3 bxa3 26 lib 1 .:tb2, when ... .ibS! is variation and the specific one here is
coming. that White has his bishop on gS. This
24....l:.c8 25 ltb3 weakens the dark squares behind the
25 .!DxbS llxcl+ 26 .!Dxcl l:.e3! and pawns, and this drawback is excellently
Black wins. exploited by Nunn with the following
25 . . . l:txb3 26 axb3 :ae! move.
White cannot keep the rook out for- 1 0 ...�h51
ever. Obviously with the threat of ...h7-h6.
27 ltb1 ? 1 1 '6'f3
27 :cz! l:.al+ 2B .!Del looks fragile, This is probably the most realistic
but it was the only chance. The rook is move, but black can still take over the
simply too passive on b 1 . initiative. Nunn gives the following line'
27 . . .l:ta3! 1 1 g4 f6 1 2 gxhS fxgS 1 3 hxg6 (1 3 fxgS
Preparing ... b5-b4, but without want­ .!DeS 1 4 llg1 axbS 1S .ixbS c4! 1 6
ing the rook to be trapped on the out­ .!Dxc4 .!Dxc4 1 7 .ixc4 '6b6 1 8 :gz
side. After 27 . b4 28 .!Da4 .ixa4 29
.. .ixc3+ 1 9 bxc3 'i'e3+ and black wins.)
bxa4 .l:txa4 30 .!Del White is in a horri­ 1 3...l:txf4 14 gxh7+ 'ifi>h8 and the domi­
ble situation but still defending. Now nation of the dark squares as well as the
the game is very easy for Black. much better placed king secures black
28 lt:ld1 :la2 29 �f4 b41 30 lDd3 an excellent position.
.l:e21 31 g3 i.b5 32 lDxb4 .:e1 + 33 1 1 g3 h6 12 .ih4 axbS 1 3 .ixbS .ia6
�g2 .id4 34 lDc2 lte2+ 0-1 14 .ixa6 :Xa6 1 S.O-O :b6 1 6 .:tb1 :b4
is also unclear - Nunn.
Exercise 23: Black to move 1 1 . . .f6 1 2 i. h4 ltJxf4 1 3 1i'xf4 g5
Korchnoi-Nunn 14 'ift2
Reykjavik 1 988 1 4 .ixgS fxg5 1 5 'ilfxgS .!DeS would
give black a very good compensation
because of the strength of his bishop.
1 4 . . . gxh4 1 5 1i'xh4 �e5 1 6 .i.e2
axb5 1 7 .i.xb5 'ifa5!
Black is better.
18 'i*'g3 1i'b4 1 9 :b 1 l:.xa2 20
tDxa2 li'xb5 2 1 liJc3?1 ifa6 22
�d1 f5 23- :n llf6! 24 exf5 i.xf5
25 lht5 Axf5 26 �c2 �h8 27
�b3? i. h6 28 .l:a 1 ? l:.t 2 + ! 29
� b1 :n + 30 �c2 :L xa1 31 �xa 1
'ii'xa1 32 'ii'h3 'ifc1 + 33 �b3 c4 +

1 04
Solu tions to Ex ercis e s

34 � a2 lt:Jd3 35 'ii'cS + �g7 36 1 2 . . . exf4 1 3 .ixf4 d6

'ii'b7 .i.g5 37 ltJe4 .i.f6 38 ltJxd6
�g6 39 ltJcs c3 40 d6 ._xb2 + 41
.xb2 cxb2 42 d7 lt:Jb4 + 43 � b1
ltJc6 44 g 4 �g5 4 5 h3 � h4 46
ltJb6 h6 47 t0d5 �xh3 0-1

Exercise 24: Black to move

Haifa 1 995

Black is fine thanks to the control

over eS.
14 �h 1 ¥! 1 5 .te2 tt:lge5 16
tt:ld4 'ile7 1 7 .C.b 1 tt:lxd4 % - %

Exercise 25: Black to move

St. Petersburg 1 995

For Black it is important that he justi­

fies putting the bishop on b4, otherwise
he is simply lagging behind in develop­
ment. Moreover White's development
lead will be used to organise an assault
in the centre or on the kingside (proba­
bly both), starting with e4-e5. The solu­
tion is to give White structural weak­
nesses and then prevent this advance.
1 0 . . . .ixc31 1 1 bxc3 e51
The simplest way to prevent e4-e5.
Note .that this fits in weU with ... .1l.xc3 An important feature here is that
as there is now no l!Jc3-d5 to look out Black will not be okay in positions such
for. as the one after, for example, 9.. .exf4 1 0
1 2 c4 .i.xf4 �c6 1 1 ...d2 l!Jg4 1 2 0-0-0 lLlge5
12 fS? dS is just bad, and 12 fxeS 13 .i.e2, which is slightly better for
lL!xe5 13 i.g5 'ifb6+ 1 4 �h1 lLlfg4l 1 5 White according to Svidler. ActuaUy I
1i'e2 1i'g6 1 6 .i.£4 d6 sees Black assume think it is worse than that. The control
full control over the dark squares (and over dS is clearly more important than
limits the scope of the d3-bishop). eS, as from dS there are possibilities for

7 05
Ex c e lling a t Positional Chess

direct hits into the enemy camp. Conse­ good reason to do here.
quently, traditionally, White is a little 1 2 i.c 1 i.g7 1 3 h3 tDe5 14 i.e3
better in these positions. Below you will g4! Yz - Ya
find the game Karpov-Polugaevsky, And i n this dynamically balanced po­
which goes some way in illustrating this sition the players agreed a draw. Actu­
concept. ally I would prefer Black's position as it
Anyway, in the diagram position the seems to have more dynamic potential
hS-pawn is exposed, so Black is actually than White's.
worse off than usual because it is not
clear where his king belongs. Karpov-Polugaevsky
9. 'i!fc7 10 f5 i.c4 1 1 .i.xc4 Wxc4 12
.. Candidates, Moscow 1 974
'Wd3 is also better for White. Look at Sicilian Difence
dS and b6.
9 lbg41 1 0 i.d2
... 1 e4 cS 2 lD£3 d6 3 d4 cxd4 4 l2Jxd4
10 'i!i'd2 ltlxe3 1 1 'iixe3 exf4 12 'ifxf4 lLlf6 5 lDc3 a6 6 .i.e2 e5 7 tiJb3 i.e7 8
lL!c6 1 3 0-0-0 .i.e7 14 �b1 g6 1 5 ltldS 0-0 .i.e6 9 f4 'i!i'c7 10 a4 t0bd7 1 1 <iifh1
.i.gS 1 6 'i!fg3 ltleS should be fine for 0-0 12 i.e3 exf4 1 3 l:txf4 lLle5 1 4 lLld4
Black. The bishop compensates for the l:.ad8 1 5 ..g1 l:td7 16 l:.d1 lte8 17 tiJfS
weakness of the king's position. i.d8 1 8 lLld4 lLlg6 19 l:.ff1 ltle5 20 .i.f4
1 0 . . .exf4 1 1 i.xf4 'i!fc5 21 tbxe6 'ifxgl+ 22 lhg1 l:.xe6 23
i.£3 lLleg4 24 :.g£1 i.b6 25 l:td2 i.e3
26 .i.xe3 liJxe3 27 l:.b1 <iiff8 28 �g1
.:.c7 29 �£2 lLlc4 30 :d3 g5 31 h3 h5
32 lLldS lLixdS 33 ltxdS lDeS 34 c3 h4
35 l:bd1 �e7 36 .!:.1d4 f6 37 aS l:.c6 38
i.e2 �d8 39 c4 �c7 40 b4 lLlg6 41 bS
axbS 42 cxb5 :cz 43 b6+ �d7 44 ltd2
llxd2 45 .l:!xd2 l:.e5

1 1 . . . g5!
This move is absolutely essential for
the concept. Now Black is able to de­
velop his dark-squared bishop to an
active post, while after 1 t ...lZk6 1 2 'it'd2
lDge5 .i.f8 there are problems finding a
good place, allowing White to retain the
better prospects. The tempo is not so
important as Black is playing for static 46 a6 �c6 47 l:.b2 ttJf4 48 a7 :as 49
features, something only White has a i.c4 1-0

1 06
Solu tions to Ex ercis e s

Exercise 26: White to move bishop well placed on b2? This is not so
Solozhenkin-Svidler dear. Secondly, does Black have a
St. Petersburg chance to use the fact that White is un­
developing? It appears that 1 4...c6!? 1 5
b 3 (1 5 'ii'd3 bS 1 6 b3 b4!? and Black has
counterplay against e4, although the
position cannot be described as any­
thing other than unclear) 1 5.....a5 1 6
'it'd3 tbdS!, for example, is successful.
1 3 . . .c5 1 4 lbb3!
14 lbf3 .ic6 15 eS dxeS 1 6 tOxeS
..ixg2 1 7 'ii'xd8 (17 'iti'xg2?! Wc7! and
Black has a better Icing's position)
1 7 ... ltaxd8 1 8 'itxg2 lbd5! is equal ac­
cording to Svidler.
This position is quite tense. Black has 1 4 . lt'lc4 1 5 e5!
. .

numerous dynamic possibilities, with Necessary. Black has so much dy­

...c7 -cS and ..lbc4 coming to mind. As­
. namic power and so much pressure
sisting Black's cause is the plan of against White's centre that the text,
.....ic6, putting pressure on e4, and ...b7- challenging for the dark squares, is ab­
b5 with the intention of pushing further solutely essential. 1 5 'ii'd3 b5! 1 6 e5
with ...b5-b4. The move Solozhenkin dxe5 1 7 .ixa8 ...xa8 is winning for
chose in the game dissolved all of the Black. Rooks have nothing to do in
problems and gave him a small advan­ such positions.
tage. I believe he chose the right path, 1 5 . . . dxe5 1 6 fxe5
but as we shall see, it is not so apparent. 1 6 .i.xb7 exf4! and White has won
1 3 i.f2! nothing, or 16 lbxc5 exf4 1 7 gxf4 and
1 3 b3?! c5 14 tbde2 i.c6 1 5 Wd3 b5 Black has many ways to get an equal,
and White is not really well co­ albeit attractive position (17 .. .'i'c7).
ordinated. Additionally, 1 3 'ii'd3 cS 1 4 16 lt'lxe5
• . .

tbde2 (14 tb £3 .ic6 1 5 e S dxe5 1 6 Also possible was 16...l:.xe5J? 17

'ii'xd8 l:.axd8 1 7 tbxeS .i.xg2 1 8 'iti'xg2 :1xe5 lDxeS 18 lDxcS .ic6 19 'i'xd8+
ltJdS! is just equal) 14 ...b5!? presents (19 ll'lxb7 Wxd l+ 20 llxd1 .i.xb7 2 1
Black with several aggressive opportuni­ .ixb7 ltb8 2 2 .ixa6 :Xb2 23 .1d3 is
ties, knowing that 1 5 'i'xd6? tfr4 1 6 slightly better for White according to
'it'xc5 .l:tc8 1 7 ..a7 l:.c7 1 8 'it'd4 tbxe4 Svidler, and perhaps he is right, but af­
19 'ii'd3 lbxb2 wins for Black. ter 23 ...lbxd3 24 llxd3 Lc2 it is hard
Rather interesting is 13 g4!? with the to see it as winning chances) 1 9 ...1:.xd8
following idea: 1 3. ..lbc4 14 .ict , and 20 ...txc6 tbxc6 21 :dt l:xdt+ 22 lbxd1
now White wants to play b2-b3 and t0b4 23 c3 lbxa2 24 lbxb7 with a draw
.i.b2. But there are some questions that likely.
need to be put to this plan. First, is the 1 7 lt'lxc5

Ex c e lling a t Po sitio n a l Chess

17 .ixb7 .i.xh3! and Black is better. the dark squares. After 1 0 dS? Black has
1 7 . . .'i'c7 no problems on the dark squares and
Here Svidler prefers 1 7 . .ic6! 1 8
.. can use his slight lead in development
'ifxd8 l:taxd8 1 9 .i.xc6 lbxc6 2 0 lbxb7 on the kingside to open up the position
l::tb8 21 lbd6 Zlxel+ 22 lhe1 :Xb2 23 immediately with 10 . £5!, when the tac­
. .

l::te2 with equality. tical justification is 1 1 ex£5 .i.x£5 12 g4

1 8 lbxd7 ttJexd7 1 9 :xeS + ! :xeS e4! and Black has the advantage.
20 'i'f3 1 0 . . . dxc5
White has emerged from the opening Now Black is planning ...t2Jb8-c6-d4,
with a small edge. Now the greater play­ after which his position would be okay.
ing strength decided the game. White has to prevent this.
20 . . .liJc5 21 Ae1 ! l:txe 1 + ? ! 22 1 1 1i'd51
.txe1 b5?! 23 a3! h51 24 .tf2 tbfd7 Forcing Black s next.

25 'WaS+ �h7 26 �5?1 'We5 27 1 1 . . .liJa6 1 2 0-0-0

VcB 'i'xb2! 28 liJe7 i.d4 29 1Vg8+ White has an edge thanks to the two
�h6 30 'i'xf7 .txf2+ 31 'i'xf2? bishops. Now . fl-fS also becomes ir­

'i'xa31 32 1i'f4+ �g7 33 'Wd4+ �7! relevant as ex£5 would leave Black with
34 �c8 'Wxg31 35 tbd6+ �e7 36 a wounded pawn structure.
liJcS+ �dB 37 liJd6 �7 38 tbe8+ 1 2 . . .'i'f6
�b6 0-1 As is often the case there is a cross­
fire between tactics and the ·_positional
Exercise 27: White to move aspects of the position, and when one
Yakovlch-Solozhenkin player has accumulated positional ad­
St. Petersburg 1995 vantages even razor sharp tactics have a
tendency to go his way. The following
line is a good illustration of this -

12 ... i.e6 1 3 .d6 .f6 1 4 lLlc3 btfd8 1 5

liJdSI 'i'h4 1 6 'i'e7 'i'xe7 17 lbxe7+
�£8 18 liJdS and White has a better
1 3 liJc3
The alternative 1 3 j_xcS? .i.e6 1 4
.d6 ltJxcS 1 5 'i'xcS l:.fc8 i s obviously
wrong. .
1 3 ...tLlf4 14 1Wd6 .te&?l
14 ......xd6 15 !hd6 lLle6 16 tiJdS
This exercise involves a simple deci­ lbd4 1 7 f4 leads to only a modest plus
sion. White will have to decide between for White.
d4-d5 and dxcS. And it is not a very 1 5 'i'd2!
difficult decision in fact.
, Now White h as a clear advan tage as
1 0 dxc51 the knight has nowhere sensible to go
White is able to us.e his superiority on from f4 (d4 is suddenly far away).

1 08
Solu tions to Ex ercis e s

1 5 . . .lt:Jb4 1 6 ..txc5 l:fd8 1 7 ,..e3 tion as the bishop is clearly White's

l:[xd 1 + 1 8 ltlxd1 tt::lxa2+ 1 9 �b1 worst placed piece.
"ii'd8 20 ltlf2 l:tc8 21 Wxa2 'W'a5+ 1 9 tt::lxg6 tt::lxg6 20 ftJf5
22 .ia3 b5 23 g3 b4 24 gxf4 exf4 White has the advantage.
25 ,..c 1 bxa3 26 b3 l:[bS 27 .id3 h5 20 . . ..ic5 21 l:d2 :tad8 22 l:[ad1
28 ,.c2 ,.c5 29 lbd 1 a5 30 �c3 �f4?!
'it'e3 31 'W'e2 'ifd4 32 l:c1 J:!.d8 33 The cause of Black's coming prob­
tt::ld5 %tb8 34 l:tb 1 1 -0 lems. 22. . .l:e5 23 .1xc5 'l'xcS 24 l:td4 is
a shade favourable to White, whose
Exercise 28: White to move knight will be well placed on e3, while
Svidler-Dreev Black has difficulty generating counter­
Novosibirsk 1995 play.
23 'W'a4 ..txe3 24 fxe3 ftJe6 25
l:txd5 h5?! 26 lbd6 l:f8 27 'i'b4 g6!
28 c4 b6 29 "ii'c3 tt::lc5 30 Wf6 liJe6
3 1 b3 b5 32 c5 b4 33 tt::le4 l:txd5
34 l:txd5 l:tc8 35 Wf1 ! Wc6 36 Wa1
l:d8 37 lbd8+ .!Llxd8 38 Wd4 ftJe6
39 •xb4 f5 40 .!Llc3 Wxc5 41 'i'xc5
tt::lxc5 42 b4 tt::ld3 43 tt::ld5 �7 44
a3 tt::lb2 45 �2 h4 46 We2 g5 47
�b6 �e6 48 a4 �d6 49 �d2 1 ·0

Exercise 29: Black to move

The situation can easily be explained. Apicella-Svidler
White nee ds to both complete devel­ Yerevan 1 996
opment (a 1 -rook) and to generally im­
prove his pieces. But there are some
problems as 1 6 l:tad1 tbxd4! followed
by .. . .tx£3 is not right. The problem
here is, in fact, the knight on £3, which
is not very well placed, so...
1 6 ftJh4!
The only alternative, 1 6 i.e2, meets
with 1 6.. :i.g6 when White is not com­
pletely happy.
1 6 . . .tt::le5 1 7 .ie2l
Now White is able to bring the rook
to e 1 , the best square! It is interesting that none of my pu­
1 7 . . . ..txe2 1 8 %lxe2 tt::l7g6 pils was able to solve this at the first
18 . ..tbc4?! 19 l:ae l ! helps only White. attempt. Ivo Timmermans found the
The exchange on e3 is not really an op- first two moves the second time around

1 09
Exc elling a t Positional Chess

but then failed on the main idea. Still, it Exercise 30: White to move
fascinates me that pure logic will give Svidler-Leko
you the ftrst two moves of a three move Tilburg 1 997
forcing line! White has two active pieces
in the shape of the advanced knight and
bishop. Ivo decided that he had to evict
both of them...
1 4. .. f6!
Kramnik's new idea. Previously
14 ... 1i'b8 1 5 c3 aS 16 a3 a4 17 lDct e6
18 tLle3 lLlf6 1 9 tbd3 lDe 7 20 tLl£2 had
been played in Xie-Tisdall, USA 1 99S,
with an unclear game.
1 5 .th4 a6 1 6 �3
Stage 1 has now been completed. The
main reason why Black is not in trouble Of the three possible candidat�
(note that f4-f5! is threatened) is the moves in the diagram position, old the­
next move. ory considered only the weakest of the
1 6 . . . g5! three, 1 2 f4?!. There are two main con­
As the £3-bishop is pointing the other cepts in the position. For Black · it is to
way and the other will be doing likewise keep his strong knight on eS; he has
on g3, Black has no reason to fear a made considerable positional conces­
kingside attack. Therefore he can accept sions to do so. White, on the other
this slight structural weakening in return hand, wants to attack the kingside
for occupying the eS-square. pawns before Black can catch up in de­
1 7 .tg3?! velopment. Therefore White wants to
Svidler evaluates 17 fxgS fxgS 18 play h3-h4 and to deprive Black of
i.g3 lOdeS 1 9 .ihS : rn 20 We2 We7 natural development and the stronghold
21 c3 as unclear but I would say that on eS.
Black looks fme. 1 2 lt:lf3!
1 7 . . . gxf4 1 8 .txf4 li::lde5 1 9 i.h5 Also possible is 12 h4!?, but here
AfS Black has more resources as he does
Black is better. not have to make weird manoeuvres to
20 c3 ._.e7 21 'i'e2 lOg& 22 i.g3 keep e5. 1 2 ...llJfxg4! is the most logical
.!bce5 23 li::ld2 �hS 24 l:U2 l:ad8 25 - 1 3 hxgS l0xe3 1 4 Wxe3 llk4 1 5 ...g3
Abf1 .th6 26 l0g4?! lt:lxg4 27 .txg4 (1 5 1i'e2 i..g7 seems to b� just ftne for
d51 28 ..th5 dxe4 29 .!bxe4 f5 30 Black) 1 5 ... llg8 1 6 f4 hxgS 1 7 fS! sees
..txg6 hxg6 31 .!bd6 .taB! 32 ..te5+ White with numerous dangerous
�g8 33 .:td 1 l:d7 34 'i'd3 AfdS 35 threats, but Black has his domination of
'i'g3 'i'g5 36 "l'xg5 ..txg5 37 l:d3 the dark squares and a stronghold on eS
..ta4 38 l:h3 ltxd6 39 l:hS+ �7 40 to try to level it out. I think White is
J:th7 + �es 41 AhB+ �d7 0-1 better, but not much.

7 10
Solu tions to Ex ercis e s

1 2 . . . tt:lg6 tage. Now f2-f4 is possible.

Black has no useful moves. Re­ 1 6 tbd4 ..i.g7l?
routing the f6-knight takes a long time. Black gives up a pawn for some ac­
1 2 ... b5?! 1 3 ll:lxe5 dxe5 14 h4 l:tg8 1 5 tivity, which would not be present after
hxgS hxgS 16 W f3 i s very annoying for 16 ...h5 1 7 f3 ltJ4eS 1 8 �bl ll:lc4 19
Black. Full development is still far, far .te l etc.
away. 1 7 •xg4 i.xd4 1 8 gxh6 ltle5 1 9
1 2... ll:lxf.3 13 .txf3 lbd7 14 h4 llg8 ...e 2 lbc4 20 .Z:h31 ll:lxd2 2 1 'i'xd2
1 5 hxg5 hxgS 1 6 1i'd2! lbe5 17 .i.e2 f6 i.e5 22 f4 ..tf6 23 1fxd6 ...xd6 24
1 8 f4 gxf4 1 9 .txf4 followed by g4-g5 .Z:xd6 .Z:dg8 25 i.f3?
gives White an overwhelming advan­ Whoops. After 25 eS l:lxg2 26 exf6
tage. Black has been able to keep his :b7 27 lbe4 llg6 28 l:thd3 ..tc6 29 lbgs
knight on e5, but at the cost of every­ l:thxh6 30 lbxe6! '1tb8 31 lDd4 llxf6 32
thing else going wrong! ll:lxc6+ bxc6 33 llx£6 llxf6 34 l:td4
1 3 h41 ll:lxg4 1 4 hxg5 i.d7 White wins.
1 4...ll:lxe3 looks more logical but the 25 . . .e5 26 f5 l:.g 1 + 27 .Z:d 1 i.g5+
lead in development gives White an 28 <Ji>b 1 l:xd1 + 29 i.xd 1 l:.xh6 30
advantage after 1 5 'i'xe3 hxg5 (lS . .'ifcS
. l:xh6 ..txh6 3 1 i.h5 f6 32 i..f 7
1 6 Vd2 hxgS 17 ltxh8 ll:lxh8 1 8 eS dS Unfortunately for White this end­
19 lbe4! and Black is busted) 1 6 llxh8 game cannot be won.
ll:lxh8 17 'iVxgS and Black does not 32 . . .�d8! 33 i.d5 1i/c7 34 ll:le2 b6
really have control over the dark 35 c3?! 'li>d6 36 �c2? i.xf5 37 i.b7
squares in the centre. i.g6 38 ll:lg3 i.f4 39 ll:lf5+ i.xf6

Exercise 3 1 : Black to move

Bad Homburg 1 998

1 5 i.d21
White does not want to exchange this
bishop. The knight on g4 looks silly and
there is still pressure on the kingside.
1 5 . . . 0-0-0
15 ...hxg5 16 llxh8 lbxh8 17 lbxgS The centre is about to collapse yet it
gives White an overwhelming advan- is unclear what the implications will be.

1 11
Ex celling at Positio n a l Chess

Black has good prospects for most of Exercise 32: White to move
his pieces, with the d7-knight being Khalifman-Hubner
clearly his most inactive piece (if we do Munich 1 992
not count the rook). White has prob­
lems with his knight on a3 and, perhaps,
with the co-ordination. However, after
Black's next White is able to bring his
knight into play and thus gains an ad­
20 . . . .i.b7?!
20...cxd5! is correct. The sacrifice is
of the worst placed black piece and the
gain is total control of the centre. Then
21 �xd7 dxe4 22 lilli2 d5 23 ltlg4 is
good for White, as is 21 ...�c8 22 i.xc8
:Lxc8 23 l'Db5 dxe4 24 l'Dh2 d5 25 l'Dg4 This is a very complex endgame
'ii'a6 26 l'Dxf6+ Wxf6 27 icS!, but per­ where, initially, I believed there was
haps Black can play 21.....i.b7!? 22 l'Db5 only one logical way to maintain the
l:.d8 23 i.g4 d4 with counterplay. This initiative. But as it turns out there are
leaves 21...d4! 22 tbxd4!? (22 i.cl c3 23 two. I would probably still go for the
b5 i.b7 is hardly any better for White, first at the board, as it is a risk free op­
the pawns are very strong and will never tion.
be lost) 22... exd4 23 i.xd4 i.xd4 24 23 .l:g1 !
'i'xd4 i.cB! 25 .1xc8 Wxa3 26 i.g4 Introducing the rook into . the
'i'xb4 27 llcl with a more or less level proceedings. The important point here
game. is that if Black takes on eS the bishop
21 tLlxc4 cxd5 22 �xd6 should recapture to maintain pressure
Now the knight is fabulous. White on g7. Thus the text is quite logical in
has an advantage. that it activates the rook and exerts
22 . . . dxe4 23 tt:lh21? pressure on Black's main weakness. But
23 tt:lxb7 'ifxb7 24 1fxd7 1Vxd7 25 the creative 23 exf6 is also interesting.
1Lxd7 exf3 26 gx£3 was also pretty good Then 23.. .'ittf7 is the reason why most
for White. But Svidler must think that people reject this immediate capture,
his pieces are better co-ordinated and but after 24 fxg7!? 'ifi>xe7 25 h6 the
goes for the dynamic exploitation of his situation is far from clear. Here are two
advantage. possible continuations: 25 ... i.b3 26 llg1
23 . �c6 24 b5 i.d6 25 b6 .l:d8 26
.. i.g8 27 l:tg4 l:.c6 28 .:tf4 llxh6 29 1:£8
..ixd7 1:xd7 27 tt:lxe4 �h8?! 28 .tf7 30 gB'if i.xg8 31 .:txg8 .ILth2 32
tt:lxf6 gxf6 29 'i'h51 �g6 30 tt:lg4 llg7+ �d6 33 llxa7 and White has all
l:ld6 31 l:ld 1 ! f5 32 ..tc5 tLlf4 33 the chances, although a draw is likely, or
'il'xf6 ..ib3 34 Wxe5+ f6 35 'W'xf4 25. .1i>£7 26 l:tgl 'ittg8 27 l1g4 i.b3 28

1 -0 l:f4 i.f7 29 :f6! and White seems to.be

1 12
Solu tions to Ex ercis e s

much better. 20 . . .i.b8!

23 . . . 'it>f7 With the idea of ...'i'd6. Hence
23 ... fxe5 24 �f6 l:c7 25 ..txeS l:[d7 White's next.
26 �e2 gives White permanent pres- 21 g3!
sure. The pressure on dS protects the h3-
24 �d6 lte8 25 f4 .i.c2 26 �d2 pawn for the time being.
i.e4 27 'it>c31 f5 21 . . . i.a7!
27 ... fxeS 28 .1xe5 g6 29 hxg6+ hxg6
30 :dl and White has control of the
open flle s.
28 �c4 a6 29 h6! gxh6?
This is the decisive mistake. After
29 ...g6 30 h4 Black is under increasing
pressure but can at least offer some re­
30 j,.c7 lte6 31 j,.d81 b5+ 32 c;t;>xc5
ltc6+ 33 c;t;>b4 ltc2 34 b3 lbh2 35
.if& i.d5 36 .l:.g7 + �8 37 l:td7
i.e& 38 ltd6 'it>f7 39 :.xa6 l:tb2 40
lta7+ 'it>g6 41 �c5 ltxb3 42 ltg7+ This move should have been antici­
c;t;>h5 43 c;t;>d6 i.c4 44 e6 ltxa3 45 pated in order to have dealt successfully
e 7 :as 46 j,.c3 b4 4 7 .ixb4 'ifi>h4 with this exercise. After g2-g3 the
48 e4 fxe4 1 -0 bishop is no longer of any use on the
b8-h2 diagonal; the squares d4, e3 and
Exercise 33: Black to move f3 are more important now.
Gurevich-Khalifman 22 g4?
Biel 1 993 Creating a serious weakness. Better
was 22 h4, although Black remains bet­
22 . . . h6! 23 .ig2 d4! 24 exd4 ..tb3
25 l:td2 i.b8! 26 1i'f3 ..td5! 27
tLlxd5 .l:.xc1 28 lLlxf6+ gxf6 29 l:td1
lbd 1 30 1i'xd 1 ..te5 31 d5 J.xb2 32
J.d2 ..i.xa3 33 .txh6 'i'e 7 34 .tt3
.id6 35 .te3 We5 36 � 1 b6 37
1i'c2 b4 38 1i'a4 l:b8 39 i.d1 1i'xd5
40 i.b3 1i'b5+ 41 'i'xb5 l:txb5 42
'it>e2 De5 43 'it>d3 .tc5 44 i.d2
.txf2 45 .txb4 l:e3 + 46 Wc2 :txh3
This is an easy one. Black should re­ 47 .ic4 l:tg3 48 .ixa6 l%xg4 49 j,.d2
locate the bishop on e5 as it is the only l%d4 50 .id3 'ii>g7 51 ..tc3 l:d6 52
piece not doing anything sensible. .id2 .ih4 53 .ib4 lZd5 0-1

1 13
Ex c e lling a t Pos itio n a l Chess

Exercise 34: Black to move White still has attacking possibilities on

Nikolic-Khalifman the dark squares around _ Black' s_ king,
Ter Ape! 1 994 and the position is therefore by no
means clear.
23 •g4+ 'ii>h8 24 W'xb4 :d7 25 h4!
h6 % - %

Exercise 3 5 : White to move

Germany 1 995

Black has the more harmonious posi­

tion. He has an extra pawn for the
queen in addition to the rook and mon­
ster knight, and only one weakness - g7.
White, on the other hand, has several
weaknesses: eS, £2 and h2, along with _a ·

dodgy king position. The correct course

is therefore for Black to slowly improve Here Black has a wonderful situation
his position. It can be done best by in · the centre - a passed pawn, a well
driving the white queen away from the placed queen and pressure against the
attractive post on e4, and by doubling blockaded passed pawn. On the king­
rooks on the c-tile. But Khalifman had side he has serious problems, not only
another idea that, unfortunately for him, with his king safety, which is actually
is logically flawed. minor, but also with the h-pawn. These
20 . . . i.b4?! disadvantages will never fully dis�ppear,
Titis is based on a direct attack but the advantages in the centre will.
against the king. The only problem is 27 tzld2! -.xd4 28 lllxc4 lbc6 29
that White, by sacrificing the rook, ruins lbe3
the enemy king position and trades off White has a modest edge. This is al­
his not so effective pieces. 20 . b5!, with
. . ways an annoying situation to have to
the idea of ...%tc4 and .. .l:tdc8, is the defend, and in the game Black did not
most logical way to iffiprove the pieces. succeed.
Black would probably be a little bit bet­ 29 . . .W'a1 + 30 �h2 "i'e5+ 31 g3
ter. llld4 32 'it'c8+ �g7 33 "i'd7 illf 3+
21 .:xg7+ ! �xg7 22 .txb4 lbxb4 34 �g2 lllg 5 35 'ii'g4 h5 36 "i'f5
22... :tc1+ 23 <itte2 :tgl looks active, lbe6 37 h4 llld4 38 •d3 �g8 39
but the problem is that after 24 �d2! �f1 �f8 40 �g2 �g8 41 �f1 f5?!

1 14
Solu tions to Ex ercis e s

It transpires that this exchange· and, !might on g4.

_later, the exchange of queens, is no help 23 lt:lfh2!
for Black, as White now can bring his Also the least active piece in the at­
king into play without thinking about tack. The g5-square is for the bishop.
the random factors queens produce. Note thai: the text cle� the third rank
42 'ii'c4 f4?! 43 _fi'd51 'ii'xd5 44
· for the rook.
lt:lxd5 fxg3 45 fxg3! �g7 46 �f2 23 . . .it.xg4
�g6 47 �e3 �f5+ 48 �f4 f6 4� Alternative lines look - like this:
�f3! �g7 50 lt:lf4 �h6 5 1 lLle2 lLld6 23 ...'i'xh4? 24 :h3; 23... h5 24 J.g5 f6
52 �e3 �g6 53 lt:lf4+ �h6 54 25 !Dxf6+ .1xf6 26 e5 and White wins,
�d3! as is the case after 23 ... £5 24 ex£5 :Xel +
Two weaknesses. Now a7 is officially 25 'i'xel i.x£5 26 .ixh6 .ixh6 27
a target. - !Dxh6 �xh6 28 .ixfS gx£5? 29 'W'e6+
54 ltlf5 55 lLle2 �g6 56 �e4 !fJe7
... �h7 30 1i'xf5+ etc.
57 lLlf4+ �h6 58 a3 lLlcS 59 �d5 24 lLlxg4 'i'e7 25 l:txa8!
lLlb6+ 60 �c6 lLlc4 6 1 a4 lLld2 62 25 hS? :Xa3 26 bxa3 !D4xd5 gives
a5 lLle4 63 �b7 lLlxg3 64 <j;xa7 lt:lf5 Black counterplay through the c3-
65 a6 1 -0 square.
25 . . . :xa8 26 h5 lLlc4 27 e51
Exercise 36: White to move White has a considerable advantage
Khalifman-Gavrllov and went on to Win.
.St. Petersburg 1 994 27 ...dxe5 28 b3 lL!b6 29 hxg6+
fxg6 30 lLlxe5l i.xe5 31 -.,5 'Wg7
32 _Axe5 lLl4xd5 33 :e6l lg8 34
:xg6 1 -0

Exercise 37: Black to move

Van der Werf-Khalifman
Wijk aan Zee 1 995

This position is �ard to solve, al­

though I feel it should have been very
easy. - The white knight is very well
placed on g4, all the other · pieces help
with the attack and h4-h5 is annoying
for Black. The only problem is that the
offensive seems to end before it has
begun. The right idea is to keep the This is also a very difficult position,

1 15
Ex c elling a t Positio n a l Chess

apparently, but let me try to make it gerous, it merely shows that White is
simple. Black has lasting advantages in drifting. 1 8 �f6 :gs 1 9 lbge2 lbc6 is
the form of a superior pawn structure. also preferable for Black. but had to be
Quite simply, the eS-pawn is weak. played.
White, on the other hand, has a lead in 1 8 . . .�c61
development and some attacking The only idea for White is to play b2-
chances against the king, which is still b4 and win the e6-pawn, but this is
stuck in the centre. In many lines a rather easily prevented with the most
piece sacrifice on d5 opens the position natural Black moves.
and brings the knight on g3 into close 1 9 .te3?!
contact with the enemy king via e4. So This does not work. Now Black
what are we to do? Well, the threat of might play 1 9 ... d4!?, when the loose 20
getting mated should prompt us to b4 is the only move. But the game is
trade queens. also clear enough. After 1 9 l:b 1 lbb4!
1 6 . . .1i'c5! the fight for that idea is over.
There are no acceptable alternatives. 1 9 . . . .te7 20 .txc5 .txc5+ 2 1 �h1
1 6.)Dc6 17 lbxdS! exd5 18 'ifxdS <J;e7 22 :taf1 Ahf8 23 Axf8 :xta 24
seems too dangerous, while 16 .lbf5 1 7
.. ltxf8 <Jrxf8 25 lbce2 lbxe5 26 �f4
ftlxf5 gxf5 also fails to solve Black's �e 7 27 �ge2 �c4 28 b3 �e3 29
defensive problems (18 'lh4.! then c3 �6 30 h4 e5 31 �d3 .td6 32
looks like the strongest). After <J;g 1 e4 33 tildf4 .txf4 34 lbxf4

16 . 'ifxe5 17 'i'xeS tt)xeS 1 8 i.£6 White

. . �e5 35 g3 ti)f5 36 ti)e2 d4 0- 1
wins the exchange and 1 6 . lbxe5 17
. .

.lael ! brings the inactive piece into play Exercise 38: White to move
- 1 7 lb7c6 18 'W'h4 with problems for
... Khalifman-Filippov
Black (18 lLlxdS?! exdS 19 'ifxdS i.e?! Kazan 1995
1 7 -.xc5 �xc5

White needs to develop fully before

starting an assault on the black king. His
1 8 ltf6? two remaining moves are castling
This is the test but, as it is not dan- queenside and the development of the

1 16
Solu tions to Ex ercis e s

knight. This�should, of course, be done 23 ...e3! was better.

with the consideration o f Black's only 23 'ifxf4+ 24 l:xf4 :as 25 l:ldf1

idea - to put something on f4. Hence :ads 26 h4 lbb8 21 'iPc2 tDd7 2s

White's next. �c3 tLlfS 29 hS h6 30 a4 ltJh7 3 1
1 0 ttlh3! b4 a 6 3 2 c 5 lbg5 3 3 .tc4! .txc4 34
Planning 0-0-0 with advantage. After '*xc4 g&?: 35 hxg6 �g7 36 l:f7+ !
the immediate 10 0-0-0 ...£4!, as in �xg6 3 7 l:xb7 lbxe4 38 l:f4 ttld2+
B eckhuis-Haba, Pardubice 1 994, Black 39 �d3 �g5 40 Ilf2 1 -0
- will have to face a lot less pressure. In­
teresting, but not the strongest, is 10 Exercise 39: Black to move
lb£3 tbf4 1 1 0-0!? tbxe2+ 12 'iil>h1 with Luther-Khalifman
compen sation accordiflg - to Seq,er. Ac­ Hastings 1996
tually White does have some initiative
for the piece here.
1 0 . . .f6! ?
I willignore the theoretical discussion
here, as the exercise is related to finding
the right plan and nothing more. All I
will say now is that the only game where
Black had a good result was the follow­
ing: 1 0... c5 1 1 .1.xc5 W'xg2 1 2 0-0-0
tbd7 1 3 tbgS 'Wc6 1 4 .i.a3 f6 1 5 �hS
fxgS 1 6 ...xgS <ii£7 1 7 'i!fe7+ �gB 1 8
...e8+ tiJdf8 1 9 i.xg6 hxg6 20 ...xf8+
c.ith7 21 ..f4 lte8 22 lthg1 eS 23 1i'h4+ Black seems to be worse. The knight
�gB 24 :d6 'i'c7 2� ltdxg6 i.fS 2� on f5 is strong and £7 is a potentially
1:6g5 (26 l:.h6l) 26 ...W'f7 27 b3 (27 serious weakness. Trading knights on f5
:hS!) 27... llac8 28 �5? (28 :hS ..e6 also seems dangerous for Black as his
29 ...gS llxc4+ 30 bxc4 ...xc4+ 31 �dl bishop is clearly the inferior one, al­
still wins, but now it is going the other though this is not completely clear if the
way) 28 ...:Xc4+ 29 bxc4? (29 �b2 is pawn structure alters just slighdy. White
the lesser evil) 29 .. .''i'xc4+ 30 �d2 also has some problems - the e4-pawn
:ds+ 31 'it>et 'Wc3+ 32 'iil>fl i.h3+ 33 is likely to become a weakness later -
1:1g2 'i'al+ 34 �e2 ..d l+ 35 �e3 Wd3 and the queen is only temporarily well
mat�, Jacob-Prang, Germany 2000. placed on c4. But here we notice that it
Also very dangerous for Black is will go to b3, unless Black is ready to
1 O.. .'i!fxg2 1 1 0-0-0 i.d7 1 2 tbgS etc. advance his a-pawn immediately. So
1 1 0-0-0 Wf7 1 2 f3 •h4 1 3 .tc5! again we have a situation where White's
e5 14 ttlf2! f5 1 5 g3 1i'f6 1 6 l:lhf1 1 advantages are mainly shan-term while
l;)a& 1 7 .ta3 .te6 1 S lt'le4! fxe4 1 9 Black's chie f advantage i� long-term.
fxe4 �f4 20 gxf4 - l:lhdS 2 1 .i.d6 1bis leads us to the solutiQD. to Black's
�gS 22 fxe5 'ifh4 23 1i'f4?! problems without altering the issue

Excelling a t Positio n a l C h e s s

concerning the e4-pawn. 28 lbe3 was better.

23 . . .'i'b51 28 . . .l:a5 29 .i.e1 tbf6 30 :l.d3 l:tca8
The exchange of queens creates a 31 a3 l:t5a6 32 �a2 bxa3 33 b4
situation where two pawns will keep a J:lc6 34 l%c3?! :lacS 35 l:txc6 l:txc6
good eye on three and where the e4- 36 c3 J:lc4 37 �xa3 lbxe4 38 tt:Jxe4
pawn will remain slightly weak. How­ .:.xe4 39 JLf2 f5 40 gxf5 gxf5 41
ever, White should be able to hold the l:d 1 f4 42 .:.d5 l:te2 43 .i.a7 f3 44
balance. l:txb5 �f71 45 �b3 .i.d6 46 l:td5
24 ...xb5 axb5 �e6 47 l:td 1 e4 48 .:.11 .i.g3 0-1

Exercise 40: White to move

Russia 1996

25 J.a7
The alternatives are interesting. 25 gS
hxgS 26 .ixgS f6 27 .id2 lbd6! leaves
Black with the superior structure. The
exchange of knights does nothing about People have a way of not solving this
the weakness of the e4-pawn but does exercise despite the fact that it is ex­
limit White's chances of attacking g7. tremely logical and clear. Very often the
Black is better after both 25 lidS :XdS solutions to these exercises tend to be
26 exd5 lbd6! and 25 lld3 lbf6 26 ltlg3 in some way dependant upon the ftnal
l:l.dc8, intending .. .:c4. Finally 25 :xd8 evaluation of a resulting position, which
l:txd8 26 'ificl ltl£6 27 l!Jg3 produces can be somewhat complicated - espe­
equality. cially when you have to give reasons for
With the text White did not realise in your intuitive decisions.
time that he had to exchange a pair of But let us tum to the position. .f:low
rooks to avoid the incoming threats on should we approach the situation? Re­
the a-file. member - each position has a primary
25 :as 26 J.t2?!
••. concept, something that you want to
After this White is in trouble. Neces­ achieve. 1bis is the aim of calculation
sary was 26 .l:txd8 llxd8 27 .ie3 lbf6 28 and problem solving in over-the-board
lbg3 with equality. chess. First you define your agenda,
26 . . .l:tdc8! 27 .:.d5 g6 28 tbg3?! then you find a way to pursue it. Here

1 18
Solu tions to Ex ercis e s

the main thing is to complete develop­ on the f-file (and this will spread to the
ment. My advice is - when you have kingside).
something you want, first try to see if 1 8 �d5!
you can do it directly! In something like 18 tlJ£5 'WgS 19 cxd4 'iWxfS! 20 .ih4
90% of cases (not in studies and prob­ t!Jh3+! 21 gxh3 l:.g8+ 22 <i>h1 "ifx£3+ 23
lems, but in tournament chess, 9 out of llx£3 is somewhat better for White, but
10 moves) you will have the option to this is not what one is aiming for. The
do it straight away. This is such a case. text is a real killer - and one of those
White wants to castle but also wants 10% moves that we find slightly surpris­
to recapture with the queen on f.3, as ing and not straightforward - and a
g2-g3 is required to later restrain Black's clever way to solve the immediate prob­
knight. Therefore the previously played lem of eliminating the knight on f4.
move was 1 6 i.c2, before Khalifman Note 1 8 exd4 tbxd3.
discovered the obvious... 1 8 . . .�xd5
16 0-01 You never fully escape calculation,
Playing the move you want to play - and you should use it when necessary.
first make it work. Here you need to see things like
1 6 . . . gxf3 18 ... i.xd5 19 'ii'xf4 tbcS 20 exdS lLlxd3
The move everybody expects here, 21 9xd4 tbeS 22 .ih4, which is win­
but they do not try to make the moves ning.
they want to make work! White is al­ 1 9 exd5 �5
ready strategically winning. 1 6......g5 1 7 19 ...0-0-0 20 cxd4 and Black is posi­
fxg4 h S 1 8 lDdS! ...xg4 19 'i!ixg4 hxg4 tionally lost. The last chance, 20 ... b5,
20 tbxf4 exf4 21 i.e1 also gives White a serves only to illustrate how hopeless
convincing position according to the situation is. 19 ... dxc3 produces a
Khalifman. more complicated solution. Most of this
1 7 'W'xf3! could probably have been decided upon
The point. The bishop is not hanging with an intuitive decision and need not
as the natural 1 8 lbfS! is practically deci­ have been calculated - 20 .i£5! tbes
sive. This is really not difficult to see (20...0-0-0 21 .l:.ae l ! "ii'gS 22 .id4 :hg8
when you have the idea. And how do 23 .tf6) 21 'ii'h3! WgS (21 ...tbxc4 22
you get the idea? You ask yourself - :ae1 ttleS 23 J.h4) 22 .ih4 •d2 23
· what moves do I want to make without :ae1 1fd4+ 24 'ifr>h1 l:tg8 25 :le4 1i'd2
looking at the immediate tactical as­ 26 l:lxeS+! dxeS 27 J.d7+ �£8 28
pects? With this knowledge you will be :xf7+! '1Pxf7 29 'i'fS+ 'iPg7 30 .i£6+
more atuned to seeing such little tactical and mate (all lines given by Khalifman).
twists. 20 'W'e4 dxc3?1
1 7 . . exd4
. This does not meet with Black's ideal
1 7 ...tbxd3 1 8 tb£5 'ifgS 1 9 i.h4 agenda - completing development by
'W£4?! 20 lLlg7+ �£8 21 tbe6+! fxe6 22 bringing the king into safety. Yemelin is
'i'xd3, or 17 ....:g8 18 .ic2! 0-0-0 19 a true fighter more than. anything else,
i.g3 and Black is under serious pressure but fighters are often helpless when it

1 19
Exc elling a t Positio n a l C h e s s

comes to very accurate play. Khalifman and this factor is an important part of
gives the following lines: the solution. White is breaking up the
20 ... f5 21 'ifxfS! l:t£8 22 �5+ 1i't7 centre, so Black will need to do some­
23 'ii'xh6 winning, 20 ...ttlxd3 21 ..xd3 thing about the fact that the c4-pawn
0-0-0 22 �xd4 :he8 23 .i.f6 ..e4 24 will soon be in trouble.
jt'xe4 llxe4 25 i.xd8 'itxd8 26 :Xt7 22 .:fe81

l:.xc4 27 l:.d1 with excellent chances Tactics are used to solve a purely
and 20...ttlg6 21 W'xe7-+l ttlxe7 22 l:.ae1 static problem. Here these are based
�fB 23 �h4 l:.e8 24 cxd4 with an around a tactical trick with ...lle 1+, thus
overwhelming positional plus. prompting White to trade queens and in
21 .ih4 ..d7 22 .i.f6 turn 'develop' Black's queen's rook.
The struggle is over. You need not 23 'i'xb6?
see further than this to be certain. White is on his way to very serious
22 . . .l:lg8 23 l:lae1 c6 trouble. Correct is 23 exd5! (23 eS?
23.....g4 24 llf4 'it'g6 25 ..e2 and ..i.x£3 and Black is better thanks to 24
wins. After 23. ..C�f8 24 �4 'ii'g4 25 exf6 lel+Q 23 ... .i.x£3 24 'iixb6 (forced)
'it'xh6+ �e8 26 llf2 Black has not 24... axb6 (24... i.xd1 is too adventurous
solved a fraction of his problems, only since after 25 'i'd4 l:.el+ 26 �h2 ttlg4+
lost the h-pawn and seen the white rook 27 'i'xg4 �xg4 28 l:xe1 Black loses) 25
enter the attack. gx£3 b5 and Black is probably a little
24 .lixe5 dxe5 25 'ii'xe5+ �8 26 better already. His pieces are perform­
.if5 •ds 27 .i.e6 l:.g7 28 .txf7! 1 -o ing better and his pawns are better pro­
White wins in a million ways. One of tected.
them is 28.. .lht7 29 �8 mate. 23 ...axb6 24 exd5

Exercise 4 1 : Black to move

Groningen 1 997

24 . . . b5!
With this move order Black has been
freed from the illogical exchange of the
bishop for a misplaced and pinned
Yet another situation in which a knight. Black has a very pleasant game.
player has yet to complete development, 25 l:.c3 l:.ad8 26 lte3 l:lxe3 27 fxe3

1 20
Solu tions to Ex ercis e s

.l:.a8! 28 e4 l:txa3 29 e5 lbd7 30 e6 but what about the rooks? The d-file, of
fxe6 31 dxe6 ..txe6 32 lbd4 .i.g4 33 course, which means that the queen is
lbxb5 l:tb3 34 l:tc1 �xb4 35 i.a2 in the way. All this leads to the exploita­
lbb6 36 lbc3 .te6 37 lb1 :xb1 + tion of cS.
38 .i.xb 1 lbd5 39 lbxd5 .i.xd5 40 1 5 i.a7! :as
�2 �f7 41 �e3 .i.xg2 42 'it>d4 ..tn The line 1 5 ...l1b7 1 6 'iie3 i.d6 1 7
43 .i.e4 'iPf6 44 .i.f3 �5 45 ..td1 tl'ld2!?, intending ll.Jb3 when the a-pawn
�g5 46 ..tf3 h5 0-1 is hanging, is what Black is defending
against by putting the rook on a8. How­
Exercise 42: White to move ever, after 1 7 ... 'ii'a8 1 8 i.cS <l;fl here I
Short-Smvslov think Black would have a better fighting
Subotica 1 987 chance than in the game, although
White is much better.
1 6 'i'e3 'i'b7 1 7 i.c5 lb8 1 8 b3
i.xc5 1 9 'ifxc5 'i'b6
1 9... a4 20 l:td2 and White is better.
Incidentally Smyslov is famous for us­
ing an exchange of queens in poor posi­
tions as a defensive tool.
20 'i'xb8 cxb6 21 .l:.d6
The endgame is much better for
21 . .. lbe7
21...'1ii>e7 22 l:hd1 ! l:thd8 23 :Xc6
Another logical situation, although :Z.xdl+ 24 li::Dcdt! and Black does not
this time prophylaxis plays a part. It have any compensation (see /1Jd1-e3-
does not take long to see that Black will f5).
attack down the b-file, forcing White to 22 l:.hd1 rJlf7 23 lbe1 .tea 24 lbg2
nudge the b-pawn forward one 'square i.g4 25 : 1 d3 llhc8 26 lbe3 .i.h3
and thus weaken the dark squares in 27 lbe2 :tc7 28 f4 exf4 29 lbxf4
front of the king. Here it is natural to .it.g4 30 lbxg4 hxg4 31 �d2?1 :as
believe that .i.e3-cl will at one time be 32 lbe6 l:tcc8 33 c4 lbg6 34 lbd8+
necessary because Black's dark-squared �8 35 lbxc6 l:txe4 36 h5! lbh8 37
bishop cannot be allowed a free hand lbd8! l:lb8 38 lbe6+ �g8 39 l:td8+
on the queenside. Furthermore, White l:txd8 40 l:.xd8+ 'it>h7 41 l:te81 .l:te5
needs to think about the organisation of 42 lbf8+ �h6 43 l:txe5 fxe5 44
his forces. The pawn advance f2-f4 is a lbd7 lbf7 45 lbxb8 lbd6 48 h-3
logical plan, but a move like 15 tl'lh2?! is lbf5+ 47 '1Pf2 g6 48 hxg8 <hg8 49
so ugly that we should avoid it on lbd5 lbd6 50 lbc3 rJlf6 51 We2 �e6
purely intuitive grounds. The knights 52 �d3 lbf5 53 llJe2 �6 54 a 3
will not have any good squares before �c6 55 '1Pe4 lbd6+ 56 '1Pxe5 a4 5 7
f2-f4 and an opening of the position, lbd4+ 'it>d7 58 �d6 lbf5 59 . b4

Exce/Ung a t Positio n a l Chess

lllxg3 60 tbc6 t0e2 6 1 tOeS+ <l;c7 difficult to develop.' White -is clearly
62 t0xg4 t0c3+ 63 <i;d4 t0b 1 64 better.
t0e3 t0xa3 65 �c3 1 -0 1 4 . . .'i'h4+ 1 5 g3 -.e7? 1 6 0-0 h5
1 7 J:ae1 0-0-0 18 t0c1 ! 'i'b7 1 9
Exercise 43: White to move tbd3 h4 20 g4 h 3 2 1 a4 •c7 22 b3
Short-Adianto J..e 7 23 g5 <i;b7 24 •t2 e5 25 fxe5
Jakarta (Game 5) 1996 J..x g5 26 bxc4 dxc4 27 lbf4 J..h4
28 llb1 + �aS 29 •e3 'W'd7-.30 "c3
.tg5 3 1 e6 'i'cB 32 exf7 llh4 33
'i'e6 1 -0

Exercise 44: White to move

Novgorod 1 996

This exercise concerns captw:es and

recaptures. Black is threatening to win
the eS-pawn, which White would nor­
mally fight to avoid. But there are other
issues here. If Black takes on eS there
will be numerous exchanges, after
which Black's dark-squared bishop has
no obvious square. White's, on the This is a prototypical example of a
other hand, will be more than happy on situation where looking at the pieces
d4. And this is how it worked out for individually will lead to a general con­
Short. clusion. First we start by stating the ob�
1 0 t0fd4! .ixe2 1 1 -.xe2 t0dxe5 1 2 vious. The weaknesses in the position
f4! are the pawns on g6 and h7 (which
Remember that the key is to establish White's bishop will constantly target)
the bishop on d4, from where it helps and those on d4 and £2. There is no
White control the board - primary con­ reason for White to alter the pawn
cept. structw:e immediately because after 24
1 2 ...tl:!c4 1 3 tbxc6 bxc6 1 4 .td4 exfS l:txe l 25 :.Xel itJxd4! the �t�ck on
Short: 'Utut had overlooked/mis­ the dark squares proves strong. So let us
assessed this continuation. My litd.e look for the ideal squares. All of Black's
Gennan friend, Fri� does not under­ pieces are go�d and are ready for action,
stand it either. The white bishop domi­ which can also be said of White. The
nates the board, making it fiendishly bishop could not be better placed than

Solu tions to Ex ercis e s

on c2, from where the currently impor­ 'ii'a6 l:ta7 4 1 'ii'c4 liJf7 42 'ii'c3 liJe5
tant e4-square can _ be _ monitored. 43 'ii'b2! - l:te7 44 Ae1 'it>g7?1 45
White's queen sits well on c4, adding lLld41 h5! 46 lLlf3 Af7 47 �e4 lLlxf3
weight to the d4-d5 push and, thanks to 48 'ii'xf6 + �xf6 49 �xf3 Ae7 50
the central location, not far from any Ab1 .:S7 5 1 l:tb2 !l.z -%
action on either flank: (b7 could be a
target, as could Black's king). White's Exercise 45: Black to move
knight is under a little pressure, standing Short-Sokolov
on the f-file and having to offer support Groningen 1 997
to the d4-pawn.
Finally White's rooks are quite happy
occupying the centre fl.les. It would ap­
pear that this is a fruitless exercise as
nothing can be improved, but never
forget the king! There are two points to
be made regarding the monarch - it
stands on a dark square and it can, in
fact, have an effect on the short line
given above.
24 'it>f1 ! fxe4 25 �xe4 lLld8 26 b4
Black's situation is critical. Neverthe­
less, as we know, most situations are
critical in a hard fought game, or at least
there are five or so very complex situa­

tions in every game. I remember Yer­

molinsky writing that for 90% of his
moves he was making a clear judgement
based on analysis of the position while

for the remaining 10% he was basically

just guessing.
The reason why these exercises are
more difficult than combinations is be­
27 l:te3! cause you can never be sure about your
White has a ·modest advantage. From answer. You have to develop a strong
here on the position changes and the intuition and a method of analysis of
pieces find new and better squares. more positional aspects. In this position
27 J:fe8? 28 l:tde1 Wh6 29 a4 c6!?
.•. it is crucial to evaluate correctly the dif­
30 d5 b5 31 axb5? cxb5 32 'ii'a2 ferent positions that arise after three
'ikg77 33 Wxa6 �c3 34 l:tc1 �xb4 moves.
35 •xb5 �c5 36 l:.e2 _.f6 37 �d3 1 7 �h4?

l:txe2 38 �xe2 l:.e4 39 �d3 l:te7 40 Clearly not the right choice. Black

Ex c e !Ung a t PosWo n a l Ch ess

will now have to play ...d6-d5 and the 1 7 ...cxb2!? 18 exf6 1i'xf6 has been
bishop looks stupid on h4. It is also suggested by some students, but I feel
hard to find a decent square for Black's that this is far from good. White can
knight, which is slightly passive on c6 simply play 1 9 .i.d3 and it is difficult to
because only after e4-e5 does it have imagine a situation where the pawn will
prospects of coming back into play. seriously threaten to promote. The key
1 7...lbxe51 1 8 fxeS �xeS is the right is that Black does not have anything else
choice for positional/intuitive reasons. going in the position, and White has
But what about the material evaluation? i.e3-d2-c3 coming, so something needs
Well, material is normally reflected in to happen soon or White will be doing
the abilities of the remaining pieces (the fine. Also 1 9 :Xd6 Wg6+ 20 'i'g3 Wxc2
key reason for positional sacrifices - if a 21 li)d2 ltfd8 22 i.d3 'ii'c3 23 i.b6 is
piece does not play it does not count). not so easy to get a hold on. Black is
Here White would be in a great shape if going to be a rook down, but what are
after 19 bxc3 he could follow up with the consequences? So the bottom line is
li)d4 and i.f3, leaving Black with only that I feel that this line is inferior for
two pawns for the piece. But Black can Black after 1 9 i.d3 (from a principal
prevent this by developing a piece with point of view) but that this should be
tempo with 1 9...l:c8!, when Short had tested in a five game match with Fri�
apparently overlooked that after 20 Nonetheless, this is better than ...i.h4.
lbd4 Black has 20...:Xc3! with a strong 1 8 it'xc3 d5
position. Therefore White will have to I do not like this move. 1 8...li)e7 1 9
play 20 c4, after which Black has 20... £5! llxd6 Wc8 i s better, offering Black
with the idea of mobilising the only re­ some compensation. Here we are in the
maining inactive piece - the rook on £8. traditional case of forcing play, where
Note that the queen is more or less ac­ the forcing line seems to give White
tive as its ideal square is h4 and it can good prospects, and Black should there­
go there in one move. If a piece is ca­ fore try to steer the game into more
pable of making threats like this it will murky waters. It would be silly to vol­
be able to enter the attack with great untarily set off on a track that might
strength as the opponent cannot pro­ lead to a forced loss in the hope that the
ceed with his own active ideas. opponent will not be able to see it. I
Slightly passive is 17...�e7 1 8 Wxc3 tried to explain this to a junior earlier
(18 exd6?! chooses too soon for no rea­ this year, but he was so convinced
son and favours Black after 18... hd6 19 about his own abilities that he somehow
'l'xc3 lbe7!) 18...'1'b8 19 exd6 �xd6 20 forgot that I could play, that what he
�5, which is playable but White has a saw I would see too.
serious (though not enormous) edge in 1 9 tlJc5 _Wc7 20 f5!
the shape of the bishops, particularly White has a serious attack now. Black
since with this structure it is easy to see has nothing.
passed pawns on both sides in the end­ 20 d4?
. . .

game. -And now he just loses in one move.

Solu tions to Ex ercis e s

20 ... 1&.c8! was necessary. Take a look at position is that his knights have no ideal
the original position and we can see that squares. The bS-knight can find two
Black's prospects have seriously wors­ squares in f5 and dS, but which one is
ened since then, with both bishops hav­ best? The solution is based on tactics.
ing declined in worth. White has some Note that d6 is not really an ideal square
advantage. as White can never fully gain control
2 1 J:xd4 ltlxd4 22 -.xd4 i.d5 over a dark square that is so far into the
Losing, although Short points out enemy camp.
that after 22 ... 1&.e7 23 tLlxb7 ...xb7 24 1 4 ltld6?!
f6! Afd8 25 'Wg4 �f8 26 fxg7 1&.e7 27 Leading to trouble. The thing about
'i"hS Black loses too much material. this move is that it changes none of the
23 -.xh4 -.xe5 24 _.d4 -.xd4 25 concepts listed above and therefore
i.xd4 e5 26 .te3 a5 27 J:d 1 1 -0 does little good.
1 4 c!DaS 1&.h6+ 1 5 �b1 tz'lcS looks
Exercise 46: White to move good for Black. The knights on aS and
Seeman-Short bS contribute little to the position in a
Tallinn 1 998 positive way, and tactical stuff like 1 6 b4
lba4 appears to be more problematic
for White than for Black.
The right option must be to eliminate
Black's advantage of the bishop pair
and improve the control over the light
squares with 14 tz'lc7!, when 1 4. :. .th6+
1 5 �b 1 l:r.ad8 16 lZ'ldS+ .ixdS 17 l:lxdS
tz'lcS! 1 8 �c4 .:XdS 19 .ixdS :ds 20 c4
tz'lxb3 saw Black only very slightly
worse in the endgame (which he man­
aged to draw) in Popov-Tseshkovsky,
Russia 1 997.
This position is rather dangerous for 1 4 . . .b6 1 5 i.b5
White. Let us try to address the con­ White is short of 'easy' moves. 1 5
cepts that are immediately apparent. �d3? i s a n example o f the Esben Lund
Black has the two bishops. Black has expression: When I think I am clever I am
the open g-fde. Black has a pawn major­ realfy sftpid. I know that, with regards to
ity in the centre. White has a lead in deciding on 15 i&.d3, we can follow my
development. Did I miss anything? Per­ claim that 90% of the moves you really
haps the weakness of the fS-square, but want to play can simply be played. But
can it be exploited? here, unfortunately, we are dealing with
And now to weaknesses. Black might the remaining 10%, since after
be a little bit vulnerable on the light t S...'ithd6 16 �e4+ �c7 1 7 .ixa8
squares, but g2 is also a potential target. .ih6+ Black wins.
The main problem for White in this 1 5 -tc4 .ih6+ 16 �bl l:hd8 is good

1 25
Ex c elling a t Positio n a l Ch e s s

for Black, I think. After .i.xe6 fxe6 Black can develop normally and hope
White's knight sits pretty on d6 with that White, after a trade on e5, will try
nowhere to go; perhaps the bishop will to cling on to the e5-pawn with passive
eventually tell. moves. But if White instead tries to
1 5 . . . l:ld8 1 6 J:ld3 i.h6+ 1 7 �b1 quickly organise a counterattack against
tt:lb8!? dS he will be able to hold the balance.
Fritz suggests an even stronger con­ So instead Black should look at the
tinuation in 17 ... a6 18 .ic4 e4!! (Short: pawn structure and put the question to
'which was way too difficult for a patzer White as to the efficacy of his bishop.
like me') 1 9 l:.d4 exf3 20 gx£3 l!JeS 21 As a student pointed out, this works
.ixe6 fxe6 and Black wins a pawn, and very well with Silman's idea of imbal­
most likely the game. In the game White ances.
could have put up more resistance but 1 1 . ....i.xe5!
still had a difficult task ahead of him. From here on the knights will domi­
1 8 l:lhd1 nate. The bishop cannot find a good
1 8 tLle4 f5! and after ...lftf6 Black's square. The alternatives give White a
position is favourable, with the two chance to hold the balance.
bishops and the open g-file. 1 1 ...ll:lec6!? 12 l!Jc3! ll:lxeS 13 dxe5
1 8 ... a6 1 9 .i.c4 :txd6 20 .l:xd6 i.xeS 14 l:tfel! lllc6 (14... .ixf4 1 5
i.xc4 21 J:lxb6 ..i.a31 22 l:lb4?! :Xe8+ 'i'xe8 1 6 'ifxf4 c 6 1 7 'i'c7 W'd7
22 l:tb7+ is preferable. 1 8 .:e1 f6 1 9 l:te7 is more than just
22 . . .J:lc8 23 tba5 i.b5 24 a4 i.c5 compensation ...) 15 �xeS :xeS
25 :th4 i.d7 26 J:ld3 i.d4 27 J:lb3 (1 S ...lbxe5 16 'ifd4 'i'gs 17 �fl :tad8
tt:lc6 28 tt:lxc6+ i.xc6 29 c3 ..i.d5 1 8 lZlxdS c6 looks good for Black but
30 :ta3 i.f2! 31 J:lh5 J:lg8 32 c4 White seems to be holding everything
i.xc4 33 g3 i.d4 34 J:lxh7 l:tb8 0-1 together with tactics after 1 9 h4!!, with
the key idea of 19 ...'i'h5 20 llle7+! and
Exercise 47 : Black to move White is doing well) 1 6 :XeS l!JxeS 1 7
Gurevich-Shon 'ii'd4 with equality.
Manila 1990 1 1 ...lllbc6 12 tz.lc3!
· i.xeS
(12 ... lllxd4?! 13 lllx f71 looks good for
White) 1 3 �xeS tLlxeS (13...tLlg6 14
.i.g3 is interesting but, compared with
the game, the knight seems less appro­
priate on c6) 14 dxe5 lllg6 1 5 llfe 1 ! and
White keeps the balance as in the lines
1 2 i.xe5 tt:lg6
Now Black has a slight but serious
1 3 .ig3
13 ii'g3 is not good. The bishop is

Solu tions to Ex ercis e s

still exposed on eS. Black can use the Here White, rather unwisely, has
momentum with 1 3.)Zk6! 1 4 .ixc7 weakened his kingside by advancing
'it'd7 when White is in trouble. A possi­ with g2-g4 and h2-h3. This has not only
ble line would be 15 :d 1 :ac8 1 6 .id6 compromised the kingside structure but
l:te6 1 7 .1La3 :le4 and White is still also slowed down his development. Not
s truggling with development. Note that surprisingly Black is able to exploit this
White does not have rime for 1 3 llel in factor through rapid development and
view of 1 3. .. f6t, winning a piece. pressure again st £'2. Thus Black should
1 3 . . .lbd7! try to generate an initiative.
Coming to the ideal square on e4 8 . . . lba6!
with gain of time. Gaining rime and, after the forced
14 'Od2 lLlf6 1 5 'iWf3 c6 1 6 'it'b3 reply...
'Wb6 1 7 'ii'xb6 axb6 · 9 a3
Black has a better endgame. The ... Black can gain even further time
bishop simply cannot find a good way with
to join the game. 9 . . .dxc4 1 0 i.xc4 b5!
1 8 a3 lbe41 1 9 lbxe4 Axe4 20 Afd 1 Again time is crucial. When you at­
b5 2 1 'otf1 f6 22 f3 l:.e6 23 Ae1 tack it is always good to send your
<JJf7 24 Axe6 �xeS 25 .:te1 + �d7 pieces into battle while simultaneously
26 �e27! h51 27 �d3 h4 28 i.h2 pushing your opponent back.
lbe7 29 .tf4 itJf5 30 .td2 b6 3 1 1 1 .i.e2
l:te2 c5 3 2 .le3 b41 3 3 axb4 c4+
· 1 1 .ia2 b4 1 2 lLla4 �h8 clearly fa­
34 �c3 lbd6 35 lte1 lta4 36 �d2 vours Black according to Short.
Axb4 37 .:ta1 :xb2- + 38 .:a7+ 1 1 . .. b4 1 2 lba4 bxa3 1 3 bxa3 l2Je4
'ote6 39 Axg7 b5 40 .tf2 b4 41 �c 1 Short prefers Black bu t I am not sure
c3 42 .lxh4 itJf5 0-1 that this is the whole story. White is
uncoordinated, has a ruined pawn struc­
Exercise 48: Black to move ture and is seriously behind in develop­
Gelfand-Short ment. In practise it was not hard for
Tilburg 1 990 Short to finish off Gelfand.
1 4 lbf3 'iWa5 + 1 5 1itrf1 tbc7! 1 6 tbb2
.ta6 1 7 tbc4
17 �xa6! 'it'xa6+ 18 'ittg2 was the
best defence according to Short.
1 7 . . .'ii'd5 1 8 ltg 1 fxg4 1 9 Axg4
1 9 ... l!Jxf2! 20 �xf2 .ixc4 21 e4
1t'xd4+ 22 .ie3 'it'xal 23 'ii'xc4 'ii'xa3
wins direcdy.
20 .:txg3 i.xg3 21 �g2! .th4 22 e4
'it'h5 23 lbxh4
23 l!Jce5! .ixe2 24 'ifxe2 was better.

Ex c e lling a t Po sitio n a l Chess

23 . . ....xh4 24 .tel l:.f6 25 l:.h 1 dark squares on the kingside White also
l:.af8 26 l:th2? l:.xf2+ 27 .i.xf2 toys with the idea of c3-c4 in the event
'ii'xf2+ 28 �h 1 •e1 + 0-1 of Black castling queenside.
1 3 .:9a

Exercise 49: White to move Overprotection. White was threaten­

Short-Kamsky ing 1 4 lbc4! here.
Tilburg 1 990 1 4 tbb3

Once again it is Short who is alone in 1 4 . . ...ixd3?!

fmding the right move in a_ difficult po­ Giving White a clear and lasting
sition. When we talk about squares here advantage. Correct is 1 4 ...'1i'e7 15 c4
we are particularly concerned with the ..txe3 16 fxe3 0-0-0 17 cxdS exdS 1 8 aS
dark squares on the kingside, where a6 19 lbdcS when White certainly has
Black has adopted an aggressive stance. far more dangerous threats against
He is trying to push further with . g4-g3
.. Black's king, with sacrifices in the air as
to prise open White's defensive wall and well as the rook manoeuvre .:tal -a4-b4
damage the dark squares. Furthermore 1116 .txd3 a5?1 1 6 f4! gxf3 1 7 l:.xf3

the h6-knight will soon come to f5 to b6 1 8 i.xg5 'trxg5 1 9 'i6'xg5 l:lxg5

exert more pressure on the dark 20 l:.h3 �e 7 21 l:lxh4 tDg4 22 l:.e 1
squares. l:tag8 23 g3 c5 24 tDd2 c4 25 .i.c2
White has two fairly poor pieces - his f5! 26 exf6+ .!Llgxf6 27 tbf3 l:.h5 28
d2-knight and the queen. Consequently l:.xh5 �h5 29 �2 �d6 30 tDe5
these two unemployed workers should ll:ldf6 31 .td1 tbg7 32 g4! l:.b8 33
join in the struggle for those dark i.f3 b5 34 axb5 l:.xb5 35 l:te2 a4
squares, for if White wins this fight he 36 h4 a3 37 bxa3 l:lb3 38 l:tc2
will have a clear advantage due to the l:txa3 39 h5 tbh7 40 �g3 �e7 41
long-term concessions Black has made l:lb2 l:tb3 42 l:la2 .:tb7 43 'it>f4 ll:le8
in his search for an early initiative on 44 g5 tbd6 45 g6 tbf6 46 h6 �f8
the kingside. Hence White's next. 47 l:.a8+ �e8 48 �g4 �g4 49
1 3 •c1 ! ! .txg4 l:te7 50 �e5 1 -0
A s well as preparing to monitor the

1 28
Solu tions to Ex ercis e s

Exercise 50: White to move Exercise 51 : White to move

Short-Gelfand Gelfand-Short
Brussels 1 991 Brussels 1 991

White is not helped by the awkward White has no choice regarding which
placing of his knight, but it is not diffi­ side to castle. His h2-bishop is mis­
cult to find a superior square £3. Thus
- placed for a kingside offensive and,
the solution to this exercise is rather generally, attacking on the kingside after
straightforward. queenside castling is not very good for
1 5 'ii'h 1 J White in the Classical Exchange lines in
White should not be tempted to play the QGD. But Black is threatening to
1 5 f5 as after 15 . . gxf5 Black has im­
. solve all his problems with ... i.d6, after
proved his pawn structure and the situa­ which a level game would arise. There­
tion is far from clear. However, White fore White is practically forced into re­
has a good position after 1 6 lLlf4 ..e7 acting correctly.
1 7 lL!hS, but making a mess of things in 1 2 tt:le5!
this way seems to be quite unnecessary. 12 0-0? .td6 with equality.
1 5 . . . a5 1 6 a4 'i'd6 1 2 . . . i.d6 1 3 f4!
But now if Black plays 16 ...�6 the White has an absolute advantage. The
queen has left the kingside and it is time bishop on h2 looks funny but this is not
for 1 7 f5! gxf5 1 8 t'Llf4 with much more permanent because the kingside pawns
dangerous threats than before. will keep on rolling (f4-f5 is guaran­
1 7 ti:lg 1 i.d7 1 8 ti:lf3 teed!). And Black is really struggling to
White has a slight advantage and get something out of his minor pieces.
went on to win the game: White's advantage is best illus�te.d by
1 8 tt:lb4 1 9 -.121 1i'c5 20 i.c3!
•.. considering future plans.
tt:lc6 21 l:tae1 b6 22 i.d2 tt:lb4?! 23 1 3 . . .c5 1 4 0-0 c4 1 5 �e2 i.b4
'it'g3 b5 24 f51 exf5 25 tt:le5 i.e8 1 5...h5!? 16 lL!xg6 fxg6 1 7 'i'xg6
26 axb5 'ti'xb5 27 :lxf5 �h8 28 l:.xe3 1 8 i.xhS lL!xhS 19 'i'xhS .ie6
:lxf6! :lxf6 29 tt:lg4 :lf5 30 ti:lh6 with advantage for White has been sug­
l:th5 31 9f4 1 -0 gested by Short as a means of genet'at-

Ex c elling a t Po sitional Ch e s s

ing counterplay. But the text is hope­ comes 18 i.c6 followed by ... g7-g5 and

less. .)t�e4, also with equality. Therefore

16 f5 tl:!f8 1 7 ..tf3 i.xc3 1 8 bxc3 there is only one good move.
.id7 1 9 g4 1 8 .tc 1 !
The strength of White's concept has White is not trying to prove anything
been illustrated and the rest is just an­ in particular but simply posts his bishop
other game of chess. on the most natural square, b2, retaining
1 9 .tc6 20 Wg2 lDad7 21 g5
•.. some advantage due to the bishop pair.
tl:!xe5 22 ..txe5 l0e4 23 ..txe4 dxe4 As Black also is not fully developed the
24 h4 �h8! 25 f6 g6 26 h5 J:tg8 27 loss of time is by no means critical.
�f2 Wa5! 28 .l:lh1 Wxc3 29 J:tae1 1 8 . . .i.c6
.id7 30 hxg6 lbg6 31 Wxe4 �gB
32 Wxb7 Wc2+ 33 J:te2 Wf5+ 34
J.f4 l:tcB 35 Wf3 c3 36 e4 Wa5 37
d5 c2 38 e5 l:tc3 39 Wh5??
39 'i'e4 wins.
39 . . . h6 40 e6 .taB 41 ..tc 1 Wxd5
42 e 7 J:txg5! 43 i.xg5 Wf5+ 44
�e1 c 1 W + 45 ..txc1 l:lxc1 + 46
�d2 J:tc2+ 47 �d1 Wd3+ 48 <i>e1
Wg3+ 49 ci>d1 Wd3+ 50 �e1 Wc3+
51 ..t>"f2 Wd4+ o-1

Exercise 52: White to move 1 9 i.b2

lvanchuk-Short The pressure on the long diagonal
N ovgorod 1994 will force Black either to weaken his
kingside with ... f7-f6 or make him feel
quite uncomfortable. All endgames also
give White good wincing chances.
1 9 . ..Wd6 20 J:te 1 l:le8 21 JbeB+
�xeS 22 f3 a6 23 a4 �f6 24 ..td3
24...i.d5, with advantage to White, is
better. Now White gets a winning posi­
25 .ixa6! bxa6 26 'i'xc6 Wd2 27 h3
W'e3+ 28 'it>f1 Wd2 29 ..txf6 gxf6
30 'i'xc7 h5 3 1 h4 Wd 1 + 32 �2
White needs to make a choice. If he Wd2+ 33 �g3 We1 + 34 <i>h3 Wh1 +
plays 1 8 i.xf6 he has no advantage at all 35 Wh2 Wd1 36 Wg3+ �h7 37 Wf4
as the resulting weak pawns cannot be �g7 38 We4 a5 39 c4?? Wxb3 40
attacked. In the event of 1 8 �h4 there c5 Wc3! 41 Wd5 'ii'a 1 42 Wc4 'W'b1

1 30
Solu tions to Ex ercis e s

43 �gJ 'ii'e1 + 44 �h2 1lfe5+ 45 Now White is doing well.

�h3 Wf5+ 46 �gJ 'ii'e5+ 47 �2 20 . . .i.b5 21 c4 b6 22 l:ta1 .tc6 23
1lfb2+ 48 �e3 'ii'xg2 49 c6 Wg 1 + l:txa6 g5 24 .tg3 ll:lh5!
50 �e4 'ii'e1 + 5 1 �d4 'i'e5+ 52 24. ..lod7 25 lta7 is uncomfortable.
�d3 Wf5 + 53 �d2 'ifxf3 54 c7 25 lt:'lxe5
Wf2+ 55 �d3 1lfg3+ 56 �e4 'ifg4+ 25 .i.xeS? g4 and the king is exposed.
57 �d3 Wg3+ 58 �c2 1lff2+ 59 25 . . . .tb7 26 l:ta7 lt:'lxg3 27 l:xb7
'it>b3 'ifb6+ 60 �a2 •f2+ 61 �a3 1lfxe5 28 fxg3 Wxg3 29 11Ve3 We5
We3+ 62 �b2 ft6+ � -% 30 c5!? b5?
30 ...g4 3 1 h4 bS would have kept
Exercise 53: White to move Black in the game. Now it is critical for
Short-Speelman him.
London 1 991 31 :t1 :d7 32 l:txb5 c6 33 :be
l:g6 34 b4 �g7 35 Wt3 h5 36
'W'xh5 •d4+ 37 �h1 g4 38 l:tb8
l:d8 39 l:b7 'W'f6!? 40 :xt7+ 'ifxf7
41 :xt7 + �xf7 42 hxg4 l:e8 43
Wh7+ �f6 44 11Vd7 l:te6 45 b5 �e5
46 b6 l:g8 47 c3 1 -0

Exercise 54: White to move

Debrecen 1 992

The weakest point in Black's camp is

the e5-pawn. The worst placed white
piece is the rook on b 1 . The ideal
square for this piece is on aS. Therefore
the correct move is logical.
1 8 l:ta 1 !
Playing with all the pieces - always
remember to do that!
1 8 . . .J:tg8
Better was 1 8. . .:ta8!? with the same

line as in the game, only now Black How should White, to move, com­
does not lose the a6-pawn White would
. plete his development?
probably play 1 9 c4 with some advan­ Here White could develop normally
tage. Black has parried White's idea, but with 1 5 0-0 but this does not lead to any
White has improved his rook and Black particular advantage. The strongest way
has worsened his. to develop an initiative is by bringing
1 9 dxe5 dxe5 20 :as the Icing's rook into play via the 3rd

Ex c e lling a t Positio n a l Chess

rank. As �e l-fl will work as well as Black won tells us something about his
castling, White has no problems with greatness.
the king. How should one find this 1 7 . . .l:td7 1 8 f4 l:tad8 1 9 �d2 l:te7
plan? Well, very simple. White is slightly 20 b4 r!be6 21 l:th4 e5 22 l:teh 1 h6
better placed and is reasonably active, 23 bxc5 bxc5 24 �xh6+ gxh6 25
while Black is passive. Therefore White l:txh6 r!bf8 26 f5 f6 27 :xf6! c4 28
should attempt to prove an advantage, .1e2 c3+ 29 �e2 :g7 30 g6 .:te7
and this is done by finding the most 31 g4 i.c6 32 g5 :ba 33 'M3 :teeS
serious weakness in the enemy camp. In 34 .tb3?
this case it is the kingside. After... 34 :xf8+! llxf8 35 f6 l:b7 36 .i.f5
1 5 h41 �d7 37 g7 and wins.
...White is trying to establish four 34 . . .l:txb31 35 axb3 i.b5 36 e4?! e2
pieces against two on the kingside. This 37 .le1 .lc3+ 38 �4 dxe4 39 :f7
is the most dangerous approach. 1 5 e3 40 l:tb7 a6 41 :tb8?! ..td3! 42 d5
.ia6 is a tricky line, when Black must .le5 43 �e5 l:tb5! 44 l:td8?
make a choice: 44 l:xbS! axb5 45 'itr>d4 .i.xf5 46
a) 15 ... .ixa6 16 .:.Xc6 lbeS? (16 ...h5 <it>xe3 lDd7 47 �d2 lLlcS - 48 ltxc2
17 'i'f4! gS? 18 ...f3 wins for White - l0xb3+ 49 <it'c3 .i.xc2 50 �xc2 lLld4+
Short - but 17 ... .!DeS! transposes to 5 1 �c3 lbf5 52 �b4 lbd6 53 �c5
1 5...h5) 17 dxeS 'i'xeS leads to an un­ draws - Short.
clear game according to Shon. I believe 44...l:txb31 45 l:th1 l:tb6! 46 .:tea e2
White has nothing to fear here, as after 47 f6 .1xg6 48 .:tc7 e H W + 49 l:txe1
1 8 l0h6+! rwitf8 1 9 'i'b4+ �e8! (19 ...'i'e7 l:tb1 50 J:teJ c H i 51 l:txc1 J:txc 1 52
20 'ii'xe7+ �xe7 21 lD£5+ �f8 22 l0d4 l:ta3 l:tg 1 53 .:txa6 llxg5+ 54 'OtJd4
gives White a clear advantage) 20 Wa4 �7 0-1
<it>f8 21 'i'a3+! �e8 22 'i'xa6 gxh6 23
0-0 White holds the advantage. Exercise 55: White to move
b) 1 5 ...h5! was found by a clever stu­ Short-Karpov
dent, Jan Hondenbrink. This leads to a Linares 1 992
draw after 16 'i'f4 .i.xa6! 17 :.Xc6 lDc5
1 8 dxc5 'i'xb2, when White has nothing
better than 1 9 lDe7+ �f8 20 lDg6+ �g8
21 lLle7+ with a perpetual.
1 5 . . ..!Df8 1 6 'l'g5!
The key move. 16 l::th3 �c8 17 :f3
.i.xf5 allows Black to eliminate the pri­
mary attacker and thereby equalise.
1 6 .. .'i'xg5 1 7 hxg5
This position is structurally better for
White. The knight on f5 is great and
Black has weaknesses on h7 and c6, as
well as some other weak squares. That Here we are concerned with identify-

1 32
Solu tions to Ex ercis e s

ing weaknesses. White has only one Exercise 56: White to move
weakness - the b2-pawn - while Black Short-Kramnik
suffers from having some rather weak Novgorod 1995
pawns and one weak square: a6, c6 and
dS. The defender of all these squares is
the knight on b4 which, consequently,
keeps Black's position intact. Therefore
Short came up with a brilliant idea.
1 8 ltJa21
Simply eliminating the prime de­
fender. After this Black is in trouble.
The knight on e4 is not as well placed as
the one on b4, despite its appearance.
1 8 . . . c5
Karpov tries an active defence, but
Black cannot save this position even This i s a very difficult exercise. It is
with perfect play. also very instructive because it tells us
1 9 dxc5 ltJxc5 20 ltJxb4 l:.xb4 21 something about winning positions, or
jLc6 at least gives me the chance to say
something about winning positions.
The most important rule in winning
positions (here I am talking 11bout stati­
cally winning positions, where the struc-
.. ture or the material makes it winning) is
that if nothing happens, then you win.
This is simple logic. So what should you
do? Prevent all counterplqy- if possible.
Another rule is that the presence of
opposite coloured bishops makes it
easier to win once the position is
. genuinely winning, as Bent Larsen
21 . . .'it'b6 explained. In our case this means that
Black also loses after 21. .. .i.e4 22 White already has the ideal situation. All
lbd4 (with the plan 23 £3) 22 .'i'b8 23
.. he has to do is prevent counterplay, aad
i.xdS .i.xdS 24 lbf5 'ilfxeS 25 ltxdS! Black has counterplay only connected
etc. to the advance of the f-pawn, so the
22 i.xd5 :Xb2 23 .,c4 llc2 24 'Wg4 right move would have been 26 f4!,
.,c7 25 ltJd4 D.c3 26 ltJc6 lle8 27 fixing the pawn and taking the eS­
jLd4 llc2 28 ltJb4! lld8 29 ltJxc2 square away from Black's bishop.
i.xc2 30 e6 i.f8 3 1 exf7+ �h8 32 Instead Short played carelessly.
J:te1 i.g6 33 J:te8 J:txe8 34 fxeB'it' 26 J:tfd 1 ?!
i.xe8 35 i.xc5 i.xc5 36 'ii'a6 1 -0 26 LaS is also an acceptable move,

1 33
Ex c elling a t Positional Chess

as after 26... f4 27 gxf4 :Xf4 28 l:ta8+ sonable chances of a successful defence.

White willprobably win anyway. Never­ 54 Wf5 i.c7?? 55 Wg5+ 1 ·0
theless it is inexact. And in a winning
position it is precision that is of the Exercise 57: White to move
highest importance as this prevents Yusupov-Short
situations occurring in which you have Novgorod 1 995
to play good moves to earn your full
point - something that might fail, as
millions of players have experienced
over the years...
Instead, 26 f4! 'l'c7 27 ltfdl ltf6 28
•hs with a clear win is the line Short
prefers, and he is right in doing so.
Black has no counterplay and the oppo­
site coloured bishops are not a drawing
factor, as Black's is not playing. More­
over this feature accentuates White's
advantage by increasing the threats to
the black king. Yusupov, quite understandably, had
26 . . .i.e5 27 ltxa5 ltb7 28 :tad5?! some problems finding a plan in this
'l'f6 29 f4 exf3 30 Wxf3 ltg7 3 1 position. Indeed it is by no means easy.
<t>h 1 Wh6 However, with the aid of logic one
White is probably still winning, but should be able to find the right move,
there are different kinds of winning po­ though it does not lead to an over­

sitions. There are those you will win whelming advantage. It is clear that
with optimal play, those you will win White should try to exploit the weak­
most of the time and there are positions ness of the fS-square in some way as it
where you will always win. This exam­ is not easy to prevent Black's main plan
ple does not fall in the latter category, of ...tbb6, ...i.d7 and ...tba4 followed
which it would have done after 26 f4!, by ...a7-a6 and ... b7-b5, opening the
of course. For practical results, this is a queenside. If Black takes on g3 then
very important lesson. after 1 6 fxg3 White has good control
32 ltg1 l1g6 33 :d3 ltg5 34 aS l:.h5 down the f-ftle and good control over
35 'fi'f2 f4 36 g4 lth3 37 ltf3 ltxf3 the strongholds e4 and £5.
38 Wxf3 i.d4 39 ltd1 .ie3 40 a6 The other idea is to open the centre
1ff6 41 Wd5 f3 42 Wxd6 'itg7 43 in some way that utilises the two bish­
i.d5?! i.a71 44 'itb-4?! f2 45 .i.g2 ops. The move for this is f2-f4, but this
:as 46 :n 'ite5 47 i..c 6! lte7 48 does not work if Black can just take on
1fd2 �g7 49 b4 lU7 50 b5 i.b6 51 g3, as the extra power on the dark
'l'd5 'l'f4 52 'i'e4 Wd6 53 .idS squares then fades away. This, appar­
:te7? ently, did not help Yusupov find the
53...l:f4 would have given Black rea- appropriate plan, although he thought

1 34
Solu tions to Ex ercis e s

about pushing the f-pawn. The thing is .l:c6 l!Jd2 55 J:la6 �b1 ! 56 �5
that after 15 f3 Black does not necessar­ �xc3 57 l:ta7+ �e6 58 �c7+ �d6
ily take on g3 because the bishop would 59 l!Je8 + �c6 60 J:ta6+ 'it>b7 0- 1
not be a major threat on £2. Here pat­
tern recognition and knowledge of these Exercise 58: White to move
kinds of Nimzo/Queen's Indian posi­ Timman-Hi.ibner
tions comes in handy. It is often seen Tilburg 1988
that White chooses to recapture with
the f-pawn on g3 in order to use the f­
ft.le. True, then the rook is often already
on fl , but still it is an idea worth re­
Therefore the right move is 1 5 h3!,
intending 1 5 . ..lbb6 16 .i.h2! and l:tfdl,
with g2-g4 and f2-f4 on the way, when
White has achieved something on the
dark squares.
1 5 .l:de1 ?!
With the idea of f2-f4 but, appar­
ently, Yusupov was unable to make this This exercise is about squares and
work. those moves that we know are inevita­
1 5 . . .lbdf6 1 6 f3?! ble. Sometimes we can .grow much
1 6 f4 ttlxg3 17 hxg3 .td7 with the wiser by just pointing out which devel­
idea of ...llae8 is slightly better for opments we know are bound to come.
Black according to Short. 16 h3! was Here . . .l2Jxb4 is coming, to which llxb4
still the best move. Now Black has _ is the most comfortable answer, but
equalised. then there is a problem with .. .'ii'xa3.
1 6 .. .�xg3 1 7 hxg3 i.d7 1 8 :h2 h5 Then after i.xfS exfS, lDd4 the fS-pawn
19 �b1 g4 20 'itla1 J:lhg8 2r f4 h4!? comes under fire and, in turn, g7 might
22 ltxh4 exf4 2 3 gxf4 �xd5 24 be a weakness, not forgetting that we
.l:lhh 1 �b6 25 e4 rl.h8 26 .l:hf1 J:lh2 will be left with knight against bishop.
27 lLlb3 rl.ah8 28 e5 i.c6 29 i.e4 .This means that White will have the
�xc4 30 i.xc6 bxc6 31 exd6+ advantage on the dark squares while
ii'xd6 32 ._.f5? ji'd5 33 .:te7+ �b6 Black will have more influence on the
34 ii'xd5 cxd5 35 :txf7 l:.h 1 36 ltb1 light squares. Consequently White's
.l:lxb1 + 37 'itlxb1 J:lh 1 + 38 'itlc2 l:h2 queen will be best placed on the dark
39 Af6+ �c7 40 t0xc5 l:txg2+ 41 squares, so the ideal square - which has
�d 1 g3? 42 .Ug6 .Ud2+ 43 �e1 g2 yet to be found - is f4.
44 lLle6 + ! �d7 45 �4 .:txa2 46 1 4 -..c 1 !
lile2 a5 47 J:lxg2 a4 48 .Ug7+ �d6 This move is in tune with all these
49 J:lg6+ �e7 50 l:ta6 a3 51 �d4? considerations and, in my opinion,
l:b2 52 �c6+ �f7 53 lLlb4 d4 54 clearly the strongest of all the possible

1 35
Ex celling a t Positio n a l C h e s s

continuations. However, the others Of course we have learned from Fn"tz to

should still be investigated: take such pawns, and we have also im­
a) 1 4 0-0 l0xb4 1 5 :Xb4 (1 5 axb4 is proved our defensive skills, but this still
probably better, but this is not the way looks like it is too much for Black to
we want to play) 1 S . .'i'xa3 1 6 .tbS+
. survive. The weakness of the king does
�fB and losing the right to castle does not even seem to be a dynamic, but
not seem to justify the sacrifice of two rather a static advantage.
pawns. 1 4. . . b6
b) 14 .txf5 l0xb4! (14 ...exf5 1 5 0-0
l0xb4 1 6 l:txb4 'i'xa3 17 'i'd4 is good
for White) 1 5 axb4 (1 5 :Xb4 ...xa3 1 6
l:tg4 exf5 1 7 l:txg7 � fB 1 8 l:tg5 'i'b4+
looks good for Black; White needs to
castle) 1 5 ... exf5 1 6 'l'xdS 'ffxc2 1 7 0-0
0-0 18 :1fd1 'l'e4 and it is hard to imag­
ine that Wlllte should have more than
just enough compensation for the pawn
c) 1 4 g4? falls for the tactical trick
1 4...a5! when, although the position
remains unclear after 1 5 ic5 'i'xg4 1 6 1 5 c41
l:lg1 'iff4, i t does not give the impres­ Another advantage of 14 'it'cl , which
sion that White has in any way bene­ did not have to be foreseen to make the
fited from these developments. Actually decision, so I did not comment on it.
I prefer Black. 1 5 .. .t'Llxb4
d) 1 4 h3, to prepare g2-g4, seems 1 5 ...dxc4 1 6 ...xc4 .ib7 1 7 .ixfS exf5
slow and I am not even sure that push­ 1 8 e6 is terrible for Black.
ing the g-pawn is in Wlllte's interest. 1 6 :txb4 'ifc6 1 7 0-0 0-0
After 1 4 ...l0xb4 15 axb4 id7 16 g4 1 7 ....ia6 1 8 .ixfS exf5 19 cxdS 'iWxdS
l0e7 followed by ...�g6 Black holds the 20 ltd 1 is a very uncomfortable position
balance - at least. for Black.
e) 1 4 ...e2!? is a move suggested by 1 8 'iff4! .tb 7 1 9 .txf5 exf5 20 ltJd4
Junio,. 7. I do not like it as much as 1 4 'ifc5?
'i'cl because it does nothing o n the 20...1i'c7! 21 cxd5 .ixdS 22 �xfS
dark squares. But it does make uncom­ ie6 23 �d6 would have left Wlllte
fortable threats ro the black 9ueen and with only a minor edge, now it is all
offers some chances of an advantage over.
after 1 4 ...a6 1 5 0-0 lbxb4 1 6 axb4 fol­ 21 lLlxf5 .tea 22 �xg7! �xg7 23
lowed by b4-b5. Here 1 4...�fd4 1 5 'i'g5+ �h8 24 'i'f6+ c;t>g8 25 'ifg5+
�d4 �xd4 1 6 'i'g4 lLlxc2+ 1 7 .ixc2 �h8 26 'i'f6+ �g8 27 l:lb3 :tea 28
'ifxc2 1 8 0-0 'i'g6 19 'i'h4 f6 20 f4! llg3+ �f8 29 llg7 :te7 30 e6!
looks extremely dangerous for Black. i.xe6 31 llxh 7 1 -0

1 36
Solu tions to Ex ercis e s

Exercise 59: White to move 20 ... h5 21 h3 a5 22 a3 l:td7 23 l:tf3

Short-Kasparov 'ii'dB 24 l:tb1 J..g5 25 l:tbf1 axb4 26
Amsterdam 1996 axb4 J..e7 27 l:tb1 ?7
Now that White cannot further im­
prove his position there should be an

indication that direct action is appropri­

ate. W'hoo ps. White could have won
with 27 gxhS ttlxhs 28 .:x£7! :Xf7 29
:X£7 �x£7 30 'i'xg6+ �ffi 31 'i'h6+
liJg7 32 .ig6! and mate will follow.
27 h4 2B _.e3?! �B 29 J..e4 fiJc7

30 .i.c6 lbd5 31 'ili'e4??

White is still winning after 31 'ilr£2
:c7 32 .ixbS 'i'c8 33 ltb3.
31 . . . rtJxc31 ! 32 J..xc3 ft6+ 33 �g2
White has a clear static advantage. . l:tc7 34 i.aB .l:!.c4 35 'ii'b7 'ii'xb7 36
His isolated eS-pawn is a strength rather i.xb7 l:tbB 37 .ta6 l:tb6 3B l:ta1
than a weakness as it controls key l:tcc6 39 J..xb5 lbb5 40 .:taB+ <it>g7
squares in the heart of Black's camp, 41 :a7 �B 42 .:taB+ �g7 43 :S7
denying access to the defender's pieces. �B � - �
Additionally White has more territory
under his control as well as the advan­ Exercise 60: Black to move
tage of the two bishops. Therefore the Topalov-Short
plan should be to prevent counterplay, Amsterdam 1 996
improve the position to its maximum
and then the winning combination
should come about by itself.
So the first part of the plan is to pre­
vent counterplay. The only way for
Black to gain any is to eliminate the
bishop on d4, hence White ' s next.
1 9 b41
This is better than 19 'l'e3 because the
queen is already fantastic on e4 and there
is no need to be tied down to monitoring
cS. One should also calculate 1 9.. .ltc8 20
b4 i.xb4. Not so many choices here, but it is
1 9 . . .�7 20 g4! stillvery important to be exact.
Not to be allowed is ttlg7-f5xd4 etc. This is a simple position where Black
Now Black's counterplay has been has only two moves that do not lose on

nipped in the bud, leading us to the im­ the spot. It is a matter of simple calcula­
provement phase. tion.

Ex c elling a t Positio n a l Chess

20 . . . .td7! Another case of prophylaxis. The

Now Black will have the better game. core of the position is whether or not
But after 20 ... lle1+? 21 'iiff2 'it'e8 22 Black will be allowed to play ... c7-c5.
'it'xe8+ l:txe8 23 i.b6! he cannot Timman comes up with a cunning idea
develop his queenside, with a dreadful to prevent this.
defensive task to look forward to. 1 3 .:e1 1
21 h3 Aiming at the potentially vulnerable
Topalov sees that there is nothing. 21 piece on e7. The standard 1 3 lLlc3 lLld8
.l:td3 h6! 22 i.e6 'i'c7 and 21 'ii'd1 'it'c8! leads to no advantage, as can be seen
are fine for Black, while 21 'i'f7 WaS! from the following: 1 4 b4 (14 e4 cS 15
sees the mate on e 1 come into play. .if4 'iia7 1 6 dxcS 'i'xcS!? is unclear
These things are important in tactics according to Short) 14 ... a5 does not
(21....:te1+ 22 �£2 and White wins). promise White an advantage:
Then after 22 c3 .1c6 23 .i£8 'i'h6+ a) 1 5 bxaS ftxaS 16 lltb1 (16 lLldS?
Black wins. .ixdS 17 i.xaS 'ii'a8 18 .ib4 cS and
21 . . . .:es 22 .:xe5 fxe5 23 i.f7 .tea Black wins) 1 6 ...1Wa8 17 'ilfd1 c6? (an
24 .tg6 h6 25 .txe8 'irxe8 26 improvement is 1 7....ic6!, with advan­
'i'xe8+ AxeS 27 .td6 .:da 28 .t.xe5 tage to Black) 1 8 a4 and White was bet­
l:.d5 29 .t.c3 Axf5 30 .td4 b5 31 c3 ter in Kobalia-Kiriakov, Dubai 2002.
h5 32 .t.f2 �g8 33 �1 �7 34 b) 1 5 'ilfb2 axb4 1 6 axb4 ltle4 17
�e2 �e6 35 i.d4 g6 36 .t.b6 �d5 tLleS lLlxc3 1 8 'iixc3 .ixg2 19 'i1txg2
37 .tc7 q.,e4 38 .td8 J:.d5 39 .t.h4 i.d6 20 ll:ld3 'ii'b7+ with equality in
J:.d3 40 .tf2 a5 41 .te1 a4 42 .t.d2 Chetverik-Ponomariov, Briansk 1995.
.:d7 43 .t.h6 Ad5 44 .tc1 .:tc5 45 c) 15 .:.ab1 axb4 16 axb4 ltlc6 1 7
.te3 a3! 46 .tc1 b4! 0-1 lLla2 (17 lLlxbS i.a6 1 8 lLlh4 lLldS! 1 9
The further . .. b4-b3! is coming. i.xdS exdS 20 'ifxc6 i.xbS 2 1 'i'xdS
i.xh4 22 gxh4 .ixf1 with advantage to
Exercise 61 : White to move Black - Short) 1 7 . ..l:ta6 1 8 lLlct 'ifa8 1 9
Timman-Hjartarson 'ilfdt ! lla3 20 ltle1 eS! 2 1 lbc2 (21
Amsterdam 1989 i.xc6? i.xc6 22 dxeS lbe4 favours
Black) 21....:ta4 22 dS (22 i.xc6? i.xc6
23 dxeS i.f.3 24 'iie 1 lLlg4 and Black is
better) 22. ..lLld8 23 e4 c6 24 dxc6 .ixc6
with equality in Karpov-Short, Tilburg
1 988.
1 3 . . . a5
Now Black cannot play for the break
with the c:.pawn because opening the
centre after 13 ...lLld8 1 4 e4 cS 1 5 dS!
illustrates the problems of having
packed all the pieces together on the
queenside. After 1 5 .. . exd5 16 exdS 'i!fc7

1 38
Solu tions to Ex ercis e s

(16 ...i.d6? 1 7 i.c3! tLlxdS 1 8 tLlgS g6 1 9

. s ary was 26... tLlf4 27 tLle2! when, to
lte8+ i.fB 2 0 i.eS 'i'a7 21 lLlxh7 gives avoid transposition, Black must play
White a dangerous attack) 17 tLlc3 b4 27 ... tLlxe2+ 28 'i'xe2 tLld7! with an edge
1 8 i. f4 'i'd7 19 lLleS 'i'e8 20 tLla4! it for White.
appears that White has clearly the better
chances due to his freer play.
1 4 ll:\c3 ll:\d8 1 5 e4 b4 1 6 ll:\a4 ll:\d7
1 7 i.f4 i.d6 18 i.e3 ..tc6 1 9 d5 b3
20 Wc4 Wb5 21 Wxb5 i.xb5 22
ll:\c3 i.d3 23 ll:\d4! ll:\c5 24 l:tad1
l:tab8 25 dxe6 fxe6 26 i.h3 i.c2?!
27 l:td2 ..txe4 28 ll:\db5?
A tactical mistake. After 28 lLlxe6!
.!Lldxe6 29 .!Llxe4 l:r.d8 30 tLlxd6 l:r.xd6 31
l:xd6 cxd6 32 l1d1 White simply wins.
28 . . .i.c21 29 i.f1 ll:\f7 30 ll:\d4
% -% 27 llle 21 ll:\t4
After 30 ... tLle5 31 f4 tLlg4 32 lt)xc2 It is difficult to suggest other moves,
bxc2 33 .l:!xc2 .!Z.lxe3 34 l1xe3 :tb6 the but this leads by force to a clear advan­
position is level. tage to White, so Black should probably
step down from the fight for the f4-
Exercise 62: White to move square.
Timman-Portisch 28 ll:\xf4 Wxf4 29 g3 Wt3
Antwerp 1 989 29 . 'ifg5 30 �g2 does not
. . improve
Black's life. After 29 ...'i'f6 30 i.c3! the
f-pawn can no longer can be held back,
although Black can try something like
30... We6 3 1 f4 .xh3!? 32 fxeS ...xg3+
33 tLlg2 cS!? and the position is rather
messy, despite the extra piece. Also
possible here is 30 �g2!? followed by
30 ll:\g4! Wxd1 31 ll:\f6+ �h8 32
l:tfxd 1 :led8 33 .i.a2?l
White should have gone for 33 J.xeS!
dxeS 34 tLld7 .1J..g7 35 l:td3 with a clear
White would have a clear advantage if advantage. It is hard for Black to get his
he could drive the knight away from e5 minor pieces to work.
with f2-f4. White's forces are almost all 33 . ....tg7 34 . i.xe5 dxe5 35 llld 7
ln place for this, so White played... l:tc7 36 lLlc5? ! l:td4 37 ll:lb3 l:lxd1 +
26 :tt1 l Wg5?! 38 J:txd 1 i.f6 39 ll:\c5 �g7 40 i.e6
This proves to be less good. Neces- i.e7 41 l:td7 J:txd7 42 .ixd7 �f8 43

1 39
Ex c e lling a t P o s i t io n a l Chess

�d3 ..td6 44 f4 f6 45 f5 g5! 46 h4

'i;;e7 47 .i.e6 ..tb7 48 �f2 'it>d8 49
�f3 ri;e7 50 <lr>e3 <lr>d8 51 �e2 �e7
52 <lr>f3 �f8 53 �g4 <lr>g7 54 ..td7
11.e7 55 �f3 'i!ff8 56 �e3 .ltdB 57
�c5 i.b6 58 'iti>d3 i.xc5 59 bxc5
'it>e 7 60 .i.e6 �d8 61 �e3 �e 7 62
�d2 Yz - Yz

Exercise 63: White to move

Portisch-Tim man
Antwerp 1 989
22 . . .'ii'xa2
22 ... g5? 23 'ile4! followed by .td3 re­
sults in a strong attack for White.
23 'ir'c4! 'ifb2
23 ...'i'xc4 24 �xc4 l:ad8 25 tbe1 .a.d7
26 tbd3 is a perfect illustration of
compensation. Though Black is ahead
on points he still suffers badly from
missing files for his rooks and a missing
diagonal for h.is bishop. \Xlhite can
consider a plan such as g2-g4-g5 to bury
the bishop completely, and then_ take the
This position is not easy to evaluate, rook around to c6. What Black can do is
bur for White there is simply no way to a more difficult proposition.
avoid sacrificing the exchange, as all 24 ..ttJ .!:tabS 25 i.c6
other moves give Black a very pleasant Patience, my friend. After 25 il.d5?!
game. That the exchange sacrifice not llf8 26 .txf7+? :X£7 27 e6 llfb7! 28 e7+
only offers White compensation but 'it>hB there is nothing.
even a very strong position is a plus. 25 . . . lted8 26 1fxc5?!
1 9 .l:bb5! Here Timman gives the following line,
1 9 tba3 b4 20 tClc2 tbb6 21 e5 lba4 leading to a clear advantage: 26 tbe l !
creates a very double-edged position in 'ifc3 27 .tdS 'ii'xc4 2 8 .txc4 a6 29 bxa6
wh.ich White's knight has little future on :b4 30 i.dS l;ta4! 31 lL\d3 J:.xa6 32
c2. tt::lxcS :as! (32 ... l:!b6 33 :a 1 l:.bS 34
1 9 . . .€ld6 20 e5 l2lxb5 21 cxb5 "W'a5 .tx£7+ 'it>x£7 35 l:ta7+ ..ti>g8 36 ltxg7+
22 d6! 'it>xg7 37 lbe6+ ..ti>gB 38 tbxd8 d3 39
The knight will find a good square on 'it>£2 and White has all the chances) 33
c4 and the bishop on c6, wh.ile it is not lbb7 lhd6! 34 i.x£7+! <t>xf7 35 lbxaS
immediately apparent how Black will be l:tdS 36 tt:k4 gS, when Black's task is to
able to mobilise h.is own forces. draw.

Solu tions to Ex ercis e s

26 . . . d3?! 27 lt:lb4! d2 28 liJd3 'ii'b3 inaccurate.

29 liJf2 'i'a4! 30 g3?1 a6! 31 b6! In endgames it is very important to
l:tdc8 32 b7 l:txc6 33 'ifa7 J:[xb7 34 fix weaknesses in the opponent's camp,
'*xb7 :c1 35 Wf3? Wd4 36 �g2 and a weakness is defined as a point you
.:e1 37 tbd 1 ?! .ltf8 38 'iff2 'ifd5+ can attack. Here that weakness is h7,
39 'Oti>g 1 lbf 1 + 40 'it>xf1 f6- + 41 adding weight to the h6-pawn. This
exf6 .txd6 42 'i'e3 cJ;;f7 43 �e2 needs to be proven in exact analysis, of
.tc5 44 'i'c3 'i'e4+ 45 <t>f1 W'h 1 + course, as Karpov has done: 30 l1hd 1 !
46 \Pe2 'ii'xh2+ 47 �f3 'ii'h 1 + 0-1 (30 �cd 1 allows 3 0. . .l:tgd4 31 :Xd4
cxd4!) 30...l:tgd4 (30 ... l:txd1 31 l:.xd1
Exercise 64: White to move l:th4 32 i.bS lhh6 33 :d7+ presents
Karpov-Hjartarson White with new attacking fronts on the
Tilburg 1 988 queenside; the pawns on a7 and b6 will
fall quickly, after which White's a-pawn
will be quick) 31 l:1xd4 l:txd4 32 J.gB
l'.ld7 33 �c2 followed by I:ld 1 and
White is close to winning.
30 hxg7 .:txg7 3 1 l:!.cf1 .l:td6 32 l:th6
White is winning.
32 . . . e4 33 .l:thxf6 h5 34 .l:t6f4 .l:td4
35 .l:.f7 + .l:.d7 36 l:Xg7 .l:.xg7 37 l:[f4
:g3+ 38 �c2 l:tg2+ 39· �c3 l!g3+
40 �d2 l:tg4 41 .l:tf7 + 'ifo>d6 42 'it>e3
a6 1 -0

One very important rule in the end­ Exercise 65: White to move
game is to keep your pieces active. Here Karpov-van der Wiel
White seems to be having - difficulty Tilburg 1988
bringing his rooks into play. Addition­
ally Black has a very healthy structure
on the kingside (it seems), as well as
passed pawns. Nevertheless White can
indeed undermine the kingside, activate
his rooks and put Black in a terrible
situation in only one move!
29 h6! l:txg2?!
One student correctly pointed out
that Black has no choice but to play
29 ...g6 because otherwise there will be
too many weaknesses to defend. This is
true, but his assessment that after this White has no weaknesses. Black has
move the pawns start rolling is quite two - cS and c6. White's e2-bisbop and

14 1
Ex c elling a t Po s itio n a l Ch e s s

knight should be working in that direc­ 26 Wt8 27 �f1 .!Lld7 28 llkl3l �e7

tion, so some reorganisation is neces­ 29 .J:la3 4:\b6 30 �e1 :ca 31 b3

sary. .J:lcc7 32 l:lda1 .i.b7 33 .id1 ! i.cS
1 9 lbd2! 34 bxa4 bxa4 35 .txa4 f6 36 i.b3
The knight is going to cS via e4 and l:l.xa3 37 l:lxa3 g5 38 �d2 �d6 39
the bishop is coming to £3. There is an l:.a5 .J:le7 40 lbc5 f5 41 lbd3 lbd5
alternative in 1 9 �f1 !?, with the idea of 42 f3 llb7 43 .txd5 exd5 44 .J:la8
19 ... e5 20 dxeS lDxeS 21 :xd6 when 1 -0
'White wins - a common accident. I
believe nudging the king to the centre is Exercise 66: White to move
a good move but, basically, I think Karpov-Yusupov
Black improves his position just as 1988 USSR Championships
much with 19 .. �f8. and it is better for

White to follow in the footsteps of

Karpov with 1 9 lDd2.
1 9 . . . 84
1 9 . e5 20 lDe4 J.e7 21 dxe5 l£lxe5 22
. .

:Xd8+ leaves Black in a dilemma. Re­

member prophylaxis -it often happens
when having your wishes granted and
preventing what you want to prevent
that a positional advantage materialises.

1bis is a very famous position and

hardly a difficult exercise. White needs
to activate his forces and can best do so
through a pawn sacrifice.
1 6 d5! !
16 .i.xe6 fxe6 1 7 'l'a3+? �f6 offers
White nothing since after 1 8 'i'xa7?
'lbS+ 19 'ot>g1 l:la8 the queen would be
trapped. Nor is 16 'it'a3+ �f6 17 .i.d3
'i'dS 1 8 'l'xa7 .i.g4 any good. as Black's
20 .i.f3 .J:la6 21 ttle4 .te7 22 a31 pieces are suddenly useful, while White
Preparing J.b4. is still undeveloped. Also harmless is 1 6
22 . . .:a7?! 23 .ib4! �gS after 1 6. ..'.fid8! 1 7 J.xe6 (17
White is close to winning. �xe6+ fxe6 1 8 J.xe6 'fibS+ 19 �g1
23 . . . i.xb4 24 axb4 lbb6 25 lbc5 l:l.e7! and ... l:l.ae8) 17 fxe6 1 8 lDxe6+

.taS 26 .J:la 1 �c8 19 'ifb3 l:l.e7, when Black is not

A move typical of Karpov's style. worse.
Now ... tt:k4 can be met with b2-b3. 1 6 . . . cxd5 1 7 .tb51

1 42
Solu tions to Ex ercis e s

Precise chess. The d4-square will not Exercise 67 : White to move

run away, so White should not be too Karpov-Hjartarson
eager. After the basic 1 7 llld4 'iVeSI Seattle 1 989
White's advantage evaporates because 1 8
'iia3+ can easily b e met with 18.. 'iid6

and Black is doing well.

1 7 . . . a6?!
This does not look good as now
White wins almost by force. 1 7 .. .'�£8?!
18 'ii'c3, intending li:ld4xe6, is hardly
advisable, either. 1 7 . .d4!?, as suggested

by Ka.rpov, seems logical. The bishop is

freed a little and the knight will be less
free on d4 than with a black pawn on
dS. This is what I would call unforcing
thinking. White has accumulated nu­ In his excellent book Technique for the
merous advantages and , now presents Tournament P�er Mark Dvoretsky fea­
some threats, so Black will have to get tures a chapter called Exploiting an Ad-
away from all kinds of forced lines as , , vantage. Here he investigates different
these allseem to lead to his end. techniques, one of them being Do not
1 8 'ifa3+ Wd8 Hurry. This means that improving static
1 8. . .'�£6? 19 .i.xd7 �xd7 20 1fc3+ advantages slowly can be a very useful
and White wins. way to prepare for an attack. But some­
1 9 'ii'a5+ 1 �e7 times you also have to jump and change
White also wins after 1 9 ... �c8 20 the nature of your advantage in order to
:ct+ �b8 21 ..c7+ Wa8 22 li:ld4 ..f6 make the most of your position.
23 .i.xa6 l:lb8 24 'i!fa5 'ii'xd4 25 .i.xb7+ Dvoretsky argues very well the case
�b7 26 J:tc7 mate! that Ka.rpov and Flohr, players with
20 'iib4+ Wf6 great technique, too often waited to see
Zaitsev gives 20... �d8?! 21 lDd4 'i'f6 what their opponents would do in de­
22 i.xa6! bxa6 23 .:.ct ! Z:teg8 24 J:tc6! fence, thus losing their advantage
etc. (e.g. 24 ... 1fe5 25 :lxe6Q. through their failure to act when neces­
21 'W'd4+ <3te7 22 .td3 'W'h5 23 h4! sary.
Wd8 24 lbg 5 l::thf8 25 J..e 2! 'iih 6 26 This diagram position is an exception
.tf3 l:te7 , 27 'ifb4 lt.Jf6 28 'iid 6+ to this, however, as Karpov makes the
l:.d7 29 'i'f4 lLlg8 30 J..g4 weB 3 1 most of his static advantage (the back­
.txe6 fxe6 32 llc 1 + �dB 3 3 ward c-pawn) through a tactical ex­
fue6 + We7 34 'ii'xfB+ •xta 35 change sacrifice.
lbxta wxta 36 l:th3 t:'iJe7 37 h5 ct>g7 1 6 .ixd41 exd4 1 7 Wc6!
38 h6+ �6 39 l:tf3+ ..te6 40 l:te1 + This is again tactical. 17 li:lxd4 tllxb4
�d6 41 l::tf6+ rj;c7 42 g4 lbc6 43 18 axb4 'ii'xd4 19 bS also provides White
l:.eB 1 ·0 with some advantage but if Black is able

1 43
Ex c elling a t Pos itio n a l C h e s s

to neutralise the b-pawn then White will l:c8+ �h7 43 l:c7+ �g6 44 .l:tg7+
not have anything but trouble, e.g. 19...£5 �h5 45 f3! 1 -0
20 e3 (20 �d2!? f4 21 'i'c6 .id6 22 �f3
might improve) 20...'Wd7 21 �c3 .ib4 Exercise 68: Black to move
should offer Black reasonable chances to Hjartarson-Karpov
defend. Seattle Match 1 989

1 7 . . ...xc6 1 8 :xc6 i.d7 1 9 �xd4! I remember seeing this game in Chess

The point. Black has no alternative Informant in 1990 and not understanding
but to part with two pawns for the ex­ anything despite reading the annota­
change, after which White's domination tions. Only now, when I am 1 2 years
of the light squares gives him a clear ad­ older (and 7 Elo points lower) do I un­
vantage. Please note the enormous ad­ derstand it.
vantage the opposite coloured bishops 30 ...l:l.c5! 1
are to White in this position. 19 l%.cc1 cS! A wonderful move, with a particular
would not be better for White. focus on ideal squares. Black's light­
1 9 . . . i.xc6 20 ll:lxc6 .l:lce8 squared bishop looks good but, in real­
Probably the only move. 20..l:tf7 21 ity, it is not doing much. The rook on
�gS! and 20...'iirf7 21 �xa7 :as 22 c1 also looked okay but was working
tDd6+! �e6 23 it)dbS are decisive. Here alone. The other bishop should proba­
you could imagine something like bly be on g7 but, again, nothing is clear.
23....id6 24 �c6 t'iJe7 25 �cd4+ �d7 What is obvious is that White will be
26 .ixa8 and White willcertainly win. able to attack both f1 and aS, and Black
21 .l:lc1 f5 22 tLld2 ttlf6 23 ll:lxa7 will at best make a draw in the nonnal
i.d6 24 e3 c5 25 �c4 i.b8 26 l0c6 course of the game. But after the text
b5 27 lb4a5 cxb4 28 axb4 lbd7 29 move, everything changes. The rook
d4 g5 30 ttlxb8!? .l:txb8 31 llc7 �f6 will find a- wonderful outpost on £5,
32 ll:lc6 l:b6 33 ll:le7+ �h8 34 where it defends f1 and attacks the f­
tt:Jxf5 lla6 35 .l:lc1 .l:la2?! 36 h3! llb2 pawn. The bishop will suddenly be able
37 e4 l:xb4 38 g4 h5 39 e5 hxg4 to attack the king from eS and the rook
40 exf6 gxh3 41 i.xh3 llxf6 42 on e8 is free to do whatever. White con-

1 44
Solu tions to Ex ercis e s

tinued logically. 29 ll:ie2!

31 .:d7 With the threat of lDd4-f5. The
Also possible is the more passive 31 alternative 29 lLld1?! is less good as Black
l:td3 �g7 32 b3 :t£5 but then Black is in can activate his rook with counterplay
very good shape. The main threat is, of with 29 ...gxf4 30 gxf4 l:.bS!, when White
course ... .i.x£3, and after both 33 "ii'b4 does retain some advantage after 31 :b2
l:te2! and 33 'Wg4 :Z.e2 34 l:td8+ rJolh7 f5 32 e5 but this is less clear than the
Black is doing well The main idea here, �c 5
though, is 35 tbgS+? :txgS 36 'Wxe2 Tills is the move that dissuades people
:Xg2+ and Black wins. from t'Lle2
3 1 . . ..txf3 32 'i'xf3 IUS % -% 30 ll:ic3!
Black offered a draw as it suited his Very nice. N ow the weak light squares
position in the match. After 33 'Wg3 in Black's camp are plain to see.
i.g7 34 f4 �xb2! Black has a slightly 30 . . . l:d6 3 1 l:d5! l:bd5 32 ll:ixd5+
better endgame. But not 34 .. Jhf4?! 35 �e6
'i'xf4 .i.eS 36 l:.xf7 with an immediate But could have put up more resistance
draw. with 32...<it>f7 33 tbe3 rJ;g7, though after
34 a4 it is obvious that White has control
of the position and Black has many
weaknesses. I cannot believe that such a
position can be saved.
33 �e3 b5 34 ll:lf5 ll:lb6 35 ll:lxh6
lLlc4 36 lLlfS �f7 37 b3 lbd2+ 38
�e3 1 -0

Exercise 70: White to move

Thessaloniki Olympiad 1988

Exercise 69: White to move

Skelleftea 1 989

Here is a classical case of wanting and

executing. To my delight most of my
students quickly realised that the pri­
mary idea for White in this position is
to bring the knight to fS. The only ques­
tion is that of which route to take. The
ans'wer is a little complicated, and most
of them, unfortunately, failed to find A classical worst · placed piece
the right path. scenario. The a1-rook is not very well

1 45
Ex c e lling a t Positio n a l Chess

placed and would very much like to 4 1 .tg8! fxe5 4 2 lUxe5 ltlb4 43
come out. Meanwhile 20 lbc5 can be l2Jc4 h4 44 i.h7 hxg3 45 hxg3 �e7
answered quite well with 20 . tbd7l, giv­
. . 46 i.e4 �f6 47 g4 'li>e6 48 �f2
ing Black a tenable position. ll:Jd5 49 �f3 ltlb4 50 g5 c5 51 f5 +
Sometimes it is a good idea to look �e7 52 �e3 �8 63 �d2 �e7 54
for the opponent's next move as this �c3 1 -0
might tell us something about what we
should do. In this case this is 20. ..ie6.
. Exercise 7 1 : White to move
Knowing this, it should not be too dif­ Karpov-Timman
ficult to find our way, even if White's Amsterdarn 1 99 1
next is rather odd.
20 Aa3! 1
Best. O f course it takes some expla­
nation why this is better than the tempt­
ing line many of my students carne up
with, namely 20 .:CcS!? .ie6 21 lbxaS!
lbd7 22 lt.lxb7 when Black can either
play with an exchange against two
pawns or a piece against three pawns
after 22 ...l:b8 23 .:txc6 .:Xb7 24 aS - in
my opinion this is a very dangerous po­
sition for Black, and the acceptance of
the exchange is far more logical. Never­ I remember that, when collecting the
theless White seems to retain all the raw material for these exercises, only a
chances. few positions made me aware of seeing
However, I believe 20 .:ta3! is a much something that I did not initially under­
better move as it results in a clear ad­ stand. 1bis is one of those. If we
vantage without taking any chances. quickly compare pieces, then we will
White already has a static advantage and soon find out that the e7-bishop is
then brings his pieces into play in a very Black's worst piece and the cl-bishop is
harmonious way. I willleave it to you to White's worst. We also see that White's
make up your own mind on this, but at knight would be very well placed on d6,
least you know why I mean what I and in the case of an exchange and the
mean. transformation of the e-pawn into a
20 . . .i.e& 21 ll:Jc5 i.f7 22 ::td3 :ae8 passed pawn, the bishop on g6 would
23 i.h3 ll:JcB 24 i.d7 AdS 25 Acd1 be misplaced.
ltlb6 26 �g4 Axd3 27 ::txd3 �8 2S It is my experience that most stu­
AdS+ :ea 29 Ad4! ::tb8 30 Ad71 h5 dents overrate the bishops in a situation
31 i.h3 �e8 32 .:tc7 ll:Jas 33 :lxb7 like this, and refrain from i.f4 simply
.:xb7 34 ltlxb7 ltlb6 35 lLld6+ ! �e7 because of the possible exchange there.
36 ltlxf7 l0xa4! ? 37 'OhSI ltlxb2 3S However, a look at the knight on d5
e5 a4 39 ltlg6+ �eS 40 .ie6 ll:Jd3 suggests that this might not be some-

Solu tions to Ex ercis e s

thing we should fear. Therefore the his . a8-rook. However, this position is
right course of development revolves uncomfortable for Black. White has the
around finding the ideal square for the two bishops might plan an exchange on

dark-squared bishop, and that is f4. d5, might start a kingside offensive and
There is no other square to offer the still there is a lead in development Of
bishop an immediate future. course Black has some chances to de­
1 6 .tf41 fend.
With the intention of tLld6. Black now 1 7 �6! .txd6 1 8 exd6 :tea 1 9
replied... :xe8+ l:txe8 20 .ixd51 cxd5 21 d71
1 6 . . ...85 :e7 22 :tc1 :txd7 23 :tc8+ :td8 24
...And quickly went downhill. But b41 Wb6 25 i..c7 l:!.xc8 26 .txb6
could he have defended better? I have axb6 27 Wxd5 h6 28 Wxb7 l:!.c1 +
analysed two alternatives. 29 <li>h2 :c2 30 'i'xb6 :xa2 3 1
The first is 16...lLlxf4 17 'i'xf4 Le4 Wd41 1 -0
18 l:.xe4, when there is no reason to But why was it that this example was
look any further. The f7-pawn is terribly so fresh for me? It was simply the enor­
weak, as are the light squares around mous power of the passed pawn on d6. I
Black's king. The old rule concerning did no t realise it before playing through
opposite coloured bishops is like this - the game. Did you?
in a pure opposite bishop· endgame they
can be a drawing factor if the defender Exercise 72: Black to move
is successful in putting his pawns on the Gelfand-Karpo�
same colour complex as his bishop Reggio Emilia 1 991
(while the attacker will put his pawns on
the colour of the opponent's bishop). In
the middlegame opposite coloured
bishops favour the attacker (in a situa­
tion like this) as he will simply have one
more piece involved in the fight for the
squares of his bishop's colour. Finally
there is a minor rule - if you have a
winning position there is no easier way
to transform it into a full point than
with opposite coloured bishops Of .

course you should be very careful in this

case as you might end up in a drawn With his latest move, 1 8 b4, White is
endgame if you are not fully aware of trying to establish a strong blockade on
what is going on. d4. Black, naturally, willnot sit back and
Next we have 1 6.. .i.xe4!? 17 llxe4
. let this happen, and he forces White to
'ii'aS which, in my opinion, is the only occupy the square with his OWn pawn!
possible defence for Black. He prevents 1 8 d4!

lL:.d6 and prepares the development of There followed a forced sequence.

Ex c elling B t Positio n B I Chess

19 exd4 cxb41 20 axb4 .txt3 21 74 �h 1 l:.e2 75 �g1 J:tc2 76 l:.b4+

.txt3 l:.xb4 22 lbd5! 'it>e3 77 l:.b3+ �e2 78 l:.b1 l:td2 79
White cleverly seeks salvation in a l:.a1 l:.b2 80 l:.f1 �e3 81 l:.a 1 l:.g2+
rook endgame that is uncomfortable 82 �h1 g3 83 l:ta3+ ¢>14 84 l:ta4+
but perhaps possible to hold. 22 �c6?! · ¢>15 85 hxg3 l:.xg3 86 'it.>h2 l:tg4 87
meets with 22 . .:ds.
. l:.a5+ �f4 88 .:ta4+ �g5 89 l:.a3 f2
22 . . . ll:lxd5 23 .txd5 ll:Jf6 90 l:.t3 l:.f4 91 l:.xf4 � - �
Here Black could have played more
ambitiously with 23 . l:ld8 24 �c6 tl:l£8
.. Exercise 73: White to move
25 dS llxf4 26 llal ltd6 with a slight Kamsky-Karpov
advantage according to Karpov. I think Reggio Emilia 1 991
that White should be able to make a
draw quite easily with 27 life 1! but there
is no disputing who is under pressure.
24 i.c6 .l:ld8
24 . . llc8!? 25 dS :Xf4 26 l:.al l:.c7

was a possible attempt to keep some life

in the position, but I feel that the
passed pawn will soon let White escape
into the same endgame as in the game.
25 d5 l:.xt4 26 :a1 l:.d6 27 J:txa7
h5 28 f3 ll:lxd5
28 ... tl:le4!? is met with 29 l:.a4! l:.g6+
30 'i!th 1 l:.gf6 31 l:.b 1 and perhaps Theme: Always keep your eyes
White is on his way to being better. peeled. One of the things that should
29 .txd5 l:.xd5 30 l:.a3 .l:lg5+ 31 always be imprinted in your mind is
�h1 �h7 32 l:.e3 �h6 33 .l:lg 1 l:.a5 hanging pieces. Here White starts with a
34 .l:ld3 .l:lf6 35 .l:lc3 l:.ff5 36 .l:ld3 g6 rather basic combination.
37 l:.c3 .l:la4 38 .l:ld3 �g7 39 l:.c3 25 .txd41 cxd4 26 d6!
�6 40 .l:le3 J:te5 41 .l:lge1 l:1xe3 42 White wins a piece, but matters are
l:txe3 �f5 43 .:te2 �f4 44 �g2 g5 not always that easy!
45 �2 f5 46 l:.b2 g4 4 7 fxg4 hxg4 26 . . .1i'xd6 27 'W'xa5 fxg3
48 .l:lc2 .l:la3 49 .l:lb2 �g5 50 l:.b8 Watch these pawns fall.
l:ta2+ 51 �g1 l:.d2 52 :as �4 53 28 hxg3?
l:.a3 .:te2 54 .l:la 1 �f3 55 l:.f1 + �e4 Permitting Black to generate play
56 l:.a 1 f4 57 l:.d 1 l:.c2 58 l:.e1 + with his rook. The right move was 2 8
�3 59 l:.f1 + �e3 60 l:.e1 + l:.e2 61 'ifdS!, whep. White has the advantage.
.l:la1 .l:lc2 62 .:te1 + ¢>d3 63 l:.f1 f3 28 l:.e3 29 'ti'd5 •xd5 30 .txd5

64 .l:la1 �e2 65 l:.b1 l:.a2 66 l:.f1 l:.xg3+ 31 �h2 l:.xd3 32 l:txf7 .l:txf7
�e3 67 .l:lb1 l:ta4 68 .l:lb3+ �4 69 33 .l:lf1
l:.b8 .:ta 1 + 70 �2 l:.a2+ 71 �g1 33 llxb7? l:ld£3 would be a grim sur­
l:.g2+ 72 �h 1 l:.d2 73 �g 1 .l:lg2+ prise.

Solu tions to Ex ercis e s

33 . . . b5! 34 llxf7 l:c3 35 l:f4+ �h8 tually the opposite of all this. Opening
36 l:lxd4 g5 37 i.b7 bxc4 38 .i.xa6 the game on both the queenside and in
38 a4! was · nec��sary to keep the
- the centre is particularly important, acti­
hopes of winning �ve, but it seems that vating his bishops without opening lines
Black has good chances of survival after for the white bishops (therefore I dis­
38...a5. count ... d6-d5). All this leads directly to
38 . . .l:l.xa3 39 .ixc4 l:lf3 40 l:ld7 l:lf4 Karpov's choice in the game.
41 i.d3 �g8 42 �g3 �8 43 l:ld6 1 8 . . . b51 1 9 axb5
h5 44 l:d7 �e8 45 i.b5 l:b4 46 It is important to remember that after
l:d5 + �e7 47 .ie2 h4+ 48 �h3 1 9 cxbS axbS 20 aS cS! Black is opening
�6 49 l:ld3 �5 50 l:la3 Ad4 5 1 even further. White's king and g2-
:as 'iti>f4 5 2 l:f8+ �e5 53 .i.f3 llb4 bishop seem to be all dressed up with
54 :as l:ld4 55 lla3 �f4 56 i.h5 nowhere- to go. Meanwhile the other
'it>e5 57 ..tg6 �f6 58 i.h7 l:f4 59 bishop will also soon find that its posi­
l:le3 Wg7 60 i.d3 � 61 ..ta6 �g6 tion is compromised.
62 i.e2 Yz -% 1 9 . . . axb5 20 cxb5

Exercise 74: Slack to move

Reykjavik 1 991

20 . . . ..tb6!
This is the final finesse that your
analysis requires. After 20... 1Lxb5 21
Aa.S! c6 22 Aha1 Black is tied down in a
Here I see the following: White has very uncomfortable way, and the white
no control over the light squares on the rooks are active.
queenside. White's king is uncomfort­ 21 l:l.hd1 .txb5 22 �e1 liteS
able and he therefore wants the centre Black has a good position, and all of
to remain closed. White is preparing a his pieces will eventually find goodposts.
pawn storm on the kingside but his mi­ Karpov suggests in his notes that
nor pieces don't properly support this 22 ... c6!? is also possible. This is most
adventure, so it will take a long time to likely with the idea of . .d6-d5 at a given

take off. moment, but the move played in the

Black, on the other hand, wants vir- game seems more clear-cut. Now the

1 49
Ex c elling a t Positio n a l Ch e s s

knight c an return to the game via f8 and opening ftles. This cannot be started
e6, as it should. This is much better than with a2-a3, not only in view of ... tt:lb3,
conceding control of the dark squares but also ... a5-a4, fl.xing the pawns in an
with ... exf4 at some point. unfortunate way. There are therefore
23 l:xa8 J:l.xa8 24 l:.a1 l:td8! 25 'it>d2 two equally good first moves. The ques­
c6 26 'iPc2 liJf8! 27 tLlg3?1 tion is in what order should they come.
White is forced to play 27 £5 in order To have solved this exercise you need
to fight for survival. to have found the plan more than the
27 . . .ltJe61 28 liJf5 Wc7 29 fxe5 execution.
dxe5 30 i.f1 c5 3 1 bxc5 Wxc5 32 1 8 i.f3!
..t>b2 lt:lf4?! 33 d4! exd4 34 l!Jxd4 18 b3!, with the idea of i.£3-dl , is
i.xf1 35 J:txf1 i..c7 36 e5 l:xd4 0-1 equally good. It also prevents a possible
... a5-a4, although I don't feel that this is
Exercise 75: White to move anything to prevent. After 18 ... g6 1 9
Karpov-Salov .i.h3 gS 20 .i.g4 there is really do differ­
Reykjavik 1991 ence with the game. White has a com­
fortable advantage.
18 . . . g5
1 8... g6 1 9 .i.g4 favours White.
1 9 b3 i.c8 20 i..d 1 !
Putting the bishop to use. Notice
how all the pieces are playing, and how
the bishop is in reach of both ways out
of the pawn chain (a4 and h5).
20 . . .i..d 7 21 a3
White has the advantage.
21 . . . l:a7 22 b4 liJb7 23 J:tf2 liJd8
24 We3 axb4 25 axb4 liJf7 26 h4
This is a classical example of the �h81 7 27 � 1 ! l:tg8 28 l:cc2 Wf8
worst placed piece. Not surprisingly, 29 J:ta2 lba2 30 J:l.xa2 gxh4 31
however, there is more to it than that. gxh4 Wg7 3 2 �e1 ! liJh6 33 .:ta7
White has two standard plans at his dis­ i..e8 34 �d2 'itg2+ 35 �c1 lllf 7 36
posal, or so it seems. The first is ad­ l:txc7 l:tg3 37 1i'd2 1i'h3 38 'iPb2
vancing on the kingside with the pawns lt>g7! 39 J:tc8 39 . . .id7 40 l:c7
to induce weaknesses in the enemy i..eB 41 i..e 21 Wxh4 42 c5 bxc5 43
camp. But most likely Black will stop bxc5 �fB 44 c6 'ith2 45 llcB 1i'f2!
this before it is even started with ... g7- 46 l:tb8 _'i'c5 47 'i'c2 l!Jg5 48 l:.b3
g5, so there is really no need to start l:tg2 49 l:b5 -.a7 50 'i'd3 h5 51
putting the pawns on light squares. 'iic4 'it'd4 52 �b3 llg3 53 �c2 l:tg2
The other plan, which is more logical, 54 �b3 l:g3 55 Wc2 J:l.g2 56 'it'd3
is the slow advance of the queenside 'it'a7 57 l:b7 'it'c5 58 :tb5 'it'a7 59
pawns, driving the knight away and l:b7 'i'c5 60 'iPb3 h4 61 'it'c4

1 50
Solu tions to Ex ercis e s

'irxc4+ 62 �xc4! tbf3 63 �b51 ply 1 5...J.xd4 1 6 ll:lxd5 J.x£2+ 1 7 l:Z.x£2

tbd4+ 64 �c4 h3 65 c7 tbxb5 66 i.xd5 18 J..xd5 exdS 19 l:td1 ::ld8 20 e4
cs• h2 67 'ife6 1 -0 d4 21 :Xd4 lbd4 22 i.xd4 lllc6 it does
not appear to me that \Vhite has any
Exercise 76: White to move advantage at all. The weaknesses of e4,
Karpov-Korchnoi b3 and c4 will fully make up for the
Biel 1 992 advantage of bishop against knight.
1 5 b4 is, as I said, fine, but after
1 5 ... J.xd4 1 6 J..xd4 lllc6 1 7 ii.b2 l:tac8
Black has developed with gain of time
and is close to equality.
1 5 ...tbc6?!
Karpov does not put this down as a
mistake but I believe that after this
Black has no real chance of saving the
game. After 1 5 ... J.c6? 1 6 ..ixd5 Black
loses, but there seem to be saving
chances after 1 5 ... a6, where the main
idea, of course, is 1 6 lllc7 lla7!, which is
Both players need to complete the practically forced. I find it difficult to
development of their queenside, and the understand that Karpov considers only
manner in which this is conducted is 16 ...lLlxc7 1 7 i.xb7, when Black will
essential. Because White has the move suffer needlessly from the power of the
he has the opportunity to win himself two bishops. Returning to 1 6...l:ta7, af­
some territory in the centre and thereby ter 1 7 lllxd5 i.xdS 1 8 J..xdS exd5 19
limit the amount of space for the black l:tdl l:td7 20 b4 J.e7 21 llld2 \Vhite's
pieces. For this reason exchanges are advantage is more obvious than it is
not desirable, as a principle. ! 5 llld2 is overwhelming. This is actually one of
not a very good move because the the principles of defence that I am
knight is not going anywhere useful. But working on. Very often one player has a
1 5 b4 is okay, although Black is solid simple line that leads to an advantage -
and has a reasonable position. as in the game - but the opponent can
Karpov came up with a more surpris­ play some irregular moves and avoid it
ing move, exploiting the weakness of because the main objective of the de­
the long diagonal. fence tends to be to avoid forcing lines,
1 5 tbb51 as these will inevitably lead to pain and
Now \Vhite can advance his pawns in suffering.
the centre and gain space with tempo. 1 6 b4!
The alternatives are less convincing. 15 Now the pawns start rolling, and
ttk3 is a tactical way of trying to make Black is simply pushed back.
the move you want to play work, but it 1 6 . . . i.e7 1 7 e4 �f6 1 8 e5 tLld5 1 9
allows simplifications, and after the re- � 1 c3! lt'lxc3 2 0 i.xc3

Ex c e lling a t Positional Chess

White's advantage is beyond doubt. As White has no immediate way of

Just compare the pieces. breaking down the defences, and as
20. . .tDd8!? Black has no ways of creating counter­
20...l:fd8 2 1 l:tfd1 a 6 22 .!Lld6 �xd6 play, White should find a way of
23 exd6 l:d7 24 l:tac1 ::i.ad8 25 �f6 strengthening his position quietly. And
gxf6 26 �xc6 �xc6 27 l:xc6. we know that this is done by improving
21 i.xb7 the worst placed piece. Here it is the
21 J:ac1 ! king. One might overlook this basic
2 1 . . .lbxb7 22 l:lfd1 l:lfd8 23 l:lxd8+ manoeuvre due to the queens still being
.txd8 24 l:ld1 a6! 25 l:.d7 axb5 26 on the board, but this would be a grave
l:txb7 �8 27 i.d4 l:la6 28 h4! h5 mistake. This is an endgame, a queen
29 �1 �e8 30 �e2 �8 31 l:lb8 endgame. There might be some threats
�88 32 l:lb7 'iti>t8 33 �11 �.a 34 against the king but there is no reason
'ili>g2 �8 35 ltb8 �e7 36 ltb7+ �8 to fear being mated.
37 ..li.e3 �e8 38 �f1 �8 39 l:.b8 31 'iti>g31
�87 40 J.g5+ f6 41 exf6+ gxf6 42 Simply improving the pos1t10n.
.txf6+ �xf6 43 l:txd8 .l:xa3 44 Moves like 31 .1e2? don't work ·out
l:th8! l:lb3 45 .l:!.xh5 .l:xb4 46 �g2 well, if for no other reason than the
�g6 47 lt85 �f6 48 f4 l:tb2+ 49 break 3 1 ...c5!? (probably not necessary)
�3 b4 50 ltb5 b3 51 �g4! l:tb1 52 32 bxcS aS, which gives Black several
�h5 b2 53 g4 1 -0 ways of organising counterplay, e.g. 33
'fla7 W'd8! 34 �gl �c6 and the advan­
Exercise 77: White to move tage is in doubt, or 33...W'b7 34 ..xb7
Karpov-Lautier �xb7 35 .1b5 W£8 and Black has, for
Biel 1992 instance, ... .1d5-b3; he should probably
be able to draw this, although f7 is a
little weak.
31 . . .'i'b7?
Leading to a lost endgame. Actually it
is lost in a very ordinary way. White will
simply create an extra weakness on the
kingside. However, 31 ...'ilh8 32 �f4 is
also highly uncomfortable as Black has
no way of improving his position while
White will slowly but surely get closer
and closer.
32 W'xb71
Quite a technical position. Black has Karpov does not miss these chances.
an extra pawn but it is of little impor­ 32 . . ...li.xb 7 33 �4 �8 34 �g5
tance. Of greater significance is the One of my students suggested 34
pressure on the long diagonal, White's �e3 rife7 35 �d4 �d7 36 �cS �c7 37
active forces and Black's passivity. .1xh5!! gxhS 38 g4 and White wins.

1 52
Solu tions to Ex ercis e s

This would be very nice if Black did not opponent's breathing space.
have the equally cunning 35...c5+! 36 1 0 f4!
�xeS .ix£3 37 gx£3 gS!, and it is Black Very problematic for Black to meet.
who is winning. 1 o. .g4 closes the kingside, after which

34 . 'iite7 35 i.e4 .iaS 36 f3!

. . White will be able to go to £2 with the
Like clockwork. After g2-g4 Black king, should Black organise something
cannot prevent White from creating a on the kingside. Therefore Short swal­
passed pawn on the kingside. lows his pride and settles for a more
36 . . . .ib7 passive strategy. This does not alter the
36. .c5 37 .i.xa8 cxb4 38 .ic6 b3 39
. situation much, though; Black is in
i.a4 b2 40 i.c2 is analysed by Karpov. trouble.
White wins. 10 'ii'b3!? is an interesting attempt to
37 g4 .taB 38 gxh5 gxh5 39 f4 exploit the weakness on b7. Black can­
.ib7 40 .if3 .iaB 41 'iftxh5 1 -0 not play 1 0 ... tl'la6 because of the nice
little combination 1 1 i.xb7 .:tb8 12
Exercise 78: White to move .ixc8! with an extra pawn, or a winning
Karpov-Short advantage for the queen. The right an­
Linares 1 992 swer is 1 0...c6! (Black would be seri­
ously worse after 1 O . tbc6?! 1 1 lt)bS! a6
. .

t2 lt)xc7 [1 2 .i.xc6 axbS 1 3 i.xbS wins

a pawn, but the bishop is completely
astray] 1 2...tl'lxd4 13 1id1 l:tb8 1 4 e3
tl'lc6 1 5 tl'lxa6 bxa6 1 6 i.x"c6 and I can­
not see sufficient compensation here)
1 1 dS eS! (1 1 ...cxd5 1 2 cxdS eS 1 3 tl'lxe5
is an important fmesse, but Black is not
forced to take on dS) 1 2 dxc6 (1 2 cS
�h81) 12 ... tl'lxc6 13 tl'ldS lbxdS 1 4
i.xdS+ tlrh8 and Black has n o serious
worries. Actually White has made little
A classical strategic situation, not or no progress on the queenside, while
unlike a nwnber of positions in the Black willsoon stan advancing.
King's Indian Defence actually (I have 1 0 h6 1 1 d5 ltJa6 1 2 b4! exd5 1 3
. . .

added some examples below to illus­ liJxd5

trate). It is a race on both sides. White 1 3 cxdS also gives White a substantial
will advance on the queenside and Black advantage due to the pressure on the c­
on the kingside, and it is a simple matter f.tle, the extra space and the better co­
of who will be first. ordination of the pieces (look at the
But then again, slowing down the poor fellow rotting away on a6). But I
opponent is an important ingredient of have noticed that Karpov loves playing
this kind of situation. One of the technical positions where the pieces are
important aspects of this is limiting your the main actors, while he is less happy

1 53
Ex c B IIing a t Po sitio n a l Chess

with positions where the more strategic

aspects of the positions are dominant.
Here, of course, the great masters are
Botvinnik and Korchnoi, but a player
like Gelfand also springs to mind. My
definition of strategy is play with pawn
1 3 . . . tt:Jxd5 1 4 .i.xd 5 + �h7 1 5 b5!
tt:Jc5 16 tt:Jxc5 dxc5 17 'it'c2 a& 1 8
a41 llb8 1 9 fxg5! hxg5 20 lla3 c6
21 .i.g2 i.f6 22 i.e3! .i.d4 23 i.xd4
cxd4 24 e3 dxe3 25 :Xe3 .i.e& 26
g4! .:bes 27 bxc6 bxc6 28 c51 •f6 9 a5 1 0 a3 c5!?

29 i.xc6! llb8 30 gxf5 �f7 3 1 i.g2 10 lLld7 is probably

... better but it is
llb2 32 "ii'c3 •xc3 33 lxc3 IdS 34 the idea that is important here.
c6 lldd2 35 .i.e4 .l:!.e2 36 c7 Axe4
37 c8"ii' 1 -0 Exercise 79: Black to move
Romero Holmes-Karpov -

Examples from the King's Indian De­ Madrid 1992

1 d4 tt:Jf6 2 c4 g6 3 tt:Jc3 i.g7 4 e4
d6 5 tt:Jf3 0-0 6 .i.e2 e5 7 0-0 tt:Jc6
8 d5 tt:Je7 9 tt:Je1 tt:Jd7 1 0 tt:Jd3 f5 1 1
f3 f4 1 2 .i.d2 g5

There are a few issues for Black to

consider here. How .to develop, for ex­
ample. How to generate counterplay on
the queenside to match the attack White
will surely establish on the kingside in
1 3 g4! ? the not too distant future. And is it pos­
sible to use the pin on the e-file for any­
And another one: thing?
1 d4 tt:Jf6 2 c4 g6 3 tt:Jc3 i.g7 4 e4 Let us start with the last first. No, it is
d6 5 tt:Jf3 0-0 6 .i.e2 e5 7 0-0 tt:Jc6 not After 1 1...d5? 12 llJxdS lLlxdS 13
8 d5 tt:Je7 9 tLld2 'i'd3! White is on his way to winning a

1 54
Solu tions to Ex ercis e s

pawn. Still, the idea could work if the 13 ..• a4 14 bxa4 ..t.xa4
bishop on b4 were not hanging. 1his is Black is obviously first. White is in
useful to remember. big trouble.
The second is also obvious. The 1 5 lbb 1 !? i.c5!? 1 6 i.e3 b6 1 7 i.d4
weakness of White's king position is on i.d7 1 S g4 .l:l.a5 1 9 lbf4 i.c6 20
the a-file, and a quick attack with the a­ :hg 1 lbd7 21 ltlh5 g61 22 'i'd2?
pawn, supported by the bishop from J.xd4 23 'W'xd4 gxh5 24 gxh5+ �fS
d7, would quickly open files on the 25 •o7+ �e7 26 .l:l.g5 .-as 27 b4
queenside. The second and the third :xg5 2S 'W'xg5+ f6 29 Wg7+ �dS
answer the first question. The bishop 30 Wxh7 "iYa2 3 1 Wh6 �cS 32 •f4
belongs on d7 and the a-pawn is to be .l:l.hS 33 h6 •us! 34 .l:l.d3 Wg5 o-1
pushed forward, and the need to protect
the bishop on b4 decides the order of Exercise SO: White to move
moves. Karpov-lllescas Cordoba
1 1 . . . a51 Dos Hermanas 1 994
The best kind of move to play. As
explained, 1 1 ... a5 is both prophylactic
and aggressive. White now (unwisely)
decided to lose more time.
1 2 'i'f2
After 12 tt'lge2 dS Black's prospects
are certainly preferable but White
needed to develop and fight in some

White has the advantage but there is

still the matter of how to exploit it.
· Well, in an endgame it is often a good
idea to create as many weaknesses as
possible in the enemy camp. It is also
good to think about the future of all the
pieces. Here this led Karpov to lose
25 i.a3! !
1 2 . . .J.d7! 1 3 lbge2 The alternative is something like 25
Development is necessary. Karpov ltcl <j;f7 26 l::tc6 .:Xc6 27 dxc6 l:tc8 28
also gives 13 'fth4 a4 14 �gS a3! (with .:.c1 �e6 and Black willbe all right. He
the threat of ...a3-a2!) 1 5 tt'lge2 a2 16 will round up the c-pawn but probably
�d2 ltaS with the better chances for be forced to make some concessions on
Black. the way. Karpov says it is equal, which

1 55
Ex c e lling a t Positional C h e s s

is probably not too far off. 33 .. .lt:lc5 34 c7 lt:le6 35 l:c4 �d7

So why is 25 �a3 so obvious? Well, 36 l:lxb4 lt:lxe7 37 llc4! lt:le8 38
first of all it forces Black to put his ltxe8 �xeS 39 �d4 �b7 40 ..i.a5
pawn on a more exposed square, and �c& 41 'iic41 �d7 42 ic3! h5 43
that is a dark square! Secondly, it opens a4 �e6 44 id4 f5 45 exf5+ �xf5
a path for the white king to travel to the 46 �d5! �4 47 b4 �xf3 48 �c6!
queenside, where it can become an im­ g4 49 b5 axb5 50 a5! �e4 51 a&
portant player. Thirdly it introduces the 1 -0
possibility of l:cl-c4 later to attack the
b-pawn, which is what happened in the Exercise 81 : Black to move
game. Morovic-Karpov
25. . b4
. Dos Hermanas 1 994
Black cannot live with 25.. .:td8 26
.:tel ! followed by .:tc6.
26 ib2 'iif7 27 le1 lt:lf4+ 28 �e3
Karpov gives the following convinc­
ing variation as a possible outcome of
major exchanges: 28. ..ltlh3 29 l:txc8
:xeS 30 ltc1 .:.Xc1 31 .ixc1, with the
idea of 32 �d2, and the pawns on the
queenside will be eternally weak (and on
dark squares).
29 id4!
A simple move, improving the worst It is always important to be able to
placed piece. spot the weakest point in the oppo­
29 . . .'�e7 30 ltc&! nent's position Here it must be c4 as

After this Black is forced to allow only the queen offers protection, which
White a passed pawn. is hardly convenient for White. In fact
30 . . .ltxc6 31 dxc& . ltc8 32 l:lc 1 White has just played 1 5 dxe4?, when 1 5
lt:le& 33 .ib&! ltl xe4 would have been level. Not sur­
White does not give away the advan­ prisingly Karpov succeeds in exploiting
tage of bishop over knight. After 33 the weakness of c4.
.:tc4? llJcS! 34 �xeS dxc5 35 :XeS �d6 1 5 . . .lt:lc8!
36 ll£5 (the pawn ending after 36 �d4 There are some questions that need
would win if the white pawn stood on to be answered before this move is
h3 instead of h2; now Black is cruising picked out. First, which is Black's worst
after 36 ... .:txc6 37 .:txc6+ �xc6 38 e5
placed pie ce? The knight, obviously. It
£5! 39 e6 'itd6 40 e7 <it>xe7 41 'itte5 f4 42 has no future on e7 and needs to find a
'it£5 h6 and White loses in view of 43 better square. The c4-square, which is
'itg6 g4!) 36...�e6 Black will be able to already in our sights, is the premium
hold the endgame. choice! But what about the rooks -

1 56
Solu tions to Ex ercis e s

which one should go to d8? Well, at the dxc4 25 a5 i..f8 26 a6 f6 27 f4 c5

moment White has this enormous 28 l:lxd8 'i'xd8 29 'i'e1 'i'b6 30
chunk of pawns in the middle, and fxe5 fxe5 3 1 i..xe5 l:xa6 32 l::td1
there is no reason to rush to exchange i.g4?1 33 l:lb1 'it'e6 34 i..c 3 l:la2!
them with .fl-f5. Meanwhile the
.. 35 bxc5 i..x c5 36 .i.d4 .:txg2+ ! 37
queen's rook might be able to prove �xg2 1i'xe4+ 38 �g1 0-1
itself useful where it is, so the decision
is rather easy to make. Exercise 82: Black to move
1 6 l::Ud1 l:d8 1 7 lba4 Adams-Karpov
Preventing . ..lbb6. After something Las Palmas 1 994
like 17 l:.d2 ll:lb6 1 8 cS ll:lc4 1 9 :l.d3
'iie7 Black would have profited 100%
from his strategic choices and would be
ready to find a new plan for improving
his position.
1 7 . . .'ii'e8 1 8 l:d2

Subject: Ideal squares and improve­

ment of the worst placed piece. The
most problematic piece to bring into
play is the d7-bishop, and something
awkward like ...'ifi>b8, ....tc8 and . .b7-b6

is both slow and easily preventable

1 8 . . . a61 (i.bS at the right moment, for exam­
Black does not want to lose control ple).
over dS in order to control c4 hence
- The ideal square for this bishop is,
this preparatory move. Instead 18 ...b5? unfortunately, already occupied, so the
19 cxbS cxbS 20 tZlc3 i.c4 21 'iidl right plan is a rearrangement of the
would not be bad for White. pieces.
1 9 .1c3 b5 20 ll:lb2 ll:lb6 21 cxb5 1 4 . . .tllb8!!
axb5 This is the logical solution, although
Black has a strong position. White has not completely natural. The knight
poor co-ordination among his minor needs to move and d7 is the best avail­
pieces and some serious weaknesses. able square,. so in this way the bishop
Now Black opened up the position to his and knight simply switch places.
advantage. 14 .. ll:lb4? is logical too, in order to go

22 'ii'd 1 d5 23 a4 tt:lc4! 24 tt:lxc4 to dS, but after 15 1Qd6+! cxd6 1 6 'ii'xb4

Ex c elling a t Positional Ch e s s

Black is in trouble. position is fS, while White's is h2. An­

1 5 lbf6 other simple observation: the ideal
1 5 tbd4 .tc6 1 6 tbf6 'i'c5 is difficult square for the £3-knight is fS. So the
for White. first move can be selected - as always -
1 5 . . ...tc6 1 6 ie2 �d71 1 7 lbh5 based on the sum of the most impor­
17 tbxd7 was probably better as the tant observations (of course it is diffi­
knight is now lost on the edge, but Black cult to know what are the most impor­
is already doing better. tant observations!).
1 7 . . . ..i.xf3 1 8 iLxf3 .txe5 1 9 *e4 1 5 h3!
c6 20 l:xd7 ..txb2+ ! 21 'it>xb2 l:xd7 Simple chess. Now everything has
22 'W'e3 l:hd8 23 a3 l:d4 24 g3 been achieved. 1 5 .tb2?! tbc:S! �6 tbxe5
*c5 25 .l:.e1 .l:.c4! 26 *xc5 :xc5 27 .txe5 17 h3 l:fe8 would give Black
l:e2 l:d 1 ! 28 �f4 l:b5+ 29 �a2 equality.
l:tbb1 30 .th5 l:a 1 + 31 �b3 l:db1 + 1 5 . .. lbe5 1 6 lbh4 lUeS 1 7 i.b2
32 �c4 l:txa3 33 ..i.xf7 :xg3 34 *d7 1 8 l:te2
i.xe6+ 1;c7 35 lbd3 a5 36 l:tf2 White is slightly better.
l:e3! 37 l:tf7+ �b8! 38 l:te7 l:te4+ 1 8 . . .l:tad8 1 9 :ae1 lbxc4 20 bxc4
39 q;,c3 l:tb5 0-1 i.c8 21 lbd1 .:xe2 22 .:txe2 lbh5 23
l:te41 .:es 24 �e3 lbe4 25 1ixe4 · ·

Exercise 83: White to move *e7 26 'W'xe7 .txe7 27 lbhf5 .tf8

Lautler-Karpov 28 g4 lbf4 29 .i.e5 �h3+ 30 1;g2
Groningen 1 995 lbg5 31 iLc7 iLd7 32 .txb6 .i.xa4
33 d6 lbe4 34 lbe 7 + .txe 7 35 dxe 7
f6 36 f3 lbd6 37 i..xc5 lbe8 38 lbf5

Exercise 84: White to move

Dortmund 1993

This is a case of pure prophylaxis.

Black wants to play ...tbe5 and ex­
change a misplaced knight. In cases like
this, with White having a space advan­
tage, it is quite natural for him to avoid
Another thing is weaknesses. In this
position the weakest point in Black's Tills position holds one major ques-

1 58
Solu tions to Ex ercis e s

cion for White. Should he play lLleS di­ 1 -0

rectly or does he want to exchange
queens first? Well, the b7-bishop is per­ Exercise 85: White to move
fect on the long diagonal and White Karpov-Adams

really has no alternative but to eliminate Dos Hermanas 1993

it. Moreover 'White is happy to do so
despite the location of his own bishop,
as the c6-square is the potentially weak­
est link in Black's position.
White's advantages all support an ex­
change of queens - a distant passed
pawn, the weakness of c6 and the
domination of the d-file.
1 9 •e5!
19 lbe5 .txg2 20 'iPxg2 :fd8 21 :ac11
h6 is also better for White, but the with
the queens still on Black can create coun­
terplay against the king, something which W'henever we see an unprotected
is impossible after the text. piece we must keep our eyes peeled be­
1 9 . . ...e7! cause this is one of the most important
The endgame after 1 9...'ii'xe5 20 ingredients of a combination. Here
lLlxeS . .i.xg2 21 �xg2 offers Black no White has a very tempting move that
prospects of counterplay, with White should be seriously considered before
having the superior forces and a passed the more prudent alternatives and, since
pawn. it resulted in a close to winning advan­
20 a3 llbc8 21 Aad 1 J:.c7 tage, Karpov executed it.
More natural is 2 1 ...llfd8 22 !lxd8+ 21 o!bd51
!lxd8 23 llxd8+ Wxd8 24 ltJd4 when A simple launch, but very logical.
White is only slightly better. White's pieces are ideally placed - unlike
22 ll'lh4 Afc8? Black - so the fact that decisive action is
22 ... .i.xg2 23 ltJxg2 l'lfc8 24 ltle3 is in the air is not surprising.
correct, with an edge for White, but 21 . . ..txd5
now White has a practically winning Forced. 21...lLlxd5? 22 cxdS ..txgs 23
combination. dx:c6 .i.xd2 24 cxb7 is obvious. The
23 ll'lf5!! exf5 24 •xe7 Axe7 25 same goes for 21...exd5 22 ltxe7 bxc4
.txb7 l:.fS 26 .ta6! g6 27 c5 bxc5 23 .i.xf6 gxf6 24 'ii'h6, or something
28 bxc5 llc7 29 llc1 lieS 30 c6 1 along those lines.
'it>g7 31 'iitf1 l:te5 32 i.b7 l:ta5 33 22 cxd5!
Acd 1 ! l:le5 34 Ad7! l:.e7 35 :Xc7 22 hf6 .i.xf6 23 cxdS looks better
Axc7 36 AdS :te7 37 f3 g5 38 .:as for White but there is no reason to enter
g4 39 fxg4 fxg4 40 Axa 7 ll'ld5 41 into this exchange. It is simply a matter
J:.a4 ll'le3+ 42 �g1 <it>f6 43 llf4+ of calculation.

1 59
Ex c e lling a t Positional Ch ess

22 . . .lbxd5 ous once it is seen. ..

22.. ...xd5? 23
. ...xdS lDxdS 24 llxdS 25 . . .Wt8!
and White wins. Black protects the weak dark squares
23 .te4! .i.xg5 around the king and puts pressure- on
Again the only move. After 23 ..i.b4?
.. a3. The situation is highly unclear.
24 Wxb4 and 23. .:es 24 .txe7 llxe7
. 26 a4
25 i.xdS White wins. Around here White might possibly
24 Wxg5 Wb6 25 .i.xd5 exd5 26 have a better way to play, but this in no
J:e7! Wg6+ 27 'iixg6 hxg6 28 :xd5 way alters the correctness of Black's
White has an obvious advantage. treatment.
28 . . .lla7 29 :td6 �h7 30 :te4 :tc7 26 . . . lba3! 27 .ixa3
31 :txa6 d5 32 .:h4+ �98 33 :d4 27 hS i.c2! is good for Black.
�7 34 J:f4+ �g8 35 l:d4 �f7 36 27 ....txa3 28 l:g3! r h5 29 lLle3
:be J:c4 37 .:d2 b4 38 b3 llf4 39 .ib2! 30 lla2 .ic1 1
�c2 g5 40 h3 J:d7 41 �d 1 1 -0 The infiltration is complete and Black
has the advantage.
Exercise 86: Black to move 31 Wg5 1ib4 32 lLlf4? .ixe3 33
Shirov-Karpov fxe3 J:c1 ! 34 l!Jxg6 .l:.xd1 + 35 �h2
Las Palmas 1 994 fxg6 36 'iixg6 1ke7 37 :f2 'iixh4+
38 llh3 'iixf2 0-1

Exercise 87: Black to move

Germany 1994

With this posrtlon we have gone

from the technical to the more com­
plex. It is, of course, Alexei Shirov who
has created this mess with the white
pieces. White is threatening to attack
the enemy king, simultaneously trying to One of the things that I find my pu­
keep the queenside closed from a possi­ pils have a . problem with is fluent de­
ble invasion. But Black has a simple velopment. Then there is the significant
move that helps on both sides of the problem of how to identify important
board. Apparently this is not too easy to weaknesses and, subsequently, relate to
find, but should still appear quite obvi- them. In this position some pupils start

1 60
Solu tions to Ex ercis e s

with ... b7-b5 and ... d6-d5 in order to 1i'e5 :c1 1 46 1i't4 hxg3+ 47 'ii'xg3
challenge c4. But this is wrong as c4 is 1i'xg3 -i- 48 �xg3 g51 49 .ia6 l:lg1 +
the only significant weakness in the po­ 50 �h2 :tb1 51 i.f1 J:tb3 52 i.g2
sition, so we do not Vlish to allow White �g6 53 .txa4 txe4 54 �g2 b5 55
to trade it offl The appropriate course Ac2 �h5 56 �f2 �h4 57 J:tc5
of action involves finding a way to exert :tb2+ 58 'ifte1 b4 59 q,d 1 b3 60
extra pressure on this pawn, thus con­ Ab5 �xh3 61 :xg5 :tg2! 0-1
tinuing to keep White busy with unwel­
come attention for the rest of the (mid­ Exercise 88: White to move
dle)game. This also answers the basic Karpov-Polgar
question of how we get the f8-rook into Linares 1 994
16 ••. :c7!
Opening up several avenues through
which to hit c4. In the game Karpov
chooses to play ...:fc8 and ....i.a6 to
force b4-b5 and thereby secure the cS­
square for himself and retain c4 as a
weakness to target. But also possible is
something like ....i.b7-c8-e6, where the
bishop can take aim at c4.
1 7 1i'b2 'tfe6 1 8 a4 .ta6 1 9 b5 .tb7
20 a5
White is trying to generate counter­ For this exercise I was so happy with
play on the queenside and create a the solution from Ivo Timmermans that
weakness on b6, but this is obviously I decided to include it in the book (with
more difficult to achieve than for Black only a few comments):
to harass c4. 'White has a definite advantage. He
20 . . .h6 21 J:td1 :tfc8 has prevented castling and is ready, after
Bringing the rook into play and the preparatory 'it>ht , to attack in the
changing the pawn structure in his fa­ centre Vlith £2-f4-f5. The bishop on a3
vour. He has obviously improved his shines through the black position.
position. There is still the additional issue of de­
22 h3 lt:ld7 23 lt:\d2 lt:\ec5! 24 axb6 veloping the queenside pieces , though.
axb6 25 f3 e4! 26 fxe4 lt:\xe4 27 However, the at -rook is already on an
lt:\xe4 i.xe4 28 .ig3 ii'g6 29 �h2 open file and the knight on b1 has op­
d5! 30 J:tac1 lt:lt6 31 .i.t1 :c5 32 portunities to go to a3, c3 and d2. So
cxd5 lt:lxd5 33 l:lxc5 ltxc5 34 'W'd4 the question is: where should the queen
'ili>h7 35 .:td2 h5 36 l:lf2 f6 37 .te2 and king's rook go?
..'Llc3 38 .if1 lt:ld5 39 .ie2 f5 40 What can Black do in the meantime?
.i.e5 lt:lc3 41 .tn .!Dxb5 42 1i'b2 He cannot casde and should take care
..'Llc7 43 i.xc7 J:txc7 44 g3? h4! 45 of his a-pawn. When comparing pieces

16 1
Ex c e lling at Positio n a l C h e s s

we see that .1a3 versus .1e6 and lf.Jb 1 moves are easy to overlook because we
versus Ci:Je7 are strongly in White's fa­ feel the need to be more aggressive.
vour. There is counterplay only with White has a clear plus.
... c6-c5 and this should be prevented by 1 5 . . . h5 1 6 tbd2 f6 1 7 exf6 gxf6 1 8
White. For instance if White starts with .i.b4
1 �h1 , then 1 ...c5 comes, and perhaps White misses the chance for 1 8 'ti'xb6
. .%lc8 thereafter, and suddenly there is
. axb6 19 i..c S!! with the important little
life in Black's minor pieces. White detail of 1 9... .:tb8?! 20 .1xb6! - for
should play 1 'i'd4 to prevent ... c6-c5 some reason this is also easy to over­
and to go into a great endgame after look. Still, White is doing extremely
1 ...'i'xd4 2 cxd4; the pawns on a7 and well.
c6 are weak and llf1-c1 and lt1c3-a4 (or 1 8 .. .'ii>f7 1 9 l:ta41 Wxc5 20 .ixcS a6
simply lf.Jb1-d2, defending b3) are fan­ 21 f3 h4? ! 22 .:tfa1 .icB 23 �2
tastic manoeuvres. So Black should do .:tbB 24 b4 .:teB 2 5 �b3 .ib7 26
something else: l ...c!D£5 2 'iic S (2 �cS �a5 .ia8 27 .:t4a2 l:te6 28 .ia7
or 2 Wb-4, with the threat of g2-g4, also .:tb5 29 �b3 .ib7 30 g4 hxg3+ 31
look attractive on this Ifeel Ius certain ­
- hxg3 .:tea 32 g4 �d6 33 �d4! 1 -0
Aagaard) and 3 .1b4 still keeps the black
king in the middle and maintains the Exercise 89 : White to move
positional advantages on the dark Karpov-Topalov
squares. Finally, no solution is 1 ...'i'xb3 Linares 1 994
as White has more than one way to get
an attack, e .g. 2 i..xe7 �xe7 3 :xa7+
l:txa7 4 'i'xa7+ Wd8 5 Ci:Jd2 followed by
:bt , or (possibly stronger) 2 ireS 'ib7
3 lt1d2 and an invasion of rooks on the
b-ftle. In conclusion, White plays 1 'i'd4
and has a big advantage.'
I agree with lvo on more or less
everything above. And it is a good
chance to show a sample of the solu­
tions people handed in. The method of
deduction, the solution and the lines are
all powerfully explained. White has numerous ways to secure
1 4 'ili'd4! ltJf5 an advantage. 1 1 'iid2, to recapture with
14 ......xb3? 1 5 i..xe7 �e7 16 'iicS+! the queen on f4, looks attractive, and so
leaves Black finished. does 1 1 .1e3, but none of them can
1 5 'ili'c5 1 1 compare- to the way Karpov decides to
This move is apparently hard to find alter the pawn structure!
for some people. I think this is the phe­ 1 1 e3!
nomenon I describe in Excelling at Chess Maintaining perfect control over the
called Unforcing Play. Such 'small' centre. As far as ideal squares are con-

1 62
Solu tions to Ex ercis e s

cerned, here it is the di sappearance of Exercise 90: Black to move

them . Now Black is unable to find any Georgiev-Karpov
good squares for his pieces. 1 1 e4 tbxf4 Tilb urg 1 994
12 gxf4 isless good b ecause White s '

bish op is angry and Wlsighted and his

kingside has -for no reason - been
weakened on the dark squares. Black
might not be able to exploit this imme­
diately, but nor does he need to - this
weakness is permanent. Finally there
will be no attack on the g-file here as
Black has the bishop to protect the g7-
1 1 . . . ll:lxf4 1 2 exf4
Now Black has no easy way to de­
velop. In fact examination of the p osi­ It is time for action. White is threat­
tion indicates that there is no difficult ening to take on cS and Black is on the
way, either! wrong side of an unp leasant pin on the
h4-d8 diagonal. The d3-bishop looks
quite a lot better than Black's, while
White's knight looks silly. White would
like to increas e the pressure on Black's
centre, and has only two pro blems of
his own - the weakness of the d4-
square and the poorly placed knight.
This Black can use to his advantage.
1 3 . . . cxd4?1
1 3... g5! 14 .ig3 cxd4 is the most ac-
curate move order from a practical
point of view since in the game White
1 2 . . . i.d7 1 3 'ii'd2 ii'b8 1 4 l:tfe1 1 g6 has the extra possibility of 1 5 'it'd.
1 5 h4 a6 16 h5 b5 1 7 hxg6 hxg6 Quite messy is 13 ... :Ic8 1 4 dxcS d4 1 5
1 8 tLlc5! dxc5 1 9 "ii'xd7 l:c8 20 'ifc2, when i t i s debatable that Black has
.:txe6! Aa7 21 Axg6+ fxg6 22 made any progress. Rather he has
1i'e6+ �g7 23 .ixc6 l:td8 24 cxb5 opened the position for the white
i.f6 25 �e4 i.d4 26 bxa6 'ib6 27 pieces. 1 3 ... c4!? 1 4 J.c2 i.c8 15 i.xf6
.:td1 'ifxa6 28 Axd4 :Xd4 29 'ii'f6+ 'i'xf6 16 lLlf4 l4d8, with approximate
�g8 30 'W'xg6+ �f8 31 'it'e8+ �g7 equality, is also possible.
32 'ife5+ 'it>g8 33 tl::!f6+ ..tf7 34 1 4 exd4 g5!
.ie8+ �fa 35 'i'xc5+ 'it'd& 36 'ifxa7 1 4. .tt:le4? 15 i.xd8 tt:lxc3 creates a

Vxf6 37 i.h5 l:td2 38 b3 ltb2 39 desperado siruation in which more than

�g2 1 -0 one piece is hanging, usually resulting in

1 63
Ex c e lling a t Positio n a l Ch e s s

being sold at the maximum price - 1 6 lbf4 lbxd4 20 �hS is to be avoided.

.i.xb6! and White has the advantage. 1 6 .l:te 1
16 •d1 C£Jg7 1 7 .1g3 f5 1 8 .ies
CiJxeS 19 dxeS l:tc8 20 f4 g4 21 lb£2 d4
is better for White according to Fri�
and clearly better for Black according to
Karpov. As I see it the truth is some­
where in between. I prefer Black too,
but how much? Hard to judge.
1 6 ... 'i'f6
1 6...li:Jxd4 1 7 'Wd1 lbg7 1 8 .i.g3
lbdfS 19 .i.eS gives White some com­
pensation in the form of active bishops
and positional targets in Black's weak
1 5 'ifc 1 ! dark squares.
White should do something - any­ 1 7 tLlxg5 hxg5 1 8 .txg5 'i'xd4 1 9
thing - as after 1 5 .i.g3 �e4 1 6 .i.xe4 .l:te3 f5! 20 :113 lLle51 21 .tc2 lllg 7
dxe4 Black looks better. For example 1 7 22 ..ie3 'i'c4 23 .th6 .l:tf7 24 'i'd2
dS!? (probably the best move, a s 1 7 :tea 25 .tb3 'i'c6 26 .l:tg3 .:tcc7 27
l:tad1 'ii'dS [1 7...�e7!? also looks natu­ .l:te1 �g4 28 .tf4 llce7 29 Ac1 _,5
ral] 1 8 f3 l:.ae8 gives Black a good game 30 h3 lllf6 31 .tc2 'i'e2 32 'i'd4
- the opening of the long diagonal for 'W'c4 33 •xc4 dxc4 34 ..ig5 b5 35
his bishop is on its way, as it always b3 f4! 36 llc3 lL'ld5 37 .l:tf3 lle5 38
would be) 1 7 .. .'ii'xd5 1 8 llad1 'ii'f5 1 9 h4 �e6 39 bxc4 bxc4 40 .tg6 llg7
:d6 l:tac8 20 :Xh6 f6 2 1 ..td6 :n 0-1
(21 ...�g7? does not work in view of 22
l:.hS! 'iPg6? 23 �f4+! arid White wins) Exercise 91 : Black to move
22 f4 .1a6! (the position is messy and Sendergaard·Aagaard
both players have problems to address Simultaneous Display,
here, but I have a feeling that White's Denmark 2002
position will crack - despite the fact
that Fritz is less certain) 23 :te 1 �eS 24
._.g3 tfri3 25 fxgS fxgS 26 1lh5 Wf6 27
l:hgS+ :g7 28 l:lb1 'ifxb2 is the result
of a quick discussion with Fri� and
clearly White only just hanging on here.
1 5 . ..�h5
1 5...�e4!? 16 f3 �xd4 17 'ii'dl €\£5
also looks better for Black, but not
much. But 15 ...gxh4?? 16 'ii'xh6 'ii'd6
(1 6 ... :te8 1 7 l!Jg5 :te4 18 f3! and White
wins) 1 7 .1h7+ �h8 1 8 i.£5+ 'iPg8 1 9

1 64
Solu tions to Ex ercis e s

This is a trap. Sorry. Sometimes you lZJf6 is definitely not the same anymore.
should test people's awareness of tac­ 1 5 . . .i.xc3 1 6 bxc3
tics. Black wins the exchange by choos­ 1 6 .txc3 lZJf4.
ing the accurate move order. 1 6 . . .lDf6
1 . . . i.xd41 2 i.xd4 lDb3! Black is simply a lot better here. The
Look at how many white pieces are damage to the White king's safety is
hanging (in one way or another). White permanent.
could limit the loss with l:!.Sxb3, but the 1 1 c4?! 'ii'b6! 1 8 f4 lLled7 1 9 i.d3
game is still lost. lLlc5 20 e5 lDa4 21 i.b4 dxe5 22 c5
3 cxb3? 'i"xb5 4 i.e3 .:tc2 5 'W"d 1 'ii'c7 23 .i.c4 a5 0-1
.:a2 6 .:c 1 :xc1 7 'ii'xc1 'ii'e2 8
'ii'c8+ �g7 9 i.d4+ f6 1 0 'ii'h3 'ii'e1 Exercise 93: White to move
mate Karpov-Gelfand
Sanghi Nagar 1995
Exercise 92: Black to move
Dos Hennanas 1995

This is a really difficult exercise, so

don't be disappointed if you failed to
get to grips with it. Black has a good
Again this is a question of tactical set-up but it is of a rather defensive na­
awareness. Black has a forcing advan­ ture. White has potential weaknesses in
tage with... bZ and d4, but it is difficult to imagine
1 4 . . . .i.xh3! any way that these should come under
14 ...i.xc3 15 .txc3 .txh3 is a faulty serious fi.te. Black has a weakness on b6
move order due to 1 6 g4! .txg4 17 fxg4 (and perhaps e6) but this is also hardly
lZJf6 18 ...gZ and the two bishops and enough to worry about. So White

the open files compensate for the pawn. should tty to create further weaknesses
1 5 'ii'xh3? in Black's camp if he is to generate an
15 gxh3 'iVxh4 16 :hg1 h6 17 �b1 is advantage. This is best done with the
much better. Black is better after 17 ... b5 pawn advance h2-h4-h5 (g3-g4-g5
18 i.e! 'iVf6 19 lZJd5 cxdS 20 .txaS but weakens the dark squares around the
still it is a mess. 1 5 g4? i.xg4 16 fxg4 white king just as much as those around

1 65
Ex c elling a t Positio n a l Ch ess

its opposite number, thus achieving 28 . . .ltJeB? ! 29 tLle3 tLlg7 30 tLlc3!

nothing). But this is not the best option tLlxc3 31 J:txc3 g5!? 32 hxg5 fxg5
in this position just yet. There is a great 33 lZJg4 gxf4 34 gxf4 i.d6 35 .:tf3!
improvement to be made before. i.eS 36 tLle5 tLlf5?! 37 d5! f i.xe5
As is the case in so many situations it 38 fxe5 l::tb 7 39 �h2! :l.g7 40 i.h3
is important to improve your worst i.h5 41 l:f4 �h8? 42 i.xf5! exf5
placed piece, or the piece that can be 43 i.h4 'W'eS 44 i.f6 i.g4 45
best improved. In this respect it would i.xg7+ �xg7 46 J:tc7 + �hB 47 e6!
be really nice to be able to do some­ 'W'h5+ 48 �g 1 ..g5 49 �2! 'i'h4+
thing for the f2-bishop, but right now it 50 �e3 1 -0
does a valuable job from £2, and no
ideal square is apparent. But what about Exercise 94: White to move
the el-rook? The e4-knight is not going Salov-Lautier
to move, and the e6-pawn is hardly Wijk aan Zee 1991
weak. The following manoeuvre is
therefore both logical and necessary, for
an attack on Black's king cannot be se­
riously considered before White has
mobilised all his forces to their opti­
25 :l.e2!
The rook is brought to the c-fJ.le
where it will do much more good.
White has time to do this as Black has
no obvious improvements to make.

This position is not so much posi­

tional as it is tactical, although it is
grounded in positional evaluation.
Clearly the best move is:
22 .:td6!
Here is the reasoning. The tactical
exchange that now follows is more or
less forced and we reach an endgame
with an outside passed pawn and a
weakness on e6. This is the principle of
two weaknesses, which in most cases is
25 . . . i.e7 26 :ec2 l:b8 27 'W'd2 l:tf8 enough-to win an endgame. If you had
28 h4 problems evaluating the end of this line
Look at the position from a few I suggest that you either search your
moves ago and compare h4 in that posi­ database for en dgame s with outside
tion with h4 in this. White is better. passed pawns and see how they work,

1 66
Solu tions to Ex ercis e s

or play the position a few times with often as it does in quiet positions, but it
Fritz with both colours. is still wonh investigating if the move
22 .1:lxd6
••• you want to play for positional reasons
This is forced due to 22 ...ltJ£8 23 works tactically. Here it does. White is
lUgS! and all of :Xd8, 'ilxf7, lUxf7 and all set up for b2-b4-bS but cannot pro­
.td7 are threatened. White wins. tect his c-pawn with moves other than
23 cxd6 'i'xd6 24 .i.xe6 'ifxe6 b2-b3. So it is now or never, as this will
After 24 ...fxe6 2S .txa7 White is also not change.
threatening ltJgS, and the inclusion of 1 2 b4!
the queens in the position in no way 12 fue6 'ilxe6 13 b3 Aad8 14 .ia3
offers Black serious hopes of a perpet­ :res 1 S 'W'c2 lUc7 1 6 :bet dS was
ual check. equal in Horvath-Santo Roman, Novi
25 'ifxe6 fxe6 26 .txa7 !Df6 27 a4 Sad Olympiad 1 990.
!Dd7 1 2 . . ...txc4?!
Salov gives the following, very con­ A very risky decision. It is not diffi­
vincing line: 27...lDxe4 28 aS lLlc3 29 a6 cult to understand what Black was
liJdS 30 ..tcS �c7 31 a7 and White is thinking: if White is allowed to execute
winning (e.g. 31 ...e4 32 lDgS) . his plan at least I should bag a pawn.
28 a5 ..tf8 29 a6 c5 30 !Dd2 �f7 Instead 12 ...l2Je4 13 lbxe4 fxe4 14
3 1 lDc4 1 -0 lDxe6 ..xe6 l S 1i'c2 l'lfe8 16 .ie3 lDc7
1 7 :rdt leads to a comfortable advan­
Exercise 95: White to move tage for White (equal acc?rding to Fri�
Salov-Gurevich which is rather irrelevant). However,
Reggio Emilia 1992 this seems to be the best option.
1 3 b5! cxb5 1 4 lildxb5 l:.d8 1 5 .ta3
Here White has a superior way to
prove his advantage in l S tt:lxd6! ..e6
1 6 .ia3 lL!e4 1 7 �cxe4 fxe4 1 8 'I'c2!,
when White is winning according to
Salov, who writes the following in his
annotations: 'In the post-mortem to­
gether with Mikhail we discovered the
following beautiful line: 1 8... .idS 1 9
lDe4 :res 2 0 liJgS! 'ile2 21 .i.d5+ lidS
22 l:tbell! 'ilet 23 l:et l:tel 24 <ltg2
llg5 25 'ili'c4+! �h8 26 'it'c8+ and mate
This position comes under the 90% in two moves.'
rule. Remember that the move you want 1 5 d5! 1 6 'Dd61 1fe5 1 7 !Dxc4

to play for positional reasons normally dxc4 1 8 'W'c2 !Dc5 1 9 i.xc5??

has a 90% chance of working tactically. Very bad and very sad. After 19 lDa4!
Of course that does not count for sacri­ b6 20 tt:lxcS bxc5 21 'i'xc4+ White
fices and other brutal approaches as wins.

Ex c e lling a t Positio n a l Chess

19 .. .'it'xc5 20 ltb5 'ii'd6! 21 l:txb7 better game, when it is not easy to see
�h81 22 l0b5? 'ifc5! 23 a4 aS 24 how White can improve his position,
�c3 �g4 25 h3 l0e3 26 fxe3 whereas Black has a lot of moves com­
'ii'xe3+ 27 �h2 f4?? ing.
Life is tragic. 1 8 h4! c5 1 9 dxc5 bxc5 20 'ii'xc5
28 l:tf3! fxg3+ 29 lbg3 .te5 30 'ii'xc5 21 l:txc5 i.xg4 22 h5 g5 23
l:txh7 + 1 1 -0 i.f5! i.xe2 24 �xe2 ..txb2 25 l:txd5
l:te5 26 l:thd1 a4 27 �g3 l:ta5 28
Exercise 96: Black to move l:txa5 l:txa5 29 i.c21 �e6 30 l:td7
Salov-Van dar Sterren i.g7 31 l:l.b7 i.f8 32 �3 �g7? 33
Hilversum 1 993 i.g6 �d8 34 l:td7 �g8 35 l:txd8
fxg6 36 hxg6 �g7 37 lt>g4! i.a3!
38 �h5 1 g4+ 39 �xg4 �xg6 40
l:tg8+ �f7 4 1 l:tc8! l:tg5+ 42 �f3
i.f8 43 l:c7+ �e6 44 l:l.c6+ Wf7 45
e4 l:l.b5 46 lLlf5 llb2 4 7 e5 1 -0

Exercise 97: White to move

Wijk aan Zee 1997

In this position White has little con­

trol over the queenside - an area in
which he normally places his hopes in
the Queen's Gambit Declined. A trans­
fer of his forces to the queenside seri­
ously compromises the kingside, where
g2-g4 has created certain commitments.
For this reason White will have to play
on the kingside, and as Black has a
dark-squared bishop it will most likely White has a lead in development and
be on the light squares. All of this basic he cannot use it to create an attack on
logic points us in the direction of the the kingside (the pieces are not headed
threat in the position. White will play 18 that way, and Black has no obvious
h4! and it will give him the advantage. weaknesses). Consequently the correct
Black should prevent this, after which approach is to put immediate pressure
he will have time to slowly improve his on the centre, which is in fact Black's
position. only weakness. If not, then the lead in
1 7 . . . b6? development will soon be history, and
17 . .th4! would have given Black a
. . White will be left with no structural ad-

1 68
Solu tions to Ex ercis e s

vantage to compensate for the loss. Exercise 98: White to move

1 1 dxe5 dxe5 1 2 'Wc3! Salov-Karpov
The point. Now Black has problems Hoogovens 1998
with both the dark squares and co­
ordination. 12 ..tbS c6 does not seem to
achieve anything as the knight is far
from d6.
1 2 . . .li:lc6
After 12 ...e4 13 lDd4 Black still has
problems with his development as c7 is
hanging, and thus White has a small
advantage. White wins a pawn after
1 2 ... l:te8? 1 3 'ifxeS! etc.
1 3 .ib5 l:l.e8!
Black cannot prevent the doubled
pawn for after 1 3 ... ..td7? 14 ..txc6 One of the important things to do
ii.xc6 1 5 ltJxeS i.xg2 1 6 l:thg1 ..te4 1 7 before formulating a plan is to take all
lDc6 White wins. aspects of the position into considera­
1 4 l:l.d2? tion. One of these is spotting weak­
This is silly . After 14 1l.xc6 bxc6 15 nesses. Here Black has a clear weakness
lhd2 White has a structural advantage. on a6, which might be very hard to de­
But not 1 5 'i!i'xc6 l:.b8 and Black has fend. Black also has a problem with his
some compensation. development, and \Vhite would love to
1 4 . . . li:lb4! prevent ...ii.e7 followed by ...0-0 (at
Perhaps this is the move that White least for the time being). It is often the
missed. case that a move that addresses some of
1 5 �b1 c6 1 6 a3 cxb5?! the immediate considerations in a posi­
Instead of this careless capture, the tion also serves other, deeper objectives.
move 16 ...a5! is a considerable im­ Here the attack on a6 happens to also
provement. prevent ... i.e7 .

1 7 ._xb4 'ii'xb4 1 8 axb4 f6! 1 9 1 5 .l:!.e3!

l:hd1 i.e6 20 b3 ritf7 2 1 e4 l:ac8?! The most obvious way to attack a6,
22 'itrb2 l:l.c 7 23 lLJe 1 I �e 7 24 lbd3 and it turns out that Black cannot de­
g5 25 h41 gxh4 26 l:l.h 1 h3 27 f4! velop freely.
.l:!.g81? 28 gxh3 l:td7 1 ? 29 fxe5 fxe5 1 5 . . . lLld7
30 Ae2 l:l.d4 31 lLlxe5 Axb4 32 l:l.a 1 1 If 1 S. ii.e7? 1 6 :b3 'ii'c7 1 7 :xb8+

.l:!.g3 33 l:l.xa7 �d61 34 lbd3 l:l.xb3+ 'i'xb8 Black has a loose piece on c6 and
35 cxb3 llxd3 36 b4 l:l.b3+ 37 �c2 White wins after 1 8 e5! i.xg2 19 exf6.
l:txh3? 38 llxb7 .ib3+ 39 �d2 i.c4 1 6 l:tb3 'i'a7 1 7 l:l.ba3 i.b7 1 8 b4
40 l:l.e3 Ah2+ 41 �c3+- l:th1 42 'W'b6 1 9 'ii'a4!
e5 + �c6 43 l:l.a7 l:l.b1 44 Aa6+ Black has a more or less permanent
'it>b7 45 l:l.xh6 1 -0 problem with his development. Now

1 69
Ex c e lling a t Positio n a l C h ess

the knight on d7 is in trouble. 1 .!t:Jxa 7! .l:a6

1 9 . . ..ie7 20 .i.b2 .i.f&?
20 ... e5 is a disgusting move, but
might still be the best option.
21 e51
Sealing Black's fate.
21 . . . .ixe5 22 c5 'ilc7 23 Lxe5
dxe5 24 .i.xb7 J:lxb7 25 c6 lDb6 26
'ilxa& 'ifxc& 27 lDc4!
The piece cannot be saved.
27 . . . l:.bS 28 l0xb6 0-0 29 .!t:Jc4 'ife4
30 l:te3 'ifd5 31 .!Dxe5 AfcS 32 'ifd3
1 -0
2 b411
Exercise 99: White to move This move has to be found - it will
Lund-Husted not appear from nowhere. __ _ ..

Denmark 2002 2 . . ..:txa7 3 lDb5 .l:a6 4 l0c7 + �d& 5

lDxa6 bxa6 6 �c41 �c6
6.. e4 7 b5 and White wins.

7 f3
Followed by b4-b5 with a winning
pawn endgame.
7 . . .�d6 8 b5 axb5+ 9 �xb5 �c7
1 0 �c5 �b7 1 1 �d5 �a& 1 2 �xe5
�xa5 1 3 �f6 and so on.

Exercise 1 00: White to move

Skellefta 1989

When calculating it s useful to end a
line like you started - looking for extra
possibilities. Obviously the first move
that comes to mind is seldom the only
one in the position. It might be the best
in 50% of the cases, but just following
your first idea will then result in making
the correct choice only 50% of the time!
Here it would not do you any good.
White played 1 lLle4 and eventually lost,
although the position is, of course, play­
able for him. But he could have won in
an instant. .. Apparently White has a strong posi-

1 70
Solu tions to Ex ercis e s

cion. He has an influential knight on eS 21 . . ..l:.fe8

and the queen seems well posted on f4, 21 ...ll:ld7!? 22 ll:lxd7 :Xd7 23 'ifxc7
while both target f7, one of the poten­ Lc7 gives White a better endgame due
tially weak squares in the black camp. to the control of the open £ile.
Additionally, Black has moved his b­ 22 h4 �h7 23 �g2 g5 24 hxg5
pawn, which creates a weakness on c6. hxg5 25 'ih12 f6 26 �f3 'it'xh2+ 27
But then when you scrape at the surface �xh2 <t.>f7 28 �g3 lDf8 29 a3
a little you realise that Black is about to �e7?! 30 lbd2 tt:lg6 31 lt:le4 lbd1
exchange rooks on the d-£ile, after 32 .l:.xd 1 :hs 33 b4! lt:le51 34 bxc5
which he might gain control over what lt:lxc4 35 .l:.c1 b51 36 a4 lt:la3?
is the only open £ile on the board. So is 36 ...a6! 37 axbS axbS 38 .:at lld8
the first assessment wrong? When I try would have kept the position together.
to solve such positions I normally trust Here there is a little bonus exercise:
my initial evaluation 100%. If I cannot
find a move that supports this evalua­ Exercise 1 01 : White to move
tion I might take a practical decision Salov-Ehlvest
but, mainly, I try to find something that Skellefta 1 989
backs up my evaluation. Here Salov
found something for me.
21 g4!
Black's weakest square is fl. There is
nothing else that "White can seriously
expect to attack. Note that problem of
mate on d1 has also been addressed
with this thrust. Thus, overall, White
retains his advantages and solves his
problems, which is enough for an edge.
Instead 21 tlJg6? l:lxd1+! 22 :Xd1 l:td8!
is horrible (now something like 23 l:te 1
is forced ... urgh!). Now it is all about the control of c8.
If White just advances the c-pawn it is
bound to be tamed in the near future,
when Black will be able to generate
counterplay on the queenside or simply
make a draw as all the pawns get vac­
uumed off the board. With the text
White takes command of c8 and simply
pushes his pawn to the 8th rank.
37 lld1 ! 1 f5
What else? 37 bxa4 38 c6 l:d8 39

llh1 tL!bS 40 :h7+ �f8 41 c7 and wins,

or 37 ...lidS 38 J:h 1 eS 39 llh7+ �e6 4D

Ex celling a t Po s i tio n a l C h e s s

lha7 and Black is in big trouble. Finally knight manoeuvre to e4) 24...g6 25 lt:Jd2
after 37 ... l:tc8 38 llh1 ! there is, appar­ with a clear advantage to White, whose
ently, no decent continuation for Black. pieces have excellent squares at their
38 gxf5 exf5 39 lbd6 f4+ 40 exf4 disposal, while Black is left trying to
gxf4+ 41 �g41 keep his game together.
Preventing counterplay, which might 22 g3 i.f5
result from 41 �xf4 l:th4+ 42 <itfg5 Black is forced to give up the pawn
l:txa4 43 c6 'iti>d8!, when Black has rea­ now. 22... fxg3+ 23 l:txg3 .:th6 24 i.c4
sonable chances of saving the game. l:.f4 25 i.xe6 .:xh4 26 i.xd7 �xd7 27
41 . . . bxa4 42 c6 lLlc2 43 c7 �e6 .:cgl and White wins, while 22...:th6 23
43 ... a3 44 lt:Jf5+. i.c4!? looks similar.
44 �8! AxeS 45 l:d8 1 -0 23 gxf4 l:l.f7 24 c4 �b8 25 d5!
cxd5 26 cxd5 ..'Db6
Exercise 1 02: White to move After 26 ... exd5 White's forces enjoy
Kramnik-Hertneck new found freedom with 27 lt:Jd4 lt:Jb6
Germany 1 995 (27...lt:Jf8 28 lt:JxfS l:.xfS 29 i.h3, or
27 ... l:.e8 28 lt:JxfS .:r.xf5 29 i.h3 etc.) 28
e6 l:f6 29 :Xg7 i.xe6 30 lt:Jxe6 l:.xe6
31 :.cc7 and Black is busted.
27 d6
White has a decisive positional ad­
27 ..'Dd5 28 ..'Dd4 %thf8 29 �g3!

.tg4 30 fxg4 :xf4 31 tbxe6 l:txg4+

32 �h2 l:xh4+ 33 i.h3 1 -0

Exercise 1 03: Black to move

Perhaps this is difficult to solve, but Horgen 1995
it is easy to explain. \Xlh.ite has a very
poor piece on h 1 and needs to activate
it. Ideally White would like an open file,
which does not exist at the moment. If
only the g-file were open - then we
could target g7...
21 %tg 1 1 .:ct8?!
2t ....i.f5 22 g3 fxg3+ 23 .:txg3 l:tcg8
(23 ...g6 24 i.h3 leaves g6 seriously
weakened) 24 .i.e2! (24 i.h3 allows the
unnecessary 24... .ixh3 25 l:.xh3 g5,
although White is probably better after
26 hxg5 l:txg5 27 .:ch 1 followed by a Here Black can play 17 ...lt:Jf6 with an

1 72
Solu tions to Ex ercis e s

equal position. But the pawn sacrifice is still not without problems for W'hite,
far more interesting. but a draw is the most likely result.
1 7 . . . b3! 29 . . . l:txc2 30 Axc2 l:lb 1 + 31 tt::lt1
With this move Black creates several 1i'b6
weaknesses in the white camp, namely Now White is severely tangled up.
b2, b3 and d3. Both b-pawns are likely 32 h3 tt::lc6 33 •c4 tt::ld4 34 l:la2 h5
to fall eventually and, with the bishop 35 h4 �h7 36 �h2 l:tb4 37 Wc1
on a6, he will be able to put White un­ lt:lc6! 38 tt::le3 .ih6?
der severe pressure. 38 ....!L\e5 with the idea of ... J..h6 was
1 8 cxb3 tt::lge5 much stronger.
This is the position Black was head­ 39 f4 l:tb3 40 l:ta31 l:tb1 41 l:.a6!
ing for. If you look for the ideal squares •xa6? 42 'W'xb 1 -.e2 43 tt::lt1 ! tt::ld4
for the white pieces you will not really 44 'Wb7! �g8 45 'i'c8+ .itS 46
find any. There are no weak squares in •c3 .ig7 47 We3 Wb2 48 Wd2 'i'b3
Black's camp besides d6, and there is no 49 tt::le3 % -%
convenient way for W'hite to get his
pieces to attack d6. Exercise 1 04: White to move
1 9 .id21 Kramnik-Vaganian
lvanchuk decides to improve his Horgen 1 995
most inactive piece, which is the right
course of action in a position under
pressure like this. An important detail is
that 19 f4?! meets with 19 ... lLlb41 20
l:l.a3 lLled3 with the beginning of an
invasion. 19 l:a3 'iib 6! 20 lbc4 lLlxc4
21 bxc4 .i.xb2 looks slightly better for
Black, while 1 9 lbc4 .!Lib4 20 .:a3 .!Lied3
followed by ... d6-d5 will deprive White
of his only good square for a kfught.
1 9 . . . 'i'b6 20 .ic3 .ia6 21 'i'd2
.ixe21 22 'i'xe2 tt::ld4 23 .ixd4
Forced. 23 'i'd l ?! 'i'xb3 24 'i'xb3 Here W'hite could play 15 b4 with
lLle2+! 25 �h 1 :xb3 is clearly better for some advantage, but he has the oppor­
Black. tunity to develop a serious initiative
23 . . .'i'xd4 24 .l:d1 'i'c5 25 l:la6? thanks to his lead in development. This,
25 'ii'dZ! llfd8 26 .:a6 ltxb3 27 .:Xd6 together with the great squares he can
l:Xd6 28 'ifxd6 'ifxd6 29 l::txd6 was the get for his pieces, as well as all the
safe path to a draw. weaknesses in the black camp...
25 . . .l:.xb3 26 l:.axd6 l:txb2 27 l:l6d2 1 5 c5! bxc5
l:lfb8! 28 l:lc21 l:t8b31 ? 29 l:ldd2? 1S ... J..c 7 16 cxb6 J..xb6 17 tlk4 is
29 :XeS! :xe2 30 .:c2 .:Xc2 · clearly better for White.
(30 ... l:lbxe3 31 ltc8+) 31 lLlxc2 .i.£8!? is 1 6 �c4 We7

1 73
Ex c e lling a t Positional Ch e s s

1 6....i.e7 17 dxcS t0d7 1 8 b4 gives g3. The other bishop would be better
White a clear advantage. off on c2. White might consider the
1 7 dxc5 i.xc5 18 b4 i.d6 1 9 .i.b2 advance e3-e4 but, currently, all the
..tc7 20 l:tfd 1 black pieces are posted to prevent this,
White stands much better. Black and playing it would open up for them
cannot develop freely and White needs in the most self-destructive way.
only to play 21 :tact to dominate. 1 S ..td1 !
20 . . . c5 Intending a3-a4 after 1 8 ....i.c6 and
20...t0d7? 21 'i'xc6 tDb6 22 t0xb6 19 .. .t0b6. In this way Wbitc:_ _!ceeps his
.i.xb6 23 :d7 and White wins. queenside together while manoeuvring
21 bxc5 Wxc5 22 l:tac1 'fke7 23 on the kingside.
Wb5 i.b6 24 a4 1 8. . .lbb6 1 9 i.c2 i.c6 20 a4
24 .i.eS! is preferable. White stands better.
24. . . i.c5 25 .i.d4 i.xd4 26 :txd4 a6 20 . . .'i'd7?1 21 a5 tl)cS 22 ..th4!
27 'W'b6 l:ta7 2S tbd6! l:td7 29 l:tcS! lt:lh7 23 �f2 lbd6 24 �g4! �f5 25
:xeS 30 tbxc8 'W'a3 31 'W'xe6+ �fS .tf2 f6 26 h31 lOde 27 :te 1 :a�S
32 'W'xf5+ 'iteS 33 'ile6+ q,.d8 34 28 e4! �b5 29 'W'd2 h5- 30 liJe3
'ilb6+ ! ¢>e8 35 tbd6+ 1 -0 dxe4 31 d5 f5? 32 tbxf5! •xd5? 33
'W'xd5+ i.xd5 34 i.a4! a6 35 ..txb5
Exercise 1 05: White to move axb5 36 fxe41 i.c6 37 lL!d6 �f6 3S
Kramnik·Tiviakov a61? c3 39 a7 c2 40 e5 �e4 41
Las Vegas 1 999 �xeS :xeS 42 l:tec1 1 -0

Exercise 1 06: White to move

Dortmund 1999

This is a combination of the mis­

placed piece and prophylaxis. Black is
about to play ...t2Jb6 and . . ..i.c6 to put
these two pieces to good use. The idea
of . ..lt)a4 is obvious. White would like The situation in this position can be
to get all his pieces into play. His knight explained quite simply as being a matter
needs to go to £2 at some point, which of future structures, and how the minor
means that the bishop must go to h4 or pieces work with them. If Black is al-

1 74
Solu tions to Ex ercis e s

lowed to take on e4 he will secure both l:txa4 lbcB 37 J:.b4 liJa7 38 bxa7
f5 and dS for his knights and White will l:laS 39 c6 l:laxa7 40 ltc 1 1 -0
have three weak pawns on b4, d4 and f4
- in other words White will be worse. If Exercise 1 07: White to move
White decides to take on d5 to avoid Kramnik-Svldler
this he will still have to concede the £5- Linares 1 999
square (or weaken his king and the f­
pawn with g2-g4) and probably also the
e4-square (due to the dS-pawn), while .
the weakness of e3 becomes apparent.
With these alternatives it seems natural
to sacrifice a pawn to deprive Black of ·

all these squares and simultaneously

create a majority on the queenside,
1 6 f51 exf5 1 7 exd5 cxd5?1
This recapture leaves the e7-knight
rather poor, while the knight on c3 is
clearly perfectly placed. Black had no
choice but to play 1 7 ...lDxdS 1 8 �xd5 We arrive here via a razor-sharp line
(1 8 lLlc4 t£Jxc3 1 9 Axc3 0-0 20 �6 in the Griinfeld Defence in which
�f6 seems equal) 18 ... cxd5 19 'ii'b3 White sacrifices a pawn in order to gain
�f6 20 t£!£3 0-0 21 �eS, when White a lead in development, a potential
has good compensation for the pawn . passed pawn on the d-Ele, put the black
but Black does not have the space prob­ queen in trouble and, often, secure the
lems he experiences in the game. bishop pair. By now the opening has
1 8 b5 0-0 1 9 b6 'i'd8? ! just about come to an end and it is time
1 9... 'i'c8! is necessary in order to for White to find a natural way to get
keep an eye on the light squares on the his initiative rolling. For if Black is given
queenside, which is where White now the time he will consolidate his extra
gets a powerful attack. pawn, improve his co-ordination and •
20 li:lb3 maybe even win the game. The main
On the way to the ideal square. problem for White is that the fl -rook
20 .lbf6 21 lba5 l:tb8 22 a4 lbe4
•. and £3-bishop have little scope. The
23 .!Oa2! solution to this problem, then, is to try
White plans to win on the queenside, to push the d-pawn with a pawn sacri­
and it is not apparent how Black is go­ fice.
ing to stop him. 1 7 e5! lbc4
23 . . . f6 24 llJb4 i.eS 25 'Wc2 g6 26 Now White is a whole lot better. But
l:tb1 l:!.f7 27 lbxb7 l:!.xb7 28 a5 .!Oc6 1 7 . i.x�5 also looks dangerous, when

29 lbxc6 i.xc6 30 i.xa6 l1b8 31 one way to gain a clear advantage is 1 8

.tb5 •cs 32 i.xc6 -.xc6 33 a6 d6!? i.xd6 1 9 :a1 'i'c4 20 :XaS
�g7 34 l:tb4 lLld6 35 Wa4 '*'xa4 36 i.xh2+ 21 �hl and the active white

1 75
Ex c elling a t Positional Chess

pieces will pick off Black's '.�teal:: �1\S +o·the. pressure:.

and perhaps mount a kingsi(k ��.a. c.k. l\ bxc.3
. . .

1 8 d6 t2Jxe5 1 9 i.d5! •a3 20 .Sxe::} There was a possible alternative in

i.f8 21 i.xb7 i.xe7 22 he.� i.xa6 14 ... lbxd5 1 5 lL:lxdS 'ii'xdS 1 6 'i'xdS
23 i.d5 :e7 24 <i>h1 h5 25 h3 �g7 exdS 1 7 l:lac l , when White keeps the
26 Wd2 i.c7 27 l:bd 1 tbd7? 28 initiative without the · queens. Or here
i.xf71 :xn 29 Wxd7 :txd7 30 1 S ... exdS 1 6 l:lcl and White wins mate­
l:txd7+ �h6 31 :xc7 •d3 32 �g1 rial, one line being 1 6 ...:a7 1 7 l:xe7!
'1Wd4? 33 l:c2 1 -0 :.Xe7 18 'ii'xdS i.d7 1 9 �eS and Black
is busted. Alternatively, 1 4 ... exd5 15
Exercise 1 08: White to move lbe2 followed by tbed4 leaves White
Kramnlk-Karpov with long-tenn dynamic compensation
Frankfurt 1 999 as Black is uncoordinated.
1 5 d6 l'Dd5
The alternative 1S ...cxb2 might look
strong, but it all depends on finding the
right moves. This normally means that
you do not take any moves for granted,
but look for options on every move.
Here White is close to winning after 1 6
dxe7+ �xe7 17 'ifc2! etc. .
1 6 dxe7 + 'iVxe7 1 7 i.e51
Keeping up the pressure. It is not
easy for Black to free himself. 1 7 li'xdS?
exdS 1 8 .i.d6 ..i.e6! 1 9 .i.xe7+ �xe 7 20
White has put all his chips on the bxc3 :hc8 favours Black.
blood red colour of the initiative. When 1 7 . . . i.b7
you opt for a dynamic advantage over 1 7 . cxb2 1 8 i.xb2 lt::lc4 1 9 i.d4! with

the static (long-tenn) advantages that very strong compensation for the pawn.
your opponent is most likely accumulat­ In practice White will be a rook up for
ing along the way you cannot drop your about five to ten moves. If he plays
pace. Here White will only just keep an with enough vigour this should at some
equal game after 14 lbe4, but this move point be translated into a permanent
is irrelevant as there is another move advantage. 1 9 .i.xg7+?! �xg7 20 'ii'xdS
that should work - unless White really exdS 21 :Xe7, with a very slight advan­
is worse. But why should White be tage, is Huzman's line.
worse? He is better developed, better 1 8 bx_c3 .l:td8?! 1 9 ti)d4 ll:lc4? 20
co-ordinated, his king is safer and he ..txg7 + ! q,xg7 2 1 ti)f5+ exf5 22
has sacrificed nothing thus far. l:.xe7 0.xe7 23 'i'e2! t2Jg6 24 'Wxc4
1 4 i.f4! .:td2 25 .tb3 i.d5 26 •xa6 l:td8 27
This is the logical move. An inactive i.xd5 l:t8xd5 28 h3 ti)e5 29 a4 f4
piece is brought into the battle, adding 30 a5 f3 3 1 'Wb7 fxg2 32 a6 1 -0

1 76