Aagaard - Queen's Indian Defence (2002)

queen's
Indian
defence
by Jacob Aagaard
EVERYMAN CHESS
Everyman Publishers pie www.everymanbooks.com
First published in 2 002 by Everyman Publishers plc, formerly Cadogan Books plc,
Gloucester Mansions, 140A Shaftesbury Avenue, London WC2 H 8HD
Copyright© 2 002 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 300 4
Distributed in North America by The Globe Pequot Press, P.O Box 480,
2 46 Goose Lane, Guilford, CT 06 437-0480.
All other sales enquiries should be directed to Everyman Chess, Gloucester Man
sions, 140A Shaftesbury Avenue, London WC2 H 8HD
tel: 020 7539 76 00 fax: 02 0 7379 4060
email: [email protected]
website: www.everymanbooks.com
EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Chief advisor: Garry Kasparov
Commissioning editor: Byron Jacobs
Typeset and edited by First Rank Publishing, Brighton.

Production by Book Production Services.
Printed and bound in Great Britain by The Cromwell Press Ltd., Trowbridge,
Wiltshire.
Everyman Chess
Popular opening books:
18574 4 218 0 Unusual QG Declined Chris Ward

18574 4 253 9 Alekhine's Defence Nigel Davies
18574 4 256 4 Queen's Gambit Declined Matthew Sadler
18574 4 232 6 French Classical Byron Jacobs
18574 4 2 81 4 Modem Defence Speelman & McDonald
18574 4 292 x Symmetrical English David Cummings
18574 4 290 3 c3 Sicilian Joe Gallagher
18574 4 2 42 3 Offbeat Spanish Glenn Flear
18574 4 262 8 Classical Nimzo-Indian Bogdan Lalic
18574 4 2911 Sicilian Grand Prix Attack James Plaskett
1 8574 4 252 0 Dutch Stonewall Jacob Aagaard
18574 4 257 1 Sicilian Kalashnikov Pinski & Aagaard
18574 4 276 8 French Winawer Neil McDonald
Books for players serious about improving their game:
1 8574 4 226 1 Starting Out in Chess Byron Jacobs

·
18574 4 2318 Tips for Young Players Matthew Sadler
18574 4 236 9 Improve Your Opening Play ·Chris Ward
18574 4 2 41 5 Improve Your Middlegame Play Andrew Kinsman
18574 4 2 46 6 Improve Your Endgame Play Glenn Flear
18574 4 223 7 Mastering the Opening Byron Jacobs
18574 4 228 8 Mastering the Middlegame Angus Dunnington
18574 4 233 4 Mastering the Endgame Glenn Flear
18574 4 238 5 Simple Chess John Emms
Books for the more advanced player:
18574 4 233 4 Attacking with 1 e 4 John Emms

18574 4 233 4 Attacking with 1 d 4 Angus Dunnington
1 8574 4 219 9 Meeting 1 e 4 Alexander Raetsky
18574 4 22 4 5 Meeting 1 d 4 Aagaard and Lund
18574 4 273 3 Excelling at Chess Jacob Aagaard
queen's
Indian
defence
by Jacob Aagaard
EVERYMAN CHESS
Everyman Publishers pie www.everymanbooks.com
CONTENTS I
Bibliography 6
Preface 1
Introduction 9
White plays without g2-g3
1 Opting for Nothing: 4 �f4 and 4 �gS 16

2 The Petrosian System: 4 a3 24
3 5 'ii'c2 and 5 �gS 62
4 5 'ii'b3 76
5 4 e3 87
Lines with g2-g3
6 4 g3 i.b4+ and 4 ... �a6 without 5 b3 93

7 4 g3 i.a6: Main Line with 5 b3 111
8 4 g3 i.b7 121
Index of Complete Games 143

PREFACE I
The Queen's Indian Defence (arising after 1 d4 lLif6 2 c4 e6 3 lLi£3 b6) is one of the most
popular openings ai: both club and international level. The reason for this is obvious - ever
since its introduction at the beginning of the century, the Queen's Indian has proved to be the
most dependable defence for Black against 1 d4, together with its companion, the Nimzo
Indian (3 lLic3 J.b4). Additionally, most players like to create some kind of imbalance and
avoid structures of a symmetrical nature, such as those found in the Queen's Gambit Declined
and the Slav, although these openings also have their followers.
When I was originally asked to write a book about the Queen's Indian I asked if I could
write a two volume work in order to delve deep into the different positional aspects of the
opening. My publisher judged that most people would be more int\:rested in a single volume,
standard work, which is what I have ultimately agreed to. In order to achieve this I had to
make some decisions concerning the structure of the book because it was evident that certain
things which, in an ideal world, merited space, would have to be omitted. First, it was impossi
ble to provide a deep positional study in a naturally restricted book form, as I did with the
Dutch Stonewall. In this case I counted 19 different pawn structures that would have to be
discussed, and the only way this could be done would be via a multitude of arrows and short,
abstract words - something other writers have done before, but for which I have found little
use for myself and, consequently, believe is the same for others. Therefore, other than brief
discussions about the isolated pawn structures, below, this book is more a traditional work on
the Queen's Indian. Moreover, when it came to the lines to include I had to make numerous
decisions, minimising the material as much as possible. In ECO there are about 1 20 pages de
voted to the QID and the positions that can result via transpositions which, in a book like this,
would be around 300 pages. Meanwhile, ECO includes only those games that were previously
published in Informant, so they actually exclude quite a lot of important games! I decided to
include games from the very highest level -more or less ;n the games in this book, as well as all
the sidelines, originate from Grandmaster competition. But this only got me down to 20,000
games! Then I made the obvious decision: only critical lines should be represented. But merely
eliminating mistake after a mistake was far from enough. I also had to decide that some moves
were not critical in any way and that, therefore, they had no role in the book because nobody
would try to memorise the continuation, while looking it up afterwards also appears to me to
7
Q u e e n 's In dia n D e fe n c e
b e rather indifferent. Finally, in 'normal' positions, those 'normal' moves that give Black sev
eral ways to equalise have been omitted.
The result is a traditional work of theory, aimed at the tournament player and thus designed
for practical use. I feel that I have been good at finding the critical lines in modem play, and
picking out the most important games. I hope you will find this a useful and enjoyable book.
Jacob Aagaard
Copenhagen
July 2002
8
INTRODUCTION I
Theoretical Overview possibilities.

Here, briefly, is a an overview of the status of 4 . . d5!? leads to standard positions after
.
the various lines found in the following chap 5 cxdS exdS; these are normally very slightly
ters. better for White, but are ultimately a matter
of taste. Such positions are discussed
Chapter 1 throughout the book and are examined in
4 .tf4 is by no means an attempt to create some detail in Chapter 1 .
an opening advantage. I have a feeling that 4 . c6? ! has been played only once at the
..
the bishop is not very well placed in this top level and will 'rarely be repeated. White
variation as Black can still play ... d7-d6 in should secure a small plus in a symmetrical
order to limit its scope. In fact Black can position, which is always unpleasant for the
equalise in more than one way. second player.
4 .tg5 should normally lead to transposi 4 . c5!? leads to a sort of Benoni set-up
..
tions because the best move for White on the with the bishop on a6 after 5 dS �a6 where,
next move is 5 lilc3. Ideas with 5 lilbd2 ap for the time being, theory promises Black a
pear to be nothing other than a means of good game (yet nothing more). I have a feel
avoiding opening theory and promise White ing that White will eventually find a way to
nothing. prove a small plus here, although this has yet
to happen. This line has gained significantly
Chapter 2 in popularity thanks to Kasparov's convinc
4 a3 is, generally, an interesting system. As ing victory against Gelfand.
there is no line that guarantees an advantage 4 . .ta6 is the most popular system these
..
against the Queen's Indian White should days, leading to a variety of different posi
select his approach based on style and taste, tions, goth hedgehog and the more flexible
although I do feel that 4 a3 is not the most systems with ... cS and cxd4. This should lead
annoying move order for Black. 4 lll c 3, with to equality and pleasant play for Black if he is
the idea of meeting 4... �b7 With 5 a3!, ap sufficiently prepared, but it is a system that is
pears to avoid some of Black's most popular, still open for improvements and new ideas.
fashionable lines. 4 . . . .tb7 is the other main line (besides
Anyway, after 4 a3 Black has the following 4... �a6), and a solid reply. White has tried a
9
number o f different ways to gain an advan move an exclamation mark simply because it
tage but it still seems that Kasparov's idea is the most unpleasant for Black to meet.
from the 80's with 5 ltic3 dS 6 cxdS exdS 7 1bis hitherto offbeat line became a main line
'ilfc2! is the only real chance of giving Black a with the match Timman-Seirawan, Hilver
headache. I believe that these positions are, sum 1 990, where Seirawan employed it in all
perhaps, equal after correct play, but it his White games with convincing results.
should be home in mind that this correct Since then it has been played by some of the
play could often prove difficult to produce. other world's elite, with varying degrees of
success. Black has three main options.
Chapter 3 5. . . a5 does equalise but the resulting po
4 lll c 3 i.b7 5 Wc2 is harmless and Black sitions seem to be a little passive for Black
equalises easily, but the chapter is also mainly because he will often have to play on against
about 5 i.g5. If Black has played 4... i.b7, the two bishops.
then 5 a3! is the best way to fight for an ad 5 . . . We7 is also sufficient for equality.
vantage, and the move order I consider the However, one should always remember that
most dangerous for Black (what White if Black first has to fight to solve various
should play after 4... .ib4 is another question problems before equalising, then this only
- see below). 1bis particular move order means that both players have equal chances
gives Black the chance to equalise with 5 .. h6 . in the position itself, the assessment exclud
6 i.h4 i.e7!, after which there are no prob ing other factors such as familiarity with the
lems for the second player. position, time used on the clock and playing
4 ...i.b4 is a logical response, often played strength. Consequently, in the reality of
to avoid the 5 a3 transposition after 4... i.b7. competition, it is not necessarily always
These positions can arise after 1 d4 ltif6 2 c4 'happiness' when one equalises.
e6 3 ltic3 .ib4 4 ltif3, where both 4... cS and 5...c5 is the main move and, it seems,
4... b6 are main lines, the latter obviously also the best. After 6 a3 .iaS (the point)
transposing to the QID. Black is allowed to keep the two bishops for
5 i.g5 and now Black has ):wo ways to the time being, and he will exchange only if
equalise. The first is 5 .. h6 6 j_h4 i.b7 7 e3
. doing so leads to a clear gain of some sort.
(the move that has been proven to give Black Then comes 7 i.gS i.b7!, as played several
most problems, unlike7 1i'c2, for example, times in the match between Seirawan and
which does not impress) 7 ... .ixc3 8 bxc3 d6 Adams in Bermuda 1 999. Black should be
when a Nimzo-Indian structure has emerged, able to keep the game level, but this line is
with Black being quite comfortable. The still in its early childhood, and new discover
other comes after 7 ... gS 8 .ig3 ltie4. Al ies will most certainly emerge with time. It
though this has been played many times and should be noted that this line is full of all
the 'opening' theory can continue to 30 kinds of tactical tricks and options, which is a
moves or more, I still feel that there is room major reason why I recommend it. Better
for improvement and that Black should not preparation will pay off.
trust the verdict of equality completely. It
should be noted that White must play 9 'ilfc2 Chapter 5
and that the gambit with 9 ltid2 seems rather 4 e3. 1bis chapter provides some easy to
dubious. use ways for Black to equalise against this
(from a theoretical point of view) harmless
Chapter 4- system. 1bis might be a good place for a little
4 lll c 3 i.b4 5 Wb3! - I have given this warning: If you (as Black) do not know what
10
Intro du ctio n
you are doing against this system, you can 5 'i'c2 is completely harmless. Black
quickly find yourself in terrible trouble. See, probably has several ways to equalise, but the
for example, Yusupov-Beliavsky (Game 33), simplest is 5 ... dS!, when there is no doubt
where top players produce moves about about the evalua�on of the position.
which a deeper theoretical knowledge would 5 'i'b3 might look stupid and, after
have made them think twice. 5 ... �c61 (best) Black is able to exploit the
misplacement of the queen immediately to
Chapter 6 further his own development. However,
4 g3 sees the beginning of the great saga of White's move should not be mistaken for
White's most popular. I have always played 4 being bad since Black only equalises - noth
g3 when facing the QID, but I have never ing more.
mistaken the great taste I had for the posi 5 �bd2 is my own choice, although it
tions with an advantage, and neither, I am should not give Black any problems. Black
sure, have the majority of the players who has some different and interesting choices
employ the fianchetto. Black has a consider here. 5 ... .tb4!? is very interesting and leads to
able variety of choices, but I would recom an unbalanced situation in which, while
mend playing something along the main lines White might be slightly better, the stronger
with 4 ... i.a6 while, for White, I would seri player will be in the driving seat. All the lines
ously ask you to consider 4 �c3. Of course, with a quick ... c7-c5 seem a bit dangerous to
perhaps you play these lines to avoid fun, just me, but (of course) I might be mistaken. It all
like I do on a bad day. comes down to an evaluation of Game 39
4 . . . .i.b4+: The lines with ....tb4+ gener and the lines featured therein. Finally 5 ... dS
ally benefit from the inclusion of the prepara should equalise without too much effort, and
tory 4... .ta6. The only exception is 4 ... i.b4+ that is the main reason why this is not a main
·
5 i.d2 .txd2+ 6 'i'xd2 i.a6!, which trans line.

poses to 4... i.a6 5 b3 i.b4+ after 7 b3, which 5 'i'a4 is popular but simply lacks punch.
is the best move. Generally these lines seem 5 ... c6 is rather dubious, but after 5 ... cSI Black
to be me to be slightly worse for Black, but has several ways to generate a good game.
never anything less than that. They are not Actually I am normally pleased when I see
very exciting and the second player rarely has this move from Black as I simply cannot see
any reason to hope for a full point. how White can play for a win in the resulting
4 . . .ta6 is probably a better move than
. positions -there is no pressure.
4... i.b7, and certainly more modem. The
reason why this appears to be so clever is Chapter 7
that almost no matter which way White pro 5 b3 is the main line. And actually the only
tects the c4-pawn, Black can always play argument I see for playing the sytem with 4
5 ... i.b7 and claim that the white pieces are g3 is to prove that this move is an advantage
no better placed than they were the previous for White rather than a weakening of the
move. The idea certainly appeals to many dark squares. Having investigated the rele
players. Black should equalise in all the main vant V!lriations I must say that I still cannot
lines; perhaps there is a theoretical plus in the see this as anything other than a weakness,
main line, but that is only something for and it is for this reason that I recommend
players to worry about if they cannot draw that everyone dump 4 ... i.b7 and choose a
drawn endgames ... We will come to this but, line from below.
first, let us look at White's options to be pre 5 . . . d5!? is not in itself poor but the idea
sented in this chapter. is to answer 6 i.g2 with 6 ... dxc4, which is
11
unsound. 6 ... .i.b4+ might still be theoretically which is highly respected yet no longer has
okay, but I do not know many people who any place among the living. The story of the
like to play such positions. main line is a funeral - I know that this is an
5 c5 also looks slightly fragile, although
..• easy thing to say for a believer, so let us keep
it might work as a surprise. But if you are it blurred for now, and see if the evaluation
looking for good positions, look elsewhere. should ever change.
5 . . . b5!? is Adorjan's move, about which a
positional verdict is quite simple. If Black Chapter 8
finds success in ... a5-a4xb3 he can look for 4 . . . .tb7 is still popular with some top
ward to a good position. But why should grandmasters. Only the absolute elite such as
White allow this? Against a logical treatment Kramnik, Anand and Karpov stick to
Black should be suffering from his inferior 4 ... .i.a6, while Tiviakov, Timman and
pawn structure and be slightly worse. How Korchnoi still have some affection for this
ever, 5 ... b5 is not a poor move and it does old main move. Personally I have no doubt
offer Black chances to steer the game away about the correctness of 4 ... .i.b7, but I do
from more theoretical lines. feel that the play after 4... .i.a6 is more flow
5 . . . i.b7 6 i.g2 i.b4+ is an interesting ing and should suit most players - at least
option which promises something close to those who I know - better. But there are
equality after both 7 i.d2 a5 (Adams' some players with a very relaxed attitude to
choice) and 7 . . . c5 (Korchnoi's choice) . The chess, and these do best with the calm and
struggle often hinges on deep strategic un strategic 4 ... .i.b7. Anyway, let us turn to the
derstanding, and I would not recommend lines arising after 5 i.g2:
that you undertake the responsibilities of 5 . . . c5 is simply positionally suspect and
Black if this book is the only material you will Black has nothing to show for it but a very
look at. But if you work through the games modest - and very temporary - initiative. Stay
and find them interesting, then I will certainly away from this move.
not discourage you from seeking further 5 . . . i.e7 is the main move (5 ... .i.b4+ be
knowledge and subsequently including these longs to the comments given above for
lines in your repertoire. In fact most players 4... .i.b4+). Now White has two paths. First,
with White are unaware of what is best after after 6 0-0 0-0, there are some minor side
5 . . . i.b7 and can thus find themselves on lines.
unfamiliar territory when Black refrains from 7 b3 is harmless. Actually, I play this oc
entering the long main lines. casionally in order to reach an equal endgame
The main lines after 1 d4 lbt6 2 c4 e6 3 (you know - hot summer days when you
lbf3 b6 4 g3 i.a6 5 b3 i.b4+ 6 .td2 would like to be at the beach ... ). Nonetheless,
i.e7 7 i.g2 c6 have been debated endlessly do take a quick look at the easiest equalising
for the last decade, and the verdict is - not methods.
surprisingly - that it is very difficult for White 7 d 5 ! ? was once wildly popular. First Po
to gain an advantage; and if he succeeds it is lugaevsky used it, then Kasparov. Now -it's a
only of 'scientific' value - that is if Black is joke. Black has several ways to equalise and
well prepared! I have looked at 8 0-0 d5 9 can probably also play for a win without any
'il'c2, which is the new idea at the top level, great risks -if he is well prepared.
but it seems to be harmless, and if anybody 7 J:e 1 was modem a few years ago. The
takes any risks in the games played, it has problem seems to have been solved with
mainly been Black. 8 .i.c3 d5 9 llle 5 is the 7 ... llla 6!, after which the new try is 8 a3.
main line and has been treated as something Come on guys! Call that an advantage?
12
Intro ductio n
7 �c3 raises nwnerous issues. First White

has 6 �c3 with some different ideas, and this
is the most popular line at the moment. In
the event of 7 �c3 I would go for the highly
exciting 7 ... �a6!? - which is quite compli
cated - rather than the somewhat drawish
7 ... �e4. Anyway, after 6 �c3 Black usually
replies 6 ... �e4. Then 7 'ifc2 is hannless. In
stead after 7 .td2 it seems to me that only
7 . . .tf6! guarantees Black a good game. Of
.
course this is tested again and again at the

top level, but it seems to me that those lines
that are given as leading to an advantage for
White (8 1i'c2 �xd2 9 'i'xd2 is an example) This is it! Black has accepted an i solated
are often the lines I would fear least, while pawn on dS. As is usually the case this gives
those given as hannless might conceal a trick Black considerable· free play for his pieces,
or two. There are some tricks involving de but he will have to keep them active or use
laying castling but after 7 ...tf6 these do not
. them appropriately before White's perma
pay off (as they do against all the other lines). nent advantage of a superior pawn· structure
That's all folks. Enjoy your games, no begins to tell. Normally exchanges tend to
matter what faith or choice or line you might favour White, as the removal of pieces serves
take. to enhance the significance of the static as
pects of a given position. This does not nec
The isolated d5-pawn essarily mean that endgames are bad for
The QID is characteristic for its many differ Black. Typically White might have weakened
ent lines and prototypical positions. It can be his queenside with b2-b3 at some point,
compared to the Sicilian, where the struc when Black will be able to gain counterplay
tures after the opening vary from, for exam with a knight placed on e4, heading for c3.
ple, the wildly different Sveshnikov, Dragon Gelfand-Karpov, Game 31 is an example of
and Hedgehog variations. In the QID I have how activity can last long into the endgame.
found 1 9 different prototypical variations In this game Black creates compensation
which, of course, might themselves contain for the isolated d-pawn by advancing his g
nwnerous variations. Only one pawn struc pawn, creating some potential threats on the
ture (along with its distinct variations) is tran light squares around White's king. Gelfand
scendental universal and appears, in some then decides to eliminate the g-pawn and, in
form, in all lines. I am referring to a situation return, takes a weak pawn on e3 (which is
·
such as the following example: just as weak as dS).

1 5 . . . lllce4 1 6 J:ac 1 ..i.c5 1 7 e3 lllx c3 1 8
Gelfand-Leko 'i'xc3 llle4 1 9 °ifd 3 "iie 6 20 ..i.e5 J:ac8
Cannes 2002 21 ..i.d4 ..i.e7 22 °iff1 J:xc 1 23 J:xc1 J:c8
24 'ifd 1 i.d6 25 ..tn °iff5 26 �g2 J:xc 1
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 i.a6 5 27 'ifxc 1 f6 28 ..i.d3 �7 29 'ifc2 g5 30
"iia4 ..i.b7 6 ..i.g2 c5 7 dxc5 ..i.xc5 8 0-0 "ife2 . h5 3 1 llld 2 g4 32 f3 gxf3+ 33
..te7 9 lll c 3 0-0 1 0 ..i.f4 lll a 6 1 1 J:fd 1 "ifxf3 %-%
lll c 5 1 2 "ifc2 "ii c 8 1 3 J:d4 d5 1 4 cxd5
exd5 1 5 J:dd 1 In the following game Black places all his
13
pieces in a more o r less ideal way. I recom good eye on f2 and also prevents White from
mend you play through this game carefully putting his pieces completely at ease. The
and note how Black's pieces generate pres bishop on cS exerts pressure on d4 and e3
sure. and will, in the long run, prevent White from
playing f2-f3, or at least serve to remind
Kramnik-Hi.ibner White that e3 would then become a weak
Yerevan 1 996 ness. The bishop on b7 looks bad but can
often come into play with ... i.a6 or, perhaps,
1 lllf3 lllf 6 2 c4 b6 3 g3 i.. b 7 4 i..g 2 e6 even the pawn sacrifice ... d5-d4 on some
5 0-0 .te7 6 lll c 3 0-0 7 J:e 1 occasions. Only the heavy pieces are not easy
1bis is Kramnik's way of playing the 7 to place. Usually the rooks should be at d8
l:tel variation. The threat of 8 e4 often and e8 if Black is playing for ... d5-d4, and at
makes Black respond with 7 ... dS, but 7 ... cS, c8 and e8 if he (like here) has more or less
along the lines of the hedgehog set-up, is a ignored the fight for the d4-square.
sensible move. White, on the other hand, tends to have
7 . . . d5 8 cxd5 exd 5 9 d4 llla 6! problems in finding a good square for the
1bis is where the knight belongs. Black dark-squared bishop. Principally it is best
has no reason to fight for d4 in these posi placed on b2, but that, of course, cannot
tions because it is a fight he will only very happen without White weakening his queen
rarely win. Rather he focuses on bringing the side with b2-b3 at some point. As I men
knight to an active post, the ideal spots being tioned above, this will give the knight, which
cS and, primarily, e6. is usually placed on e4, a future. Additionally,
1 0 .tf4 c5 1 1 dxc5 lllx c5 1 2 J:c1 lllfe4 the bishop on g2 can often find that it is
1 3 lll b 5 llle 6 1 4 i..e 5 'i'd7 1 5 lll bd4 nicely placed, but with the knight on e4 it
J:ac8 1 6 l:lxc8 J:xc8 1 7 .i.h3 .tc5 1 8 a3 tends to be looking at a piece that cannot
a5 1 9 e3 l:le8 20 .i.g4 realistically be removed (the exchange would
be far from ideal anyway) for fear of the re
sulting weaknesses on the light squares.
Let us take a look at another variation of
the structures after ... d7-d5 and cxdS exdS,
where White is in no hurry to exchange on
cS.
Karpov-Spassky
Riga 1 975
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 i.. b 7 5

i.. g 2 .te7 6 lll c 3 0-0 7 'i'c2 d5 8 cxd5
lllx d5 9 0-0 llld 7 1 0 lll x d5 exd5 1 1 l:ld 1
Here a draw was agreed. ll:lf6
Where should White place his pieces here?
If we look at the final position we see that I could use this as an exercise for my pupils
all Black's pieces have a function. The knight because the answer is not so obvious if you
on e6 monitors both d4 and f4 and can, po are unfamiliar with the position. Well, the
tentially, jump to cS, from where an invasion queen is fine (£5 or d2 look appropriate for
might be possible. The knight on e4 keeps a later), the queen's rook goes to cl, the other
14
Intro du ctio n
rook is already perfectly placed, the queen's The same manoeuvre is also possible even
bishop has no obvious post and the other if the pair of knights has yet to be exchanged.
bishop is good. But what about White's The following example illustrates how White
knight? As Black will not fight for d4 the handles the possibility of ... c5xd4.
knight is not required on £3, but where would
we prefer it to be? The correct answer is f4, Beliavsky-Chuchelov
from where it exerts considerable pressure Ohrid 2001
on dS. The following manoeuvre is standard
and will not have come as a surprise for 1 d4 ibf6 2 c4 e6 3 ibf3 b6 4 g3 .ta6 5
Spassky. b3 d5 6 cxd5 exd 5 7 ..i.g2 ..i.b4+ 8 ..i.d2
..i.e7 9 0-0 0-0 1 0 ibc3 ..i.b7 1 1 ..i.c 1 iba6
1 2 ..i.b2 l:le8 1 3 l:lc 1 ..i.f8 14 ibe5 c5 1 5
ibd3 cxd4
1 2 ibe5 c5 1 3 dxc5 .i.xc5 1 4 tbd3 .i.d6

1 5 .i.f4!
This bishop has no ideal square so seeking
to remove it is quite logical. Black's trade in the centre is probably not
1 5 . . . .t!.eS 1 6 e3 ibe4 1 7 ..i.xd6 'iix d6 1 8 the best. Clearly Black was familiar with the
ibt4 following manoeuvre, but it still seems to be
White has an advantage. The d4-square is a particularly unnecessary concession to give
a good outpost for the heavy pieces and both up the d4-square when having to face a well
the knight and the bishop are well placed. placed knight on f4.
But most importantly all the play will re
- 1 6 ibb5!
volve around White's pressure on the dS The main idea. In this position the knight
pawn, so Black will not be able to generate transfer is rather straightforward, but some
sufficient active play to compensate for the players have a tendency to forget about this if
weakness. White has the option of recapturing on d4
1 8 . . . l:!.acS 1 9 'ii'a4 'ii'e 7 20 'it'xa7 ibxf2 with the queen. Now White has an edge.
21 ibxd5 ..i.xd5 22 'it'xe7 ibxd 1 23 .l:k 1 1 6 . . .'ii'd 7 1 7 ibxd4 ibe4 1 8 ibf4 fled8 1 9
i:tb8 24 'it'b4 .i.xg2 25 �xg2 ibxe3+ 26 ..i.h3 'i'es 20 a3 ibg5 2 1 .tg2 ibc5 22 b4
'it>g 1 i:te6 27 'ii'f4 l:ld8 28 'ii'd4 l:!.de8 29 iba4 l3 ..i.a1 ibe4 24 l:!.c7 l:lab8 25 ibf5
'ii'd 7 ibg4 30 l:!.c8 ibf6 31 l:!.xe8+ i:txe8 d4 26 'iic 2 l:bc8 27 ibxd4 a5 28 b5
32 'ii' b 7 l:le6 33 'ii' b 8+ ibe8 34 a4 g6 35 i:txc7 29 'ii'x c7 'it'd7 30 'ii'c 2 ibac5 31
b4 �g7 36 'ii'b 7 h5 37 h3 �f6 38 �g 2 ibd3 :ea 32 ibc6 ibxd3 33 .i.xe4 ibc5
l:d6 39 a5 bxa5 40 bxa5 l:!.e6 41 a6 ibc7 34 .i.xh7+ �h8 35 ..i.f5 ibe6 36 l:!.d 1
42 a7 l:!.e7 43 'ii'c 6+ �e5 44 'it>f3 1 -0 'ii'e8 37 'it'e4 1 -0
15
CHAPTER ONE I
Optin g for Nothin g :
4 i.f4 and 4 i. g 5
1 d4 lDf6 2 c4 e6 3 lDf3 b6 4 J.f4/4 J.g5 ties exist, take a look at them once or twice
There are some players, no matter what and then let your understanding of the posi
their rating, who always try to avoid main line tion be your guide. The three games I have
theory. This can be for different reasons, of presented here should more than cover what
course - one is simply a surrender to the you need to know about the system.
modem reality of the enormous amount of
games in a standard database, while another
is the need for an independent and original
way of playing, based more on psychology
than actual playing style.
Both 4 .i.gS and 4 .i.f4 are ways to stay
clear of all established theory in the Queen's
Indian. Neither is dangerous for Black, but if
you are content to get your pieces out and
then play a game on practically alien soil,
then 4 .i.f4 in particular makes good sense,
as can be seen from :Miles' and Bareev's rela
tive success with this approach.
4 J.b7 5 e3 J.b4 +
...
Game 1 This is an interesting choice. The simpler

Miles-Unzicker 5 ... .i.e7 is presented in Game 3.
Johannesburg 1979 6 lLJfd2
The key reasoning behind this retreat is as
1 c4 lDf6 2 lDf3 b6 3 d4 e6 4 i.f4 follows:
This move is mainly played by those want 1) The queen's knight belongs on c3 and
ing to avoid the extensively explored waters not d2, and
of opening theory. It is neither good nor bad, 2) White does not want doubled pawns.
just a plausible option. The thing to do when There are two alternatives, each contra
you are preparing an opening system like the dicting these points. Somewhat similar to the
QID is to acknowledge that these possibili- Nimzo-Indian and the lines in this book fea-
16
Optin g fo r No t h in g : 4 i. f4 a n d 4 i. g 5
turing 4 l£1 £3 and 5 i.g5 are the positions l:.el i.d6 17 'fif3 with a slightly worse posi
arising after 6 l£ic3 i.xc3+ 7 bxc3 d6, when tion for Black.
there are two set-ups for Black to choose 1 0 lll xd 5 i.xd5 1 1 ';ifc2
between: Here White has a serious alternative in 1 1
a) 8 i.d3 l£ibd7 9 0-0 'fie7 10 l£id2 e5 1 1 l:.cl . Then 1 1 ...i.d6!? 1 2 i.xd6 cxd6 looks
i.g3 0-0 1 2 l:.e 1 c5 sees Black trying to solid, while Miles-Hubner, England-West
dominate White's centre and bishops. In fact Germany 1 979, continued 1 1 ...c5 12 dxc5
this seems to be rock solid, and Black had a i.xc5 13 i.c4 i.b7! 14 0-0 i.d6 with
good game in Ostenstad-Hjartarson, Gausdal approximate equality. Instead 13 ... i.xc4 14
1 985. l£ixc4 'iWxdl+ 1 5 �xdl l£Jd7 1 6 �e2 l£if6 1 7
b) 8 h3 l£ibd7 9 i.h2 'fie7 10 l:.bl 0-0 1 1 i.d6 led to a lasting ending advantage for
i.e2 l£ie4 1 2 l:.b3 f5 brings about another White in Miles-Ligterink, Amsterdam 1 978,
traditional set-up from the Nimzo-Indian. It while 13. ..l£ic6 14 b4 i.e7 15 i.b5! favours
is not clear that i.h2 is, after all, such a great White as both i.xc6 and e3-e4 are threat
piece, but it is easy for Black to find good ened . .
squares for his minor pieces, e.g. 13 0-0 e5!? 1 1 . . . c5
(limiting the scope of White's bishop, al Equally good looks 1 1 ...i.d6 12 i.xd6
though it was also possible to ignore it) 1 4 c5 cxd6 13 i.d3 g6 14 0-0 l£Jd7 1 5 l:.acl l:.c8 1 6
i.c6 and Black is fine, Agdestein-Yrjola, 1Wa4 l:.c7 wi th equality in Meduna-Unzicker,
Gausdal 1 985. Moscow 1 982. Here 14 e4 i.b7 1 5 0-0 l£ic6
Meanwhile 6 l£Jbd2 0-0 7 i.d3 d5 8 0-0 1 6 l£1£3 'iWf6 also seems to fail to give Black
l£ibd7 9 l:.cl i.d6 10 l£ie5 l£ie4 1 1 cxd5 problems in finding good squares for his
exd5 12 l£ixd7 'iWxd7 gave Black full equality pieces.
in Matera-Dzindzichashvili, New York 1 980. 1 2 dxc5 i.xc5 1 3 llle4
6 . 0-0 7 a3 i.e7 8 lll c 3 d5
.. 13 i.d3 h6 14 0-0 l£Jd7 1 5 l:.fdl l£if6
looked fine for Black in Meduna-Brynell,
Leon 2001.
1 3 . . . llld 7!
9 cxd5 lllx d5 !
This is the safe path to equality. Recaptur
ing with the pawn would be ungrateful to the
bishop on b 7, although it has been tried. Black chooses to complete his develop
After 9 ... exd5 10 i.d3 c5 1 1 0-0 l£ic6 1 2 l£1£3 ment. This is more important than the two
White probably has a slight edge. In Meduna bishops. Now, thanks to the activity of the
Vilela, Leipzig 1 980, play continued 1 2 ... a6 knight and the possibility of exchanging
13 l£ie5 cxd4 14 l£ixc6 i.xc6 15 exd4 l:.e8 16 dark-squared bishops, Black can equalise.
17
1 4 lid 1 'i'c8 1 5 .td3 .txe4 1 6 .txe4 ltif6 7 ... ltixg3!? 8 hxg3 .if6 9 'i'c2 g6 1 0 0-0-0
1 7 .td3 'i'e7 1 1 .id3 lLic6 1 2 .i.e4 0-0-0, leading to
The trick was, of course, 17 .i.xa8 .i.b4+!, an original position where both players had
and Black wins. good chances in Rodriguez-Fedorowicz,
1 7 . . . .te7 1 8 'i'e2 Wc5 1 9 h3 ltid5 20 New York 1 987 .
.tg3 llad8 21 0-0 .td6 22 llc 1 ft'a5 23 8 fic2
.txd6 J:lxd6 24 llfd 1 l:fd8 25 .tb5 ltie7 In this game Bareev elects, quite logically,
26 llxd6 .l:lxd6 27 J:lc3 lld5 28 a4 Wb4 to bury the bishop on b7 by advancing d4-
29 J:lc7 a5 30 g3 g6 31 �g2 'i'd6 32 d5. If he had not done so he might have
lld7 fic5 33 J:la7 lids 34 fif3 ltif5 35 e4 played along the following lines: 8 .ie2 g6! (a
Yz-Yz standard reaction; now the knight won't be
After ...ltid6 Black has nothing to fear. hanging after surprises like ltif3-e5) 9 0-0 d6
------ 1 0 llcl ltid7 1 1 lLid2 lLixg3 1 2 hxg3 a6 1 3
Game 2 .i.£3 c6 1 4 g4 'i'b8 1 5 a4 d5 1 6 cxd5 cxd5 1 7
Bareev-Karpov .i.e2 'i'd6 with equality in Agdestein-
Moscow (blitz) 1993 Adorjan, Oslo 1 984.
._____________... 8 . . . d6 9 d5
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 .tf4 .tb7 5 Essentially the same position as in the
e3 .te 7 game, with the same players, arose after 9
1bis move is more natural than 5 ... .ib4+ .i.e2 g6 1 0 0-0-0 lLid7 1 1 d5 e5 1 2 ltid2 in
and should lead to a good game without too Bareev-Karpov, Paris 1 992. Now Black did
much trouble. not play the best move, 12 ... ltig7!, which he
6 ltic3 llih5! must have realised later (which brings us to
the present game). Instead 1 2 ... lLidf6 13 �bl
a6 1 4 £3 lLixg3 1 5 hxg3 h5 1 6 e4 .ic8 1 7 lLifl
lle8 1 8 ltie3 .i.f8 1 9 'i'd3 'i'e7 20 llh2 .i.d7
21 lldhl .ig7 22 ltic2 .i.h6 23 ltib4 llec8 24
ltic6 'i'e8 25 g4 saw White being allowed to
create a strong attack that was enough to win
the game...
9 . . . e5 1 0 .te2 g6 1 1 lLld2
If White fails to address the well-being of

the bishop then Black should, of course,
attack it. Now White will anyway have to lose
a tempo...
7 .tg3 0-0
Karpov's strategy is logical and simple.
The key idea is to develop, eliminating the
bishop only when he is somehow forced to
do so. Another possibility is to delay castling 1 1 . . . ltig7 !
by exchanging th e bishop immediately, e.g. The correct decision. The g3-bishop is not
18
Optin g for No th in g : 4 i.. f4 a n d 4 i.. g 5
impressive and White would benefit tremen i..d 1 J:.fe8 32 Wf2 J:.xe 1 33 tbxe 1 b5
dously from the opening of the h-file, as can
be seen in the following game: 1 1 ...ltixg3?! 12
hxg3 ltid7 1 3 0-0-0 Wg7 1 4 g4 a6 1 5 g3 c6 16
i.£3 l:tc8 1 7 dxc6 i.xc6 18 i.xc6 l:txc6 19
'ii'e4 with a clear advantage to White in
Bareev-Karpov, Paris 1 992. It should be
noted that all these games were in 'active'
chess of some form, and not played with the
usual, longer time control.
1 2 e4 tbd7 1 3 0-0
1bis looks very suspicious.
1 3 . . . i..g 5!
Compare the bishops on gS and g3. Black
has at least equalised. 34 cxb5 i..x d5 35 a4 J:.e4 36 i..f 3 .l:.e5
1 4 b4 h5 1 5 h3 'i'f6 1 6 tbf3 i..f4! 37 tbd3 IH5 38 tbt4 i..c4 39 .l:.e 1 �f7
0-1
Game 3
Nikolic-Lautier
Moscow (Fide knockout) 200 1
1 d4 ll:if6 2 tbt3 e6 3 c4 b6 4 i..f4 i..b 7 5
e3 i..e 7 6 h3
Preserving the bishop, but this costs time,
so Black decides to act at once in the centre.
6 . . . c5!
Exchanging the bishop to win the eS

square and making way for the g-pawn.
1 7 i..xf4
17 ... gS, with an attack, was a serious
threat.
1 7 . . . exf4 1 8 ii'd2 g5 1 9 e5!?
White is worse and he knows it, so he de
cides to sacrifice a pawn in a bid to change
the nature of the position and win some
squares for his pieces.
1 9 . . . tbxe5 20 tbxe5 'i'xe5 21 J:!.ae 1 'i'f6
22 h4 gxh4 23 tbe4 'i'f5 24 ii'd4 f6 25 Till s is the easiest way to equalise. White's
llld 2 h3! dark-squared bishop has left the queenside
Black uses his additional material to ruin (g1 -a7 and everything on the queenside as a
White's king position. Black is winning. result of e2-e3) so Black then operates on the
26 gxh3 'i'xh3 27 'ilixf4 J:.ae8 28 'i'g3 dark squares there. 1bis is perfectly logical.
'i'xg3+ 29 fxg3 .l:.e5 30 tbf3 J:.e3 31 The old way to play this was 6 . . 0-0 7 ltic3
.
19
d 5 but this might result in a modest advan son-Hansen, Gentofte 1 999 went 1 0 0-0 d6
tage for White in some cases. After 8 cxd5 1 1 ll:ib3 'flc7 12 :et ll:ibd7 with equality,
ll:ixd5 9 ll:ixd5 'ii'xd5 1 0 a3 c5 1 1 dxc5 'ii'xc5 although Black's game is probably the easier
12 ..ie2 Z%d8 13 'ii'b l ..ia6 14 ..ixa6 'ii'as+ 1 5 to conduct.
c;ti ft 'ir'xa6+ 1 6 <it>gl ll:id7 17 <it>h2, a s in Me b2) 8 ... 0-0!? is provocative but if White
duna-Lemer, Lvov 1 981 , White's minor cannot exploit it then it is by far the most
pieces are well placed, although Black should logical response. Miles-Kupreichik, Reykjavik
be okay. Perhaps if White plays something 1980 continued 9 ll:idb5 ll:ie8 1 0 'ii'd2?! (this
slower in the opening Black will have to face move is obviously just losing time; better was
some problems. 1 0 ll:id6 with an unclear position) 1 0 ... a6 1 1
7 d 5? ll:id6 ll:ixd6 1 2 .ixd6 .txd6 1 3 'ii'xd6 b5! 14
This is completely suicidal. It is difficult to a3 bxc4 1 5 Adl ll:ic6 1 6 .ixc4 'ii'g5 and
imagine a top GM playing something like Black has a useful lead in development.
this. There are two essentially different set 7 . . . exd5 8 lll c 3 dxc4!
ups here.
a) 7 dxc5 bxc5 8 ll:ic3 0-0 9 ..ie2 ll:ic6 1 0
0-0 d6 1 1 'ir'c2 'ii'b6 is a universal approach
for Black in the QID and almost always
guarantees an equal position. Lputian
Mikhalchishin, Frunze 1 979 continued 1 2
l:.adl l:tfd8 (1 2 . . .:ad8 1 3 :d2 h 6 14 l:tfdl
l:td7 1 5 ll:iel l:.fd8 was equal in Langeweg
Timman, Leeuwarden 1 980) 1 3 l:td2 d51? (an
interesting decision, sharpening the position)
1 4 :fdl ll:ib4 1 5 'ii'b t dxc4 1 6 :xd8+ :Xd8
1 7 ll:id2 (1 7 ..ixc4 ..ixf3 1 8 gxf3 ll:ibd5)
1 7 ... ll:ibd5 1 8 ll:ixc4 ll:ixc3 19 bxc3 'ii'x bl 20
:Xbl ..ie4 21 l:tb3 ll:id5 and here it is White Black is testing the pawn sacrifice. Or he
who is fighting to keep the balance (which is just taking the pawn. You decide.
Lputian failed to do). 9 lll b 5
b) 7 ll:ic3 cxd4 and now a possibility is 8 It is too late to play for compensation
exd4, leading to a position similar to the with 9 .ixc4 because 9 ... d5 1 0 .ib5+ i..c6
Panov/Nimzo-Indian after 8 ... 0-0 9 .td3 d5 gives White very little.
1 0 0-0 dxc4 1 1 .ixc4 ll:ic6 1 2 l:.cl Zlc8 1 3 9 . 0-0!
..
.td3 ll:id5, Rivas Pastor-Hubner, Linares Simple chess. White should not really take
1 985. Here the moves h2-h3 and .if4 were the exchange so Black simply makes the
clearly not very aggressive. Black has equal move from which he most benefits.
ised. This leaves us with 8 ll:ixd4 and a fur 1 0 lll c 7?
ther branch: The strategically decisive mistake. The
bl) 8 ... a6 9 .ie2 0-0 and now 10 .tf3 is only move here is 10 .tc7! 'flc8 1 1 i..d6,
logical, dealing with Black's bishop, which with .ixc4 to come, trying for some kind of
has considerable influence on the centre. compensation.
Gretarsson-Stefansson, Stockholm 1 998 1 0 . . . d5 1 1 lllx a8 .txa8
continued 10 ...l:ta7 11 'ii'a4 .ixf3 1 2 ll:ixf3 Black has two big central pawns and an
d6 1 3 0-0 'ir'c8 14 l:tfdl :d8 1 5 :d2 :c7 1 6 enormous lead in development. The advan
'ii'd t with an equal position, while Gretars- tage is not in doubt.
20
Optin g fo r No t h in g : 4 i. f4 a n d 4 i. g 5
4 . . . i.b4+
This is just as natural as 4 ... .i.b7, although
Black now loses the option of 5 ... .i.e7 in
stead of ... .i.b4. Of course if Black is not
interested in 5 ... .i.e7 the text is quite a funny
move, threatening to engineer a win, as hap
pens in this game.
4... .i.b7 5 l2Jbd2?! fails to impress. After
5 ... .i.e7 6 e3 h6 7 .i.h4 0-0 8 .i.d3 cS 9 0-0
cxd4 10 exd4 dS!, as in Wuts-Matthias, Ger
many 1 996, Black has at least equalised. The
long-term potentials in the position are in
Black's favour and the knight on d2 is not
1 2 i.e2 ©c6 1 3 0-0 b5 1 4 a4 b4 1 5 well placed in terms of the fight for the dS
'i'b1 ©e4 1 6 :d 1 ©a5 1 7 ltie5 ©b3 1 8 square. Now Black will be able to take con
i.xc4 trol over this quite easily and thus develop
Who cares about pieces here? The pawns more freely than is usual in such positions.
are doing the job! 5 ©bd2
1 8 . . . ©ed2 1 9 'ifc2 dxc4 0-1
This move gives the line independent

...----. value, but it is not good. 5 lLic3 .i.b7 leads us
Game 4 back to Chapter 3.
Teske-Paehtz 5 . . . h6 6 i.h4
East Germany 1988 6 a3!? is a possibility. If you are an occa-
._______________.. sional Bogo-Indian fan you can play
(When one is confronted with a game 6 ... .i.xd2+ 7 .ixd2 .i.b7 with a good ver
such as this one always wonders whether the sion!? Otherwise the simple 6 ... .i.e7 should
draw was pre-arranged, as well as the moves equali�e directly.
themselves - often this is the case, but you 6 . . . g5 7 i.g3 g4 8 ltle5?
can never be really sure) Losing. The only move is 8 a31, with good
1 d4 ©f6 2 c4 e6 3 ©f3 b6 4 i.g 5 prospects for Black after 8 . . . .i.e7 9 lLieS d6
This move has less independent relevance 10 'i'a4+ c6 1 1 lLid3 'i'd71 followed by
than 4 .i.f4, as it often transposes to 4 lLic3 ... .i.b7 and ... c6-c5.
.i.b7 5 .i.gS, which is treated in Chapter 3. 8 . . . ©e4 9 .if4 'i'f6 1 0 ©d3 'i'xd4!
21
Capturing the pawn without releasing the would take the risk!
pressure. 1 2 exd4
1 1 e3 Now we have a perpetual.
1 2 . . . �f3 + 1 3 �e2 �xd4+ 1 4 �e3
�f5 + 1 5 �e2
1 5 <i>e4 .tb7+ 16 ci>eS d6+ 1 7 <i>f6 �d7
mate(!) is a sour way to end your days.
1 5 . . . �d4+ 1 6 �e3 �f5 + 1 7 �e2 �d4+
%-%
·1 1 . �xd2??
..
What a blunder! After 1 1 ....txd2+! 1 2

'ii'xd2 �xd2 1 3 exd4 �xf1 1 4 :Xf1 .tb7
Black is clearly better. White has some com
pensation on the dark squares but I, for one,
22
Op tin g for No th in g : 4 1.. f4 a n d 4 1.. g 5
Summary
4 i.f4 is a rather harmless move and after 4 ... i.b7 5 e3 both 5 i.b4+ and 5 . i.e7 lead to safe
... . .
equality. Still it is a good idea to check out the most important ideas here, or White might be
able to pose Black some problems. 4 i.gS should not gain any independent significance, but
simply transpose to Chapter 3. After S lL!bd2 Black is always going to be fine.
1 d4 tL'lf6 2 c4 e6 3 tL'lf3 b6 4 .tf4 (D)

4 i.gS (D) i.b4+
S lL!c3 i.b7 - Chapter 3
5 lLi bd2 Game 4
-
4 . . . 1.b7 5 e3 1.e7 (D)

5 i.b4+ 6 lLifd2 Game 1
... -
6 tL'lc3 Game 2
-
6h3-Game3
4 1.. f4 4 .tg5 5...1.. e 7
23
I CHAPTER TWO I
The Petrosian System:
4 a3
1 d4 ltlf6 2 c4 e6 3 .!bf3 b6 4 a3 10 i.f4 tl:ihS, which currently looks like the

The Petrosian System had its heyday in most promising system for Black in the
modem times during the 1980's, when Kas 4 ... .ta6 line. In Game 1 3 it is time for
parov - still a boy - blew away the world's 10 ... i.cS, which seems to give White an edge.
strongest GMs. Today the variation does not
enjoy the same status and is regarded as one Game 5
of several options, not necessarily offering Portisch-Nikolic
White an advantage, b�t leading to positions I.jub§ana 1985
where many players feel comfortable. And,
of course, new ideas are constantly produced. 1 d4 .!bf6 2 c4 e6 3 .!bf3 b6 4 a3
In Game 5 we see what happens when
Black pushes with ... d7-d5 and recaptures
with the e-pawn. White might have a slightly
more comfortable game but Black should do
okay with decent play. Game 6 features an
idea, 4... c6, tried once and not again (for
good reasons), while Game 7 introduces a
very interesting, dynamic system, namely
4... c5l?, a line that resembles the Benoni,
which seems to be totally playable at the
moment. In Game 8 we just glance at what
happens if White is unambitious against
4 ... c5 - and not surprisingly, nothing happens.
Game 9 includes the minor lines against 4 d5!?
...
4 ... .ta6, which is one of Black's main replies, This move is not the standard theoretical
while we discuss the direct assault 5 'ii'c2 recommendation, but such great players as
.ib7 (5 ... c5 leads to Game 7) 6 tl:ic3 cS 7 d5?! Petrosian, Larsen, Karpov and Bronstein,
in Game 1 0. Game 1 1 features one of the have given it a strong vote of confidence,
main lines, 7 e4 cxd4 8 tl:ixd4 i.cS, and simply by employing it against fellow players
Game 1 2 investigates 8 ...tl:ic6 9 tl:ixc6 i.xc6 from the world's elite. Besides this, 4... i.b7
24
Th e Pe tro s ia n Sys tem: 4 a 3
and 4... i.a6 are the standard moves. 4 ... c6 treated under 7 °ii'a4+ in Game 1 S.
and 4 ... cS are treated in the following games, 5 . . . exd5
while 4 ... .!De4 S .!Dfd21 dS 6 e3 i.b7 7 cxdS! S ... .!DxdS 6 e4 is not relevant.
exdS 8 .!Dxe4 dxe4 9 .!Dc3 i.d6 10 i.bs+ 6 ttlc3 i.e7 7 i.f4
c;ilf8 (1 0 ... c6 1 1 i.a41 and e4 is in trouble, This is one of the two standard ways to
unless Black accepts something along the place the pieces. 7 i.gS is also possible, of
lines of 1 1 ...fS 1 2 i.b3!, and the king is course, with a position from the Queen's
caught in the middle) 1 1 'ii'c2 'ii'gS 12 g3 a6 Gambit Declined rather than the QID. The
1 3 i.ft f5 1 4 i.g2 gave White the advantage real altemative, and simply a matter of taste,
in Glek-Rozentalis, Lvov 1 98S. is 7 g3 0-0 8 i.g2 cS 9 0-0 i.b7 10 i.f4, but
4 ... i.e7 S .!Dc3 0-0 is a strange and altema White cannot count on a significant advan
tive way to play. After the logical 6 dS!? (tak tage after 1 0 ... .!Da6 1 1 .!Des (1 1 dxcS bxcS 1 2
ing control over the light squares in the cen °ii'b3 °ii'b6 1 3 .!Dd2 'ii'xb3 14 .!Dxb3 l:ad8 1 S
tre, hoping to bury the bishop on c8) 6 ... l:e8! .!DaS i.a8 was equal in Lputian-Nikolic, Sara
(directed against 7 e4) 7 g3 i.a6 8 i.g2! jevo 1 998, but 1 1 l:ct I seems to be the most
i.xc4 (8 ... d6 9 dxe6 fxe6 1 0 .!Dd4 seems logical move) 1 1 ....!Dc7. Note that here
risky) 9 .!Des i.xdS 10 .!DxdS exdS 1 1 i.xdS 1 1 ... .!De4? meets with 12 .!Dxe4 dxe4 1 3 dxcS
lLJxdS 12 'ii'xdS c6 13 'ii'xf7+ 'ath8 14 'ii'h s .!DxcS 1 4 'ii'c2 .l:.c8 1 S .l:.fd1 'ii'e 8 1 6 i.h3
'ii'c 8 1S i.f4 dS 16 .l:.ct i.. £8 Black might .!De6 17 'ii'b 3, which put Black under great
have been okay in Gofshtein-Vitolinsh, pressure in Vaganian-Gulko, Lvov 1 978. As
Beltsy 1 979, but the general impression of we shall see it is a characteristic of these posi
the game is that Black had to take risks with tions that ... .!De4 tends to be a clear positional
out ever having a genuine opportunity to error, and Black should be well aware of
obtain more than an equal position .. when it is playable and when it is not!
Returning to 1 1 ...�c7, with 1 2 dxcS White
chooses a forced- continuation which ends
with a completely level position, but ... cxd4
was a real threat now that the knight has
control over bS. After 12 ... bxcS 13 .!Dc4 Abs
14 i.xc7 'ii'xc7 1 S .!DxdS i.xdS 1 6 i.xdS
l:fd8! (the most active, although 1 6 ... AbdS is
also good enough) 17 e4 .!Dxe4 we have the
following:
a) 1 8 'ii'hS?I g6 1 9 i.xf7+ is given by
Parma as leading to an advantage, but
1 9 ...'atg7 20 'ii'e2 .!Dd2! 21 l:.fe1 (21 .!Dxd2?1
l:.xb2 22 .l:.ad1 ti;xf7 and here the bishop
5 cxd5 and the passed pawn must give Black a fa
White decides not to give his opponent vourable position) Z1...i.f6 22 i.e6 .!Dxc4 23
the chance to retum to the main lines with S 'ii'xc4Jhb2 followed by ... i.d4 might even
.!Dc3 i.b7, although most players would favour Black.
probably stick with their choice and play b) 1 8 'ii'e2 .lhdS 19 'ii'xe4 'ii'd 7 20 Afet
S ... i.e7, with a transposition. However, after i.f6 21 l:tac1 hS 22 h4 i.d4 23 l:.c2 aS 24
the usual 6 cxdS Black can remain within the l:.ce2 .l:.fS with equality and a draw in
lines here with 6 ... exdS, but there are some Browne-Ljubojevic, Brasilia 1981.
things with 'ii'a4+. All these nuances are 7 ...0-0 8 e 3
25
bS 14 lLieS 'iib 6 1S i.gS :£e8 1 6 f4 earned

White an advantage in Wilder-Larsen, New
York 1 984.
d) 8 ... c6!? offers White some chances of a
slight advantage if he plays 9 i.d3. But after
9 .!Lies i.£5 1 0 g4!? i.e6 1 1 i.d3 ltifd7 1 2
'ii'c2, a s in A.Petrosian-Short, Yerevan 1 984,
the situation is more difficult to evaluate,
although White looks good here, too ..
8 c5
. . .
The most active, but Black has more than

one reasonable move here.
a) 8 ... i.£5 seems perfectly logical because
the exchange has already been performed on
dS and, consequently, b7 is no longer a good
square for the bishop. Still, the move does
abandon the light squares on the queenside.
Van der Sterren-Short, Baku 1983, continued
9 i.e2 (9 i.d3 also looks okay) 9 ... ltie4 1 0 9 llle 5
'iib3 c 6 1 1 0-0 i.d6 1 2 i.xd6 'ii'xd6 1 3 :ac1 It may sound strange but this is the most
ltid7 with approximate equality. flexible move available to White. Note that it
b) 8 ...i.b7. This way of holding back the is the only move that White is sure he will
c-pawn makes sense. All the pieces can de make, whereas the decision to put the bishop
velop freely, so why not? Timman-Sunye on e2 or d3, for example, can be made later.
Neto, Las Palmas 1 982, went 9 i.e2 (9 However, many games have been played
.llc 1 !?, forcing Black to make a decision, is using these moves directly:
interesting) 9 ... ltibd7 10 .!Lies (1 0 0-0 must be a) 9 i.e2 is the most serious alternative.
more testing) 1 0 ... ltie4 1 1 ltixe4 dxe4 1 2 l:tc1 After 9 ... i.b7 10 .!Lies (1 0 0-0 ltic6 or
ltixeS 13 i.xeS i.d6 14 i.xd6 cxd6 1S 'ii'a4 1 0 ... ltia6, or 1 0 ... ltibd7 1 1 .!Lies transposes)
(this move does not impress; 1 S 0-0 must be Black must make a choice - two options are
better) 1 S ... a6 1 6 0-0 i.dS 1 7 l:tc3 bS 1 8 'ii'c2 interesting, the third less good.
'iib 6 and Black looks more than fine. Bacrot al) 10 ... ltia6!? is natural. It has not been
Roder, Spain 2001 , saw positionally unsound properly tested in practice, only in this one
play from Black: 9 ... a6 10 b4 i.d6 1 1 i.g3 game, Bacrot-Kurajica, Yerevan 1 996, where
'iie 7 1 2 'iib3 ltibd7 1 3 0-0 i.xg3 1 4 hxg3 White achieved nothing by playing 1 1 'iia4
'iid 6 1 S :fc1 :£e8 1 6 bS c6 1 7 bxc6 i.xc6 ltic7 12 ltic6 i.xc6 1 3 'iixc6 with the two
1 8 \ib4 and Black had weaknesses in his bishops, being behind in development and
pawn structure, while his bishop had no already fielding threats - after 1 3 ... ltie6 14
prospects. i.eS cxd4 1 S i.xd4 (1 S exd4!? is the testing
c) 8 ... i.e6 seems illogical. The bishop is move) 1 S ... l:tc8 1 6 \ibs ltixd4 1 7 exd4 ltie4
less well placed here than on f5 and b 7. 9 h3 1 8 ltixdS i.cS 1 9 dxcS :xcS 20 \ib4 'iixdS
cS 1 0 i.d3 ltic6 1 1 0-0 a6 1 2 :cl c4 1 3 i.b1 was equal. Instead 19 :dl ! 'ii'xdS 20 dxcS
26
'ifxc5 21 'ifxc5 llxc5 22 0-0 lies 23 llfel choose a very interesting plan in 1 1 b4!?, with
gives White a tiny lead with bishop against the idea that c3 will be a fine blockading
knight in the endgame, although this will , in square, and later e3-e4 will undermine Black's
99.9% of all cases, not influence the result. centre. Ostojic-Abramovic, Banja Vrucica
a2) 1 0 ... tLlbd7 seems to do less well with 1 991 was unclear after 1 1 ...a6 1 2 0-0 b5 13
... c7-c5 than the other knight moves. After tLle5 .tb7 14 i.f.3 .
1 1 0-0 l:r.cS 1 2 dxc5 tLlxc5? (1 2 ... bxc5!? 9 .i.b7
. . .
should keep the balance) 13 tLlb5! a6 14 tLld4 Once the pawn has arrived on c5 this is
White had the advantage in Ribli-Gligoric, the appropriate square for the bishop. After
Lucerne 1 982. 1 1 ...tLlxe5? is almost always 9 ... i.e6 the c6-square is left weakened, some
bad when White's bishop is on e2: 1 2 dxe5 thing well illustrated in the following game:
tLle4 13 tLlxe4 dxe4 1 4 Wa4 a6 15 llfdl 'ife8 1 0 i.d3 'ii'c 8 1 1 0-0 c4 12 i.c2 tLlbd7 1 3
1 6 1i'h3 b5 17 e6 with advantage to White, i.a4 a 6 1 4 tLlc6 l:leS 1 5 tLlxe7+ :Xe7 1 6 f.3
Lputian-Zagrebelny, Lucerne 1 993. The e4- b5 1 7 i.c2 Wc6 1 S Wd2 and the two bishops
square will always be weak in the endgame. gave White good chances for an advantage in
a3) 1 0 .. . tLlc6 1 1 0-0 cxd4 1 2 exd4 tLle4 (as Bisguier-Villarroe� Caracas 1 970.
always with the bishop on e2 Black ends up 1 O .i.d3 lLic6
in trouble after 1 2...tLlxe5?! 13 dxe5 tLle4 14 10 ... tLlbd7 is quite passive -as I mentioned
tLlxe4 dxe4 1 5 'ifa4 a6 1 6 l:lfdl We8 1 7 Wb3, earlier, ... c7-c5 and ...tLlbd7 do not combine .
as in Timman-Larsen, Tilburg 1 9S2) 13 well. 1 1 'ii'f.3 lies 1 2 0-0 a6 1 3 Wh3 created
tLlxc6 i.xc6 14 llcl 'ifd7 15 i.a6 l:lad8 1 6 promising attacking chances for White in
Wd3 i.d6 and now Farago-Matanovic, Vi Tai-Bronstein, Tbilisi 1 982.
enna 1 9S6 continued 1 7 tLle2?! i.xf4 1 S 1 1 0-0
tLlxf4 'ii'd 6, when Black appeared to have
survived the opening. With 1 7 tLle2?! White
completely lacked a plan; better was 1 7 i.xd6
Wxd6 1 8 llc2! to use the c-file in the near
future!
b) 9 i.d3 was played in Plaskett
Stefansson, Reykjavik 1 992. There followed
9 ... tLlc6 1 0 llet (1 0 tLle5 transposes to the
main line) 1 0 ... i.b7 1 1 0-0 h6 1 2 tLle5 tLlxe5
13 dxe5 (1 3 .lxe5, with equality, is most
likely superior) 13 ... tLle4 14 'ii'h 5 Wes!, pre
paring ... f7-£5 in the case of a sacrifice. After
1 5 tLlxe4 dxe4 1 6 i.c4 b5 17 i.a2 c4 Black
was better. 11 cxd4
. . .
c) 9 h3 tLlc6 1 0 i.d3 should not be dan This is the main line in ECO and other
gerous for Black, but after 1 0 ... cxd4 (a pre places, but I recommend 1 1 .. .tLlxe51 12 dxe5
mature decision; 10 ... i.b7 offers good (1 2 .i.xe5, with equality, might be the best
chances of equality) 1 1 exd4 i.d6 1 2 i.xd6 move) 12 ... tLle4!, the difference between this
'ifxd6 1 3 0-0 lies 14 i.b5! i.d7 1 5 i.xc6 and other situations being that the bishop is
i.xc6 1 6 tLle5 White had an edge in Schan now on d3. Browne-Frias, Santiago 1 981
dorff-Larsen, Denmark 1 9S9. continued 13 'ii'c2 tLlxc3! 14 bxc3 (1 4 i.xh7+
d) Interesting is 9 llet tLlc6 10 i.e2 c4 �hS 1 5 bxc3 g6 1 6 i.xg6 fxg6 1 7 'ii'xg6 'ii'e S
(1 0 ... i.b7 is more natural). Now White can and, as Black has control over all the light
27
squares, there will be no attack, only three in the endgame, but only if he manages to
scattered pawns for the piece), and now not penetrate to the 7th rank will he be able to
14 ...g6? 15 i.h6 :e8 16 i.b5, when White win.
was winning, but 14 ... h6, with an unclear 20 'ifd3 g6 21 J.. b 3 'iPg7 22 :tae1 :ea
position that Black should not fear. The al 23 f3 l0d6 24 'ife3 l0e4? !
ternative 1 2...ltJe8?! was tried in Ligterink Forcing this exchange only benefits White.
Nikolic, Wijk aan Zee 1 984, when 1 3 'ii'g4 g6 After a more careful move Black should be
1 4 :ad1 h5 1 5 'ii' £3 ltJc7 1 6 i.h6 was slightly only slightly worse.
better for White. 25 J.. xe4 dxe4 26 :ted 1 'ifg5!?
1 2 l0xe6 ..be6 1 3 exd4 J..d 6? !
After this move Black will always face
some problems. The exchange is illogical as it
highlights the difference between the bishops
on d3 and c6. An improvement is 13 ...'ii'd 7!
14 '1'£3 l':.ae8! (the argument for using this
rook lies in the following line: 14 ....l:[fe8 1 5
l:.fe1 i.d6 1 6 i.g5! ltJe4 1 7 ltJxe4 dxe4 1 8
i.xe4 :Xe4? 1 9 :Xe4 'ii'b7 20 :ae1, where
.. f7-f5 has no protection - with the rook on
.
f8 this is not the case) 1 5 .l:[fe1 i.d6 1 6 i.e5

(1 6 i.g5? ltJe4! 17 �xe4? dxe4 18 i.xe4
:Xe4! 1 9 :Xe4 'ii'b7 and Black wins)
1 6 ... i.xe5 17 dxe5 d4 1 8 ltJe4 was Petursson Introducing a tactical endgame in which
Larsen, N:estved 1 988, and now Larsen Black is on the receiving end.
writes that Black can keep the balance with 27 'ifxg5 hxg5 2a d5 J.. b 7 29 :te5 :te5?
1 8 ... �xe4 1 9 i.xe4 i.xe4 20 'i'xe4 :e6!, Here 29 ... �f6 is the appropriate move, af
which looks correct as far as I can see. 21 ter which White maintains some pressure.
:ad1 l':.d8 22 f4 d3 23 f5 l':.ee8! is an impor 30 :txg5 f5 31 h4 'iPf6 32 d6 :te5 33
tant line, when the e-pawn is getting as weak 'iPf2 :te6 34 l0e2 :tdxd6 35 l0f4 :txd 1
as the d-pawn, at least. Black should avoid 36 :txg6+ 'iPf7 37 :txe6 :td4
23 ...l':.c6?? 24 e6 fxe6 25 fxe6 'ii'e 8 26 e7
.l:[dc8 27 'i'xd3 and White wins.
1 4 J..g 5
1 4 'i'£3!? i s also natural, preparing a future
i.e5!.
1 4 .. .. .. h6 1 5 J.. h 4 :tea 1 6 J.. e 2 J..e 7 1 7
:te1 l0e4
This is the only move that, in the long run,
frees Black from having to push with ... g7-g5.
1 a J.. x e7 :txe7 1 9 'ife2! :td7? !
Very passive. Now White gets time to
strengthen his position without any worries.
1 9 ...l':.e6 20 i.xe4 dxe4 21 d5 i.xd5 22 :ad1
·
.l:[d6 23 'i'd2 'i'h4 24 �xd5 l':.ad8 25 g3 'ii'g4 3a :th6 :txf4 39 'iPg3 :td4 40 :th7 + 'iPg6
26 ltJe7+ �f8 27 'ii'xd6 :Xd6 28 :xd6 <lhe7 41 :txb7 :td2 42 :txa7 :txb2 43 :te7 b5
29 l':.d3 might give White a small advantage 44 :te6 + 'iPg7 45 :te5 'iPg6 46 :tee+
28
Th e Pe tro s i a n Sys tem: 4 a 3
<iL'g7 47 .J:.c5 <iL'g6 48 h5+ <iPxh5 49 7 . . .a 6 8 l:.c 1 ! .tb7

l:xf5+ <iL'g6 50 .J:.c5 <iL'f6 5 1 a4 bxa4 52 8 ... i.e7? is very bad in view of 9 llia4!!,
.J:.xc4 a3 53 l:.a4 l:.a2 54 l:.a5 <iL'g6 55 f4 when Black is forced into 9 ..i.d7 1 0 i.c7
.
'iPf6 56 .J:.a6 + <iL'e7 57 �f3 <iL'f7 58 g4 i.xa4 1 1 'ii'xa4+ 'ii'd7 12 'i'b3 bS 13 llieS
<iL'e7 59 g5 1 -0 with a significant advantage to White. After
The lesson provided by this game, from a e2-e3 comes a3-a4.
theoretical point of view, is that when White 9 e3 ll)c6? !
puts his bishop on d3 Black should consider Failing to appreciate the immediate danger
... llixeS followed by ... llie4, but with the on the c-file. With this in mind 9 ... llibd7
bishop on e2 Black will have to unravel more followed by ... l:.c8 is absolutely necessary.
slowly, although he should be able to main Now White springs a wonderful combina
tain the balance. tion.
Game 6
D reev-Seirawan
Wijk aan Zee 1995
1 d4 ll)f6 2 ll)f3 e6 3 c4 b6 4 a3 c6? !

One of those ideas that is tried once for
the sake of originality, and never tried again
for the sake of results ...
5 ll)c3 d5
1 0 ll)b5! .J:.c8
10 ... axbS 1 1 i.xbS :cs 12 llieS sees
White win the piece back with interest.
1 1 'ii'a4 ll)d7
1 1 ...llihS 1 2 llic7+! l:r.xc7 13 1.xc7 'il'xc7
14 i.xa6 i.xa6 1 5 :xc6! and White wins.
1 2 ll)c7 + .J:.xc7 1 3 .txa6!
6 cxd5!
Letting Black choose at once.
6 . . . cxd5
This was the reasoning behind 4 .. c6, of
.
course. Otherwise the move would have

contributed nothing to Black's cause. 6 ... exdS
meets with 7 i.gS, when White has the free
dom to play as he likes in the centre.
7 .tf4!
The bishop takes up residence on what is
now the most dangerous diagonal for Black. Precisely played. After 13 !i..x c7? 'i'xc7 1 4
29
i.xa6 ltJdb8 1 5 i.xb7 11i'xb7 Black is ready exd5 i.e7 Black is starting to unravel.
with ... b6-b5, breaking the pin. And after 1 6 22 exd4 i::tta 23 a4!
11i'b5 i.d6! Black prevents ltJe5.
1 3 . . . i..x a6
There is no alternative. 13 ... ltJdb8 1 4
i.xb7 l:txb7 1 5 i.xb8 b 5 1 6 11i'a6 and White
regains his piece, or 1 3 ... e5 1 4 i.xb7 :Xb7
1 5 11i'xc6 and White wins.
1 4 i..x c7 'i'xc7 1 5 J:!.xc6 'i'b7 1 6 lbe5
i..e 7 1 7 b4! f6
17 ... i.c4 would lead to serious trouble af
ter 1 8 :Xb6!, but not 1 8 ltJxc4 dxc4 1 9 b5 as
indicated by Dreev. Then 19 ... ltJb8! 20 l:txc4
11i'xg2! by no means looks bad for Black!
1 8 b5?
Now Black can do nothing to stop the a
pawn.
23 . . . i..g 5 24 J: 1 c2 J:!.f4 25 a5 J:!.xd4 26
g3 J:d 1 + 27 �g2 :a1 28 axb6 lbb8
28 ...11i'a6 29 11i'xa6 l:.xa6 30 b7 l:txc6 31
l:txc6 i.f6 32 l:td6 and White wins.
29 J:d6 + 1 -0
Game l
Gelfand':'Kasparov
Novgorod 1997
1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 a3 c5!?
This is simply a mistake. After 1 S l:txb6! The alternative continuation 4... i.a6 5
11i'xb6 1 9 11i'xd7+ 'iii> f8 20 ltJc6 11i'b7 21 11i'xe6 11i'c2 c5 6 d5 is seen just as often.
11i'c7 22 �d2! 11i'd6 23 11i'xe7+ 11i'xe7 24 ltJxe7 5 d5
'iii>xe7 25 a4 White has very good chances in 5 e3 is considered in Game 8.
the endgame. 5 ... i..a 6 6 'i'c2
1 8 . . . i..x b5 1 9 'i'xb5 fxe5 20 0-0
20 l:txe6 exd4 �1 0-0 'iii> fl 22 l:tc6 ltJf6 is
okay for Black.
20 . . . exd4?
The decisive mistake. Dreev must have
not seen that Black is absolutely fine after
20 ... 0-0 21 l:tfcl ! i.d8 22 l:txe6 e4, instead
giving 21 l:txe6 as a simple explanation, but
after 21...i.xa3 22 l:tal l:.c8! all Black's pieces
are playing.
2 1 J:tc 1 ! �dB? !
Another grave error. After 21.. .i.d8 22
:Xe6+ �fl 23 l:.cc6 ltJf6 24 e4!? 11i'd7 25
30
Th e Pe tro s ia n Sys t em: 4 a 3
6 b3 has been played a few times but the Elista 1 995. And 7 ... ll:\xd5 8 'i!ie4+ llle7 9
pawn structure is not really suited for a 'ii'xa8 lllec6 1 0 llle5 'ii'c7 1 1 i.f4 lllxeS 1 2
bishop on b2 as d6 is the weak spot in 'ii'e4 simply wins fo r White, a s in Einarsson
Black's camp. Thorhallsson, Reykjavik 1 996.
6 exd5
. . . 8 t2Jc3
Black can try to prevent e3-e4 with There is an alternative set-up both here
6 ...'ii'e 7?! but experience shows that this and on the following move, designed to take
move, which hinders development, is rather immediate advantage of the weaknesses of
risky if White reacts with determination. the d6-square and the a4-e8 diagonal. Luckily
Miles-Kudrin, London 1 982, went 7 i.g5 for Black, he also has his resources. 8 i.f4!?
exd5 8 lll c 3! i..xc4 (8 ... i.b7 9 cxd5 h6 1 0 d6 9 ll:\c3 i..g7 10 'i!ia4+ 'ii'd 7! 1 1 i.xd6
i..xf6 'ii'xf6 1 1 e4 was a shade better for 'ii'xa4 12 ll:lxa4 lllxd5 13 0-0-0 llle 7! is very
White in Vyzmanavin-Rozentalis, Lvov important, for if the knight goes to f6 it
1987) 9 e4! (Kasparov's contribution) 9 ... h6 blocks the bishop and does not possess any
10 i.xf6 'ii'xf6 1 1 exd5 i.xf1 1 2 <t>xf1 d6 1 3 mobility. About 6-7 serious games have been
l:.e1+ i.e7 1 4 'ii'a4+ � £8 1 5 'ii'g4 and Black played with 1 1 i.xd6, and in all of them
was under a lot of pressure. Returning to the where the knight went to f6 (mainly after 1 3
position after 9 e4!, Black has a couple of e4) Black did badly, while i n th e rest, where
alternatives to 9 ... h6. 9 ...'ii'e 6 1 0 i..xc4 dxc4 the knight went to e7, Black equalised.
1 1 0-0-0 lll c6 12 e5 lllh 5 1 3 llld S .l:c8 14 Browne-Timman, Las Palmas 1 982 contin
'ii'xc4 i.e7 1 5 g4 was terrible for Black in ued 14 e4 i..xf1 1 5 :men lllb c6 16 lll c3
Hernandez-Ortega, Cuba 1 982, but 9 ... dxe4! i.xc3! 1 7 bxc3 l:td8 1 8 i.. c7 .l:d7 1 9 i.f4 f6
10 0-0-0 i.e6 1 1 lllxe4 'ii'dB 1 2 i.c4 i.e7 1 3 with an equal ending which Black went on to
i.xf6 i.xf6 1 4 lll d6+ ct> f8 1 5 i.d5 i.xd5 1 6 win. Less ambitious is 1 1 'ilixd7+ �d7,
:Xd5 gives White compensation fo r th e ma when Black should , be doing fine. In Garcia
terial rather than an enormous advantage. Trobat-Rodriguez, Malaga 1 991 White even
7 cxd 5 g6 got into trouble after 1 2 g3 �e7 1 3 i.g2
lllbd7 1 4 0-0 .l:he8 1 5 h3 lllh 5 1 6 i.e3 �£81
etc. Finding a good move for White here is
by no means easy.
8 .i.g7 9 g3
. . .
White can also give up the right to castle

in order to gain immediate control over the
centre with 9 e4 i.xf1 1 0 �f1 , but this ne
glects the fact that in Benoni positions Black
is normally thrilled to exchange the light
squared bishop for any piece. Consequently,
without any weak light squares, this makes
little sense from a positional point of view.
Of course this is my way of thinking. A
This only serious move. The bishop never number of the world's top players think dif
belongs on e7 in a Benoni set-up and win ferently, and the variation has been tried by
ning a pawn with 7 ... i.b7?! 8 e4 'ilie7 9 i.d3 better players than I. Arencibia-Almasi, Elista
ll:lxd5 looks unhealthy. After 10 0-0 lllc 7 1 1 1 998, for example, continued 1 0... d6 1 1 i.f4
i.g5! f6 1 2 i.f4 d 6 1 3 lll c3 ll:ld7 1 4 b41 0-0 1 2 h3 l:te8 1 3 g4 b51 14 lllxb5 lllxe4 1 5
Black was under pressure in Sakaev-Ruban, l:t e1 a 6 1 6 ll:\c3 lllxc3 1 7 :Xe8+ 'ilixe8 1 8
31
bxc3 �d7 1 9 �g2 (White should avoid 1 9 Benoni set-ups is that Black has not yet
i.xd6 c4 2 0 i.g3 �f6 2 1 d6 �e4, where the played ... a7-a6 or ...�d7. Now the knight
. importance of the inactive rook is all too comes to b4 in a second, and White experi
evident) 1 9 ... c4 20 .l:tel 'ii°d 8 21 i.xd6 �f6 ences problems with the a6-f1 diagonal (d3)
22 i.eS 'i'xdS 23 'i'd2 'i'xd2 1/2-1/2. Cebalo and the dS-pawn. 1 1 ...i.b71 1 2 �d2 �a6 1 3
Tukmakov, Pula 1994 saw Black enjoy a 0-0 �b4 14 'i'b3 'i'e7 1 5 l::t d l i.a6 gave
pleasant position after 14 litg2 'i'b6, with all Black a clear advantage in Matjanovic
his pieces being able to find good squares. Timman, Bled 1 979.
Who knows - Black might eTren be a little bit 1 1 . . .:es
better here. This is the exact developing move. It is a
9 . 0-0 1 0 i.g2
. . question of timing and the deeper details in
The automatic square for the bishop, and the position. After both possible knight
also the best. However GM Neverov has moves Black comes under pressure.
experimented with 10 i.h3 bS!, but now 1 1 ...�bd7?1 12 'iia4! i.b7 (1 2 ...'iic 8?1 1 3
Black does not rush with ... d7-d6 as there is i.h3! and Black i s under irritating pressure,
no pressure on the g2-a8 diagonal. Play with i.f4! coming soon) 1 3 'iWh4! and Black
might continue 1 1 0-0 l::te 8 (Also good looks has some problems, as has been demon
1 1 .. .b4 12 axb4 cxb4 13 �dl �xdS 14 i.gS strated in many games, one such being
'i'c7 1 5 'ii'd2 i.b7 1 6 l::tc l 'Was 17 �e3 Grooten-Riemersma, Holland 1 993.
�xe3 1 8 i.xe3 i.x£3 1 9 ex£3 �c6 with a 1 3 ...�xdS? loses at once to 14 �xdS i.xdS
very complex position, which ended with a 15 �gSI, for example.
draw in Neverov-Onischuk, Donetsk 1 998) Nor does 1 1 ...�g4?! 1 2 .l:.el improve
1 2 b4!? (the e2-pawn needs protection, and Black's situation because after the natural
ruling out ... b5-b4 safeguards the dS-pawn) 1 2 ... �d7?? White might consider 1 3 'i'a4,
1 2... d6 1 3 i.f4 �hS 1 4 i.gS 'i'c7 1 5 bxcS when Black resigned in Korotylev-Kiselev,
'i'xcS 1 6 l::t fcl i.c8 1 7 i.g2 and now instead Moscow 1 996. Sometimes it is nice to see
of 17 ... �a6, as in Neverov-Nisipeanu, 2545 players blunder like we might do in 1 -
Koszalin 1 998, 1 7 ... i.b7!? looks best, with at minute games on the Internet Chess Club! It
least equality for Black. makes me feel that the game is still alive for
1 0 . d6
. . humanity...
1 2 J:e 1
Clearly the most natural move, and by far
the most popular. Others:
a) 12 i.f4 �hS 13 i.gS 'ii'c8 14 .l:.fel
�d7 15 g4?1 �hf6 16 i.f4 (1 6 h3 hS! is also
bad for White) 16 ... �xg4 17 �e4 :t.xe4! 1 8
'ii'xe4 i.xb2 and Black was better in Zhu
Chen-Ehlvest, Beijing 1 998. Look at the
prospects for the queenside pawns!
b) 12 b41? was original and not unsuccess
ful in the following game but, as we shall see,
Black had some promising improvements:
1 2 ...�bd7 (1 2 ... �xdS 1 3 i.gSI, or 12 ... �e4
1 1 0-0 13 �xe4 i.xal 14 i.gS f6 1 5 :Xal fxgS 1 6
An understandable mistake would be 1 1 �fxgS, followed by �e6 o r 'i'c3, with good
a4?. The problem here compared to normal compensation) 1 3 i.b2 i.c4 14 .l:tfdl cxb4
32
(1 4...b51?) 1 5 axb4 l::tc 8 (1 5 ... a5, gaining c5 for lLJc3 looked better for Black in Riazantsev
the knight) 1 6 'ii'd2? (1 6 e4 lLJxe4 17 lLJxe4 Mitenkov, Moscow 1 998) 1 6 ...°ii'f6 1 7 l:ad1
i.xb2 1 8 lLJxd6 with a mess) 1 6 ... i.xe2? i.c6 and White's pieces are all tangled up,
(1 6 ... lLJxd5! wins a pawn� 1 7 lLJxe2 lLJe4 1 8 Szymczak-Panczyk, Poland 1 982.
'ii'd3 i.xb2 1 9 :xa7 and the position was a3) 14 l:ad1 lLJe4! 15 "ii' a4 (1 5 lLJxe4
rather unclear in Krasenkow-Emms, Hast 'ii'xe4 1 6 °ii'd2 "ii'e7 with equality) 1 5 ... i.xc3!
ings 1 990. 16 bxc3 i.b7 17 l:d3 5 and Black is by no
c) 12 l:bl lLJbd7 13 :e1 b5 14 b4 l:c8 1 5 means worse, Jusupov-Timman, Tilburg
°ii'd3 lLJg4 1 6 lLJe4 "ii'e7 saw White already 1986.
experiencing problems in Zhu Chen b) 13 e4 lLJg4 14 i.g5 "ii'c7 15 i.f4 lLJde5
Alterman, Beijing 1 997. 16 l:ad1 b5 gave Black excellent prospects in
1 2 lL!bd7
. . . Popov-Kastanieda, St. Petersburg 1 998.
1bis is the standard move. There are some c) 1 3 'ii'a4?! i.b7 1 4 'ii'h4 does not work
transpositions with 12 ... b5, but they are not now as the rook is no longer on ffi! After
so relevant. Yet one idea does look interest 1 4...lLJxd5! 1 5 lLJxd5 i.xd5 1 6 lLJg5 i.xg2 1 7
ing: 1 2...h61?, playing against the cl -bishop. °ii'xh7+ � ffi 1 8 �g2 lLJf6 1 9 'fi'b4 d 5 Black
13 e4 i.b7 14 i.f4 lLJhS 1 5 i.e3 lLJa6 1 6 has the advantage (... d5-d4 and ...°ii'd 5 are
l:adl °ii'd7 led to an interesting position in coming, with total domination).
Touzane-Bauer, France 1 996. However, I 1 3 b5
. . .
have a feeling the bishop's value on b7 has Clearly the best move. After 1 3 ...lLJe5?1 14
declined here. lLJxe5 :xe5 1 5 e4 l:e8 16 i.e3 lLJd7 17 f4
Black's forces had problems finding squares
in Yusupov-Timman, Tilburg 1 986.
1 4 e4
1bis was supposed to be the move for
White until the present game was played.
Now nobody really knows. Just as frequently
played has been 1 4 i.f4, but Black should be
fine if he continues 14 ...'i'b6 1 5 e4 l:ac8 1 6
i.e3 i.b7!?, remembering to push with . . .a7-
a5 and then ... i.a6 in order to fight for con
trol of the c4-square. Yevseev-lonov, St.
Petersburg 2000, went 17 lLJd2 a6 1 8 a4 b4
19 lLJc4 (1 9 a5 "ii'c 7 20 lLJcbl c4 and Black is
1 3 h3! okay) 1 9 ...'ii'c7 20 lLJbl a5! 21 lLJbdZ i.a6 22
By far the most logical option. White i.f4 i.ffi 23 l:ad 1 lLJh5 with a sound posi
wants to play e2-e4 without allowing ... lLJg4- tion for Blaek. In Moreno-Emms, Mondariz
e5-d3. Others: 2000, 1 6 ...'i'b7 1 7 i.fl lLJb8! also proved to
a) 1 3 i.f4 "ii'e 7, and now: be fine, e.g. 1 8 i.f4?! b4 1 9 lLJdl b3 20 'i'b1
a 1) 1 4 e4 lLJg4! 1 5 i.g5 f6 1 6 i.f4 lLJge5 i.xfl .f 1 �fl 'fi'a6+ 22 �g2 lLJbd7 23 lLJc3
17 lLJxe5 lLJxe5 1 8 l:adl °ii'd 7 and Black has c4 and Black was better.
the better prospects, Kramnik-Timman, 1 4 'ifcS ! !
. . .
Linares 1 993. White's pieces are nicely I have used this position a s an exercise in
placed, but they are not really doing anything. my positional training program on the Inter
a2) 1 4 "ii'a4 i.b7 1 5 lLJb5 lLJxd5 1 6 i.xd6 net. The key idea is that the bishop should be
(1 6 lLJxd6 i.xb2 17 lLJxb7 i.xal 1 8 :Xa1 supported so that Black can play ... b5-b4 as
33
soon a s possible. Kasparov's move i s the need to see variations to see that 1 7 'fib 1? is
only one that does this. The automatic bad. Just look at the white pieces!
14 ... :c8 1 5 ..ie3 'fic7 (1 5 ...b4 1 6 axb4 cxb4 1 7 lL!xe4 1 8 'it'c2
•••
. 1 7 :Xa6 bxc3 1 8 li)d41 followed by li)c6 has 1 8 li)c3? l::tb 8 19 'ii'c2 .l:.xb2! pockets a
long been known to give White a good posi pawn for Black.
tion) 1 6 ..tn 'fib7 17 ..if4 gave White a good 1 8 . . . lLidf6
position in Van Wely-Kamsky, Amsterdam
Donner 1 996. 14 ...'fib6 1 5 ..ie3! also slows
down the ... b5-b4 idea.
Here White needs to do something to

challenge the power of the e4-knight. The
way he does it is terrible, as it serves only to
1 5 .tf4 weaken his king.
The aggressive approach, but after this it 1 9 g4? !
is White who must try to keep the balance. 1 9 li)g5! was the only move.
However, other games suggest that Black is 1 9 'it'd7 20 g5? !
• • •
absolutely fine. 15 ..ifl!? b4 1 6 li)b5 ..ixb5!? Still 20 li)g5!? i s necessary.

17 ..ixb5 l::tb 8 18 ..ic6 b3 19 'fid3 li)e5 20 20 lLih5 21 .th2
• • •
li)xe5 l::txe5 21 <li>g2 l::te 7 led to an interesting 21 l::txe4 .l:.xe4 22 'ii'xe4 l::te 8 23 li)xc5
position (which Black went on to win) in l::txe4 24 li)xd7 .l:.xf4!? leads to a very clear
Markos-Nisipeanu, Cappelle la Grande. The advantage for Black. White's pieces could
question is what is the most important here - hardly be worse.
the bishop on c6 or the weaknesses of the 21 . . . f51
b2-pawn? The debate is most likely not over.
1 5 ..id2? b4 1 6 li)dl b3 1 7 'fibl c4 was
plainly good for Black in Tunik-Ionov, St.
Petersburg 2001 .
1 5 b4 1 6 lba4
• • •
1 6 li)dl b3 1 7 'fibl 'fic7 1 8 li)c3 l:.ab8

did not look impressive in Zhu Chen
Zatonskih, Istanbul 2000.
1 6 b3!
• • •
This is the idea - now White loses control

over e4.
1 7 'it'xb3
I am one of those people who does not
34
With complete domination. nos Aires 1 980.

22 �c3 l:.ab8 23 l:.ab 1 i.xc3! 6 �c3 cxd4 7 exd4 d5 8 cxd5
Securing the difference between Black's There has been an attempt to finish Black
majestic knight and the bishop on h2. off immediately by weakening his light
24 bxc3 l:.xb 1 25 l:.xb 1 i.c4 26 �d 2 squares with 8 i.g5 i.e7 9 i.xf61? i.xf6 10
lpxd2 27 1"xd2 f4! 28 l:.e 1 l:.e5!? 29 cxd5 .txd5 1 1 llixd5 'ikxd5 1 2 'ii'a4+ ri;e7 1 3
l:.e4? l:.xe4 30 i.xe4 'ifxh3 31 i.g2 1"g4 l:tcl , but rime has shown that thi s i s not pos
32 'ife 1 �g7 ! 33 f3 'ifxg5 34 'ifb 1 �f5 sible. Christiansen-Gurevich, Parsippany
35 1"b8 + �g7 36 'ifxa7+ �h6 37 'iff7 1 996 went 13 ...l:td8 14 .td3?! a6! 1 5 0-0 l:ta7
i.f 1 ! 38 �xf 1 �e3 + 39 �e1 'ifh4+ ! 40 1 6 l:tfel .txd4! 17 i.xh7 Wb5 1 8 'ikxb5 axb5
�e2 'ifxh2 41 �d3 �f5 0-1 19 llixd4 :Xd4, when Black was better
placed for the endgame, while 14 i.e2 a6 1 5
Game 8 l:tc7+ l:.d7 1 6 l:tc8 J:ld8 1 7 :Xd8 ri;xd8 1 8
Silman-Gheorghiu 0-0 llic6 results in equality according to
Palo Alto 198 1 Gurevich.
8 . . . �xd5
1 d4 �f6 2 c4 e 6 3 �f3 b 6 4 a 3 c 5 5 e3 Tbis is clearly the most flexible move, and
as White has no way of troubling Black
straight away there is no need to resort to
8 ... exd5?! 9 i.bs+ llibd7 1 0 ltie5 i.e7 1 1
ltic6 .txc6 1 2 .txc6, when White stood bet
ter in the game Eingom-Lauber, Bad Wor
ishofen 1 997.
9 i.b5+ i.c6 10 i.d3 �d7 !
Tbis move is less ambitious than 5 d5, and

Black has no problems equalizing.
5 . . . i.b7
5 ... g6!? 6 liJc3 .tg7 7 .te2 0-0 8 0-0 .tb 7
9 b4 (9 d5?1 exd5 1 0 cxd5 l:te8 gives Black a
good version of the Benoni - where does
White's queen's bishop belong and how is
the other bishop contributing?) 9 ... cxd4 1 0 The bishop appears t o b e exposed o n c6
llixd4 (10 exd4!? d 5 1 1 c 5 bxc5 1 2 bxc5 ltie4 but White has no way to exploit it, so Black
should not certainly not be worse for Black) maintains the balance.
1 0 ... ltic6 1 1 llixc6 i.xc6 1 2 i.b2 d5 1 3 'iib3 1 1 o:o i.e7 1 2 i.d2? !
(1 3 cxd5 llixd5 with at least equality, Tbis clearly shows that White lacks suffi
Browne-Timman, London 1 980) 1 3 ... dxc4 1 4 cient uriderstanding of the positions arising
i.xc4 'ike7 1 5 l:tadl l:tfd8 with an equal posi with the isolated pawn. 1 2 'ike2 is an im
tion (but also somewhat boring, with chances provement.
only for a draw) in Fanno-Andersson, Bue- 1 2 . . . 0-0 1 3 l:.c 1 �xc3 1 4 l:.xc3 i.b7 1 5
35
�f4 ltif6 1 6 �e5 :ea 1 7 .J:xea Wxea 1 a

•b 1 g6 1 9 We1
5 •b3
The queen is quite obviously less well
This looks just like a bad endgame, and placed here than on c2. Other moves have
guess what it is . . . ? also been tried.
1 9 . . .Wxc 1 20 .J:xc 1 tlld 5 21 �e4 f6 22 a) 5 e3 is really harmless, and often trans
�g3 .J:ea 23 .J:xca+ .ixe8 24 .ib8 a6 25 poses to the 4 e3 line after 5 ... dS, where
�xd5 exd5 26 .ic7 b5 27 b4 .if5 White has the extra move a2-a3, which
Here the bishop is better than the knight. hardly does him any good (Black can also
Black has good practical chances, although play 5 ... cS 6 �c3 i.. e7 7 i..e 2 cxd4 8 �xd4
White might have been able to hold it with i..b 7 9 i.. £3 �c6 with total equality, as in
best play. Hort-Ljubojevic, OHRA 1 988). And now:
28 ltie 1 �f7 29 � 1 h5 30 �e2 �e6 3 1 at) 6 b3 i.. e7 7 i..d3 0-0 8 0-0 cS 9 i..b2
.if4 g5 3 2 .i c 1 �d6 3 3 h 3 g 4 34 h 4 g3 dxc4 10 bxc4 �c6 1 1 �eS �aS 12 �d2 l:tc8
35 f3 .ib1 36 tlld 3 �f5 37 ltie5 .ia2 38 13 l:tct �d7 and White's queenside weak
tllx a6 .ic4 + 39 �e 1 .if4 40 .ixf4 �xf4 nesses began to tell in Smolej-Siegel. Ger
41 tll c 5 �e3 42 a4 bxa4 43 tllx a4 �xd4 many 1 985.
44 ltic5 �c3 0-1 a2) 6 �c3 i.e7 7 �eS r 'ii' a4+ c6 8 cxdS
..------.. i.xft 9 �xf1 exdS, Flesh-Romanishin, Lvov
Game 9 1 981, is simply better for Black - look at
Barlov-Sakaev White's bishop) 7...0-0 8 i..e 2 c6 9 0-0 �fd7!
Budva 1996 and Black has solved all his opening prob-
_______________.. lems, Browne-Benjamin, USA 1 984.
1 d4 tllf 6 2 c4 e6 3 tllf3 b6 4 a3 �a6 a3) 6 �bd2 is the only move for people
The ... i.. a6 system has gained much popu who want to play. Kholmov-Naumkin,
larity over the last decade and is now the Voskresensk 1 990 went 6 ... i..d6 7 b3 0-0 8
primary system for Black among top players. i.. d3 i..b7 9 'ii'c2 �bd7 1 0 i..b2 cS 1 1 0-0
I think this has less to do with the actual l:lc8 12 l:lfel bS 13 dxcS i..xcS 14 'ii'b l with
worth of the move (both 4...i.. b7 and a draw, although I must say that as Black I
4... i.. a 6 probably lead to equal play) and would have considered playing on here.
more to do with the double-edged positions b) After 5 �bd2 Black plays 5 ... i..b 7! as
that we find in the critical lines. There are the new arrival on d2 is poorly placed.
simply more chances to outplay your oppo Rohde-Miles, USA 1 989 continued 6 'ii'c2 dS
nents in this variation. 7 cxdS exdS 8 g3 i.. d 6! (controlling eS and
36
better than the old 8 ... i.e7) 9 i.g2 tiJbd7 1 0 ble problem on the first rank and on b2.
0-0 0-0 1 1 lLlh4!? l:.e8 1 2 lLlc4?! (not good; 21 . . . �xg5 22 �xg5 'ilxg5 23 J:l.d 1 'fle5
12 tiJdf3 gives both players chances) 12 ... i.f8 Black has a pawn and very good winning
13 i.f4? lLle4! and Black was much better chances, which proved enough...
due to the threat of ... g7-g5. 24 b3 �g7 25 a4 'flf4 26 'ild7 + .U.f7 27
5 . . . g6! 6 tl\c3 i.g7 7 e4 0-0 8 i.e2 c5 9 'ild2 'ilf6 28 h3 e5 29 J:l.e 1 'i/f4 30 'ile2
d5 d6 e4 31 'ilb2+ �h6 32 'ilc2 J:l.e7 33 'ilc4
9 ... exd5 would be too soon. After 1 0 exd5 'ile5 34 'ilg8 'ilf5 35 'ild8 'ilf7 36 'ild2+
l:.e8 1 1 'it'c2 White might be better as the �g7 37 'ile3 'ild 5 38 'ilg3 'ild4 39 'ilh4
bishop is misplaced on a6. .U.e5 40 'ilg3 :ts 41 'flh4 :e5 42 'ilg3
1 0 0-0 tl\bd7 a6 43 �h2 :e7 44 'ilh4 'fle5 + 45 �g 1
Black is just developing and, as White has b5 46 axb5 axb5 4 7 'i'g4 c4 48 bxc4
no active plans, the position is equal. Now bxc4 49 'ile2 c3 50 'ilc2 :d7 51 :xe4
White goes over the edge in an attempt to do :d 1 +
something.
1 1 dxe6? ! fxe6 1 2 e5 dxe5 1 3 tl\g5 'fle7
1 4 �f3 e4!
Black will not allow White to occupy e4
with a knight without first opening up for his
bishop on the long diagonal.
1 5 tl\cxe4?
1bis seems to be bad, but 1 5 ltJgxe4
lLlxe4 1 6 i.xe4 lLle5 1 7 lLlb5 l:.ad8 would
also benefit Black.
1 5 . . . tl\e5 1 6 tbxf6+ i.xf6 1 7 i.xa8
�xc4 1 8 'ilg3 i.xf 1 1 9 �e4 i.d3!
52 'ilxd 1 'flxe4 53 'ild7 + �h6 54 'ilea

c2 55 'ilf8+ �g 5 56 h4+ �h5 57 'i'c5+
�xh4 58 �h2 g5 0-1
Game 10
Dzhandzhava-Chernin
Lvov, 1987
1 d4 tl\f6 2 c4 e6 3 tl\f3 b6 4 a3 �a6 5
'ilc2 �b7 6 tl\c3 c5 7 d5?
1bis should not be considered as a serious
alternative to 7 e4, and nor should 7 dxc5
A very important simplification. Now bxc5, e.g. 8 i.f4 lLlh5 9 i.g5 i.e7 10 i.xe7
White's hopes of an attack against the weak 'iixe7 l 1 e3 0-0 1 2 i.e2 lLlf6 13 0-0 d6 14
ened king disappear together with the minor l:.fd1 l:.d8 1 5 l:.d2 liJbd7 1 6 l:.ad1 and White
pieces. realised that he had no advantage at all in
20 i.xd3 tl\xd3 21 'ilxd3 Gavrikov-Yemelin, Tallinn 2000. Another
The last finesse is that 21 lLlxh7 'i'xh7! 22 game went 8 i.g5 i.e7 9 e3 d6 10 l:.d1 0-0
'iixd3 l:.d8! 23 'ii' e2 'ii'h 5 gives White a terri- 1 1 i.e2 tiJbd7 1 2 0-0 'it'b6 1 3 l:.d2 l:.ad8 14
37
Q u e e n 's In dian D e fe n c e
l:tfdl l:tfe8 1 S h3 lLif8! with the idea of White must get going. The alternatives are
... lL!g6 and ... h7-h6, winning the two bishops. hopeless. After 14 f4 .i.e71 1 S lL!f5 g6 there is
After 16 .i.f4 lL!g6 17 .i.h2 l:.d7 18 'ii'b l no compensation for the sacrificed material,
:ed8 Black was no worse in Portisch while 14 'ii'e4 'if e7! 1 S 'ifxa8 'ii'xe3+ 1 6 .i.e2
Timman, Reykjavik 1 988. lL!c6 17 'iWb 7 'ifd2+ 1 8 'ifi>fl 'ii'xd6 1 9 l:.d 1
lL!d4 is very close to winning for Black. The
domination of the dark squares and the mis
placement of White's king will decide the
game.
1 4 . . . ltic6? !
Here Black could have buried 7 dS once
and for all by simply playing 14 ...'i!i'e7!, e.g. 1 S
'ife4 'i!i'xd6 1 6 'ifxa8 lL!c6 1 7 'i!i'b 7 .i.xe3, or
1 S lLifS 'it'eS 16 e4 g6, when Black has a
pawn plus an overwhelming positional ad
vantage.
1 5 f4?
Here White could have fought back a little
7 . . . exd5 8 cxd5 tbxd5 9 i..g 5 bit with 1 S h4! he31 1 6 lLixfl :.Xfl 1 7
9 'ife4+ 'ife7 10 'ifxdS is a silly blitz line. .i.xfl+ �h8 1 8 .i.dS, and Black has no clear
Black just takes the money and runs. way to conduct the attack, although he has
9 . . . i..e 7 1 0 ltib5 0-0 1 1 ltid6 good chances. Now the struggle is effectively
over.. .
1 5 . . . i..xf4! 1 6 ltixf7
1 1 . . . ltie3! !
This was a serious new move at the time.
Now Black takes over the initiative. 16 exf4 'ife7+ 17 lL!e4 lL!d4 1 8 'ifd3 l:tae8
1 2 fxe3 19 .i.dS 'it'h4+ 20 'ifi>fl (20 g3 'it'hS 21 .i.b7
Practically forced as 12 'ifd3 .i.xf.3 1 3 dS) 20 ...'ii'x f4+ 21 'iti>gl 'if fS 22 .i.b7 dS 23
.i.xe3 .i.g4 1 4 'ii'e4 .i.xd6 1 S 'ii'xa8 'iff6 lLif2 'it'eS 24 g3 c4 2S 'ii'd l f5 and the attack
looks extremely dangerous for White. wins by itsel£
1 2 . . . i..xf3 1 3 exf3 1 6 . . . 1fh4+ 1 7 g3
13 .i.xe7 'ifxe7 14 lL!fS 'it'e4 and Black No alternatives: 1 7 �e2 'ifhs+ 1 8 g4 (1 8
wins. 'iltel dS� 1 8 ... 'ii'xg4+ 1 9 �d2 .i.xe3+! 20
1 3 . . . i..xg5 1 4 i..c4 �xe3 'ifd4+ 21 'ifi>e2 l:tae8+ 22 �ft dS and
38
1 7 cJ.?d 1 i.. x e3! 1 8 tllh6+ cJ.?h8 1 9 tll f7+ .l:t.xf7 8 . . . .i.c5!?

20 i.. x f7 tt::l d4 21 'iid3 'ii'g4+ 22 �el 'tixg2 Here the main line is 8 ... tt::l c 6, but the text
both win for Black. move and 8 ... d6!? have gained in popularity.
1 7 . . . .i.xg3+ 1 8 �d2 After 9 i.. e2 i.. e7 White has played:
Black would also win after 18 �e2 'ii'h5+ a) 1 0 i.. e3 a6 1 1 g4?! has been tried occa
1 9 cJ.?d2 i.. £2 ! (funnily enough Chemin makes sionally but it seems to be too dangerous for
the mistake of mentioning 1 9 ... d5 as winning, White in view of the weakening of his queen
but 20 hxg3! is worth a try!� 20 .l:t.afl i.xe3+! side with a2-a3. Dreev-Lemer, Vienna 1 996
21 �xe3 d5 and Black wins in the attack. continued 1 1 ...tll fd7 12 0-0-0 tllc 6 13 tt::lxc6
1 8 . . . .i.f2 1 9 �d 1 i..xc6 14 f4 'ii'b 8 15 g5 0-0 16 h4 b5 17 h5
1 9 'ii'd3 .l:t.xf7 20 i..xf7+ �xf7 21 'ii'xd7+ b4 18 axb4 'ii'xb4 19 g6 i.. f6 20 gxf7+ (20
tt::l e7 22 .l:t.afl .l:t.£8 and Black wins. gxh7+ �h8 21 i.. d4 i.xd4 22 .l:t.xd4 tllc5
1 9 . . . d 5 0-1 would also give Black a serious attack)
20 ... .l:t.xf7 21 i..d3 i.xc3 22 'ii'xc3 (22 bxc3
'ii'a3+ 23 cJ.?d2 i.. a4 24 'ii'c l 'ii'a2+ 25 i.c2
tllc 5 and Black is better) 22 ...'ii'xc3+ 23 bxc3
tll c5 24 i..xc5 dxc5 and Black was much
better and went on to win.
b) 10 0-0 0-0 1 1 i.. e3 a6 1 2 .l:t.fdl 'ii'c7 13
.l:t.acl tt::lbd7 14 £3 .l:t.fc8! 1 5 i.. fl 'ii'd 8 1 6 'ii'f2
.l:t.ab8 17 �hl i.. a8 1 8 'ii'g3 .l:t.c7 1 9 b4 'ii'e 8!
with equality in Ljubojevic-Piket, Monaco. In
these hedgehog positions a2-a3 tends to have
pros and cons. It provides White with the
opportunity to advance with a3-a4-a5 later to
generate an attack against a6, but it also
--------.. weakens the defence of a2-c4, and any pawns
Game 1 1 that might be placed on that diagonal.
Bacrot-Adams 9 ll'lb3 ll'lc6 1 o .i.g5
Cannes (rapid) 200 1 1bis game illustrates quite clearly that this
line is okay for Black. Here are the altema-
1 d4 ll'lf6 2 c4 e6 3 ll'lf3 b6 4 a3 .i.a6 5 rives:
'iWc2 .ll b 7 6 ll'lc3 c5 7 e4 cxd4 8 ll'lxd4 a) 10 i.. d3 0-0 1 1 0-0 h6!? 12 tt::lxc5 bxc5
13 i.. e3 d6 14 £3! .l:t.b8 15 .l:t.fdl 'ii'e7 16 i.fl
.l:t.fd8 17 .l:t.ab l ! with a very small advantage
to White in Sakaev-1\:filos, Sao Paulo 1 991 . In
Lputian-Adams, Wijk aan Zee 2000, 1 1 i.. f4
e5 12 i..g5 h6 1 3 i..h4 i.e7 14 0-0 tt::lh5 1 5
i...xe7 'ii'xe7 1 6 .l:t.adl tt::l f4 1 7 tt::l d5 'tig5 1 8
tt::lxf4.exf4 left Black slightly better.
b) 10 tt::lxc5 bxc5 1 1 i.d3 d6 12 0-0 0-0 13
h3 e5 1 4 'ii'd l h6 1 5 .l:t.bl a5 1 6 f4 exf4 1 7
i..xf4 tt::l e5 and Black seems to b e fine,
Khalifman-Salov, Amsterdam 1 995.
c) 1 0 i.. f4! is the real test of 8 ... i.. c 5, when
wild and gruesome is 1 0 ... 0-0 1 1 tt::lxc5 bxc5
39
1 2 i.. d6 o!iJd4 1 3 9d3 e5 14 i..x c5! (the ex 1 1 �xd4 J.xd4 1 2 �b5
change does not appear to be so interesting 1bis is the test, but as we shall see it seems
when all the remaining pieces are good for to be in Black's favour. 12 i.. d3 'itb8 and the
Black) 14 ...l:te8 15 i..xd4 exd4 1 6 'ir'xd4 control over the dark squares guarantees
ltJxe4 1 7 0-0-0 and White had the advantage Black equality.
in Piket-Salov, Wijk aan Zee 1 997. In Kram 1 2 . . . .tc5!
nik-Psakhis, Debrecen 1 992, 1 0 ... i.. e7 1 1 12 ... i.. e5 1 3 £4, with unclear play, was the
i.. e2 d6 1 2 l:tdl Wb8 was too slow, and main line before this game. But the bishop is
White had 13 c5 e5 14 cxd6 exf4 1 5 dxe7 better placed on c5.
ltJxe7 1 6 J.b5+ c;t>f8 17 0-0 with a small ad 1 3 e5 h6! 14 exf6
vantage. 1bis leaves 1 0 ... e5 1 1 i.. g5 h6 1 2 Also fine for Black is 14 i.. e 3 a6! 15 exf6
i..h4, and now Black has the following op axb5 1 6 fxg7 .l:.g8, e.g. 1 7 'ir'h7 �e7 1 8
tions at his disposal. 'ii'xh6 i.. xe3 1 9 'ir'xe3 bxc4 and Black ap
cl) 12 ... 0-0 13 f3 i.. e7 14 i.. £2 with advan pears to be better as g2 is weak. 17 i..xc5
tage to White according to Dautov. And trust bxc5 18 'ii'h7 c;t>e7 19 'ir'xh6 'ii'a5+ looks
me, he is right, although Black should hurry good for Black. White is severely behind in
with 1 4... a5!, with the idea of ... a5-a4 and development and the potential pressure
... ltJd4 and, perhaps, ... d7-d5 to use his lead against g2 is strong.
in development to blow the centre apart. 1 3 1 4 . . . hxg5 1 5 fxg7 l:.g8 1 6 'ifh7 </;e7 1 7
o!tJxc5?! bxc5 1 4 i.. d3 ltid4 1 5 'ir'dl a 5 1 6 0-0 b4 a6 1 8 bxc5 axb5 1 9 'i'h6 f6
g5 17 i.. g3 d6 18 :et i..c 6 was better for
Black in Christiansen-Seiravan, USA 1984.
c2) 12 ... i.. e7 13 0-0-0!? 0-0 14 f3 was bet
ter for White in Notkin-Kiselev, Cappelle la
Grande.
c3) 12 ... d6 13 0-0-0 a6 (1 3 ... 0-0 14 f3 a5 1 5
'ir'd3! gave White a serious advantage in
Akopian-Salov, Wijk aan Zee 1 993) 14 f3
l:tc8 1 5 @bt ltia5, Gruenenwald-Bischoff,
Bundesliga 1 988, and here White should
·
have played 1 6 ltixc5! bxc5 1 7 ltid5 with an

advantage.
1 0 . . . �d4!
White has attacked with his lone queen
and now his king is stuck in the centre. Black
now wins almost by force from here.
20 J.d3 J.xg 2 21 l:.g 1 'i'c7 ! 22 cxb5
22 .:r.xg2 'ii'es+ is easy.
22 . . . 'i'xc5 23 'i'h7 J.f3 24 l:.b1 l:.xa3 0-1
Game 12
Bareev-Eingorn
Kiev 1986
1 d4 lbt6 2 c4 e6 3 lbt3 b6 4 a3 J.a6 5
'ifc2 J.b7 6 �c3 c5 7 e4 cxd4 8 �xd4
40
lDc6 9 lDxc6 refutatioi:i is not 1 1 ...lDxe4?1, as played in

In practice this is the only move played by Lputian-Psakhis, Sochi 1 987, but 1 1 ....ie7!
the top players; there are no serious alterna 12 .1g2 lDxe4!1 13 .1xe7 lDxc3 14 .1xc6
tives. After 9 .i.e3 lDg4! Black gains the two Wes+ 15 'iit f1 dxc6 16 Ae1 'iid4 1 7 .ih4 h61!
bishops without any concessions. 18 g4 (1 8 'ifxc3 'i'xc3 19 bxc3 g5 20 Ae4
9 txc6
. . .. l:.d8! and Black will have a better endgame)
1 8 ...'ii'xg4 1 9 .1g3 "ii'xc4+ 20 �g1 l:td8 21
'ii'xc3 'ifxc3 22 bxc3 l:r.d3 23 l:r.e3 :Xe3 24
fxe3 rj;e7, and the three pawns look stronger
than the bishop. Instead 1 1 i.d3 .i.e7 1 2
'iie2 h 6 1 3 .id2, Klimov-Yemelin, St. Pe
tersburg 2000, looks like a tiny edge for
White.
1 0 lDh5
. . .
Whether this is the best move or not is

very difficult to tell. The alternative, 10 ... i.c5,
is treated in Game 13.
1 1 i.e3
1 0 i.f4
This is the main move, but not the only
serious one.
a) 10 .1e2 does nothing about the impor
tant dark squares. In fact White should now
be careful if he is to maintain the balance:
al) After 1 0 ... 'ifb8 we have 1 1 0-0?! .1c51
12 �h1 (1 2 b4 .i.d4 13 .ib2 h5 and the at
tack continues) 12 ... h5 1 3 'ifd3 lLJg4, when
Black was already much better in Fe
dorowicz-Miles, USA 1 989, while 1 1 .1e3
was equal in Christiansen-Miles, Linares
1 985. 1 1 .td2 'ii'h 8 12 g3 f5 13 .td3 .i. d6 14
a2) 10 ... .1c5! looks best. Then 11 0-0?! is 0-0-0 f4 gave Black good control over the
too dangerous. Now Black attacks on the dark squares in Polugaevsky-Arnason, Reyk
dark squares (but what should White do?): javik 1 987. Super-GM Curt Hansen evaluates
1 1 ...'ifc7 1 2 �h1 h5! 1 3 f3 h4 14 b4 .i.d4 1 5 the position as unclear, but I feel that Black
.id2 lDh5! with a decisive attack in prospect, should not fear such unclear situations.
Farago-Grooten, Sas van Gent 1 988. 1 1 i.c5?!
. . .
b) 10 .id3 'ii'h 8 1 1 'ife2! (no castling This is not the appropriate strategy. Kar
here!) 1 1 ... .id6 12 .id2 0-0 13 h3 .1f4 1 4 pov �veloped the correct approach over a
.1xf4 'ifxf4 1 5 0-0 d6, a s in Sak.aev few tries in the early 1 990's. The first was
Poluljahov, Vrnjacka Banja 1 996, should be 1 1 ....id6??, which ended quickly with 1 2
equal. 'ii'd 1 !, when Black resigned in Christiansen
c) 1 0 .1g51? might give White something. Karpov, Wijk aan Zee 1 993. But the second
It all comes down what happens after attempt, 1 1 ...'ii'h 8 !, is here to stay. The strat
10 ...'ifb8. While 1 1 g3?! is not very good, the egy involves a fight for f4 with .. f7-f5 and
.
41
Q u e e n 's In dia n D e. fe n c e
. . .i.f8-d6 coming. Black should equalise: c3) 12 ... tDf61 is probably the only good
a) 12 g3 fS! 13 0-0-0 (1 3 l:r.gl !? fxe4 1 4 move here. Then 13 f4 tlJg4! does not work,
tDxe4 i.xe4 1 5 'ii'xe4 lDf6 1 6 'ii' f3 i.cS 1 7 so Arencibia-Bischoff, Havana 1 998 contin
i.f4 eS 1 8 i.gS 0-0 with equality in San Se ued 13 i.e2 i.d6 14 g3 i.eS 1 5 l:r.hel 0-0 1 6
gundo-Langeweg, Zaragoza 1 995) 13 ... tDf6 f3 a6 17 rJi;bl :c8 1 8 i.d4 i.xd4 19 l:r.xd4
1 4 i.d3 �7 1 5 f3 (1 5 tlJdS fxe4 1 6 i.xe4 dS with equality.
exdS 17 cxdS tlJxdS does not work) 1 5 ... fxe4 1 2 ..txc5 bxc5 1 3 g3 0-0
1 6 tDxe4 tDxe4 1 7 i.xe4! (1 7 fxe4?! hinders 13 ... fS 14 0-0-0 f4 worked out well after
the bishop on d3; after 1 7 ... i.d6 Black was 1 5 g4?1 in Dreev-Supatashvili, Moscow 2001 ,
better in Lutz-Karpov, Dortmund 1 993) but White could have used hi s lead in devel
17 ... i.xe4 1 8 fxe4 i.e7 19 l:r.hfl 0-0-0 with opment more aggressively with 1 5 i.e2! 'i'gS
an unclear game according to Karpov. (1 5 ... lDf6? 16 gxf4 'fic7 1 7 eS! i.xhl 1 8
b) 12 l:r.dl validates 12 ... i.cS! 1 3 i.xcS l:r.xh l followed by tDe4-d6 and White is prac
bxcS 14 tlJbS 'ii'eS 15 tlJd6+ rJl;e7, when tically winning) 16 h4 'ii'eS 17 g4 tDf6 1 8 gS
White has the most problems with his king. tDxe4 1 9 i.hs+ g6 20 l:r.hel gxhS 21 tDxe4
Bykhovsky-Soffer, Tel Aviv 1 994 went 1 6 g3 i.xe4 22 l:r.xe4 'ii' fS 23 'ii'c 3, with an initia
l:r.ab8 1 7 :d2 gS 1 8 i.g2 f5 1 9 0-0 f4, with tive and the safer king position.
preference for Black. 1 4 ..td3 lLlf6 1 5 0-0-0 d6 1 6 f4 l:!.b8 1 7
c) 12 0-0-0 with a further branch: l:hg 1 e5 1 8 g4
cl) 1 2... i.cS 13 i.xcS bxcS 14 g3 0-0 1 5
f4 d 6 1 6 i.g2 tlJf6 1 7 l:r.d3 l:r.c8 1 8 l:r.hdl
tDe8 19 g4!, Gurevich-Korchnoi, Biel 1 993,
with aggressive ideas, as g4-g5 and l:r.h3 look
very good for White.
c2) 12 ... i.d6 13 g3 i.eS (1 3 ... fS 14 i.e2
tDf6 15 exfS! i.xhl 16 l:r.xhl 'ii'b7 17 l:r.dl
i.e7 18 g4 0-0 19 gS tDe8 20 i.d3 illustrates
how the light squares can suddenly become
so important; White is clearly better) 14 i.d3
'iib7 (1 4... tlJf6 1 5 f4! i.xc3 1 6 'ii'xc3 'ii'b 7
[or 16 ... i.xe4 17 i.xe4 tDxe4 1 8 'ii'xg7 with a
clear advantage, and the same goes for
1 6 ... 0-0 1 7 i.d4U 17 l:r.hel and White looks Here White's attack seems to be the most
much better) 1 5 l:r.hel tDf6 and now, instead promising, as none of Black's minor pieces
of 1 6 f4, as in Kramnik-Lutz, Dortmund seems able to help on the kingside.
1 993, White could have claimed a large ad 1 8 . . . lbd7 1 9 f5 'ifa5 20 i..e 2 J:l.b7 21
vantage with 16 i.f4! d6 (1 6 ... i.xf4+ 17 gxf4 J:l.xd6 J:l.fb8 22 J:gd 1 J:l.xb2 23 Vxb2
d6 1 8 eS dxeS 1 9 fxeS tlJd7 20 tDe4 and J:l.xb2 24 'iPxb2 'ifc7 25 g5 'iPf8 26 f6
Black is in deep trouble) 1 7 i.xeS dxeS 1 8 gxf6 27 gxf6 'iPe8 28 'iPa2 a6 29 i..h 5
i. fl 0-0 1 9 f3 l:r.fd8 20 'ii' £2 . This structure is 'ifb7 30 lbd5 1 -0
simply very bad for Black. Once all the heavy
pieces are traded off the endgame will give Game 13
White a strong passed pawn on the queen Lautier-Timman
side and, potentially, pressure against the eS Amsterdam 1996
pawn, or e6 if the first is protected with . . f7-.
f6. 1 d4 lLlf6 2 lbt3 e6 3 c4 b6 4 a3 ..ta6 5
42
Th e P e tro sian Sys tem: 4 a 3
1"c2 .i.b7 6 l0c3 c5 7 e4 cxd4 8 l0xd4 i.b4 i.e5 1 6 i.f3 £5 , with strong compensa
l0c6 9 l0xc6 .i.xc6 1 0 .i.f4 .i.c5 1 1 .i.e2 tion. For more cautious players I recommend
13 ... .l:.e81 14 e5 (1 4 i.d3 e5 looks even better
for Black) 14 ... i.xc3 1 5 'ii'xc3 leie4 with
equality.
1 2 . . . lOeB
The best move. White does not want to

allow 11 0-0-0 leih51, when he has no advan
tage after 1 2 i.d6 i.xd6 1 3 !txd6 0-0 14 g3
£5 1 5 i.g2 (1 5 f4?! fxe4 1 6 leixe4 i.xe4 1 7
'i'xe4 ltc8 i s even more dangerous for Also possible are one decent alternative
White) 1 5 .. .f4 1 6 !thd1 'i'g5 17 'i'd2 'i'e5, as and two mistakes:
in Greenfeld-Kindermann, Pardubice 1 994. a) 1 2 ... ltc8 13 'it>bl a5 (1 3 ... leie8 14 i.g3!?
1 1 . . 0-0 1 2 0-0-0
. f6 15 b4 i.e7 16 i.g41 leic7 17 b5 i.a8 1 8
This is by far the most promising con 'i'd3 i.xa3 1 9 'ii'xd7 'i'xd7 20 !txd7 !t f7 21
tinuation. After 12 .:td1 a5 Black appears to !thd1 with a clear advantage for White) 1 4
be doing all right. Play might develop as fol i.g3 leie8 1 5 !thel 'iie7 1 6 i.d3 f6 17 f4
lows: 13 i.g5!? (1 3 i.g3!? 'i'e7!? 14 e5 leie8 �h8 1 8 i.£2 should supposedly give White a
15 leie4 £5 1 6 exf6 lDxf6 1 7 i.d3 i.xe4 1 8 small advantage, as in Greenfeld-Yu, Beijing
i.xe4 leixe4 1 9 'i'xe4 and Black cannot be 1 996, but it does not appear to be that clear,
that much worse, if at all, Gurevich does it?
Chuchelov, Germany 1 995) 13 ... h6 14 i.h4 b) 1 2... e5? 13 i.xe5 leig4 14 i.xg4 'ii'gs+
e5? (Black is okay after 1 4 ...'ii'c7 1 5 i.xf6 1 5 'it>bl 'ii'xg4 1 6 leid5 i.xd5 1 7 cxd5 ltac8
gxf6 1 6 ltd3 'it>h7) 1 5 0-0 'ii'e7 1 6 �h1 g5 17 1 8 f3 'iig6 19 i.c3 left White with an extra
i.g3 l:.ab8 1 8 i.f3 i.d4 1 9 leie2 i.c5 20 pawn in Krasenkov-Hellsten, Malmi:i 1 995.
leic3 i.d4 21 ltfe1 g4 and now instead of 22 c) 1 2 ... i.xf2?! 1 3 i.d6 l:.e8 14 e5 i.xg2 1 5
i.e2?, as in Dreev-Adams, London (rapid) lthg1 i.xg1 1 6 !txg1 i.b7 1 7 °iid3 l:r.c8 1 8
1995, White could have won with 22 leid511 'iig3 g6 1 9 'ii f4, with a very dangerous posi
i.xd5 23 exd5 'ii'c5 (23 ...gxf3 24 ltxd4 and tion for Black (Piket).
Black is on the wrong side of a strong attack) 1 3 .i.g3
24 'ii' f5 gxf3 25 'it'xf6 fxg2+ 26 �xg2 'iixc4 1his appears to be the strongest. The al
27 !txd4 'iixd4 28 i.xe5 'ii'g4+ 29 �h1 �h7 ternative 13 �b1!? is probably not best met
30 ltg1 etc. (Vaisser). with 13 ...f6 14 i.g3 e5 1 5 f4 i.d4 1 6 leib5!
12 e5 meets with 12 ... leih5! 13 i.xh5 'i'h4 i.xb5 1°7 cxb5 ltc8 18 'iia4 leic7 19 ltc1 ,
14 i.xf7+ l:r.xf7 1 5 i.g3 'ii'xc4, while in the with an advantage to White in Greenfeld
vent of 1 2 0-0 Black has 1 2 ... i.d4!? with the Tunk, Beersheva 1 996. Instead Black has
wild idea of 13 i.d6 Lc3 1 4 i.xf8 leixe4 1 5 1 3 ... e5 1 4 i.xe5 i.xf2 1 5 i.d3 f61? 1 6 i.f4
43
i.d4 t 7 :lift with close to equality, or even 1 998.

t3 ... i.xf2! t4 i.d3 i.c5 t 5 e5 f5 and the 1 7 l:.xd4! ? exd4 1 8 l:.d 1 d 5 ! ?
. pawn sacrifice gives White compensation, Black i s trying to complicate the position.
but not too much. t 8 ...1Wc5 t9 b4 'ifxc2+ 20 'ifiiixc2 :tc8+ 2t
1 3 . . . e5 'ifiiib 2 it)f6 22 .l:lxd4 .l:lfe8 23 i.d3 gives White
a small advantage in the endgame due to the
two bishops.
1 9 exd5 ltid6
t 9 ... it)f6 20 d6 'ifd7 2t .l:xd4 is just very
good for White.
20 .1d3! h6?
A grave error. After 20 ... :tfc8! 21 'i'e2 (2t
'i'a4 .l:lc5!) 21...1i'xe2 22 i.xe2 the chances
are even (Lauri.er) .
21 1ka4 1kf6?!
Black continues to hesitate about seizing
the open file, and soon it is too late. After
2t ...:ac8 22 'iixd4 .l:lfe8 23 'i'f4 l:ed8 24 h4
Here you can also take the direct route to :c5 25 a4 White is slightly better.
the insane asylum with t 3 ... f5!? t4 exf5 exf5 22 l:.c 1 ! ltif5?
t 5 £3 'ii'g5+ t 6 'itiiib t f4 t 7 i.d3 fxg3 t 8 Still not doing too well.
i.xh7+ rif fl t 9 f4 1Wf6 20 hxg3 it)d6 2 1 g4 23 .1c7! ltih4? 24 f4! ltixg 2 25 .1e5 1kh4
'iiiie 7, and who can tell what the hell is going 25 ... 'ifd8 26 'iixd4 f6 27 'iie4 wins for
on here without scanning all the other games White immediately.
in the database and spending two weeks with 26 1kxd4 ltie 1
this position?! Piket-Timman, The Hague 26 ... f6 27 'ife4 fxe5 28 'i'h7+ riffl 29
t 995 ended in a draw. .l:lc7+ 'ifiiie 8 30 'ifg6+ and White mates.
1 4 �b1
t4 i.xe5 'ifgs+ t 5 f4 i.e3+ t 6 ritbt i.xf4
t 7 i.d4 i.e51? with equality according to
Tiviakov.
1 4 . . . .1d4 1 5 ltib5 .1xb5 1 6 cxb5 1ke7
Here I prefer t6 ... :c8 t 7 1Wa4, when
Black has a couple of options. t 7 ... it)f6 t 8
i.h4 h 6 ( t 8. . .d 5 t 9 f4 .l:lc5 was played in
Watson-Browne, USA t 996, and now White
could probably have gained an advantage by
bringing his last piece into play with 20
:lift !) t 9 f4 g5 was Van Wely-Adams, Wijk
aan Zee t 998, and now White would have
the advantage after 20 i.g3! it)xe4 (20 ...gxf4 2 7 1kg 1 !
2t hf4 it)xe4 22 i.xe5 i.xe5 23 'ii'xe4 27 i.xg7?! meets with 27 ... it)£3!, when
appears even more dangerous) 21 fxe5 it)xg3 matters are less clear.
22 hxg3 i.xe5 23 l:txh6. Alternatively 27 . . .f6 28 1kxe 1 ? !
t 7 ...1We7 t 8 Act it)d6 t 9 i.d3 h5 20 h4 led Here Lauri.er later felt his queen was better
to equality in Van Wely-Timman, Breda placed and that 28 :Xet fxe5 29 fxe5 would
44
Th e P e tro s ia n Sys tem: 4 a 3
have won easier as both the pawns march i.d3, e.g. 8 ... d5!? (8 ... i.xd3 9 'Wxd3 d6 1 0 b4
forward and the idea of .l:tel -e2 followed by c6 1 1 0-0 0-0 12 e4 ltJd7 13 i.f4 gave White
'Wgl -g6-h7 seems to have deadly powers. a safe plus in Browne-Trois, Buenos Aires
28 . . .'i'xe 1 29 J:l.xe1 fxe5 30 fxe5 �f7 3 1 1 979) 9 i.xe4 dxe4 10 liJd2 £5 1 1 £3 i.d6!?
h4! g 5 (1 1 ...c5?! 12 fxe4 cxd4 13 'Wa4+! �fl 14 liJ£3
31 ....l:tad8 32 d6 �e6 33 i.c4+ �d7 34 favours White, as does 1 1 ...e5 12 fxe4 exd4
e6+ �xd6 35 e7 and White wins. 1 3 exf5 0-0 14 e4) 1 2 'Wa4+ c6?! (1 2 ... liJd7!
32 hxg5 hxg5 33 .�.f5 J:l.ad8 13 fxe4 'Wh4+ 14 �dl 'Wg4+ 15 �c2 'Wxg2
33 ...�g7 34 i.d7! followed by the e 1 6 .l:tfl also gives White a grip on the light
pawn's march forward wins for White. squares, but still it was a fighting chance -
34 d6 �g7 35 .i.g4 .1:1.hS now White is just much better) 1 3 0-0 'Wh4
35 ... .l:tf4 36 e6! .l:txg4 37 e7 and wins. 1 4 f4 0-0 1 5 c5! bxc5 (1 5 ... i.c7 1 6 cxb6
36 �c2 J:l.h4 37 J:l.e4 J:.xg4 i.xb6 17 ltJc4 is also clearly better for White)
37 ... �fB 38 �c3 .l:th2 39 b3 �e8 40 a4 1 6 liJc4 'We7 (1 6 ... .ie7 1 7 dxc5 liJd7 1 8 b4
and the king marches to the centre to assist and Black has no compensation for his struc
the pawns forward. tural weaknesses) 17 dxc5 i.xc5 1 8 b4 i.d6
38 J:l.xg4 �f7 39 J:l.xg5 �e6 40 �d3 J:l.h8 19 i.b2 with a clear advantage for White in
41 �e3 J:l.h 1 42 J:l.g8 1 -0 Kramnik-Vaganian, Horgen 1 995.
6 'i'c2 .i.xf3
Game 14 This is the only move that makes sense.
Gershon-Anastasian After 6 ... .ig7?! 7 e4 The bishop is not very
Saint Vincent 2000 good on b7. 7 ... d5 (7 . . . 0-0 8 .ig5!? h6 9 i.e3
d6 1 0 h3 liJbd7 1 1 .ie2 was very good for
1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 a3 .i.b 7 5 White in Dreev-Sorokin, St.Petersburg 1 993)
lbc3 8 cxd5 exd5 9 e5 ltJe4 10 .id3 £5 (White also
had a solid advantage after 10 ...liJxc3 1 1 bxc3
0-0 1 2 0-0 c5 13 .ig5 'i'c8 14 'Wd2 i.a6 1 5
i.xa6 liJxa6 1 6 a4 .l:te8 1 7 i.h6 in Farago
Podlesnik, Bled 1 996) 1 1 exf6 liJxf6 1 2 'iie2+
'iie7 1 3 'ii'xe7+ �xe7 1 4 0-0 .l:te8 1 5 i.f4
�d8 1 6 liJb5 and Black is in trouble,
Christiansen-Schroll, Vienna 1 99 1 .
7 exf3
Also good is 7 gxf3 liJc6 8 'Wdl !? i.g7 9
.ig5 0-0 10 f4 d5 1 1 cxd5 exd5 1 2 i.g2 liJe7
1 3 b4, when White had the advantage in
Garcia Ilundain-Epishin, Manresa 1 995. Ro
mero Holmes writes that White is better in
5 . . . g6 view of the superior bishop and the weak
This is the most serious alternative to ness Qf c6.
5 ... d5, although Black hardly ever equalises. 7 . . . .i.g7
The other option worth mentioning is This move has to be played, so just get it
5 ... liJe4!?, a variation I have tried without any out. After 7 ... liJc6 8 i.e3 .ig7 9 0-0-0 ltJe7
pleasure. Black has no real chance to fight for (9 ... 0-0 10 d5 liJa5 1 1 h4 a6 12 g4 was good
equality. A good, solid way for White to gain for White in Dreev-Korchnoi, Yalta 1 995) 1 0
an advantage is 6 liJxe4 i.xe4 7 e3 i.e7 8 g4 d 5 1 1 h4 h6 White had the upper hand in
45
Slobodjan-Speelman, Lippstadt 2000. Black is trying to defend. Active counter

8 .i.e3 play with 1 3 ... c5 14 hxg6 hxg6 1 5 dxc5 l:r.c8
Tbis is the most nonnal move. The alter invites 16 g5, undermining the defence of the
natives just seem to be too premature. 8 g4 centre.
0-0 9 i.e3 d5 1 0 g5 ltle8 1 1 cxd5 exd5 12 f4 14 hxg6 hxg6
ltlc6 13 ltle2 ltle7 14 ltlg3 c5 1 5 dxc5 d4 14 ... fxg6 15 g5 ltlh5 16 f4 and White has
gave Black good counterplay in Kozul the advantage due to the weak light squares
Romanishin, Yerevan 1 996, while 8 d5 0-0 9 in the centre.
i.e2 exd5 1 0 cxd5 c6 1 1 dxc6 ltlxc6 1 2 'iia4 1 5 .i.d3
a6 1 3 0-0 b5 14 'ii'h4 ltlh5 1 5 i.g5 i.f6 was 15 i.e21? makes more sense in some ways
certainly not worse for Black in Lalic because it monitors the h5-square. But from
Romanishin, Germany 1 996. d3 the bishop might be sacrificed on g6 later!
8 . . . 0-0 9 h4 1 5 . . . c6
Here it is actually most natural to castle Before he can think of any counterplay
first. In the following game Black chooses a Black has to protect the d5-pawn
good set-up compared to the main game: 9 1 6 l::r. h 3!
0-0-0 d5 1 0 g4 dxc4 1 1 i.xc4 c6 1 2 h4 ltld5 16 i.h6? i.xh6+ 1 7 l:.xh6 ltlxg4! gives
1 3 ltle4 ltld7 14 h5 'iic7 1 5 hxg6 hxg6 1 6 Black some interesting opportunities. For
i.xd5 exd5 1 7 ltlc3 'iid6 with a good game this there is no reason.
for Black, Huzman-Van der Wiel, Pula 1 997. 1 6 . . . lbc4 1 7 .i.g5 'ii'd 6
1 0 cxd5 ltlxd5 11 ltlxd5 'iixd5 1 2 i.c4 'iid6
13 d5 e5 would be the critical position. I
prefer White due to the control over the light
squares.
9 . . . �c6
Here Black has 9 ... d5!? 10 cxd5 ltlxd5 1 1
ltlxd5 'iixd5 1 2 i.c4 'iid6, with a slight edge
for White.
1 0 0-0-0 d5
18 .txg6! ?
Tbis sacrifice does present Black with
more problems than most players can solve
at the board. But it does seem that as White
does not have anything directly winning that
it is not the objectively best continuation. 1 8
l:tdhl ltld7 1 9 ltle2 followed by ltlf4 is also
very dangerous for Black.
1 8 . . . fxg6 1 9 'ii'x g6 �f8
1 1 cxd5! exd5 Better is 1 9 ... ltld71 20 'ii'b.7+ <i;fl 21 l:r.h6
1 1 ...ltlxd5 1 2 ltlxd5 'iixd5 1 3 i.c4 'iid6 l:r.e6 22 'i!ifS+ ltlf6 23 'iig6+ �£8 24 l:r.dh1
14 d5 and now the knight is misplaced on c6. l:r.ae8 25 l:th8+ ltlg8 26 'i!ifS+ ltf6 27 i.xf6
1 2 g4 �a5 1 3 h5 :ea 'iixf6 28 l:t8h5, when White has some advan-
46
Th e Pe tro sian Sys tem: 4 a 3
tage. With a mating attack. Black resigned.

20 :dh 1 l:.e6
20 ...liJg8 21 'iWf5+ liJf6 22 i.h6 l:te7 23 g5 Game 15
and Black is just dead lost. 20 ...'iWe6 seems to Neverov-Stefanova
lose quite quickly to pressure on the 6th rank Rry�javik 2002
with 21 l:th6! etc.
21 .bf6 'iif4+ 1 c4 e6 2 d4 lllf 6 3 lllt3 b6 4 a3 i.b 7 5
21...l:txf6 22 l:th8+ i.xh8 23 l:txh8+ <Ji;e7 lll c 3 d5
24 'iWg7+ l:tf7 25 'iWg5+ 'iWf6 26 'iWxf6+ l:txf6
27 l:txa8 l:txf3 28 l:txa7+ <Ji;e6 and Black has
some good chances of a defence, although
White has an obvious material advantage.
22 cJi>b1 :xt6
22 ... 'iWxf6 -see the note to 21 st move.
23 l:.hS + St.xh8 24 1.txhS+ @e7 25
'iih 7 + ! l:.f7 26 'iih 4+ @d6?
26 ...'iWf6! -see note to move 21 .
27 l:.xa8 'iid 2 28 'iid S+
Here moves were made in order to win
time, as the game was being played with time
increments.
28 . . . �e6 29 'ii e S+ �f6 30 'iih S+ 'iit e6 1bis is, of course, the main move here.
31 :ea + �d7 32 l:.dS + <ot>e6 33 'iie S+ Now White has a variety of choices, which
<ot>t6 34 'iih S+ �e7 35 l:.e8 + ? will be discussed in the following seven
Still not finding the right path. Luckily the games. The first o( them is the popular...
win did not go away. 6 St.g5
35 . . .�d7 36 l:.dS + 'iite 6 37 'iie S + <ot>t6 ...which should give White an advantage.
38 llle4+ ! 6 . . . St.e7
6 ... dxc4!? also appears to be fully playable.
After 7 e4 i.e7 8 i.xf6 i.xf6 9 e5 i.e7 1 0
i.xc4 0-0 1 1 0-0 c 5 1 2 d 5 exd5 1 3 i.xd5
i.xd5 14 ltJxd5 ltJa6 15 'iWb3 ltJc7 1 6 l:.fdl
ltJxd5 17 l:.xd5 'iWc8 1 8 l:.adl 'iWe6, as in Van
Wely-Korchnoi, Wijk aan Zee 1 997, Black
cannot be much worse, although White still
dictates events.
7 'iia4+ ! ?
1bis was the major line in the 1 990's. The
key idea is that after 7 ... c6 8 cxd5 exd5 9 g3
0-0 1 0., i.g2 White has gained a slightly im
proved version of the 4... d5 line. Here White
Addressing the ltJc4 problem and there has had many successes in practical play, but
fore winning the game. experience has shown that Black completely
38 . . . dxe4 39 'iix c6 + 'iitg 7 40 'iixc4 exf3 equalises after...
41 'ii c8 'iixf2 42 l:.gS+ 'iith 7 43 l:.hS+ 7 . . . 'iid 7! 8 'iic 2
�g7 44 'iig S+ 'iitf 6 45 :th6 + 1 -0 1bis is what Gelfand invented when Black
47
found sufficient defences after 8 'i'xd7+ White in Gelfand-Karpov, Dortmund 1997,

tDbxd7 9 lDb5 .id8! (the most logical, and when Black was still far from finishing his
. also the height of fashion), e.g. 1 0 cxd5 (10 development. Here 11 0-0-0?! meets with
.if4? dxc4 11 tDxc7+ j_xc7 1 2 j_xc7 0-0 13 1 1 ...cxd4 1 2 :Xd4 1i'c8 13 llh4 h6 etc. In
llc1 .l:.fc8 1 4 .if4 b5 and Black is better. The stead 9 ... .ixf3! 10 gxf3 b5 is best, when 1 1
control over the centre and the queenside .ixf6 hf6 leaves White to make a choice.
majority is far more important than the dark 1 2 'i'e4 0-0! 13 'i'xa8 tDc6 14 'i'x£8+ (14
squares, Van Wely-Gelfand, Monaco 2001) 'if'b7 .l:.b8 15 'i'a6 .l:.b6 and the queen does
10 ... lDxdS 11 e4 lD5f6 12 .l:.c1 with a draw not escape) 14 ...cJi;xfB 15 lDxb5 'i'd5 sees
agreed in Dreev-Anand, Moscow Knockout Black win the £3-pawn, with dangerous
2001. Then 12 ... 0-0?1 13 lDxc7 .l:.c8 14 lDb5 threats on the dark squares.
.l:.xc1+ 15 .ixc1 leads to the win of a pawn This leaves 12 a4 c6 1 3 axb5 cxb5 14 'ii'e4
for White, but Bacrot-Gelfand, Leon 2001 0-0 1 5 f4! (White is playing for structure;
continued 12 ... .ixe41 13 tDxc7+ .ixc7 1 4 after 15 .l:.a5 b4 16 tDa4 .id8! 1 7 lDcS 'i'd5
:Xc7 lDd5 1 5 l:l.c1 h6 1 6 .id2 .ixf3 1 7 gxf3 1 8 l:l.a4 tDc6 1 9 .ig2 j,, b6 20 f4 .l:.fd8 Black
cJi;e7, and the question here is whether is ideally placed according to Gelfand), and
White's two bishops can compensate for the now time has validated 1 5 ... .l:.c8! (1 5 ...tDc6,
poor pawn structure. Perhaps the answer is as in Gelfand-Lautier, Biel 1 997, has also
yes, but Black appears to have full control as been played). Now White has more than one
c7 is his for keeps, meaning White canno t possibility:
threaten him. a) 1 6 'ii'xa8 tDc6 1 7 1i'xc8+ 1i'xc8 1 8 .ig2 a6
8 . . . dxc4 19 0-0 g6 and the primary factor in the posi
tion is Black's future on the queenside.
b) 1 6 .ig2 tDc6 1 7 0-0 (1 7 lDxb5? lDb4 1 8
'i'b7 'i'd8 and White i s not well co
ordinated, nor fully developed) 17 ... .l:.ab8 1 8
.ih3 shows that White can eam no advan
tage from 1 6 i..g2. Bacrot-Gershon, Ber
muda 1 999 continued 1 8 ... b4 1 9 lDa4 lld8 20
lDc5 'i'd5 21 .ig2 tDxd4 22 exd4 .ixd4 23
lDa4 1/2-1/2.
c) 1 6 .ih3!? attacks on the light squares,
where White is better. All other moves are
harmless. Piket-Khalifman, Wijk aan Zee
2002, went 16 ... lDc6 17 tDxb5 .l:.ab8 18 tDc3
9 e4? ! .l:.xb2 1 9 0-0 .l:.b3 with a very promising posi
This is an opening experiment we will not tion for Black, but best seems 19 ... lDxd4!
see Gelfand try again. White permanently with the idea of 20 exd4 'i'xd4 21 'i'xd4
sacrifices the c4-pawn and hopes to generate .ixd4 22 .l:.ac1 ? (otherwise three pawns and
an initiative in the centre. I feel that it is the better structure should be important, too)
doubtful whether he can prove any substan 22 ....l:.b3 and Black wins. 1 7 d5 lDd8 1 8 0-0 is
tial compensation for the pawn. But the main a suggestion - without an evaluation - of
line here has, by now, been more or less es Gelfand. Black should be fine if he chooses
tablished as equalising for Black. After 9 e3 to play 18 ...b4 1 9 lDa4 exd5 20 .ixd7 dxe4
Black needs to go for the pawn, as 9 ... 0-0?! 21 .ixc8 l:l.xc8, when the passed pawns are
10 .ixc4 c5 1 1 dxc5 was somewhat better for of paramount importance, and White is
48
Th e Pe tro sian Sys te m : 4 a 3
struggling for survival. passed pawn on dS, while Black's develop

9 . . . b5 1 0 l:l.d 1 a6 1 1 .ie2 h6 1 2 .th4 ment difficulties remain.
Wd8 1 3 0-0 0-0 14 ltie5 1 8 . . . f5 1 9 ltie3 f4 20 ltlt5 Wg5 2 1 ltid6
ltie5 22 f3 l:l.a7 23 a4 Wd8 24 ltice4
Wb6 25 g3 ltied7 26 �h 1 ltif6 27 gxf4
ltibd7 28 ltif5 ltixd5? !
28 ...lDhS looks better, with a complex
struggle ahead. Now White slowly takes over.
29 l:l.xd5 .ixd5 30 ltie7 + �h8 3 1 ltixd5
We6 32 l:l.d 1 l:l.f5 33 ltidc3 b4 34 l:l.d6
Wf7 35 ltid5 c3 36 i.c4 Wh5 37 �g2
l:l.f8 38 bxc3 ltie5 39 We2 ltixc4 40
Wxc4 bxc3 41 Wxc3 l:l.b 7 42 ltig3 Wh4
43 l:l.e6 J:l.d7 44 We5 Wd8 45 J:l.xh6+
1 4 . . . c5?
Here Black loses the chance to gain a clear
opening advantage. After 1 4... lbc6! White
has no better move than the exchange of
knights, as d4 is hanging. After the trade
Black needs slightly less space in which to
manoeuvre, and the pawn grows in stature.
1 5 d5! exd5 1 6 exd5 ltifd 7 ? !
1 6 ...lbxdS!? 1 7 .i.g3 lbxc3 1 8 l:txd8
lbxe2+ 1 9 'ii'xe2 i.xd8 with a situation that
is very difficult to assess. Black is, in theory,
better placed, with the large majority on the 4 5 ... �g8 46 We6 + J:l.ff7 47 J:l.h5 Ad6 48
queenside, but there will be considerable Wh3 J:l.h6 49 J:l.xh6 gxh6 50 We6 'i'f8 5 1
problems completing development. f 5 c 4 5 2 f 6 �h7 5 3 We4+ �h8 54 'i'xc4
1 7 .txe7 Wxe7 1 8 ltig4 'i'b8 55 We4 a5 56 'i'g6 J:l.b7 57 ltie7
J:l.b2+ 58 �h3 Wts 59 ltlgt5 1 -0
Game 16
Piket-Anand
Monte Carlo (blindfold) 1997
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 ltic3 .ib7
5 a3 d 5 6 Wc2!?
1hls move has been known ever since
1 950, but it was only when Gelfand used it
against Karpov in 1 992 that it became fash
ionable. As with the 6 i.gS line its span in
the great circulation of modem theory was
Here White finally has a very interesting short, as Black quickly found the right way to
position. For the pawn he has a strong respond. In this game we shall see two such
49
examples. c) 9 lDxd4 has become the critical move,

although Black has nothing to fear after
9 ... tbxd5 1 0 lDxd5 i.xd5 1 1 e4 i.b7, e.g. 1 2
i.e3 a 6 1 3 f3 b5 1 4 'ii'd l i.e7 1 5 i.e2 0-0 1 6
0-0 l:.c8 1 7 'ii'e 1 i.g5 1 8 i.£2 tbe5 19 l:. d1
'iie 7 and Black had equalised so effortlessly
that he became too ambitious in Dreev
Karpov, Cap d'Agde 2000. Cramling-Xu Jun,
Yerevan 1 996 went 12 i.a6 i.xa6 13 'ii'xa6
i.c5 14 i.e3 0-0 15 0-0 tbf6 1 6 l:.adl 'ii'c 8!
with equality.
6 ... i.e7 is dealt with in the next game.
7 e4 c5 8 d5
8 dxc5 seems to be rather harmless. After
6 . . . dxc4 8 ... i.xc5 9 i.xc4 lbbd7 10 0-0 'iic7 1 1 i.d3
This is the current trend, but I prefer a6 12 i.g5 tbe5! Black already has a pleasant
6 ... c5, although it is a matter of taste. How position. In fact in Sokolov-Polugaevsky,
ever, not good for Black is 6 ...lbbd7?! 7 cxd5 Holland 1 994 the sequence 1 3 tbxe5 'i'xe5
tbxd5 8 tbxd5. Then 8 ... exd5 9 i.g5 gives 14 i.h4 lDh5! 1 5 'i'a4+! �f8 1 6 'ii'd7 g5!
White a small edge, but 8 ... i.xd5? 9 e4 i.b7 gave him a great position.
10 i.b5! c6 (otherwise i.c6 with a great ad 8 . exd5 9 exd 5 a6!
. .
vantage) 1 1 i.xc6 l:.c8 1 2 d5 'ii'c7 is a differ

ent story. Now instead of 1 3 tbd4? as in Gel
fand-Karpov, Moscow 1 992, White should
play 1 3 i.e3! i.xc6 14 l:.cl exd5 (1 4 ...lbb8
15 dxc6! tbxc6 16 0-0 e5 17 b4 and Black is
still struggling to complete his development)
1 5 exd5 lbb8 1 6 dxc6! 'ii'xc6 1 7 'ii'd l 'ii'e6 1 8
'ii'a4+! with a lasting initiative.
Returning to 6 ... c5!, this is the tactician's
preference. After 7 cxd5 cxd4 8 'ii'a4+ lbbd7
White has three logical options (of different
value):
a) 9 'ii'xd4?! i.c5 1 0 'i'a4 exd5! 1 1 i.g5
0-0 1 2 e3 i.e7! and White needs to develop This is the way they are doing it these
quickly, but chose not to in Gofshtein days. After the passive 9 ... i.d6 10 i.g5 0-0
Altennan, Israel 1 997. 1 1 0-0-0 lbbd7 12 i.xc4 'iib 8 13 �bl a6 14
b) 9 dxe6 dxc3 10 exd7+ 'ii'xd7 11 'ii'xd7+ tbe4 tbxe4 1 5 'ii'xe4 l:.e8 1 6 'ii'g4 tbf8 17
tbxd7 12 bxc3 i.e7 13 i.e3! l:.c8 14 i.d4 i.d3 i.c8 18 'ii'h 5 'ii'c 7 1 9 l:.cl White had a
0-0 1 5 e3 tbc5 16 tbd2 (1 6 i.e2? lbb3 1 7 little pressure in Lautier-Karpov, Monte
l:.bl [17 l:.a2? i.d5 and Black is getting some Carlo 1 996.
action] 1 7 ... tbxd4 1 8 cxd4 i.xa3 and Black is 1 0 .i.g5! ?
better due to the pawns and bishops) The beginning of a wild stream of compli
1 6 ...lba4! 17 c4 i.c5 with good compensa cations that ends with an equal position.
tion for the pawn, Lputian-Dautov, Budapest 10 i.xc4 b5 1 1 i.a2 i.d6 1 2 i.g5 0-0 1 3
1 996. 0-0 lbbd7 1 4 l:.adl 'iic 7 1 5 i.bl l:.fe8 and
50
Black was in no way worse in Avrukh while White often prefers 8 e4, transposing
Anand, Haifa 2000. Instead 1 1 ...i.e7? is poor to 7 ... lL!xc3) 7 ... lLid7 8 'ii'c2. Now 8 ... cS
since the bishop belongs in front of the d seems safe enough, albeit a little boring.
pawn. Sakaev-Veingold, Moscow 1 994 went Akopian-Gurevich, Haifa 1 995 continued 9
1 2 0-0 0-0 1 3 i.g5 lL!bd7 1 4 l:!ad1 'ii'b6 1 5 e4 lL!xc3 1 0 i.xc3 cxd4 1 1 lL!xd4 a6 1 2 g3
lL!h4 wi th an advantage to White in view of 'ii'c7 13 i.g2 i.. e 7!? 14 0-0 0-0 1 5 l:!ac1 l:!ac8
15 ...g6? 16 lL!xg6 hxg6 1 7 d6 i.xd6 1 8 1 6 'ii'e2 l:!fd8! (the x-ray finds no practical
'ii'xg6+, when White wins. use, so Black ignores it) 1 7 lL!b3 lL!e5 1 8
1 o . . . b5 1 1 0-0-0 i.d6 1 2 ll:\e4 i.e7 1 3 i.b4 'ii'd7 1 9 i..xe7 'ii'xe7 with equality. In
i.xc4? ! stead 8 ... i.e7 9 e4 lLixc3 10 i.xc3 0-0 would
This attack seems incorrect. 1 3 i.xf6 be a very natural transposition to our main
i.xf6 14 lLixc5 0-0 (14... i.c8 1 5 'ii'e4+! with game, which seems more interesting and
an attack) 1 5 lL!xb7 'ii'c7 1 6 g4!? was a try for equally sound.
an advantage. 6 . . . i.e7 7 cxd 5 lbxd 5
1 3 . . . bxc4 1 4 l:.he 1 0-0 1 5 d6 i.xe4 1 6 7 ... exdS leads to positions similar to the
l:.xe4 i.xd6 1 7 ll:\e5 ll:\bd 7 ? ! 4 ... d5 lines. There is no need for exact theo
1 7 ... c3! would have disturbed White con retical knowledge there.
siderably. Now we are heading for a draw. 8 i.d2 0-0 9 e4 lbxc3 1 0 i.xc3 lbd7 1 1
1 8 :xd6 ll:\xe4 0-0-0
1 9 i.xd8 liJxd6 20 lbxd7 .=.txd8 2 1 lbb6 1 1 . . . c6!

:ab8 22 lbxc4 lbxc4 %-% 1 1 ...'ii'c 8 1 2 h4 1ld8 13 l:!h3 h6!? 14 d5 1 5
ltJgS! (threatening e4-e5) 1 5. . .hxg5 1 6 hxgS
Game 1 7 lL!g4 (1 6 ... lL!e8?! 1 7 f4 gives White a very
C ramling-Al masi strong attack as the black pieces are not play
Horgen 1995 ing) 17 f4 i.cS! 1 8 i.xg7!? �xg7 1 9 'ii'c3+
and White had a very interesting attack in
1 d4 ll:\t6 2 c4 e6 3 lbt3 b6 4 a3 i.b 7 5 Akog.ian-Granda Zuniga, Groningen 1 993.
ll:\c3 d5 6 'ii' c 2 Best for Black here is 19 ... �£8! 20 l:!h7 �e7,
Another way to lead to the positions in bringing the king to safety, when 21 i.e2
the game (in fact the most common) is the exdS 22 f5 is messy indeed.
·
following: 6 cxdS lLixd5 7 i.. d2 (7 'ii'c2 i.. e7 1 2 h4 b5 1 3 �b1

8 i.d2 has been seen, but the most popular More direct is 13 l:!h3!? a5 1 4 d5 cxdS 1 5
replies to 7 'ii'c2 are 7 ... c5 and 7 ...lL!xc3, i.xb5, as in Kamsky-Anand, Sanghi Nagar
51
1 994. Black defended well by taking the cen No improvement is 1 8 dxe6 fxe6 1 9 .ib5
tral squares with 1 5 ...lll f6 1 6 ltlg5 'ii'b 6 1 7 (1 9 hxg5 'i'e7 and Black's position is com
exd5 l:ac8! 1 8 .id7 .ixd5! 1 9 .ixc8 :xc8 20 fortably superior, with, among others, ... ltlb6
l:xd5 (the bishop seems to be stronger than and ... c5-c4 coming) 19 ... .ic6! 20 .ixc6 :Xc6
the rook, so this exchange is quite natural) 21 hxg5 'i'e7 and White is very weak on the
20 ... exd5 21 'ii' f5 'ii'a 6, and now instead of highly important light squares.
the adventurous 22 c,t>d2?! White should have 1 8 . . . exd5 1 9 exd5 g6 20 .i.b5
settled for the repetition after 22 c,t>b1 'ii'f1+ 20 'ii'b3 'fllxg5 21 'iih3 ltlf6 and Black is
23 c.t>a2 'i'c4+. close to winning. 20 f4 is met by 20 ... ltlb6.
1 3 . . . a5 1 4 .i.e 1 ? 20 . . . lLlb6
1bis move hurts my eyes. I cannot see Black is clearly better according to Almasi.
how the bishop will ever be a useful piece on 21 d6 c4?
e1 . 21 ... .ixg2! 22 :h2 .if3 and White is fin
Better, in my opinion, is 1 4 .id2 b4 1 5 a4 ished.
c5 1 6 d5?! (this pawn sac seems wrong; 1 6 22 f3! c3! 23 b3
.ig5! i s obviously superior, getting all the Forced. 23 d7 b3! wins on the spot for
pieces into action) 16 ... exd5 17 exd5 .ixd5 Black.
1 8 ltlg5 ltlf6, Mikhalchishin-Petrosian, Lvov 23 . . . ltld5 24 .i.f2!
1 994, when Black has the initiative and a More tricks: 24 d7 ltle3! and Black wins.
pawn and, therefore, the better prospects (his 24 . . .'ilxg5 25 d7 .z:lb8 26 .i.h4?
pieces are only temporarily awkwardly placed Here White had the chance to gain good
in the centre). counterplay with 26 .id4! lll e3 27 .ixe3
14 d5 cxd5 1 5 lllg5 (1 5 .ixb5 l::tc 8 1 6 'ii'xe3 28 .l:.he1 1i'g5 29 l:e8, although
exd5 .ixd5 1 7 .id3 'ii'b 6! looks good for Black's chances are preferable after 29 ... l:d8!
Black) 1 5 ...dxe4 16 .ixb5 .id5 17 lllxe4 'fllc 7 etc.
was good enough for Black in Tregubov 26 . . .'ilf4 27 .z:lde 1 ?
Akesson, Cap d'Agde 1 994, although White White is on the way, not to sacrifice an
has quite a bit of compensation for the pawn. exchange; but to win a piece. The only prob
1 4 . . . b4 1 5 a4 :ea 1 6 ltlg5 c5 1 7 d 5 lem is that a bishop on d5 together with the
.i.xg5! strong c3-pawn combine to provide Black
1 7 ...e5? fails to 1 8 ltle6! fxe6 1 9 dxe6, with decisive compensation. 27 .if2 was
when White is better. better according to Almasi.
1 8 hxg 5 27 . . . ltle3 28 .z:lxe3 'ilxe3 29 .z:le 1 'i'b6 30
52
Th e Pe tro sian Sys tem: 4 a 3
J:e8?! .i.d5! 3 1 �a2 J:bxe8 32 dxe8'i' development and, in fact, this is the main
J:xe8 33 i.xe8 'i'd4! reason why I cannot fully recommend the
system with 4... i.b7 followed by 6 ... i!llx d5. I
just do not feel comfortable about the result
ing positions.
7 . . . ltlxc3
7 ... c5 is dealt with in the following game.
8 bxc3
This has become the only serious move.
In the late 1980's 8 1Wxc3!? was also played
quite a lot, but several methods to equalise
were quickly found. The simplest is 8 ... i!lld7 9
i.g5 (9 i.f4 i.d6 is no problem at all for
Black) 9 ... i.e7 10 i.xe7 cJilxe7! (the king is
not exposed, so this yields no problems) 1 1
White has a piece but, as promised, Black e3 :c8 1 2 i.e2 c5 1 3 dxc5? (this starts a
is very close to winning. kamikaze journey for the queen which leads
34 i.g3 'i'd2 35 �b 1 'i'e3! 36 i.b5 only to problems; 1 3 0-0 is better) 1 3 ... llxc5!
i.xb31 37 'i'xb3 'i'g 1 + 38 �a2 'i'xg2 + 14 'ilxg7 ltg8 15 'ii'xh7 l:lxg2 (the white king
3 9 �a1 'i'xg3 4 0 'i'd 1 'i'f2 4 1 i.e2 b3! is more awkward in the centre as Black is
42 'i'xb3 'i'xe2 43 'i'xc3 'i'd 1 + 44 �b2 well developed) 1 6 1Wh4+ lll f6 1 7 ltd1 'ikc7
'i'd8 0-1 1 8 llld4 a51 and Black is clearly better accord
White has had enough. ing to Petursson. Portisch-Karpov, Biel 1 996
------- saw White blunder with 1 9 i!ll b 5?? (1 9 h3!?)
Game 18 1 9. . .l:.xb51 2 0 i.xb 5 ltg4 0-1 .
Timman-Polgar a . . . i.e7 9 e4 0-0 10 i.d3 c5 1 1 0-0
Bali 2000 Here there is a very interesting alternative.
Of course this can also be played at move 13,
1 ltlf3 lLlf6 2 c4 b6 3 d4 e6 4 a3 i.b7 5 but it is for some reason seen more often
ltlc3 d5 6 cxd5 ltlxd5 7 'i'c2! here - namely 1 1 i.b2!, when Piket-Rau,
Rotterdam 1 988 continued 1 1 ...'ilc8 1 2 1We2
i.a6 13 0-0 i.xd3 14 'ii'xd3 'ila6 (14 ... cxd4
1 5 cxd4 'i'a6 1 6 'i'e3 llld7 1 7 d5 gives White
the initiative, while 14 ...i!ll c 6? 1 5 d5 llla5 1 6
c4 is overwhelmingly better fo r White) 1 5 c4!
cxd4 1 6 lllxd4 'ii'b7 (1 6 ... ltc8?! 1 7 lllb 5 lllc 6
18 'iig3 e5 19 f4! was very promising for
White in Piket-Polugaevsky, Aruba 1 994) 1 7
'ii'g3 ( 1 7 i!llb 51? also looks good) 1 7 . . .i.f6 1 8
e 5 i.e 7 1 9 l:lad1 ltd8 20 lllb 5 llxd1 21
:Xd1 lll a6 with an unclear game. I think that
this is more unclear for White than for Black,
if you get my drift. White has some strong
This is the most popular move in this po plans involving the advance of the h-pawn.
sition with GMs today, and for good reason. Note that 1 5 c4! is far stronger than 1 5 1We3
As we shall see Black has no clear-cut path to i!ll d7 1 6 llad1 ltac8 1 7 Ad2 b5, where the
53
bishop is poor on b2. Black has at least d5!? c4 1 6 'ii'e2 exd5 17 exd5 i.f6 1 8 d6 'iWfS
equalised, and won quite soon in Flear 1 9 i.g3 �d7 20 'i'xc4, when a draw was
Hjartarson, Szirak 1987. Nor does 1 5 'ii'xa6 agreed in Piket-Van der Wiel, Holland 1 996)
�xa6 1 6 �e5 l:tfc8 17 l:tfdl cxd4 1 8 cxd4 1 5 ... �c6 16 ltfdl �a5!?, Arbakov-Levin,
.l:r.c2 present Black with anything but pros Berlin 1 994. Now White had sufficient com
pects, as in Malaniuk-Ehlvest, USSR 1 987. pensation for the pawn after 1 7 h4 �c4 1 8
1 1 . . . 'ifc8 1 2 'ife2 .ta6 'i'e2 cxd4 1 9 cxd4 i.xa3 20 �d2 �xd2 21
l:txa3! �c4 22 l:tg3, but nothing has been
decided. Personally I would not like to be
Black here, but I think it is more a question
of style than anything else ..
13 i.b2! leads to the note to White's 1 1 th
move.
1 3 . . . cxd4 1 4 cxd4 .txd3 1 5 'ii'x d3
White has no chance for an advantage
here, it seems. The alternative looks a little
clumsy and Black equalises easily: 1 5 l:txd3
�d7 1 6 i.b2 (1 6 i.g5?! i.xg5 1 7 �xg5 h6
1 8 �h3 'ii'c4, as in Peshina-Gurevich, Eger
1 987, is already better for Black; White's
1 3 J:d 1 forces are misplaced) 16 ...'i'a6 1 7 l:tadl l:tfe8
Others: 1 8 'iii'e3 l:tac8 1 9 d5 exd5 20 l:txd5 �£8 21 h4
a) 13 a4 i.xd3 14 'ii'xd3 ltd8 15 i.f4 (1 5 'iii'a4 22 l:t1d4 and a draw was agreed in
a5 bxa5 1 6 :Xa5 �c6 1 7 l:ta4 cxd4 1 8 cxd4 Bareev-Rodriguez, Sochi 1 988.
°ii'd7 with equality - Dautov) 1 5 ...'ii'b7 1 6 a5 1 5 .. .'ifa6 1 6 .tb2
�c6 (1 6 ... cxd4!? 17 cxd4 bxa5 1 8 :xa5 �c6 16 d5 'iii'xd3! 17 ltxd3 exd5 18 exd5 i.d6
would probably be fine) 17 axb6 cxd4 1 8 1 9 .i.e3 .:r.c8 20 l:te1 �d7 is equal according
cxd4 axb6 1 9 .:Xa8 :Xa8 2 0 .:r.d1 .:r.d8 with to Knaak. 1 6 'iii'e31? �d7 1 7 i.b2 ltac8 is
equality, Azmaiparashvili-Epishin, Reggio more adventurous, but Black should be able
Emilia 1 995/96. Dautov has suggested the to gain sufficient counterplay on the c-file.
following alternatives: 14 ...\i'a6 1 5 'ii'b 5 'ii'h7 1 6 . . .'ifxd3 1 7 J:xd3 �d 7 1 8 J:c 1
1 6 ltel a6 1 7 'i'e2 .:r.d8 1 8 d5 with good White gains the c-file but Black has the
prospects for White, and 14 ... cxd4!? 1 5 cxd4 possibility of a distant passed pawn and. of
�c6, which possibly equalises. I am not creating . a good square for the knight on c4.
completely sure on this one, although it looks Also equal is 1 8 d5 �c5 1 9 l:te3 .:r.fd8 20
natural. .:r.d1 :ac8.
b) 13 i.f4 gives White the advantage ac 1 8 . . .J:fc8 1 9 J:dc3 J:xc3 20 J:xc3 �f6
cording to ECO, but I will remain sceptical 21 �d2
on that one, too. Black should be fine after 21 e5? �d5 helps only Black.
1 3 ... i.xd3 1 4 'ii'xd3 l:td8 (more careful is 21 . . . b5!
1 4... �d7 1 5 a4 .:r.d8 1 6 l:tfe1 cxd4 1 7 cxd4 Black would never allow 22 a4, where the
'ii'b 7 1 8 l:lab1 :ac8 1 9 a5, when White might pawn is less weak and when he would not
have a slight positional edge but it will be have c4 waiting for the knight.
very difficult to prove, and he was nowhere 22 J:c6
even close in Piket-Polugaevsky, Aruba 1 994) 22 ltc7 'iii f8 23 .:r.b7 a6 has been assessed
1 5 'iii'e 3 (White has no advantage after 1 5 as unclear. I think White should be very care-
54
ful here that ... l:tc8 does not suddenly win 0-1
something. All in all it seems that improving
the king should be a priority. Game 19
22 . . . 'iPfs 23 f3 �ea 24 ltlb3 Kasparov-van der Wiel
24 'iifi' f2 'iitt'd7 shows that White somehow Amsterdam 1988
gained nothing by 22 l:tc6, only to have to
waste time keeping it there. 1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 a3 .i.b 7 5
24 . . . ltld7 ltlc3 d5 6 'i'c2 c5 7 cxd5 ltlxd5
24 ... 'iifi' d7 25 l:tc1 l:tc8 26 ltic5+ and White
maintains the balance.
25 �f1 ?
Creating problems later. Actually it is dif
ficult to why 25 'iifi' f2 ! was not played.
25 . . . ltlb6 26 ltlc5?
White is drifting. The knight does less
here than it seems to while the knight on c4
is really annoying. 26 ltia5 'iifi'd 7 27 l:tc2 l:tc8
28 l:r.xc8 ltixc8 29 'iifi' e2 results in an endgame
where Black has very slightly better chances
due to the prospects of creating a passed
pawn and the weakness of the a3-pawn.
However, this should be a drawable option 8 dxc5!
for White. This is the only move that is dangerous
26 . . . ltlc4! 27 .i.c 1 :I.dB for Black. The alternatives both promise
Black equality.
a) 8 e4 ltixc3 9 bxc3 and now:
al) After 9 ... ltid7 White reaches another
crossroads.
al 1) In reply to 10 ..td3 the best move is
1 0 ... 'iVc7! according to Kasparov, a draw
being agreed · in Kasparov-Sosonko, Lucerne
1 982 after 1 1 'iVd2 g6 1 2 0-0 ..tg7. Kasparov
was, in fact, so uncomfortable with his posi
tion here that he claimed in Chess Informant
that Black already has a slight advantage.
Whether or not this is too strong an assess
ment I shall not say, but the fact that Black is
Now White just loses something. If the not unhappy with the outcome of the open
king had been on f2 the game continuation ing is quite clear to me. Black is also fine after
would have worked better because there the alj:ernatives, e.g. 1 1 'iVbl g6 1 2 0-0 ..tg7
would be no check on dl . 1 3 :a2 0-0 14 l:tel a6 1 5 a4 l:r.fc8 1 6 h3 ..tc6
28 a4 17 ..te3 c4 1 8 ..tc2 b5 with equality, as in
28 d5 ltie5 29 l:r.c7 ..td6 and Black should Ostermeyer-Sosonko, Hannover 1 983.
win. Meanwhile 11 ..tb2 cxd4 12 cxd4 'i'xc2 1 3
28 . . . Axd4! 29 ltlb3 Ad3 30 axb5 ltle5 3 1 ..txc2 ..ta6 led to a carefree endgame for
:ea + �d7 3 2 :as Axb3 3 3 :xa7+ �ea Black in Yusupov-Miles, Linares 1 983, and
55
1 1 0-0 cxd4 12 cxd4 'i'xc2 13 j,xc2 l:lc8 �eS 14 �xeS j,xeS 1S j,bS+ 'it>f8 16 0-0
even gave Black some initiative in Franco 'i'd6 17 f4 j,f6 18 c4 a6 19 i.eS and White
Ocampos-Van der Sterren, Wijk aan Zee is already winning, Tukmakov-011, Kujby
1 983. shev 1 986 continuing 1 9 ... j,xeS 20 fxeS
a1 2) 10 i.f4. This bishop needs to be de 'i'xeS 21 'ii' f2 f6 22 :£e1 'ti'hs 23 j,d7 'i'f7
veloped immediately. 10 ... cxd4 1 1 cxd4 l:tc8
24 j,e6 'ii'c7 2S d6 'ii'd8 26 i.h3 j,c6 27
12 'i'b3 'ii' f6!? 13 j,g3 'ii'g6 (1 3 ... j,xe4 1 4
l:te7 1 -0) 1 3 exdS �aS 14 i.d3 j,f6 1 S
j,bS i s very dangerous fo r Black) 1 4 i.d3
j,e7 1S 0-0 0-0 16 l:tfel l:tfd8 and Black was j,xh7+?! (1 S 0-0! gives White the advantage
well placed in Van der Sterren-Van der Wiel, due to 1 S ... g6 1 6 c4) 1 S ... 'it>h8 1 6 j,e4 (1 6
Wijk aan Zee 1 986. There are both good 0-0 g6 17 j,xg6 fxg6 1 8 'i'xg6 'i'e7 and
things and bad things to say about the queen White's queen is alone in attack) 1 6 ... j,a6 17
on g6. The rather futuristic 1 2 ... 'ii'f6!? seems j,d3 %:te8+ 1 8 'it>fl �c4 1 9 j,xc4 j,xc4+ 20
to be the best, while more solid is 12 ... j,e7 c;t>g1 :e2 and Black had enough for the
13 j,d3 �f6 14 'libs+ (the only serious pawn (to put it mildly) in Portisch-Sosonko,
move - after 14 'i'bl 0-0 1 S j,d2 �e8 1 6 0-0 Tilburg 1 982. 10 ...g6 1 1 :d1 i.g7 1 2 dS exdS
�d6 17 eS �£5 1 8 j,xf5 exf5 1 9 'ii'x f5 j,x£3 1 3 exdS 'i'e7+ 14 j,e2 �eS 1 S 'ti'a4+ �d7
20 'i'x£3 'ii'xd4 Black was already slightly 16 0-0 0-0 17 l:tfel left White better devel
better in Crarnling-Karpov, Oropesa del Mar
oped in Petrosian-Sosonko, Tilburg 1 982,
1 996, and 1 4 dS?! exdS 1 S l:td1 0-0 1 6 0-0
while 10 ... .rtc8? 1 1 dS!? exdS 12 exdS 'i'xdS
dxe4! was better for Black in Miles
Polugaevsky, Sarajevo 1 987) 14 ... 'i'd7 1 3 i.d3 �eS 14 0-0-0 was just a killer in
(14... j,c6! 1S 'i'bl 0-0, with good play for Agdestein-Lau, Dortmund 1 987. I would say
Black, is probably what Karpov would have that Black's king is not completely safe ...
played against Cramling) 1 S �eS 'i'xbS 1 6 b) 8 j,gS is also not so dangerous. After
j,xbs+ 'it>f8 1 7 £3 �e8! 1 8 i.d7 %:td8 1 9 j,c6 8 ... i.e7 9 i.xe7 'ii'xe7 1 0 �xdS exdS 1 1 e3
j,c8! and Black was only slightly worse in 0-0 1 2 j,d3 g6 (taking control over the king
Timman-Karpov, Jakarta 1 993. Also possible side light squares) 1 3 dxcS bxcS 1 4 0-0 �d7,
is 10. .. j,e7, when 1 1 dS?! is too early: as in Seirawan-Timman, Amsterdam 1 992,
1 1 ...exdS 12 exdS 0-0 1 3 :d1 ?! gS! 1 4 j,cl Black secured equality. Vilela-Rodriguez,
g4 1S �d2 j,xdS and Black is close to win Cienfuegos 1 98S seemed okay for Black after
ning. Instead White should develop first with 1 1 g3 0-0 1 2 i.g2 �d7 1 3 0-0 :ac8.
1 1 j,d3, e.g. 1 1 ...Ac8 1 2 'i'e2 0-0 1 3 0-0 8 i..x c5 9 i.. g 5!
. . .
cxd4 14 cxd4 �f6 (not 14 ... l:tc3? 1 S %:tfc1 !

l:lxcl+ 1 6 l:lxc1, Gulko-Timman, Amster
dam 1 987, when White had the advantage
due to the threat of l:tc7) 1 S a4 Ac3! 1 6 eS
�dS 1 7 j,d2 l:tc7 1 8 l:tfc1 with equality in
Bareev-Chernin, Lvov 1 987.
a2) Black can also try 9 ... j,e7 10 j,bs+!
j,c6 1 1 j,d3 and the bishop is badly placed
on c6. No need for further variations ...
a3) Or 9 ... �c6, which also puts Black in
trouble after 1 0 j,b2!, e.g. 1 0 ... i.e7 1 1 dS!
exdS 12 .rtd1 0-0 (worse is 1 2. .. j,f6 13 exdS
56
Th e Pe tro sian Sys t em: 4 a 3
The most energetic. There are two differ i.d3 %:tcS 1 4 'We2 lDd7 1 5 0-0 lDc5 1 6 lDd4
ences between this and S i.g5 - Black no lDxd3 1 7 'Wxd3 Ac5 Black had the c-file to
longer has a c-pawn, so the d4-square is compensate for his structural deficiencies in
available to White, and Black has already Ehlvest-Korchnoi, Zagreb 1 9S7. In Gure
moved his bishop. Now there are three dif vich-Ionescu, Moscow 1 9S7 White played 1 3
ferent paths to choose from. I would rec i.b5! to be ready to remove Black's knight.
ommend 9 ... f6, although none of them Play continued 1 3 ... d4 14 lDxd4 i.xg2 1 5
promises equality. :g1 i.e4 1 6 lDfS! i.xfS (1 6 ...'i'e5? 1 7 lDh6+
9 'i'c8
. . . q.,hS 1 S lDxf7+! and White wins) 1 7 'WxfS a6
Others: 1 S i.d3 g6 1 9 'Wd5! with a clear advantage to
a) 9 .. .£6 10 i.d2 0-0 (Black should not White. Black's 1 3th move seems a bit too
hesitate over this logical move - after optimistic. White is just better developed for
10 ... lDd7 1 1 e4 lDxc3 1 2 i.xc3 'i'cS 1 3 0-0-0 this kind of stuff. Best was 13 ... :cs, when
0-0 14 b4 i.e7 15 'i'b3 %:tf7 16 i.c4 lDf8 1 7 Black is still fighting to equalise.
lDd4 White had strong pressure l_his king will 1 0 :c1
be fine on bl] in Plaskett-Short, Plovdiv 1 0 e3 0-0 11 lDxd5 hd5 1 2 i.d3 h6 13
1984) and now White should play 1 1 e3! i.h4 lDd7 14 0-0 a5 1 5 l:.acl a4 was harm
(intending to use the b1 -h7 diagonal), e.g. less in Farago-Horvath, Hungary 1 995.
1 1 ...q.,hS?! 1 2 i.c4! 'ile7 (1 2 ... lDxc3?! 1 3 1 o h& 1 1 .i.h4!
...
i.xc3 'We7 1 4 h41 wi th th e idea of h4-h5 with The only serious move - the bishop must
a very dangerous attack) 1 3 %:td1 !? and White remain active. For example after 1 1 i.d2
has the better prospects (Piket). An im lDf6! the bishop looks stupid on d2, e.g. 1 2
provement is 1 1 ... a5! - there is no need to go e 3 0-0 1 3 i.e2 lDbd7 14 0-0 'i'bs 1 5 b 4 i.e7
to the h-file before being forced to do so. 16 lDd4 l:.cS 17 'i'b3 a5 1 S £3 axb4 19 axb4
Gelfand-Lautier, Manila 1 990 continued 1 2 i.d6 and Black was clearly better in Chek
i.d3 c:;ihs! 1 3 l:.d1 (13 h4!? 'Wes 1 4 lDg5 £5 hov-Timoshchenko, Berlin 1 9S6, or 1 2 b4
1 5 e4 lDf6 1 6 0-0-0 lDc6! offers Black coun i.e7 1 3 lDb5 lDc6 1 4 'ila4 a5 1 5 bxa5 0-0
terplay according to Gelfand, while 1 3 with unclear play in Petursson-Amason,
i.xh7? £5 1 4 i.g6 'i'f6 1 5 i.h5 'ilh6 benefits Reykjavik 1 9S5. Meanwhile 1 1 lDxd5?1 hxg5
only Black) 13 ...lDd7 14 0-0 'We7 15 i.e4! 1 2 b4 g41 1 3 lDe5 l:.h5 favoured Black in
lDxc3 1 6 i.xc3 (1 6 i.xb7? lDxd1 17 i.xaS Petursson-C.Hansen, Borgarnes 1 9S5, and
lDxe3 and Black wins) 1 6 ...:a7 17 lDd4?! (17 Ligterink-Beliavsky, Wijk aan Zee 1 9S5 went
l:.d21? and White is perhaps a little bit better 1 1 lDe4 lDd7 (1 1 ...hxg5?? 1 2 lDd6+1) 12
Gelfand) 1 7 ... i.xd4 1 S exd4 %:tdS! 1 9 .l:fel lDxc5 lDxc5 1 3 e4 (1 3 i.d2 lDf6!, to take
i.xe4 20 'Wxe4 lLJf8 21 d5 'i'd6! with equal control of e4, guarantees Black a good game)
ity. Note that after 1 1 e4 there is no good 1 3 ... hxg5 1 4 exd5 exd5 1 5 lDxg5 'it>f8 1 6
place for the light-squared bishop, e.g. i.e2 d4! 1 7 0-0 d3 1 S i.xd3 'i'g4 1 9 lDe4
1 1 ...lDxc3 12 i.xc3 a5 13 %:td1 'i'cS 14 i.d3 i.xe4 20 i.xe4 'i'xe4 and White resigned.
lDa6 1 5 0-0 �hS 1 6 %:tfe1 e5 with equality in 1 1 a5
. . .
Vyzmanavin-Lautier, Sochi 1 9S9, or 1 3 i.c4 Pe'1iaps the altematives are better:

'Wes 1 4 'i'e2 i.a6 1 5 ..L:a6 'ilxa6 1 6 'i'xa6 a) 1 1 ...lDc6 1 2 lDxd5 exd5 13 e3 0-0 14
lDxa6 1 7 q.,e2 e5 1 S :bd1 %:tfdS with a good i.d3 a5 1 5 0-0 d4 1 6 e4 :es 1 7 i.g3 with a
position for Black in Benjamin-Korchnoi, slight advantage to White in Ftacnik
Jerusalem 1 9S6. Hjartarson, Esbjerg 1 9S5.
b) 9 ... i.e7 is slightly passive. After 1 0 b) 1 1 ...0-0 12 lDxd5 exd5 13 e3 lDd7 14
i.xe7 'Wxe7 1 1 lDxd5 exd5 1 2 e 3 0-0 1 3 i.e2 i.d6 15 'i'dl 'i'eS 16 0-0 lDc5 17 lDd4
57
'ii'e4 1 8 .1g3 with only a tiny edge for White, 1 5 ... e5 1 6 .1c4! with a large lead.
Yusupov-Chemin, Tunis 1 985. 1 6 0-0 0-0 1 7 :td 1 f6!
1 2 l2Ja4! 17 ... l:.e8?! 18 .1xa6 l:.xa6 19 'i'd2! and
The only serious option. 1 2 �e4 �d7 1 3 White wins.
'ibl 0-0 14 �d6 'ii'c6 1 5 �xb7 looks like 1 8 .i.c4! .i.xc4 1 9 'it'xc4 :t7 20 .i.g3 e5
White has never heard about development, 21 l2Jh4!
but only about the two bishops (well, the With a clear advantage for White.
king's bishop is yet to wake up). Ftacnik 21 . . . 'it'e8
C.Hansen, Esbjerg 1 985 led to Black eaming 21. ..'i'a6? 22 'ii'xf7+! <t>xf7 23 l:.xd7+ and
an enormous lead after 1 5 ... 'ii'xb7 1 6 e4? (1 6 White wins.
e3 was still equal) 16 ...�f4 1 7 .1g3 £5! etc. 22 l2Jf5 'it'e6 23 'it'e2! l:l.b8 24 l:td6 'it'e8
12 e3 0-0 13 .1e2 �xc3 14 'ii'xc3 �c6 1 5 On 24... 'i'b3 Kasparov gives the following
0-0 a4 was equal in Spraggett-Portisch, winning line: 25 'ii'g4 <t>h7 26 �xg7 (26
Montpellier 1 985. �xh6!? gxh6 27 :Xd7 �h8 28 l:.cd1 also
1 2 . . . l2Jd7 1 3 e4! looks good) 26 ...'i'xb2 27 l:.£1 l:.xg7 28 l:.xd7
�e8 29 f4 and here Black has nothing better
than 29 ...'i'd4+ 30 l:.xd4 l:.xg4 31 l:ld7+ l:.g7
32 l:.xg7+ @xg7 33 fxe5 l:.b6 with a terrible
endgame.
25 :cd 1 ! l2Jf8 26 f4! l2Jb5
26 ... exf4 27 .1xf4 �fe6 28 .1g3 'Wb5 29
'i'g4 @h7 30 l:.xe6 �xe6 3 1 �d6 and White
wins.
27 fxe5 l2Jxd6?!
Apparently 27 ... fxe5 28 l:.6d5 �d7 29
'i'd2!, with a decisive advantage to White, is
better.
28 l2Jxd6 'it'a4 29 l2Jxf7 :xb2! 30 l2Jxh6 +
White needs to get going before Black fin lfi>h7
ishes his development. After 1 3 e3· 0-0 1 4
.ib5 .1a6 1 5 .1xa6 (1 5 'i'd3 .1b4+ 1 6 <t>e2
�c5 1 7 .1xa6 'ii'xa6 1 8 'i'xa6 l:.xa6 1 9 �xc5
.1xc5 with equality) 15 ... 'ii'xa6 Black was no
worse in Salov-Timman, Belgrade 1 987.
1 3 . . . l2Jc7
Black's position is beyond repair, e.g.
13 ... �f4 14 .1g3 �h5 15 �xc5! �xc5?!,
while 15 ...bxc5 is horrible too in view of 1 6
b4! axb4 1 7 axb4 �xg3 1 8 hxg3 �b3 1 9
.ibs+ <t> f8 2 0 'ii'xc8+ l:.xc8 2 1 l:.xc8+ .1xc8
22 .1c4 and White wins. After 13 ... �5f6 1 4
.i.xf6 �xf6 1 5 .ib5+ �e7 1 6 e 5 �e4 1 7 0-0
White has a powerful initiative. 3 1 'it'h 5
1 4 l2Jxc5 bxc5 31 'ii'£3! is even stronger.
14 ... �xc5? 15 b4 and Black drops a piece. 31 . . . g6 32 'it'f3 <li>xh6 33 :n 'it'd4 + 34
1 5 .i.e2 .i.a6 lfi>h 1 l2Jh7 35 exf6 l2Jxf6 36 .i.f4+
58
Th e Pe tro sian Sys t em: 4 a 3
36 e5 �h7 37 i.£4+ '1tg7 38 i.cl and 'i'h7+ '1tf8 1 9 .i.h6+ '1te8 and the attack
Black has nothing left to do but resign. never went further in Vokac-Sax, Lazne
36 . . . r;1;>g7 37 i.g5 l:.b6 38 'i'h3?! Bohdanec 1 995. 10 ...�xc3 1 1 bxc3 .:r.c8 ap
Again 38 'iWf4! was simpler. pears to be equally good, when White should
38 . . . "1i>g8 ?! start thinking of castling. After 1 2 e4?! �c5!
38 ...'ifxe4! 39 i.h6+ rt>h71 40 i.£8+ '1tg8 13 hxg6 �d3+ 14 'ii'xd3 hxg6 15 Axhs+
41 i.xc5 gives White a pawn but at least puts .i.xh8 1 6 .i.f4 .i.g7 17 '1td2 c5 Black was
up some resistance. better in Maksimovic-Groszpeter, Berlin
39 'ii'c 8+ �g7 40 'i'c7 + li:ld7 41 'ii'f4 1 988. 9 ... 0-0 is also okay, although it seems
'i'c4 42 h3 'i'e6 43 l:.d 1 ! .:l.c6 44 i.d8! natural to wait. Here after 10 hxg6 hxg6 1 1
.!Lib& 45 :n 1 -0 �xd5 exd5 1 2 .ad2 �d7 1 3 j,, d3 l:le8 1 4
------.. .i.c3 a5 15 'ili'c2 .i.a6 1 6 .i.xa6 Black has no
Game 20 problems, Razuvaev-Rodriguez, Moscow
Khalifman-Short 1 985 already coming to a peaceful conclu
Paernu 1996 sion.
b) 8 �xd5 exd5 9 b4 is supposed to give
1 d4 li:lf6 2 c4 e6 3 li:lf3 b6 4 .!Llc3 i.b 7 White an advantage. But due to the plan illus
5 a3 d5 6 cxd5 li:lxd5 7 e3 trated in this game (... c7-c6 and ... a7-a5) no
This is the main move in ECO, but in re body adopts this approach any more. Black is
cent times it has been less popular among just fine. Shirov-Karpov, Monte Carlo 1 995
those who know. This is due to 7 ... g6, and went 9 ... i.g7 10 i.b2 c6 11 i.e2 0-0 12 0-0
for this reason I will not consider other �d7 13 'ili'b3 Ae8 14 Afcl a5 1 5 bxa5 Axa5
moves here, or give 7 e3 serious attention. 16 a4 c5 with equality.
Anyway, White plays the text with the idea of 8 . . . c6 9 i.d3 i.g7 1 0 .!Lla4!?
e3-e4. It does not seem to be so dangerous. Short wrote that this move was an inven
7 . . . g6 tion of Lobron and his long-term companion
] ack Daniels. Surely Mr. Daniels is an inven
tive guy. After 1 0 e4 �xc3 1 1 bxc3 c5! 1 2
.ag5 'ii'd6 1 3 e5 1i'd7 Black seems to d o well
as White's centre appears rather fragile. And
after 14 dxc5?! 0-0! 1 5 cxb6 axb6 Black was
even better in Kasparov-Korchnoi, London
1 983. White has three weak pawns and is
lacking in co-ordination and development. .
1 0 . . . li:ld7 1 1 e4 .!Lle7 1 2 0-0
Here (and on the next move) White has a
serious alternative in 12 i.f4 0-0 13 0-0.
Then Black needs to accept a weak pawn and
he will be okay: 13 ... c51 14 dxc5 �xc5 1 5
8 i.b5+ �xc5 l;>xc5 1 6 'ife2 (1 6 'ii'c2 �c6! and Black
Here there are other possibilities: is fine ; while 16 Acl c4! also equalises be
a) 8 h4 _ag7 9 h5 �d7 10 i.d3 0-0 1 1 e4 cause after 1 7 Axc4? i.a6 1 8 .ac7 'iWd7 1 9
(this is just not very goodQ 1 1 ...�xc3 1 2 �e5 i.xe5 20 i.xe5 i.xc4 2 1 i.xc4 'ii'x dl
bxc3 �c5! 1 3 hxg6 hxg6 1 4 dxc5 �osing, but 22 .i::r.x dl Afd8 Black has all the chances)
Black is already doing very well) 14 ... i.xc3+ 1 6 ... �c6 1 7 l:lacl �d4 1 8 �d4 i.xd4 with
1 5 rt>e2 .axal 1 6 'ii'gl 'ii'd7 1 7 'ii'h2 l:lfd8 1 8 equality in Yermolinsky-Ivanov, Parsippany
59
1 996. Here White could have kept the balance

1 2 . . . 0-0 1 a i.g5 with 25 'ii'd 61 f41 26 i.xf4 i.xg2 27 'iii'xg2
13 i.f4-see the previous note.. 'ir'xf4 28 'ii'xe7 'ii'g4+ with a perpetual.
1 a . . . h6! 25 . . . a6 26 i.e2 :.d7 27 "i'e1 "i'e6 28
i.ta i.xf3 29 :.xta i.f8 30 i.f2? !
30 i.h4! is forced. Now Black will be able
to play ... g6-g5 and prove a clear advantage.
ao . . . i.e7 a 1 c4 :.ad8 a2 :.tea g5!
1 3 ...l:.e8?1 (Khalifman) 14 l:.cl °i'b 8 15

i.h4! (going to the h2-b8 diagonal) 15 ...h6 16
b4 bS 1 7 .lbcS .lbxcS 18 Les with a strong
positional advantage for White in Khalifman
Sivokho, St Petersburg 1 996. The e-pawn is now in trouble. From here
14 i.ea �h7 1 5 :.c1 f5! Short's fantastic technique took over.
Black has equalised (Short). aa i.ea gxf4 a4 i.xf4 :.d4 as °i'f1 i.g5
1 6 exf5 exf5 a6 i.xg 5 hxg5 a7 "i'e2 �g6! a8 :.ea
16 ... .lbxfS 17 i.e4 is just wrong. :.d2 a9 "i'ta g4 40 "i'n :.8d4 41 9a :.b2
1 7 i.f4 lL!d5 1 8 i.d6?! 42 :.e2 :.xe2 4a "i'xe2 :.e4 44 °i'd2
18 i.g3 improves. "i'xe5 45 �2 :.d4 46 "i'e2 Wxe2+ 47
1 8 . . . :.e8 1 9 tLica °i'f6 20 i.ga lL!xca 21 �xe2 f4! 48 gxf4 �f5 49 :.b1 :.d6 50
bxca c5 :.n :.h6 51 :.t2 :.ha s2 �n :.ba! sa
Here Black could have rewarded himself �g2 a5 54 :.e2 a4 55 :.e5 + �xf4 56
for his strong opening play with 21...f4! 22 :.e6 :.xaa 57 :.t6 + �e5 58 :.xb6 :.ba
i.h4 'ikf7, when all White's pieces are un 59 :.a6 aa 60 :.as ltf4 0-1
prepared for ... c6-c5.
22 i.b5 :.e7 2a lL!e5 lL!xe5 24 dxe5 °i'f7
25 f4? !
60
Summary
The Petrosian system still holds great dangers for Black as it did 15 years ago when Kasparov
was championing it. The move order with 4 a3 gives Black some extra lines to choose from,
most notably 4... i.a6, therefore it can be a good idea for White players to play 4 liJc3 with the
idea of 4 ... i.b7 5 a3! with transposition to the 4 a3 i.b7 lines. After 4 a3 then 4 . . dS gives a .
typical d-pawn position, as discussed in the introduction. After 4... cS!? we get some interesting
and unusual positions. Black has so far been able to keep the game level, but perhaps White
can make his better structure count for something in the future. 4... c6 is just not very good,
while 4 ... i.a6 leads to a whole series of variations which are basically okay for Black although
precise play is needed. 4... i.b7 is still the main line, and after 5 liJc3 Black only has one good
plan: 5 ... d5 6 cxd5 liJxd5 and now 7 'iilf c2! is perhaps very slightly better for White.
1 d4 lLlf6 2 c4 e6 3 lLlt3 b6 4 a3 (DJ .i.b7

4 ... d5 5 cxd5 exd5 6 liJc3 i.e7 Game 5; 4 ... c6 - Game 6
-
4 . . . c5
5 e3 Game 8; 5 d5 i.a6 6 'iiic2 Game 7
- -
4 ... i.a6
5 'tib3 Game 9
-
5 'iiic2
5 ... c5 6 d5 - 4 ... c5
5 . . . i.b7 6 liJc3 c5
7 d5 - Game 10
7 e4 cxd4 8 liJxd4
8 . . i.cS - Game 1 1; 8 . . liJc6 9 liJxc6 i.xc6 1 0 i.f4
. . (DJ
·
1 0 . . liJh5 - Game 12; 1 0 . i.c5 Game 13

. .. -
5 lLlc3 d 5
5 ...g 6 Game 14
-
6 cxd5
6 i.g5 - Game 15,· 6 'iii c2: 6 ... dxc4 Game 16; 6 ... i.. e7 Game 1 7
- -
6 lLlxd 5 (DJ 7 'ii' c 2

. . .
7 e3 - Game 20
7 c5
. . .
7 ... liJxc3 - Game 1 8

8 dxc5 - Game 19
4 a3 1 0 .i.f4 6 . . . lLlxd5
61
CHAPTER THREE I
5 ii'c2 and 5 i. g 5
1 d4 itif6 2 c4 e6 3 itif3 b6 4 ltic3 e4, is the best way to play with this structure.
i.b7/i.b4 5 'i'c2/5 i.g5 After the not so clever 10 ... dS 1 1 0-0 liJbd7
In this chapter you will see White try to (1 1 ...dxc4 1 2 :fdl 'i'c8 13 e4 gave White an
put Black under pressure and fight for the initiative in Razuvaev-Kruszynski, Polanica
control of the e4-square - and you will see Zdroj 1 979) 1 2 :fdl :cs 1 3 :acl White
him fail. These lines do not offer White an had an edge in Sokolov-Stohl, Burgas 1 992.
advantage from the opening, yet they can 1 1 0-0 ltic5 1 2 l:fd 1 ltife4 1 3 i.xe7
nevertheless be difficult to play for Black, 13 i.g3 liJxc3 14 'ii'xc3 liJe4 favours
and if they suit you, as they used to suit Black thanks to the nvo bishops.
Kamsky, there is a definite possibility that a 1 3 . . . 'i'xe7
level opening might later lead to a superior
middlegame and, ultimately, a full point.
Game 21
Tolnai-Adorjan
Hungary 1992
1 ltif3 itif6 2 c4 b6 3 ltic3 i.b7 4 d4 e6
5 'i'c2
A rather harmless option, the game con
tinuation demonstrating the easiest way to
equalise.
5 . . . c5! 6 dxc5 i.xc5 7 i.g5
7 e4 h6 8 i.f4 i.b4 9 i.d3 0-0 10 0-0 1 4 itid2?1
i.xc3 1 1 bxc3 dS 12 cxdS exdS gave Black at Presenting Black with the opportunity to
least equality in Szekely-Daroczy, Debrecen take full control of the centre. 14 liJxe4, with
1 956. equality, improves.
7 . . . h6 8 i.h4 i.e7 9 e3 0-0 10 i.e2 14 . . . ltixc3 1 5 'i'xc3 d5 1 6 i.f3
ltia6! 1 6 b4 liJd7 17 cxdS :£c8 is nice for Black.
This move, aimed at planting a knight on 1 6 . . ..Z:.fdS 1 7 h3
62
5 'il c 2 . a n d 5 i. g 5
1 7 b4!? tlia6 1 8 a3 .l:.ac8 1 9 'i'b3 dxc4 20 was necessary.

tlixc4 i.xf3 21 gxf3 tlic7 is a shade prefer 33 . . . l'Lixb2! 34 l'Lixb 2 :c2+ 35 q;f3 :xb2
able for Black. 36 axb4 :xb4 37 ltc3 :b7 38 :a3 Aa7
1 7 . . . dxc4! 1 8 i.xb7 l'Lixb7 1 9 l'Lixc4 39 q;e4 q;d6 40 q;d4 a4 41 q;c4 f5 42
:ac8 20 'ila3 'i'xa3 21 l'Lixa3 lbc5 22 rt;b5 :b7 + 43 ..t>c4 lta7 44 ..t>b5 :b7 +
l'Lib5 a6 45 ..t>c4 :b2 46 :xa4 ltxg2 47 .f:.a7 g5
48 :as+ q;e7 49 :a1 + q;f6 50 .:l.h7
q;g6 51 Ae7 ..t>f6 52 Ah7 gxf4 53 :xh6 +
'1Pg 7! 54 :xe6 f3 0-1
Game 22
Kamsky-Yudasin
Biel 1993
1 d4 l'Lif6 2 c4 e6 3 l'Lif3 b6 4 l'Lic3 .tb7
4 ... i.b4 is treated below.
5 .tg5
·
23 l'Lic3?!
White could have equalised with 23 tlid6
:lc6 24 tlic4 tlid3 25 tlie5 .l:.cd6 26 tlic4,
with a repetition.
23 . . . l'Lid3 24 :d2 l'Lie5 25 :xd8+
25 .l:.ad1 .l:.xd2 26 .l:.xd2 @f8 is only
slightly better for Black. Now Adorjan's fan
tastic will to win comes into play
25 . . . :xdB 26 :d 1 :ea 27 Ad4 ..tits 28
..tin rt;e7 29 rt;e2 l'bc4 30 l'Lid 1 b5 3 1
:d3 b 4 32 f4?
32 .l:.b3 was forced. TIUs variation is known to be quite harm
32 . . . a5 33 a3? less but, as this game clearly illustrates, it is a
genuine struggle so long as the pieces remain
on the board. The most normal move here
must be 5 a3, transposing to the Petrosian
System. Actually, this is the way I would play
with White, as the lines with 4... i.b7 seem to
be the most fragile for Black in the Petrosian.
5 . . . h6 6 .th4 i.e7
No}V Black has this possibility, and there
fore equalises easily.
7 e3
Standard. 7 'iic2 has also been tried, when
7 ... c5! is another standard reaction, White
having neglected the dark squares. Sokolov
33 .l:.d4, with problems but nothing more, Salov, Wijk aan Zee 1 997 went 8 dxc5 bxc5 9
63
e3 d6 1 0 i.e2 �bd7 1 1 .l:dl 0-0 12 0-0 a6 1 3 0-0 1 5 �h4! White had a space advantage in
:d2 ii'c7 14 .:tfdl :fd8 1 5 h 3 �IB 1 6 �h2 Korchnoi-Ligterink, Wijk aan Zee 1 984) 12
:d7 17 i.g3 l:ad8 1 8 i.£3 �g6 19 i.xb7 h4 �d7 1 3 i.h2 ii'e7 1 4 0-0-0 0-0-0 1 5 :dn
'ii'xb7 20 �£3 d5 and, if anyone is better, it is g5 with dynamic equality in 11ikhalevski
Black. Yudasin, Tel Aviv 1 994.
7 . . . lt:le4
This is a good way to equalise. Also possi
ble is 7 .. 0-0, and now:
.
a) 8 i.d3 c5 9 0-0 cxd4 (9 .. d6 1 0 a3 �bd7

1 1 b4 d5 12 cxd5 �xd5 13 i.xe7 ii'xe7 1 4
�xd5 i.xd5 1 5 e 4 and a draw was agreed in
Onischuk-Dautov, Gennany 2000) 1 0 exd4
d5 (also possible is 1 0 ...i.x£3!? 1 1 ii'x£3 �c6
12 �e2 d5 13 .:.adl �b4 14 i.xf6 i.xf6 1 5
cxd5 �xd5 1 6 i.e4 with a draw in Seirawan
Schussler, New York 1 985) and now the
most testing is 1 1 i.xf6!, unlike the hannless
1 1 cxd5 �xd5 12 i.g3 �xc3 1 3 bxc3 �c6
14 :e1 i.f6 1 5 .l:cl .:r.c8 16 ii'e2 �a5 1 7 8 ..txh4? 1
. . .
�e5 'ii'd 5 1 8 f4 i.xe5 1 9 fxe5 :c7 with This is risky. A safe path to equality is
complete equality in Spassky-Hjartarson, .8 ... i.xe4 9 i.g3 0-0, e.g. 10 �d2 i.b7 1 1
Reykjavik 1 988. In Sokolov-Rivas Pastor, i.d3 c5! (standard when White weakens the
Leon 1 995 the position after (1 1 i.xf6) dark squares) 1 2 dxc5 bxc5 1 3 0-0 d5 14
1 1 ...i.xf6 1 2 cxd5 exd5 was reminiscent of cxd5 exd5 1 5 e4 dxe4 16 i.xe4 i.xe4 17
the Tartakower where the bishop on b7 is �xe4 �c6 18 'ii'a4 �d4 with complete
somewhat silly, and where there might be a equality in Uhlmann-Padevsky, Havana 1 964,
weakness on c6 occasionally. But the ex or 1 0 i.d3 i.xd3 1 1 ii'xd3 d6 12 0-0 �d7 1 3
change of the dark-squared bishop does also b 4 a 5 1 4 b5 £5 1 5 ii'b3 a 4 1 6 ii'c2 'ii'e 8 1 7
influence White's position. After 1 3 .:.e1 �el e5, Uhlmann-Antoshin, Leipzig 1 965,
�c6 14 i.c2 ii'd6 15 ii'd3 g6 16 a3 :re8 1 7 also equal.
h 3 i. g7 th e game was equal, Black having 9 ..td3 ..te7 1 0 'ii'b 3!
the manoeuvre ... �a5-c4. White has a territorial advantage to add to
b) 8 i.e2 c5 9 0-0 d6 (9 ... cxd4 10 exd4 d5 his superior development, which he uses to
is, of course, also possible) 10 dxc5 bxc5 1 1 generate an attack.
'ii'c2 �bd7 1 2 :fdl 'ii'b6 1 3 .:.d2 :fd8 with 1 o d6 1 1 h41? ltid7 1 2 0-0-0 c6? !
. . .
equality in Karpov-Polugaevsky, Bugojno Slightly passive. After 1 2 ... c5!? 13 dxc5

1 980. 1 0 :cl �bd7 1 1 a3 :c8 1 2 i.g3 a6 1 3 dxc5 14 �c3 wliite has a small edge (Kam
b 3 cxd4 1 4 �xd4 ii'c7, Groszpeter-Tal, So sky) .
chi 1984, was at least equal for Black. The 1 3 g4 'ii'c 7 1 4 'itrb1 a6 1 5 g 5 c5 1 6
move a2-a3 does not improve White's posi 'ii'c 3!? .:gs 1 7 :hg 1 hxg5?!
tion. Now the g5-square becomes available for
8 lt:lxe4 ! ? White, with attacking possibilities on e6 and
This should not be dangerous for Black at f7. Better is 1 7 ... 0-0-0 with a messy position
all, but neither is 8 i.g3 i.b4! 9 ii'c2 d6 1 0 where anything could happen.
i.d3 i.xc3+ 1 1 bxc3 £5 (1 1 ...�xg3?! is too 1 8 lt:lexg5!
early: after 1 2 hxg3 �d7 13 e4 ii'e7 1 4 0-0 1 8 �fxg5?! allows 1 8 ...d5! etc.
64
5 fl c 2 a n d 5 iJ.. g 5
'i'e7 with a good position for Black in

Moutousis-Nikolaidis, Glyfada 2001 . An
other suggestion for Black is 7 ... i.e7. Then,
after 8 e4, 8 ... d6 is too passive - 9 i.g3 li:Jbd7
1 0 i.e2 e5 1 1 d5 lll f8 1 2 b4 lllg6 1 3 li:Jb3 h5
14 h3 h4 15 i.h2 0-0 16 0-0 gave White the
advantage in Ibragimov-Bischoff, Pulver
muehle 2000. Black's position is very solid
but too passive, and White can hope to bene
fit from his space advantage on the queen
side. Instead 8 ...lllxe4!? is another prospect
altogether: 9 i.xe7 lllxc3 10 i.xd8 lllx d1 1 1
i.xc7 lllxb2 1 2 a4 d5 1 3 cxd5 exd5 1 4 .l:ta2
.
1 8 . . . cxd4 ! ? 1 9 exd4 iJ..f 6 20 'il*'b3! lll c4 1 5 lllxc4 dxc4 16 i.xc4 with equality.
Shooting at e6 and f7 while taking control However, this position looks a little nasty for
of the light squares on the queenside. Black, so I would feel less inclined to test this
20 . . . il.. x f3? line ...
From here on Black is defenceless on the 6 . . . h6 7 il..h 4 g5 8 il..g 3 lLle4
light squares. A lesser evil is 20 ... lll f8 21
i.e4!? 0-0-0 with a poor - but not losing -
position.
21 ltJxf3 l:!.h8 22 l:!.g4!
A strong, active post.
22 . . . �e7 23 :e 1 b5!? 24 cxb5 axb5 25
lLlg5! il..x g5 26 hxg5 g6
White was considering g5-g6 himself.
27 l:!.ge4 �dB 28 il..x b5 d5 29 l:!.4e3 l:!.b8
30 J:tc 1 iib6 31 a4 'ilt'xd4 32 'ifc2 �b6
33 l:!.c3 l:!.h4 34 l:!.c6 1 -0
Black resigned due to 34 .l:f.c6 'i'a7 35
.l:f.c8+ rl;e7 36 i.xd7 etc.
This is the simple way to equalise against
Game 23 this system. White now has no path to an
Portisch -Popovic advantage.
Ljubgana 1985 9 e3
White can also go for complications, but
1 lLlf3 lLlf6 2 c4 e6 3 lLlc3 iJ.. b4 4 d4 b6 Black is okay after 9 i.e5 f6 1 0 d5 exd5 1 1
5 il..g 5 iJ.. b 7 cxd5 i.xc3+ 1 2 i.xc3 i.xd5 1 3 0-0-0 lllxc3
5 ...h6 6 i.h4 i.b7 and now 7 'i'c2, along 14 'i'xc3 i.f?! 15 h4 g4 16 lll e5 i.h5 17 'i'c2
with 7 llld2, is aimed at gaining control of e4. fxe5 1 8. 'i'f5 'i'e7 1 9 'i'xh5+ rl;d8 with dy
Both moves weaken the d4-square and namic �quality in Portisch-Timman, Hilver
should therefore be countered by ... c7-c5. sum 1 984.
6 �c2 9 . . . iJ..x c3+ 1 0 bxc3 d6
6 lll d2 h6 7 i.h4 c5! 8 a3 cxd4 9 axb4 A natural move. Black can also choose to
dxc3 1 0 bxc3 0-0 1 1 f3 d5 12 e3 a5 13 bxa5 exchange the bishop immediately, which
bxa5 14 i.e2 li:Jbd7 1 5 cxd5 exd5 16 0-0 should lead to the same type of positions:
65
1 0 ... ll'lxg3 1 1 hxg3 ll'lc6!? (Black decides that Ael cl;g7 27 a4 1/2-1/2 And Black cannot use
•.
c4-c5 is the only White way to make sense of his activity for anything, as the rook must
the position, so he avoids creating a 'hook' stay on f2 (otherwise White has Afl), while
with d6 - 1 1 ...d6 12 c5 ll'ld7 13 cxd6 cxd6 1 4 White cannot try to win as Black is too active
a 4 'itif8 1 5 i.d3 Ac8 1 6 a 5 bxa5 1 7 e 4 cl;g7 to make a single pawn important.
1 8 0-0 ended peacefully in Ehlvest-Adams, b) 14 Af2 is the most natural.. 1bis does
Pula 1 997) 12 Abl (in Langeweg-Karpov, not mean that White immediately wants to
Amsterdam 1 981 , the following manoeuvre play Aafl , but that he can do so whenever he
did not favour White: 1 2 ll'ld2 'flle7 1 3 'f!lb2 pleases, without spending time in prepara
g4 14 ll'lb3 'ii'g5 1 5 c5 Abs 16 'ii'a3 h5 1 7 tion. Quite simply, his position becomes
Ah4 i.aS 1 8 i.e2 f5 and Black i s already more flexible. We are following Zvjaginsev
better) 1 2...'f!le7 1 3 c5 h5 1 4 i.e2 ll'la5 1 5 Timman, Biel 1 995, which continued
ll'ld2 i.xg2 1 6 l:.xh5 0-0-0 1 7 cxb6 axb6 1 8 1 4 ... 0-0-0 1 5 i.e4 l:.hf8!? 1 6 l:.el (intending
Axh8 Axh8 1 9 i.. £3 i.xf3 20 ll'lxf3 'ii'f6 fa to meet 1 6 .. fS with i.xb7 and e3-e4 with a
.
voured Black in Gheorghiu-Miles, London good game) 1 6 ... i.xe4 1 7 'ii'xe4 cl;bS 1 8
1 980. 'ii'h7 Ah8 1 9 'ii'c2 h 5 with unclear play. Now
1 1 .i.d3 tl:lxg3 1 2 fxg3! ? Black slightly misplayed his position with 20
e4 h4 21 gxh4 gxh4 (21 . ..g4 22 ll'ld2 'ii'xh4
23 g3 'f!lh6 is better, with chances for both
sides) 22 e5 with a modest lead for White. 1 6
a4!? was suggested by Timman a s a possible
improvement. It is hard to say anything
about that...
14 ... .i.xe4 1 5 'fkxe4 0-0 16 tl:ld2 c5 1 7
g4 :ac8 1 8 :f2 �g7 1 9 :an :ce8
1bis decision is quite common in these

positions. White wants to build up his artil
lery with the rooks on f1 and £2, aiming at f7.
Black should be able to meet this plan with
out fear, but some attention is required, of
course.
1 2 . . . tl:ld7 1 3 0-0 'fke7 1 4 .i.e4
Other, equally good, options are:
a) 14 ll'ld2 0-0 15 g4!? seems rather exotic, The position is more or less equal. Now
and Black gets a good position after 1 5 ... ll'lf6 the longest arms won the fight.
1 6 h3 h5! 1 7 gxh5 ll'lxh5 1 8 g4 ll'lg3 1 9 Af2 20 ii'b7 tl:lbS 21 'fkf3 f5 22 ii'g3 tl:ld7 23
f5 20 <ith2 ll'le4!, the pawn sacrifice provid �h 1 fxg4 24 'fkxg4 :xf2 25 :xf2 :ta 26
ing enough compensation to draw, but not :xf8 �xf8 27 �g 1 tl:lf6 28 'fkf3 �g7 29
more. Van Wely-Anand, Tilburg 1 998 con e4 cxd4 30 cxd4 e5 31 'ikd3 exd4 32
tinued 21 ll'lxe4 fxe4 22 :.Xf8+ Axf8 23 'fkxd4 co1>f7 33 �f 1 'fie& 34 �e2 'ikg4+
i.xe4 i.xe4 24 'fllxe4 l:.£2+ 25 cl;gl 'f!lf7 26 35 �d3 'ikxg2 36 e5 'flh3 + 37 �c2
66
5 'ii' c 2 a n d 5 i.. g 5
'ii'f 5+ 38 c,ilb3 dxe5 0-1 queenside and is overrun completely. 13 £3?

------.. is how this happens: 13 ...h4 14 i.£2 0-0-0 1 5
Game 24 h3 :dg8 1 6 i.e2 lCih5 1 7 'i'a4 <ittb 8 1 8 c5
Belozerov-Nikole nko dxc5 1 9 i.a6 i.a8 20 lCib3 c4! 21 i.xc4 £5 22
Moscow 1999 i.e2 g4! (the final breakthrough; it is clear
now that Black is the quicker) 23 fxg4 fxg4
1 d4 e6 2 lbf3 lLif6 3 c4 b6 4 lL!c3 i.. b4 24 i.xg4 lCihf6 25 lCic5 lCixc5 26 dxc5 lCixg4
5 i..g 5 i.. b 7 6 e3 h6 7 i.. h 4 i.. x c3 + 8 27 hxg4 'iWxc5 28 l:tfdl h3 29 e4 h2+ and
bxc3 d6 9 i..d 3 White resigned in Borik-Speelman, Dort
This move is less good than the theory's mund 1 981 . The alternatives are no better,
favourite, 9 lCid2!, for obvious reasons. After e.g. 15 a4 a5 1 6 l:tbl g4 17 fxg4 h3 1 8 g3
the bishop is developed White is forced to :lhg8 1 9 i.e2 lCie4 20 i.f3 lCixd2 21 i.xb7+
hurry with kingside castling as the knight <itfxb7 22 'i'xd2 :Xg4 23 'ii'e2 £5 with a clear
cannot move, and White cannot build his advantage for Black in Rohde-Benjamin,
usual impressive centre. USA 1 986, or 1 5 e4 :ldg8 (1 5 ... lCih5!? 1 6 c5!?
9 'ii'e 7
• . . dxc5 17 'iWa4 lCif4! 18 i.a6 lCib8! gave Black
a good position in Hjartarson-Stoica, Taflfe
lag-Politehnica 1 987) 1 6 h3 lCih5 1 7 l:tel lCif4
1 8 i.c2 c5 1 9 lCifl £5 20 ex£5 g4!! 21 fxg4
i.xg2 with an overwhelming attack in Ne
nashev-Tiviakov, Groningen 1 997.
1 3 . . . .C.gS
13 ... lCig4 has also been tried, but it seems
that White gets better control of the kingside
after 14 lCif3 gxh4 1 5 lCixh4 0-0-0 1 6 e4
(Beckmann-Steil, Germany 1 990) than he
normally does.
14 f3 0-0-0 1 5 hxg5 .C.xg5 1 6 i.. h 4 .C.g7
1 0 0-0
Practically the only alternative is 10 'ii'c2,
which does not pose Black problems after
1 0 ...gS 1 1 i.g3 h5 1 2 h3 (now this is forced
because 12 �gS? drops a piece to 12 ... h4 1 3
i.f4 e5, while 1 2 h 4 g4 1 3 tDd2 i.xg2 drops
a pawn) 1 2 ... h4 13 i.h2 l:tg8 1 4 e4 lCic6 1 5
lCid2 e5 1 6 d 5 and the bishop i s now buried
on h2, and White will have to do some ma
noeuvring to get it out. First the other bishop
needs to get outside the pawn chain:
1 6 ... lCib8 1 7 i.e2 tDbd7 1 8 i.g4 lCixg4 1 9
·
hxg4 lCic5 2 0 f3 i.c8 and a draw was agreed 1 1 .C.f2
in Beliavsky-Gulko, Munich 1 990. The alternative is very sharp, but perhaps
1 o . . lL!bd7 1 1 lL!d2 9 s 1 2 i.. g 3 h5 1 3 h4!
. the best way for White to place his pieces
This is absolutely forced, otherwise White might be 1 7 tDe4. If not Black has the fol
has no control over the dark squares on the lowing active continuation: 17 ... :dg8 1 8 l:lf2
67
lL!xe4! (Black will have to do something White a winning attack) 28 cxd4 would have
about the pin eventually, and this is a good kept the struggle unclear.
way - a spectacular queen sacrifice) 19 i..xe7 1 9 . . . We8 20 Wc2 lt:lg4!
lLlxf2 20 �xf2 :Xg2+ 21 �e1 l::th2 22 i.. fl A cute tactic that allows Black to break
l:.g1 23 e4 (White has nothing better since t'3 through to g2, albeit at a price.
must be protected; after 23 l::tc 1 h4! 24 l:.c2 21 l:lfe2
:Xfl+ 25 �xfl l:ht+ 26 �e2 LO+ Black 21 fxg4 l:txg4 22 l::t f4 :Xg2+ 23 �fl e5 is
should win). Now Black should choose be inferior for White since Black's attack is even
tween three moves: stronger and White's forces more exposed.
a) 23 ... fS? 24 exfS exfS 25 'ii'c 1! (address 21 . . . lt:lxe3! 22 l:lxe3 l:lxg2+ 23 *t 1 ?
ing the unwelcome rook on g1) 25 ... i.. a6 26 Losing by force. It is understandable that
'ii'e3 l:.xfl+ (26...l:hg2 27 'ii'e 61 and Black White did not want to go on to the long
cannot improve his position easily) 27 �fl diagonal with 23 �h1 , but at least this brings
l:.h1+ 28 �£2 l::tx a1 29 'ii'e 6! and White is with it the possibility of putting something
winning. Black severely lacks co-ordination. on e4. It is much harder to do something
b) 23 ... d5!? seems to be the safest policy. about g1 . Anyway, Black is probably still
After 24 exd5 (24 'ii'd3 i.. a6 25 exd5 exd5 26 better here, and it is not easy to recommend
'ii'e3 l::tgh 1 and Black wins) 24 ... exd5 25 'ii'a4 since the comer is no place to be during a
dxc4 26 l:.c1 b5! 27 'ii'xb5 i..x t'3 28 'ii'xc4 raging attack. White does avoid having to
l::tg4 29 'ii'a6+ �b8 30 i.. d3 l:tg1+ 31 i.. fl worry about a rampaging h-pawn, though,
l:tg4 a draw results. and has slightly better piece co-ordination
c) 23 ... h4 24 i..xh4 l:.xh4 25 �£2 l:tg5 with a here than in the game. 23 ...l:.2g4 24 i.. £2
complicated position in Barsov-Schoenthier, 'i'e7 25 lL!e4 l::t4g7! 26 i.. fl h4 looks good,
Germany 1 996. I am a little sceptical about with the two pawns plus good co-ordination.
Black's position as I cannot find a good 23 . . . l:lh2 24 .tf2 'ile7 25 lt:le4 l:lgg2 26
square for the knight, but perhaps most GMs : 1 e2 .txe4!
would see great options in ... f7-f5 and ... lLlf6.
1 7 . . .l:ldg8 1 8 Wa4
1 8 lL!e4 transposes to the previous note.
1 8 . . .*b8
1bis seems more healthy than 18 ... e5, al
lowing 1 9 i.. fS �b8 20 e4 when White looks
good. After 20 ... lL!f8?? 21 i..xf6 he had al
ready won(!) in Sideif Sade-Guedon, Cappelle
1 995.
1 9 l:le1 ! ?
I have most faith in closing the long di
agonal with 1 9 e4. Then Farago-Rechlis,
Beersheba 1 987 continued 19 ... 'ii'e 8! 20 c5
dxc5 21 i.. b 5 'ii'c 8 22 e5 a6 23 exf6 axb5 24 Removing a defensive piece for one not
'ii'xb5 l:tg6 25 a4 e5 26 a5 exd4 27 axb6? taking part in the attack on White's second
lLlxb6 (now Black has a clear advantage) 28 rank. Pure logic.
�h2 dxc3 29 lLlfl 'ii'fS 30 g3 l:tg4 31 'ii'e2 27 l:l.xe4
l::tb 4 32 lL!e3 'i'e5 33 f4 'ii'e4 34 lLlc2 l:tb2 35 There is nothing left to do. 27 fxe4 'i'g5
l:.e1 'ii'xe2 and White resigned. Instead 27 a6! 28 i..g3 (28 l:td2 llh1+ 29 �e2 :Xf2+ 30
i.. d 5 (27 ... i.. a 8 28 a7+ �b7 29 lL!e4 gives �£2 'ii'g 1+ 31 �8 l:.h3+ 32 �f4 'ii'xe3
68
5 ii c 2 a n d 5 i.. g 5
mate) 28 ... l:.xe2! 29 i.xe2 l:.hl+ 30 �£2 h4 8 . . . lbbd7 9 f3 h6 1 0 i.. h 4 iie7 1 1 i..d 3
and Black wins. There is no difference between this and 1 1
27 . . .iig5 e4 in practice. After 1 1 ...eS there are two
... l:.h1 is now a deadly threat. alternatives to 12 i.d3, which transposes to
28 .l%h4 the game:
The battle is over. 28 :ld2 l:.hl+ 29 �e2 a) 1 2 i.e2 lt:lf8! (this is the right way to hit
l:.xf2+ 30 �xf2 •gl+ 31 �e2 •e1 mate or the bishop) with a further branch:
28 i.g3 •xg3 29 l:.xg2 l:.hl+ and wins. al) 13 lt:\£1 lllg6 14 i.£2 lt:\£4 1 5 ll'le3 and
28 . . . .1%xf2+ 29 .1%xf2 .1%xh4 30 .1%g2 iie3 now in Khalifman-Hulak, Bled, 1 991 the
31 .1%g8 + <iii b 7 32 �g2 lbf6 33 .1%g7 iih6 natural 1 5 ... 0-0 was, for some reason, new at
34 .1%g3 l::U4 35 c5 h4 36 i.. a 6+ ! <iitb 8 37 the time. There followed 1 6 i.f1 ?! (planning
.1%g8+ lbxg8 38 'i'a4 'i'g5 + 39 �h 1 lLlf6 g2-g3 to chase away the knight, but this gives
40 'i'c6 0-1 Black a chance to generate a substantial lead
White resigned before Black could play in development) 1 6 ... l:.fe8 1 7 g3 (1 7 dS!? was
40 ...•dS. probably becoming necessary) 17 ... i.xe4!!
Game 25
Piket-Tiviakov
Wyk aan Zee 1994
1 d4 lLlf6 2 c4 e6 3 lbc3 i.. b4 4 lLlf3 b6
5 i.. g 5 i.. b 7
Normally the moves 5 ...h6 6 i.h4 i.b7
are played here, but sometimes they are de
layed. It does not really change anything, as
long as they are played eventually.
6 e3 i.. x c3+ 7 bxc3 d6
1 8 fxe4 (1 8 gxf4 exf4 1 9 fxe4 f:xe3 20 i.g3

lllxe4 gives Black a completely winning at
tack; the e3-pawn is worth a piece) 1 8 ... exd4!
(1 8 ... lllxe4 1 9 dS! keeps the position closed)
1 9 •xd4 (1 9 cxd4 lllxe4 20 i.g1 dS! with a
powerful attack) 1 9 ... cS 20 'i'd2 lllxe4 21
•c2 'i'f6!! (a brilliant quiet move) 22 i.g1
(22 gxf4 lllxf2 and White loses) 22...lllxc3 23
gxf4 dS! with a decisive attack .. According to
Khalifman 16 0-0, with equality, is an im
provement. After seeing the game one tends
to agree.
This is the move I would recommend for a2) 1 3 •a4+ •d7 14 'iib3 does not seem
Black. With sensible play from Black equality too smart. Now Black has 14 ...ll'lhS! 1 5 cS
is possible. lllg6 1 6 i.£2 dxcS 1 7 dxcS ll'lhf4 1 8 i.f1 0-0
8 lbd2 and White was on the way to trouble in
This is the serious move here. After £2-£3 Campos Moreno-Polugaevsky, Oviedo 1 991.
and e3-e4 White has a powerful centre. a3) 13 i.£2!? seems more to the point, but
69
Q u e e n 's In dian D e fe n c e
after 1 3. . .�g6 14 g3 0-0 1 5 � fl c6!, with the i.. d l c5) 14 ...�f4 15 �e3 �xd3+!? (directed
idea of ... d6-d5, Black is able to use his lead against i.. c2 and g2-g3) 1 6 'i'xd3 �h5! (not
in development constructively. In Dokhoian 1 6 ... 1td7 1 7 :d1 1Wa4 1 8 0-0 0-0 1 9 i..g3
Djuric, Philidelphia 1 989, 1 6 d5 l:.ac8 1 7 l:.fe8 20 l:.f'2 �h5 21 l:.b2, when White had
1Wd2 b5! 1 8 dxc6 i..xc6 1 9 �e3! bxc4 2 0 0-0 an enduring advantage in Sokolov
1Wd7! 21 i..xc4 i.. b 5! saw Black completely Govedarica, Belgrade 1 987; 1 6 ...g6? runs into
equalise. 1 7 i..h4Q 1 7 0-0-0 g6 1 8 1tc2 0-0-0 1 9 l:.d2
Note that here 1 2 ... g5?! 1 3 i.. f'2 �h5 1 4 f6 and the position is unclear, with chances
g3! seems to guamtee White an opening ad for both sides.
vantage, e.g. 1 4... �g7 1 5 0-0 0-0-0 (1 5 ... h5 1 6 1 3 .i..f 2 lLih5 1 4 lL!f 1 exd4?
c 5 dxc5 1 7 i.. b 5 0-0-0 1 8 1Wa4 �b8 1 9 �b3 Handing White a large advantage, but
with an attack - Epishin) 16 l:.e1 c5 1 7 a4 a5 nothing leads to equality.
1 8 llb1 l:.hf8 (1 8 ... £5 19 exf5 �xf5 20 i.. d3 a) 1 4... £5 1 5 �e3 f4 1 6 �£5 'ir'f6 1 7 g4 has
l:.hf8 21 i.. e4 is also better for White - Ep been assessed as unclear, but I find that silly.
ishin) 1 9 �fl I and White had a structural With the knight on £5 and a strong structure
advantage in Epishin-Shneider, USSR 1 990. in the centre White appears to be better.
When White's knight comes to d5 Black b) 14 ... �f4 does not work here. After 1 5
cannot really take it but, unfortunately, nor �e3 g4 1 6 0-0! White already has an enor
can he leave it! mous advantage: 1 6 ...gxf3 1 7 'i'xf3! exd4
b) 1 2 'ir'a4 is not really dangerous. The (1 7 ... �xd3 1 8 ti:Jf5! 'i'g5 1 9 i..h4 leads to a
only thing White has in this position is the substantial plus due to 1 9 ...1tf4?! 20 �g7+
pin from h4-e7, and Black can unravel with �£8 21 �e6+! fxe6 22 1Wxd3, when White
out allowing �d2-f1 -e3-f5, which at times is wins) 1 8 �d5! �xd5 (1 8 ... i..xd5 1 9 'i'xf4
very annoying. Flear-Salov, Szirak 1 987 con �c5 20 exd5 �xd3 21 1txd4 �e5 22 i..h4
tinued 1 2 ... 0-0 13 i.. e2 1We8! 14 l:tbl �h5 also wins for White) . 1 9 exd5 and Black did
with equality. After 15 0-0 �f4 1 6 l:.fe1 £5 1 7 not make it to move 30 in Khalifman
i.. fl �h8 Black's position looks like it is yemelin, Russia 1 996. The dark squares are
more pleasant to play... simply too weak.
1 1 e5 1 2 e4 g5( ? ! )
••• 1 5 cxd4 f5 1 6 lL!e3! fxe4
I a m a little suspicious about thi s move.
The hole on £5 is big enough to consume all
the pieces I have blundered during the years.
The alternatives are:
a) 12 ... 0-0-0!? is quite a normal move that
has been suggested in numerous places but
never tried.
b) 12 ... �£8! begins a logical manoeuvre.
Piket-Ljubojevic, Monaco 1 994 continued 1 3
� fl �g6 1 4 i.. f'2 (I think thi s i s the best
move because it keeps the option of g2-g3
open; in van der Wiel-Ljubojevic, Amster
dam 1 986 Black was never really worse: 1 4
i..g3 �h5 1 5 �e3 �gf4 1 6 'ir'a4+ � £8 17 1 7 lLif5!
l:.dl �xg3 1 8 hxg3 �e6 1 9 i.. e2 g6 20 �f'2 An important trick to remember.
q;g7 21 1Wa3 1Wg5 22 1Wc1 l:.ad8 23 1td2 h5 1 7 Wf7 1 8 fxe4
• • .
24 �d5 1txd2 25 :Xd2 f6 26 �e3 �fl 27 White has a very strong position. Now
70
5 'il c 2 a n d 5 j_ g 5
'ii'xh5 is a threat. 3 3 . . . ltlxa4 34 :xh8+ �d7 35 l:.xa8

1 8 . . . ltlf4 1 9 0-0 ltlxg2!? j.xa8 36 d5 1 -0
Black is trying to create counterplay as the
'normal' 1 9 ... lLixd3 20 'ii'xd3 0-0-0 21 �g3 Game 26
'ii'h7 looks very dangerous for Black. Here Miles-Timm an
White can try 22 �xd6!? cxd6 23 lLixd6+ Ti/burg 1986
'it>b8 24 'ii'g3 'it>a8 25 :n 'ii'g6 26 :an with
a strong attack against the king (Piket). 1 d4 ltlf6 2 c4 e6 3 ltlc3 i.b4 4 ltlf3 b6
20 j_g3 ltlf4 2 1 j.xf4 gxf4 22 :xf4 5 .tg5 j.b7 6 e3 h6 7 j.h4 g5! ?
'ilg6+ 23 :g4 'ilf6 24 ltlg7 + �dB 25 'This used to be the main line, championed
�h 1 ? by Timman, but these days it is seen less and
'This is unnecessary. Preferable is the logi less.
cal 25 �c2! with an advantage after 25 ... hS 8 .tg3 ltle4 9 'ilc2!
26 l:tg2 l:tg8 27 lLixh5 ::r.xg2+ 28 �xg2 'ii'g5+
29 lLig3, when Black has nothing concrete
for the pawn.
25 . . . �cS 26 j.c2 j.a6?
A terrible mistake, simply overlooking the
reply. Better are 26 ... c5! and 26 ... l:tfB!?, both
with counterplay.
27 j.a4!
Eyeing c6. 'This is much more important
than the c-pawn.
27 . . . j.b7 28 'ild3 ltlc5
9 lLid2 is dealt with in the following game.

9 . . . j.xc3+ 1 0 bxc3 d6 1 1 j.d3 f5
1 1 ...lLixg3 1 2 fxg3 leads to Game 23 by
transposition. 'This variation is a good option
for Black.
1 2 d5 ltlc5
'This is the old move, and probably good
enough for some kind of equality, but the
new 12 ... lLid7!? is certainly more logical.
Then 13 �xe4?1 gives away all the light
squares without any concessions. After
29 :f 1 ! ltlxe4 1 3 ... fxe4 1 4 'ii'xe4 'i'f6 1 5 0-0 0-0-0 Black
29 ...'ii'xg7 30 'ii'h3 !. seems to have a good position despite the
30 :xf6 lL'lf2+ 3 1 �g 1 ltlxd3 32 ltle6 pawn qeficit. 1 6 'i'xe6 1i'xe6 1 7 dxe6 lLic5 1 8
ltlb2? lLid4 l:tde8 saw Black enjoy a good ending in
Losing immediately, but after 32 ... 'it>b8 33 a few games, among them Hort-Bellon,
l:tg7 l:tc8 34 ::r.ff7 lLib2 35 �b3 a5 36 lLixc7 Hastings 1 975. The score so far is two wins
White also has a winning position. for Black and one draw. Instead White
33 :xh6! should play 1 3 lLid4 lLidc5 1 4 dxe6 and now:
No squares left for the pilgrim! a) 14 ... l:tfB 1 5 �e2 g4! 1 6 f3 lLixg3 1 7
71
hxg3 'i'g5 1 B 0�0-0! 0-0-0 1 9 i.d3 gxf.3 20 1 6 . . . ltlxe6 1 7 .txf5 ltlg7 1 8 .i.g6 + 'it>d7
gxf.3 c!l)xe6 21 i.xfS l:txfS 22 'i'xfS 'i'xfS 23 1 9 f3 l:af8!
c!l)xfS i.xf.3 24 l:txh6 i.xd1 25 �xd1 c!l)g5 26 Preparing a possible getaway to the
g4 l:tfB?! (26 ... 'ili>b71 with compensation was queenside for the king.
better according to Tisdall, but it still feels as 20 fxg4 'i'e7 2 1 e4 'iPc8
if Black is fighting for equality here) 27 'ili>e2 Also possible is 21...c!l)c51? 22 'ii'e2 'ili>cB
with an advantage to White in Zvjaginsev 23 e5 i.a6 24 e6 c6 25 :Xm+ :XfB 26 l:tf1
Hiibner, Elista 1 99B. l:txfl+ 27 'i'xf1 'ili>dB 2B 'iWf4 i.b7 29 'ii'xh6
b) 1 4...'i'f6 appears to be the most natural cxd5 30 i.xd6 1i'xd6 3 1 'ii'h s+ �c7 32
move. After 1 5 f.3 f4! 1 6 i.xe41 (forced - 1 6 'ii'xg7+ �dB 33 'i'hs+, with a perpetual in
exf4? c!l)xd3+ 1 7 'i'xd3 c!l)c5 1 B 'ii'e2 gxf4 Ivanchuk-Anand, Monaco 1 993.
was very dangerous for White in Komarov 22 'ifd 2
Mantovani, Reggio Emilia 1 996) 1 6 ... i.xe4
1 7 fxe4 fxg3 1 B hxg3 0-0-0 the situation is
unclear, although I feel a bit more comfort
able about White's position. The pawns
might look ugly, but they control many inva
sion squares.
1 3 h4 g4 1 4 ltld4 'i'f6 1 5 0-0 ltlba6
This is the only move these days since
1 5 ... c!l)xd3 16 'ii'xd3 e5 17 c!l)xfS i.cB
2 2 . . . ltlc5?!
This game was a sad episode for Black
and, later, play was improved with 22... 'ili>bB!,
e.g. 23 l:txfB+ :txfB 24 'ii'xh6 i.cB (this is the
idea behind ... 'ili>bB - the bishop is badly
needed on the kingside to stop the pawns) 25
l:te1 i.xg4 26 c5 'ii' f6 27 cxd6 i.h5 2B e5
'i'xg6 29 'i'xg6 i.xg6 30 e6 c!l)c5 31 d7
c!l)xd7 32 exd7 l:ldB 33 l:le6 i.h5 34 i.e5
was met with 1B f4!! in Miles-Beliavsky, :Xd7 35 l:lh6 i.t7 36 i.xg7 i.xd5 37 i.e5
Tilburg 1 9B6, practically winning on the spot. i.xa2 V2-V2, Kasparov-Timman, Hilversum
There followed 1B ... 'ii'xf5 19 e4 'ii'h5 20 fxe5 1 9B5, or 23 'ii'd4 c!l)eB 24 l:lt7 :Xt7 25 'i'xhB
dxe5 21 c5 �dB 22 d6 'i'eB 23 dxc7+ 'ili>xc7 :m1 26 'ii'xh6 ll:ic5 27 h5 c!l)d7 2B i.£5 i.cB
24 'i'd5 ll:ic6 25 l:lt7+ i.d7 26 :taf1 l:ldB 27 29 l:lft ll:ie5, which left Black okay in Salov
: 1 f6 �cB 2B cxb6 axb6 29 'ii'h 5 1 -0. Timman, Saint John 1 9BB.
1 6 ltlxe6! ? 23 l:xf8 + l:xf8 24 'i'xh6 'i'f6 25 .tf5 +
Thi s sacrifice i s the only way fo r White to ltlxf5 26 'ifxf6 l:xf6 27 exf5 .ta& 28 .tf2
prevent his centre from collapsing. After 1 6 .txc4 29 .id4 l:f7 30 f6 l:h7 3 1 .txc5
dxe6 Black's bishop would be a truly power dxc5 32 l:d 1 !
ful piece and Black would rock n' roll on the Keeping the bishop out of the game.
light squares. 32 . . . .tb5
72
5 W c 2 a n d 5 .i. g 5
32 ...:xh4 33 f7 l:h8 34 g5 and White Take the money and run.

wins. 1 0 bxc3 .bc3 1 1 :c 1 i.a5
33 g5 .i.e8 34 :e 1 ! This appears to be the safest path here. It
And now the king. is doubtful that White has gained much for
34 . . . �d7 35 g3 b5 36 �g 2 a5 37 �h3 his pawn. But other moves have also brought
b4 38 cxb4 axb4 39 g4 :ha 40 h5 @d6 good results.
41 J:l.xe8! 1 -0 a) 1 t . ...i.xd2+ 12 'ii'xd2 d6 13 c5 (13 £3? is
too slow: 1 3 ... lbd7 14 .i.d3 'ii'e7 1 5 0-0 f5 1 6
c5 dxc5 17 i..xc7 0-0 1 8 dxcS lbxcS 1 9 .i.e5
l::t ad8 and White resigned in Priehoda
Rozentalis, Tmava 1 988) 1 3 ... dxcS 1 4 'irh21?
(harmless is 14 dxc5 'ii'xd2+ 15 'iii>xd2 i.. a6
1 6 .i.xc7 .i.xfl 1 7 l::thx ft lba6 1 8 .i.d6 lbxc5
1 9 .i.xc5 bxc5 20 l::txc5 l::td8+ 21 'iii>e2 and a
draw was agreed in Smirin-Rozentalis, New
York 1 997) 14 ... 0-0 15 h4 g4 1 6 dxc5 'ii'dS 1 7
.i.e2 and now Black wa s too careful with
17 ... lbc6? in Stocek-Cvek, Plzen 1 997, after
which the positional aspects of the position
begin to count over material. Correct was
17 ...'ii'xg21 1 8 l::th2 'ii'd5 to ruin White's set
Game 27 up. After 1 9 'ii' f6 lbd7 20 'ii'xh6 lbxc5 21
Crouch-Harikrishna 'ii'gS+ 'i'xg5 22 hxgS we have a big mess.
undon 200 1 b) 1 t ....i.b4 1 2 h4 and now the piece sacri
fice 1 2 ...lbc6!? should be the reason to play
1 d4 e6 2 c4 ll:lf6 3 ll:lf3 b6 4 ll:lc3 i.b 7 this system. 12 ...gxh4? is still the main line
5 i.g5 h6 6 i.h4 i.b4 7 e3 g5 8 i.g3 given in ECO, but after 13 l::txh4 .i.d6 1 4
ll:le4 9 itid2?! 'ii'g4 'ile7 1 5 i..xd6 cxd6 1 6 'i'g3! they sug
This is a gambit that might have been gest to improve on 16 ... £5 with 1 6 ... d51?.
playable back in the days when people were About 10 years ago I told Steffen Pedersen
afraid of taking the pawns and defending an the refutation and he put it in his book 1 d4!:
awkward position. But these days - no. 1 7 cxd5 exd5 (what else? 17 ... lba6 is just
9 . . . ll:lxc3! plain bad) 1 8 :c 7 i..a6 1 9 :xa7 and White
wins. But even stronger is 1 9 .i.c4!!, winning
everything after 1 9 ... dxc4 (1 9 ... 'i'a3 20 .i.xa6
'ii'xa6 21 'ile5+ �d8 22 l::tc21 and White wins
material or mates) 20 l::txa7! :xa7 21 'ilxb8+
with a mate in 8 according to Junior. After
1 6 ... £5 White has 1 7 c5 d5 1 8 1i'e5! with a
strong., initiative, as in Agdestein-Hellers,
Gausdal 1 987. Returning to 13 d5, Schussler
Amason, Reykjavik 1 986 went 1 3 ... 1i'e7! 14
dxc6 dxc6 1 5 .i.eS 0-0-0 1 6 .i.c3 (1 6 l::tc2
l::txd2 1 7 l::txd2 l::t d8 and Black wins)
1 6 ... .i.xc3 1 7 :Xc3 c5 1 8 hxg5 hxg5 1 9 l::txh8
:xh8 20 e4 f5 21 l::th3 :d8 22 :d3 :us 23
73
'ir'c2 'ikfl and Black actually won this posi 'ifa3 1 7 .:Id 1 ltib4 1 8 'i'b3 'i'xa2 1 9
tion. Here I will just say that White has no Wxa2 ltixa2 20 f3 i.b4 2 1 .:la1 ltic3 22
easy way to co-ordinate his pieces, and 'it>f2 a5 23 i.f6 g4 24 i.d3 g3+ 25 'it>e 1
Black's bishop does look terribly strong. Still, b5 26 i.h7 :ta 27 i.g7 bxc4 28 i.xf8
my choice would be 1 1 . ..aas.
. . 'it>xf8 29 �xc4 f5 30 i.g6 'it>g7 31 ltie5
1 2 h4 .:lg8 1 3 'i'c2 'iie 7! ltie4+ 32 'it>f1 i.a6+ 33 'it>g 1 lDg5 34 f4
Black lives with the knowledge that c7 34 . . . i.d2!

falls. So what? White is still trapped in a ter The winning move. Once the e3-square
rible pin and g2 is still exposed. All in all it caves in the game is up.
looks as if White does not have any chances 35 .:lh4 i.xe3+ 36 'it>h 1 i.xd4 37 J:[d 1
to gain an advantage in this line, as he does i.xe5 38 i.h5 i.xf4 39 :xd7 + 'iitf6 40
not have the fluent development you usually :xt4 :ha 41 .:lh4 i.e2 42 :a1 :xh5 43
get when you sacrifice a pawn. :xh5 i.xh5 44 :xa5 f4 45 'it>g 1 i.e2 46
1 4 hxg5 hxg5 1 5 i.xc7 ltia6 1 6 i.e5 :as f3 4 7 gxf3 i.xf3 0-1
74
5 'il c 2 a n d 5 i.. g 5
Summary
5 'iic2 i s not ambitious and Black equalises without trouble. 4 lLic3 i.b4 5 i.gS i.b7 6 e3 h 6 7
i.h4 i.xc3 8 bxc3 d6 is a good reliable system for Black. Now 9 lLid2 is the only good move,
but 9 i.d3, which has led to disaster for many White players, is still seen occasionally. 7 ...gS 8
i.g3 lLie4 is also seen at the top level, and after 9 'iicz the game seems to be more or less
equal, even though this is a complex line. The gambit after 9 lLidZ is simply not to be recom
mended.
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lll c 3 (D) i.b4

4 ... i.b7
5 'iic 2 - Game 21; 5 i.gS Game 22
-
5 i.g5
5 'iib 3 - Chapter 4
5 . . . .ltb7 6 e3
6 'iic 2 Game 23
-
6 . . . h6 7 i.h4 i.xc3 +
. 7 . . gS 8 i.g3 lLie4 (D)
9 'iic2 - Game 26; 9 lLidZ Game 27
-
8 bxc3 d6 (D)
9 i.d3 - Game 24; 9 lLidZ Game 25
-
4 t'i:Jc3 8 . . . t'i:Je4 8 . . . d6
75
CHAPTER FOUR I·
5 'ifb3
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lllc 3 i.b4 Nimzo-Indian, or 5 .tg5, but because this is

5 'i'b3 what I would play myself with White in this
In this chapter we will investigate Seira position, as it is clearly the most problematic
wan's pet line which, in my opinion, is one of line for Black to face.
the most difficult for Black to face. Look out 5 . . . a5
particularly for Alterman's nice strategic idea Tbis is one of three ways to protect the
in the first main game, and that of Miles in bishop. 5 .. .'ii'e 7 is considered in the following
one of the sub-variations. game and 5 ... c5 in Game 30.
6 i.g5
Game 28 The main line. Black has no problems af
Alterman-Liss ter the alternatives:
Israel 1999 a) In reply to 6 g3, 6 ... .tb7 7 .i.g2 0-0 8
0-0 .txc3 9 1i'xc3 d6 1 0 b3 c!Dbd7 1 1 .tb2
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lll c 3 i.b4 leads us to a standard position where White
seems to have a tiny edge thanks to the po
tential pressure on the long diagonal, as in
Dreev-Kiselev, Podolsk 1992. Instead
6 ... c!Dc6! is the aggressive approach, e.g. 7
.i.g2 a4! (now this is possible) 8 'ii'c2 .i.a6 9
a3 (9 0-0 .i.xc4 10 c!De5 c!Dxe5 1 1 dxe5 c!Dd5
12 lDxa4 b5 13 c!Dc3 .i.xc3 1 4 bxc3 l:r.a4 was
fine for Black in Labunsky-Tsesarsky, USSR
1986) 9 ....txc3+ 10 'ii'xc3 lDa5 1 1 lDe5?! (1 1
lDd2 d5 1 2 cxd5 lDxd5 1 3 'ii'c2 is unclear
according to Giplis) 1 1 ...d5! 1 2 cxd5 c!Dxd5
and it is already clear that Black has consider
able power in his pieces. That he used it to
.
5 'i'b3! win a superb game in Polugaevsky
I give this move an exclamation mark not Ljubojevic, Linares 1 985, is no surprise: 13
because it is better than 5 e3, leading to the 'ii'c2 c!Db3 14 'ii'c6+ �f8 1 5 .i.g5 'ii'e 8 1 6
76
5 'il b 3
lDd7+ �g8 1 7 e4 lDxd4 1 8 'ii'xa4 cS 1 9 exdS White the advantage due to his influence in
i.bS 20 'ii'd 1 W'xd7 21 i.e3 i.e2 22 dxe6 the centre) 1 3 W'xc3 lDd7 14 lDe1 i.b7 1 5 cS
'i'xe6 23 'iixe2 lDxe2 24 i.xa8 °iic4 25 :td1 White seemed to have some initiative in
lDd4 26 i.xd4 cxd4 27 i.e4 g6 28 i.d3 Waz Miles-Budnikov, Beijing 1991, although
29 .l:.b1 W'b3 30 �d2 �g7 and Black went Black might be fine.
on to win.
b) 6 a3 i.xc3+ (6 ... a4 7 'ii'c2 i.xc3+ 8
bxc3! reminds me of the main game. 8 bxc3
has never been played before) 7 W'xc3 i.b7 8
g3 0-0 9 i.g2 a4! and Black had at least
equalised in Ahner-Bischoff, Germany 1 997
(c4 is a potential problem for White and the
queen's bishop has no good squares).
6 .tb7 7 e3
. . .
White is developing. 7 a3 i.xc3+ 8 "iixc3

d6 9 e3 lDbd7 10 lDd2 'iie7 1 1 £3 eS 12 i.e2
exd4 13 W'xd4 a4 was good enough for
equality in Zaja-Baklan, New Delhi 2000.
7 h6
. . . 9 a3! ?
There is only one really independent way Thi s builds on a very interesting concept,
to play this position, namely 7 ...'iie7 8 "iic2 similar to the one Miles used in the note to
a4!? 9 a3 i.xc3+, and now 10 "iixc3 lDe4! 1 1 Black's 7th move. Others:
i.xe7 lDxc3 1 2 i.h4 lDe4 seems to give a) 9 i.e2?! - it is not easy to see what the
Black equality due to the weakness of c4. bishop is doing here. Black is doing well after
Instead 10 bxc3! is best, the idea being that 9 ... gS 10 i.g3 lDe4, e.g. 1 1 0-0-0 i.xc3 1 2
Black will have some commitment to the a4- bxc3 lDd7 1 3 lDe1 h S 14 h 4 g4 1 5 lDd3
pawn and that White will benefit from a lDxg3 16 fxg3 i.xg2 17 .l:.h2 i.e4, and White
strong centre, as is the case in the 5 i.gS did not have enough compensation for the
lines. Miles-Arkhipov, Miinster 1 993 contin pawn in Conquest-Onischuk, Germany 1 997,
ued 10 ... i.e4 (it is not easy to decide where or 1 1 d5 lDd7 12 lDd4 lDdc5 13 'iic2 i.xc3+
this bishop bclongs, as White will certainly 14 bxc3 eS, and Black was satisfied in Stohl
play t2-8 and e3-e4 soon) 1 1 'iib z h6 1 2 Polak, Olomouc 1 998.
i.h4 0-0 1 3 lDd2 i.h7 1 4 £3 e S 1 5 e4 lDc6 b) 9 i.d3 is the main line here. After
16 i.e2 (planning to re-route the knight to 9 ... lDbd7 1 0 0-0-0 i.xc3 1 1 "iixc3 Black has
dS) 16 ... d6 17 tDn gS 18 i.t2 lDhS 19 lDe3 a variety of valid options:
lLJf4 20 lDdS! W'd8 21 g4! and White had a b1) 1 1 ...a4!? 12 'iic2 l:taS 13 lDd2 gS 1 4
very promising position. i.g3 i.xg2 1 5 .l:.hg1 i.b7 1 6 £3 "iia8 1 7
8 .th4 :tdf1 rJi; e7 1 8 i. e1 c S 1 9 lDe4 .l:. a7 2 0 lDc3
The only serious move. After 8 i.xf6 i.a6 21 f4, Miles-Bischoff, Havana 1 998, and
'iixf6 9 i.e2 0-0 1 0 0-0 i.xc3 1 1 "iixc3 d6 1 2 White; had compensation for the pawn but it
lDd2 lDd7 the position was even in Ivanov was still a battle.
Farago, Philadelphia 1 984. b2) 1 1 .. .W'c8?! 1 2 'iic2! cS 1 3 dS eS 1 4
s d6
... lDd2 gS 1 5 i.g3 'iic7 1 6 f4 and Black experi
Black should develop smoothly. After enced serious problems on the light squares
8 ...gS 9 i.g3 lDe4 1 0 i.d3 d6 1 1 i.xe4! in Khalifman-Tunik, Maikop 1 998.
i.xe4 1 2 0-0-0 i.xc3 (1 2 ... i.b7 1 3 dS! gives b3) 1 1 ...'ii'e7 1 2 lDe1 eS?! 1 3 i.5 0-0-0 14
77
c5! put Black under fire in Miles-Singh, Cal with the a4-pawn.
cutta 1 994. After 14 ... exd4 1 5 cxd6 'iixd6 1 6 1 6 0-0 ltih5 1 7 :te 1 lLlf4 1 8 .i.f 1 'ilg6
:Xd4 lDd5 1 7 'iic2 c5 1 8 lDd3 .tc6 1 9 .i.g3 19 :ad 1 f5 20 exf5 :xf5 21 .i.g3!
'iif6 20 lDe5 cxd4 21 lDxc6 he was ready to
resign.
b4) 1 1 ...g5 12 .i.g3 lDe4 1 3 1i'c2 lDdf6 1 4
l:lhe1 'ile7 with an even position in Khalif
man-Granda Zuniga, Ulcinj 1 998.
9 . . . a4
9 ... J..xc3+!, with the idea of 1 0 'ilxc3 a4!
with equality, is probably a better line. Now
White gets something.
1 0 'i'c2 .bc3+ 1 1 bxc3!
I would like to quote Alterman on this po
sition because he says it all: '1bis was my
idea. It looks as if Black gets a tempo with
... a5-a4 but, as I understand it, the pawn on All White's pieces are well organised here
a4 only presents problems for Black because and Black has managed to do nothing other
it requires constant protection.' than open up the position for them. White is
1 1 . . . ltibd7 1 2 lLld2 'ile7 1 3 f3 e5 1 4 e4 clearly better.
0-0 1 5 .i.e2 21 . . . 'ilf7 22 ltie4! <iilih 8 23 ltif2 :tS!?
Leaving the pawn to its own devices
seems a rational thing to do. The pieces need
to play before everything goes wrong.
24 'ilxa4 ltif6 25 . .i.xf4 :xf4 26 dxe5
dxe5 27 ltid3 :t5 28 :xe5 ltih5
Here Black has only one minute left with
which to reach the 40th move, but White
completely overlooks his threat, and the
game takes a sudden tum.
29 l:l.de 1 ? ? :xf3!
Here Black should find a way to organise

his pieces so that he gets out of the pin and
activates his forces. I have a feeling that Liss
does not find the appropriate route.
1 5 . . . 'i'e6?!
1 5 ...'iie 8! improves, with indirect protec
tion of a4 and the manoeuvre ...lDh5-f4 com
ing up, perhaps even ... c7-c6 and ... d6-d5 in
some situations. Liss leaves the e8-square
free for a rook, which seems sensible
enough, but the queen appears to be slightly 1bis rook is obviously taboo.
awkward on e6 and he still has problems 30 'ild 1 'i'g6 31 'ild2? !
78
5 'il b 3
31 'i'e2! is better according to Alterman. I like this move quite a lot. Th e next note
31 . . . lLlf4 32 lLlxf4 l:3xf4 33 g3 l:4f6 34 demonstrates the main reason why (... �e4) .
i.d3 flf7 35 .:e7 flh 5 36 fle2 flc5+ 37 6 a3!
fle3 .:t2 38 i.f1 ? The most challenging set-up. At least it
Presenting Black with the opportunity to earns the bishop pair. Others:
finish the game with a nice tactical blow. a) After 6 i.gS i.b7 7 e3 aS 8 a3 i.xc3+ 9
After 38 'i'xcS .l:.g2+ 39 @hl bxcS 40 i.e4 'i'xc3 �e4! Black steers the game to an end
i.xe4 41 l:r.1xe4 l:r.ff2 42 .l:.h4 .l:.a2 43 a4 ing where c4 is a slight weakness. Seirewan
:tgc2 44 .:e1 .l:.xc3 Black is close to winning, Sokolov, Candidates 1 985, went 1 0 i.xe7
but White retains some drawing chances. �xc3 1 1 i.h4 �e4 12 �d2 �xd2 13 @xd2
38 . . . flc6! a4! 14 £3 �c6 15 i.d3 �aS 16 @c3 i.a6 17
e4 0-0 1 8 i.c2 �xc4 1 9 i.xa4 c6 20 i.b3 dS
21 i.xc4 i.xc4 22 b3 with complete equality.
b) 6 g3 i.b7 7 i.g2 cS 8 a3 i.xc3+ 9
'i'xc3 cxd4 10 'i'xd4 �c6 1 1 'i'h4 �as 12
i.gS h6 1 3 0-0 0-0 14 i.xf6 'i'xf6 1 5 'i'xf6
gxf6 1 6 �d2 J:tfc8 and Black was completely
fine in Dreev-K.arpov, Dortmund 1 994.
However, I think White might be slightly
better after 7 ... 0-0 8 0-0 i.xc3 9 'i'xc3 d6 10
b3 �bd7 1 1 i.b2 i.e4 12 l:.fel . At least
Black was in trouble after 12 ... cS 1 3 l:r.adl
l:r.fd8 14 dS! exdS 15 cxdS bS 1 6 �h4! i.xg2
17 �xg2 b4 1 8 'i'£3 g6 1 9 e4 in Van Wely
0-1 Psakhis, Internet 1 9'95, the bishop making its
Now the control ofhl is deadly. presence felt.
c) 6 i.f4. The bishop is nice here, but not
Game 29 very useful. 6 ... dS 7 e3 0-0 8 a3 i.xc3+ 9
Timman-Karpov 'i'xc3 i.a6! 1 0 i.gS l:r.c8 1 1 cxdS i.xfl 1 2
Hoogeveen 1999 .l:.xfl exdS 1 3 .l:.cl 'i'e6 left Black n o worse
in Alterman-Hracek, Bad Homburg 1 997.
1 d4 lLlf6 2 c4 e6 3 lLlf3 b6 4 lbc3 i.b4 6 . . . i.xc3 + 7 flxc3 i.b7 8 e3
5 'ilb3 fle7 Here I prefer 8 g3 0-0 9 i.g2 as it puts
less restraint on the queen's bishop. Now
9 ... d6 10 0-0 �bd7 1 1 b4 is perhaps a touch
better for White, but this is difficult to prove,
e.g. 1 1 ...aS 12 i.b2 axb4 1 3 axb4 l:tfc8 14
.l:.fdl , Gurevich-K.arpov, Reggio Emilia
1991, when 14...'i'fB!, with the intention of
... .l:.xal ,and ... l:r.a8, puts Black as close to
equality as he can possibly come in positions
like this (without a mistake from White).
Indeed he should not fear such positions.
Gurevich calls it equal, but this is because
White's advantage is so minor that, at his
level, it is insignificant. 1 1 ...�e4 12 'i'c2 f5 is
79
less good fo r Black. Smyslov-Chandler, Hast weakness on the light squares, namely the c4-
ings 1988, went 13 i..b2 liJdf6 14 a4 a5 15 b5 pawn. 1 1 ...ll\e4 12 'i*'d3 f5 tries to put pres
l:tae8 16 ll\e 1 c6 17 bxc6 i..xc6 18 £3 lbg5 1 9 sure on g2, but Black is not in position to
lbd3, and White was better. Look a t all the justify this: 1 3 liJd2 'it'h4 1 4 £3 ll\g5 1 5 d5!
weak pawns on the dark squares! gave White a slight advantage in Van Wely
8 i..g5 lbe4! sees Black equalise immedi Lobron, London 1 995. The point is, of
ately. course, 1 5 ... e5 1 6 f4!, with a fight for a giant
8 0-0
. . . bishop on b2.
8 ... d6 9 i.. e2 lbe4 10 "iic2 lbd7 1 1 0-0 0-0 1 2 i.b2 a5
12 liJd2 f5 13 £3 ltJg5 14 .l:f.£2 e5 1 5 liJfl Here Black has a great range of alterna
exd4 16 exd4 lbe6 gave Black sufficient tives.
counterplay in Seirawan-Karpov, Roque a) 12 ... l:tac8 13 dxc5 dxc5 14 b5 .l:f.fd8 1 5
brune 1 992, but there is still a lot of play in a4, Becker-Pitschak, Bad Liebenwerda 1 934.
the position for White, and the bishop might This kind of position is typical for the line.
later become strong on b2, so I would not White has a very minor plus due to the
recommend playing like this. bishop's superiority over the knight.
9 i.e2 d6 b) 1 2... ll\e4 1 3 'i*'c2 and now 1 3 ... fS?I does
I think this is the most natural way to play not sit well with ... c7-c5. Now White's dark
the position, but also possible is 9 ... c5!? 1 0 squared bishop assumes great influence that
0-0 .l:f.c8 1 1 dxc5 "iixc5 (1 1 ...bxc5 1 2 b 3 fol could otherwise be neutralised later with ... f7-
lowed by 1 3 i..b2 is slightly better for White) f6 if necessary. Moreover Black's second
1 2 b4 "iie7 13 i.. b2 d5 14 °iid4 dxc4 1 5 rank is slightly more exposed, and this is felt
i.. xc4 a5! (1 5. . .lbc6?! 1 6 °iih4 and Black on g7. In Shamkovich-Kholmov, Baku 1 972,
needs to be careful) 16 lbe5 lbc6 1 7 lbxc6 14 dxc5 bxc5 15 b5 liJdf6 16 liJd2 l:tad8 1 7
.l:f.xc6 1 8 bxa5 (1 8 b5 .l:f.d6 19 °iih4 e5 20 a4 £3 lbg5 1 8 .l:f.fdl gave White the advantage .
.l:f.d2 21 i.. a3 °iid8 22 'i*'g3 i.. d 5 is no worse Better is 1 3 . .. .l:f.ac8, e.g. 1 4 dxc5 dxc5 15 b5
for Black at all) 1 8 ... e5! 1 9 °iih4 l:txa5?! was h6 16 .l:f.fd1 .l:f.fd8 17 a4 lbg5 1 8 'i*'c3 lbx£3+
Sokolov-Karpov, Groningen 1 995. Better is 19 i..x £3 lbf6 20 i..xb7 'i*'xb7 with an edge
1 9 ... bxa5, when Black is not worse in any for White in Cebalo-Hohler, Saint Vincent
way. 1 998. Often Black should not fear such posi
1 0 0-0 ltJbd7 1 1 b4 c5! tions as White's bishop tends to have diffi
culties causing any harm all by itself.
c) 12 ... .l:f.fd8 13 .l:f.fdl l:tac8 14 'i*'h3 i.. e4
1 5 ll\d2 i.. a8 1 6 dxc5 dxc5 1 7 b5 lbe4 1 8
lbxe4 i..xe4 1 9 a4 f6 2 0 a 5 ltJ£8 2 1 £3 i..g6
and, because Black's bishop looks so good
and the al -h8 diagonal is limited, Black had
equalised in Gurevich-Rozentalis, Belfort
1 997.
d) 12 ... .l:f.fe8 13 lbd2 l:tac8 14 dxc5 dxc5
1 5 b5 e5 1 6 £3 ltJ£8 17 a4 was slightly better
for White in Burmakin-Zayac, St. Petersburg
1 998. White's set-up here is probably the
most promising.
Black should seek counterplay, and the 1 3 :td 1
only way to do so is to put pressure on a 1 3 dxc5 dxc5 14 b5 .l:f.fe8 1 5 .l:f.fdl l:tad8
80
5 'ii b 3
1 6 tlld2 e5 equalises. Perhaps 1 6 tlle 5!? 26 ... l:ta8 27 .iel ! shows the strength the
Might offer some kind of an advantage. bishops have even in this semi-closed posi
1 3 . . . axb4 1 4 axb4 l:!.fb8! tion.
27 l:!.a 1 l:!.d8 28 'i'e2 tt:ld6
28 ... e5 29 f4! and the black king's defences
crack.
29 f4 �f7 30 e4 tt:lca 31 e5!
Clearing space for the queen in order to

trade major pieces on the a-file.
1 5 tt:ld2 'iid 8 1 6 f3 l:l.xa 1 1 7 ..txa 1 l:!.a8
1 8 ..td3 'ii c 7 1 9 ..tc2 l:!.c8? !
A slight positional mistake. After 1 9 ... e5! White is trying to open the position.
Black has neutralised the pressure on the 31 . . . f5
long diagonal and thus limited the scope of 31 ...tlle7 32 'ifhs+ tlleg6 33 exf6 gxf6 34
the bishop. Now White has a chance to es l:tfl ! and White is ready to launch a direct
tablish a small edge. assault on f6 with f4-f5.
20 dxc5 dxc5 21 b5! 32 °iWf3 tt:le7 33 h3 <l;g8 34 <l;h2!
The pawn structure is now clearly favour Improving the king before the attack.
able for White. His bishops enjoy open di 34 . . .tt:lfg6 35 g3 cj;f7 36 h4! 'it>g8 37 h5
agonals, but Black's knights have no good tt:lh8
squares. 37 ...tll f8 38 g4! - now this is possible as
21 . . . tt:iea Black cannot organise counterplay against f4.
21 ...e5 22 i.f5! is uncomfortable for 38 g4 l:l.f8
White. Thanks to the misplacement of 38 ... fxg4 39 'ifxg4 'ifc8 40 l:r.a7 l:r.e8 41 h6
Black's pieces, e3-e4 followed by tll fl -e3 will with a winning attack.
soon be a relevant option. 39 gxf5 tt:lxf5
22 tt:le4 f6 23 'iid 3! ..txe4 39: .. exfS 40 e6 and the bishop comes to
This move is based on the following tac life.
tics: 23 ... tll f8 24 tlld 6! tllx d6 (24 ... l:f.d8 25 40 ..txf5 l:!.xf5 41 l:ta8 + l:tf8 42 l:txf8 +
tllxe8 l:txe8 26 'ifd6 gives White a great ad c.t>xf8 43 Was + <3;17 44 Wxh8 'iib 7 45
vantage; b6 falls) 25 'ifxd6 (with the threat of h6! 1 -0
"
'ifxc7 and l:td6) 25 ... 'ifxd6 26 l:f.xd6 l:r.a8 27

i.c3 l:f.a2 28 i.d3 i.c8 29 i.fl l:ta3 30 i.e 1 Game 30
and White wins b6 for e3, which gives him a Seirawan-Timman
passed pawn on the queenside and estab Hifversum Match (5th game) 1990
lishes c5 as a weakness.
24 Wxe4 tt:lf8 25 °iWd3 <l;f7 26 ..tc3 <l;e7 1 d4 tt:lf6 2 tt:lf3 e6 3 c4 b6 4 tt:lc3 ..tb4
81
5 'ifb3 c5 1 0 .. .'.c7 1 1 1'.c4 cxd4 1 2 1'.xd5 .Jlxc3+! 1 3

bxc3 exd5 14 cxd4 1'.a6 and Black had the
advantage.
· b) 6 .Jlg5 with a further branch:
bl) Now after 6 ... llic6 there has been
much debate as to whether 7 d5 gives White
an advantage. But the basic fact that the
knight looks misplaced is possibly best illus
trated by the less ambitious, but smooth 7
e3!? h6 8 .ih4 g5 9 1'.g3 llie4 10 1'.d3
.ixc3+ 1 1 bxc3 llixg3 1 2 hxg3 d6 13 .Jle4
.id7 14 .ixc6 1'.xc6 1 5 d5, when White had
an advantage in Condie-Cullip, Edinburgh
1 989.
1bis is the main line. Actually it was this b2) 6 ...1'.b7
match between Timman and Seirawan that b21) 7 0-0-0 1'.xc3! 8 'iixc3 llie4 9 .ixd8
put the 5 'ii'b3 line on the map as a serious llixc3 10 bxc3 �d8 1 1 d5 ri;e7 gives Black
main line, instead of a decent sub-variation. an even endgame.
Seirawan used it in all three of his games with b22) 7 e3 0-0 (7 .. .ixf3? 8 gxf3 llic6 9 d5!
.
White -all of which he won. exd5?! 10 cxd5 llie5 11 f4! llif3+ 12 ri;eZ
6 a3 llixg5 13 fxg5 1'.xc3 14 'i'xc3 llih5 15 'i'e5+
1bis is the most serious move. Alterna 'i'e7 1 6 'i'xe7+ rJ:i;xe7 1 7 1'.g2 rJ:i;d6 1 8 1'.£3
tives: gave White a strategically won position in
a) 6 .Jlf4 0-0 and here White has two ways Van Wely-Miles, London (rapid) 1 995) 8
he can go: .ie2 cxd4 9 exd4 llic6 1 0 0-0 1'.xc3! 1 1 1'.xf6
al) 7 a3 .Jla5 8 l:.dl (8 dxc5 llie4 9 cxb6 (1 1 Vxc3 llie4 and Black appears to have an
axb6 1 0 g3 llic6 1 1 1'.g2 .ixc3+ 1 2 bxc3 edge after 1 2 1'.xd8 llixc3 1 3 bxc3 l:.fxd8,
1'.a6 13 0-0 llic5 14 Vc2 .Jlxc4 15 llie5 1'.b3 when c4 is exposed) 1 1 ...'iixf6 12 'i'xc3
16 Vd2 llixe5 17 1'.xe5 d5 was nice for l:tac8 13 'i'd2 llie7! 14 l:.acl llig6 1 5 l:tc3
Black in Dreev-Timman, Moscow 1 993, until llif4 and Black was doing fine in Timo
he blundered a few moves later...) 8 ... 1'.a6! scenko-Pelletier, Leon 2001 . Actually it might
and the bishop is much better placed here be White who should be careful here.
than on b7. In Garcia Palermo-Eingom, b23) 7 l:tdl !? 0-0 8 e3 cxd4 9 exd4 and
Cienfuegos 1 986 Black organised a quick now 9 ... 1'.xc3+! is a strong theoretical novelty
attack on White's centre: 9 e3 cxd4 1 0 llixd4 of Boris Gelfand which gives Black equality.
d5 1 1 1'.e2 1'.xc4 12 .Jlxc4 dxc4 13 Vxc4 1 0 bxc3 Vc71 1 1 .ixf6 1'.xf31 1 2 gxf3 gxf6
Vd5 and Black had equalised, if not more ... 1 3 'iic2 'i'f4 14 Ve4 'iixe4+ led to a draw in
a2) 7 e3. 1bis must be the way to play Alterman-Gelfand, Tel Aviv 1 999. 1bis end
with 1'.f4 but it fails to eam White an advan ing is equal.
tage. Dreev-lvanchuk, Linares 1 995 contin c) 6 e3!? 1'.a6 (6 ... llic6!? looks so natural
ued 7 ... d5 8 :d1 llibd7 9 cxd5? (9 dxc5 bxc5 here that one wonders why it has not been
10 1'.d3, with an unclear position, is better, played - when there is no 1'.g5, then d4-d5
while 9 1'.d6 :e8 10 1'.d3 1'.a6 seems dan also becomes less of a real threat) 7 a3 1'.a5 8
gerous for White - what is the bishop really 1'.d2 0-0 and now:
doing on d6?) 9 ... llixd5 1 0 1'.g5 (10 1'.d6? c4! cl) 9 l:tdl cxd4 10 llixd4 d5 1 1 llidb5
might have been what Dreev overlooked) 1'.b7! 12 cxd5 a6! 1 3 llid4 llixd5 14 llixd5
82
5 'il b 3
'i'xdS 1 5 'i'xdS .txd2+ 1 6 l:xd2 .txdS 17 f3 .te2 �e4! 9 dS! .txc3+! 1 0 bxc3 .ta6!? 1 1
l:a7! 1 8 e4 .ta8! was even in Seirawan .tb2 1"e7 1 2 lld 1 l:e8! 1 3 .td3 (1 3 0-0?
Browne, USA 1 987. A decent alternative is exdS 14 l:.xdS �c6 with ... �aS to follow)
9 ... dS!? with the following idea: 10 dxcS bxcS 1 3 ...�d6!? Black was doing very well in Sei
1 1 cxdS exdS 12 .txa6 �xa6 1 3 0-0 l:b8 14 rawan-Timman, Hilversum 1 990. 7 .tf4 ..ib7
'i'a2 c4! and Black has good play on the light 8 l:d1 0-0 9 e3 cxd4 10 �xd4 �e4 1 1 1"c2
squares. ..ixc3+ 12 bxc3 was fine for Black in Por
c2) 9 0-0-0 and now after 9 ...'i'e7 10 dS tisch-Polugaevsky, Linares 1 985. It seems
exdS 1 1 �xdS �xdS 1 2 cxdS .txf1 1 3 that the bishop is rather harmlessly placed on
:hx f1 .txd2+ 14 �xd2 d 6 1 5 e 4 �d7 1 6 f4 in these lines, as the attack on d6 and c7 is
l:fe1 �eS, Browne-Korchnoi, Chicago 1 982, not supported by other pieces.
I prefer White (slightly). 9 ... dS!? looks like an
obvious improvement. White has no way of
turning the tension in the centre to his ad
vantage: 10 dxcS?! �bd7! 1 1 cxb6 �cS 1 2
'i'a2 axb6 1 3 cxdS ..ixf1 1 4 .l:thxf1 .txc3 1 5
.txc3 �xdS and Black i s a pawn down but
certainly has the advantage!
6 . . . .i.a5! ?
Also possible is 6. . ..txc3+ 7 1"xc3 0-0,
when 8 dxcS is the best try for an advantage,
e.g. 8 ... bxcS 9 .tgS �c6 1 0 e3 h6 1 1 .th4 gS
1 2 ..ig3 �e4 1 3 1"c2 f5 1 4 0-0-0 1"f6 1 5
.td3 �xg3 1 6 hxg3 .tb7, and now in Seira
wan-Arnason, Manila 1 992, White played 1 7 7 lbc6?
. . .
.te2?. Instead 1 7 g4! gives White a powerful This move is refuted in the meanest pos
attack. When the kings are castled on oppo sible way. Others:
site flanks and there are lots of open files, a) 7 ... h6?! 8 .th4 gS (8 ... �c6? 9 0-0-0!
time is an important aspect of the game. The ..ixc3 10 'ii'xc3 cxd4 [1 0 ...gS 1 1 dSI and
alternatives are 8 .tgS .tb7 9 e3 d6 10 dxcS Black is clearly not developed to meet this]
bxcS 1 1 l:d1 'i'e7 1 2 .td3 �bd7 1 3 0-0 1 1 �xd4 �e4 1 2 'i'h3! and White won a
l:fd8, which appeared to secure Black equal quick victory in Seirawan-Timman, Hilver
ity in Malaniuk-Lerner, Donetsk 1 998, 8 e3 sum (3) 1 990, while 8 ... .tb7 is, of course,
.tb7 9 b4 d6 1 0 .tb2 �e4 1 1 1"c2 f5 1 2 best here, but because the bishop is now on
l:d1 �d7 1 3 .te2 a S 1 4 bxcS bxcS 1 5 0-0 h4 Black will have less effect with the
l:f7 16 .ta1 �b6 17 dxcS �cS 1 8 �d4 ... .txc3+ and ... �e4 trick than without ... h7-
'i'gS 19 .tf3 .txf3 20 �xf3 with equality in h6). After 8 ... gS 9 ..ig3 Black has played:
Seirawan-Amason, Reykjavik 1 986 (although al) 9 ... �c6 10 0-0-0 .txc3 1 1 dS! exdS 12
this set-up might be more dangerous than it cxdS �xdS 1 3 l:.xdS ..ig7 1 4 h4 and White
appeared here) and, finally, 8 g3? cxd4! 9 had ;a powerful initiative in Nielsen-Hracek,
'i'xd4 �c6 1 0 'flh.4 .tb7 1 1 .tg2 l:lc8 1 2 Germany 2000.
.td2 �aS! 1 3 .txaS bxaS, when Black had a a2) 9 ...�e4 10 e3 �c6 1 1 .td3 �xg3
lead in development and very active pieces in (1 1 ....txc3+ 1 2 bxc3 �xg3 1 3 fxg3! gives
Gurevich-Kasparov, Linares 1 991 . White some prospects on the f-file) 1 2 hxg3
7 .i.g5 g4 13 dS! gxf3 14 dxc6 fxg2 1 5 l:g1 with a
The aggressive move. After 7 e3 0-0 8 powerful initiative for White in Psakhis-
83
Gruenfeld, Israel 1 99 1 . and a draw was agreed in Bareev-Adams,

a3) 9 ... g4 10 lDd2 cxd4 1 1 tiJb5 i.xd2+ 1 2 Wijk aan Zee 2002. Note that 9 ... cxd4?!
�xd2 lDe4+ 1 3 �e1 with a clear advantage quickly leads to trouble: 10 l:.xd41 lDc6 1 1
for White in Malaniuk-Lendwai, Kesckement l:.f4! tt.Je7 1 2 :Xf6! and White had an initia
1 991 . tive in Seirawan-Adams, Bermuda 1 999.
b) 7 ... .tb7! and now White has a flood of b42) 8 ... cxd4!? 9 lDxd4 0-0 10 'iic2 i.xc3+
different opportunities: 1 1 'iixc3 d5 looked fine for Black in Soko
b1) 8 0-0-0!? i.xc31 9 W'xc3 lDe4 10 i.xd8 lov-Adams, Dortmund 1 999.
lDxc3 1 1 bxc3 �dB 12 d5 1;e7 13 e4 d6 1 4 b43) 8 ... h6?! 9 .th4 g5 (9 ... 0-0 10 l:.dl
i.d3 lDd7 1 5 l:.he1 l:lad8 with an equal end cxd4 11 :Xd4 i.xf3 12 gxf3 lDc6 1 3 :d2
ing in Gretarsson-Hjartarson, Leeuwarden lDe5 14 .te2 lDg6 1 5 .tg3 lDe8 1 6 'ii'c2
1 995. i.xc3 1 7 'ii'xc3 £5 1 8 'iid3 left White in a
b2) There is no time for 8 dxc5?!. Black strong attacking position in Sokolov-Janssen,
now develops a lot of threats quickly: Netherlands 1 999) 10 i.g3 lDe4 1 1 .td3
8 ... lDa61 9 cxb6 (9 'iic2 .txc3+ 1 0 'iixc3 i.xc3+ 12 bxc3 lDxg3 13 hxg3 d6 14 'iic2
lDxc5 1 1 tt.Jd2 a5 12 b4 axb4 1 3 axb4 l:.xa1+ lDd7 and a draw was soon agreed in Nielsen
1 4 'iix a1 tDa6 15 'iib 2 'iie7 gave Black a C.Hansen, Aars 1 995, but the position does
development advantage in Campos Moreno not feel completely safe for Black.
Adams, Cala Galdana 2001 -who cares about b44) 8 . . ..txm 9 gxf3 cxd4 1 0 exd4 lDc6
two bishops when the rook is still on h1 with 1 1 0-0-0 i.xc3 12 'ii'xc3 l:.cB 13 �b1 gave
no immediate prospects?) 9 ... lDc5 10 'ii'c2 White a storming advantage in Korchnoi
.te4 1 1 'ii'd 1 i.xc3+ 1 2 bxc3 1i'xb6 and Polugaevsky, Reykjavik 1 987.
Black has a strong initiative, Van Wely b45) 8 ... 'ii'e 7 9 .te2 d6 1 0 dxc5 dxc5 1 1
Seirawan, Wijk an Zee 1 995. 0-0-0 tiJbd7?! (1 1 ...i.xc3 1 2 'ii'xc3 is a little
b3) 8 l:.d1 .txc3+! 9 bxc3 'iie7 10 d5 d6 better for White due to 1 2 ... lDe4 1 3 'iixg7!
1 1 e3 lDbd7 12 i.e2 h6 13 .th4 e5 with a with advantage) 12 lDb5 0-0 13 tiJd6 .tc6 14
good position for Black in Gunawan e4 was critical for Black in Shabalov
Timman, Bali 2000. yemelin, Moscow 2002.
b4) 8 e3 and now: 8 0-0-0 .txc3
b41) 8 ... 0-0 is the most normal. 9 .:d1 d5
(again the most logical, but also possible is
9 ... .txc3+ 1 0 bxc3 ii'e7 1 1 i.e2 d6 1 2 0-0
lDbd7 1 3 lDd2 llfc8 14 f3 h6 15 .th4 with a
complicated struggle ahead in Seirawan
Adams, Bermuda 1 999) 1 0 i.e2 tiJbd7 1 1
0-0 cxd4 1 2 lDb5!? (although neither o f the
players could have foreseen it, this move
practically leads to a forced draw; anyway,
Black appears to be doing well) 1 2 ...h6 1 3
.th4 lDc5 1 4 'ii'c2 d31 1 5 i.xd3 tiJxd3 1 6
l:.xd3 a6 1 7 cxd5 axb5 1 8 dxe6 ii'e 8 1 9 i.xf6
fxe6 20 .te5 .te4 21 .:d7! (21 b4?? l:.xf3 22
gxf3 'ii'g6+) 21 . ..i.xc2 (21 . . .'iixd7 22 'iixe4 9 d5!!
gives White powerful compensation for the This i s the main point o f the fight against
exchange as his minor pieces are very ... lDc6 in this system, and why I suggest that
strongly placed in the centre) 22 :Xg7+ <itih8 Black should avoid ... tt.Jc6 altogether. 9 dS is
84
5 'fi b 3
available to White because the lead in devel o f mine illustrated when given thi s game as
opment over-rides a possible loss of a pawn. an exercise, 12 i.xf6! is more straightfor
9 . . . exd5? ward: 1 2 ... i.xf6 (1 2...'i'xf6 1 3 cxd7+ i.xd7
Opening up the position when behind in 14 'i'e3) 1 3 'iid5 0-0 14 cxd7 i.a6 1 5 e4 and
development tends to be a poor idea. White wins.
9 ... i.e5! 10 dxc6 i.c7! is the correct defence 1 2 . . . .i.xd7 1 3 e3!
(weaker is 1 0 ...'it'c7?, e.g. 1 1 g3! i.d6 1 2 .i.g2 With the idea of :Xd7 and i.b5+. Note
dxc6 1 3 nxd6!! 'iixd6 14 nd1 'it'c7 1 5 i.f4 that Black cannot get the king into safety as
'it'b7 1 6 tlle 5 i.d7 1 7 nxd7! lllxd7 1 8 i.xc6, he will lose instantly.
which led to a swift victory in Seirawan 1 3 . . JldB
Zarnicki, Buenos Aires 1 993). Then with 1 1 13 ... 0-0 ? 14 lllxe5 'ii'xe5 15 i.xf6 'if'xf6 1 6
e3! White maintains a lead in development nxd7 and White wins a piece, or 1 3. . .0-0-0
for as long as possible in order to create the 14 i.a6+ 'it>b8 1 5 tllxe5 and Black can only
greatest possible disharmony in the oppo resign.
nent's camp. Soppe-Debamot, Buenos Aires 1 4 J:xd7! J:xd7
1 991 continued 1 1 ...h6 1 2 i.h4 'it'e7 1 3 'it'a4 14 ... 'it>xd7 1 5 'it'a4+! 'it>e6 1 6 i.c4+ �fS 1 7
d6 14 .i.d3 (14 i.e2!? appears logical) 14 ...g5 'ii' c2+ 'it>g4 1 8 h3+ 'it>h5 1 9 g4+ and White
1 5 i.g3 tllh 5 1 6 i.e2 f5 1 7 tll d 2 tll f6 1 8 h4 wins.
g4 1 9 f3 h5 20 tllb 1 ! a6 21 tllc3 0-0 22 i.f4 1 5 .i.b5 .i.d6 1 6 l:td 1 0-0 1 7 i.xd7 'iixd7
:b8 23 i.h6 :n 24 i.g5 'it'f8 25 e4 with 1 8 .i.f4!
excellent prospects for White. Note that the And Black has no defence.
bishop is yet to be of use on c8 and, there 1 8 . . . c4
fore, the c6-pawn remains a constant source 1 8 ...tll e4 1 9 'ii'd5 'i'a4 20 i.xd6 :d8 21
of discomfort for Black. tll e5 tllxd6 22 tllc4 and White wins a piece.
1 0 cxd5 i.e5 1 1 dxc6 'iie 7 1 9 'iic 2! ltieB 20 ltig5! f5 21 'iixc4+
�hB 22 .i.xd6 ltixd6 23 'iid 5 l:tdB 24
ltie6 ! 'iic B+ 25 �b1 l:td7 26 'iix d6 ! 1 -0
1 2 cxd 7 + ?
This move wins the game but, as a pupil
85
Summary
The positions after 5 'W'b3 are less well known than other lines in the QID, and that is already
a good reason for playing it. Furthermore it can give White some good pressure. After 5 . . aS
.
and 5 . 11i'e7 Black is probably always going to be very slightly worse, while after the more ac-
. .
tive 5 cS - with the idea of 6 a3 .i.aS! keeping the bishop - Black should be able to find a
...
path to equality. This latter line is particularly worthy of experimentation.
1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 lbc3 i.b4 5 'iFb3 (D)

S ... aS Game 28
-
5 11i'e7 Game 29
... -
5 c5 (D) 6 a3 J.a5 7 i. g 5 (D) Game 30

. . . -
5 'ifb3 5 . . . c5 7 J.g5
86
I CHAPTER f/VE I
4 e3
1 d4 lLlf6 2 c4 e6 3 li:lf3 b6 4 e3
The 4 e3 line is often favoured by players
who do not like to study a lot of theory. Not
because knowledge of the line is limited but
because it has a tendency not to change.
Consequently in the books (as in this one)
Black is completely equal and the lines are
rather harmless, but some older GMs (who
might prefer playing with their kids) choose
this line over others that they feel will de
mand more study. A large number of draws
should be anticipated.
In fact the e3-line is rather harmless for
Black and I simply cannot recommend it for Black should play this for safety. After
White, despite strong players such as Yusu 5 . i.e7!? 6 dS! White is playing against the
..
pov and Gelfand having found it useful for bishop on b7. Nevertheless this is still play
their purposes occasionally. The system sim able for Black; only the main line is safer.
ply does not put Black under any pressure 6 cxd5 exd5 7 .tb5+ !
and therefore he should be able to equalise Played in order to limit Black's influence
without too great an effort. Of course, that is on e4.
if he is prepared. I just think that playing a 7 . . . c6 8 .td3 .te7! ?
more aggressive system will benefit your 8. . .i.d6 appears equally but we are follow
game generally, no matter who you are. ing the path of the great masters here.
9 0-0 0-0 1 0 b3
Game 3 1 Wlule is taking it slowly. 10 e4 dxe4 1 1
Gelfand-Karpov lllxe4 lllbd7 1 2 :et l:.e8 1 3 lllegS h6 1 4
Sanghi Nagar (7) 1995 lll e4 lllxe4 1 5 i.xe4 i.b4 was fine for Black
in Kurajica-Dizdar, Zagreb 1 993 .
1 c4 lLlf6 2 d4 e6 3 liJf3 b6 4 e3 .tb 7 5 . 1 0 . liJbd7
..
lLlc3 d5 This is where the knight belongs. After
87
10. . .lDa6 1 1 i.b2 c 5 1 2 l:.cl lDc7 1 3 dxc5 possible weak square on c3.
bxc5 14 lDa4 lDe4 (14 ... lDe6 1 5 i.xf6! would 1 6 l:fe1 'ifd6
be uncomfortable for Black), 1 5 i.xe4 was
played in Langeweg-Schneider, Plovdiv 1 983,
but 15 lDe5! is stronger, securing White an
advantage. The main threat is f2-£3, but
i.xe4 followed by liJd7 is also in the air.
1 1 i.b2
Black has equalised.

1 7 'iic 2 i.xb2
17 ... a5!? is also possible.
1 8 'iix b2 c5 1 9 l:ed 1 l:ac8 20 li)e2
l:ed8
20 ... a6 21 l:lc2 cxd4 22 lDexd4 gave White
1 1 . . . i.d6 a little something in Gelfand-Karpov, Sanghi
This move order is actually not so exact. Nagar 1 995.
As we shall see lD£3-£5 is a relevant manoeu 2 1 l:c2 cxd4 22 li)exd4 l:xc2 23 i.xc2
vre, and therefore Black should start with l:c8 24 l:c 1 li)e4 25 i.d3 l:xc 1 + 26
1 1 ...l:le8!, securing equality and ensuring that 'iix c 1 'iic 5 27 'iid 1 b5 28 li)e2 b4 29
dropping back to f8 with the bishop no 'ifa 1 'ifd6 30 'ikd4 'iic 5 31 'ika1 'iid 6 32
longer wastes time. h3 li)dc5 33 i.c2 'iif 6 34 'if e5 'iix e5 35
1 2 l:c 1 li)xe5 f6 36 li)f3 li)e6 37 li)e 1 cj;>f7 38 f3
Others: li)4c5 39 g4 a5 40 cj;>f2 .ta6 4 1 h4 h6
a) 12 'ii'c2 :e8 13 l:.fel l:.c8 14 e4 (1 4 42 h5 gxh5 43 gxh 5 cj;>e7 44 li)g3 li)g7
l:.acl c5 1 5 'ii'e2 lDe4 1 6 i.a6 i.xa6 1 7 45 i.g6 cj;>d6 46 cj;>g2 .tb5 4 7 li)c2 li)ce6
'ii'xa6 liJdf6 1 8 dxc5 l:.xc5 1 9 h3 'ii'd7 20 48 f4
'ii'd3 l:.ec8 was equal in Portisch-Hiibner, White has played well and Black is still de
Manila 1 990) 14 ... dxe4 15 lDxe4 lDxe4 1 6 fending. Now Karpov's superior technique
i.xe4 liJf6 1 7 i. £5 l:.c7 1 8 :Xes+ 'ii'xe8 1 9 pays dividends.
lDe5 c 5 20 'ii'd3 l:.e7 with equality, Po 48 . . . i.d7 49 cj;>f2 li)c7 50 li)d4 li)b5 5 1
lugaevsky-Karpov, Roqueburne 1 992. li)xb5 + ? ! i.xb5 5 2 c,f;>e 1 i.d7 5 3 i.d3
b) 12 lDh4! l:.e8 13 lD£5 i.£8 14 l:.cl was i.g4 54 i.e2 li)f5!
a little better for White in Petrosian Suddenly White has a difficult bishop
Taimanov, Zurich 1 953. endgame on his hands.
1 2 . . .:ea 1 3 li)e2 'iie 7 14 li)g3 g6 1 5 55 li)xf5+
'iie 2 i.a3! 55 i.xg4 lDxg3 56 i.£3 lD£5 and White
Apparently exchanging the 'good' bishop, has problems.
but in reality preventing any problems on the 55 . . . i.xf5 56 cj;>d2 cj;>c5 57 i.d3 i.d7 58
long diagonal after ... c6-c5, and creating a i.c2 .tea 59 i.g6 i.c6 60 i.c2 i.d7 61
88
4 e3
i.d 1 i.h3 62 i.f3 i.f5 63 �c 1 i.d7 64 rook now gets stuck on a1 - 1 1 ... �b4! 1 2
�d2 a 4 65 bxa4 i.b1 dxc4 13 bxc4 i.x£3 14 gxf3 'ii'xd4 1 5
6 5 �d3 a3! and tricks like . . .i.a4 will al- �e4 'ii'd8 1 6 l:r.dl 'fic7 1 7 �xf6+ i.xf6 1 8
ways be in the air. i.xf6 gxf6 19 i.xh7+ �g7! (1 9 ... �7 20
65 . . . i.xa4 66 i.e2 i.c6 67 i.f3 'i'e4+! is not something Black allows) 20 l:r.d4
67 i.d3 d4 68 e4 i.e8 69 i.e2 i.f7 and l:r.h8 21 l:r.g4+ �f8 22 'ii'b2 :xh7 23 'ii'xb4+
Black wins. 'ii'c5 and Black had a positional advantage in
67 . . . i.e8 68 i.d 1 d4 69 �d3 i.b5+ 70 Grigorian-Karpov, Moscow 1 976.
�d2 dxe3+ 71 �xe3 i.e8 72 �e4 �d6 b) 11 :el l? dxc4 (this does not look quite
73 �f5 right; after 1 1 ...l:lc8! White has nothing better
73 �d4 i.f7 74 i.b3 i.xh5 75 �c4 i.f7+ than 12 l:lc1 , transposing to the main line) 1 2
and the pawn ending i s winning fo r Black. bxc4 :tc8 13 :c1 l:le8 14 d5! (suddenly this
73 . . . �e7 74 i.e2 i.f7 75 i.d 1 i.xa2 76 is possible) 14... �b4 1 5 i.b1 exd5 16 a3 d4
�g6 b3 77 �xh6 b2 78 i.c2 �f7! 1 7 �xd4 �c6 1 8 �fS 'i'xd1 1 9 l:lcxd1 i.f8
Prevents all counterplay. After 78 ... bl'ii' 20 �d5 �d7 21 h3 l:lxe1+ 22 :xe1 l:r.d8 and
79 i.xb1 i.xb1 80 �g7 the position is Black kept his stuff together in Portisch-
drawn. Hjartarson, Szirak 1 987, although the line
79 i.g6 + �f81 80 i.d3 b 1 'ii' 8 1 i.xb 1 looks shaky.
i.xb 1 82 f5 �f7 0-1 1 1 . . . :tc8
Tbis is the most natural move. Black is
Game 32 improving his position before taking action.
Danner-Yu S haoteng 1 1 ...�b4 12 i.b1 :c8 1 3 �e5 �c6 14
Gyu/a 2000 �xc6 :Xc6 1 5 'ii'e2 l:le8 1 6 �d1 i.f8 1 7
,_______________... �e3 was agreed drawn in Gelfand-Kramnik,
1 d4 �f6 2 c4 e6 3 �f3 b6 4 e3 i.b 7 5 Novgorod, 1 997. lt seems to me that White
i.d3 d5 6 0-0 i.e7 7 b3 has more freedom to manoeuvre than is
As this game shows Black has no prob usual in this line.
lems in this line. 1 1 . ..'ii'd 6?1, as in Portisch-Adorjan, Szirak
7 . . . 0-0 8 �c3 c5 9 i.b2 cxd4 1 0 exd4 1 987, is probably best met with 1 2 cxd5!
�c6 exd5 13 :e1 .l:.fe8 14 �b5 'ii' f4 1 5 g3 'ii'b 8
16 �e5 and White has the advantage.
1 2 lle 1 !
The rook belongs here. After 1 2 'ii'e2 :e8
13 l:lfd1 i.f8 14 h3 g6 15 i.b1 i.h61 1 6 �d2
dxc4 17 bxc4 �xd4 1 8 'ii'e 1 e5 Black was
much better in Polgar-Hansen, Aabenraa
1 989.
1 2 . . . �b4!
With this move Black takes control of the
centre� Others seem to be inferior.
a) Now 1 2...:e8!? comes out okay here,
but I do not trust it completely: 1 3 cxd5
�xd5 14 �xd5 'ii'xd5 (14 ... exd5 1 5 i.fS :c7
1 1 :tc 1 16 'ii'e2 gave White an edge in Yusupov
The main line. The alternatives are: Renet, Kaufbeuren 1 993) 1 5 i.e4 'ii'h 5
a) 1 1 'ii'e 2?! is not so good because the (1 5 ...'ii'd7? 1 6 l:r.xc61 i.xc6 1 7 �e5 'ii'b7 1 8
89
.txh7+! was winning i n Portisch-de Finnian, .txd5 23 lbfl i.f8 24 lbe3 l:.c5 25 'it'd2
Reggio Emilia 1 989; while 1 5 ...'iid6 1 6 d5! 'i'h6 26 l:.dl i.d6 27 g3 i.e4 28 lbg4 'i'h3
led to White asswning the initiative in Yusu 29 £3 .txd3 30 'i'xd3 l:.d5 and White re
pov-Rodriguez, Novi Sad 1 990). Then 1 6 signed in Lukacs-Atalik, Budapest 1 99 1 .
lbe5 (certainly not forced. . .) 1 6. . .'iixdl 1 7 1 7 . . .'i'h5 heads fo r th e safe spot. 1 8 d5
l:.exdl lbxe5 1 8 i.xb7 l:.xcl 1 9 .txcl l:.d8 exd5 1 9 'iixd5 'iixd5 20 i.xd5 i.. f6 21 l:.c2
20 'it>fl l:.d7 21 i.c8 l:.d8 22 i.b7 l:.d7 23 l:.fd8 22 .txf6 l:.xd5 23 i.b2 l:.cd8 was equal
i.c8 l:.d8 24 .tb7 was agreed drawn in Pet in Shariyazdanov-Chemyshov, Djakovo
rosian-Peters, Lone Pine 1 978. 1 994.
b) 12 ... .tb4 1 3 a3 i..x c3 14 i.. xc3 lbe7 1 5 1 6 . . . i.f6 1 7 :d 1 'ifd6 1 8 cxd5
lbe5 looked better for White in Ambarcum 1 8 l:.e3 dxc4 1 9 bxc4 lba5! and Black's
jan-Shabalov, Seattle 2002, but I am sure that pieces are much the better placed.
Shabalov only wanted to depart from the 1 8 . . . 'ifxd5 1 9 i.c4 'ifh5 20 d5 lba5!
normal paths and somehow unbalance the Black sacrifices a pawn for the two bish
situation. ops and to open up the position in order to
1 3 i.f 1 lbe4 exploit White's uncoordinated pieces.
21 dxe6 lbxc4
21 .. .i.xc3? 22 'i'xc3 'i'g6 23 exf7+ 'it>h8
24 l:.d6 is given by Gershon as winning. Af
ter 24...l:.xc4 25 bxc4 'it'xf7 it is, presumably,
not bad at all, but still - why should Black
want this position in the first place if he does
not win the exchange?
22 exf7 + 'ilxf7 23 :xc4 Ibc4 24 bxc4
.bb2 25 'ifxb2 i.xf3 26 gxf3 'ifxf3
The number of weaknesses in White's
camp spells the end.
27 :d4 h6
Creating an escape square for the king so
Clearly the most active move. Also possi the rook can be activated.
ble is 13 ...l:.e8 14 a3 lbc6 1 5 cxd5 exd5 with 28 a4 :t6 29 'ilc2 :e6 30 'ii' d 2 l:tg6+
equality in Nogueiras-Farago, Aosta 1 990. 3 1 �f1 'i'g2+ 32 'ifi>e2 :l.e6 + 33 �d 1
14 a3 lbxc3 1 5 :xc3 lbc6 1 6 'ifc2? ! 'i'f1 + 34 'ifi>c2 :l.e2 35 :d8 + 'ifi>h7 0-1
White wants to prepare d4-d5, but this is
too slow and Black is excellently placed to Game 33
meet it. Better was 1 6 cxd5 'iixd5 1 7 i.c4 Yusupov-Beliavsky
(1 7 i.d3? lbxd4 18 i.xh7+ 'it>xh7 19 'iixd4 Austria 1998
'iixd4 20 lbxd4 i.f6 21 l:.d3 l:.fd8 was better
for Black in Van der Werf-Van den Doel, This game is an illustration of one line that
Holland 1 994) and now 1 7 ...'i'd6?! 1 8 d5 Black should try to avoid, and into which
lba5 (1 8 ... exd5 19 l:.d3! gives White the ini White players have occasionally had luck in
tiative) 1 9 l:.d3 lbxc4 20 bxc4 l:.fd8 21 lbd2? tricking their opponents.
(21 i.e5! 'iid7 [21 ...'iic 5! 22 d6! with an ad 1 d4 lbf6 2 lbf3 e6 3 e3 b6 4 i.d3 i.b 7
vantage to White] 22 i..xg7!! 'it>xg7 23 lbe5 5 0-0 c5
and White has a winning attack on the way) Already the problems begin. After 5 ... d5!
2 1 . ..exd5 (now Black holds the lead) 22 cxd5 Black can play 6 lbbd2 c5! and return to
90
4 e3
these solid positions against the Colle or, in After this it seems that Black is in dire
the event of 6 c4 ii.e7, employ a transposi straits. Imperative is 13. .. l:.e8 14 ii.b5! a6 1 5
tion to the previous game. ll:ic6 .ixc6 1 6 i.xc6 l:.a7 1 7 ii. £3, where
6 c4 i.e7 7 lLic3 cxd4? ! White is simply better (Yudasin).
7 ... d5! is still good enough for Black. 1 4 'i'h3 lLidf6
7 ... h6!? has also been tried (to avoid ii.xh7+
and thereby d4-d5, as in the following exam
ple) and now 7 ... 0-0?! 8 d5! exd5 9 cxd5 d6
10 e4 ll:ia6 1 1 l:.el ll:ic7 1 2 l:.bl l:.e8 13 a3
ll:id7 14 i.f4 l:.c8 1 5 •d2 gave White an
enormous advantage in Bukic-Ljubojevic,
Bugojno 1 978. The bishop is not well on b 7.
8 exd4 d5
Seems forced. After 8 ...d6 9 d5! White has
a better version of: 8 ... 0-0 9 d5! ll:ia6 10 .if4!
d6 1 1 dxe6 fxe6 1 2 i.g3, which gave White
an edge in Malaniuk-Tiviakov, Moscow 1 992.
9 cxd5 lLixd5 1 0 lLie5 0-0 1 1 'i'g4
Also possible is 1 1 �5 g6!? (1 1 .. .fS?! 1 2 1 5 d5! !
i.c4! secures White a strong position) 1 2 Opening the cl-file and thereby undermin
� 3 ll:ixc3! 1 3 bxc3 ll:ic6 1 4 ll:ig4! with ing e4 as a strongpoint.
menacing play. 1 5 . . . lLixc3 1 6 bxc3 i.xd5 1 7 .1.g 5 h6
1 1 lLif6
. . . 1 7 ... g6 1 8 �4! h6 demonstrates how bad
Perhaps it was better to play 1 1 ...ll:ixc3 1 2 things are. In Polgar-Christiansen, San Fran
bxc3 ll:id7 1 3 ii.h6 .if6 14 l:.adl , although cisco 1 991, White· was in the driving seat
White retains the initiative. after 19 i.xg6! fxg6 20 •xh6 ll:ih7 21
1 2 'i'h4? •xg6+ 'iith 8 22 .ixe7 •xe7 23 c4 �4 24
This move gives Black a chance to fight. cxd5 •xf2+ 25 'iith l l:.f6 26 •e4 exd5 27
After 1 2 •g3! 'iith 8 (1 2 ...•xd4 1 3 .ie3 �4 •xd5 l:.g8 28 h3 etc.
1 4 a3! is clearly favourable to White accord 1 8 i.xh6! gxh6 1 9 'i'xh6 'i'c7 20 'i'g5+
ing to L.Hansen) 13 .ig5 h6 1 4 •h3 ll:ig8 1 5 �h 8 2 1 'i'h4+ 1 -0
�5! Black was in serious trouble in Dizdar
Hansen, West Berlin 1 988.
1 2 lLibd7?!
. . .
Necessary is 1 2... ll:ie4! 1 3 �5 ll:if6 14

•h4 ll:ie4 1 5 �5 ll:if6 1 6 •h4 ll:ie4 17
�5 ll:if6 1 8 �4 ll:ie4 1 9 �3?! •xd4 20
il.f4 ll:if6 21 ll:ie2 •a4 22 l:.fcl i.a6 23 i.c2
•e8 and Black held out in Kishnev
Schlosser, Budapest 1 99 1 .
1 2. . .ll:ic6? 1 3 il.g5 g6 ( 1 3...h 6 1 4 .ixf6
ii.xf6 1 5 •e4 g6 1 6 ll:ixc6 •c7 1 7 .£3 il.g7
1 8 il.e4 f5 1 9 ll:ie7+ and wins) 1 4 i.a6! and
White was winning in Plaskett-Arkell, Lon
don 1 99 1 . The moves •g3+ and ll:ig6+ will soon win
1 3 lld 1 lLie4? the queen.
91
Summary
The variation with e2-e3 is not really theoretical, but it can still be a minefield for Black. The
right way to meet it has been presented in this chapter. If you play like Black in Danner -Yu
Shaoteng, or the sidelines, you should not end up in any trouble. The most important lesson is ·
not to play like Beliavsky against Yusupov. Many grandmasters have suffered in this line.
1 d4 lL'lf6 2 c4 e6 3 lL'lf3 b6 4 e3 i.. b 7 (D) 5 i.. d 3

S liJc3 dS 6 cxdS exdS 7 .i.bS+ (D) - Game 3 1
5 d5
. • .
5 c S 6 0-0 .i.e7 7 ltJc3 cxd4 8 exd4 d S - Game 33

. . .
6 0-0 i.. e 7 7 b3 0-0 8 lL'lc3 c5 (D) - Game 32
4 . . . i.. b 7 7 i.. b 5 + 8 . . . c5
92
CHAPTER SIX I
4 g 3 i.. b 4 + and
4 . . . i.. a 6 without 5 b3
1 d 4 l0f6 2 c 4 e6 3 l0f3 b6 4 g3 .i.b4 + 5 i.d2 a5

I n this chapter we are considering the The alternative is 5 ... i..xd2+ 6 'ii'xd2 i.. a6
lines arising after 4 ... i.. b 4+ and also those 7 liJa3!? (the knight is not so impressive out
that occur after 4 ... i.a6 when White tries here; 7 b3 transposes to 5 b3 i.b4+ 6 i.d2
any move other than 5 b3. 4 . . . i.. a6 is the i.xd2 7 'ii'xd2 and Game 47) 7 ... 0-0 8 i.g2
modem way of playing these Queen's In i.b7 9 0-0 d6 1 0 l:tfd1 liJbd7 1 1 l:tac1 'ii'e7
dian positions, and usually generates more 12 liJb5 a6 13 liJc3 ltJe4 14 'ii'd 3 £5 1 5 liJe1
lively play than 4 ... i.. b 7, which will be con liJdf6 16 d5 l:tae8 17 e3 b5 1 8 b3 ltJc5 1 9
sidered in Chapter 8. After 4 ... i.. a 6, White 'ii'c2 b 4 2 0 liJe2 e S with an unclear game in
has several replies besides 5 b3. They in Beliavsky-Miles, Tilburg 1 986. Garcia Pal
clude 5 'ii'c2 and 5 'ii'b 3, which are not ermo-Groszpeter, Zenica 1 987 went 10 liJh4
really dangerous for Black. Korchnoi has i.xg2 1 1 ltJxg2 c5 1 2 d5 exd5 1 3 cxd5 l:te8
tried to come up with some ideas after 5 1 4 £3 a6 1 5 liJc4 b5 1 6 liJce3 l:te5 with equal
'ii'c 2, but I find it hard to believe in them. 5 ity, while 1 0 liJb5 liJbd7 1 1 l:tfd1 a6 1 2 liJc3
'it'a4 has been played for a long time now. It ltJe4 13 ltJxe4 i.xe4 14 l:tac1 liJf6 1 5 'iie3
was originally made popular by Michael °iVb8 16 liJel i.xg2 17 liJxg2 'ii'b7 was equal
Rohde in the 1 9 80's, and has ever since in Yusupov-Kuzmin, Minsk 1 982. White's
been a serious alternative. 5 liJbd2 is also a space advantage is not so important because
popular alternative, but as this chapter illu s so many pieces have been exchanged.
trates, Black can always find a safe path for 6 .i.g2 0-0
equality in these lines. Here White has a very slight advantage af
ter 6 ... i.a6 7 i.xb4 axb4 8 ltJe5 l:ta7! 9 a3
Game 34 i.c8 1 0 liJd3 bxa3 1 1 l:txa3 l:txa3 12 ltJxa3
Magerramov-Makarichev 0-0 13 0-0, as in Browne-Spassky, Manila
Moscow 199 1 1 976. ,;
7 0-0 .i.a6
1 d4 l0f6 2 c4 e 6 3 l0t3 b 6 4 g3 7 ... i.b7 8 i.g5 (8 'iic2 h6 9 i.. f4 i.. e4 1 0
.i.b4+ ! ? 'it'dl liJc6 1 1 a 3 i..d6 12 i..xd6 cxd6 1 3 liJc3
This i s rather off-beat but does offer a d5 14 cxd5 exd5 15 e3 ltJe7 16 ltJe5 i.. xg2 17
chance for transposition to standard lines. �xg2 d6 1 8 liJd3 was also better for White
93
in Ivanchuk -Speelman, Roquebrune 1 992) d6-square secures him a slight pull.

8 ... i.e7 9 ll'lc3 ll'le4 10 ll'lxe4 i.xe4 1 1 i.xe7 1 5 . . .'i'bS? !
'ii'xe7 12 'ii'd2 d6 13 'iie3 i.xf3 14 i.xf3 1 5 ... 'ii'e 7! (Magarramov) 1 6 ll'ld6 l:ad8! 17
lta7 15 l:acl with advantage to White in l:cl cxd4 1 8 ll'lxd4 ll'lxe5 1 9 ll'lc4!? ll'lxc4 20 ·
Karpov-Van der Wiel, Brussels 1 987. bxc4 ll'lf6 21 ll'lc6 'ii'c7 22 ll'lxd8 :Xd8 with
8 b3 just about enough compensation, although it
is still White who is trying for a win.
1 6 ltld6 cxd4? !
Now White gets the chance to start an at
tack. Necessary was 16 ... h6.
1 7 'i'xd4 ltlc5
17 ... l:d8? 18 ll'lxf7! �fl 19 ll'lgs+ <j;e7
20 ll'lxe6! would eliminate Black completely.
1 8 ltlg 5! 'i'd8
1 8 ... h6? 1 9 ll'lgxfll l:.xfl 20 i.xd5 exd5 21
ll'lxfl <j;xfl 22 e6+1 ll'lxe6 23 'ifxd5 i.cB 24
l:acl and White wins.
1 9 'i'g4 l:a7
Black is trying very hard to defend.
8 . . . d5 1 9 ... h6? 20 ll'lgxfl l:.xfl 21 b41 and White
'lb.is is the main move here, but there are wins.
some important alternatives: 20 b4 ltld3 21 .i.e4! g6 22 ltlxe6 ltlxe5
a) 8 ...l:a7 9 i.f4 a4 (9 ... d5!? 10 c5 ll'le4 1 1 Black is running out of defensive re
a3 i.c3 1 2 cxb6 cxb6 1 3 l:a2 b5 1 4 ll'lg5! sources. After 22 ...'ii'd 7? White wins with 23
ll'lxg5 15 ll'lxc3 b4 16 axb4 axb4 17 ll'lb5 i.xd5 ll'lxel 24 l:xel i.c8 25 h3! fxe6 26
l:d7 1 8 'ifcl with an edge for White) 10 'i'c2 ll'lxc8 exd5 27 ll'lxa7.
d5 1 1 l:dl 'ifc8 12 ll'le5 c5 13 dxc5 i.xc5 1 4 23 ltlxd8 ltlxg4
ll'lc3 and now, instead of 1 4... b5?, which was
played in Avrukh-Mikhalevski, Beersheba
1 997, Black should play 14 ... axb3 1 5 axb3
dxc4, although White is better after 1 6 ll'lxc4.
b) 8 ... c6 9 'i'c2 d5 1 0 i.xb4 axb4 1 1 a3
i.b7 12 ll'lbd2 bxa3 13 b4 ll'la6 14 'i'b3 dxc4
1 5 ll'lxc4 ll'ld5 1 6 :Xa3 ll'ldxb4 1 7 ll'lfe5 °flc7
1 8 l:.bl gave White excellent compensation
for the pawn in Zimmennan-Letreguilly,
Budapest 1 994.
c) 8 ... i.b7 transposes to Games 48-50
with 4 ... i.a6 5 b3 i.b7 6 i.g2 i.b4+ 7 i.d2
a5 8 0-0 0-0.
9 cxd5 ltlxd5 24 ltl8xf7!
9 ... exd5 10 ll'le5 is a structural advantage A nice combination that in practice wins
for White. the game.
1 0 Ae 1 c5 1 1 e4 ltlf6 1 2 a3! �xd2 1 3 24 . . . ltldf6 25 �c2 :Z.axf7 26 �b3! ltlh6
ltlbxd2 ltlbd7 1 4 e5 ltld5 1 5 ltle4 27 ltlxf7 ltlxf7 28 l:e6 �g7 29 l:xb6
White has won the opening battle. The �c8 30 bxa5 ltlg5 31 f4 ltlh3+ 32 �g2
94
4 g 3 i. b 4 + a n d 4 . . . i. a 6 with o u t 5 b 3
.:es 33 1:.c 1 l:l.e2+ 34 �f 1 J:U2 + ? ! 35 squared bishop to concentrate on the light

cJi>e 1 llle4 36 .:c2 1 -0 squares. To be successful in this strategy he
needs to keep the bishop outside the pawn
Game 35 chain, meaning on the b1 -h7 diagonal. Lil
Anastasian-Brodsky loni-Gereben, San Benedetto 1 957 continued
Moscow 1992 1 0 .l:te1 c5 1 1 a3 J..xc3 12 bxc3 lt:ibd7 1 3 e3
'i'c7 14 'iia4?! (doing nothing about solving
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 g3 .i.a6 5 the most important problem -where does the
'i'c2!? queen's bishop belong?) 14 ... J..d3 1 5 .i.b2
(hardly a dream spot for the bishop) 15 ...b5
16 li'dl i.e4 17 J.. fl c4 18 lt:id2 .i.g6 1 9
.i.g2 lt:ib6 2 0 f3 lt:ia4 and Black was better.
b) 5 ...lt:ic6!? tries to exploit White's lack of
development. I am not sure Black equalises
with this line, but it does result in an interest
ing position, e.g. 6 a3?! (6 J..g2! must be the
right move - after 6 ...lt:ib4 7 'itb3 d5 8 lt:ie5
[8 cxd5 li'xd5 is already equal] 8 ... dxc4 9
li'dl ! [9 lt:ixc4 li'xd4 1 0 .i.xa8 J..xc4 1 1 'i'dl
'ii'xdl+ 12 'it>xdl J..c 5 is terrible for White]
9 ... b5 10 a4 c6 1 1 axb5 .i.xb5 1 2 lt:ic3 lt:ibd5
13 e4 lt:ixc3 14 bxc3 and White has a strong
1bis idea is not particularly popular. The centre to compensate for his pawn deficit).
main reason is probably 5 ... d5, which appears Returning to 6 a3, play might continue
to be a very easy route to equality. 6 ... i.b7 7 'ii'd3 eS! 8 dxe5 lt:ixe5 9 lt:ixe5
5 . . . c5 i.xhl 10 f3 .i.d6 1 1 'ii'c3 'i'e7 1 2 lt:ig4 lt:ixg4
ECO suggests this move but does not in 13 'ii'xg7 0-0-0 14 'ii'xg4 h5 15 'ii f5 .l:tdg8 1 6
clude 6 e4 in its world view. i.f4 i.xf4 1 7 'i'xf4 h4 1 8 g4 f5 1 9 gx f5 .i.g2
a) After 5 ... d5! we have: 20 f6 'ii'd6 21 'i'e4 c6 22 f7 .l:tf8 23 .i.xg2
al) 6 b3 .i.b4+ 7 J.. d2 i.. xd2+ 8 lt:ibxd2 °i'xh2 24 'ii'g4 h3 25 .tfl :Xf7 26 lt:id2 'ii'e5
0-0 9 i.g2 c5 promises Black equality, and 1 0 27 .i.xh3 .l:tg7 28 'ii'e4 'ii'g3+ 29 'it>d1 'ii'xh3
e 3 lt:ic6 1 1 dxc5 dxc4 12 bxc4 lt:ib4 1 3 'itbl 30 'it>c2 .l:tg2 31 'ii'd 3, and Black had great
lt:id3+ much more than that in Vacek winning chances in Korchnoi-Kachiani
Hausner, Bmo 1 969. Gersinska, Willingen 1 999.
a2) 6 lt:ibd2 .i.e7 7 J..g2 0-0 8 0-0 c5 9 6 .i.g2
dxc5 (9 .l:tdl lt:ic6 1 0 li'a4 li'c8 1 1 dxc5 dxc4 After this move Black can equalise. I am
12 lt:ie5 lt:ixe5 1 3 i.xa8 b5 14 'iia5 .l:td8 1 5 less sure about 6 e4!, although it has been
.i.g2 i.xc5 was crushing in Udovcic-Rakic, played only once - 6 ... cxd4 7 lt:ixd4 i.b7 8
Sombor 1957) 9 ... J..xc5 10 a3 lt:ibd7 1 1 li'a4 i.g2 a6 9 0-0 'ii'c7 10 .l:tel lt:ic6 1 1 lt:ixc6
i.b7 1 2 cxd5 exd5 1 3 b4 i.d6 14 J.. b 2 a5 1 5 'ii'xc6 1 2 b3 .i.e7 13 lt:id2 d6 1 4 .i.b2 0-0 1 5
b 5 .l:te8 1 6 e 3 lt:ie4 and Black had sufficient 'i'c3 appears to be very slightly better for
counterplay in Blom-Johannessen, Marianske White. After 15 ....l:tfcS 16 g4 lt:ie8 1 7 'ii'g3
Lazne 1 96 1 . i.f6 1 8 J..xf6 lt:ixf6 1 9 g5 lt:id7 20 .l:tadl 'i'c5
a2) 6 cxd5 i s the most logical move: 21 lt:ifl White is better, and in Korchnoi
6 ... exd5 7 J.. g2 J.. b4+ 8 lt:ic3 0-0 9 0-0 .l:te8! Lau, Hessen 1 999, Korchnoi more or less
and Black is ready to give up the dark- outplayed his opponent by advancing the h-
95
pawn. i.e3 i.h6 51 i.f5 i.g7 52 i.xh7 �xh7

6 . . . �e6 7 dxe5 .be5 a 0-0 0-0 9 i.g5 ! ? 53 i.xf6 :l.ga 54 �f1 1 -0
:ea 1 O 'ii'a4 i.b7 1 1 �e3 h6 1 2 i.f4!
'ii'e 7 1 3 a3 a5? Game 36
13 ... .i.d6 would have kept the position Barlov-Beliavsky
more or less equal. Now Black has problems. Yugoslavia 1992
14 llad 1 llfda 1 5 �e5
The b-file is a terrible source of weakness 1 d4 �f6 2 e4 e6 3 �f3 b6 4 g3 i.a6 5
for Black. 'ii' b 3
1 5 . . . i.aa 1 6 �d3! This line can at best be described as com
pletely harmless. After...
5 . . . �e6!
White has a clear advantage.

1 6 . . . �h5 1 7 i.e1 i.d4
17 ... lCid4 1 8 .i.xa8 l:txa8 19 .i.e3!? lCif5 20 ... White is forced to protect e4 an addi
.i.xc5 bxc5 21 lCie4, with the idea g3-g4 is tional time.
given by Dautov. White is close to winning. 6 .lbbd2 d51
1 a �b5 �f6 1 9 i.f4 �ea 20 ild2 i.f6 Black's forces are quickly finding a way
21 llfd 1 d5 22 exd5 exd5 23 lle2 �a7 into the game.
24 �xa7 'ii'x a7 25 llde 1 llxe2 26 llxe2 7 "ila4
"ile7 27 'ii' b 5 'ii'e 6 2a h4 g 5 ! ? 29 hxg5 Perhaps it is better to simply play 7 cxd5
hxg5 30 i.e 1 .:tba 3 1 �e 1 ! �d6 32 'ii'd 3 exd5 (7 . .'ii' xd5!? must also equalise) 8 'i'a4
.
.lbe4 33 �f3 �e5 34 'ii'd 1 °ilf5 35 i.e3 .i.b7 9 .i.g2 .i.d6 10 0-0 0-0, e.g. 1 1 lCibl
i.e6 36 lld2 lCie4 12 .i.f4 l:le8 13 l:tc1 .i.xf4 14 gxf4 lCie7
Entering some complicated tactics. 36 15 lCic3 lCif5 and Black had equalised in Gel
.i.xc5? .i.a4 37 e4 .i.xc2 38 'ii'xc2 'i'xe4 fand-Leko, Monaco 2002.
would turn the tables, but 36 b31?, as sug 7 . . . i.b7 a i.g2 i.d6 9 0-0
gested by Dautov, might improve. 9 lCie5 looks good, but after 9 ... 0-0! 1 0
36 . . . i.a4 37 llxd5 "ilxf3 3a i.xf3 i.xd 1 lCixc6 1i'd7 Black retains his piece wi th no
39 llxd 1 i.xb2 40 i.xg5 i.xa3 4 1 i.f6! minuses whatsoever, and complete equality.
Now White also has an attack on the king Chekhov-Bareev, Germany 1 992 continued
side to make him happy. 1 1 0-0 :fes 12 1i'c2 .i.xc6 13 b3 (1 3 :d1
4 1 . . . �e6? ! 42 lld7! �fa 43 lla7 �h7 44 :ad8 with equality, as in Grivas-Delchev,
i.e5 Ilda 45 i.h5 f6 46 i.f7+ �ha 47 Vama 1 994, is probably better) 1 3 ... e5! 1 4
i.e3 :ea 4a i.d2 i.e 1 49 i.e6 Ilda 50 .i.b2 dxc4 1 5 .i.xc6 'ii'xc6 1 6 1"xc4 1"b 7 1 7
96
4 g 3 i. b 4 + a n d 4 . . . i. a 6 with o u t 5 b 3
:acl ?! (1 7 dxeS .txeS 1 8 ilxeS :xeS is ac 'iig 4?!

tive and quite pleasant for Black) 1 7 ... bS! 1 8 33 g4! is simpler.
'ir'c2 exd4 1 9 ltJ £3 cS 20 b4 cxb4 2 1 .txd4 33 . . . 'ii'h 7 34 %:tc1 i.c8 35 ft'e2 hxg3 36
ltJe4 22 :fdl aS 23 ltJh4 Ji.f8 24 £3 ltJc3 2S hxg3 'i:ig6 37 g4?
.txc3 bxc3 26 e4 b4 27 :as :ed8 28 :cdl Allowing a strong combination. After 37
:xdS 29 exdS :as 30 'ir'd3 :xdS and White :c6, planning g3-g4, Black is busted.
resigned. 37 . . . i.xd4! 38 exd4?
9 . . . 0-0 1 0 a3 a5! ? 1 1 'ii c 2 a4 1 2 l:e 1 38 :xc8 :xc8 39 exd4 retains the better
i.e7? chances. Now Black takes over.
1bis move makes very little sense at all. 38 . . . e3 39 liJd3 i.xg4 40 'iic 2 i.f5 41
Correct is 12 ... eS 13 dxeS ltJxeS 14 ltJxeS fic6 i.xd3 42 'i'xg6 i.xg6 43 i.c3 l:.d6
Ji.xeS 1 S ltJ£3 .td6 16 ltJd4 .tcS 17 ltJf5 44 'it>h2 i.f5 45 .Ue 1 l:.h6+ 46 'it>g3
:es! with an unclear game. White is not very l:.g6 + 47 'it>h2 %:tg4 48 d5 l:.xf4- + 49
well co-ordinated to exploit the well placed 'ii.og 3 l:.g4+ 50 'ii.of3 l:.g6 51 i.h 1 i.e4+
knight. 52 'it>e2 i.xh 1 53 J:.xh 1 f5 54 i.b4 l:.g2+
1 3 e3 'iid 7 1 4 cxd 5 exd5 1 5 liJb 1 ! 55 'it>d3 J:.xb2 56 d6 e2 57 l:re 1 f4 58
'it>c3 J:rxb4 0-1
Game 37
Barbero-Cebalo
Caorle 1987
1 d4 ltJf6 2 c4 e6 3 liJf3 b6 4 g3 i.a6 5
liJbd2
1bis is one of the main systems and
should offer White a modest chance for an
advantage.
5 . . . i.b4!? 6 'ii b 3
6 'ii'c2 Ji.b7 7 .lig2 aS 8 0-0 dS 9 ltJeS 0-0
The knight belongs on c3. 10 cxdS exdS 1 1 ltJdf3 ltJe4 1 2 ii.e3 :es led
1 5 . . . ltJa5 ! ? to equality in Porcisch-Bronstein, Las Palmas
An interesting pawn sacrifice. 1 972, as did 6 'ii'a4 cS 7 a3 ii.xd2+ 8 .txd2
1 6 ltJe5 'ilfe8 1 7 fixc7 i.d8 1 8 fic2 liJb3 cxd4 9 ii.g2 .tb7 10 0-0 ii.c6! 1 1 'ii'd l .txf3
1 9 l:a2 :tc8 20 ltJc3 ltJe4 21 liJd3 i.f6 1 2 .txf3 ltJc6 13 .tf4 0-0 14 Ji.d6 :es 1 S
More accurate is 21...£5!, with the idea of ii.xc6 dxc6 1 6 'ii'xd4 c S 17 'ii'f4 ltJhS 1 8 'ii'e S
22 :a1 gS followed by ... f5-f4 etc. f6 19 'ii'xhS 'ii'xd6, Dizdar-Palac, Medulin
22 J:.d 1 h5 23 i.d2! 1 997.
White has the advantage. 6 . . . c5
23 . . . i.g5 24 i.e 1 h4 25 .Uaa 1 ! ? Best. 6 ... ltJc6 7 dS .txd2+ 8 .txd2 ltJe7
As the a4-pawn also falls, this makes good (8 ... ltJa,? 9 'ii'a4! ltJe4 10 .txaS bxaS 1 1 dxe6
sense. fxe6 12 .tg2 0-0 13 0-0 gave White a clear
25 . . . ltJxa 1 26 l:xa 1 fie6 27 'iiid 1 .Ufe8 advantage in Vaganian-Nogueiras, Montpel
28 ltJxa4 'iih 6 29 ltJc3 J:.cd8 30 f4! lier 198S) 9 .tc3 ltJf5 10 ltJd2 favours White.
Black is strategically lost. His rooks have Kasparov-Speelman, La Valetta 1 980 contin
no play. ued 10 ...ltJd6?! 1 1 £3! 0-0 1 2 e4 and White
30 . . . i.f6 3 1 ltJxe4 dxe4 32 liJf2 'i:ig6 33 was now clearly better. After 1 2.:.exdS 1 3
97
cxdS i.xfl 1 4 .l:.xfl aS 1 5 eS a4 1 6 'i'c2 'i'e8 Hracek, Slovakia 2000, or 1 0 'i'a4 i.b7 1 1
1 7 �£2 lllxdS 1 8 'i'd3 'i'e6 1 9 exd6 the lead lllb3 i.e7 12 .l:.d1 'i'c8 1 3 i.f4 lllb4 14 llleS
had grown to being technically winning. An i.xg2 1 5 @xg2 'i'b7+ 16 �g1 dS, Redze
improvement is 1 0... 'i'e7 1 1 0-0-0 exdS 1 2 pagic-Popovic, Novi Sad 1 985.
cxdS llle4 1 3 lllxe4 'i'xe4 1 4 £3 'i'e3+ 1 5 9 0-0
i.d2 'ii'xb3 1 6 axb3 llld4 1 7 b4 with a plus
for White (Kasparov).
6 ...'i'e7 7 i.g2 i.b7 8 0-0 i.xd2 9 i.xd2
0-0 10 .l:.ad1 gave White a slight edge in Al
exandria-Chiburdanidze, Thilisi 1 981 .
9 . .1:1.bS?
..
The rook is misplaced here due to a future

i.f4. Both alternatives are better.
a) 9 ....l:.c8 1 0 'i'c2 0-0 is interesting. We
don't know much about it because 1 1 lllb3
·1 i.g21 lll d4 12 ltibxd4 cxd4 13 b3 dS 14 ltixd4 eS
7 a3 i.xd2+ 8 i.xd2 lll c 6 9 dS (9 i.g2 1 5 a3 i.cS 1 6 ltJEi dxc4 17 bxc4 i.xc4 was a
cxd4 10 'i'a4 'i'c8 1 1 0-0 i.b7 12 b4 llle4 display of poor opening play that left Black
was a pawn in Nesis-Sanakoev, corr. 1 986, in front in Dlugy-Ehlvest, Mazatlan 1 988.
while 9 dxcS bxcS 10 'i'c2 looks best, but But this looks like the route that Black
White has no chance of an advantage) should pursue.
9 ... lllaS 1 0 Las bxaS 1 1 dxe6 fxe6 1 2 'i'c2 b) 9 ... 0-0 1 0 a3 i.xd2 1 1 i.xd2 dS 12 cxdS
.l:.b8 13 i.h3 'i'b6 and Black has the better exdS 1 3 l:fc1 was, potentially, slightly better
prospects, Fedorowicz-Seirawan, South Bend for White in Kempinski-Hansen, Hamburg
1 981 . 1 999.
7 . . . t'tlc6 1 0 'i'c2 0-0 1 1 t'tlb3! 'i'e7 ? !
7 ... i.b7 8 dxcS i.xcS 9 0-0 0-0 10 'i'c2 1 1 ...lLlaS 1 2 .tf4! d6 1 3 lLlfd2 gives White
i.e7 1 1 b3. d6 12 i.b2 lllbd7 was equal in the advantage but is nonetheless preferable.
OU-Andersson, Tallinn 1 998. But the simple 1 2 i.f4
8 0-0 looks enough for an edge. 1 2 a3! .taS 1 3 lllxaS ltixaS 14 'i'a4 ltib3
8 dxc5 15 'i'xa6 ltixa1 16 i.f4 .l:.b6 17 'i'xa7 :Xb2
8 a3 transposes to the note to White's 7th 1 8 l:xal has been suggested as giving White
move. an advantage.
8 bxc 5 ! ?
... 1 2 . . . e5
8 ...LcS 9 0-0 0-0 seems to equalise, e.g. Forced. After 1 2 ...d6? 1 3 .l:.fd1 l:lfd8 1 4 a3
10 a3 i.e7 1 1 'i'a4 i.b7 12 e4 'i'c7 13 'i'c2 i.aS 1 5 lLlgS White wins a piece.
lll e S! 14 b3 lllxf3+ 1 5 i.xf3 i.d6 16 i.g2 1 3 i.g5 .1:1.fcS? !
.l:.ac8 1 7 i.b2 i.eS with equality, Tibensky- Imperative is 13 ... ltid4 14 lllbxd4 cxd4 1 5
98
4 g 3 i. b 4 + a n d 4 . . . i. a 6 wi th o u t 5 b 3
'i'a4 .l:tb6 1 6 a3 i.d6 1 7 ti:Jd2! i.b8 1 8 b4 d6 mentarily weakness: 7 ...tt:Jxe4?! 8 ti:JeS! dS

1 9 bS i.c8 20 i.c6!, when White has some (8 ... ti:Jc3?? 9 libs g6 10 'iih3) 9 'ii'a4+ ti:Jd7
positional pluses. Now he is winning. 1 0 cxdS ti:Jxd2 1 1 i.xd2 i.xdS 1 2 i.xdS
1 4 a3 i.a5 1 5 ti:\xa5 ti:\xa5 1 6 ..W a4! exdS 13 0-0 f6 14 l:tfe1 fxeS 1 5 .:r.xeS+ i.e7
i.xc4 1 7 ..Wxa5 i.xe2 1 8 lUe1 i.xf3 1 9 1 6 i.gS?! bS! with a quick draw in Goldin
i.xf3 h6 20 i.xf6 ..Wxf6 2 1 ..Wc3 d 6 22 Smirin, Moscow (GMA) 1 989, but Avrukh
i.d5 c4 23 .l:.e3 .l:.b3 24 ..Wd2 J:.xe3 25 recommends 1 6 .l:tae1 ! (with the threat of
fxe3 ..Wg5 26 .l:.c1 ..Wg4 27 ..Wf2 �h8 28 i.gS), when 1 6 ...'it>f7 1 7 l:txdS! appears to be
..Wxf7 ..We2 29 ..Wf2 ..Wd3 30 e4 a5 3 1 decisive .
.l:.xc4 .l:.xc4 32 ..Wf8+ �h7 3 3 i.g8 + 1 -0 After 7 ... d6 8 dS exdS 9 cxdS i.a6 1 0 i.£1
------. 'i'c8 1 1 'i'a4+ ti:Jfd7 12 i.xa6 'i'xa6 1 3 'i'xa6
Game 38 tt:Jxa6 14 0-0 ti:Jb4 15 tt:Jc4 f5 16 exf5 ti:Jf6 1 7
Salov-Karpov i.gS ti:JbxdS 1 8 .l:tadl White was clearly bet
Wy:k aan Zee 1998 ter in Sosonko-Miles, Wijk aan Zee 1 981 .
8 0-0!
1 d4 ti:\f6 2 c4 e6 3 ti:\f3 b6 4 g3 i.a6 5 The modem treatment - Black cannot
tll b d2 i.b7 cling on to the pawn. Other options promise
5 ... cS 6 i.g2 .tb7 is just a transposition, nothing:
but 6 ... ti:Jc6! is much stronger, as can be seen a) 8 eS tt:Je4! 9 0-0 ti:Jxd2 1 0 i.xd2 i.xf3
in the next game. 1 1 'i'xf3 ti:Jc6 12 'i'xc6 dxc6 1 3 i.xc6+ 'i'd7
6 .tg2 c5 14 i.xd7+ 'it>xd7 was completely equal in
6 ... i.. e7 7 0-0 0-0 8 b3 normally leads to Salov-Karpov, Wijk aan Zee 1 993.
transposition after 8 ... dS. b) 8 ti:Jxd4 i.cS! and now:
7 e4! bl) 9 ti:J4b3 i.e7 10 0-0 'ii'c 7 1 1 .l:te1 d6
12 ti:Jd4 0-0 13 b3 a6 14 i.b2 ti:Jbd7 was
equal in Piket-Salov, Wijk An Zee 1 992.
b2) 9 ti:Jc2 'i'c7 10 0-0 i.e7 and now in
Agdestein-Hjartarson, Belgrade 1 989, White
should have maintained the balance with 1 1
b3. Instead 1 1 eS?! was too optimistic. After
1 1 ...i.xg2 12 exf6 i.xf1 13 fxe7 i.d3 14 'ii'f3
i.xc2 1 5 'i'xa8 'it>xe7 1 6 b3 l::td 8! 1 7 i.a3+
'it>e8 1 8 'i'f3 'i'eS! 1 9 l:tcl i.g6 Black had a
clear lead.
b3) 9 ti:JbS a6 10 ti:Jc3 'i'c7 1 1 'ii'e2 i.e7
with complete equality in Welin-Chemin,
Lugano 1 989.
Time has given this move the stamp of 8 . . . d6
approval. 7 dS!? exdS 8 ti:Jh4 g6! 9 cxdS i.. g7 The alternatives favour White:
1 0 tt:Jc4 0-0 1 1 ti:Jd6 (1 1 0-0 d6 1 2 i.f4 i.a6! a) § . . . tt:Jxe4?! has never been played but
13 l:tcl i.xc4 14 .l:txc4 bS is fine for Black) has been analysed by Cifuentes, among oth
1 1 ...i.a6 1 2 a4 ti:Je8 1 3 tt:Je4 d6 1 4 0-0 ti:Jd7 ers. After 9 ti:JeS dS 1 0 cxdS exdS 1 1 'ii'a4+
15 .l:tb1 ti:Jdf6 led to equality in van Wely ti:Jd7 1 2 tt:Jxe4! dxe4 1 3 i.h3 bS (1 3 ... i.c8 1 4
Psakhis, Leeuwarden 1 993. 'ii'c 6 .l:tb8 1 5 ti:Jxf7! 'i'f6 1 6 i.e6! .l:tg8 1 7
7 . . . cxd4 i.gS and White wins) 14 'ii'xbS i.c8 1 5 'i'c6
Black cannot exploit the e-pawn's mo- l:tb8 1 6 ti:Jxf7! White wins. Black's best op-
99
tion seems to be 9 ... .td6 10 &Dxf7 �xf7 1 1 .tc6 1 S lLic3! bS 1 6 fxeS and White had a
&Dxe4 .txe4 1 2 he4 &Dc6 1 3 1W£3+ <ii>g8 14 powerful initiative. Finally 1 0 ... eS!? 1 1 &DbS
.txc6 dxc6 1 S 'i'xc6 l:tc8 1 6 'iWe4, although a6 12 lLic3 i.. e7 13 lLif1 l:tc8 14 lLie3 0-0 1 S
White still has a pleasant structural advan a4 was slightly worse fo r Black in Cramling
tage, while 9 ... &Dd6 10 i..xb7 &Dxb7 1 1 'iWf3 Chiburdanidze, Groningen 1 997, although
does not work out well... Black's position remains solid.
b) 8 ... &Dc6?! 9 eS &Dg4 1 0 h3 &Dh6 11 &De4 1 0 l:.e 1 •c7
gives White a clear advantage. 10 ... lLibd7 11 eS! i..xg2 12 exf6 i..b7
c) 8 ... .tcS?I 9 eS &De4 1 0 &DgS! lLic3 1 1 (1 2 ...'iWxf6 13 lLixe6 fxe6 14 coftxg2 &Des 1 S
bxc3 .txg2 1 2 coftxg2 'iWxgS 1 3 tLl f3 is given &De4 lLixe4 1 6 l:txe4 wa s clearly better for
as clearly better for White by Ftacnik. In fact White in Browne-Burger, Philadelphia 1 990)
13 cxd4! looks even stronger... 1 3 fxg7 i..xg7 14 &De4 0-0? (1 4 ... i..xe4 1 S
9 lL!xd4 a6! ? .:.Xe4 0-0 1 6 l:tg4 gives White an edge) 1 S
&Dxd6 lLieS 1 6 lLixb7 'iWxd4 1 7 'iWxd4 lLif3+
18 �f1 lLixd4 1 9 l:td1 l:tab8 20 &Dd6 l:tfd8 21
&De4 f5 22 lLic3 and White was winning in
Tregubov-Grooten, Amsterdam 2001 .
1 1 a4!
Here there are a number o f alternatives

but it would seem to be an enormous waste
of energy trying to memorise them all. White
has a simple plan in a2-a4-a5 followed by a
knight transfer to aS. That is really all you
need to know, and that ... &Dbd7 often runs The reason why this is possible here and
into e4-eS after l:te1 , as we shall see. not in other hedgehog positions is the knight
9 ... &Dbd7 10 :et (1 0 a4 &Des is premature) on d2 and the excellent support it adds to the
1 0 ... 'i'c8 1 1 a4! a6 12 aS eS 1 3 lLif5 g6 1 4 c-pawn. White is also more flexibly placed to
&De3 .te7 1 S axb6 0-0 1 6 b 4 was poor for push with b2-b4.
Black in Oll-Kengis, Riga 1 99S. 12 ... .:.b8 has 1 1 lL!c6
. . .
been suggested as an improvement, but to 11 ... lLibd7 12 aS i.. e 7 13 &D2b3 l:td8 14

me it still looks as if White is a whole lot axb6 tLlxb6 15 tLlaS was better for White in
better. 10 ... .te7?! has also been investigated, Bonsch-Chuchelov, Berlin 1 996.
1 1 eS i..xg2 12 exf6 .txf6 1 3 lLixe6 fxe6 14 1 2 lL!xc6 .bc6 1 3 a5 l:.bS?!
�xg2 &Des 1 S &De4 giving White a clear ad This is probably because Karpov over
vantage according to Avrukh. Browne looked White's 1 Sth move. Now Black is in
Bradford, Dallas 1 996 continued 1 0 ... 'iWc7?! trouble. 13 ... i.. e 7, developing, is necessary.
11 &DbS 'iWb8 12 cS! (freeing the c4-square 1 4 axb6 •xb6
for the knight) 1 2 ... dxcS 1 3 lLic4 eS 14 f4 1 4 ... l:txb6 leaves the rook awkwardly
1 00
4 g3 i. b 4 + a n d 4 . . . i. a 6 wi th o u t 5 b 3
placed without solving the problem of the 22 c5 ilc7

a6-pawn. After 1 5 b3, with the idea J.b2-d4, 22 ... fib5 23 °it'xb5 axb5 24 J.xe5 and
Black has problems and is still not devel White wins.
oped. 23 i.xe5 dxe5 24 i.xb7 fl.xb7 25 c6
1 5 fl.e3! ltld7 ltlb6 26 ilxa6 ilxc6 27 ltlc4!
Now Black is in trouble as developing is
not possible. Note that 15 ... J.e?? 16 l:.b3
fie? 17 .l:.xb8+ "ilxb8 1 8 e5 .i.xg2 1 9 exf6
wins for White.
1 6 fl.b3 ila7
Or 16 ...fic? 17 :xa6 l:.xb3 18 itJxb3 'ifb7
1 9 l:.al ..ixe4 20 ltJa5 "ila8 21 ..ixe4 �xe4
22 J.f4 and the lead in development is again
a key factor.
1 7 fl.ba3
White exerts maximum pressure on the
primary weakness.
1 7 . . . i.b7
17 ... ltJc5 18 itJb3! ltJxb3 19 l:.xb3 :Xb3 Winning a piece and the game. The threat
20 "it'xb3 J.e7 21 "it'a2 and the a-pawn is is ltJa5.
difficult to protect without allowing 'it'a4+. 27 . . . fl.b8 28 ltlxb6 0-0
1 8 b4 "ilb6? ! 28 ... :Xb6 29 Wa8+ and White wins a
Black is wasting time. 1 8 ... J.e? has to be rook.
tried, although White has many strong op 29 ltlc4 ile4 30 fl.e3 °ild 5 3 1 ltlxe5 fl.fc8
portunities. 32 °ild3 1 -0
1 9 ila4! i.e7 20 i.b2 i.f6?
The lesser evil 20 ... e5 is tantamount to po Game 39
sitional resignation. Tregubov-Shaposhnikov
Samara 2000
1 d4 ltlf6 2 c4 e6 3 lLlf3 b6 4 g3 i.a6 5
ltlbd2 c5
21 e5! !
Fully exploiting the lead in development.
21 . . . i.xe5
21...dxe5 22 :d3! and Black will never es
cape as c4-c5 is coming. In my opinion this line is unnecessarily
101
risky for Black. and that the pawn on d4 is very difficult for
6 e4 Black to hang on to, so he should just play
The only serious option. 6 i.g2 lDc6 7 around it.
lDe5!? (White can also play safe, but he will 1 1 -*.e7
. . •
gain no advantage) 7 ... lDxd4 8 e3 lDfS 9 'il'a4 This move has been given a ?! by one
'il'c8 10 i.xa8 'il'xa8 1 1 .l:r.g1 i.c8 1 2 b3 lDe4 commentator but no improvement has been
1 3 i.b2 lDxd2 1 4 �xd2 i.e7 1 5 .l:f.adl f6 1 6 suggested. The following lines demonstrate
lDd3 i.b7 gave Black the better chances in the problems faced by Black:
Timman-Dautov, Forchheim 2000. a) 1 1 ... l:c8?! 1 2 l:e1 i.e7 1 3 b4 i.b7 1 4
6 cxd4 7 e5!
••• lDe4! and Black i s in trouble a s h e cannot
White needs to act now to gain an advan develop freely. After 1 4... 0-0 comes 1 5 g4!
tage. 7 i.g2?! is a waste of time. Remember lDh4 1 6 lDxh4 i.xh4 1 7 lDd6 .l:f.c7 1 8 lDxb7
that White has sacrificed the d4-pawn! von .l:f.xb7 1 9 b5 lDa5 20 i.xb7 lDxb7 21 'il'xd4
Hennan-Hiibner, Altenkirchen 1 999 went f6 22 l:a2 and the compensation for the ex
7 ... lDc6 8 e5 lDg4 9 0-0 :b8 1 0 .l:f.e1 i.c5 1 1 change is not apparent.
h3 lDe3, and White resigned. b) 1 1 .. .'il'c7 1 2 b4 i.b7 1 3 lDe4 h5 1 4 i.f4
7 lbg4 8 h3 lbh6
. . • a5 1 5 b5 lDd8 1 6 lDxd4 'il'xc4 1 7 lDxfS exfS
Now the knight is worse off and White 1 8 :et 'il'e6 1 9 lDg5 gave White a clear ad
can take time to develop. vantage in Sherbakov-Galliamova, Cheliab
9 -*.g2 lbc6 1 0 0-0 lbf5 insk 1 989.
10 ... i.e7 1 1 a3 0-0 1 2 b4 i.b7 13 lDe4 c) 1 1 ...b5 12 g4 lDh4 13 lLJxh4 'il'xh4 14
lDxb4? (1 3 ... lDfS!? is better, but White ap tDn 'il'd8 15 lDxd4 lDxd4 16 'il'xd4 looks
pears to be well placed) 14 lDf6+I i.xf6 1 5 good for White (Gershon).
exf6 lDc6 1 6 i.g5 'il'c7 1 7 i.xh6 gxh6 1 8 1 2 b4 -*.b7 1 3 lbe4!
'il'd2 �h8 1 9 'il'xh6 .l:r.g8 20 lLJg5! .l:f.xg5 21 This is the most logical move. Before
i.e4 .l:f.xg3+ 22 fxg3 'il'xg3+ 23 �h1 and White takes any action he improves his
Black resigned in Hellsten-Astrom, Ronneby pieces to the maximum. 1 3 g4 has proved to
1 998. be less dangerous.
1 1 a3! 1 3 d5
• • .
Here Black has a senous alternative in

1 3 ... 0-0!? 14 g4 d5 (only move) 1 5 cxd5! exd5
1 6 lDed2 (1 6 lDeg5 lDh4! is only in Black's
interest; the knight is misplaced on g5)
1 6 ... lDh4 1 7 b5 lDa5 1 8 lDxh4 i.xh4 1 9 tDn
i.e7 20 lDxd4 and White has an appealing
position, but Black has some good things
going for him, too. The c4-square is a
strongpoint for the knight and the c-file will
also come in handy. However, I believe that
the potential storm on the kingside is very
serious, and White should be optimistic.
1 4 cxd5 'ii'x d5
.
Menacing. 11 'il'a4 has also been played After this White gets time to irritate Black.
(Hellsten-Shipov, Gistrup 1 997 being one Preferable is 14 ... exd5 1 5 lDed2 a6I (directed
example) but never with much success. against b4-b5) 1 6 lDb3 .l:f.c8!, and now 1 7 g4?!
White has to realised that he is well placed lLJh4 1 8 lDxh4 i.xh4 1 9 lDxd4 lDxd4 20
1 02
4 g 3 j,, b 4 + a n d 4 . . . j,, a 6 with o u t 5 b 3
°flxd4 l:tc4 gives Black a strong llllttative. :acl a4 and Black was already doing fine.
Instead after 17 .ib2 1 7...0-0 1 8 g4 tLlh4 1 9 8 b3 c5 9 dxc5 i.xc5 10 i.b2 tiJbd7 1 1
lLixh4 .ixh4 White has only a slight edge. l:tcl i.b7 1 2 lLid4 l:tc8 1 3 tLlb5 °fle7 led to
1 5 l:.e 1 d3?! equality in Piket-Karpov, Wijk aan Zee 1 998.
Black is in trouble and now self-destructs. 8 . . . j,, b 7!
After 1 5 ... 0-0 16 g4! tLle3 (1 6 ... tiJh6 17 .ixh6 1bis move, followed by ... a 7-a5 and
gxh6 18 °fld2 looks very dangerous for ...tLla6, is the main option for a good reason.
Black) 17 fxe3 "flxe4 18 exd4 "fldS 19 i.e3 With the knight on d2 White has to play b2-
White is slightly better. b3 and i.b2 to develop his dark-squared
1 6 j,, b 2 0-0-0? bishop, but cannot meet ... a5-a4 very well.
However, there is another way to go, namely
8 ... c5!?, e.g. 9 cxd5 exd5 1 0 b3 l:te8
(10 ... .ib7! followed by ...tLla6 is the right
path; Black does not play the opening alto
gether well here) 1 1 .ib2 cxd4?! 1 2 tiJdf3!
(standard in these lines) 12 ... .ic5 13 a3 lLibd7
14 lLid3! and White has a very clear advan
tage, Polugaevsky-Rodriguez, Sochi 1 988. 1 0
dxc5!? i s interesting. Black should probably
not play 10 ... bxcS, when White was better in
Romanishin-Psakhis, Moscow 1 983 after 1 1
b3 .ib7 1 2 .ib2 aS 1 3 e4! d4 1 4 tLlec4 lLifd7
1 5 l:.el tiJb6 1 6 lLixb6 °flxb6 1 7 tLlc4. In
This seems completely ridiculous. stead 10 ... .ixcS! 1 1 lLid3 (heading for f4)
1 7 g4 tt'ih6 1 8 tt'itd2 'iWb5?? 1 1 ...l:.e8 1 2 lLif3 ll'ic6 secures equality. Earlier
Losing. After the necessary 18 ... °fld7 19 9 tLlb3 is possible, e.g. 9 ... .ib7 1 0 dxc5 bxc5
tLlc4 'ifi>b8 20 lLied6 Black is still in trouble. 1 1 cxdS i.xdS 12 .if4 "flb6 13 "flc2 i.xg2
1 9 tt'ic3! 1 -0 1 4 �xg2 tLlc6 1 5 lLixc6 "flxc6+ 1 6 'iti>gl tiJd5,
White wins material. which was perhaps slightly preferable for
------- White in Jusupov-Polugaevsky, Toluca 1 982.
Game 40 9 cxd5!
Lautier-Karpov
Monte Carlo (rapid) 2000
1 d4 tt'if6 2 c4 e6 3 tt'if3 b6 4 g3 j,, a 6 5

tt'ibd2 d5
1bis is the main reason why 5 lLibd2 is
not such a popular choice.
6 j,,g 2 j,, e 7 7 0-0 0-0 8 lt'le5
Here White could transpose with 8 cxd5!
exd5 9 tLle5. Other options causes no prob
lems for Black. With the move order in the
game Black has some additional options.
In Aagaard-Gavrikov, Gothenburg 1 998 I
discovered that the queen is not well placed Clearly the most dangerous option. Alter
after 8 "flc2: 8 ... .ib7 9 b3 a5 1 0 i.b2 tLla6 1 1 natives:
1 03
a) 9 'ii'a4. What the queen i s supposed to 1 1 -.c2

do out here is not altogether clear. Kempin Again I do not like this move at all. White
ski-Nisipeanu, Krynica 1 998 continued should concentrate on developing his minor
9 ... .id6 10 cxd5 .ixd5 1 1 e4 .ib7 1 2 tDec4 soldiers first before sending the queen into
.ie7 with a good opening for Black, although battle. t t .ib2 l:iJa6 t 2 l:.ct c5 1 3 l:iJd3! (an
it was about to get even better: 1 3 e5?! .ixg2 other standard manoeuvre in such positions)
14 exf6 .ixf6 1 5 'it>xg2 'ii'ds+ 1 6 tlJ£3 b5 and t 3 ...:e8 t4 l:iJf4 was played in Khalifman
Black had a clear advantage. Leko, New Delhi 2000. Such situations are
b) 9 b3 c5! 10 dxc5 bxc5 1 1 i.b2 t2Jbd7 what the main lines of the Queen's Indian
12 t2Jxd7 'ii'xd7 with equality, 011-Khalifman, Defence are all about. White might have a
Pamu 1 996. After 1 0 .ib2 cxd4! a difference small plus and can manoeuvre around, trying
from positions where cxd5 exd5 has been to create something more, while Black will
played is that White cannot delay the recap try to equalise or use any chance that might
ture - 1 1 .ixd4 t2Jc6 was equal in Uhlmann be given him to get some kind of an initia
Stahlberg, Prague 1954. Note that 10 cxd5 tive.
does not transpose to the main lines because 1 1 . . .ltia6 1 2 i.b2 c5 1 3 dxc5 bxc5 1 4
after 1 0 ... .ixd51 Black has already equalised. .:tad 1 ? !
c) 9 'ii'c2 c5 10 cxd5 exd5 1 1 dxc5 .ixc5 This move is rather careless. The only
12 t2Jdf3 t2Jbd7 13 .if4 :e8 was also level in black piece that is not really well placed is the
Sutter-Kindermann, Bem 1 995. queen's knight, and now Black is allowed to
9 . . . exd 5 bring it back into the game. 14 a3 'i'b6 1 5
The interesting 9 ... .txd5!? 1 0 e4 .ib7 1 1 :adt a4 1 6 bxa4 l:.fc8 t 7 l:[bt 'i'a7 results in
'ii'c2 (1 1 'ii'a4 'ii'e8 1 2 'i'c2 tDc61, with an a kind of dynamic equilibrium.
even game, is better) 1 1 ...t2Ja6 1 2 l:.dl 'i'e8? 1 4 . . . a4!
was seen in Dizdar-Granda Zuniga, New
York 1 997. After 12 ...'i'xd4! 1 3 tDdc4 'i'c5
14 .ie3 'i'b5 it is clear that White has genu
ine compensation for the pawn. A GM once
told me that the most important thing he had
learned from Fritz was that you should take
pawns when it seems dangerous but you
cannot actually see any real danger. If you
look through Tal's old games you will find
that he bluffed half his way to the World
Championship title.
1 0 b3 a5
Here Black can also try 10 ... c5 t t .ib2
t2Jc6!? t2 tDxc6 .ixc6 t3 :et .ib7 with Black is activating all his pieces this way.
counterplay according to Dolmatov. Instead 1 5 bxa4 ltib4 1 6 -.b 1 -.c7 1 7 e4
t t ...t2Jbd7 t2 t2Jxd7 'i'xd7 t 3 dxc5 bxc5 t 4 Black is about to advance his c-pawn and
e4 seemed slightly better for White in Topa White uses the moment to claim squares for
lov-Gelfand, Monte Carlo t 999, while his knights. t 7 a3 t:fJc6 t 8 tDxc6 .ixc6 also
t 1 ...tDa6 t 2 :et t:fJc7 1 3 dxc5 .ixc5 t4 l:iJd3 gives Black the advantage .
.id6 t 5 t:fJ£3 tDe4 t 6 l:iJd4 was a good deal 1 7 . . . d4 1 8 a3 ltic6 1 9 ltiec4 .Z:.xa4 20
better for White in Goldin-Ulibin, Kazan .Z:.fe 1 ? !
t 995. Inviting Black to exploit hi s more active
1 04
4 g 3 j_ b 4 + and 4 . . . j,, a 6 with o u t 5 b 3
pieces. It was better to try to disturb Black games.

with 20 'it'c2 l:.fa8 21 f4!?, when White's 6 ltic3 b5!?
forces are not easily kept out. The point. Passive play with 6 ... .i.e7 7
20 . . . ltia5! i.g2 0-0 8 0-0 d5 9 �e5 1i'c8 1 0 cxdS cxd5
Now this is possible. After this the de 1 1 i.f4 favoured White in Hubner-Bauer,
fence of a3 is more difficult for White. Leon 2001 .
21 ltixa5 7 cxb5 cxb5 8 ltixb5 "ii'b 6 9 tbc3 i.. b4
21 'it'c2 �xc4 22 �xc4 l:.fa8, followed by Black also fails to equalise after 9 ...�c6 1 0
...�d7-b6 with advantage. i.g2 .i.b4, e.g. 1 1 0-0 .i.xc3 1 2 bxc3 .i.xe2
21 . . . :1.xa5 13 l:.e1 "ii'a6 14 'it'xa6 i.xa6 15 �e5 l:.c8 1 6
White has no counterplay and Black has .i.xc6! dxc6 17 .i.a3, when White had a last
the c-pawn as a strong force and the a-pawn ing advantage in Nikolic-Ljubojevic, Monte
to put pressure on. Carlo 1995.
22 "ii' c 2 "ii'c 6 23 ltic4 :l.a7 24 f4 ltid7 25 1 0 j_g2 0-0 1 1 "ii'd 1 !
"ii'd 2 ltib6 26 :l.c 1 ltixc4 27 .l:.xc4 :I.dB 28
J:.ec 1 "ii' b 6 29 a4 j.a6 30 a5 "ii' b 3 3 1
J:txc5 i.. x c5 32 J:.xc5 :l.b7 3 3 i.. xd4 "ii'a4
34 �f2 :l.b4 0-1
Game 4 1
Sakaev-Berzinsh
Duisburg 1992
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.. a 6 5
"ii'a4
1bis line can easily be compared with 5
�bd2. It is by no means dangerous for Black
from a theoretical point of view but there is 1bis is the dangerous move for Black. Af
still plenty of play in the position, allowing ter 1 1 0-0 i.xc3 12 bxc3 i.xe2 1 3 l:.e1
for innovations and improvements. .i.bS!? 14 "ii'b 3 dS (Black must prevent c3-c4:
5 . . . c6 White stood better in Skembris-Timman,
Corfu 1 993 after 1 4... .i.c6?! 1 5 c4! �a6 1 6
i.f4!) 1 5 �e5 l:.c8 16 i.e3 �bd7 1 7 �4
�£8 1 8 a4 .i.a6 19 'it'xb6 axb6 with equality
in Vukic-Bronstein, Sarajevo 1 971 . Mean
while in Tukrnakov-Gulko, Leningrad 1 977,
11 "ii'c2 �c6 12 0-0 :fc8 1 3 a3 .i.xc3 14
bxc3 �dS 15 :d1 �ce7 16 .i.b2 l:.ab8 1 7
l:.abl "ii'b3 gave Black enormous compensa
tion for,,the pawn.
1 1 . . .:ca
1 1 ....i.xc3+ 12 bxc3 :cs 13 .i.d2 dS 1 4
0-0 �e4 1 5 l:.c1 leaves Black with compen
sation for the pawn, but perhaps not enough.
1bis line is somewhat dubious. The supe 1 2 i..d 2 d5 1 3 0-0 i.. xc3 1 4 i.. x c3 ltie4
rior 5 ... c5 is considered in the following 1 5 :l.c 1 ltid7 1 6 :l.e 1 !
1 05
White is taking his time to untangle. After 5e be ©ae 57 ©f5 + �ee 58 ©d4+ '3iie 7
16 ttld2 ttlxc3 1 7 bxc3 l:.c7 Black will exert 59 ©b5 ©db8 eo .tb7 ©e5 e1 .tde+
pressure on c3, and 1 8 c4 'i'xd4 19 cxdS 1 -0
l:txc1 20 'ii'x c1 :lc8! 21 'i'd1 exdS is level.
1 e . . . :tab8 Game 42
16 ... l:.c7 17 'i'a4! and Black will find it Karpov-Pol gar
hard to improve his situation due to the Buenos Aires 200 1
threat of �aS.
1 7 'i'a4 .tb5 1 8 'i'a3 'i'b7 1 9 .b5! :txe 1 1 d4 ©te 2 e4 ee 3 ©f3 be 4 g3 .tae 5
20 :txe 1 'i'a4 .tb7 e .tg2 e5 7 dxe5
White has more or less untangled.
20 . . .:te8
After 20 ... �xe2 21 l:tc7! °WWbS 22 'i'b3!!
'i'xb3 23 axb3 ttlf8 24 b4 a6 25 ttleS White
loses his pawn but, suddenly, his pieces to
tally dominate Black and he can hope for a
successful result.
21 :txe8+ 'i'xe8 22 .tf1 'i'e 1 23 <;Pg2
'i'd 1 24 'i'b3 .txe2 25 'i'xd 1 .txd 1
This is practically the only line played

these days. After 7 0-0 cxd4 8 ttlxd4 ifc8 9
ttlc3 .txg2 10 �xg2 'Wib7+ Black has an
improved version of the hedgehog system, as
with little space each exchange is normally an
advantage. Nikolic-Karpov, Monte Carlo
2000 continued 1 1 f3 �e7 1 2 l:.d1 a6 13
.tgS h6 14 �xf6 �xf6 1 5 ttle4 .te7 1 6 'i'b3
ifc7 17 l:tac1 0-0 1 8 f4 ttlc6 1 9 ttlf3 l:tfd8 20
Again Black has won back his pawn, but 'i'e3 l:tac8 21 g4 dS and Black was okay.
at a price. White has the two bishops and an 7 . . . bxe5!?
extra pawn on the queenside, so the ending More usual is 7 ... �xcS, as can be seen in
is, in fact, dreadful for Black. the two following games. After the text Black
2e .tb5 ©be 27 b3 .te2 28 ©e5 . ©de 29 is rock solid but has few chances of an ad
.tae .tb1 30 ©ee .txa2 3 1 ©xa7 ©d7 vantage.
32 b4 e5 33 .te7 ©e4 34 ©ee exd4 35 8 0-0 .te7 9 ©e3 0-0 1 0 .tf4 'i'be 1 1
©xd4 .tb1 3e .tb7 ©dbe 37 .tee 'iW8 :tfd 1 :td8
38 f3 .td3 39 �f2 �e7 40 .tb5 ©d7 4 1 After 1 1 ...d6 12 l:.d2 Black has an extra
.tee ©ebe 42 �e3 .t b 1 4 3 © b 5 .ta2 44 option in 1 2 ... ttlc6!? 1 3 .txd6 �xd6 14 l:.xd6
�d4 �ee 45 ©e3 .tb3 4e ©e2 g 5 47 'i'xb2 1 5 l:tb1 'ii'xc3 1 6 l:txb7 ttld4 1 7 'i'd1 ,
�e3 .te4 48 ©d4+ �e7 49 b5 ©a4+ 50 as in Karpov-Polugaevsky, Biel 1 990. Now
�b4 ©ae5 51 ©t5+ <;Pee 52 ©d4+ �e7 17 ... l:.ad8! would have kept the balance. No
53 �a5 te 54 ©t5 + �ee 55 ©d4+ �e7 tice it is important to use the appropriate
1 06
4 g 3 j,, b 4 + a n d 4 . . . j,, a 6 with o u t 5 b 3
rook, as 1 7 ... l:f.fd8 gives White the additional Seems too optimistic; perhaps Black failed
option of 1 8 l:f.xa7!?, with the main point to appreciate the implications of his oppo
being 1 8 ... 'i*'xc4?? 1 9 ll:ie5! etc. What worries nent's reply.
me most here is 1 2 'iWb3! l:f.d8 with transposi 21 e5! d5 22 cxd5 exd5 23 tl\xd5 tl\xd5
tion to the main game. 24 l:xd5 .:1.xa2 25 l:d7 !
1 2 'i'b3! Now th e white pieces are active, which
cannot be said of Black's. In fact Black is in
trouble.
25 . . ..:l.aa8 26 tl\g4 .:1.dS 27 tl\e3 l:l.ac8 28
tl\f5 .:l.xd7 29 l:xd7 j,,d 8 30 l:l.b7 j,, c 7 3 1
h 5 l:l.d8 3 2 tl\e3 l:l.d3 3 3 e 6 fxe6 34
j,, x c7 l:l.d7 35 1:Xb6 l:l.xc7 36 l:xe6 l:l.b7
37 I:.e5 l:l.xb3 38 l:.xc5 tl\f6 39 g4 .:l.a3
40 l:l.cS + 'it>f7 41 tl\c4 I:.a7 42 f3 'it>e6 43
J:tc5 tl\d5 44 .:!.c6+ 'it>e7 45 tl\e5 tl\f4+
46 'it>g3 I:.a4 47 lfo2 'it>f6 48 tiJd7 + <3ie7
49 tl\c5 l:l.d4 50 l:l.a2 'it>f7 51 l:a 7 + 'it>f8
52 tl\e4 tl\e6 53 I:.a6 <3ie7 54 <3if2 I:.c4
55 'it>e3 l:l.c 1 56 l:l.a7 + l:l.c7 57 l:l.a4 l:l.b7
12 l:f.d2 d5 13 cxd5 exd5 14 'i*'c2 ttJ a6 1 5 58 l:l.a3 l:l.c7 59 tl\g3 <3if6 60 tl\e4+ 'it>e7
a3 h 6 1 6 i.e5 d4 1 7 ll:ia4 'it'b5 1 8 b4? cxb4 61 tl\c3 l:.d7 62 f4 tl\c5 63 tl\e2 <3if7 64
1 9 ll:ixd4 l:f.xd4! and Black was winning in tl\d4 tl\e6 65 tl\f3 l:.b7 66 tl\e5 + <3if6 67
Olafsson-Bareev, Hastings 1 990. Instead 14 tiJd3 <3ie7 68 'it>e4 I:.d7 69 <3if5 .l:.d6 70
ll:ie5 ll:ia6 1 5 l:f.ad1 'it'e6 1 6 ll:ic4 ll:ib4 1 7 a3 l:a7 + 'it>f8 71 tl\e5 tiJd4+ 7 2 'it>e4 tl\e6
i.c6 1 8 'it'a5 d4 1 9 axb4 i.xg2 20 �xg2 73 tl\g6 + 'it>gS 74 <3ie5 I:.b6 75 .:l.e7 1 -0
cxb4 was equal in Tukmakov-Aseev, USSR
1 989, as was 14 e3 ll:ia6 15 l:f.ad1 h6 16 h4 Game 43
l:f.ac8 in Sturua-Arkhipov, Moscow 1 992. Borges Mateos-Spraggett
1 2 . . . d6 Cienfuegos 1997
12 ...'it'xb3 is to premature in view of 1 3
axb3 d 6 1 4 ll:ib5!, forcing concessions. 1 d4 tl\f6 2 c4 e6 3 tl\f3 b6 4 g3 j,, a 6 5
1 3 l:d2 'i'a4 j,, b 7 6 j,, g 2 c5 7 dxc5 j,, x c5 8 0-0
White is improving his position. 13 'i*'xb6 0-0 9 tl\c3 j,,e 7!
axb6 14 ll:ib5 tLle8 15 tLlg5 i.xg2 1 6 �xg2 The bishop is not well placed on c5, a
tLlc6 gave White nothing in Mirzoev-Dautov, square that is more suited for the knight.
Istanbul 2000, but 14 l:f.d2! improves, with a 1 0 l:.d 1 a6! ?
chance to gradually develop an initiative. This has become fashionable over the last
1 3 . . . tl\a6 couple of years. Black is trying to establish a
13 ... 'it'xb3 14 axb3 ll:ic6 15 l:f.ad1 tLle8, as hedgehog position. 10 ...tLla6 - the main line
in Olafsson-Lautier, Antwerp 1 998, is possi is considered in the following game.
bly enough for equality. But one tends to ask 1 1 j_f4 d6 1 2 l:.d2
what the rook is really doing on d1 besides With this move White is trying to put d6
attacking the d6-pawn. under pressure but, as we shall see, Black can
1 4 'i'xb6 axb6 1 5 l:ad 1 tl\e8 1 6 h4 h6 easily protect it. The alternative is 1 2 l:f.ac1
1 7 tiJh2 j,, x g2 1 8 'it>xg2 tl\ac7 1 9 e4 'it>f8 'it'c7:
20 b3 .:l.dbS a) 1 3 �c2 l:f.d8! (1 3 ... ll:ibd7 1 4 'it'd2 tLle8
107
is a standard Hedgehog but Black wants the Zaltsman-Amason, New York 1 986, met
rook on the other side of the knight) 14 'i'd2 with 17 ... ll:iaS!, a brilliant manoeuvre, albeit
ll:ie8 1 5 i.gS i.xgS 16 'i'xgS ll:id7 1 7 b3 quite simple. There followed 1 8 b3 ll:ib7 1 9
:ac8 18 'i'e3 ll:ief6 with equality, Jakobsen f3 ..ti>fB! (guarding the minor pieces and ad
Hansen, Greve 2002. dressing the king's own security), and now
b) 13 'i'b4 (more dangerous than 13 'i'c2) White should have played 20 'i'a3 with a
1 3 ... ll:ie8 14 i.gS (14 ll:ia4 ll:id7 1 5 'i'd2 ll:ic5 sensible position. Instead after 20 i.g3?1
16 ll:ic3 ll:ie4 1 7 ll:ixe4 i.xe4 was fine for 'i'c8! Black assumed the initiative: 21 i.£2 bS
Black in Hansen-Onischuk, Bundesliga 1 995) 22 cxbS axbS 23 'i'xbS dS 24 ll:ia4 ltaS 25
14 ... ll:if6 15 i.xf6 gxf6 16 li:idS!? (forcing a 'i'b6 dxe4 26 :xd8 exf3+ 27 exf3 i.xd8 28
draw, but it seems that White has no advan :xd8 'i'xd8 29 'i'xb7 :ds and Black had a
tage anyway) 1 6 ... exdS 1 7 cxdS 'i'd8 1 8 ll:id4 clear advantage.
..ti>h8 1 9 ll:ic6 'i'c7 20 li:id4 'i'd8 21 ll:ic6 1 7 . . . f6!
'i'c7 22 ll:id4 and a draw was ag-reed in lvan
chuk-Lautier, Moscow 200 1 .
1 2 . . . 'ii'c 7 1 3 :ad 1 :ds 1 4 ltig5
Others:
a) 14 ll:ie 1 i.xg2 15 ll:ixg2 ll:ic6 16 i.gS
ll:iaS 17 ll:ie3 h6 1 8 i.xf6 i.xf6 19 ll:ie4 i.e7
20 :c2 'i'b7 with equality in Bonsch
Stempin, Polanica Zdroj 1 987.
b) 14 'i'c2 h6. Black addresses the possi
bility of ll:ig5-e4 and i.gS. This is not strictly
necessary, but quite a common move when
White has no chances of a kingside pawn
storm. Grivas-Ionescu, Dubai 1 986 contin
ued 1 5 e4 ll:ihS 1 6 i.e3 ll:id7 1 7 b3 :ac8 1 8 This is not really necessary. Also possible
a 4 ll:ihf6 1 9 i.£4 ll:ieS 2 0 ll:ixeS dxeS 21 is 17 ... i.xgS!? 1 8 ll:ixgS h6 1 9 ll:if3 ll:iaS 20
:xd8+ i.xd8 22 i.e3 i.e7 23 f3 ll:id7 and :d4 (20 b3? bS! is in Black's interest)
Black was okay. Normally Black should be 20 ... :dc8 with sufficient counterplay.
careful when allowing White a pawn majority 1 8 .1e3!
on the queenside, but because the pawn is on Forced. The bishop is misplaced after 1 8
a4 there is no problem here. i.f4? ll:iaS! 1 9 b 3 f5 2 0 ll:igS bS! 2 1 'i'a3
c) 14 'i'b3 ll:ie8 15 e4 ll:id7 16 eS ll:ixeS 17 i.xgS 22 i.xgS l:f.db8, when Black has a
ll:ixeS dxeS 1 8 .:txd8 :xd8 1 9 i.xeS 'i'xeS 20 powerful initiative.
i.xb7 aS with equality, Meister-Wegner, 1 8 . . . ltie5 1 9 :d4 Ades 20 ltid2 f5 21 f3
Germany 1 987. .1f6
1 4 . . . .1xg2 1 5 <it>xg2 ltic6 1 6 ltige4 ltie8! Black is more than comfortable here.
Black should be careful and remember 22 .1f2 ltic6 23 :d3 ltia5 24 e4 g6 25
this lesson. White has a knight on c3 and .1d4 .i.xd4 26 :xd4 ltic6 27 :d3 ltie5
cannot find any good use for it, so an ex 28 :d4 ltif6 29 'ii' a 3 :ds 30 exf5 gxf5
change on e4 would be silly. Indeed 31 ltie2 ltic6 32 l:.h4 ltie5 33 'ii' e 3 :ea
1 6 ... ll:ixe4?! 1 7 ll:ixe4 eS leaves the dS-square 34 l:th6?
susceptible to future attention. What the rook is supposed to be doing up
1 7 .i.g5 here is not clear. After 34 'i'gS+ 'i'g7 35
More sensible than 17 g4!? which, in 'i'xg7+ @xg7 36 .:td4 the endgame is level.
1 08
4 g 3 i. b 4 + a n d 4 . . . .t a 6 with o u t 5 b 3
34 . . .-.g7 ! 35 l:h4 b5 36 b3 l:ad8 37 In Bareev-Karpov, Tilburg 1 99 1 , 1 4 h3?!

tllt4? tllg 6! 38 tll x g6 hxg6 39 l:d4 Ad8 15 g4? d6! left Black in the lead as White
39 'ii'd3 g5 40 l:.d4 d5! is also clearly bet had achieved nothing but weakening his
ter for Black because 41 cxd5? e5! is decisive. kingside. Karpov offers 1 5 tt:'ld4 with equal
39 . . . e5 40 l:d3 'ii' b 7? ity.
The traditional 40th move blunder... Black 1 4 . . .tll x c3 1 5 •xc3 a6 1 6 'ii'b 3 .txg2 1 7
can win with 40 ... d5! 41 cxd5 e4! 42 fxe4 fxe4 �xg2 'ii'b 7+ 1 8 'ii'f3 I:ta7 !
and White will suffer material losses.
41 'iitg 1 �f7 42 l:e 1 d5 43 cxd5 tll x d5
44 'ii'c 5 tllf 6 45 l:xd8 J::.x d8 46 •ea -.d5
4 7 l:e2 'ifd4 48 •xd4 l:xd4 49 lllf 1 'iite 6
50 l:c2 �d7 51 Wf2 tll d 5 52 l:d2 l:xd 2+
52 ...tt:'lc3 53 lk2 l:td3 54 tt:'le3 �d6 main
tains the pressure, but now we have a draw.
53 tllx d2 Yz -Yz
Game 44
Piket-Chuchelov
Netherlands 2000
1 d4 tllf 6 2 c4 e6 3 lllf3 b6 4 g3 .ta6 5 Stronger than 1 8 ...'ii'xf3+ 1 9 �f.3 l:tac8
•a4 .tb7 6 .tg2 c5 7 dxc5 i.xc5 8 0-0 20 e4, when White has more freedom to
0-0 9 tllc 3 .te7 1 0 l::. d 1 tll a 6 1 1 .tf4 manoeuvre thanks to his space advantage.
The most normal move, but dangerous is 1 9 ii'xb7 l:xb7 20 f3 l:c8
1 1 i.e3!?, with the idea of putting the bishop 20 ... tt:'lh51? 21 .i.e3 £5!? was an interesting
on d4. This has led to problems for Black but alternative.
seems to have been solved by 1 1 ...°iib BI, with 21 e4 d6
the idea of overprotecting the d6-square. 21...tt:'leB!? 22 h4 h5! 23 a4 i.f6 24 i.e3
Georgiev-lvanchuk, Tilburg 1 993 continued was slightly better for White in Piket-Karpov,
12 Aac1 llic5 1 3 'ji'c2 llice4 1 4 .if4, trans Monaco 1 999. However, as in the main
posing to the main game, but 14 ...'ii'c B 1 5 game, White failed to make anything of the
llixe4 i.xe4 1 6 'ii'a4 'ifb7! (the obvious modest lead, so perhaps Black should not
move) 17 i.ft Afe8 1 8 llie1 a6 was level. fear it so much. Nevertheless I feel that
Here 14 i.d4 .ic5 looks fine for Black. 1 0 ... d6 is the safest option for Black.
1 1 . . . -.c8 1 2 l:ac 1 tll c 5 1 3 •c2 tll ce4 1 4 22 tlle 2 tlle8 23 b3 h5!? 24 h4 g6 25
tlld4 l::.d 2 I:tbc7 26 l:d3 l:c5 27 a4 l:5c6 28
14 llixe4 i.xe4 1 5 'ii'a4 (both 15 'ii'd2 l:cd 1 I:tc5 29 �f1 .tf8 30 tll c 3 l:b8! 3 1
'fib7 1 6 i.d6 i.xd6 17 'ii'xd6 Aac8, Dautov I:t 1 d2 l:b7 ! 32 �e2 .te7 33 �d 1 �f8 34
Smirin, Daugavpils 1 989, and 1 5 �3 'iib 7 l:c2 l:c8 35 tlle 2 l:bc7 36 l:a2 l:b7 37
16 i.d6 i.xd6 17 Axd6 Aac8, Dlugy .th6+ .*08 38 .te3 .tt6 39 .tf4 .te7 40
Adorjan, New York 1 984, give Black a good ltic3 .tf6! 41 tlle 2 .te7 42 l:c2 l:bc7 43
game) 1 5 ... �7! 1 6 .id6? (this exchange tll c3 .tf6 44 tll b 1 b5?! 45 axb5 axb5 46
favours Black; 1 6 i.f1 is seen in the note tll a3 bxc4 47 l:xc4 .te7 48 .1:1.xc7 .l:l.xc7
above) 1 6 ... .ixd6 1 7 l:.xd6 Afc8! and Black 49 tll c4 l:c6 50 �c2 f5! 5 1 l:c3 fxe4 52
was better in Blees-Gershon, Tel Aviv 1 999. fxe4 lllf 6 53 tlld 2 l:xc3+ 54 �xc3 Wf7
Already the threat is ... b6-b5. 55 �d3 Yz-Yz
1 09
Su mmary
The 4 ... i.b4+ line invariably leads to a small advantage for White. 4 ... i.a6 is more modem and
more dynamic. After moves like 5 'ii'c2, 5 c!Dbd2, 5 'ii'b3 and 5 'ir'a4 Black can easily achieve
equality. The real test (if that is what you want to call it) is found in the following chapter. It is
a notable feature of these lines that, after the bishop returns to b7 at a later date, it seems that
White has used his extra move to his own disadvantage. The only real dangerous move con
sidered here is 5 c!Dbd2, but the idea of ... d7-d5 and ... i.b7 followed by ... a7-a5 and ... c!Da6
should give Black a good game.
1 d4 l'Llf6 2 c4 e6 3 l'Llf3 b6 4 g3 (D) .ta6

4. . . i.b4+ 5 i.d2 a5 6 i.g2 0-0 7 0-0 i.a6 8 b3
8 ... i.b7 - Chapter 7; 8 ... d5 - Game 34
5 'ifa4
5 b3 - Chapter 7
5 'ir'c2 - Game 35
5 'ir'b3 - Game 36
5 c!Dbd2 (D)
5 . . . i.b4 - Game 37
5 ...i.b7 6 i.g2 c5 - Game 38
5 . . . c5 6 e4 - Game 39
5 ... d5 6 i.g2 i.e7 7 0-0 0-0 - Game 40
5 . . . .tb7
5 ... c6 - Game 4 1
6 .tg2 c 5 7 dxc5 .txc5
7 ... bxc5 - Game 42
8 0-0 0-0 9 l'Llc3 .te7 1 0 l::l. d 1 (D)
1 0 . . a6 - Game 43; 1 0 . c!Da6 - Game 44
. . .
4 g3 5 liJbd2 1 0 .l::l.d 1
1 10
CHAPTER SEVEN I
4 i.a6 :
. . .
Main Line with 5 b3
1 d 4 tllt 6 2 c 4 e 6 3 tllf 3 b 6 4 g3 .ta6 5 some might consider worth a try is 5 ... cS, e.g.
b3 6 i.g2 and now:
This is where the main theoretical battle a) 6 ... dS seems to be the only serious op
ground for the Queen's Indian and maybe tion. Bach-Schilow, 1 997 continued 7 0-0
even for 1 d4 is taking place. Black has a �c6 8 cxdS exdS 9 .tb2 .l:c8 10 .l:el with
multitude of variations to choose from. (probably) an edge for White in one of those
Some of them · are more experimental and 'better player wins' situations. 7 ... .te7 8 cxdS
lead to an unbalancing of the position, while �xdS 9 .tb2 0-0 1 0 �bd2 �d7 1 1 �c4 .l:c8
others aim for long forced lines hoping to 12 e4 �Sf6 13 dS· exdS 14 exdS was better
equalise. The balance of play currently fa for White in Neverov-Sazonov, Yalta 1 996.
vours Black in the sense that White appar b) 6 ... �c6 7 dS! exdS 8 cxdS �e7?!
ently cannot engineer an advantage in the (8 ... �b4 9 �c3 .tb7 10 e4 .ta6 1 1 .tfl
main lines towards the end of the chapter. .txfl 12 <ifiixfl , with some advantage to
But if you, like me, not only desire to play White, is a lesser evil) 9 �c3 g6? (9 ... d6 1 0
perfect chess, but also want to win some 0-0 g6 1 1 i.b2 .tg7 merely points to the
games now and again, then you might find weird knight on e7) 10 i.gS .tg7 1 1 d6 �c6
more fun in some of the sidelines presented 12 �dS with a clear advantage in Popov
in this chapter, as all top players have done Zvjaginsev, St. Petersburg 1 998.
from time to time. c) 6 ... i.b7 7 dS exdS 8 �h4 g6 9 .tb2
.tg7 10 cxdS d6 1 1 0-0 bS 12 a4 b4 13 �d2
Game 45 was better for White in Tratar-Slak, Slovenia
Kasparov-Gelfand 1 994.
Novgorod 1997 6 .tg2 dxc4?1
After this move White has a large advan
1 d4 tllf 6 2 c4 e6 3 tllf 3 b6 4 g3 .ta6 5 tage. :Sut 6 ... .tb4+! 7 .td2 .txd2+! is
b3 d 5 ! ? interesting (rather than 7 ... .td6 8 �c3 �bd7
This move has been played more than it 9 e4! cS 10 exdS exdS 1 1 'it'e2+!, as in Nes
deserves, or at least it has been played with torovic-Rajkovic, Belgrade 1988). Then 8
an idea neither I nor practice can fully ap �bxd2 0-0 (8 ... cS 9 dxcS bxcS 1 0 0-0 0-0 1 1
prove 0£ Another dubious alternative that l:.el �c6 1 2 cxdS exdS 1 3 l:.cl �6 1 4 'i'c2
111
lllb4 1 5 1i'xc5 lllxa2 1 6 'ii'xb6 axb6 17 l:.c6 nikowski-Lau, Polanica Zdroj 1 986, while 1 1
was better for White in Nikolic-Yudasin, e5? .i..xe5 1 2 .i..xa8 'i'xd4 1 3 'ii'xd4 i.xd4 1 4
Tilburg 1 993) 9 0-0 lllbd7 1 0 l:.c1 c5 1 1 cxd5 lll c3 .i..xc3 1 5 l:.bl c6 is just stupid. Black is
(1 1 l:.e1 i.b7 1 2 cxd5 exd5 13 'i'c2 l:.e8 1 4 better.
'ii'b2 'ii'e7 1 5 e 3 a 5 with equality in Tischer 1 1 . . . 'll b c6 1 2 .'ll b d2 0-0
Eismont, Biel 1 994) 1 1 ...exd5 1 2 l:.e1 l:.e8 1 3 12 ... e5 13 d5 lll a5 14 �g1 lllxc4 1 5 lllxc4
e 3 l:.c8 1 4 .i..h3 ltc7 1 5 i.f1 ended peace i.xc4 16 bxc4 0-0 1 7 .i..h3 £5 1 8 'at>g2 gave
fully in Altennan-Yudasin, Haifa 1 995, al White an overwhelming advantage in Kas
though I have a feeling that White might be parov-Atlas, Catonsville (Simultaneous Dis
slightly better here. Instead 8 'ii'xd2 0-0 9 0-0 play) 1 996.
'ii'e7 10 cxd5 exd5 1 1 lllc3 .i.. b7 12 ltac1 1 3 �g 1 b5 14 'llxd6!
lll a6 13 l:.fdl l:.fd8 14 'ii'h2 c5 1 5 e3 was a An improvement at the time.
prototypical position in Chernin-Razuvaev, 1 4 . . . cxd6 1 5 h4 'ii' b 6
Tilburg 1 994. 15 ... d5 is suggested as an improvement by
7 'll e 5 i.. b 4+ 8 c.t>f 1 i..d 6 Dokhoian, Kasparov's second, but after 1 6
Here 8 ... lllbd7?? 9 lllc 6! is embarrassing. h51 Gelfand does not like Black's position.
The other alternatives are: 1 6 h5!?
a) 8 ... lll fd7 9 lllxc4! c6 1 0 i.b2 0-0 (after Here White could play even better with 1 6
1 0 ... b5 1 1 llle 3! Black appears to have noth d5!, with an advantage.
ing better than 1 1 ... 0-0, transposing) 1 1 1 6 . . . h6?!
lllbd2 b5 1 2 lll e3 .i..b7 1 3 'ii'c2 'ii'b 6 1 4 ll:\£3 16 ... lllxd41 17 h6 g6 1 8 lll c4 lll e2+! 19
tll f6 15 lll e5 .i.. e7 16 lll3g4?! was Karpov 'ii'xe2 bxc4 lead to complications that are not
Timman, Kuala Lumpur (8) 1 990. Karpov necessarily in White's favour.
himself indicates that 1 6 l:.c1 ! followed by 1 7 d5 'lle 5 1 8 l't)f1 b4 1 9 i..d 4 'ifa5
.i.. £3 and �g2 was better, securing an advan 19 ... .i..e2 20 i.xb6 .i..xd1 21 l:.xdl axb6 22
tage. f4 lllg4 23 dxe6 l:.xa2 24 l:.xd6 gives White a
b) 8 ... c6 9 bxc4! 0-0 10 .i..b 2 .i.. d6 1 1 tlld2 considerable advantage.
'ii'c 7 1 2 'ii'c2 secured White a small plus in 20 'lle 3 .Z:.ac8 21 .Z:.h4 .Z:.c7 22 'ii'd 2
Bykhovsky-Gofshtein, Tel Aviv 1 995.
9 'llx c4 'lld 5 10 e4 'lle 7
White i s firmly in the driving seat.

22 . . . .Z:.c3?
1 1 i.. b 2 22...lll d 3.
11 .i.. £3 ?! lll bc6 1 2 'at>g2 .i..xc4 1 3 bxc4 e5 23 i..x c3 bxc3 24 'ifd4
14 d5 lll d4 15 .i.. e3 c5 is fine for Black, Jas- Now it is over.
1 12
4. . . i. a 6 : Ma in L in e with 5 b 3
24 . . . exd5 25 exd5 •c7 26 •d 1 :ea 27 Korinthos 1 998.

i.e4 •bs 2a :t4 i.b7 29 :c 1 •as 30 9 . . . 0-0 1 0 a3 i.e7
:c2 ;.,ha 3 1 i.g2 i.a6 32 :a4 •b6 33 If Black exchanges the bishops on d2 he
ltic4! i.xc4 34 bxc4 ltif5 35 .l:xc3 ltid4? will experience permanent problems on the
36 c5 :xc5 37 •xd4 1 -0 cS-square: Korchnoi's 10 ... .txd2 1 1 'ii'xd2
--
---. .tc6 12 lLJc3 dS 13 lLJeS .tb7 14 .l:.fcl , with
Game 46 a pleasant position for White, illustrates this
Tkachiev-Nisipeanu well.
Naujac-sur-Mer 2000 1 1 ltic3 i.c6
1 1 ....i.a6 12 'i'c2! dS 13 l:fcl lLJbd7 14 e4
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.a6 5 gave White the advantage in Karpov-
b3 b5!? 6 cxb5 i.xb5 7 i.g2 i.b4+ Christiansen, Wijk aan Zee 1 993.
1 2 •c2 .d 5
12 ... .tb7!?, with the idea 13 e4 dS 1 4 eS
lLJe4 with counterplay has been suggested by
Gershon, but I think that White has the ad
vantage after the simple 1 5 l:.fcl.
1 3 ltie5 i.b 7 1 4 ltia4 ltifd7 1 5 ltid3
7 ... cS 8 0-0 .tc6 9 .ta3 lLJa6 1 0 lLJbd2

.te7 1 1 .l:tc1 cxd4 1 2 .txe7 'ii'xe7 1 3 lLJc4
0-0 14 'ii'xd4 dS 1 5 lLJfeS l:.fc8 1 6 lLJxc6
:.Xc6 17 lLJeS l:.c7 18 :.Xc7 'Wxc7 19 lLJd3
gave White a slight advantage, due to the
superiority of bishop against knight, in van
der Sterren-Timman, Holland 1 987. White has the advantage.
7 ... dS 8 0-0 lLJbd7 9 .ta3 .txa3 10 lLJxa3 1 5 . . . i.c6 1 6 ltiac5 ltixc5 1 7 lLixc5 :a7
.ta6 1 1 lkl ! 'We7 12 lLJbl l:.c8 1 3 .l:tc2 was 1 a :tc 1 i.b5 1 9 e3 lLid7?
Wells-Dautov, Bad Worishofen 1 997, but Losing.
White could have played better with 1 3 l:.c6! 20 a4! i.e2 21 i.xa5! ltixc5 22 dxc5
.tb7 14 l:c2 with an edge according to :xa5 23 b4 i.f6 24 :a3 i.c4 25 bxa5
Dautov. •bs 26 l:l.b 1 •a7 27 i.f1 ! •xc5 28
8 i.d2 a5 9 0-0 l:l.ab3 1 -0
Here White has a refutation according to
all the books, but after 9 .txb41? axb4 10 Game 47
'ii'd2 lLJc6! (the new move) 11 'WgS l:.aS 1 2 Bogdanovski-Ca brilo
'ii'xg7 ri;e7 1 3 'ii'h6 'ii'a 8 1 4 'Wd2 lLJe4 1 5 Bije/jina Dvorovi 2002
'Wb2 d S Black has enough compensation for
the pawn and had a good game, until a draw 1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.a6 5
was settled in Belakovskaia-Mastrovasilis, b3 i.b4+ 6 i.d2 i.xd2+ ! ?
1 13
1bis sideline is not as popular as one dxe4 1 3 lbgS lbc6! 14 .:!fdl lbxd4 1 5 lbxf7
might imagine when we look at the evalua lbxe2+ 1 6 'ii'xe2 .i..xe2 1 7 .:!xd8 .:!axd8 1 8
tion after the opening. I think this is because lbxd8 .:!xd8 1 9 .i..xe4 cS left White struggling
the position is rather dull and Black does for equality in Eingom-Gelfand, Moscow
have a slight structural disadvantage in hav 1 990) 1 1 ...i..b 7 12 0-0 cS 1 3 e3 'ii'e7 14 'ii'c2
ing a c-pawn compared to the e-pawn (both lba6! 1 5 .:!acl lbc7 1 6 .:!fel lbe6- with equal
pawns are in half-open files). One difference play in Hansen-Epishin, Tilburg 1 993. In
is that £2 is a better protector for the e-pawn Georgiev-Christiansen, Las Palmas 1 993
than b6 for the c-pawn. It is also significant Black chose the less accurate 10 ... lbe4?!, the
that White's king can assist the e-pawn better game continuing 1 1 lbxe4 dxe4 1 2 lDg5 lDc6
than the Black can assist the c-pawn. These 13 i.xe4 lbxd4 14 i.xh7+ �h8 15 i.d3
are, of course, relatively minor issues but, .i..xd3 1 6 'ii'xd3 'ii'xgS 17 'ii'xd4 .:!ae8 1 8 e3
structurally, White does possess a slight ad 'iib S 1 9 ..Wc4 (Dautov's 1 9 0-0-0!? cS 20 'iVdS
vantage. It should not matter theoretically, 'ii'b4 21 �b2 looks even stronger) 1 9 ... 'ii'xc4
but in practice such things often do. 20 bxc4 .:!eS!, and now Dautov gives 21
7 'ifxd2 �e2! .:!cS 22 .:!acl .:!d8 23 h4 .:!d6 24 .:!hdl
The only serious move. 7 lbbxd2 i..b7 8 .:!dc6 25 .:!d4 .:!aS 26 l:k2, retaining a clear
i..g2 cS 9 0-0 0-0 1 0 'ii'c2 lbc6 1 1 'ii'b2 advantage.
lbxd4 1 2 lbxd4 i..xg2 1 3 �xg2 cxd4 14 9 . . . 'ife7 1 0 'ifxe7+ <Jilxe7 1 1 lLic3 l:le8
'ii'xd4 dS was completely equal in Anders 1 1 . ..lbbd7 12 i.h3 .:!he8 13 0-0 �f8 14
son-Short, Na:stved 1 985. .:!fcl c6 was equal in Hjartarson-Agdestein,
7 . . . d 5 8 cxd5 Gausdal 1 987.
Also interesting is 8 i..g2 and now: 1 2 J:lc 1 c6 1 3 i..g 2 tLlbd7 1 4 0-0 �8 1 5
a) 8 ... 0-0 9 0-0 cS 10 lbc3 lbc6 1 1 .:!fdl J:fe 1 i.. b 7 1 6 i.. h 3 J:lad8 1 7 e3 h6
.:!c8 1 2 cxdS exdS 13 dxcS bxcS 14 lbxdS
ltJxdS 1 5 'ii'xdS i..xe2 1 6 'ii'xd8 .:!fxd8 17
:Xd8+ .:!xd8 1 8 .:!el .:!dl 1 9 :Xdl i..x dl 20
lbd2 lbb4 21 a4?! was Gligoric-Szabo, Mos
cow 1 956. Chekov suggests 21 a3! lbc2 22 a4
ltJd4 23 .i.. d S �f8 24 f4 f6 25 �£2 �e7 26
�e3, with a slight advantage, as being more
prudent. After 9 ... c6 10 'ii'c2 lbbd7 1 1 .:!dl
.:!c8 12 lbeS?! (1 2 lbbd2 cS 13 dxcS is a
shade better for White) 1 2... cS 1 3 lbxd7
'ii'xd7 14 cxdS cxd4 1 5 'ii'b 2 eS Black
emerged ahead in Zhu Chen-Christiansen,
Seattle 2001 .
b) 8 ... cS 9 cxdS exdS 1 0 'ii'e3+ 'ii'e7 1 1 Black has an equal position. However, this
'ii'xe7+ �xe7 1 2 lbc3 lbc6 1 3 dxcS bxcS 1 4 endgame is far from decided and White just
lbh4 lbd4 1 5 0-0-0 .:!hd8 1 6 .:!hel i..b7 1 7 plays on and on, his longer arms deciding the
�b2 g6 1 8 e4, Ruzele-Balashov, Boeblingen game ...
1 998, and White looked slightly better, al 1 8 b4 g 5 19 i..f 5 lLie4 20 h4 gxh4 21
though it is very very slight. tLixh4 lLid6 22 i..d 3 tLlf6 23 a4 J:le7 24
8 . . . exd5 9 'ife3 + a5 b5 25 tLlf5 tLixf5 26 i..x f5 lLie8 27
9 lbc3 0-0 1 0 i..g2 .:!e8 1 1 lbeS! (White lLie2 i.. c 8 28 i.. d 3 .i.d7 29 �g2 lLic7 30
must be careful here: 1 1 0-0?! lbe4! 12 lbxe4 l:th 1 J:le6 3 1 i..f 5 J:ld6 32 .i.xd7 .1:8xd7
1 14
4 . . . i. a 6 : Ma in L in e with 5 b 3
33 ltif4 ltia6 34 ltid3 .l:!.e7 35 .l:!.h5 .l:!.ee6 The main line, 9 ll'ic3, is considered in the
36 '1t>f3 <iit9 7 37 We2 liJbS 3S ltif4 .l:!.e7 next two games.
39 l:e5 �S 40 J:xe7 <iitx e7 41 .l:.h 1 ltid7 Another option is 9 'it'c2, when Black can
42 94 ltif6 43 f3 liJ9S 44 <iitd 3 l:.f6 45 equalise with 9 ... c5!, e.g. 10 l:tdl cxd4 1 1
l:h3 �S 46 <iite 2 l:d6 47 ltih5 ltie7 4S ll'ixd4 .ixg2 1 2 'it>xg2 'it'c7 1 3 .ig5 'it'e5 1 4
lti93 l:96 49 <oitf2 l:f6 50 l:h 1 l:.d6 5 1 .ixf6 'it'xf6 1 5 ll'ic3 ll'ic6 1 6 ll'idb5 l:fd8 1 7
l:c 1 <oi?97 52 @e2 <iit9 6 5 3 .l:!.9 1 W97 54 .l:.acl 'it'e7 1 8 'it'e4 f5 1 9 'it'f3 .ic5 20 ll'ia4
@d3 �S 55 e4 dxe4+ 56 fxe4 l:.96 57 ll'ie5 21 'it'c3 ll'ig4 22 ll'ixc5 'it'xc5 23 'it'd4
ltih5 liJgS 5S ltif4 .l:!.95 59 ltih3 .l:.g6 60 ll'if6 with a draw in Brenninkmeijer�Hiibner,
@e3 <>ties 6 1 ltif4 .l:!.95 62 <iitt 3 ri;e7 63 Wijk aan Zee 1992. Black should avoid 9 ... h6
.l:!.c 1 ri;d7 64 ltid3 t6 65 l:h 1 Wd6 66 10 a3, when 10 ... .ie7 1 1 ll'ic3 d5 1 2 .l:.fd l !
ltit4 .l:!.g7 67 ltih5 l:g6 6S lti93 '1t>e6 69 'it'c8?! (1 2 ...ll'ia6 improves) 1 3 cxd5 exd5 14
ltif5 Wd7 70 ri;f4 We6 7 1 J:c1 ri;d7 72 .if4 l:td8?! (1 4... ll'ih5) 1 5 ll'ih4! left White in
.l:.c5 a6 73 d5 cxd5 74 l:xd 5 + Wc7 75 charge in Karpov-Lobron, Baden-Baden
l:d6 '1tib7 76 .l:td7+ @bS 77 ltid6 l:95 7S 1 992, while 10 ... .ixd2 1 1 ll'ibxd2 d5 12 cxd5
.l:tb7 + was 79 .l:!.f7 @bS so l:.b7 + <iita S exd5 13 .l:.acl ll'ia6 14 e3 c5 1 5 .l:.fd 1 fol
S 1 .:r.h 7 <oi?bS S2 ltif5 .l:!.96 S3 ltid4 l:g5 lowed by ll'ib1 -c3 gives White a slight advan
S4 ltic6 + <oi?cS S5 ltid4 <iitb S S6 ltie6 l:e5 tage according to Dautov.
S7 ltic5 'itcS SS ltixa6 ltie7 S9 ltic5 ri;c7 9 . . . a4! 1 0 bxa4
90 l:.xh6 ltic6 91 l:h7 + .l:!.e7 92 l:.xe7+ 10 a3?! axb3! does not benefit White.
ltixe7 93 a6 <iitb 6 94 ltid7 + <iitx a6 95 1 0 . . . h6 1 1 a3 .i.a5! 1 2 i.xf6 'ii'xf6
ltixf6 <l.>b7 96 ltid5 lti96 + 97 'itf5 ltih4+ Black has at least equalised.
9S 'itf6 <oi?c6 99 95 @d6 1 00 96 ltixg6 1 3 e3
1 -0 13 'it'd3 is probably better.
In TWIC this game is quoted as ending in 1 3 . . . c5!
a draw here. I would say that this is very
unlikely, but could have happened as a result
of time trouble. Whether the result was 1/2-1/2
or 1 -0 is of no great importance except, of
course, to the players. White outplayed Black
from an equal position, which is all we need
to know.
Game 48
Petursson-Polugaevsky
Reylefavik 1987
1 d4 lbt6 2 c4 e6 3 lbt3 b6 4 g3 i.a6 5
b3 i.b7 6 i.g2 i.b4+ 7 i.d2 a5 ! ? 1 4 .:r.a2
This i s the choice of the ambitious player. Anoths:r Polugaevsky game goes 14 ll'ibd2
The typical positions in this line might be .ixd2 1 5 ll'ixd2 .ixg2 16 �xg2 cxd4 17 exd4
marginally better for White, as all positions in 'it'xd4 1 8 .l:.bl .l:.a6 19 .l:.el l:c8 20 .l:.e3 ll'ic6
the QID tend to be, but results are almost 21 llbb3 ll'ie5 22 l:e4 'it'd6 23 'it'e2 'ii'c 6 24
always decided by who is the better player on �gl d6 25 'it'e3 'it'c5 26 l:tb4 'i'xe3 27 llxe3
the day. and Black won the ending in Epishin
s 0-0 0-0 9 i.g5 Polugaevsky, New York 1989.
1 15
1 4 . . . .:l.a7 1 5 llle 5?! 0-0 9 lll c 3

1 5 dxc5 bxc5 1 6 �bd2 with approximate
equality.
1 5 . . . cxd4! 1 6 'i'xd4 d6! 1 7 lllf3
17 ltlg4!? 'ii'xd4 18 exd4 i.c8! 19 ltle3
i.d7 also gives Black something. The bishop
on a5 is potentially very strong.
1 7 . . . e5 1 8 'i'd 1
9 . . . d5
The more flexible 9 ... d6 is considered in
the next game. The main reason for not play
ing this move is an attempt to push ... d7-d5
in one go, investing the extra tempo some
where else. After the alternative 9 ... l:te8 1 0
'ii'c2 h 6 1 1 l:tfe1 i.xc3 1 2 i.xc3 i.e4 1 3
1 8 . . . llld 7 'iVb2 d 5 14 i.f1 Black has tried:
Preferable is 18 ... 'i'e61 1 9 �fd2 i.xg2 20 a) 14 ... i.xf3!? 15 exf3 �bd7 16 i.d3 a4
�xg2 d5 21 'i'b3 d4! with a definite advan 17 l:ad1 was Polugaevsky-Anand, Roque
tage for Black. brune 1 992, and 1 7 f41 would have secured
1 9 lll fd2 .txg2 20 ltxg2 'i'e6 21 .:l.b2! White an edge.
lll c 5 22 .:l.b5 .:l.fa8 23 lll b 3 lllx b3 24 b) 14 ...�bd7 1 5 �d2 'iVe7 16 l:tac1 c6 1 7
'i'xb3 .:l.c8! 25 .:l.d5?! e 3 was slightly better for White in Pinter
25 J:r.d1 :Xc4 26 .l:xd6 'ii'xd6 27 'ii'xc4 Garcia Trobat, Spain 1 993.
.l:c7 keeps White's disadvantage to a mini c) 1 4... c6 15 ltld2 i.g6 16 a3 gave White a
mum. tiny advantage in Alterman-Korchnoi, Beer
25 .:l.ac7 26 .:l.fd 1 .:1.xc4 27 .:l.xd6 'i'f5
. • . Sheva 1 992.
28 :I.dB + lth7 29 .:l.xc8 .:1.xc8 30 llld 2 1 0 "it'c2 llla 6 1 1 .:1.ad 1 ! h6
e4! 3 1 ltg 1 ? 1 1 ....l:c8 12 cxd5 exd5 13 i.h3 favours
31 �f1 l:r.c3 32 'ii'b 1 .l:d3 might be terri White.
ble but it is White's only chance of survival. 1 2 llle 5 'i'c8 1 3 .th3! ? "it'e8 1 4 cxd 5
3 1 . . . .txd2! 32 .:1.xd2 .:l.c1 + 33 ltg2 exd5
33 l:td1 'ife6! picks up a rook. 14 ...ltlxd5 15 �xd5 i.xd5 16 e4 i.b7 17
33 . . . 'i'f3+ 34 lth3 .:l.g 1 0-1 i.f4 is not to be recommended for Black.
1 5 .tf4 :I.dB 1 6 'i'c 1 ! .tc8
Game 49 16 ... �e4 17 ltlxe4 dxe4 18 'ii'c4 is only a
Karpov-Salov little better for White.
Unares 1992 1 7 .tg2 .tb 7 1 8 a3 .td6 1 9 lll c 4!
The knight seeks fresher pastures. Black is
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 .ta6 5 in a bad situation.
b3 .tb7 6 .tg2 .tb4+ 7 .td2 a5 8 0-0 1 9 . . . .txf4 20 'i'xf4 .tc8 2 1 llle 3 .te6 22
1 16
4 . . . i. a 6 : M a in L in e with 5 b 3
tt::\t 5 i.xt5 23 �xt5 •ea 24 •d3 tiJbB 25 1 1 . . . i.xc3 1 2 i.xc3 i.e4 1 3 �b2
b4 axb4 26 axb4 tt::\c 6 27 b5 tt::\a 5 28 The great expert of the QID-g3, GM
:te 1 :tea 29 :c1 •d6 30 :c2 :e7 3 1 Dautov, plays the more flexible 13 °ili'cl !?,
e 3 g6 32 J:ec 1 �g 7 3 3 tt::\e 2 :as 34 tt::\t4 when the queen is yet to commit to a flank.
:a7 35 :a2 ltd7 36 �a3! •xa3 37 .C.xa3 Dautov-Romanishin, Essen 2001 continued
�ta 38 tt::\d 3 �ea 39 tt::\ b4 h5 40 :ac3 13 ... h6 14 i.h3 i.h7 15 tiJd2 tDe4 1 6 tDxe4
�dB 41 :c6! tt::\e4 42 i.h3 :d6 43 i.xe4 1 7 f3 i.h7 1 8 e4 a4 19 °iWh2 with a
:xd6+ tt::\x d6 44 i.f1 ! tt::\e4 45 tt::\x d5 pull, while 13. .. "ili'h8 14 i.h3 c6 1 5 tiJh4 h6
tt::\ b 3 46 :c2 g 5 47 i.e2! :a 1 + 48 11i>g2 16 f3 i.h7 17 e4 b5 18 °ili'd2 gave White
tt::\c 1 49 t3 tt::\d 6 50 i.t1 g4 51 txg4 hxg4 similar prospects in Dautov-Zeller, Boe
52 tt::\t 6 ::ta2 53 :t2! :xt2+ 54 �xt2 blingen 1 996.
tt::\a 2 55 e4 tt::\ c 3 56 �e3 tt::\ c xb5 57 d5! 1 3 . . . a4
c5 58 dxc6 tt::\ a 7 59 e5 tt::\d c8 60 c 7 + ! Just one of several possibilities.
�xc7 6 1 i.c4 tt::\ c 6 62 tt::\xg4 1 -0 a) 13 ... c6!? 1 4 i.f1 b5 15 ttJg5 i.g6 1 6
---- i.g2 d5 17 cS?! h6 1 8 tiJf3 ttJe4! led to equal-
Game 50 ity in Pinter-Almasi, Budapest 1 997. Avrukh
T opalov-Adams suggests 1 7 f3!? as an improvement. Sjoberg-
FZ!Jitsu Siemens Masters Frankfurt 2000 Almasi, Malmo 1994 went 1 5 d5?!, and now
the strongest seems to be 1 5 ... exdS! 1 6 cxb5
1 d4 tt::\t 6 2 c4 e6 3 tt::\t 3 b6 4 g3 i.a6 5 c5 1 7 i.xf6 tDxf6 1 8 tiJd2 i.g6 and Black
b3 i.b7 6 i.g2 i.b4+ 7 i.d2 a5 8 0-0 has a pleasant grip on the centre.
0-0 9 �c2 d6 1 0 tt::\c 3 tt::\b d7 1 1 :te 1 b) 13 ... d5 14 i.h3!? i.£5 (14... dxc4 1 5
bxc4 c5 1 6 d5! exd5 17 cxd5 i.xd5?! 1 8
.l:tadl gives White a dangerous initiative ac
cording to Chemin) ' 1 5 i.xf5 exf5 16 tiJh4
tDe4!? 17 f3 tDxc3 1 8 "ili'xc3 f4 1 9 .l:tacl with
a slight edge for White in Chernin
Romanishin, Taastrup 1 992.
c) 13 ... °ili'b8 14 i.fl (in Petursson-Lemer,
Moscow 1987 White tried 14 i.h3 bS 1 5 dS
e5 16 tiJd2 i.g6 17 cxb5 °ili'xb5 18 e4 ttJc5 1 9
i. f1 °iWb6 and was already struggling) 1 4. . .c5
1 5 .:t.adl .l:!.d8 1 6 i.h3! a4! 17 tiJd2 axb3 1 8
axb3 i.b7 1 9 d5 with compensation for the
pawn, should Black decide to take it, in van
This is the most difficult fo r Black to Wely-Karpov, Tilburg 1 996. 14 ... i.xf3!? 1 5
meet. Others: exf3 d5 1 6 .l:!.acl c6 was Hulak-Polugaevsky,
a) 1 1 .l:!.adl i.xc3 12 i.xc3 i.e4 13 °ili'cl Zagreb 1 987. Now 1 7 f4 should give White
a4 14 i.h3 b5 15 tiJd2 axb3 16 axb3 bxc4 1 7 some advantage.
bxc4 .l:!.a2 led to full equality in Morovic Fer 1 4 .tn c5
nandez-Adams, Istanbul 2000. The position 1 4 ...°iWb 8!? 15 tiJd2 b5 1 6 f3 Jf.. c 6 17 b4
is about equal, perhaps a bit better for Black. bxc4 1 8 tDxc4 d5 1 9 tiJa3 .l:!.e8 gave Black a
b) 1 1 tDel "ili'c8 12 l:kl l:.e8 13 tiJb5 good game in van der Sterren-Timman,
i.xg2 14 ttJxg2 'ib8 1 5 a3 i.xd2 1 6 "ili'xd2 Gouda 1 997. But we should still question
c6 17 tiJc3 d5 with equality in Korchnoi whether White could have played better.
Adams, Wijk aan Zee 2000. 14 ... i.b7 15 tiJd2 ttJe4 16 i.g2 tiJdf6 1 7
1 17
tt::lxe4 tt::lxe4 1 8 d 5 tt::lxc3 19 'Wxc3 e5 2 0 b4 6 1 g6 ll.lb5 62 g7 ll.lc3 63 i.g4 ll.la4+ 64

f5 offered Black sufficient counterplay in �b1 i.ga 65 e5 i.h7 + 0- 1
Pinter-Romanishin, Balatonbereny 1 995, but
it seems that White was not out to test his Game 5 1
opponent in this game. Ehlvest-Chandler
1 5 ll.ld2 i.b7 1 6 b4 Lucerne 1989
1 6 i-h3 'ir'c7 17 :adl :a6 1 8 d5 e5 19 f4
l:tfa8 was the course of Kragelj-Romanishin, 1 d4 ll.lt6 2 c4 e6 3 ll.lt3 b6 4 g3 i.a6 5
Pula 1 994, with a good game for Black. b3 i.b7 6 i.g2 i.b4 + 7 i.d2 c5!?
1 6 . . . a3 1 7 fib3 cxd4 1 a i.xd4 e5 1 9
i.e3 fic7 20 l:l.ac 1 'iWc6 2 1 t3 fia4 22
fib 1
Initially this seems quite strange but, when

you think about it, Black is trying to gain
some control over the dark squares on the
22 . . . d5! queenside, as well as opening the c-file. If
It is obvious that Black has outplayed his you compare it with 7 ... a5, then 7 ... c5 better
opponent in the opening and now has a bet meets 9 tt::lc 3.
ter game. a 0-0
23 i.t2 e4 24 i.d4 :I.tea 25 cxd5 i.xd 5 Harmless is 8 dxc5 i.xc5 9 4.Jc3 tt::le4 10
26 t4 e3 27 ll.lc4 i.e4 2a 'iWa 1 'iWxb4 29 tt::lxe4 i.xe4 1 1 i.c3 0-0 12 0-0 tt::lc 6 13 'ir'd2
ll.lxe3 b5 30 i.h3 l:.ada 31 :!.ed 1 i.aa 32 d5 14 cxd5 'ir'xd5 with complete equality,
l:.b 1 fia4 33 'iWc3 ll.lta 34 i.e5 l:.xd 1 + Hjartarson-Korchnoi, Barcelona 1 989.
35 l:.xd 1 ll.ld5 36 l:l.xd5 i.xd 5 37 i.xg7 a . . . 0-0 9 i.c3
hl.xe3 3a 'iWxe3 'iWd 1 + 39 �t2 �xg7 40 9 dxc5 transposes to 8 dxc5. 9 tt::l c3 cxd4
'iWe5 + t6 41 'iWe7+ i.t7 42 'iWxa3 'iWd4+ 1 0 4.Jxd4 i.xg2 1 1 'it>xg2 d5 results in
43 'iWe3 'iWxe3+ 44 �xe3 i.xa2 immediate equality, a draw being agreed in
As you can see it is not only club players Olafsson-Petursson, Reykjavik 1 989, for
who play on in embarrassing positions. 1bis example.
probably has something to do with a fighter's 9 . . . d5 1 0 cxd 5
attitude... Or 10 tt::l e 5 4.Jc6! 1 1 tt::lxc6 (1 1 i-xb4
45 �d4 i.c4 46 e3 ll.le6 + 47 �c3 ll.lc5 tt::l xb4! does not trouble Black) 1 1 ...i.xc6 1 2
4a i.t5 h6 49 h4 i.e6 50 i.c2 ll.la6 5 1 i.xb4 (1 2 e3, a s in Goldin-Douven, Palma
i.d3 b4+ 52 �d4 ll.lc7 5 3 e 4 b3 5 4 �c3 de Mallorca 1 989, leads to equaliity after
�ta 55 i.e2 ri;e7 56 g4 �d6 57 g 5 hxg5 1 2... i.xc3 13 tt::lxc3 :c8 14 dxc5 dxc4!)
5a txg5 txg 5 59 hxg5 �c5 60 �b2 �b4 1 2 ... cxb4 13 'ir'd3!? dxc4 1 4 bxc4 i.xg2 1 5
1 18
4 . . . i. a 6 : M a in L in e with 5 b 3
rJolxg2 b5! 1 6 cxb5 'iid5+ 1 7 f3 l::t fd8 with a 1 5 ... ltJc6? 1 6 dxc5 i.xc3 1 7 i-xc3 bxc5 1 8
complex position which should be okay for b4 is very good for White.
Black, Nikolic-Salov, Brussels 1 988. 1 6 J:!.c1 'it>h8 1 7 e3 i.xc3 1 8 i.xc3 'ii'g 5
10 i-xb4 cxb4 11 ltJbd2 ltJc6 1 2 e3 l::tc 8 19 i.b2 J:!.ad8 20 f4! ? exf3 2 1 J:!.xf3 'iie 7
13 'iib 1 l::t c7 14 l:tc1 ike7 1 5 'tib2 l::t fc8 1 6 22 'iid 2 J:!.xf3 23 .i.xf3 J:!.e8 24 J:!.e1 lLlc7
lDe5 was equal in Tukmakov-Amason, Reyk 25 i.g2 lLle6 26 .:lf 1 lLlg5 27 J:!.f5 h6 28
javik 1 990. I suspect that Black might have J:!.e5 'ii'f7 29 .=.xe8+ 'ii'xe8 30 'ii'f 2?
the better practical chances here. Allowing the creation of a passed pawn.
1 0 . . . exd5 Instead 30 dxc5! bxc5 31 b4 is unclear.
Only move. 1 0 ...ltJxd5 1 1 i-b2 cxd4 1 2 30 . . . c4! 31 bxc4 dxc4 32 d5 .i.xd5 33
ikxd4 'iif6 1 3 a3! 'iixd4 14 lDxd4 i-c5 1 5 h4 lLle4 34 'ii'f 5?
ltJb5! lDc6 1 6 b4 i- e7 1 7 e4 lDf6 1 8 e 5 lDd5 34 i-xe4 i-xe4 35 'if'f4 is an unattractive
19 lD1c3! was a solid plus for White in Kar lesser evil.
pov-Korchnoi, Biel 1 992, and 10 ... i-xd5 1 1 34 . . . c3!
a3 i-xc3 1 2 lDxc3 i-b7 1 3 dxc5 bxc5 1 4 Now this is possible.
ikxd8 l::txd8 1 5 lDe5 i-xg2 1 6 rJolxg2 was 35 'ii'xd5
uncomfortable for Black, if not a great deal This loses on the spot. There is still some
worse, in Kishnev-Romanishin, Munich resistance left in 35 i.a1 c2 36 i-b2 lDf6!? 37
1 992. i.xd5 ikxe3+ 38 rJiih2 c1\i' 39 i.xcl 'it'xc1,
1 1 i.b2 with an extra pawn for Black.
11 dxc5 i.xc5 12 i.b2 ltJa6 13 lDc3 ltJe4 35 . . . lLlf6! 36 'ii'd 4 cxb2 37 'ii' x b2 'ii'x e3 +
14 l::tc 1 ltJc7 1 5 ltJd4 l::te 8 favoured Black in 38 'ii'f 2 'ii'x a3 39 'ii'f4 'ii'c 5+ 40 'it>h 1 b5
Novikov-Hjartarson, Tilburg 1 992. 4 1 'ii'd 2 'iic 7 0-1
1 1 . . .lLle4 1 2 a3
Brenninkmeijer-Amason, Groningen 1 990 Game 52
resulted in a level game after 12 dxc5 bxc5 1 3 Dokhoian-Romanishin
a3 i-a5 1 4 b 4 itb6!? 1 5 lDc3 lDxc3 1 6 i-xc3 Yerevan 1989
cxb4 17 axb4 i-xb4 1 8 i.xb4 'iixb4 19 l::tb 1
'iie7 20 ltJd4!? 'iid 7! etc. 1 d4 lLlf6 2 c4 e6 3 lLlf3 b6 4 g3 i.a6 5
1 2 . . . i.a5 1 3 lLlfd2 f5!? b3 i.b7 6 .i.g2 .i.b4+ 7 �d2 c5 8 0-0
0-0 9 .i.xb4 cxb4
Black has had a successful opening. The

situation is unclear. 1 0 'iid 3!
1 4 lLlxe4 fxe4 1 5 lLlc3 lLla6! The most serious move.
1 19
a) After 10 tllbd2 d6 1 1 'i'c2 it is obvious bxa3 13 l:.xa3 e5 14 lllh4 i.xg2 1 5 lllxg2

that the queen is less well placed on c2. d5!? 1 6 llle3 dxc4 1 7 tlldxc4 gave White an
Shariyazdanov-Baklan, Neum 2000 contin advantage in Ftacnik-Romanishin, Maribor
ued 1 1 ...tll c6 1 2 e4 e5 13 d5 tllb 8 1 4 a3 bxa3 1 995) 1 2 l:.adl ilc7 1 3 e4 e5 1 4 lllh4 lllc6
(14 ... llla 6!, a suggested of Rabinovich, seems was equal in . Gruenberg-Hjartarson, Novi
to be more natural) 15 b4 a5 1 6 c5!? (this Sad 1990. 13 e3!? stays flexible in the centre
double pawn sacrifice was, of course, and seems to be stronger. After e3-e4 ... e6-e5
planned, and is rather complex) 1 6 ... lll a6 17 Black is always equal.
c6 lllxb4 1 8 'i'c3 i.a6 and Black had good 1 1 'i'e3 �c6 1 2 �bd2 d5 1 3 �xe4?
counterplay in this very double edged posi Later this variation was refuted with 13
tion. cxd5! exd5 (1 3 ... i.xd5 14 'i'd3 gives White
b) 10 a3 and now Black has: the advantage as e2-e4 is coming) 1 4 llle5
bl) 10 ...bxa3 1 1 lllxa3 a5 12 lll e l i.xg2 i.xg2 15 lllxc6 'i'd7 16 tlle 5 'i'h3 17 l:.fcl
was a short draw in Bauer-Baklan, Germany with a clear advantage for White in Dautov
1 999. It seems that Black should be equal. Romanishin, Kecskemet 1 989.
There is also 1 2 tllb 5! to consider, fighting 1 3 . . .�xe4 1 4 �d2
for e4. This seems more dangerous, although Preferable is 14 a3 bxa3 15 l:.xa3 lllb4 1 6
I believe Black is fine. l:.cl a5, with equality in Lemer-Romanishin,
b2) 10 ... llla6 1 1 e3 'iie7 12 a4!? llle4 13 a5 Nikolaev 1 995.
b5! 14 lll fd2 tll c 3! 1 5 ilc2 i.xg2 16 �xg2 1 4 . . . �c3
lllx bl 1 7 l:.axbl l:.fc8 with sufficient coun Now the threat of ... lllxd4 dictates White's
terplay for Black in Epishin-Petursson, Bem moves, and Black already have an advantage.
1 9 9 1 . 1 1 axb4 lllxb4 12 lll c3 llle4 13 'i'c1 1 5 �b1
lllxc3 14 'iixc3 a5 15 tlle 5 was a draw in
Yusupov-Benjamin, Saint John 1 988, while
1 1 lll e 5 i.xg2 1 2 <ifiixg2 'i'b8 13 tll d3 bxa3
14 :xa3 'i'b7+ 1 5 �gl d5 1 6 tlld 2 lllb 8 1 7
e 3 lll c 6 was yet another draw in Hracek
Dorfman, Hamburg 2001 .
1 0 . . . .te4?
Serving only to create problems for Black.
Both alternatives are superior:
a) 1 0 ... d5 and now:
al) 1 1 lllbd2 tllbd7 12 l:.fcl 'i'e7 13 'i'e3
(1 3 lll e5 l:.fc8 14 f4 tll e8 1 5 e4 tll df6 was
more or less equal in Psakhis-Lutz, Baden
Baden 1 992) 13 ...l:.ac8 14 tll e5 lllxe5 1 5 1 5 ll\£3 seems to be more natural.
dxe5 llld7 1 6 cxd5 i.xd5 1 7 i.xd5 exd5 1 8 1 5 . . . �xd4! 1 6 �xc3 �c2 1 7 'i'c 1 �xa1
ll\ £3 :c5 1 9 iid4 l:.e8 with equality (or 18 �b5 a6 19 �d4 �xb3 20 axb3 .:l.c8
something very close) in Nikolic-Amason, 21 .:l.d 1 'i'f6 22 e4 .:l.fd8 23 'i'b 1 dxc4?
Moscow 1 990. 23 ...dxe4! 24 lll c2 l:.xdl+ 25 'i'xdl :d8
a2) 1 1 llle 5 lllbd7 12 lllxd7 'i'xd7 13 26 'iie l 'i'b2! secures Black a clear advan
lll d 2 .i::r. fd8 1 4 l:.fdl dxc4 1 5 lllxc4 with tage.
equality in Gavrikov-Korchnoi, Swiss League 24 �c6 .:1.xd 1 + 25 'i'xd 1 e5 26 �xb4
1 995. cxb3 27 �d5 'i'd6 28 'i'xb3 .:l.c 1 + 29
b) 10 ... d6 1 1 tllbd2 a5 (1 1 ...lllbd7 12 a3 .tn b5 30 �g2 'i'c5 31 .td3 'i'd6 32
1 20
4 . . . i. a 6 : Main L in e with 5 b 3
i.c2 'i'c5 33 i.d3 'i'd6 34 i.c2 'i'c5 GM between 2450 and 2550 who view chess
%-% as a random profession, playing a sort of
accountant's chess. These people often play
Game 53 such technical lines, which are equal but not
N ielsen-Anand dead, and then they make their 75% score
Moscow 200 1 without any real risks. Here the position is
just equal.
1 d4 ltif6 2 c4 e6 3 ltif3 7 . . . c6
An important point here is that this can Black has two major alternatives:
lead to transposition to the Catalan with 3 g3 a) 7 ... i.b7 8 �c3 0-0 9 0-0 d5 (best -
i.b4+ 4 i.d2 i.e7 5 i.g2 d5 6 �£3 0-0 7 0-0 9 ... �a6 1 0 e3 d5 1 1 'i'e2 c5 1 2 l:lfd1 �c7 13
c6 8 'i'c2 b6 9 b3 i.a6 1 0 l:.dl �bd7, and �e5 i.d6 14 i.el 'ike7 1 5 f4 l:lfd8 1 6 g4 was
you have the move after 10 ... 0-0. Of course better for White in Van der Sterren-Sokolov,
Black can also delay castling in this line but, Amsterdam 1 994, while 9 ... c5 10 d5! exd5 1 1
as we shall see, this has no positive sides. �el �a6 1 2 cxd5 d6 1 3 �d3 �c7 1 4 e4
3 . . . b6 4 g3 i.a6 5 b3 i.b4+ 6 i.d2 i.e7 �d7 15 f4 was quite pleasant for White in
Timman-Franco Ocampos, Mar del Plata
1 982) 10 cxd5 exd5 1 1 i.cl l �a6 1 2 i.b2 c6
13 �e5 i.d6 14 �d3 l:.e8 1 5 l:lc1 �c7 1 6
l:lel �e6 1 7 e 4 with a slight advantage for
White in a typical position, Karpov-Kurajica,
Tilburg 1 994.
b) 7 ... d5 has a reputation for achieving
equality, but a poor .record when it comes to
practical results. The position arising after 8
cxd5 �d5 9 �c3 �d7 1 0 �xd5 exd5 1 1
0-0 0-0 1 2 l:lc1 l:le8 1 3 l:.el c5 1 4 i.e3 i.b7
15 i.h3 cxd4 1 6 i.xd4 �f6, as in Karpov
Sokolov, Linares 1 987, might be equal, but I
1bis is the main line everywhere, al would not like to have such a position with
though, as we have seen, not Black's only Black, where White has a simple blockade of
reasonable line. the isolated pawn. All in all it might be equal
7 i.g2 from an objective point of view, but Black
White also has 7 �c3 0-0 8 e4 as an inter will find it hard in practice.
esting possibility, but time has shown that 8 0-0 d5 9 'i'c2 ! ?
after 8 ... d5 9 cxd5 i.xfl 10 <!ixfl exd5 1 1 e5 1bis variation was previously thought to
�e4 1 2 <!ig2 'i'd7 1 3 'i'e2 �xc3 1 4 i.xc3 be harmless, but after 8 i.c3 d5 9 �e5 had
�c6 1 5 .z:thel �dB 16 �g1 �e6 1 7 f4 g6 1 8 been drained of new ideas, White tried to go
�£3 �g7, as in Tukmakov-Vaganian, Odessa elsewhere to find an advantage. I do not
1 989, Black is fine, although, of course, there thing that this is the way, but nor can I sug
is a lot of play in the position. The main line gest another path. Other moves have been
is marked by there being very little place for tried here, but none so serious that it de
individual performance or any other kind of serves a mention in a book like this.
competition between the two players, so 9 . ltibd7 1 0 l:d 1
. .
such lines do attract a lot of players who see The most precise approach. The alterna
chess as 'going to the office'. There are many tives were:
121
Q u e e n 's India n D e fe n c e
a) 1 0 a4 0-0 1 1 aS ltie4 12 axb6 axb6 13 in Gelfand-Lautier, Biel 2001 . White will

.if4 bS 14 ltibd2 ltixd2 1 5 ltixd2 bxc4 1 6 expand with f2-f4 and e2-e4 and send his
bxc4 .i.f6 with equality Gurieli-Sokolov, knight to d6, while Black maintains his extra
Basle 2001 . pawn.
b) 1 0 .if4 l:tc8 1 1 ltibd2 (1 1 l:tdl is, of 1 1 ..tf4! c5?!
course, still the main line) 1 1 ...cS 1 2 dxcS 11 ... 0-0 still transposes to the 1 Oth move.
ltixcS 1 3 l:tacl 0-0 14 'i'b2 bS 1 5 cxdS ltixdS 1 2 dxc5 .bc5 1 3 lbc3!
1 6 .ieS f6 1 7 .id4 b4 with a better position
for Black in Babu-Prasad, Kuala Lumpur
1 992. But White did not play the best possi
ble lines. Meanwhile 12 ... i.xcS 13 'i'b2 dxc4
1 4 ltixc4 .ixc4 1 5 bxc4 0-0 1 6 l:tadl 'fie7
was equal in Tjiam-Van der Wiel, Holland
1 993.
1 0 . . . :cs
Unless some improvement is found over
the main game here, Black will have to go
back to castling and slow play: 1 0 ... 0-0 1 1
.if4 l:tc8 1 2 ltic3 and now:
a) 1 2 ... ltihS 1 3 .icl ltihf6 1 4 e4 dxc4 1 5
ltid2 b S 1 6 bxc4 bxc4 1 7 ltia4 (1 7 'fia4 i.bS This remarkable piece of home prepara
1 8 'fic2 'i'aS 19 lti£3 h6 20 .id2 .ib4 21 tion should have refuted Black's set-up and
l:tabl .ixc3 22 .ixc3 i.a4 23 'fid2 'i'hs and will, in the future, make Black return to
Black has a good position, Van Wely-Adams, 1 0 ... 0-0. The latest try at the top level looked
Wijk aan Zee 2001) 17 ... cS 1 8 dS exdS 1 9 like this: 1 3 'i'b2 0-0 14 ltic3 .ib4! 1 5 ltieS
exdS ltieS 20 .ib2 .id6 resulted i n a com bS 1 6 a3 .ixc3 1 7 'i'xc3 bxc4 and Black had
plex struggle in Georgiev-Mitkov, Skopje the advantage in Van Wely-Polgar, Hoogev
2002, where the best player won. een 2001 .
b) 12 ...bS 13 cS b4 14 ltia4 .ibS 1 5 ltib2 1 3 . 0-0 1 4 e4!
. .
ltie4 1 6 ltid3 'i'aS is extremely complex, The notes that follow are based on Peter
Hansen-Emms, Esbjerg 2000. Perhaps Black Heine Nielsen's annotations in the splendid
is equal, but it is a considerably difficult posi Swedish magazine Schacknytt.
tion to handle. 1 4 . . . lbg4
c) 12 ... dxc4!? seems to be the strongest, al 14 ... dxc4 15 eS ltig4 16 ltigS g6 17 ltige4
though both the alternatives are fine. Kram would also give White the advantage.
nik-Leko, Hungary 2001 continued 13 ltid2 1 5 exd5
cS 14 dxcS .i.xcS 15 ltide4 ltixe4 1 6 ltixe4 15 l:td2!? is a more careful alternative ac
cxb3 17 'i'b2 bxa2 18 l:txa2 .ic4 19 l:txa7 cording to Nielsen, but after 1 5 ... eS! it seems
.idS 20 ltixcS bxcS 21 e4 .ic6 22 l:td6 l:ta8 to be just as complicated - only now I prefer
23 'i'a3 l:txa7 24 'i'xa7 'fic8 25 .ifl ltif6 26 Black.
.ia6 'i'e8 27 £3 .ibS 28 'i'xcS i.xa6 with a 1 5 . . . lbxf2 1 6 :n e5
draw. White obviously had compensation, 1 6 ... ltih3+ 1 7 �hl ltixf4 1 8 gxf4 exdS 1 9
but hardly an advantage. ltig5 ltif6 20 ltixdS g6 2 1 l:tadl gives White a
Earlier White has also tried 1 1 a4 cS 1 2 winning attack. The plan is 'i'c3-h3, with
ltic3 dxc4 1 3 bxc4 .ixc4 14 ltieS ltixeS 1 5 numerous threats.
dxeS ltidS with a complex but even position 1 7 i.. c 1 e4 1 8 lbh4 e3!
1 22
4 . . . .i. a 6 : M a in l in e with 5 b 3
b3 .i.b4+ 6 .i.d2 .i.e7 7 .i.g2 c 6 a .i.c3

d5 9 �e5 �fd7 !
9 ... 0-0 10 �d2 �fd7 1 1 �d3 has been
known to be an unnecessary extra possibility
for White since Lputian-Portisch, Manila
1 990. White is slightly better.
1 O �xd 7 �xd7 1 1 �d2 0-0 1 2 0-0 :ea
This has been the main line almost for
ever, but the Portisch line with 1 2.:.�f6 13
e4 b5 has brought Black good results. Kar
pov-Timman, Hoogeveen 1999 continued 1 4
Ae1 dxe4 1 5 1i'c2 Ac8 1 6 Aad1 �d5 1 7
i.xe4 (1 7 cxd5 cxd5 1 8 a3 i.xa3 gives Black
In their preparation Peter's team had fo good compensation for the pieces, while
cused on 1 8 ...'i'f6 19 i.b2 �d1+ 20 <it>h1 Black also had excellent compensation after
�e3 21 'ii'e2 �fl 22 Axfl 1i'd8 23 �e4, 1 7 i.b2 f5 1 8 cxd5 cxd5 1 9 1i'b1 b4 20 �fl
and White has overwhelming compensation. 1i'b6 21 f3 i.f6 in Karpov-Polgar, Buenos
1 9 �a4 .i.d4 Aires 2000) 1 7 ... �xc3 1 8 'ii'xc3 i.f6 1 9 1i'c2
1 9 ...�3+ 20 i.xh3 e2+ 21 �xc5 exf11i'+ g6 20 �£3 bxc4 21 bxc4 'ii'a 5 22 c5 i.b5
22 i.xfl �xc5 23 i.b2 is not as promising with a level game.
as earlier exchange sacrifices but still leads to 12 ... b5?! 13 Ae1 bxc4 14 bxc4 �b6 1 5 c5
a clear advantage for White. One line is �c4 16 e4 led to a plus for White in No
23 ... b5 24 �£5 bxc4 25 i.xg7, and White's vikov-Hellers, Copenhagen 1 991 .
attack decides. 1 3 e4 c5
20 .i.b2
20 l:.b1 ?! �h3+ 21 i.xh3 e2+ 22 �h1
exfl'ii' + 23 i.xfl i.f6!, followed by ... b6-b5,
gives Black reasonable chances.
20 . . . .i.xb2 21 'ifxb2?
A blunder. 21 �xb2! g6 is necessary ac
cording to Anand. 22 1i'c31 l:.e8 23 Aae1
leads to a clear advantage to White according
to Curt Hansen.
21 . . . b5 22 �f5 'ifg5
The move White had missed.
23 �d6 bxa4 24 �xca .i.xc8 25 'ifd4
�f6 26 J:ae 1 :ea 27 d6 �h3 + 2a .i.xh3
.i.xh3 29 l:f4 'ifa5 30 J:xe3 :Xe3 3 1 The main move. 1 3 ...b5 14 Ae1 clxc4 1 5
'ifxe3 axb3 3 2 d7 .i.xd7 3 3 axb3 .i.h3 bxc4 bxc4 1 6 1i'a4 i.b5 1 7 1i'c2 l:.e8 1 8 a4
0-1 i.a6 1 9 :ad1 i.£8 20 l:.e3 1i'c7?! (20 ... c5 21
----- d5 exd5 22 exd5 with a slight advantage is
Game 54 probably better) 21 i.fl c5 22 d5 exd5 23
Kamsky-Karpov exd5 Axe3 24 fxe3 �e5 25 1i'e4 and White
Elista 1996 was well in control in Grabliauskas-
,________________. Shariyazdanov, Swidnica 1 997. 14 ... dxe4 1 5
1 d 4 �f6 2 c 4 e 6 3 �f3 b6 4 g3 .i.a6 5 i.xe4 c5 1 6 cxb5 i.xb5 1 7 d5 exd5 1 8 i.xd5
1 23
tiJb6 1 9 �g2 was also favourable for White tiJxb6 is the only chance, but of course Black
in Shipov-Gagarin, Moscow 1 994. After cannot hope for anything other than a draw)
1 9 ... �f6?! 20 �xf6 ir'xf6 21 l2Je4 'i'g6 22 25 .l:r.dxf7! 'i'xf7 26 .l:r.xf7 .l:r.xf7 27 'i'd8+ l:r.f8
'i'd2! Black was in trouble, with 'i'a5 threat 28 'i'd5 and White won, Chernin-Veingold,
ened. Seville 1 993.
1 4 exd5 exd5 1 5 dxc5 dxc4 1 6 c6 cxb3 24 l:l.f4 'ii'e 6
1 7 l:l.e 1 i.. b 5?!
1 7 ...b2! was once thought to be refuted
and · thus replaced with the text move, but
current trends go in the opposite direction.
17 ... b2 is considered in the next game.
1 8 axb3 i..x c6 1 9 i..x c6 l:l.xc6 20 :xa 7
i..f 6
20 .. Jhc3? 21 tiJbl ! and White has a clear
advantage, e.g. 21 ....l:k7 22 !hc7 'i'xc7 23
.l:r.xe7 .l:r.d8 24 'i'd5 'i'c1+ 25 'iii>g2 .l:r.f8 26
:Xd7 'i'xb1 27 .l:r.d6 and Black has a horrible
ending to face which, in practice, is drawn
very rarely.
21 lll c4! i..x c3 25 lldxf7!
21....l:r.xc4? 22 bxc4 �xc3 23 .l:r.e3 and Picking up a pawn.
White wins material. The best chance lies in 25 . . . :es
21...tiJc51 22 'i'xd8 .l:r.xd8 23 �xf6 gxf6 The sacrifice could never be accepted:
(23 ... .l:r.xf6 24 b4 t2Je6 25 t2Jxb6 gave White a 25 ... l:r.xfl 26 'i'd8+ and 25 ...'i'xfl 26 l:hf7
clear advantage in Karpov-Beliavsky, Linares l:r.xf7 27 'i'd8+ l:r.f8 28 'i'd5+ are both
1 993) 24 t2Je3!, an improvement over a pre straightforward winning lines.
vious game. Van Wely-Karpov, Cap d'Agde 26 'i'f3 i..f 6 27 l:l.b7 h6 28 c,i;lg2 �hS 29
1 996, went 24...ttJxb3 25 tiJf5 .l:r.c1 26 :Xc1 h4 c,i;>gS 30 �h2 �hS 31 'i'h5 l:l.dS 32
liJxc1 27 .l:r.b7 b5 28 'iii>g2 h5 29 :Xb5, with a l:l.f7 i.. d 4 33 llfS+ .::l.xfS 34 l:l.xfS + c,i;lh7
promising ending for White. 35 'i'f3 i.. c 5 36 l:l.f5 J::.c S 37 h5 .:I.dB 38
22 l:l.xd7 'i'f6 l:l.e5 'i'd7 39 'i'e4+ �hS 40 �g2 l:l.fS 41
22 ...'i'c8 23 .l:r.e3 �f6 24 'i'd5 b5 25 t2Ja3 f4 l:l.dS 42 'i'f3 i..d 4 43 l:l.e2 b5 44 llld 2
b4 26 l2Jc4 h6 27 'iii>g2 .l:r.c7? 28 tiJb6! and i.. b 6 45 llle4 'ii'd 1 ? ! 46 lllf 2 'i'b 1 47 lllg 4
White was clearly better in Khalifman-Van 'i'f5 48 llle 5 �g8? 49 lllc 6 1 -0
der Wiel, Ter Apel 1 993. Instead 27 ... .l:r.c5 28
'i'b7 'i'xb7+ 29 .l:r.xb7 �c3, with an endgame Game 55
with no other prospects than a fight for a Gheorghiu -Cserna
draw, was the only chance. Berlin 1986
23 l:l.e4 'i'f5?
Karpov and his seconds prepared this po 1 d4 lllf 6 2 lllf 3 e6 3 c4 b6 4 g3 i.. a 6 5
sition before the game and none of them saw b3 i.. b4 + 6 i.. d 2 i.. e 7 7 i.. g 2 c6 8 i.. c 3
the coming combination. The most amazing d5 9 llle 5 lllfd7 1 0 lllx d7 lllx d7 1 1 llld 2
thing about this oversight is that a similar 0-0 1 2 0-0 l1c8 1 3 e4 c5 14 exd5 exd5
combination had already been played in the 1 5 dxc5 dxc4 1 6 c6 cxb3 1 7 l:l.e 1 b2!
following game: 23 ... .l:r.e6 24 .l:r.f4 'i'g6? Once refuted, now the drawing line, kill
(24...'i'xf4 25 gxf4 .l:r.e1+ 26 'i'xe1 �xe1 27 ing all life in the position.
1 24
4 . . . .i. a 6 : Main L in e with 5 b 3
1 8 .i.xb2 lDc5 according to Khalifman.

c2) 22 l:.e2? �xb2 23 l:.xb2 .tf6 wins for
Black.
c3) 22 i.e5! �xe1 23 :Xe1 .tf6! 24 .txf6
(Komarov gives 24 .th3? 'ii'd2 25 'iif fl l:.ce8)
24...'i'xf6 25 c7 and a draw was soon agreed
in Georgiev-Komarov, Yugoslavia 1 997.
1 9 . . . .i.d6?
19 ... .tb5?! 20 �b3 .tf6 21 �xc5 bxc5
was Gheorghiu's recommendation, but after
22 l:.c1 ! 'i'xd1 23 .l:exd1 .txc6 24 .txc5
White seems to have some chances. The a
pawn is probably falling.
1 9 ... .tf6! 20 l:.ct .tg5 21 l:.c2 i.b5 22
18 ... �f6 19 �b3 .tb4 20 l:.e3 'i'�d1+ 21 �b3 'ii'xd1 23 .l:xd1 .txc6 24 �xc5 .txg2
.l:xd1 l:.fd8 22 .l:d4! gives White a clear ad 25 rJ:;xg2 .te7 26 .l:dc1 .l:tfe8 27 .l:tc3 bxc5 28
vantage due to the pawn on c6. .txc5 .txc5 29 .l:xc5 l:.xc5 30 .J:.xc5 g6 31
1 9 .i.a3 l:r.c7 l:r.e2 32 l:xa7 h5 33 'iif f3 l:r.c2 34 h4 r/;g7
1bis was the refution of the ... b2 line, but 35 l:r.a6 was the course of Dautov-Alterman,
improvements have been found. Others: Germany 1 998. 1bis endgame is a theoretical
a) 19 �b3!? �d3 20 l:e2 .tb4! 21 .td4 draw. Black can even lose a tempo once the
.tc5! was a draw in Gelfand-Anand, China a-pawn starts rolling and still make the draw.
2000. The reason is 22 �xc5 (22 .tc3? �f4! However, many players find it difficult to
23 .l:d2 �e2+ 24 'itih1 'i'e7! 25 .tb2 .txf2 is draw such endings in practice. One example
enormous for Black) 22... bxc5 23 .tc3 .tb5! is Georgiev, who lost it to Akopian in Las
24 .l:e3 c4 25 'i'g4 g6 26 a4 i.xc6 27 'i'xc4 Vegas 1 999. That game can be found in my
.txg2 28 'i'xd3 with equality. Here 21...�c5? book Excelling at Chess, page 1 04. Alterman
22 .l:c2! .td3 23 �xc5! .txc2 24 1i'g4 gave also lost this endgame, but those intending to
White a strong position in Dautov-Sax, play this line should look this ending up and
Germany 1 998. then they would find that there is nothing to
b) 19 1i'g4?! .tf6 20 .txf6 'i'xf6 gave fear. 20 .txc5 bxc5 21 l:r.c1 .tb5 22 l:.xc5
White nothing but problems in Timo was agreed drawn in Shneider-Anand, Bastia
schenko-Bagaturov, Enakievo 1 997. 2000. Taking the difference of strength of
c) 1 9 �c4 .txc4 (White is better after the two players into account, one might sus
both 1 9 ... �d3 20 .l:xe71 .txc4 21 .l:d7 �xb2 pect that Black was not altogether confident
22 .l:xd8 �xd1 23 l:.dxd1!, and 1 9 ... .tf6? 20 about his position, but I think that if Anand
'i'xd8 .l:fxd8 21 .txf6 gxf6 22 �e3) 20 'ii'g4 really wanted to win against Shneider, he
.tg5! 21 'ii'xc4 �d3 with a further branch: would have chosen a less forcing system.
cl) 22 .tc3 �xe1 23 l:.xe1 .l:e8 24 .l:xe8+ 20 lDb3 lDd3
(24 .l:b1 .l:e6 25 h4 .tf6 26 .tb4 'i'd4 Barus 20 ...�xb3 21 .txd6 �xa1 22 .txf8 and
Shneider, Jakarta 1 997 and Black has a clear White can hope for a considerable advantage.
advantage according to Khalifman) 2 1 .i.xd6 'i'xd6 22 lDc 1 ! J:fd8 23 lDxd3
24 ...'i'xe8 25 'ii'g4 .th6, planning to return .ixd3 24 1i'a4 a5 25 J:ad 1 'i'c7 26 .i.h3
the exchange, gives Black a better game b5 27 1i'h4 1 -0
1 25
Summary
After 5 b3 dS 6 i.g2, game 45 shows that 6 ... dxc4?! is past its sell-by date. White has found the
right path to a small but very clear advantage. After S ... bS!? White will always be a little better,
as his pawn structure is slightly superior. Still this line has brought many players good practical
results, and is good for winning attempts, as the position easily becomes unbalanced. Also
interesting is 5 ... i.b4!? with the idea to exchange on d2. As far as I can see this line is theoreti
cally okay, even though the practically results indicate that it is easier to play the White side.
Both 7 ... aS!? and 7 ... cS!? are probably very slightly worse for Black, but again they give posi
tions where there remains much play, and where the better player most often succeed. The
long lines with 5 ...i.b4+ and 6 ... i.e7 seems to be the main reason for Black to play 4...i.a6.
There are good reasons for this: If White plays the tough lines, Black can liquidate the game to
a draw, and if not, Black is probably fine all the same.
1 d4 liJf6 2 c4 e6 3 liJf3 b6 4 g3 ..ta6 5 b3 (DJ i.. b4 +

5 ... dS 6 i.g2 dxc4 - Game 45
5 ... bS - Game 46
5 ... i.b7 6 i.g2 i.b4+ 7 i.d2 (DJ
7 ... aS 8 0-0 0-0
- 9 i..gS Game 48
9 ltJc3
. - 9 . . dS Game 49; 9 . d6 - Game 50
. .
7 ... cS 8 0-0 0-0

- 9 i.. c 3 Game 51; 9 i.xb4 cxb4 10 'it'd3 Game 52
-
6 ..td2 ..te7
6 ... i.xd2+ - Game 47
7 i.. g 2 c6 8 ..tc3
8 0-0 dS 9 'ii' c2 - Game 53
8 . . . d 5 9 ltJe5 ltJfd7 1 0 ltJxd7 liJxd7 1 1 ltJd2 0-0 1 2 0-0 .l:c8 1 3 e4 c5 1 4 exd5 exd5
1 5 dxc5 dxc4 1 6 c6 cxb3 1 7 :e 1 (DJ
1 7 ... i.. b S - Game 54; 1 7 ... b2 - Game 55
5 b3 7 ..td2 1 7 .l:.e T
1 26
CHAPTER EIGHT I
4 g 3 i.b7
1 d 4 lbf6 2 c 4 e 6 3 lt:if3 b6 4 g 3 .i.b7 al) 6 ... c5 7 J..xb4 cxb4 8 0-0 0-0 9 tLlbd2
4 ... J..b 7 is the old classical move, which d6 10 l:te1 'ilc7 1 1 e4 e5 is very similar to
was much played in the days when few Game 52, except that White has not played
people believed in such modem moves as b2-b3. This is an advantage that secures
4... J..a 6. In this chapter we will consider all White the better prospects in this position.
the classical lines as well as some of the 12 'ilb3 a5 13 'ile3 tLlc6 and now, instead of
sidelines. As you will find out, if you choose 14 d5, which was seen in Gavrilov-Seferjan,
to play anything other than 5 J.. g2 J.. e 7 6 Rjazan 1 993, Gavrilov recommends 14 h3
tLlc3 then there is really no way to play for ltfe8 1 5 d5 tLlbB 1 6 tLlh4 J..c 8 1 7 @h2 tLla6
an advantage, or even create an interesting 1 8 f4 'ilc5 1 9 'ile2 with a slight pull.
game. All these d5, b3 and l:e1 lines are not a2) 6 ... a5 7 0-0 0-0 8 i.f4 J.. e7 9 tLlc3 tLle4
dangerous for Black, once they have been · 10 'ilc2 tLlxc3 1 1 bxc3 l:a7 12 e4 d6 13 l:tfe1
looked at under the microscope. tLld7 14 l:tad1 'ile8 with a modest advantage
So the real battleground is in the main to White in Piket-Korchnoi, Nijmegen 1 993,
lines and here it is a matter of taste. All lines after 15 a4.
should lead to an even game. But what kind a3) 6 ... J..e7 7 tLlc3 0-0 (7 . . c6 8 e4 d5 9
.
of even game suits you best? Take a look cxd5 cxd5 10 e5 tLle4 1 1 0-0 0-0 12 lte1
and then decide. tLlxd2 13 'ilxd2 tLlc6 14 J.. f1 favoured White
in Savchenko-Hulak, Portoroz 1 996) 8 0-0
Game 56 tLla6 9 l:tc1 lte8 10 l:e1 d5 1 1 cxd5 exd5 12
Karpov-Gavrikov J.. f4 c5 1 3 'ila4 tLle4 1 4 lted1 with another
USSR 1988 edge in Filippov-Kalinin, Moscow 2002.
a4) 6 ...'ile7 7 0-0 i.xd2 8 'ilxd2 0-0 9
1 lt:if3 lt:if6 2 c4 b6 3 d4 e6 4 g3 .i.b7 5 tLlc3 d5 10 tLle5 tLla6 1 1 l:tfd1 , Ilincic
J.g2 c5 DamljanO'vic, Niksic 1 997, and we have yet
This move is more or less refuted (this another of these typical positions where
game). Other alternatives are: White is only very slightly better.
a) 5 ... J.. b4+. The difference with the b) 5 ... g6?1 6 0-0 J..g7 7 'ilc2 0-0 8 tLlc3 d6
Bogo-Indian is hard to find after this. After 6 9 e4 tLlbd7 1 0 b3 e5 1 1 i.b2 exd4 1 2 tLlxd4
i.d2 Black has four possibilities: l:e8 13 l:tfe1 simply led to a clear plus for
127
White in Milev-Simagin, Russia 1 959. The White will soon regain the pawn, after
weakening of the light squares on the queen which the superiority of both his bishop and
side does not work well within such a King's his pawn structure will guarantee an advan
Indian structure. tage.
6 d5! 1 7 . . .fl.ab8 1 8 Wxc5 .l:l.b6?!
1 8 ... .l:lb5 is better. This slow play presents
White with an opportunity to complete de
velopment without any concerns.
1 9 fl.ad 1 ltlb8 20 i.d5!
Now the game is more or less decided.
The rest, as they say, is a matter of tech
nique ...
20 ... 'it'b5 2 1 'flc7 'fla6 22 fl.c1 'it'a5 23
fl.fd 1 fl.b5 24 'ifd6 'flb6 25 'fle7 'it'g6 26
i.e4 'ii'e 6 27 i.h7 + �xh7 28 'flxf8 'flxa2
29 'it'd6 a6 30 'ii'd 3+ f5 31 fl.b1 'ii'e 6 32
fl.xb5 axb5 33 'flxb5 ltlc6 34 fl.d5 �g6
35 'it'c5 'fle4 36 fl.d6+ �h7 37 'it'd5
White asswnes control of the centre. Ei 'ii' b 1 + 38 'ii'd 1 'ii'e4 39 'ifd3 'it'g4 40 fl.d5
ther Black will have to live with an undesir 1 -0
able structure or fight for d5. As this game
shows, White is 'fitter' for this fight. Game 57
6 . . . exd5 7 ltlh4 b5 8 0-0! Polugaevsky-Wojtkiewicz
8 cxd5 merely justifies Black's play this far. Haninge 1990
8 . . . bxc4 9 ltlc3 i.e7 1 0 ltlf5!
1 0 i.. g5?! is inferior. Now White clearly 1 d4 ltlf6 2 ltlf3 e6 3 c4 b6 4 g3 i.b 7 5
has the advantage. i.g2 i.e7 6 0-0 0-0 7 b3 d5
1 0 . . . 0-0 1 1 ltlxe7 + ! 'it'xe7 1 2 i.g5 h6
1 2...ll:la6?! 13 lllxd5 i..xd5 1 4 i..xd5 .l:lab8
1 5 Lc4 secures White a very clear advan
tage. The knight is badly placed on a6.
1 3 i.xf6 'it'xf6 14 ltlxd5 i.xd5 1 5 'it'xd5
ltlc6 1 6 'it'xc4 'ii'x b2 1 7 e3
8 cxd5 exd5
Black has an easier way to equalise in
8 ... ll:lxd5! 9 i..b 2 ll:ld7 (9 ... c5 is a good
alternative: 10 lll c3 cxd4 1 1 lllxd5 i..xd5 1 2
lllxd4 i..xg2 13 �xg2 i.. f6 1 4 e3 i..xo4 1 5
i..xd4 gave White some advantage in Aa-
1 28
4 g3 �b 7
gaard-Christensen, Copenhagen 1 999, but 45 l:!.xa7 .i.f1 + 46 'it>f2 .l:lxd6 47 'itxf1

simply 14 ...�ds+! would have kept the game .:l.b6 48 a5 .l:txb3 49 a6 1 -0
drawish) 1 0 lDbd2!? c5 1 1 .l:r.cl .l:r.c8 1 2 lDc4
ltJ5f6 with equality in Dautov-van der Ster Game 58
ren, Ter Apel 1 994. Freitag-O ral
9 .i.b2 Graz 1998
9 lDc3 lDa6 10 ii.b2 �c8 11 .l:r.cl .l:r.d8 12
a3 c5 13 dxc5 bxc5 14 b4 �d7 1 5 bxc5 1 d4 lt:lf6 2 c4 e6 3 lt:lt3 b6 4 g3 .i.b 7 5
lDxc5 1 6 lDe5 �e6 17 lDa4 ltJfe4 1 8 lDxc5 .i.g2 .i.e7 6 0-0 0-0 7 d5!?
ii.xc5 19 �a4 ii.b6 20 4:)£3 .l:r.ac8 led to
equality in Nikolic-Khalifman, Manila 1 990.
9 . . .:ea 1 O lt:lc3 lt:la6
10 ... �c8, with equality, is mentioned
above.
1 1 lt:le5 c5 1 2 .l:lc 1 �f8
12 ... cxd4 1 3 lDb5! favours White.
1 3 e3 lt:lc7 1 4 lt:la4!?
This pawn sacrifice had its heyday in the

1 980's. These days it is not considered to be
,
very dangerous but it must be remembered
that it is more messy than the main lines.
7 . . . exd5
Also possible is 7 ...lDa6!?, as tried in Haus
rath-Tiviakov, Zwolle 2000. I suppose this
was employed as a means to steer the lower
White has emerged from the opening with rated opponent out of his 'book' ... 8 lDc3
a small plus. ii.b4 9 lDe5 ii.xc3 10 bxc3 lDc5 1 1 ii.a3 �e7
1 4 . . . lt:le6 1 5 dxc5 bxc5 1 6 lt:lg4! lt:le4? ! 12 lDd3 d6 13 lDxc5 dxc5 14 e4 e5 1 5 .l:r.e 1
1 6 ...lDxg4! 1 7 �xg4 �e7 1 8 .l:r.fdl .l:r.ed8 lDe8 1 6 ii.fl ltJd6 and Black was doing well.
was better according to Polugaevsky.
1 7 f3 lt:ld6 1 8 f4! lt:le4 8 lt:lh4!
1 8 ... d4!? was apparently necessary. This was Polugaevsky's discovery after the
1 9 lt:lf2! lt:lt6 20 f5 lt:lg5 21 lt:lxc5 .i.xc5 following game: 8 lDd4?! ii.c6! 9 cxd5 ii.xd5
22 .:1.xc5 lt:lge4 23 lt:lxe4 lt:lxe4 24 'ifd4 10 ii.xd5 lDxd5 1 1 e4 ltJb4 1 2 lDc3 ii.f6 1 3
f6 25 .:l.c2 'iVd6 26 .l:lfc 1 .l:le7 27 'iVa4 lD 5 .l:r.e8 14 f4 d 6 1 5 �g4 lD8c6 1 6 e 5 dxeS
'iVd8 28 �a3 .l:lf7 29 .l:ld 1 'ifb6 30 'ifd4 17 ltJe4 exf4 1 8 ltJh6+ �fB 1 9 lDxf6 �xf6
'ifa5 31 .i.b4 'iVb5 32 �xe4 ! dxe4 33 20 .l:r.xf4 .l:r.el+ 21 �g2 lDe5 22 .l:r.xf6 lDxg4
.l:lc7! ! .l:lxc7 34 'iVd8 + 'ife8 35 'ifxc7 .i.c6 23 :Xf7+ �e8 24 :Xg7 lDxh6 25 a3 .l:r.d8 26
36 .:ld6 .i.b5 37 a4 :ea 38 'iVa5 .i.d3 39 axb4 .l:r.d7 27 .l:r.xd7 �xd7 and the endgame is
'iVd5 + 'it>h8 40 .:l.d7 h6 41 g4! 'ife5 42 a disaster for White, as was seen in-the game
'ifxe5 fxe5 43 .i.d6 .l:lc 1 + 44 'itg2! .l:ld 1 Polugaevsky-Korchnoi, Buenos Aires 1 980.
1 29
8 . . . c6! compensation for the pawn, but Black has

The only good move. After 8 ... .!lle4 9 his chances too and did okay.
cxd5 .i.xh4 10 .i.xe4 .i.f6 1 1 'i'c2 White was White had previously played 1 3 'iWg4 i.f6 14
slightly better in Gheorghiu-Unzicker, Wijk .!llc3 i.c8 1 5 .i.xe4, but after 1 5 ... .i.xf5
aan Zee 1 981 . 8 ... .!lle 8 9 lll f5 llld6 1 0 .!llxe7+ (1 5 ....!lld 5? 1 6 .th6 g6 17 .i.xf8 <ittx f8 1 8
'iii'xe7 1 1 cxd5 c5 12 .!llc3 .!lla6 13 .i.f4 fa .i.xd5 .i.xf5 1 9 'i'b4+ .i.e7 20 'iii'b3 cxd5 21
voured White in Barlov-lvanovic, Yugoslavia .!llxd5 .i.e6 and Black resigned in Barlov
1 984. Abramovic, Yugoslavia 1 994 - simply 22
9 cxd5 �xd5 :Xe6 fxe6 23 'ii' f.3+ followed by .!llxe7 wins)
Also good is 9 ... cxd5!? 10 .!ll c3 .!lla6 1 1 1 6 'i'xf5 g6 17 'i'f.3 .!ll e6 1 8 i.h6 .!ll d4 19
.!ll f5 .!llc7 12 .tf4 :e8 13 .i.d6 i.£8 1 4 'iii'b 3 'ii'g2 l:le8 20 .i.e3 a6 21 .!lle2 l:.a7 22 lllxd4
.i.c6 1 5 i.xf8 :xf8 1 6 l:lad 1 :e8 17 l:lfe 1 .i.xd4 23 .i.xd4 'ii'xd4 24 .i.xc6 :Xel+ 25
l:le5, with equality in Romanishin-Razuvaev, l:lxel .!llxc6 26 'iii'xc6 <Jl;g7 27 b3 there was
Jurmala 1 987. In fact I think this is probably little life left in the position in Pigusov
the safest line for Black. On the other hand, Tiviakov, Beijing 1 997. Of course a man like
it does not offer much in the way of winning Pigusov is well known for his love for ex
chances. changes. Like my first trainer, who said that
1 0 �f5 �c7 the sacrifice of a pawn would only make you
try to win it back, so therefore it was a waste
of time ...
1 1 . . .d 5
1 1 ...lll e 8 12 .i.f4 .!ll a6 1 3 'i'd2 d5 14 e4
.!llac7 1 5 l:r.ad1 i.f6 1 6 exd5 lllxd5 1 7 lllxd5
cxd5 1 8 llle3 .!llc7 19 i.xc7 'ii'xc7 was al
ready drawn in Timman-Karpov, Tilburg
1 983.
1 2 e4 lle8! ?
The new move. Previously known was
1 2 ... .i.f6 13 :et i.c8 14 'ii'f.3 .i.e6 1 5 g4
.!lld7 16 'i'g3 dxe4 1 7 .i.xe4 with compensa
tion but no more, Romanishin-Epishin, Bel
Or 10 ...i.f6!? 1 1 e4 .!lle7 1 2 .!lld 6 .i. a6 1 3 grade 1 988.
l:.e1 i.e5 14 f4 .i.xd6 1 5 'iii'xd6 c 5 1 6 b4 13 .i.f4 �e6 14 'i'g4 �d 7
cxb4 1 7 'iii'xb4 .!llbc6 1 8 'i'a4 'i'c8 with un
clear play in Milos-Sunye Neto, Sao Paulo
2000. Actually I would prefer to be Black in
this line, and in the game, too, but I know
that a creative player will also be able to get
something out of White's position.
1 1 �c3
One of the two main lines. 1 1 e4 d5 1 2
lte t l i s Barlov's variation, which appears to
lead to an unclear but equal game: 1 2... dxe4
13 lllc 31? i.c8 14 .!llxe7+ 'i'xe7 1 5 .i.xe4
.i.e6 1 6 .tf4 :d8 1 7 'i'h5 h6 1 8 :e3 was
Piket-Van der Wiel, Rotterdam 1 997, with
1 30
4 g 3 j,, b 7 .
1 5 e5 jahov-Tiviakov, St. Petersburg 1 998. White

This looks like a positional error, and now has not been able to establish the powerful
Black does not have to worry about his cen centre that secures him play in Benoni struc
tre. tures. Necessary here was 1 3 .ie3 'itb7+!? 1 4
1 5 . . . g6 1 6 lll h 6+ 'ot?g7 1 7 tllt 5+ �g8 1 8 ttlf3!? cxd4 1 5 .ixd4 with complete equality.
tll h 6+ 'ot?f8 1 9 :te 1 j,,g 5 20 h4 j,,xf4 2 1 10 ttle5 has also been played, Bator
gxf4 f5 22 exf6 tllx f6 2 3 'iff 3 j,, c 8 24 Stefansson, Mariehamn 1 997, continuing 10
:e5 �g7 25 :ae1 �xh6 26 f5 gxf5 27 ttle5 hg2 11 c;l.>xg2 'Wc8!? 1 2 e4 1i'b 7 (al
:xt5 �g7 28 :e3 a5 29 'ifg3+ �h8 30 ready equal) 1 3 ttlxd7 :fd8 1 4 ttleS .if6 1 5
'ife5 tllg 7 0-1 .if4 .ixe5 1 6 .ixe5 f6 1 7 .if4 e 5 1 8 .id2
--------.. .l:.xd4 and Black had improved his prospects.
Game 59 10 d5!? is another try. Then the safe
Van Wely-Korchnoi 1 0 ... exd51? 11 cxd5 c5 1 2 dxc6 dxc6 1 3 ttle5
Hoogeveen 200 1 .ixg2 14 �xg2 'Wc8 leads to approximate
equality, while the ambitious 1 o ... .if6 1 1
1 d 4 tllf 6 2 c4 e 6 3 tllf 3 b 6 4 g 3 j,, b 7 5 !LJd4 hg2 1 2 c;l.>xg2 leaves Black with a
j,,g 2 j,,9 7 6 0-0 0-0 7 :e 1 modest lead in development, meaning he
This line was very popular in the mid- must act quickly in order to avoid e2-e4-e5.
1 990's, when Kramnik championed it. It is Delchev-Marin, Istanbul 2000 went 12 ... c5
now known to be harmless in view of... 1 3 ttlb5 exd5 1 4 'Wxd5 ttlb4 15 'Wdl a6 16
7 . . . tlla 6! ttld6 b5 1 7 e4 bxc4 1 8 e5 ttld3! 19 exf6 'ii'xf6
20 .if4 g5 21 'Wg4 ttlxf4+ 22 gxf4 'Wxd6 23
'ii'xgs+ 'Wg6 with a draw.
b) 8 a3!? is an interesting try. White is aim
ing for the standard position and invests a
tempo to do so. 8 ... d5 (8 ... c5 9 d5 exd5 10
ttlh4 ttlc7 11 ttlc3 d4 1 2 .ixb7 .l:.b8 1 3 .ig2
dxc3 1 4 bxc3 leaves all the white pieces well
placed) 9 cxd5 exd5 10 ttlc3 c5 1 1 .if4 ttlc7
12 l:tct ttle6 13 .ie5 ttlg4 was seen in
Wojtkiewicz-Zagrebelny, Dhaka 1 999. In
many ways this fits in with White's plan to
increase pressure on the d5-pawn. White was
willing to surrender the bishop for the knight
It is this move that has more or less done and now it is Black who expends energy to
away with any modem interest in 7 :et . arrange this. 13 ...'ii'd 71? is a sensible alterna
Instead 7 ... dS 8 cxdS exdS 9 ttlc3 might give tive, preparing to put more weight on the d
White a little plus. file. After (1 3 ...ttlg4) 14 dxc5 bxc5 1 5 h3
8 tlle 5 ttlxeS 1 6 ttlxeS i.f6?! 1 7 ttlg4 .ixc3 1 8 bxc3!
Other attempts at an advantage are: White had a slight but distinct advantage, but
a) 8 ttlc3 ttle4 9 ttlxe4 (9 i.d2 dS 10 cxdS the improvement 1 6 ... ttlc7!?, keeping the two
exdS was equal in Barsov-Llgterink, Hoogev bishops, would have given Black a fine game.
een 1 997) 9 ... .ixe4 1 0 'ii'a4 'Wc8 1 1 ttle5 1 2 .ixc7 'Wxc7 1 3 :et 'ii'd7 1 4 lileS 'We6
.ixg2 12 �xg2 cS! 13 dS?! exdS 14 cxdS 1 5 ttld3 c4 1 6 ttleS .l:.ab8 17 e4 dxe4 1 8
ttlb4! 1 5 .l:.dl ! d6 1 6 ttlf3 'itb7!? 1 7 a3 bS 1 8 ttlxe4 .ixe4• 1 9 .ixe4 ttlxe4 20 l:txe4 'ii'dS 2 1
'itb3 ttla6 with a good game for Black, Polul- 'ii'e2 f5 22 .l:.e3 b S wa s the course o f Ibragi-
131
mov-Pushkov, Elista 2001 , with a level game. :d1 + c,t>t8 35 llld s i.h6 % - %
Tb.is shows that the idea is just to get a posi
tion and then play, rather than attempt to Game 60
refute Black's plans. Fom inyh-Chernyshov
8 . . . d5! Smolensk 2000
The improvement over 8 ... .ixg2. The dif
ference with 7 ... dS is that White's knight is 1 d4 li:lf6 2 c4 e6 3 lllf3 b6 4 g3 i.b 7 5
not very well placed on eS so early, so Black i.g2 i.e7 6 0-0 0-0 7 ll:\c3 ll:\a6 ! ?
has no problems in equalising.
9 cxd5 exd5 1 0 lll c 3 c5 1 1 i.f4 llle4 1 2
lllxe4 dxe4 1 3 dxc5 ll:\xc5 1 4 1i'c2 1i'c8
An interesting sideline. Tb.is illustrates

how good this is against 7 l:tel . When it is
good here, it must be brilliant there! 7 ... dS 8
Black has equalised. The ES-square is a llle S lll a6 9 .if4 transposes to 8 .if4 in the
brilliant outpost for the queen, as is e6. note to White's 8th move.
1 5 :ed 1 g5 1 6 i.e3 1i'e6 1 7 ll:\c4 :ac8 8 a3
1 8 b3 h6 1 9 i.d4 i.d5 20 i.b2 :td8 2 1 Others:
:ac 1 b5 2 2 1i'c3 f 6 2 3 llle 3 a) 8 b3 l:te8 9 i.b2 dS 10 lll eS cS 1 1 e3
'iic 7 12 cxdS exdS 13 lZ:lbS 'iic 8 14 a3 lllc7
led to a balanced game in Zaremba
lbragimov, Chicago 2002.
b) 8 �f4 d6! (8 ... dS 9 lZ:leS cS 10 cxdS
exdS 1 1 l:tcl lll e4 12 lllxe4 dxe4 13 dxcS
lZ:lxcS 14 lll c4 lll e6 1 5 �eS f6 16 llld6 �xd6
17 .ixd6 gave White a little something in
Paunovic-Tiviakov, Cacak 1 996) 9 a3 cS 10
dS lllc 7 11 e4 a6 1 2 b4 exdS 1 3 exdS cxb4 14
axb4 bS with a complicated game with
chances for both sides in Ribli-Markowski,
Moscow 1 994.
c) 8 'iia4 'iic 8! (protecting the knight on
23 . . . lll x b3 24 Wxc8 :xc8 25 :xc8+ a6 just once more) 9 J:tdl lll e4 1 0 lllxe4
Wxc8 26 :xd5 lll c 5 27 lllf 5 i.f8 28 i.h3 .ixe4 1 1 .if4 d6 1 2 l:tacl 'iib 7 1 3 llle l
c,t>h7 29 c,t>g2 h5 30 llle 3 Wc6 3 1 i.f5 + �xg2 1 4 lllxg2 cS with equality in Tatai
c,t>g7 32 i.d4 c,t>f7 3 3 i.xc5 i.xc5 34 Tiviakov, Arco 1 998.
1 32
4 g3 �b 7
d) 8 .i.g5 c5 9 d5 exd5 10 cxd5 d6 1 1 liJh4 30 f4 �xd4+ 3 1 �f1 :aa 1 32 �xd5+

1i'd7 12 1i'd2 b5 13 e4 b4 14 liJd1 .i.d8, �8 33 :d 1 �c3 34 �e2 :xc 1 35 :xc 1
Shneider-Tiviakov, Moscow 1 992, was very :xc 1 36 �d3 �b2 37 :b6 :d 1 + 38
messy. �c4 ltia3+ 0-1
8 . . . ltie4 9 �d2 f5 1 0 b4 �f6 1 1 :b 1
1 1 l:r.c1 c5 1 2 dxc5 bxc5 1 3 b5 liJc7 does Game 61
not look dangerous for Black. Andersson -Karpov
1 1 . . .ltixc3 Ti/burg 1983
1 1 ...c5 12 .bxc5 liJaxc51 13 dxc5 liJxc3 1 4
.i.xc3 .i.xc3 1 5 1i'd 6 .i.c6 i s another route to 1 ltif3 ltif6 2 c4 b6 3 g3 �b7 4 �g2 e6
equality. 5 0-0 �e7 6 d4 0-0 7 ltic3 ltie4 8 'i'c2
1 2 �xc3 �e4 1 3 :b3 This line is so classical that many players
still play it with some odd kind of respect for
the old masters. The newer and more dan
gerous 8 .i.d2 is treated in the following
games.
8 liJxe4 .i.xe4 9 liJh4 is the ultimate draw
ing line. After 9 ... .i.xg2 1 0 liJxg2 d5 1 1 'ii'a4
dxc4 1 2 1i'xc4 c5 13 .i.e3 cxd4 . 1 4 J.xd4
1i'c8 the position is completely equal. Apart
from a blunder once deciding a game be
tween Polugaevsky and Andersson (White), I
cannot recall this line being dangerous for
anything other than one's interest in chess.
Yet this characteristic is precisely the reason
1 3 . . . c6! why some people play it. By somehow reach
Part of a magnificent but not too complex ing a technical position where they cannot
plan. Black wants to establish a giant knight create any winning chances they leave the
on d5. floor free, as it were, for their opponents to
1 4 �b2 create losing chances. Andersson-Karpov,
14 b5?! cxb5 15 cxb5 liJc7 favours Black. Reggio Emilia 1 991 continued 1 5 l:.fc1 Wxc4
1 4 . . . ltic7 1 5 �h3 1 6 l:lxc4 liJa6 1 7 liJf4 :fd8 1 8 .i.e3 l:.d7 1 9
15 a4 b5 16 axb5 cxb5 17 c5 liJd5 18 e3 liJd3 f6 2 0 l:lac1 l:lad8 21 <iltg2 <J;; fl with
a5! and Black has a positional plus. equality.
1 5 . . . b5! 1 6 ltid2 bxc4 1 7 ltixc4 �d5 1 8 8 . . . ltixc3 9 'i'xc3 c5
'i'd3 'i'b8 1 9 �g2 This is the most drawish line here. Equally
19 £3!? Wb5 20 l:lc1 l:r.fb8 21 e4 .i.xc4 22 good is 9 ... £5, e.g. 10 d5 .i.f6 1 1 'ii'c2 liJa6 1 2
1i'xc4 fxe4 23 fxe4 a5! leads to equality. l:ldl 1i'e7 1 3 liJd4 liJc5 1 4 .i.e3 .i.xd4 1 5
1 9 . . .'i'b5 20 :c 1 :tb8 21 f3 a5! .i.xd4 d6 1 6 b4 liJd7 1 7 a4 a5 1 8 bxa5 l:lxa5
Black has the initiative. White was not able with equality in Krogius-Kholmov, Kiev
to counter quickly enough. 1 964. Meanwhile Andersson-Akesson,
22 e4 �xc4 23 :xc4 d 5 ! 24 exd5 exd5 Skelleftea 1 999 went 10 b3 .i.f6 1 1 .i.b2 d6
25 :c5 'i'xd3 26 :xd3 axb4 27 :xc6 1 2 l:ladl a5 13 liJe1 .i.xg2 14 liJxg2 liJc6 1 5
bxa3 28 �xa3 :b 1 + 29 �c 1 ltib5 'ii'd2 'ii'd7 1 6 d 5 liJd8 1 7 .i.xf6 l:lxf6 1 8 dxe6
Black's pieces are literally springing to life liJxe6 1-9 liJf4 liJxf4 20 'ii'xf4 l:le8, again with
with powerful effect. complete equality. It should be said such
1 33
positions can be seen a million o r s o times in

GM games, and some players, like Anders
son, continue to play them because they
know them so well.
1 0 l:ld 1
Black also has nothing to fear after the
line 1 0 b3 i.f6 1 1 i.b2 cxd4 1 2 liJxd4 i.xg2
1 3 @xg2 liJc6 14 'i'e3 'i'c8 1 5 l:tfdl l:d8, as
in the encounter Beliavsky-Adams, Belgrade
1 995.
1 o . . . d6 1 1 b3 .i.t6 1 2 .i.b2 tlld 7
Ibis is the most flexible choice, but an
other route to equality is well known, namely
1 2 ... 'i'c7 1 3 'i'd2 l:d8 14 dxc5 dxc5 1 5 'i'f4 .., . . . .i.f6!
liJa6 1 6 'i'xc7 ltJxc7 17 ltJe5 (1 7 i.xf6 gxf6 I believe this is the best way to deal with
1 8 liJel i.xg2 1 9 @xg2 @£8 20 l:td3 l:txd3 these lines. 7 ... 0-0 8 d5!? is considered in the
21 liJxd3 @e7 22 l:tdl liJe8 23 g4 was equal next game. Note that if 7 ... 0-0 8 0-0 then it is
in Hort-Karpov, Buenos Aires 1 980) interesting to play 8 ... d5!?, for if you want to
1 7 ... i.xg2 1 8 �xg2 i.xe5 1 9 i.xe5 liJe8 20 play such positions this is certainly the right
f4 f6 21 i.c3 @f7 22 g4 l:txdl 23 l:txdl �e7 place to push to d5. Actually these lines are
24 <t>£3 l:td8 with a draw to follow in Tim considered to give White a slight plus in
man-Karpov, Amsterdam 1 980. ECO, but I guess that is primarily because
1 3 'ifd2 'ile7 14 dxc5 .i.xb2 1 5 'ifxb2 there was originally a guy who tried a line and
dxc5 1 6 l:ld3 l:lad8 1 7 l:lad 1 secured a modest advantage and then he has
been quoted for it ever since. Perhaps White
is a little better, but then he always is in these
situations. Korchnoi-Polugaevsky, Moscow
1 972 continued 9 cxd5 exd5 10 l:tcl liJd7 1 1
.i.f4 c5 1 2 ltJxe4 dxe4 1 3 liJd2 £5 1 4 liJb3
i.d5 1 5 dxc5 with a draw.
8 ... d6 9 d5! liJxd2 1 0 liJxd2 e5 1 1 f4! exf4
12 gxf4 liJd7 1 3 liJ£3 i.f6 14 liJd4 l:te8 1 5 e4
gave White a substantial advantage in Va
ganian-Furman, USSR 1 971 .
8 . . . £5 is considered in the last game.
8 0-0
8 'i'c2 liJxd2 9 'i'xd2 looks like a loss of
1 7 . . . ill b 8 1 8 h4 l:lxd3 1 9 l:lxd3 l:ld8 20 time, and I suspect that that is what it is.
'ild2 l:lxd3 21 'ilxd3 Ya - Ya After 9 ... 0-0 10 0-0 d6 1 1 e4 (1 1 d5 i.xc3 1 2
.------ 'i'xc3 exd5 13 liJh4 d4! 1 4 'i'xd4 i.xg2 is
Game 62 equal, as pointed out by Makarichiev)
Pinter-Tompa 1 t .. .liJd7 1 2 .:r.ad1 g6 1 3 h4 i.g7 14 h5 'i'e7
Hungary 1993 15 .:r.fe1 l:tfd8 16 hxg6 hxg6 17 i.h3 c5 1 8
______________... d5 ltJe5 1 9 ltJxe5 i.xe5 Black had no prob-
1 d4 ltif6 2 c4 e6 3 illf3 b6 4 g3 .i.b7 5 lems whatsoever in Tukmakov-Nielsen, Co-
.i.g2 .i.e7 6 ltic3 � 7 .i.d2 penhagen 1 996.
1 34
4 g 3 i.. b l
8 .l:r.c1 has not been tested since Black 1 9 .l:r.a1 .l:r.e8 only Black had realistic pros
came up with 8 ... ..ixd4! 9 il)xd4 il)xc3 1 0 pects of pushing for a win in Vyzmanavin
..ixb7 il)xd1 1 1 .l:r.xd1 c6! 1 2 ..if4 0-0 13 Khalifman, USSR 1 990.
i.d6 .l:r.e8 1 4 ..ixa8 'ii'c 8 1 5 b4 il)a6 1 6 b5 1 3 . . . i.. e 5!
'ii'xa8 17 bxa6 c5 1 8 ll)£3 'ii'e4 with sufficient 1 3 ... ..ie7?! had been played quite often be
counterplay in Karpov-Salov, Rotterdam fore 13 ... 1'.e5 took over. Actually, I wonder
1 989, a game Black went on to win after a why? The bishop does not look better than
fierce struggle. the knight once it gets to e 7. (I would suspect
8 . . . 0-0 9 l'lc1 c5 that even 13 ... il)a6 is better than this) . White
9 ... d5 is a tempo too late. After 10 cxd5 earned himself a clear advantage in Karpov
exd5 1 1 ..if4 il)a6 1 2 ..ie5 .l:r.e8 1 3 ..ixf6 Salov, Linares 1 993: 14 f4 il)d7 1 5 g4! a6 1 6
'ii'xf6 14 e3 c5 1 5 il)e5 'ii'e 7 1 6 .l:r.e1 il)c7 a4 :e8 1 7 g5 ..i f8 1 8 ..tiih 1 ! etc. Shneider
White should not play 1 7 il)d3?! (Korchnoi Beliavsky, USSR 1 990, went 1 6 g5 b5 1 7
Salov, Belgrade 1 987) but 1 7 'ii'a4!? il)xc3 1 8 il)g3 .l:r.e8 1 8 h4 ..i f8 1 9 h 5 with the better
.l:r.xc3, with the idea o f 1 8. . .c4 1 9 l:xc4 b5 20 game for White. Let us compare the bishop
:Xc7 bxa4 21 :Xe7 .l:r.xe7 22 .l:r.c1 with a with the knight now -why run away from the
small advantage to White according to exchange like that?
Korchnoi. 1 4 'ilfd2
1 0 d5 exd5 1 1 cxd5 .!Lixd2 14 f4 ..id4+ 15 ..tiih l ..ia6 16 .l:r.£3 .l:r.e8
Also possible is 1 1 ...il)xc3!? 12 ..ixc3 d6 (1 6 ... 'ii'e7 is even better according to Almasi)
13 il)d2 .l:r.e8 14 il)e4 ..ie7!? 15 g4 b5 with a 17 g4 'ii'e 7 1 8 g5 il)d7 1 9 e3 ..ixc3 20 il)xc3
messy game in Rogers-Speelman, Oropesa b5 21 l:g3! £5! and Black was slightly better
del Mar 1 996. in the game Yusupov-Almasi, Germany
1 2 .!Llxd2 1 994. It should be noted that the line itself
1 2 'ii'xd2 .l:r.e8 1 3 h4 d6 14 il)g5 ..ia6 1 5 does not give Black the advantage - White
il)ge4 ..ie5 1 6 l:.fe1 g6 1 7 a4 'ii'e7 1 8 b 3 il)d7 should be able to improve on this, but the
resulted in stone-cold equality in Epishin positive trend seems to be with Black, none
Oms Pallise, Sevilla 2000. theless.
1 2 . . . d6 1 3 .!Lide4! 1 4 . . . i.. a 6 1 5 �h 1
1 5 .l:r.fe1 g6 1 6 f4 ..id4+ 1 7 e3 ..ig7 1 8 a3!?
'ii'e7 19 b4 il)d7 20 g4! h6 21 g5!? was un
clear in Brenninkmeijer-van der Weil, Hol
land 1991.
15 ... 'ilfe7 16 f4 i..d4 17 l'lfe 1
The most natural. After 17 .l:.£3 il)d7 1 8 e3
..ixc3 1 9 il)xc3 b5 20 ..ifl l:ab8 Black had
good counterplay in Horvath-Zagrebelny,
Budapest 1 993, while 1 7 b4 g6 1 8 b5 ..ic8 1 9
e 3 introduced an early draw in Yermolinsky
De Firmian, Seattle 2000.
1 7 . . . g6 1 8 e3 i..g 7 1 9 g4
1 9 a3 il)d7 20 b4 cxb4 21 axb4 ..ic4 22 il)e2
1bis is the only dangerous move. White .l:r.fc8 23 il)g5 ..ixe2 24 .l:r.xe2 a5 gave Black a
should remember how strong the bishop is better game in Van der Sterren-Yemelin,
on f6. After 1 3 il)c4 ..ia6 1 4 'ii'b3 ..ixc4 1 5 Hamburg 1 997.
'ii'xc4 il)d7 1 6 e 3 a6 1 7 a4 'ii'e7 1 8 'ii'e2 .l:r.fb8 1 9 . . . h6 20 g5 hxg5 2 1 .!Llxg 5
1 35
The game i s rather unclear. 7 . . ..i.b4

.
2 1 . . . llld 7 22 'ii'f 2 ..i.h6! 7 ... tiJa6 8 0-0 tlJc5 9 tlJe5 a5 10 �e3 exd5
22 ... tiJf6 23 �4 followed by e3-e4-e5 1 1 cxd5 �a6 1 2 'i'c2 �d6 13 f4 gave White
would be very dangerous for Black. the advantage in Atalik-Stoica, Herculane
23 h4 ..i.xg5 24 hxg5 f6! 25 gxf6 'ii'h 7+ 1 996.
26 �g 1 lllxf6 27 'ii'g 3 :tae8 2 8 ..i.h 3 �g7 8 0-0 ! ?
29 :tcd 1 :th8 30 ..i.g2 'ii'h 4 3 1 'ii'x h4 This line allows the exchange on c3, after
:txh4 32 a4? ! g5! 33 fxg5 c!lld 7 34 :ta1 which Black is okay. The critical lines arise
..i.d3 35 ltlb5 :tg4 36 ltlxd6 .:ta 37 e4 after 8 �d2! and now:
:txg5 38 ltlf5 + �6 39 :te3 ltle5 40 a) 8 ...Lc3 9 �xc3 exd5 1 0 tiJh4 tlJe4
:tg3! :tfg8 41 :txg5? :txg5 42 ltle3 ..i.xe4 (10 ... c6 1 1 tiJf'S dxc4 12 'i'd6 �h8 13 tlJxg7
43 .:n + �e7 44 :tf4 ..i.xg2 45 ltlxg2 c.fiixg7 14 'ii' f4 h6 1 5 g4 gave White a promis
llld 3 46 :te4+ �d6 47 �f 1 lllx b2 48 a5 ing attack in Halasz-Lendwai, Miskolc 1990)
:tf5+ 49 �g 1 :txd 5 50 a6 :td 1 + 51 �f2 1 1 cxd5 tLixc3 12 bxc3 d6 13 0-0 tiJd7 14 e4
.:a1 52 lllf4 :ta4 53 :te6 + �d7 0-1 tLic5 1 5 1i'd4 �a6 1 6 l:lfdl 1i'd7 1 7 f4 with a
· slight advantage to White in Karpov
Game 63 Ivanchuk, Linares 1 993.
Romanishin-Korchnoi b) 8 ... c6 9 0-0 �xc3 10 �xc3 (1 0 dxc6!?
Lvov 2000 dxc6 1 1 �xc3 tiJbd7 12 'i'c2 also looks bet
ter for White and the two bishops) 10 ... cxd5
1 d4 lllf 6 2 lllf3 e6 3 g3 b6 4 ..i.g2 ..i.b7 1 1 tiJh4 tlJe4 1 2 cxd5 tiJxc3 13 bxc3 'ii'c7 14
5 c4 ..i.e 7 6 ltlc3 0-0 c4 tLia6 1 5 l:lct with an advantage according
This move makes a lot of sense. Black to Ribli.
wants to see if White's move order holds any c) 8 ...tiJa6!? 9 0-0 tLic5 and now 10 t:De1
real threats to his wellbeing. l:le8 11 t:Dc2 �xc3 12 �xc3 exd5 13 cxd5
7 d5! t:Dce4 14 tbe3 t:Dxc3 1 5 bxc3 :Xe3! 1 6 fxe3
7 'ili'c2 is, as always, answered with 7 ... c5! d6 1 7 'ii'd4 tbd7 1 8 l:lf4 'ili'e7 19 .l:taf1 t:De5
and Black equalises with 8 d5 exd5 9 tlJg5 gave Black brilliant play for the exchange in
tLia6 1 0 cxd5 tiJb4 1 1 'ili'd2 tiJfxd5 1 2 tlJxd5 Razuvaev-Tiviakov, Rostov 1 993. It is hard
�xg5 1 3 'ili'xg5 1i'xg5 14 �xg5 �xd5 1 5 to find any use for the rooks as all the files
�xd5 tlJxd5 1 6 0-0-0 tiJb4 1 7 a3 tlJc6 1 8 are closed or semi-closed. An improvement
l:.xd7 .l:tfe8 1 9 �dl f6 20 �d2 l:lad8, a s in for White is 1 0 tbh4!, e.g. 1 0 ... l:le8 1 1 l:lct a5
Bogdanovski-Marin, Yerevan 1 996. 1 2 a31 (1 2 .l:r.el ?! d6! was good for Black in
1 36
4 g 3 i. b 7
Van Wely-Tiviakov, Gausdal 1 992) Black had nothing to fear in Romanishin

1 2 ... .i.xc3 1 3 .i.xc3 exd5 14 cxd5 .i.a6! 1 5 Markovsky, Biel 1 995.
l::te 1 with a modest plus fo r White thanks to 14 . . . l0ac5 1 5 i.f2 i.a6 1 6 l0f3
the strong and uncontested dark-squared White decides to forget about the c4-pawn
bishop. 14 ... lDce4 has been analysed by Bar and generate his own chances on the king
lov. After 1 5 llJf5 lDxc3 1 6 llxc3 .i.a6 1 7 d6! side.
White has the initiative. Note 17 ... .i.xe2 1 8 1 6 . . . e5!
dxc71 1i'c8 1 9 liJd6, and White wins in a mil 1 6 ... .i.xc4 1 7 llJd4 liJd3 18 .i.ft lDxf2 1 9
lion ways. White is on top after 1 5 ... lDxd5? �xf2 .ixft 20 lDc6 gives White something
16 .i.xg7 1i'g5 1 7 g4 etc. to bite on. Now the d4-squares is protected.
8 . . . i.xc3 1 7 f5! ?
There are altematives, but this is the
move.
9 bxc3 l0a6 1 0 l0h4 :tb8!
White refuses to compromise.

1 7 . . . .txc4 1 8 l:te3! b5 1 9 g4 f6 20 h4
:tea 21 l0e 1 c6 22 dxc6 :txc6 23 l0c2
Time has shown this to be the strongest :tc7 24 l:tg3 l0b6 25 g5 fxg5?
option. 10 ... e5 1 1 e4 d6 1 2 f4 exf4 1 3 lD£5 I qualify this move as a suicide attempt!
.i.c8 14 lDxg7?! <l;xg7 1 5 1i'd4 �g8 1 6 gxf4 Actually, Black built up his position in order
lDc5 1 7 e5 lDg4 was good for Black in Filip to be able to push ... d6-d5, and he had to do
pov-Tiviakov, Elista 1 995, but simply 1 4 it now! Then White would be at a crossroads:
.i.xf4 would have kept White slightly better. 25 ... d5 26 .txc5! :Xc5 27 1i'h5 <it>h8 28 lDe3
10 ...lLJc5 1 1 .ta3 l:tb8 1 2 .i.xc5 bxc5 1 3 1i'e7 29 l:ldl d4 30 cxd4 exd4 31 :Xd4 lDd7
l::tb 1 exd5 1 4 cxd5 .i.a6 1 5 l:lel d6 1 6 llxb8 with equality is a suggested improvement
1i'xb8 17 1i'd2 l:le8 1 8 e4 gave White some from Korchnoi.
pressure in Shneider-Brynell, Aars 1 995. 26 hxg5 d5 27 i.xc5! :txc5 28 'ii'h 5 l0d7
1 1 e4 d6 1 2 i.e3 29 f6?
12 f4 lDd7 13 lDS e5 1 4 fxe5 (1 4 .i.e3 29 lDe31.
lDac5 15 lDd2 .i.a6 is equal) 14 ...lDxe5 1 5 29 . . . d� 30 fxg7 i.f7 31 'ii'h 3 :txc3 32
lDxe5 dxe5 1 6 .i.a3 lDc5 1 7 .i.xc5 bxc5 1 8 l:txc3 d xc3 33 'ii'x c3 'ii'x g5 34 l0e3 l0f6
1i'a4 1i'd6! with a good game for Black in 35 :tf 1 l0xe4 36 'ii'a 3 l0g3 37 :tf2 'ii'x g7
Filippov-Savon, St. Petersburg 1 996. 38 'ii'x a7 :tf8 39 'ii'b 6 'ii'g 5 40 'ii'a 7 .te6
1 2 . . .:tea 1 3 :te 1 l0d7 14 f4 41 l0d5! l:txf2 42 �xf2 l0h 5 43 'ii' b 8 +
After 14 lDS lDac5 1 5 1i'c2 1i'f6 1 6 .i.d4 �g7 44 'ii'x b5 l0f6 45 'ii' b 7+ �h6 46
1i'h6 1 7 h3 .i.a6 1 8 .tf1 exd5 1 9 exd5 'ii'g6 l0xf6? 'ifxf6 + 47 'ii'f 3 'ifxf3 + 48 �xf3
137
..i.d5+ 49 �g3 ..txa2 5 0 ..te4 ..tf7 5 1 llld4 e5 1 5 tlldb5 l:.cS 1 6 e4 was a little bet
�g4 ..tg6 52 .ib7 �g7 5 3 �g5 h6+ 54 ter for White in Ivanchuk-Salov, Linares
�g4 �f6 55 ..tea ..tt5 + 0 - 1 1 993, although Black can improve) 1 3 ... llla6
14 dxe6 dxe6 15 l:.fdl l:.acS 1 6 'ii'a4 l:.fdS
Game 64 was completely equal in Epishin-Korobov,
Yermolinsky-Gulko Ohrid 2001 .
Scattle 2000 1 1 0-0 tL!xc4 1 2 dxe6 tLlb4
12 ...dxe6 13 'ii'b3 i.. d5 14 lllxd5 exd5 1 5
1 d4 tLlf6 2 c4 e6 3 tLlf3 b6 4 g3 ..tb7 5 llld4 'ii'd7 1 6 'ii'h5 'ii'xb5 1 7 lllxb5 l:.adS l S
..tg2 ..te7 6 tLlc3 tL!e4 7 ..td2 0-0 8 d 5 !? b3 gives White a small plus.
f5 9 'i'c2 1 3 'i'c 1
9 0-0 transposes to the main lines. 1 3 'ii'b 3? llla 5! 14 'ii'a4 dxe6 is better for
9 tLld6
• • . Black.
It is also possible for Black to play for 1 3 . . . d5
equality with 9 ...exd5 10 lllxd5 lll c6 1 1 0-0 Black cannot enter 1 3 ... dxe6 1 4 l:.dl ! llld5
a5 1 2 l:tadl i.. £6 13 i.. e3 �hS 14 lllxf6 l:Lcf6 15 lll d4 'ties 16 lllxe6 lllx f4 17 'ii'xf4 as he
1 5 lllg5 lllb4 1 6 °i'hl 'ties 1 7 a3 llla 6, as in clearly has severe problems.
Horvath-Atalik, 1991, which ended in a draw 1 4 b3 tL!a5
after 26 moves. 1 4 ... llld6 15 a3 llla6 16 lll e5 looks good
1 0 ..tf4 for White.
1 5 tL!e5 .:t6?!
15 ... c5!? 1 6 l:.dl i.. £6 and Black should be
able to keep the balance. Now he has to play
some precise moves in order to equalise.
1 6 'i'b 1 !
A possibly better shot for an opening ad

vantage is 1 0 b3 i.. £6 1 1 l:.dl 'ii'e 7 12 0-0
lll a6 13 llle l c6 14 dxc6 dxc6 1 5 e4 lllb4 1 6
°i'b 1 c 5 1 7 i.. £4 lll fl 1 S lllb5 e 5 1 9 i.. c l
l:.adS with a complex struggle in Van der
Sterren-Salov, Wijk aan Zee 1 99S, where 1 6 . . . d4
White might get to occupy d5 at the right 16 ... c5 17 a3 lll a6 lS ndl and d5 is weak.
moment if he plays his cards right. 1 7 ..txb7 tLlxb7 1 8 l:td 1 c5? !
1 0 . . . tLia6 1 S ... l:.xe6 1 9 'ii'x f5 l:.£6 20 'ii'g4 i.. d 6! 21
10 ... lllxc4 1 1 lllb 5 i.b4+ 1 2 �fl i.xd5 llle4 i.. xe5 22 i..xe5 l:.g6 23 'ii' B lllc 6 gives
13 lllxc7 lllc 6 14 lllxd5 exd5 1 5 lllg5 gives Black an acceptable position, although it is
White some initiative, while 10 ... i.. £6 1 1 0-0 still White who is dictating events.
'ii'e 7 12 i.xd6 cxd6 13 l:tacl (13 l:.fdl g6 1 4 1 9 e3 .:xe6 20 'i'xf5 .:ta 21 'i'g4 tLld6?
1 38
4 g 3 .i. b 7
This is just bad. Necessary was 21...�d6! move was played before and brought Black
22 exd4 cxd4 23 �bS when White's pros good practical results. I think it's strategically
pects are slightly preferable. dubious ... '
22 exd4 cxd4 23 llxd4 ltic2 24 .!tic& The alternative is 1 0 ... �a6 1 1 0-0 �d6
(1 1 ...�acS 12 �bS 0-0 13 �xc7 l:lac8 14
�bS exdS 1 5 cxdS �xdS 1 6 'i'b1 gave
White a plus in Naumkin-Marinelli, Formia
1 995) 1 2 �bS (1 2 dxe6!? dxe6 13 'i'a4+ rJitfl
1 4 �e3 favoured White in Petursson
Rozentalis, Malmo 1 993) 12 ...�xbS 1 3 cxbS
�cS 14 dxe6 dxe6 1 5 �b4 �e4 1 6 'ilc1 and
White was slightly better, Olafsson
Rozentalis, Lyon 1 994. Rozentalis generally
plays a lot of · these typical QID positions
without too much hesitation, as he knows
them very well.
1 1 cxd5
24 . . . 'ii'f 8 25 ltixe7 + 'ii'x e7 26 ltid 5 'ii'f 7 1 1 �xdS?! �xdS 12 cxdS 'ilcS! 1 3 'ilxcS
27 ltixf6 + 'ii'xf6 28 llxd6 'ii'x a 1 + 29 lld 1 bxcS 14 �c1 �a6 and Black is better
'i'f6 30 W'd7 llf8 3 1 W'd3 llc8? 1 32 (Olafsson-Arnason, Kopavogur 1 994).
'i'd5+ Wh8 33 .i.e5 'i'g6 34 'i'd8 + W'e8 1 1 ... cS?!
35 'ii'd 7 1 -0 This move finds a refutation in this game.
Better is 1 1 ...0-0 12 0-0 �a6 (probably the
Game 65 best -White has only a slight advantage now)
Khalifman-Korchnoi 1 3 �f4 �c3 14 bxc3 l:lae8 1 5 �d4 'i'cS 1 6
St Petersburg 1997 'i'b3 �h8 17 �e3 with an for White accord-
._______________.. ing to Khalifman.
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 .i.b7 5 1 2 dxc6! dxc6
.i.g2 .i.e7 6 ltic3 ltie4 7 .i.d2 f5 8 d 5 .i.f6
I am not too fond of this line. 8 ... 0-0
transposes to Game 64.
9 'ii'c 2 'ii'e 7
9 ...�d6 gave White a better game after 1 0
�f4 'ile7 1 1 �xd6 cxd6 1 2 0-0 0-0 1 3 l:lfdl
g6 14 �d4 eS 1 5 �dbS l:lc8 1 6 e4 in lvan
chuk-Salov, Linares 1 993.
1 0 lld 1
White is concentrating his play on the dS
pawn. Both alternatives are worse. 10 0-0
J.xc3 1 1 �xc3 exdS 12 cxdS �xdS looks
acceptable for Black, Hulak-Naumkin, Palma
·
1 989, while Tyomkin gives 10 �xe4? fxe4 1 1 1 3 ltixe4!
'W'xe4 exdS 1 2 'ilf4 �a61 1 3 �d4 0-0 1 4 �£5 This was a new move. The e-pawn is now
'W'cS as being better for Black. Clever guy. terribly weak.
1 0 . . . exd5 1 3 . . . fxe4 1 4 ltih4! .i.xh4 1 5 gxh4 0-0!
Khalifman writes in his notes: ' ... the text 1 6 'ii'c 4+ 'ii'f 7 1 7 'ii'xf7+ l:txf7 1 8 .i.xe4
1 39
White is close to winning the ending. 1 0 l:ad 1

1 8 . . . c5!? 1 9 .bb7 l:xb7 20 ..tc3 lt:lc6 2 1
l:g 1 :ea 2 2 l:g5 l:e4 2 3 h5 l:be7 24 e3
li:ld4 25 'lfi>f1 h6 26 l:d5 lt:lb5 27 ..te 1
l:f7? ! 28 l:d8 + 'lfi>h7 29 l: 1 d 7! l:f5 30
J:d5 .:as 31 J:bS! l:d6 32 :t5 :ts 33
l:xf6 gxf6 34 l:d8 c4? 35 a4 c3 36 axb5
cxb2 37 l:d 1 l:c4 38 ..td2! l:c5 39 l:.b 1
l:xb5 40 ..tc3 l:c5 41 ..txb2 llb5 42 �e2
a5 43 'lti>d3 a4 44 'lfi>c2 J:c5+ 45 ..tc3 b5
46 �d3 l:xh5 4 7 ..txf6 1 -0
Game 66
lvanchuk-Timman
Wljk aan Zee 200 1 1 o . lt:la6
. .
1bis is certainly not the main move, and in

1 li:lf3 li:lf6 2 c4 b6 3 g3 ..tb 7 4 d4 e6 5 this game Black is nowhere near equality .
..tg2 ..te7 6 lt:lc3 0-0 7 0-0 lt:le4 8 ..td2 The alternatives are:
f5 a) 10 ...l2Jxc3 is the critical line given in
Because of the trouble I feel Black faces in Chess Informant, and also the most natural.
this line I highly recommend 8 ... i.f6. But I think Black needs to come up with a
9 'i'c2! new concept somewhere to play like this. 1 1
1bis quieter variation is an attempt to in i.xc3 i.e4 (1 1 ...'ii'e7 allows 1 2 d5! i.xc3 13
crease the tension. After 9 d5 i.f6 1 0 l:tcl 'ii'xc3 l2Ja6 14 liJd4 l2Jc5 1 5 b4 l2Je4 1 6
l2Ja6 Black normally equalises completely. i.xe4 fxe4 1 7 dxe6 dxe6 1 8 b 5 , when White
The theory goes as follows: 1 1 a3 c6! 12 dxe6 is just a whole lot better, Ovsejevitsch-Rehm,
dxe6 1 3 'i'c2 c5 1 4 l:tfdl 'i'e7 1 5 lDel (1 5 Berlin 1 997) 1 2 'iicl!, Azmaiparashvili
l2Jxe4 i.xe4 1 6 'i'a4 liJb8 1 7 b4 l2Jc6 1 8 b5 Gulko, Elenite 1 995. The queen is much
liJd8 1 9 i.c3 i.xc3 20 l:txc3 liJfl was equal better placed here than on d2 because it
in Beliavsky-Brodsky, Koszalin 1 998) leaves the way clear for the traditional plan
15 ... l2Jxd2 1 6 i.xb7 'i'xb7 17 'i'xd2 l:tad8 1 8 with i.h3 followed by liJd2, driving the
'i'e3 'i'e7 with equality in Hansen-Korchnoi, bishop away from e4 and seizing control of
Malmo 1 996. Or 1 1 i.el c5! (the new way to the centre. Now 1 2...'i'e8, to prevent i.h3, is
play) 1 2 dxc6 dxc6 1 3 'i'b3 l2Jac5 1 4 'it'c2 answered with 13 l:tfel with the idea of i.fl .
'i'e7 1 5 b4 l2Jxc3 1 6 i.xc3 i.xc3 1 7 'i'xc3 However, it could b e argued that the inclu
l2Je4 1 8 °Wb2 c5 with complete equality, Be sion of these points benefits Black as well as
liavsky-Stefansson, Istanbul 2000. White. Instead the game continued (1 2 'i'cl !)
9 . . . ..tf6 1 2 ... d6 1 3 i.h3! 'it'e8 (Gulko does not like
9 ... l2Jxc3 10 i.xc3 i.e4 1 1 'i'a4 d6 1 2 this move and improves it with 1 3 ... l2Jd7!? 1 4
l:tacl 'i'd7 1 3 'i'dl ! c 5 14 'i'd2 liJc6 1 5 l:tfdl liJd2 i.b7 1 5 e 4 c5! wi th counterplay - he is
and White had a little something in Skem probably correct, although it is far from clear
bris-Huzman, Beersheba 1 993. that Black has a potentially equal position,
9 ... l2Jxd2 10 'i'xd2 i.b4 1 1 l:tacl i.e4 1 2 and 16 d5 seems good for White to me) 14
a3 i.xc3 1 3 l:txc3 'ii'f6 14 ltJ e1 i.xg2 1 5 l2Jg5 'ii'h 5 1 5 lDxe6! and now White found
lDxg2 lDc6 was very slightly better fo r White the strong combinatorial continuation
in Stahlberg-Castaldi, Hamburg 1 955. 1 5 ...'i'xh3 1 6 f3 i.b7 17 lDxc7 and Black
140
4 g 3 j,, b 7
was in trouble, although he went on to win. Interesting was 15 ... cSI? 16 dxc5 lDxc5 1 7 b4
Earlier 1 2 'ii'd2 leads to a more or less d6, which Fritz came up with, although it
equal position after 12 ... d6 13 lDe1 .ixg2 1 4 looks like White is better after 1 8 :Xd6 (1 8
lLJxg2 1i'e7!, when th e idea behind develop bxc5 dxe51 gives Black a good game)
ing the queen first is seen after 1 5 1i'c2, when 1 8 ...'ii'e4 1 9 'iid2 .ixe5 20 .ixe5 1i'xe5 21
Black has the deep reply 1 5 ...1i'f7! intending f4! (21 bxc5 1i'xc5 and Black is fine) 21 ...'ii'e4
1 6 e4 fxe4 1 7 1i'xa4 d5! followed by 1 8 ... lDc6 22 bxc5 bxc5 23 'i'd3! etc.
and Black has good and natural counterplay. 1 6 dxe5 l:f7
Karpov -Polugaevsky, Amsterdam 1 981 went
1 5 lDf4 lDd7 1 6 d5 .ig5 1 7 dxe6 .ixf4 1 8
1i'xf4 1i'xe6 1 9 :fe1 lDc5 and Black fine.
b) 10 ... 1i'c8 1 1 d5! lDd6 12 b3 lDa6 1 3
.if4 favoured White in Lundin-Skold, Swe
den 1 967.
c) 10 ... d6 11 d5 .ixc3 12 .ixc3 lLJxc3 1 3
'ii'xc3 e 5 1 4 b 4 saw White assume an advan
tage on the queenside in Larsen-Levitt, Lon
don 1 990.
d) 10 ... c5 1 1 d5 .ixc3 12 .ixc3 exd5 1 3
cxd5 d 6 1 4 lDh4 lDxc3 1 5 bxc3 1i'f6 1 6 f4
lDa6 1 7 .ih3 g6 1 8 e4 was also better for
White in Navrotescu-Marin, Bucharest 2001 . 1 7 b4 c6 1 a l:d6 b5 1 9 l:ed 1 bxc4 20
1 1 .!Lle5! a3 :ea 21 j.e 1 .!Llba 22 a4 c5 23 b5
11 a3 c5 1 2 d5 .ixc3 13 .ixc3 exd5 1 4 We4 24 Wxe4 fxe4 25 j.c3 �fa 26 �g2
cxd5 d 6 1 5 lDh4 lDxc3 1 6 bxc3 achieved l:c7 27 j.d2 :cs 2a g4 c3 29 j.e3 a6
nothing in Gritsak-Brodsky, Ordzhonikidze 30 l:c 1 axb5 31 axb5 �ea 32 l:xc3 c4
2000. 33 h3 :ta 34 j.a7 �e7 35 l:d4 e3 36
1 1 . . . .!Llxc3 1 2 j.xc3 j.xg2 1 3 �xg2 'ii'c a fxe3 l:c7 37 i.b6 l:b7 3a i.c5 + �ea 39
14 l:fe 1 Wb7 + 1 5 �g 1 j.xe5 b6 l:ga 40 l:a3 d5 41 exd6 e5 42 l:a7
This is a tough decision, after which the 1 -0
pressure down the d-file and the potentially A tough struggle, where the knight never
very strong bishop gives White an advantage. got into the game.
14 1
Summary
4 ... .tb7 still seems to be playable according to theory. The older attempts to get an advantage
with 7 dS and 7 b3 do not trouble Black and neither does 7 l:.et lll a 6!. The most dangerous
tries are the logical developing moves with 6 lll c3 and 7 .id2 in reply to 6 ... llle4. Black should
have a good grasp of the theory to play these lines, but if he does, then he can also count on
equality. My investigations have clearly indicated that ... .tf6 is a better defence than . fS, but
. .
who knows what time will show.
1 d4 .!bf6 2 c4 e6 3 .!bf3 b6 4 g3 .i.b7 5 .i.g2 {D) .i.e7

5 . cS - Game 56
..
6 lLJc3
6 0-0 0-0 (D)
7 b3 - Game 57
7 dS exdS 8 lllh4 - Game 58
7 l:.et - Game 59
7 lll c3
7 ... lll a6 - Game 60
7 ... lll e4
8 .id2 - 6 lllc3
8 'fi'c2 - Game 6 1
6 .!be4
• • .
6 . 0-0 7 dS i.b4 - Game 63

..
7 .i.d2 (D) 0-0

7 ... .tf6 - Game 62; 7 . fS 8 dS .if6 - Game 65
..
8 0-0
8 dS f5 9 'ii' c2 - Game 64
8 f5
. • .
8 ... .if6 - 7 ... .tf6

9 'if c2 Game 66
-
5 .i.g 2 6 . . . 0-0 7 .i.d2
142
INDEX OF COMPLETE GAMES
I
Alterman-Liss, Israel 1999 ............................................................................. 76

Anastasian-Brodsky, Moscow 1992.............................................................. 95
Andersson-Karpov, Ti/burg, 1983 ............................................................... 133
Bacrot-Adams, Cannes (rapid) 200 1 .............................................................. 39
Barbero-Cebalo, Caorle 1987 ........................................................................ 97
Bareev-Eingom, "Kiev 1986 ........................................................................... 40
Bareev-Karpov, Moscow (blitz) 1993 ............................................................. 18
Barlov-Beliavsky, Yugoslavia 1992 ............................................................... 96
Barlov-Sakaev, Budva 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .�
................... . 36
....
Belozerov-Nikolenko, Moscow 1999 ........................................................... 67

Bogdanovski-Cabrilo, Bijefina Dvorovi 2002 ........................................... 1 13
Borges Mateos-Spraggett, Cienfuegos 1997 ............................................. 107
Cramling-Almasi, Horg,en 1995 .................................................................... 51
Crouch-Harikrishna, London 200 1 ............................................................. 73
Danner-Yu Shaoteng, Gyula 2000 .............................................................. 89
Dokhoian-Romanishin, Yerevan 1989 ..................................................... 1 19
Dreev-Seirawan, Wyk aan Zee 1995 ............................................................ 29
Dzhandzhava-Chernin, Lvov, 1987............................................................ 37
Ehlvest-Chandler, Lucerne 1989................................................................. 1 18
Fominyh-Chemyshov, Smolensk 2000 ..................................................... 132
Freitag-Oral, Graz 1998 .............................................................................. 129
Gelfand-Karpov, Sanghi Nagar (7) 1995 ...................................................... 87
Gelfand-Kasparov, Novgorod 1997 .................... , .......................................... 30
Gershon-Anastasian, Saint Vincent 2000 ........: .......................................... 45
Gheorghiu-Csema, Berlin 1986 ................................................................. 124
lvanchuk-Timman, Wyk aan Zee 200 1 .................................................... 140
Kamsky-Karpov, Elista 1996 ...................................................................... 123
Kamsky-Yudasin, Biel 1993 ......................................................................... 63
1 43
Karpov-Gavrikov, USSR 1988 ................................................................... 127

Karpov-Polgar, Buenos Ains 200 1 .............................................................. 106
Karpov-Salov, Unans 1992 ......................................................................... 1 16
Kasparov-Gelfand, Novgorod 1997 ............................................................. 1 1 1
Kasparov-van der Wiel, A.msterdam 1988 . . . . . . . . . . . . . . . . . . . . . � ............................ 55
Khalifman-Korchnoi, St Petersburg 1997 .................................................. 139
Khalifman-Short, Paernu 1996...................................................................... 59
Lautier-Karpov, Monte Carlo (rapid) 2000 ................................................. 103
Lautier-Timman, Amsterdam 1996.............................................................. 42
Magerramov-Makarichev, Moscow 1991 ................................................... 93
Miles-Timman, Ti/burg 1986 ........................................................................ 71
Miles-Unzicker, Johannesburg 1979 .............................................................. 16
Neverov-Stefanova, Reykjavik 2002 ............................................................ 47
Nikolic-Lautier, Moscow (Fide knockout) 200 1 ............................................ 19
Petursson-Polugaevsky, Rey�javik 1987 ................................................. 1 15
Piket-Anand, Monte Carlo (blindfold) 1997 .................................................. 49
Piket-Chuchelov, Netherlands 2000 ........................................................... 109
Piket-Tiviakov, Wijk aan Zee 1994 .............................................................. 69
Pinter-Tompa, Hungary 1993 ..................................................................... 134
Polugaevsky-Wojtkiewicz, Haninge 1990 ............................................... 128
Portisch-Nikolic, I..:fub!Jana 1985 ................... .-. ............................................ 24
Portisch-Popovic, l./ubb'ana 1985 ................................................................ 65
Romanishin-Korchnoi, L.vov 2000 ........................................................... 136
Sakaev-Berzinsh, Duisburg 1992 . .� . . .
. .......... ...
. .... .. ......... . . 105
. .........................
Salov-Karpov, Wijk aan Zee 1998 ................................................................. 99

Seirawan-Timman, Hilversum Match (5thgame) 1990 ............ :.................. 8 1
Silman-Gheorghiu, Palo Alto 198 1 . � . .
... .... ........ .
...... . . 35
.............. .............. . ....
Teske-Paehtz, East Ge1'111a'!J 1988 ............................................................... 21

Timman-Karpov, Hoogeveen 1999 ................................................................ 79
Timman-Polgar, Bali 2000 53
...........................................................................
Tkachiev-Nisipeanu, Naujac-sur-Mer 2000 ............................................. 1 13

Tolnai-Adotjan, Hungary 1992 . . .
........... ........... . . . .· . . . . 62
............................. ... . ...
Topalov-Adams, Fujitsu Siemens Masters Frankfurt 2000 ........................ 1 1 7

Tregubov-Shaposhnikov, Samara 2000 .................................................. 101
Van Wely-Korchnoi, Hoogeveen 200 1 ......................................................... 13 1
Yermolinsky-Gulko, Seattle 2000 .............................................................. 138
Yusupov-Beliavsky, Austria 1998 ............................................................... 90
1 44

Aagaard - Queen's Indian Defence (2002)

Uploaded by

Copyright:

Available Formats

Aagaard - Queen's Indian Defence (2002)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Aagaard - Queen's Indian Defence (2002)

Uploaded by

Copyright:

Available Formats

queen's

Copyright© 2 002 Jacob Aagaard

British Library Cataloguing-in-Publication Data

ISBN 1 85744 300 4

EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Typeset and edited by First Rank Publishing, Brighton.

Popular opening books:

18574 4 218 0 Unusual QG Declined Chris Ward

Books for players serious about improving their game:

1 8574 4 226 1 Starting Out in Chess Byron Jacobs

Books for the more advanced player:

18574 4 233 4 Attacking with 1 e 4 John Emms

White plays without g2-g3

1 Opting for Nothing: 4 �f4 and 4 �gS 16

Lines with g2-g3

6 4 g3 i.b4+ and 4 ... �a6 without 5 b3 93

Index of Complete Games 143

Theoretical Overview possibilities.

5 i.d2 .txd2+ 6 'i'xd2 i.a6!, which trans­ line.

7 �c3 raises nwnerous issues. First White

course this is tested again and again at the

such as the following example: just as weak as dS).

1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 i.. b 7 5

1 2 ibe5 c5 1 3 dxc5 .i.xc5 1 4 tbd3 .i.d6

Game 1 This is an interesting choice. The simpler

If White fails to address the well-being of

Exchanging the bishop to win the eS­

This move gives the line independent

What a blunder! After 1 1 ....txd2+! 1 2

1 d4 tL'lf6 2 c4 e6 3 tL'lf3 b6 4 .tf4 (D)

4 . . . 1.b7 5 e3 1.e7 (D)

4 1.. f4 4 .tg5 5...1.. e 7

1 d4 ltlf6 2 c4 e6 3 .!bf3 b6 4 a3 10 i.f4 tl:ihS, which currently looks like the

bS 14 lLieS 'iib 6 1S i.gS :£e8 1 6 f4 earned

The most active, but Black has more than

f8 this is not the case) 1 5 .l:[fe1 i.d6 1 6 i.e5

<iL'g7 47 .J:.c5 <iL'g6 48 h5+ <iPxh5 49 7 . . .a 6 8 l:.c 1 ! .tb7

1 d4 ll)f6 2 ll)f3 e6 3 c4 b6 4 a3 c6? !

course. Otherwise the move would have

White can also give up the right to castle

Here White needs to do something to

absolutely fine. 15 ..ifl!? b4 1 6 li)b5 ..ixb5!? Still 20 li)g5!? i s necessary.

1 6 li)dl b3 1 7 'fibl 'fic7 1 8 li)c3 l:.ab8

This is the idea - now White loses control

With complete domination. nos Aires 1 980.

Tbis move is less ambitious than 5 d5, and

�f4 ltif6 1 6 �e5 :ea 1 7 .J:xea Wxea 1 a

52 'ilxd 1 'flxe4 53 'ild7 + �h6 54 'ilea

1 7 cJ.?d 1 i.. x e3! 1 8 tllh6+ cJ.?h8 1 9 tll f7+ .l:t.xf7 8 . . . .i.c5!?

have played 1 6 ltixc5! bxc5 1 7 ltid5 with an

lDc6 9 lDxc6 refutatioi:i is not 1 1 ...lDxe4?1, as played in

Whether this is the best move or not is

f6. 1 d4 lLlf6 2 lbt3 e6 3 c4 b6 4 a3 ..ta6 5

The best move. White does not want to

i.d4 t 7 :lift with close to equality, or even 1 998.

Slobodjan-Speelman, Lippstadt 2000. Black is trying to defend. Active counter­

tage. With a mating attack. Black resigned.

found sufficient defences after 8 'i'xd7+ White in Gelfand-Karpov, Dortmund 1997,

struggling for survival. passed pawn on dS, while Black's develop­

examples. c) 9 lDxd4 has become the critical move,

5 i.d2 .txd2+ 6 'i'xd2 i.a6!, which trans line.

Exchanging the bishop to win the eS

Slobodjan-Speelman, Lippstadt 2000. Black is trying to defend. Active counter

struggling for survival. passed pawn on dS, while Black's develop

vantage) 1 1 i.xc6 l:.c8 1 2 d5 'ii'c7 is a differ

:acl ?! (1 7 dxeS .txeS 1 8 ilxeS :xeS is ac 'iig 4?!