Sadler - Queens Gambit Declined (2000)

~.
ueen's
•
~a ~I""
•
~ec Ine~
by Matthew Sadler
EVERYMAN CHESS
Published by Everyman Publishers pic, London
First published 2000 by Everyman Publishers pIc, formerly Cadogan Books
pic, Gloucester Mansions, 140A Shaftesbury Avenue, London WC2H 8HD
Copyright © 2000 Matthew Sadler
The right of Matthew Sadler to be identified as the author of this work has
been asserted 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 256 3
Distributed in North America by The Globe Pequot Press, 6 Business Park

Road, P.O. Box 833, Old Saybrook, Connecticut 06475-0833.
Telephone 1-800-243 0495 (toll free)
All other sales enquiries should be directed to Everyman Chess, Gloucester

Mansions, 140A Shaftesbury Avenue, London WC2H 8HD
tel: 0171 539 7600 fax: 0171 379 4060
email: [email protected]
website: www.everyman.uk.com
To Steve Giddins, my first coach!
The Everyman Chess Opening Guides were designed and developed by First
Rank Publishing.
EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Chief Advisor: Garry Kasparov
Advisory Panel: Andrew Kinsman and 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.
CONTENTS
1 d4 d5 2 c4 e6 3 l2Jc3
Bibliography 4
Introduction 5
Part One: Main Line Queen's Gambit Declined

(3 ... tDf6 4 tDf3 ~e7 5 ~g5 0-0 6 e3)
1 Lasker Variation (6 ... h6 7 ~h4 tDe4) 10

2 Orthodox Variation (6 ...ttJbd7): Old Main Line with 7 %tcl c6 28
3 Orthodox Variation (6 ... ttJbd7): Other Systems after 7 %tcl 40
4 Orthodox Variation (6 ... ttJbd7): 7 'iVc2 and Other Seventh
Move Alternatives for White 50
5 Tartakower Variation (6 ... h6 7 i.h4 b6): Fixed Centre Plans 62
6 Tartakower Variation (6 ...h6 7 i.h4 b6): Development Plans 89
Part Two: Exchange Variation

(3 ... tDf6 4 cxd5 exd5 and 3 ... .ie7 4 cxd5 exd5)
7 Exchange Variation: Systems with ttJf3 109

8 Exchange Variation: Flexible Systems 134
Part Three: Other Variations
9 Systems with ~xf6 148

lO 5 i.f4 Variation 161
QGD 'General Knowledge' 170

Index of Complete Games 174
BIBLIOGRAPHY
Books
Encyclopaedia o/Chess Openings vtJiume D, 3rd edition (Sahovski
Informator, 1997)
Queen's Gambit Declined 5 iL/4 Colin Crouch (Everyman, 1999)
Periodicals
In/ormator
New in Chess
British Chess Magazine
Chess Monthly
Schach
Acknowledgements
I would like to thank Mark Dvoretsky and Steve Giddins for their help with
this book, as well as the manager of the Mall Cafe, Pentagon Centre,
Chatham, for his kindness and the delicious Danish pastries!
INTROTJUCTION
1 d4 d5 ... i..c8-f5 or -g4 first and only then ...e7-e6

White's aim was to follow 1 d4 with 2 e4, and ...i..f8-e7. This costs an extra tempo for
establishing a double pawn centre. With development, but in this way, all of his pieces
1...d5, Black prevents White from achieving would be on active posts. Black could then
this goal. seek to improve his position.
Where does Black want to put his pieces? QIestinn 3. 'Improve his position'. What
When working out a scheme of does that mean?
development, there are always three Anm.er 3. At the beginning of the game,
questIons: this does not mean anything dramatic. You
1. Can I fmd an active post for each of my put pressure on the opponent's centre, you
minor pieces? gain just a little more territory, and complete
2. How will I be able to improve my the mobilisation of your forces.
position aftelWards? Q1estion 4. So how does Black do this
3. What about my opponent's aims? here?
Black wishes to develop his kingside and Anm.er4. Black's main idea is to play ...c7-
castle his king to safety there. Consequently, c5, striking at White's d4-pawn and thus
it is clear that the moves ...ttJg8-f6, ...e7-e6 gaining a little central and queenside space.
and ...i.f8-e7 (or -d6/-b4 in some cases) will He will then develop the rest of his
occur at some stage. queenside pieces probably starting with
QIestinn 1. Is there a drawback to this ...ttJb8-c6.
method of development? Q1estion 5. And after that?
An..9ZW" 1. Although this development is Anm.er 5. Now we're going too far! That
kind to the kingside pieces, it causes some depends a lot on what White has done, but if
problems for Black's queenside light-squared your minor pieces are active and you know
bishop. The move ...e7-e6 restricts its access how to start your search for activity, then
to the c8-h3 diagonal, leaving it with only the there will always be things for you to do in
d7-square, from which it performs no useful the position.
function. However, it is White's move and with
Qlestion 2. What is the solution? 2 c4
An..9ZW" 2. Ideally, Black would like to play he throws a spanner into the works.
5
Queen's Gambit Declined
begun his central counterplay with ...c6-c5

(trend 2). However, White has achieved his
goal of a double pawn centre (trend 3), as
well as a distinct lead in development. This
collision of ideas is vel)' typical of modem
chess and leads to vel)' sharp play.
Q«;stion 6. What is White's aim?

An.9Za'Y 6. White wants to take over the
whole centre by removing the only brake on
his ambitions: the dS-pawn. Thus he intends
3 cxdS 'ifxdS 4 l2Jc3 'ifd8 5 e4 with total
domination of the centre.
As a general answer to our questions so Most recently, however, the emphasis of
far, there are three noticeable trends: the top players, notably Kramnik (the most
1. Black's 'problem piece' is the light- prominent Semi-Slav expert of the 1990s)
squared bishop on c8, since the natural and Kasparov, has switched to the opposite
development of the black kingside shuts it end of the spectrum.
inside the pawn chain. Qu:stian 8. You mean, they think that the
2. Black will nonnally search for third point is the most important?
counterplay by playing ...c7-c5. An.9Za'Y 8. That's right! Black's immediate
3. White wants to remove the black dS- task is to prevent White from occupying the
pawn in order to occupy the centre with centre with pawns on e4 and d4.
pawns on d4 and e4. Qu:stian 9. So how does Black do this?
Qiestion 7. How should Black respond? Armrer 9. By playing the Queen's Gambit
An.9Za'Y 7. Black's response depends on his Declined (QGD) move...
interpretation of the relative importance of 2 ...e6
these three trends. For example, let us
consider the Semi-Slav which became the
most popular opening against 1 d4 in the
mid-1990s: 1 d4 d5 2 c4 c6 3 ttJf3 l2Jf6 4
l2Jc3 e6 5 e3 ttJbd7 6 i.d.3 dxc4 7 i.xc4 b5 8
J..d3 i.b7 9 0-0 a6 10 e4 c5.
see folio wing diagram
It is clear that Black has concentrated on
the flrst two factors and discarded the third.
Black has solved the problem of his light-
squared bishop by developing it on the long
a8-h1 diagonal (trend 1) and he has already
6
In troduc tion
By supporting dS with the e6-pawn, Black Gambit Accepted (QGA) after 1 d4 dS 2 c4

maintains his control over the e4-square as 3 dxc4, for example, which White can easily
cxdS can now be met by 3... exd5. Moreover, avoid by playing
Black allows the dark-squared bishop to 1 c4
develop and thus begins immediately to The QGD player, however, simply plays
prepare the development of his kingside and 1 ... e6!
kingside castling.
Qtestion 10. But you block the light-
squared bishop inside the pawn chain don't
you?
Anm..er 10. Right again! In order to prevent
White from achieving his plan quickly, Black
inevitably has to offer a concession of his
own: in this case, the passivity of the light-
squared bishop at the start of the game.
Black's contention is that this is only a
temporaty feature that will quickly be
rectified in the ensuing middlegame.
Move Order 2 d4 ciS reaching the QGD! Alternatively,

if
The key position of the Queen's Gambit 1 tiJf3 d5 2 c4
Declined (QGD) arises after then after
3 tiJf3 liJf6 4 tiJc3 iLe 7 2 ... e6!
and this will be the main focus of our 3 d4 ttJf6, the QGD is again reached.
attention. However there are several move- Strangely enough, the biggest move-order
order questions to be resolved before we can debate for Black arises when his opponent
proceed. plays the straightforward
The flexibility of the QGD is its greatest 1 d4 d5 2 c4 e6 3 tiJc3
asset. Whether White begins with 1 ttJf3, 1
see following diagram
c4 or 1 d4, if at any stage he intends to play
both c2-c4 and d2-d4, then he cannot avoid Black now has two choices - 3...lbf6 or
the QGD. This is in contrast to the Queen's 3 ... .i.e7.
7
Queen'5 Gambit Declined
frequently, but Short and Ivanchuk have

played 3...ttJf6. I would recommend learning
3...ttJf6, simply for its flexibility.
Qiestion 11. What is the difference?

Anszrer 11. 3...tiJf6 allows 4 cxd5 exdS 5
i.g5
Q.lestion 14. What do you mean?
Arma'Y 14. Many players aim for the QGD
via a cunning move-order: 1 d4 tiJf6 2 c4 e6.
reaching a vanatlon of the Exchange

QGD where White has not yet committed
his king's knight to the f3-square. This allows
him to play a souped-up version with the Qtestion 15. What's the idea?
knight on e2. Amzrer 15. The idea is to exploit White's
Qiestion 12. Aha, and 3... i.e7? own repertoire: after 3 tiJc3, as well as 3... dS,
Arz.flW" 12. By reversing the order of his transposing back into the QGD, Black can
kingside development (...i.f8-e7 before play 3...i.b4 leading to the Nimzo-Indian
...tiJg8-f6) Black prevents 4 cxdS exdS 5 Defence.
i.g5, and therefore encourages White to play Qtestion 16. But I don't want to learn the
an Exchange variation with the bishop on f4 Nrrnzo-Indian as well!
rather than g5: 4 cxdS exdS 5 i.f4. Amzrer 16. You don't have to! The point is
see following diagram that many White players do not allow the
Nrrnzo-Indian and instead play 3 ttJf3,
Qiestion 13. So what do strong players do? aiming for a Queen's Indian after 3...b6.
Arz.flW" 13. Opinion is divided - Kasparov Then you play simply 3...d5 and...
and Karpov have both played 3....i.e7 quite Qtestion 17. I'm into a QGD without
8
In troduc tion
allowing any It:lgl-e2 plans! the original queen's pawn opening; modem
Anszrer 17. Exactly! If they do go 3 tiJc3, systems such as the Semi-Slav or the QGA
then you go 3... d5, but I guarantee that your have developed by taking features of the
opponent will have wasted a couple of QGD and accelerating them, e.g.
minutes thinking over your move-order! Of compromising king safety in order to free the
course you cannot play the 3... i.e7 via this light-squared bishop in double-quick time as
move-order which is why I recommend we saw in the Semi-Slav example. The aim
learning 3... ttJf6. This gives you the flexibility therefore has been to give some insight into a
of two move orders to the QGD: 1 d4 d5 2 range of 1 d4 openings - the Nunzo-Indian 4
c4 e6 and 1 d4 ~f6 2 c4 e6! e3 system, the Chigorin, the Semi-Slav to
The theme for this book has been to name but a few - and thus to reveal
highlight the links between the QGD and the something about the whole queen's pawn
other queen's pawn openings. The QGD is complex as well as the QGD itself.
Q
CHAPTER ONE
Lasker Variation
(6 ... h6 7 i..h4 ttJe4)
1 d4 d5 2 c4 e6 3 lDc3 J..e7 4 lDf3 tiJf6 World Champions of all time, Emmanuel

5 J..g5 Lasker. It is an extremely important line
The 5 i.g5 variation is White's most solid because the themes within it recur
attempt for the advantage. It has two basic throughout the QGD.
ideas: QIestian 2. This looks like an aggressive
1. White puts his dark-squared bishop move!
outside the pawn chain in order to allow the Amuer 2. In fact, this is one of the quieter
development of his kingside with e2-e3 and lines of the QGD!
.ltfl-d3/e2. QIestian 3. What is the point of ... lLJf6-e4?
2. White interferes with BI~s desire to Amuer 3. Firstly, since Black's position is
play the freeing move ...c7-c5. slightly cramped, he will generally wish to
Qiestion 1. How so? exchange pieces. The fewer pieces he has in a
Anszw-l. In two ways. Firstly, it attacks a restricted space, the easier his development
major defender of the ciS-pawn - the knight becomes. Moreover, by solving his own
on f6. Secondly, it exerts pressure along the space problems, he also reduces the
h4-d8 diagonal; for example, were the black importance of White's space advantage.
bishop to be distracted from e7 after ...c7-c5, Secondly, by exchanging White's dark-
d4xc5 ...i.e7xc5, then the black knight on f6 squared bishop and his queen's knight, Black
would be unpleasantly pinned to the queen removes the pieces that were pressuring his
ond8. centre (the knight directly attacking the dS-
5 ... 0-0 pawn, and the bishop indirectly attacking the
Black can also tty the similar 5... h6 6 .lth4 dS-pawn by threatening to capture the knight
lLJbd7 (6...0-0 7 e3 is simply a transposition on f6). This releases the immediate pressure
to the main line) 7 e3lLle4, as in Game 9. from his position, allowing Black more
6 e3h6 flexibility in his development.
The immediate 6...lLJe4 is less effective - QIestian 4. Sounds like this just equalises
see Game 8. for Black!
7 J..h4lDe4 Amuer 4. Not so fast! There are a number
This move introduces the Lasker of drawbacks to this idea:
variation, named after one of the greatest 1. By moving the knight twice in the
10
Lasker Variation
opening, Black gives White an extra tempo and Miladinovic agreed a draw in Karditsa
for his own development. Moreover, ...ttJf6- 1995, but the position is quite unclear. As
e4 exchanges the only minor pieces that compensation for the two bishops, White
Black has developed! Consequently, this has ideas such as g3-g4-gS and 0-0-0.
manoeuvre does not further Black's 8 ...'ifxe7 9 l:c1
development in the short-term.
2. With his central pawns on light squares,
Black exchanges off his 'good' dark-squared
bishop. Consequently, Black may suffer from
weak central dark squares.
Qtestinn 5. It sounds a bit stupid to swap
off your good bishop!
Arzsuer 5. Black feels that these exchanges
will make it much easier for him to achieve
the freeing break that will liberate his 'bad'
bishop. As with 2...e6, when Black shut in his
light-squared bishop in order to hold back
White's centre, so here Black also has to give
something up in order to get closer to his This is Kramnik's and Karpov's choice
ultimate goal. Black's judgement is that when and is White's main attempt in this position.
he achieves his fmal goal, then this will (The alternatives 9 cxdS and 9 ,.c2 are
compensate for any small concessions he has considered in Games S and 6, and Game 7
to make. respectively.)
Qtestinn 6. These advantages and QIestion 7. Why is this?
disadvantages all sound a little subtle to me! Amzar 7. Due to the tension between the
Arzsuer 6. I know what you mean! At the c4- and dS-pawns, the c-file is likely to
moment, there is no scope for wild kingside become semi-open either by c4xdS or by
attacks or sacrifices. Both sides are quietly ... dSxc4. It is therefore a good positional
accumulating the 'evidence' for their decision to place a rook on this file.
assessment of the position: White looks at Moreover, with this move White makes it
his slight space advantage and Black's tactically impossible for his opponent to play
undeveloped queenside pieces to claim he is the desirable freeing break ...c7-cS, as 9...c5
better, whereas Black shows what he has (or 9...ltJxc3 10 :hc3 dxc4 11 i..xc4 c5 12
neutralised in White's position to claim he is dxcS 'ii'xcs 13 i..xe6!) 10 cxdS! ttJxc3
heading for equality. The QGD always takes (10 ...exdS 11 ttJxdS) 11 ':xc3 exdS 12 ':xc5
a little while to get going! costs Black a pawn. Finally, the pressure
along the c-file interferes with Black's
Gamel development. Thus the natural 9 ... ttJd710ses
Karpov-Yusupov a pawn to 10 cxdS ttJxc3 11 :xc3! exdS 12
Dortmund 1997 ':xc7.
Qtestion 8. So is Black in trouble now?
1 d4 tiJf6 2 c4 e6 3 tiJf3 d5 4 tiJc3 i.e 7 Ansrrer 8. Stay calm! Let's work this out!
5 i.g5 h6 6 i.h4 0-0 7 e3 llJe4 8 .txe7 Since Black cannot achieve an immediate
8 ~g3!? is a relatively unexplored idea. ...c7-c5, it is clear that Black needs the
After 8...c5 9 i..d3 cxd4 10 exd4 ttJxgJ 11 support of his undeveloped queenside pieces
hxg3 dxc4 12 hc4 lZ)c6 13 'iWd2, Skembris in order to create any counterplay.
11
QIestion 9. But how? You said I can't play c4xdS lines as we shall see in Game 4.
9...ltJd7. 9 ... c6 10 .i.d3
Amzrer 9. Well how about the preliminary Forcing the knight from e4 as 10.. .f5 11
9...c6? This places the c-pawn on a defended ltJeS! probes the sensitive g6-square.
square and so prepares ...ltJb8-d7. 10 ...liJxc3 11 :xc3
QIestion 10. It looks a bit slow! Qlestion 12. What is Black aiming for now?
Amuer 10. I understand, but look at An.9lW" 12. It is important to notice that
White's position. Is he ready to launch a huge Black has two central breaks: ...c6-c5 and also
offensive? Can he punish me for spending a ...e6-e5. The latter is very kind to the light-
tempo on a consolidating move? squared bishop as it reopens the c8-h3
QIestion 11. I suppose the answer is no! diagonal. Both these breaks will require the
Amzrer 11. Correct! I had a lot of trouble support of the queen's knight from d7. Thus
understanding the rhythm of these positions there are three distinct methods of play for
when I fIrst analysed the QGD as a Black.
youngster. In all my other lines - Sicilians 1. The solid ...dSxc4 with ...c6-c5. This is
and King's Indians - there was never any the choice of both Yusupov and Kasparov,
time to spare! If I wasn't going forward all and it is featured in this game.
the time, then I was getting pushed back into 2. The riskier ...dSxc4 with ... e6-e5.
submission! The QGD is different. From the 3. The slower ...ltJb8-d7 delaying a central
start, Black has not conceded White any conurutment and reserving the right to break
central space and thus has managed to keep in the centre without a prior ... dSxc4. This
White's pieces at 'ann's length' from his idea is seen in Game 4.
position. Consequently, Black can afford a 11 ... dxc4
consolidating move or two because White is QIestion 13. Why does Black give up his
not 'close' enough to launch a major attack occupation of the centre in this way?
There are two main move orders at this An.9lW" 13. The precise reasons in this
point: 9...lZJxc3 10:.xc3 c6 has been played particular case will be explained later, but in
(transposing to the game after 11 i.d3) but general this is a typical idea By activating
Kasparov's preferred 9...c6 seems the most himself with ...c6-c5 or ...e6-e5, Black
natural, as there is little point in moving the inevitably weakens his protection of his dS-
knight until one is forced to do so. For pawn. Without a prior ... dSxc4, White gets
example, 10 tiJxe4 dxe4 11 tiJd2 f5 (11...e5!? the chance to play c4xdS and then d4xe51 c5
12 d5 [12ltJxe4? exd4 13 'iVxd4 :td8! wins a saddling Black with an isolated queen's pawn
piece] 12.. .f5!?) 12 c5 (intending ltJd2-c4-e5) (IQP). By abandoning his occupation of dS,
12... tiJd7 13 tiJc4 e5 is equal according to Black frees himself from protecting his
Beliavsky. central pawn which makes his central breaks
The fmal idea is the solid 9... tiJf6!?, but a lot easier to achieve. Moreover, by playing
White has a steady edge in all variations. For ... dSxc4, Black removes an obstacle from the
example, 10 'iVc2 (10 'iib3!? l:.d8 11 .1e2 a8-h 1 diagonal. Thus when Black
dxc4 12 'iVxc4 a6 13 0-0 bs 14 'iVb3 i.b7 15 fianchettoes his light-squared bishop on the
a4 b4 16 as! was a little better for White in long diagonal with ...b7-b6 and ....1c8-b7 and
Beliavsky-Short, Belgrade 1987) 10.. .ti)bd7 plays .. c6-c5, the black bishop will stand
11 cxdS exdS 12 i.d3 c6 13 0-0 l:te8 14 "bl, actively on a clear long diagonal.
intending b2-b4, as in Portisch-Kholmov, Note that Black only captured on c4 once
Kecskemet 1962. The move ...h7-h6 is a White's bishop had been developed to d3. In
defmite weakness when White plays into comparison to the straightforward 9...ltJxc3
12
Lasker Variation
10 l:xc3 dxc4 11 ~xc4, Black has gained the Arzsrrer 15. Yes, he can do that.
useful extra move ...c7-c6. 'This 'fight for the Qiestion 16. But ... hasn't White just
tempo' (making White's bishop take two achieved his aims now? You said that White
moves to reach the c4-square) is typical both wanted to completely occupy the centre with
in the QGD and queen's pawn openings in pawns on d4 and e4?
general. A rzsrrer 16. Yes I did but...
12 i..xc4 li::Jd7 13 0-0 Qlestion 17. Well, then Black's opening has
We have in fact transposed to a position failed!
from the QGD Orthodox which is usually Arzsrrer 17. Not so fast! This was White's
reached via 1 d4 dS 2 c4 e6 3 ttJc3 i.e7 4 early opening aim, but we are now in the
ttJf3 lLlf6 5 i..g5 0-0 6 e3 ttJbd7 7 l:c1 c6 8 early middlegame and the situation has
i.d3 dxc4 9 i.xc4 ttJdS 10 i..xe7 'ii'xe7 11 changed.
0-0 ttJxc3 12 lIxc3. The difference is the Qlestion 18. How?
extra move ...h7-h6 for Black on the Arzsrrer 18. The key point is that Black has
kingside. exchanged off two of his minor pieces. A big
Qiestion 14. Is this good for Black? pawn centre has one major strength: it can be
Anszrer 14. It depends on the set-up that used to brush aside the enemy pieces,
Black chooses. In the game, where White chasing them from their secure posts, thus
plays his queen and bishop to the b 1-h7 gaining territory for your own pieces while
diagonal, ...h7-h6 is very useful since White ruining the layout of the opponent's pieces.
does not gain a tempo by attacking a pawn For example, had Black not managed to
onh7. exchange pieces, we could have reached this
13 ... b6 type of position:
13 ... eS is the subject of Games 2 and 3.
Here we see the power of the pawn

Black quietly develops 'inside his shell'. centre. At every step, e4-eS is a dangerous
The move ...b7-b6 frees b7 for his light- threat, chasing the knight from its good
squared bishop; Black then only has to play defensive post on f6, while ...c6-c5 allows d4-
...c6-cS to complete his aims: he will have dS! ...e6xdS, e4xdS with the threat of dS-d6 .
found an active post for the bishop and he The conjunction of White's central strength
will have begun his central counterplay with with the fact that Black has too many pieces
... c6-cS. for the space he has available causes Black
Qiestion 15. Wait a minute. Am I going some problems. Now let's go back to our
crazy or can White just play 14 e4? position after 13... b6 14 e4.
13
development advantage, or to extract a

concession from his opponent.
The text preys on Black's temporary
weakness along the a8-h 1 diagonal (the c6-
pawn is undefended and the bishop is not yet
mobilised on this diagonal) by preparing to
transfer the light-squared bishop to the e4-
square to combine against the c6-pawn with
the rook on c3.
Now the obvious 14...i..b7 is strongly met
by 15 ..te4! (preventing ...c6-cS) IS .. J:Hc8 16
'ii'c2! followed by 17 :fc 1 with enormous
pressure against c6.
In this case White's pawn centre cannot 14 ... c5
interfere with Black's pieces as they are well Forced.
out of range and comfortable within their 15 i.e4!
space. Moreover, after 14...i..b7 15 :e1 c5! This disrupts Black's plan of development
16 dS exds 17 exds 'i'd6 by preventing .....tc8-b7. In this way, White
maintains a small initiative.
15 ...:b8
Instead lS ...i..a6 16 i..xa8 i..xf1 17 i..c6!
i..a6 (17...tbb8 18 'ii'xfl! tbxc6 19 dxcS bxc5
20 'iibS! is awkward for Black) 18 'i'a4! (18
i..xd7 'ii'xd7 prevents d4xc5) 18 ...ttJb8
(forced) 19 dxc5 bxcS and now 20 h3
(Dolmatov) or 20 i..bS leads to a pleasant
advantage for White due to the weakness of
the c5-pawn.
16"a4
Since the queen has to retreat back to c2
later, it is logical to investigate the old move
what is wrong with Black's position? 16 'ii'c2. However, this is less forcing and
Because he has exchanged two of his minor gives Black some extra possibilities. For
pieces, the advance of the d-pawn causes no example, 16...e5 (Dolmatov mentions
problems for the hannony of Black's pieces. 16.....ta6 17 :dl tbf6!? 18 dxc5 tbxe4 19
For example, there is no bishop on e7 facing 'ii'xe4 bxcS 20 b3 with a slight edge for
execution by the ds-d6 push. Black now White) is thematic: 17 dxcS (17 i..fS!? i..b7!
intends simply ...ttJd7-f6 ganging up on the 18 ..txd7 exd4 19 exd4 'ii'xd7 20 dxcS bxc5
ds-pawn. with countetplay as 21 lhc5 i..xf3 22 gxf3
14.td3! 'iib7! is fme for Black) 17... tbxc5 18 b4 (not
Qtestion 19. So what is White's idea then? 18 ..th7+ ~h8 19 b4 tba6!, intending ... g7-g6
A173Ue" 19. This is a crucial moment for to trap the bishop on h7) 18...liJxe4 19 'ii'xe4
White as Black is poised to complete his :e8 with ...i..b7 to follow is perfectly okay
opening mission with ...i..c8-b7 and ... c6-cS. for Black.
White has just a couple of moves in which to 16 ... .i.b7 17 .txb7 :xb7 18 "c2!
either realise an aspect of his slight space and This is Kramnik's move.
14
Lasker Variation
pawn on cst .
An.mer 21. Exactly. The pawn on c5 1S
attacked by the white pawn on d4. Combined
with White's rook on c3 and the queen on
c2, this pressure ties the black knight on d7
to the defence of cS-pawn. The obvious
course would be to release the tension by
...c5xd4. However, observe the effect after
ttJf3xd4. White gains total domination of the
c-file while his knight eyes the queenside light
squares c6 and bS. These factors cause Black
grave discomfort.
Qlestion 22. So what does Black want?
Q4estion 20. Black has swapped off his bad Ansu.er 22. Black's middlegame aim, now
bishop: isn't he just equal now? that he has fulfilled his opening plan, is to
Anszrer 20. It may appear so at first sight, neutralise this c-flle pressure. He has several
but in actual fact White still maintains a ways to attempt this:
small, stable plus. Black's practical results at 1. The ideal would be to play ... c5xd4 and
the highest level have been quite poor here. then block the c-fIle with ...ttJd7-cS. The
A few draws, an appreciable number of problem, however, is that this knight can
losses and quite a bit of pain for the Black easily be driven away by b2-b4.
player! 2. The advances ...b6-bS and c5-c4 would
The black light-squared bishop was bad release the pressure on the c-pawn and
because Black's central pawn chain (c6, dS, activate Black's queenside pawn mass.
e6) was all on light squares. Black's goal was However, this is extremely difficult to
to activate this piece in order to complete his arrange.
development. In the pursuit of this aim, 3. So the easiest to achieve his goal is to
Black had to loosen his pawn structure: he play ...e6-eS, to swap pawns on d4 and thus
gave up his pawn occupation of dS, he to open more flles. White's control of the c-
played his queenside pawns from the light flle only matters so long as it is the premier
squares c6 and b7 to the dark squares b6 and open fIle on the board If a number of others
cS, and finally he managed to get his bishop are opened, e.g. the e-file by ...e5xd4, then it
on the a8-hl diagonal. At that moment, the loses its value.
bishop ceased to be bad! It became a good Qlestion 23. I don't understand I thought
bishop due to Black's efforts and that is why that by exchanging pieces, I would just avoid
White exchanged it, just when Black was any problems!
about to reap the fruits of his endeavours! Arl3ta'r 23. The exchange of pieces has
Ironically, Black's queenside structure is now made you safe. By swapping off pieces, Black
slightly weak without this bishop. The neutralised any of White's aspirations for a
queenside light squares on c6, b5 and a6 are quick kingside or central attack. The flip side
targets for both White's queen and his is that by making himself safe, Black has also
knight. robbed himself of his potential to create
White's claim for an advantage lies in the trouble by stirring up counterplay: he just
combination of Black's weak light squares doesn't have enough pieces for the job.
and the problem Black has with the c-file. Consequently, he has to continue as he
QIestion 21. What problem? He's got a started: neutralising White's initiative.
15
Qiestion 24. So what was the point of 18 achieve its objective of sidelining the black
'iWc2? knight.
An.mer 24. By removing the queen from 21 ...:tbdS
a4, White side-steps any attempt from Black 21. ..e5 22 dxe5 liJxe5 23 ltJxe5 'iWxe5 24
to play either ...b6-bs and ...cS-c4 or ...cSxd4 :cd3 gives White control of the only open
and ...ltJd7-cS with tempo. Moreover, as file and a slight advantage according to
Kramnik points out, White actually prevents Karpov.
the freeing break 18... eS due to 19 'ii'e4! 22 :'cd3 :'cS
White also gains the threat of 19 dxcS ltJxcs 22... cxd4 23 :Xd4! (23 liJxd4ltJe5! is fme
20 b4!, driving the knight back from its for Black) 23 ...ltJc5 24 b4 axb4 25 axb4lDa6
desired post. 26 1i'c4 is good for White according to
1S ... a5!? Karpov due to the poorly-placed black
An improvement on 18...:c8?! 19 :c1 knight on a6.
:bc7 20 b4! eS (20...c4 21 bS a6 22 ltb1 and 23 d5! exd5 24 .l:xd5lLlf6 25 :'e5! 'iNc7
20...ltJf6 [intending ...ltJf6-dS] 21 e4! are 25 ...'i'b7 26lhe8+ :Xe8 27 a4 'ii'e4 was a
clearly better for White according to more active defence according to Yusupov.
Kramnik) was the continuation in Kramnik- 26 :'xeS+ .l:xeS 27 a4!
Kasparov, Las Palmas 1996, and now 21 Here Karpov claims a clear advantage.
bxcS! exd4 22 exd4 bxc5 23 :c4! ltJb8 24 Qtestion 26. Why?
:XcS :xc5 25 dxc5ltJa6 26 c6ltJb4 27 'i'a4 Ansrrer 26. In effect, White is almost a
was winning for White according to pawn up. Black's queenside pawn majority is
Kramnik. powerless to expand as it is tied down by the
Qiestion 25. I don't understand this a4 pawn. Moreover, the queenside structure
18...aS!? move. is weak: the as- and c5-pawns are held up by
A~ 25. With this move, Black takes a pawn on b6 that is a perfect target for a
control of b4 in order to prevent White from knight on c4. White's kingside majority has
driving away the black knight with b2-b4 no such impediments and so it is much easier
when it comes to c5. The downside is that it for him to create a passed pawn than for
further weakens Black's queenside structure. Black.
19 a3! I understand that I am talking very
This typical move renews the possibility of breezily about something that is incredibly
b2-b4 in response to ...c5xd4 and ...ltJd7-c5. subtle and requires the highest level of
19 ...:'eS!? technique. Yusupov is one of the best
Black wishes to use plan 3 above (the ...e6- endgame players in the world, but Karpov
e5 break) and thus protects his queen in makes this position look like a forced win!
order to negate White's possibility of 1i'c2- When considering whether to play a variation
e4. like this, you have to consider the strength
20 :'d1! l:bbS and inclinations of your opponent. An all-out
Since 20...e5 is met by 21 'ife4 exd4 22 attacking player would not like the white
'ii'xb7 dxc3 23 bxc3lDe5 24 'i'xb6ltJxf3+ 25 position after move 18 and would be unlikely
gxf3 1i'g5+ 26 ~f1!? with a clear advantage to cause many problems. However, if you do
according to Karpov. get the chance to be Black against Karpov,
21 h3 don't tly this line!
Removing any back-rank tricks. As 27 ...:'dS 2S :'xdS+ 'iNxdS 29 ltJe5 'fNd5
Karpov shows, the hasty 21 dxc5 ltJxc5 22 30 lbc4 lLld7 31 b3 f5 32 ~f1 ~f7 33 f3
b4 axb4 23 axb4 ltJa6! 24 b5 ltJcs does not ~e7 34 ~e2 1i'e6 35 1i'c3 ltJf6 36 ~f2
16
Lasker Variation
~d7 37 g4 ~c6? 38 .e5! 'ii'xe5 39 dxc4 12 i..xc4 ll)d7 13 0-0 e5

ll)xe5+ ~d5 40 ll)c4 fxg4 41 ll)xb6+!
In contrast to 13 ...b6, 13...e5 strikes

41 ...<li'c6 42 ll)c4 gxf3 43 ~xf3 ~d5 44 immediately at White's centre and reopens
ll)xa5 g5 45 ll)c4 h5 46 ll)d2 ~e5 47 e4 the c8-h3 diagonal for his light-squared
ll)e8 48 ~e3 ll)c7 49 ll)c4t <it>f6 50 ~f2 bishop on c8. However, Black still has a
ll)a6 51 <it>g3 lLlb4 52 h4 ll)c6 53 a5! development problem.
lLlb4 54 lLld2 lLlc6 55 a6 gxh4t 56 ~xh4 Qiestion 28. What do you mean?
~e6 57 <it>xh5 <it>d7 58 <it>g6 <it>c7 59 lLlc4 Anszrer 28. To complete his development,
<it>b8 60 <it>f6 <it>a7 61 e5 ~xa6 62 e6 Black must involve his light-squared bishop
~b5 63 e7 1-0 and queen's rook in the game. However, the
Qiestion 27. I'm a bit confused. Is this a light-squared bishop is blocked by the knight
good or a bad variation? on d7, which cannot move without leaving
An.suer 27. Karpov wrote the following: the e5-pawn undefended. Consequently,
'Of course anyone who chooses Lasker's Black must release the central tension before
Defence is hardly in danger of earning the completing his development which is a
whole point and must be prepared for a concesslon.
prolonged defence in the battle for a half Qiestion 29. Why is that?
point.' TIlls sums up how the top players feel Anszrer 29. The longer you can keep the
about facing it: they don't know whether they central situation unresolved, the longer you
will win, but they do hope to at least make keep the opponent guessing, and so the less
you suffer! At a lower level, however, where time he has to prepare himself for your
the level of technique is less exalted, this is a eventual plan.
useful line to have in your repertoire, 14 i..b3
particularly against an aggressive all-out A typical Karpov move! After releasing
player who doesn't like endings! the central tension, Black's most natural
continuation is ... ttJd7-b6 to free the bishop
Game 2 on c8, while gaining a tempo attacking
Karpov-Yusupov White's bishop on c4. 14 i..b3 anticipates
London {8th matchgame} 1989 ...ttJd7-b6 and waits for Black to commit
himself.
1 d4 lLlf6 2 c4 e6 3 ll)f3 d5 4 ll)c3 i.e7 There are three basic central scenarios:
5 .tg5 0-0 6 e3 h6 7 .th4 ll)e4 8 i..xe7 1. Black plays ... e5-e4
-'xe7 9 ':c1 c6 10 .i.d3 ll)xc3 11 l:.xc3 2. Black plays ...e5xd4.
3. 'White captures on e5. In the game, Black chose a more

Qlestion 30. 14...e4 15 ttJd2 looks dangerous option. The more solid 14...:e8 is
aggressive for Black! the subject of the next main game.
Anszrer 30. This is slightly deceptive. The 14... exd4 15 exd4!
move ...e5-e4 does have the idea of starting a
kingside attack: it drives away the knight
from f3 and weakens White's defence of his
kingside, in particular the hZ-square.
However, in order to exploit such a
weakness, Black really needs a dark-squared
bishop raking along the b8-h2 diagonal,
opening up the possibility of ... ~d6xh2+
sacrifices, for example. Without this piece,
Black does not have the firepower to attack
on the kingside. His position consequently
lacks flexibility which promises White a small
stable advantage.
Qetion 31. What will White aim for? Qetion 33. What? Are you sure about
A nszrer 31. White has several typical plans: this?
1. f2-f3 removing Black's centre pawn. Armrer 33. Absolutely! White's voluntarily
White will aim to advance his e-pawn and to accepts an IQP for two reasons:
use the half-open f-fUe. 1. Black's temporary headache is his
2. f2-f4 to close the kingside completely development: his knight blocks his light-
and to thus remove any lingering hopes squared bishop which in tum imprisons his
Black might have of an attack there. queen's rook. Once this problem is solved,
3. The queenside minority attack with b2- White will have nothing. Consequently White
b4-b5. must open lines and 'get at' his opponent
Qlestion 32. 14 J..b3 doesn't seem to help before Black can develop. The text fulfils this
with the last idea! task brilliantly: the rook on f1 will come to e1
Anszrer 32. True. The extra point of 14 with a tempo on the black queen, and the
J..b3 is stated by Karpov who notes: ' ...if the rook on c3 can swing across to f3, g3 or h3
centre is blocked by 14...e4, White has once the knight occupies the outpost on e5
already cleared the c-file.' lbis allows White that the pawn on d4 provides. 15 exd4!
to harass his opponent on the dark squares dramatically increases the activity of the
and to give him serious problems defending White position.
the e-pawn via a later :c3-c5-e5. For 2. IQP structures where Black has a pawn
example, 15 ttJd2 Ci1f6 (1S .. 53ih8 [intending on c6 rather than e6 are generally favourable
...f7-f5] 16 'iWh5! Ci1f6 17 'iWh4 followed by for White. This is known from the analysis of
f2-f3 is annoying for Black) 16 'i'c2 J..g4 the QGA line 1 d4 d5 2 c4 dxc4 3 e3 eS 4
(16 ... J..e6 17 Ci1xe4!) 17 :b1!? ~e2 18 i.c4 J..xc4 exd4 5 exd4 Ci1f6 6 tiJf3 i.e7 7 0-0 0-0
..txc4 19 lhc4 followed by b2-b4-b5 gives 8 ttJc3
White good chances. White should also aim
to exchange queens: this removes Black's
best defender of his dark squares and and the same factors apply here. Without
forestalls any possible hope of a kingside the cover of a pawn on e6, the black f7 -pawn
attack for Black. is exposed to the combination of a bishop on
18
Lasker Variation
the a2-g8 diagonal and a knight on the eS for 2S 'iWxfS fxeS to hide the black king
outpost. Moreover, the e-file is more useful behind the white g-pawn would have been
as an attacking me than the c-file, and refuted by Karpov's fantastic 2S gxf6+!!
although Black can use it for exchanges, it is
also a valuable entry channel for White.
2S .....txg4 26 lhg4+ ~h8 27 tiJf7+ ':xf7

28 ':xeS+ ':f8 29 f7 tiJf6 30 ':xfS+ 'iWxfS 31
Qiestion 34. How about Black's extra ... h7- ':gS+ tiJxg8 32 fxg8'iW + 'iWxgS 33 ..txgS with
h6move? a winning endgame!
Anszrer 34. As mentioned earlier, this 25 g6 i..xg6 26 dxe5 'ile6 27 i.xd5 cxd5
position is a direct transposition to a line of 28 'iixg6+ 'iixg6 29 :'xg6+ ~h 7 30 :'d6
the Orthodox QGD with the extra move :'c8 31 :e3 :'c2 32 :'d7+ ~g6 33
...h7-h6 included Here, this is a distinct :xb7+- fle8 34 a3 d4 35 :d3 flxe5 36
disadvantage for Black. With the knight on :'xd4 :95 37 :d6+ ~h5 38 :'h7+ ~g4
eS, the bishop on b3 and a possible queen 39 :d4+- ~f5 40 fld5+ ~g6 41 fl97+
excursion to the kingside, White will exert ~xg7 42 :'xg5+ ~f6 43 :b5 a6 44 :'b6+
severe pressure against Black's kingside light- ~e7 45 ~h2 ~d7 46 ~h3 ~c7 47 :'b3
squares. The move ...h7-h6 weakens g6 and ~d6 48 g4 ~e5 49 ~h4 ~f6 50 rlb6+
thus Black's whole kingside structure. ~g7 51 ~h5 a5 52 flb7+ ~98 53 a4 1-0
15 ... ~f61? 16 :e1 'ii'd6 17 ~e5 ~d5 This fantastic game has caused 14...exd4
Beliavsky suggested 17... i..e6 18 ..txe6 to disappear from tournament play.
fxe6 19 'iWb3 'i'xd4 20 'i'xe6+ ~h7 with Although Black may be able to play a little
equality in ECD, but Makarichev's 19 ':g3!? more accurately, it is clear that White enjoys a
is quite annoying for Black due to the very dangerous initiative.
weakness of his kingside. Qiestion 35. One thing puzzles me: does
18 :g3 .tf5 Black really have to rush with 14...exd4 or
Risky. 18.....te6 was more solid though 14... e4?
Zaitsev's 19 'i'dl gives White a pleasant An.mu 35. This is where the third scenario
initiative. Black can never drive the white comes in!
knight from eS with .. .f7-f6 due to the
weakness of g6. Game 3
19 'ii'h5! i.h7 20 'ii'g4! g5 21 h4 f6 22 Cifuentes Parada-Korneev
hxg5 hxg5 23 f41 Aae8 24 fxg5! fxe5 Malaga 1998
This leads to a winning endgame for
White. The more cunning 24.....tfS!? aiming 1 d4 d5 2 c4 e6 3 ~f3 ~f6 4 ~c3 i..e7
. 19
5 ~g5 h6 6 i..h4 0-0 7 e3 lLle4 8 i..xe 7 White's idea is vel)' simple: his next move
'ifxe 7 9 ':c 1 c6 10 .i.d3 lLlxc3 11 l:lxc3 is f4-fs. This has two strong points:
dxc4 12 .i.xc4 lLld7 13 0-0 e5 14 .i.b3 1. White stops ....i.c8-fs and thus prevents
:e8 Black from actively completing his
development.
2. White intends fS-f6 with a strong attack.
Again, we have the transposition to a
QGD Orthodox line (with the extra move
...h7-h6) This line was thoroughly tested in
the 1930s and 1940s and in this case Black
has stumbled into an inferior line. After
17.....e4 (17...'ii'f6 18 fS! ':d8 [to develop the
bishop with ...i..c8-d7] 19 l:.d3! gives White a
huge advantage as does 17.....e7 18 f5! i..d7
19 f6!) 18 fS! Black has big development
problems as 18...i..xfs loses to 19 i..c2!
Q6t:ion 3 7. Wait a minute, couldn't White
This is a vezy reasonable idea Black wants play 14 t'Dxes instead of 14 i..b3?
to play ...esxd4 and then transfer his knight A7l3Zre" 37. He could indeed. This is a
to the solid defensive square f8, covering the much better version for Black however.
weak g6-square. Then... i..c8-e6 will follow to After 14 ttJxeS t'Dxes 15 dxe5 'ii'xeS 16 f4,
swap off the light-squared bishops. The 16...'ii'e4!
14...l:.e8 idea was ftrst played in P NIkolic-
Yusupov, Belgrade 1989 Gust after
Yusupov's match with Karpov} when 15 dS
cxdS 16 'ii'xdS ttJf6 17 'ii'cs ttJe4 18 "xe7
':xe7 19 lk2 ttJgS! gave White nothing. As
you will see, the game continuation was not
too inspiring for White either.
Q6t:ion 36. Wow! So what can White do?
Ansla'r 36. This is the time to go into the
third scenario and play 15 ttJxeS ttJxes 16
dxes 'ii'xeS 17 f4!
is best. Since White does not have the

immediate i..b3-c2, White must fIrst protect
e3 to drive the queen from e4. The standard
line is 17 'ii'e2 (mtending i..c4-d3) 17...i..fS!
18 i..d3 'ii'dS 19 e4 "d4+ 20 'ii'f2 (20 ~h1
l:.fe8 is fme for Black) 20.....xf2 + 21 ~xf2
i..d7 22 l:.d1 l:.fd8 23 i..c4 i..e8 with a small
but not very exciting edge for White.
The only other attempt I have seen from
Black is 14...l:.d8 in Zakharevich-Bezgodov,
Penn 1997, when 15 :e1 exd4 16 exd4 'ii'd6
20
Lasker Variation
17 d5 cxdS 18 'ii'xdS'ii'xd5 19 i..xd5lDf8 20 5 i.g5 h6 6 i.h4 0-0 7 e3 lDe4 8 i.xe 7

i.b3 led to a draw. 17 :ce3 lDf8 18 lDeS is 'ilxe7 9 l:tc1 c6 10 i.d3 lDxc3 11 :xc3
much stronger (but not 18 1:1e7?? lDe6! 19 lDd7!?
i.xe6 fxe6 which is rather embarrassing)
when 18...i..e6 19 .txe6 (19 ll'lxf7 i.xf7 20
i.xf7+ ~ 21 'iWb3+ 'it>g6 [21... 'it>f6 22
1:1f3+ 'it>g6 23 1i'c2+ is curtains] 22 1:1e6+
ttJxe6 23 :1xe6+ 'iWxe6+ 24 "i'xe6+ 'it>h7 is
flne for Black) 19...fxe6 (19 ...ll'lxe6 20 ttJxf7
'it>xf7 21 lhe6 'ii'xe6 [21.. ...xd4 22 "b3!] 22
lhe6 <it>xe6 23 trb3+Q 20 'ii'h5 is powerful.
15 h3 exd4 16 exd4 tDf8 17 d5 cxd5 18
i..xd5 %-%
My final thought is the try 13 ... cS!?
Qetion 38. What is the point of this move

order?
Amrrer 38. Black will meet 12 0-0 with the
immediate 12...eS!, threatening ...eS-e4. After
13 dxeS, Black plays the intennediate move
13 ...dxc4!
As far as I can see, this is not mentioned

in any reference book. I just remembered it
from a book I read when I was 10 years old:
7he Road to Ox:ss Mastery. That game occurred
via the Orthodox move order and White
caused trouble with i.c4-d3, "i'd1-c2 and
later lDf3-gS gaining time against the h7-
pawn. Here, with the pawn already on h6,
Black side-steps all these problems so the After 14 i.xc4 liJxeS, Black has
idea may be worth a go. It is a vezy flexible transposed back into the 13 ...eS line having
idea: Black can either capture on d4 and play side-stepped Karpov's dangerous 14 i.b3
...liJd7-b6/f6 or he can switch back to the line! White's only other attempt is 14 :txc4
...b7-b6 plan if necesscuy. lDxe5 15 :te4, but after 15 ...tDxf"3+ 16 ..xf3
r---------------.
Game 4
J.e6 17 .tc4 :ad8 18 J.xe6 fxe6 19 We2
:d5 White has absolutely nothing: a draw
P.Nikolic-Vusupov was agreed in Sadler-Kramnik, Tilburg 1998.
Horgen 1994 12 cxd5! exd5 13 0-0
The position of the pawn on h6 gives
1 d4 tDf6 2 c4 e6 3 lDf3 d5 4 lDc3 i.e7 White a pleasant edge in this typical position.
71
lila1+ 38 ~h2 :a2 39 ~g1 :a1+ 40 ~h2

lilf1 41 :xh4 nxf2 42 ~g1 e3 43 d5 ltd2
44 ~f1 :f2+ 45 ~e 1 :xg2 46 ltf4 ~f6
47 :f3 :d2 48 :xe3 :xd5 yz-yz
We will now examine the currently less
popular White choices.
Game 5
Kramnik-Lutz
Germany 1994
1 tDf3 d5 2 d4 tDf6 3 e4 e6 4 ltJe3 i.e 7
5 1i.g5 h6 6 1i.h4 0-0 7 e3 ltJe4 8 1i.xe7
Qtestion 41. Why? 'ilxe7 9 exd5 ltJxe3 10 bxe3 exd5
Anszrer 41. Black's knight will have to
move to let the bishop on c8 develop. When
that happens, White's knight can move to e5.
It will be very hard to dislodge with .. .f7-f6
due to the weakness of g6. For example, after
13 .. /~Jf6 14 ttJeS! (Yusupov) 14... ttJd7 15 f4!
f6, 16 tLlg6! ~xe3+ 17 ~h1 :e8 (17 ...~xd4
18 ttJe7+ ~h8 19 'iihS! is very dangerous for
Black) 18 ttJeS!! is extremely strong: White
threatens both i-d3-h7+ and i-d3-b1
trapping the queen! Consequently, Black has
to be careful.
13 ...:e8 14 'iib1 tDf6
14... aS!? stops b2-b4 but weakens the In this w'J:)', White gets to pl'J:)' the
queenside and sets up the a-pawn as a target Queen's Gambit twice! White intends c3-c4
for White's major pieces, for example with to exchange Black's ds-pawn and remove the
:c3-a3 intending b2-b4. brake on his central expansion. Moreover,
15 b4 White gains the semi-open b- and c-files on
15 ttJeS is also possible. which he can pressurise Black's queenside
15 ... tDe4 16 j.xe4 dxe4 17 ltJd2 j.e6 18 pawns.
.l::Ue1 i.d5 19 b5 :ad8 20 a4?! Qlestian 40. Just sounds good for White!
Intending as-a6, but this is rather slow. 20 Ansrrer 40. This line does have the ring of
bxc6 bxc6 21 :c5! intending ttJd2-c4-eS logic about it! However, the corollary to
attacking the weak c6-pawn would have removing Black's centre pawns is the space
given White a slight advantage according to Black gains in which to activate his pieces.
Yusupov. For example, without the e6- and dS-pawns,
20 ... h5! 21 :1e2 h4 22 h3 f5 23 bxe6 the light-squared bishop gains the open c8-
bxe6 24 tDe4 e5! 25 dxe5 'iixe5 26 tDa3 h3 and the a8-h 1 diagonals. Moreover, due to
'iib6 27 'iixb6 axb6 28 ltJb5 1i.f7 29 tDd4 the semi-open e- and d-fues it is very difficult
g6 30 :e6 :e7 31 :b2 :a8 32 :b4 for White to use his central pawns as a
:ea7 33 :bxb6 :xa4 34 :xg6+ 1i.xg6 positive force by pushing them forwards.
35 :xg6+ ~f7 36 nh6 nxd4 37 exd4 Meanwhile, Black fmds it much easier to
22
Lasker Variation
organise ...c7-cS to 'dilute' the white centre ~d4 b5 35 h4! b4 36 :b6 a5 37 ltb5
with ...cSxd4. Black's activity and White's 4:Jd3 38 ltxa5 4:Je1+ 39 'itrg3 ttJc2 40
static central pawns seem to allow Black to 4:Jb3 4:Ja3 41 lta4 ltc4 42 4:Jd4 ~c2 43
hold the balance. 4:Jf3 ltc5 44 :a7 g5 45 h5 ~g8 46 4:Jd2
11 'iib3 4:Ja3 47 lDe4 ltc2 48 lIb7 ltxa2 49 :xb4
White gains a tempo against the d-pawn ltc2 50 :b6 ~h7 51 l:.b7 ~g8 52 liJd6
while tying down the light-squared bishop to l:.c6 53 ~xf7 ~c4 54 1I1d7 :f6 55 1I1d4!
the protection of h7. 1-0
11 ...:d8 12 c4 dxc4 13 i.xc4 ttJc6!
Threatening both .. /~Jc6xd4 and ...ttJc6- Game 6
as. Karpov-Yusupov
London (6th matchgame) 1989
14 i.e2
The alternative 14
the next main game.
"c3 is considered in
1 d4 ttJf6 2 c4 e6 3 ttJf3 d5 4 ttJc3 .i.e 7
14... b6! 5 i.g5 0-0 6 e3 h6 7 i.h4 ~e4 8 i.xe7
An excellent plan according to Kramnik. 'ii'xe7 9 cxd5 ttJxc3 10 bxc3 exd5 11
Black activates his bishop on the long 'fib3 :d8 12 c4 dxc4 13 i.xc4 ttJc6 14
diagonal and prepares to free his position 'iic3
with ...tl'lc6-aS and ...c7-cS. The alternative way of preventing ...tl'lc6-
15 0-0 i.b7 16 1I1ac1 ttJa5 17 'ii'b2 1I1ac8 as.
14 ... i.g4 15 0-0
Unfortunately White cannot prevent the
doubling of his f-pawns, since 15 .te2 fails to
15.. .txf3 16 .txt3 tl'lxd4 with a strong
attack.
15 ...i.xf3 16 gxf3 'fif6 17 .i.e2 ltac8!
Kramnik considers the position equal

here. Unfortunately that doesn't guarantee a
draw against Kramnik!
18 h3 c5 19 dxc5 ':xc5 20 1I1xc5 'ii'xc5
21 1I1c1 'iie7 22 ttJd4 'iig5 23 i.g4 'ii'd5
24 .i.f3 'iid7 25 i.xb7 liJxb7 26 ~c6
1I1a8 27 'iid4 ttJc5? A world-class move. Black prepares ...b7-
Kramnik feels that Black should keep the b6, ...tl'lc6-e7 and then the typical ...c7-c5 to
queens on here with 27.....e6! 28 "c4 "e8! 'prune' White's centre. The position bears a
with ...tl'lb7 -cS to follow. He now ruthlessly distinct resemblance to the Chigorin Defence
prosecutes a small advantage. (1 d4 d5 2 c4 tl'lc6).
28 'iixd7 ttJxd7 29 :d1 ~c5 30 g4 g6 18 %:tab1 b6 19 %lfc1 4:Je7 20 ~h1 fLd5!?
31 ~g2 ~g7 32 lId2 a6 33 ':d6 :tc8 34 21 -.c2 -.h4! 22 f4!!
)
23
This excellent defensive move is the only 120-0

way to deal with the threat of ...:d5-hS.
White just manages to hold the balance, but
12 "a3 ttJd7 13 ~b5 b6 14 ~c6 :b8 15
0-0 (15 'iixa7 cxd4!) is slightly better for
Black has all the chances. White according to Beliavsky, but 15... i.b7
22 .. :ii'xf2 23 J..g4 'ii'xc2 24 :xc2 f5 25 16 ~xb7 :lxb7 looks like a safe version of
J..f3 l:td7 26 :bc1 tLld5 27 J..xd5+ :xd5 the 9 ~ellines as a quick ...e6-e5 will follow.
28 l:txc7 l:txc7 29 l:txc7 :a5 30 d5 ~f8 12... cxd4!?
31 d6 ~e8 32 ':xg7 l:txa2 33 ~g1 a534
12...ttJd7 is more sensible, intending either
l:te7+~d8 35 e4 fxe4 36 l:tb7 e3 37 ~1
...b7-b6 or ...c5xd4. For example, 13 !':tael b6
14 ~b5 ~b7 15 ~xd7 ~xf3 16 gxf3 "xd7
a4 38 ':xb6 a3 39 ':a6 l:tf2+ 40 ~e 1 a2
41 f5 ~d7 42 f6 ~e6 43 ':a8! ~xd6 44
17 dxc5 bxc5 should be fine for Black as 18
f7 :xf7 45 ':xa2 ~c5 46 :a6 V2 -V2
'ii'xc5 'ii'b7! regains the pawn.
--------------. Strangely enough 12...ttJd7 actually
Game 7 transposes to the game Polugayevsky-
P.Nikolic-Lputian Yudasin, Groningen 1993, with the extra
Yerevan Olympiad 1996 (helpful) move ...h7-h6. That game ended
r.-_ _ _ _ _ _ _ _ _ _ _ _ _• quickly in a draw after 13 dxcS ttJxc5 14
1 d4 d5 2 c4 e6 3 tLlc3 i.e7 4 tLlf3 ttJf6 !':tfdl ~d7 15 ttJeS :lfc8. The actual move
5 J..g5 h6 6 J..h4 0-0 7 e3 ttJe4 8 i.xe 7 order in this game was a Lasker hybrid -
'ilixe7 9 'ii'c2 5... ttJbd7 6 e3 ttJe4 7 i..xe7 'ii'xe7 8 "c2
In contrast to 9 :lel, White plays to place liJxc3 9 'iixc3 dxc4 10 ~xc4 c5 11 0-0 0-0.
his queen rather than a rook on c3. Although 13 ttJxd4 i.d7 14 'iib3 e5 15 ttJe2
the queen discourages ...e6-e5 plans, it is less Perhaps 15 'iixb7!? exd4 and now not 16
effective at stopping ...c7-c5 plans. 'ii'xaS ~c6! but 16 i.d5! (Fritz) is crucial.
9 ...ttJxc3 10 'ilixc3 dxc4 11 J..xc4 15... lDc6 16 i.d5 ttJb4 17 ttJc3 tLlxd5 18
Lputian-Vaganian, Yerevan 1996, saw the ttJxd5 ~-~
interesting 11 'iixc4 b6 (11...c6 intending
... ttJb8-d7 and ...e6-eS is vel)' sensible, now Lasker Hybrids
that the queen has moved from c3) 12 :lel A major weapon on the Black side of the
c5 13 dxcS i..a6 14 'ii'h4 "xh4 15 liJxh4 QGD is move order, and it is one that all the
:c8 16 ~e2 bxcS 17 liJf3 ~b7 18 0-0 ttJd7 leading experts use to confuse their
with a small edge for White. opponents. The essential Lasker move is
11 ... c5! ... ttJf6-e4. It is typically played after castling
and after ...h7-h6. Aside from the main lines
that we have looked at, there are also three
variants that all use the trademark move
... ttJf6-e4, but in slightly different settings:
1. Black plays ...0-0 without ...h7-h6.
2. Black plays ...liJbd7 without ...h7-h6.
3. Black plays ...liJbd7 with ...h7-h6.
Black plays ... 0-0 without ... h7-h6

QIestion 41. I meant to say! You said so many
times that ...h7-h6 was a weakness!
Amzrer 41. The 'natural' continuation for
White leads to a direct transposition to an
24
Lasker Variation
Orthodox QGD: 5...0-0 6 e3 ttJe4 7 ~xe7 9.. .fS is possible here, but this is not a
~xe7 S l:c1 c6 9 i.d3 ttJxc3 10 ':xc3 dxc4 great Dutch for Black.
11 i.xe4 ltJd7 12 0-0 eS is the same as 10 :xc3 dxc4 11 :xc4 lbd7 12 0-0
6...ttJbd7 7 l:c1 c6 S i.d3 dxc4 9 i.xc4 ttJdS Natural, but there is another idea in this
10 i.xe7 ~xe7 11 0-0 tLlxc3 12 ':xc3 eS. position: 12.e2!?
Although Karpov seems happy to play this as QIestion 43. What's the point?
White, there are more critical tests of the Ansrrer 43. In the game, Dizdar met
Orthodox, so it is natural for White to look White's pressure against h7 with ... g7-g6
for something better. rather than ...h7-h6. He did this in order not
QIestion 42. So what's he got? to weaken the kingside dark squares so that if
Amrw- 42. I'm glad you asked me that! White ever played ttJf3-eS, he could still meet
White has two ideas: it with .. .f7-f6. However, in the game
1. 7 i.f4!? Griinfeld-Vanden Bosch, Amsterdam 1936,
after 12 'ii'c2 g6, White played 13 ttJeS!?
preventing the freeing ...e6-eS. After
13 ...ttJxeS 14 dxeS, not only are Black's
kingside dark squares weak, but White has
the makings of a dangerous attack with h2-
h4-hS! If Black plays 12...h6 then 13 ttJeS is
less effective (though still possible) but after
13 0-0 eS 14 l:e 1, White has a slightly better
version of Portisch-Dizdar!
12 ... e5 13 'i'c2 g6 14 :e1 ltJb6
As Dizdar points out, the obvious
14... l:eS fails rather embarrassingly to 15
ttJxeS ttJxeS 16 dxeS 'ii'xeS 17 l:e4! winning
Without ...h7-h6, ~gS-h4 White can foil a rook! He also suggests 14...l:dS!?
Black's plan of exchanging the dark-squared 15 :c5 ltJd7 16 :c3 :eS
bishops, without having to give up the This is possible now that the rook has
bishop pair. There are no recent examples of been chased back to c3.
this move between strong players but it looks 17 .tf1 e4 1S ltJd2 tDf6 19 :c5 tDd5 20
very reasonable. .tc4! b6! 21 .txd5 cxd5 22 :xd5 16!
2. 7 iJ..xe7 .xe7 S l:c1 c6 9 i.d3 ttJxc3 10 Dizdar claims compensation for the pawn
:Xc3 dxc4 11 :Xc4 for Black as the rook is surprisingly short of
White's idea is to gain a tempo for squares. The position is extremely complex.
development with ~c2, hitting h7. This may 23 'i'xe4!? 'ii'xe4 24 ltJxe4 :xe4 25
be enough to tum the line in his favour, as lldS+ ~f7 26 :c1 :eS 27 :c7+ ~fS 2S
we see in this next game. :xe8+ ~xe8 29 :xh7 a5 30 h4 a4 31 f3
r--------------... :a5 32 :hS+ ~d7 33 :gS :b5 34 :xg6
Game 8 'it'e6 35 1:g7 i.d7 36 e4 :xb2 37 h5 :c2
Portisch-Dizdar 38 d5+ ~d6 39 :gS :c7 40 :dS ~e5
Sarajevo 1986 41 d6! l:1c1+ 1-0
1 d4 lLlf6 2 c4 e6 3 ltJf3 d5 4 ltJc3 i.e7 Black plays ...ltJbd7 without ... h7-h6
5 .i.g5 0-0 6 e3 ltJe4 7 .i.xe7 'i'xe7 8 QIestion 44. What is the point of an early
:c 1 c6 9 .i.d3 lbxc3 ...ltJbS-d7?
25
A1l3ZI8" 44. In general, the difference Black plays ...lL'lbd7 with ... h7-h6
between ...0-0 and ...tDb8-d7 is not so great. lbis will be very similar to normal lines.
For example, after S...tDbdl 6 e3 ltJe4 7 After S...h6 6 i..h4 ltJbd7 7 e3 ltJe4 8 i..xe7
i..xe7 (7 i..f4 is again possible: note that (8 i..g3!?) 8.. :iWxe7 9 :tel, the likelihood is
7... gS!? 8 i..g3 hS 9 cxdS! ltJxc3 10 bxc3 exdS that the game will transpose to the main
[10...h4 11 dxe6 fxe6 12 i..eS!] 11 h4 is good lines. An independent continuation for White
for White) 7.. :ilxe7 8 :tel tDxc3 9 :txc3 c6 was seen quite recently.
(9 ...dxc4 10 i..xc4 cS 11 dxcS ltJxcs 12
i..b5+! forces the awkward 12...tDd7 as Game 9
12,..iLd7 13 :XcS wins) 10 i..d3 will Sadler-Short
transpose into Portisch-Dizdar. British Ch. playoff, Torquay 1998
1 d4 e6 2 c4 ttJf6 3 liJf3 d5 4 ttJc3 i..e7
5 .i.g5 h6 6 ~h4 tbbd7 7 e3 tbe4 8
i..xe7 'iixe7 9 tbxe4!? dxe4 10 ttJd2 f5
11 'iih5+!
The point. Since Black cannot now play
...g7 -g6, he must allow an exchange of
queens.
11 ......f1 12 'iixf7+ ~xf7 13 f3
A new move although this is a suggestion
of Korchnoi's. 13 cS has been played until
now with good results for Black. After 13 ...eS
14 i..c4+ 'it'e7 150-0-0 exd4 16 exd4ltJf6 17
QIestion 45. I thought that White could :the1 :td8 18 dS b5!? the position was un-
not get the advantage in these lines! clear in Korchnoi-Andersson, Brussels 1988.
Answer 45. The difference is that Black has 13 ...exf3 14 liJxf3
already committed his knight to d7. 14 exf3 eS 15 f4!? was assessed by
Remember that Black was playing ... tDb8-e6 Korehnoi as slightly better for White. The
in the main line. Consequently, Black's most text is more modest, but keeps a small
active lines are cut out here. initiative. Just please avert your eyes around
After 10 'ii'b3, there are not really any move 33 ...!
presentable games in this move order. 14 ... b6 15 i..d3 i..b7 16 0-0 g6 17 e4
However, 10...liJf6 11 c4 c6 12 i..d3 iLe6 13 r:Ji;g7 18 exf5 exf5 19 d5 c6 20 dxc6
0-0 :tc8 14 :tabl :te7 15 cxdS i..xdS 16 'ifa4 .i.xc6 21 ~d4 i..b7 22 b4 1:he8 23 1:fd1
0-0 was played by Andersson against ttJe5 24 i..f1 1:e7 25 c5 bxc5 26 bxc5
P.Nikolic at Leningrad 1987 (with an extra 1:f8 27 1:ac1 Itf6 28 i..b5 ttJc6 29 a4 a6
...h7-h6) and is assessed in Infarmator as 30 i..xc6 i..xc6 31 lL'lxc6 1:xc6 32 1:d6
unclear. In general, I feel that this type of 1:ec7 33 1:e1?? 1:xd6 34 cxd6 :d7 35
position favours White slightly. Note that 1:e5 ~f6 36 :a5 1:xd6 37 h4 g5 38
10...lDb6 (to stop c3-c4) 11 a4 as 12 i..bS+! hxg5+ ~xg5 39 r:Ji;h2 ~g4 40 Itc5 f4 41
iLd7 and now either 13 ltJeS i..xbS 14 1:c8 Itd4 42 :c6 1:xa4 43 1:xh6 a5 44
'i'xbS+ 'it'f8 or 13 i..xd7+ 'ilxd7 14 ltJeS Itg6+ ~f5 45 :c6 :a3 46 Itc8 a4 47
'ild6 15 'ii'bS+ ct;e7 16 0-0 :be8 17 'ifd3 g6 :f8+ ~e4 48 1:e8+ ~d4 49 1:f8 ~e3 50
18 e4 as in Vidmar-Furlani, Ljubljana 1938, is Ite8+ ~d2 51 :e4 f3 52 gxf3 1:xf3 53
very awkward for Black. :xa4 Yz-Yz
26
Lasker Variation
Summary
In conclusion, the Lasker is solid but slightly passive for Black. It is the type of line that can be
recommended against an all-out attacking player who will not fmd the patient play demanded
of White to his liking. White's best hope of a lasting advantage lies in the 9 :c1lines seen in
Games 1-4.
1 d4 d5 2 c4 e6 3 tiJc3 tiJf64 tiJf3 iL.e7 5 iL.g5 0-0

S...h6 6 i..h4 tDbd7 7 e3 tDe4 - Game 9
6 e3 h6
6... tDe4 - Game 8
7 iL.h4 tiJe4 8 iL.xe7 'iixe7 (D) 9 !.1c1
9 cxdS tDxc3 10 bxc3 exdS 11 'ifb3 :d8 12 c4 dxc4 13 i.xc4 tDc6
14 i..e2 - Game 5
14 'iVc3 - Game 6
9 'iVc2 - Game 7
9 ... c6 10 iL.d3 tiJxc3 11 ~txc3 (D) dxc4
11...tLld7 - Game 4
12 J..xc4 tiJd7 130-0 (D) b6
13 ... e5 14 i..b3
14... exd4 - Game 2
14...:e8 - Game 3
14 J..d3 - Game 1
8 ... 'iVxe7 11 :'xc3 130-0
27
.
CHAPTER TWO
Orthodox Variation (6 ... liJbd7):

Old Main Line with 7 :c1 c6
1 d4 d5 2 c4 e6 3 lDc3 tDf6 4 tDf3 i.e7 by liquidating the centre. This includes

5 .i.g5 0-0 6 e3 tDbd7 systems with an immediate ...c7-cS, or with
fIrst ...d5xc4 and then ...c7-cS.
2. Black first develops his position by
exchanges or quiet manoeuvring before
striking back at the centre. This includes the
Oassical systems introduced by ...c7-c6, and
lines with a preliminary ... a7-a6, to follow up
with ... d5xc4 and...b7-bS. (I think of this as
the QGA option, since the idea of freeing b7
for the light-squared bishop is frequently
seen in that opening, e.g. 1 d4 d5 2 c4 dxc4 3
ltJf3 tLlf64 e3 e6 5 ii.xc4 cS 6 0-0 a6 7 'iie2
bS 8 i.b3 i.b7).
It is clear that ... d5xc4 is an integral part of
Q1estion 1. What is Black aiming for in this most of Black's plans. This leads to some
line.~ opening subtleties which Tartakower called
Arz..wrer 1. As always in the QGD, Black's the 'fIght for the tempo'.
general opening aims remain the activation of Qlestinn 2. What does that mean?
his light-squared bishop and the creation of Armrer 2. White will complete his
central counterplay with ...c7-cS or ...e6-eS. development by moving his light-squared
With the flexible 6...tLlbd7, Black provides bishop to d3 or e2 and then castling kingside.
support for both central breaks while However, White would prefer to meet
retaining the option of the Lasker manoeuvre ...d5xc4 with ii.flxc4 rather than to waste a
...ltJf6-e4 to free his position with exchanges. tempo fIrst with .tfl-d3 and then reach c4 in
The Orthodox QGD offers a large choice two moves after ...d5xc4, ii.d3xc4.
of development schemes for Black. These This is the reason why White most often
fall broadly into two categories: plays 7 .tIel or 7 'ifc2 rather than 7 .td3:
1. Black strikes quickly against the centre, White makes an extra useful move and waits
intending to solve his development problems for Black to commit himself with ...d5xc4. In
28
Orthodox Variation (6 . .. li:Jbd7): Old Main Line with 7:'c 1 c6
tum Black often also attempts to play useful Answer 4. In fact, this is a typical and
strengthening moves before playing ...dSxc4: excellent way of crossing Black's plans. Black
for example ... a7-a6 or .. .lHS-eS. This little strengthened his centre with ...c7-c6 in order
battle is a sub-plot to Black's main opening to exchange the dark-squared bishops with
auns. ...ttJf6-e4, or ...d5xc4 and ... ttJf6-d5 before
striking back at the centre. By retreating the
Game 10 bishop to f4, White avoids his opponent's
Karpov-Campora plan and transposes back to as .tf4 system,
Villarrobledo (rapidplay) 1997 against which ... c7-c6 systems are not very
effective.
1 d4 lLlf6 2 c4 e6 3 lLlf3 d5 4 lLlc3 i.e 7 Q«stion 5. Can't Black just chase the
5 i.g5 0-0 6 e3 lLlbd7 7 l:c 1 bishop with 9...ltJh5?
By activating his queen's rook, White Answer 5. White then plays the typical 10
prevents his opponent from using the Lasker .teS! when 10...ttJxe5 (otherwise White will
manoeuvre as 7...ltJe4 S i..xe7 'it'xe7 9 cxdS play hl-h3 and iLe5-h2, leaving the black
ltJxc3 10 :txc3 exdS 11 .s.xc7 loses the knight in limbo on h5) 11 dxe5! gives White
undefended c-pawn. a clear advantage due to the terminally
7 ... c6 offside knight on h5. For example, Thomas-
This is the Old Main Line of the QGD. Lasker, Nottingham 1936, continued 11...g6
By placing the c-pawn on a protected square
and consolidating his centre, Black renews
12 0-0 .td7 13 -.d2 dxc4 14 .txc4 'it'c7 15
ltJe4 l:tadS 16 -.c3 with a mighty position
the idea of ... ltJf6-e4. The drawback is that for White. Black must consequently fmd
Black's freeing break ...c6-c5 will take two another way to liberate his position.
moves instead of just one. The alternative S -.c2 is the subject of
S i.d3 Games 13-16.
Q«stion 3. I'm a bit confused. Isn't White S... dxc4 9 i..xc4 lLld5!
just losing your 'fight for a tempo'?
Amuu 3. Black's choice of the
consolidating 7...c6 in response to 7 :tel
means that if Black subsequently plays the
...c6-c5 break, he will have done so in two
moves instead of just one. Consequently,
White sees nothing wrong in playing the
bishop to cl3 now, since the tempo lost on
.tfl-d3xc4 will be regained if Black plays
...c6-c5. S .ic13 also restricts Black's options
by preventing any attempt to transpose to a
Lasker system: S...ltJe4 9 ltJxe4! dxe4 10
iLxe7 ii'xe7 11 iLxe4 wins a pawn, while
S...h6 is met by 9 i..f4! (9 i..h4 ltJe4! is more The standard, but ingenious solution!
than Black deserves; but 9 cxdS!? is a typical 10 i4.xe7 'iixe7 11 0-0
and interesting idea as 9...hxgS 10 dxe6 fxe6 11 ttJe4 is also popular - see Games 11
11 ltJxg5 gives dangerous compensation for and 12.
the piece). 11 ... lLlxc3 12 ltxc3
Q«stion 4. Isn't it strange to play first We analysed this position in the Lasker
iLc1-gS and then i..g5-f4? system, but with Black's h-pawn on h6 rather
?O
than h7. as g6 is covered by the h7-pawn) 18 tiJc4

Q«:stion 5. Who does this favour? 'iWf4 19 'ifu5! looks very promising for
Ansza'Y' 5. This factor is in White's favour White. The text is vety risky.
in the 12...b6 system since White will gain a 171:.f3?!
useful tempo on the h7 -pawn with a future I would have been vety tempted by 17
fld1-c2 and iLc4-d3. However, it is ltJxf7 :xf718 ~xf7+ ~719 'ili'b3+. White
undoubtedly in Black's favour in the 12... e5 is going to pick up the whole black
system. queenside: for txample 19...~f8 20 flxb7
Qlestion 6. Why is that? l:.b8 21 "xal Ihb2 is met by 22 l:.xc6! as
A nmx!Y 6. White's kingside tnltlatlve in 22 .. :iVxe6 23 fia3+! forks the king and rook.
Karpov-Yusupov flowed against Black's 17 ...i.g4?
kingside light squares. Consequently, the h- A fatal blunder.
pawn is a much better defensive unit on h7, 18 1:.xf6 li'xe5 19 dxe5 i.xd1 20 i.xf7+!
where it covers the g6-square, than on h6 1:xf7 21 :xf7 ~xf7 22 1:.xd1 ~e6 23 f4
where it is merely a target for attack. a5 24 ~f2 :a6 25 ltd6+ ~e7 26 ~e3
The question is whether this factor is llb6 27 b3 a4 28 :d3 ~e6 29 g4 c5 30
sufficiently important to enable Black to h4 1:.b4 31 h5 h6 32 bxa4 ':xa4 33 1:d6+
neutralise his opponent's initiative. ~e7 34 :b6 ':xa2 35 :xb7+ ~e6 36
12 ... e5 13 i.b3 exd4 14 exd4! ttJf6 15 :b6+ ~e 7 37 :c6 1:g2 38 ~3 :c2 39
1:e 1 "d6 16 ttJe5 'it>e4 c440 1:c7+ ~d8 41 1:.xg7 :d2 42
1:.a7 c3 43 1:a3 :e2+ 44 ~f5 :e3 45
1:a4 c2 46 :c4 :e2 47 g5 hxg5 48 txg5
~e7 49 h6 1:.h2 50 ltc7+ <&t>d8 51 :c5
<&t>e7 52 ~g6 1-0
This is a simple and fairly effective
method against the Old Main Line. White
has many other options however.
Game 11
Topalov-Vermolinsky
Yerevan Olympiad 1996
1 tiJf3 tiJf6 2 c4 e6 3 ttJc3 d5 4 d4 ttJbd7
16 ... i.f5? 5 i.g5 i.e7 6 e3 0-0 7 1:c1 c6 8 i.d3
In this position 16... .te6 can be met by 17 dxc4 9 i.xc4 ttJd5 10 i.xe7 'iixe7 11
ii.xe6 fxe6 (17...'ifxe6 18 ltJg6!) 18 fib3 as ttJe4!?
after 18...'ifxd4 19 'ii'xe6+ the black king
does not have h7 available in this line.
Obviously, Black should protect his b-pawn Qiestion 7. What does this do?
with either 18...:ab8 or 18...fle7. In this Anszrer 7. The first achievement of this
case, White's best plan is to double rooks on move is to deter Black from using his central
the e-flle and transfer the knight to eS via d3. breaks as 1L.eS loses a pawn to 12 i..xdS
White's position is the more pleasant, but exdS 13 lDxc5, while 1L.eS 12 dxeS ttJxeS
Black does only have one weakness and 13 ttJxeS fixe5 14 ~xdS cxdS 15 ttJc3 :d8
some potential activity along the f-fue. This is 16 fld4leads to a typical endgame advantage
Black's best as 16.. ,1!ijdS 17 l:.g3 f6 (possible for White (knight vs. bad bishop).
30
Orthodox Variation (6 .. .t£Jbd7): Old Main Line with 7 l:tc 1 c6
ltJ6d5 18 dxcS a6 19 ~e2 bxc5 20 a3 g5!! (I

love this moveQ 21 l:tfdl g4 22 ~el and now
22.....g5 would have given Black a very good
game according to Ljubojevic. Amazing
activity! The logical 13 ltJe5 is suggested by
Ljubojevic: once the knight on d7 is
removed, ...c6-c5 is difficult to achieve, but
there is scope for ideas here.
12liJg3!?
This was Alekhine's ambitious idea White
continues to avoid the exchange of knights
and prepares e3-e4-e5 followed by l'Dg3-e4-
d6. However, White loses his control over
White also understands that his opponent his opponent's pawn breaks. The alternative
wishes to exchange pieces to ease his 12 ~xf6+ is the subject of the next main
cramped position. Consequently, White game.
avoids the exchange of pieces as the best way 12... e5
of crossing Black's plans. White intends to The simplest. 12 ...'iWb4+ 13 "d2 "xd2+
use his pawn centre and by avoiding 14 ~xd2 gives White a small edge, while
exchanges, he makes sure that it will have 12...:d8 13 0-0 cS 14 e4 cxd4 15 e5 lDe8 16
targets to attack and brush aside as it Itel (16 'iVxd4 ~b6 17 "e4 J..d7! equalises
advances. according to Ehlvest, as 18 'iVxb7 ~xc4 19
Qtestion 8. Sounds good! Itxc4 l:tdbs followed by ...~d7-b5 wins the
A1'lJra'r 8. In theory, yes. However, the exchange) 16...~f8 17 ~xd4 ~g6 18 "d2
plan is time-consuming: e4 is after all a b6! was a little cramped but playable for
temporaty square for the knight. Since Black in Ivanchuk-Ehlvest, Yerevan
White's plan is to activate his central pawns, Olympiad 1996.
the knight will have to move again to allow 13 0-0 exd4 14 tt)f5
the e-pawn to advance which will cost 14 tDxd4 is met by 14...g6 15 l:te1 l:td8
another tempo. Consequently, Black gains with ...c6-c5 to follow according to
some time for his development. Yermolinsky.
11 ... tt)5f6 14...•d8 15 tt)5xd4
Black persistently looks for exchanges. 15 ~3xd4 (15 'i'xd4 ~b6 16 ~d3 'iVxd4
11...b6 is a more ambitious idea Black uses 17 tD3xd4 :dS IS l:.fdt ~f8 gave White
the time that White spends with his queen's very little in Alterman-Hertneck, Bad
knight for development rather than Wiessee 1997) 15...l'De5 16 ~b3 ~xf5 17
exchanges. My appetite for this variation was tDxf5 was the famous game Alekhine-Lasker,
whetted by the fantastic game Portisch- Zurich 1934, and now 17...g6! (instead of
Ljubojevic, Milan 1975, which continued 12 17.. :i'b6? 18 "d6Q 18 'iVd4 'iVxd4 19 l'Dxd4
0-0 iLb7 13 tLlg3 (mtending e3-e4) 13 ...cS 14 was agreed drawn m Flohr-Euwe,
e4 tLlf4! 15 .i.bs (preventing lS ... cxd4 due to Nottingham 1936.
16 :c7; 15 dS exdS 16 exdS :ad8 17 "d2 15...tt)b6 16 .td3 'ike7 17 'iic2 .tg4 18
'iif6 18 :fe1 lUe8 19 :e3 ~eS 20 l'Dxe5 a3 :ad8 19 :fe 1 ltJbd7
lhe5 21 :f3 lDh3+ 22 ~f1 "g5 was very Black has equalised according to
active for Black in Groszpeter-Almasi, Yermolinsky.
Budapest 1992) lS...:fd8 16 1Wa4ltJf6 17 eS 20 ltJg5 h6 21 tt)h7 :1e8 22 h3 .te6 23
L~ 1
llJxf6+ 'iixf6 24 i.f1 i.d5 25 f3 llJe5 26 Qiestion 11. Why does White play 13 ~b3
'ilff2 a6 27 b4 b5 28 lted1 llJc4 29 e4 before castling?
i..e6 30 i.xc4 i..xc4 31 ltd2 ltd7 32 An.sm:r 11. Maybe Korchnoi wanted to
ltcd1 lted8 33 f4 g5 34 fxg5 'iixg5 35 avoid 13 ...c5 after 13 0-0. This isn't really a
ltJf3 ltxd2 36 ltxd2 ltxd2 37 'iVxd2 'iVe7 problem though as 14 dxc5 ttJxc5 15 b4!
38 'iVd4 a5 39 "e5 'iVa7+ 40 llJd4 axb4 looked very pleasant for White in Schmidt-
41 axb4 i.e6 42 ~h2 'iid7 43 ~g3 ~h7 Prandstetter, Prague 1984.
44 ~f4 'iVa7 45 'iic5 'iVc7+ 46 ~f3 'iid7 13 ... e5 140-0 exd4
47 "'e5 ~g8 48 ~g3 ~h7 49 ~h4 'iVe7+ 14...':d8 15 ttJxe5! ttJxe5 16 dxe5 ':xdl17
50 ~g3 'iVd7 % -% exf61hc1 18 ::txc1 gxf6 19 ::td1 ~f5 20 e4!
i.xe4 21 ::td7 is slightly better for White
Game 12 according to Korchnoi.
Korchnoi-Hubner 15 'iVxd4! 'iixd4
Biel 1986 This has been the exclusive choice, but it
brings the white knight to a strong square in
1 !Df3 d5 2 c4 e6 3 d4 llJf6 4 !Dc3 i.e 7 the endgame for free. 15 .. :iVe7!? is very
5 i..g5 0-0 6 e3 lDbd7 7 l:c1 c6 8 ~d3 sensible, intending to develop with ... ttJd7-
dxc4 9 i..xc4 lDd5 10 i.xe 7 'iixe 7 11 b6/f6 and ...i.c8-e6.
lDe4lD5f6 12lDxf6+ 'iixf6 13 i..b3 Qtestion 12. Hasn't Black just lost time
with .. :iVe7-f6-e7?
An.sm:r 12. Yes, but in return he has
avoided the dangerous IQP lines with e3xd4
and thus greatly curtailed White's activity.
16lDxd4
Qtestion 9. It doesn't seem very logical to

avoid the exchange of knights with 11 ttJe4
and then to agree the move after!
Amrrer 9. As we saw, 12 l2Jg3 was too
time-conswning so the text is very sensible.
Qetion 10. It just looks like a worse Qetion 13. Hasn't Black just equalised
version of the 11 0-0 line. here?
Amrrer 10. White does have fewer An.sm:r 13. Unfortunately not!
attacking chances in this line: his rook is less Qtestion 14. But Black has done everything
active on c1 than c3, and after ...e6-e5xd4, right!
e3xd4 White cannot gain a tempo on the An.sm:r 14. As I mentioned when analysing
queen with ::tfl-e1. However, the position of the Lasker system, by exchanging pieces,
the queen on £6 gives White a tempo for an Black makes himself safe from an immediate
endgame possibility. assault but he does not guarantee himself
32
Orthodox Variation (6 .. .tiJbd7): Old Main Line with 7 lie 1 e6
equality. This game is a good example of a :d2 e5 22 llJf3 i..e8 23 :fd1 lIxd2+ 24
practical method of play that White can use 1:.xd2 lIe8
against the most brazenly exchange- 24.. .f6! was better according to Korchnoi.
orientated lines of the QGD. White doesn't The text allows an audacious pawn grab.
seek to get the maximum out of his position, 25 llJgS e4 26 i.e2 ~e7 27 llJxh7 f6 28
but he plays for a small dw-able edge and the lId4 i..f7 29 i..f5 :eS 30 g4 lIb5 31
prospect of torturing his opponent. White's :d7+ ~e8 32 :d2 1:.a5 33 a4 a6 34 ~g3
stable advantage is based on three factors: lIxf5 35 gxfSltJe7 36 e4 i..g8 37 llJxf6+
1. A lead in development. gxf6 38 lId6 bS 39 axb5 axb5 40 :Lb6
2. Better minor pieces. For example, ltJe8 41 :xb5 lDd6 42 :b8+ ~7 43 ~3
compare active White's light-squared bishop ~g7 44 :b6 1-0
on b3 and Black's bishop on c8. Black's light- Qiestion 16. Is 8...dxc4 the only way that
squared bishop has few tempting squares: Black can look for countetplay?
White's knight takes away fS and e6. An.su.e- 16. No, with 8. .. a6 Black can
3. The most important factor. the attempt to revert to the queenside plans
dynamism of the respective pawn majorities. normally introduced by 7...a6. White's most
It is dear that neither side will be able to aggressive try is 9 cS which transposes into
engineer a quick breakthrough with their the next chapter, but White has other moves:
pieces - it just isn't that sort of position. 9 a4 dxc4 10 i..xc4 bS!? (10 ...lDdS 11 i..xe7
Consequently, the initial aim for both sides is 'i'xe7 12 0-0 1lJxc3 13 :xc3 eS gives White
to gain space; in the future, the pawns will be the extra possibility of a4-aS fixing the black
used to drive the opposing pieces from their queenside structure) 11 axbS (11 i..d3 bxa4!
ideal defensive posts and thus create space [intending ... a4-a3] is irritating as 12 ttJxa4
for your own pieces to exploit. Due to his 'i'a5+ is disruptive - 13 t'lJd2 is not possible
lead in development and more active pieces, here as the bishop on gS is loose) 11...cxbS
it is much easier for White to expand on the 12 i..d3 i..b7 13 0-0 was slightly better for
kingside than it is for Black to expand on the White in Csonkiks-Velvan, Hungarian Team
queenside. This inevitably gives White the Championship 1994, as Black cannot
early initiative in the endgame. organise any pawn pressure against the white
Qiestion 15. Oh no! Sounds bad! centre, while 9 b3 is my personal favourite.
A1'l9r.rer 15. It isn't all doom and gloom! Qtestian 17. What does this do?
Black doesn't have any real weaknesses so his Amzrer 17. Now ...dSxc4 is simply met by
disadvantage is manageable. However, if you b3xc4! After 9...bS 10 0-0 (10 cst? is also
are not prepared to suffer a bit in order to interesting as after 10...eS 11 dxeS, White
secure the draw, then playing this position gains the idea of t'lJf3-d4 hitting the
can be thoroughly demoralising! undefended c6-pawn) 10...bxc4 11 bxc4 dxc4
16 ... lId8 17 lIed1!? 12 i..xc4 cS 13 'i'e2 with l:tf1-d1 to follow,
TIlls is aimed against the development of White has a slight edge.
Black's knight. 17 l:tfd1 lDf8 18 f3 was 8 i..d3 is an uncomplicated route to a
normal and is assessed by Korchnoi as slight advantage. However, White can also
slightly better for White. try more ambitiously to win the 'battle of the
17 ... ltJeS tempo' by further delaying i..f1-d3. He can
17...tDc5 18 tDxc6 l:txd1 19 l:txd1 i..g4 20 do this with either 8 'tic2 or 8 a3.
l:td4! bxc6 21 lhg4 lDxb3 22 axb3 is clearly Q«:stion 18. Okay, 8 'W'c2 I understand,
better for White according to Korchnoi. but why is 8 a3 useful?
18 f4 ltJg6 19 h3 i..d7 20 ~f2 ~f8 21 Amzw- 18. In general, 8 a3 adds to the
33
'comfort' of the White position: it provides a 13 ...'ifa3 (13 .. :iWc3!?) is thus normal
retreat on a2 for the light-squared bishop in preventing 14 i..d3 due to 14...lDf6! winning
the event of ...d5xc4 and ...b7-b5, and denies a piece. After 14 i..e2 (14 1:b3 "c1+ 15 ~e2
the use of b4 to the black pieces while e5 16 g3 was Agdestein-Prandstetter, Taxco
supporting the space-gaining b2-b4. Interzonal 1985, and now Prandstetter
8...dxc4 is not tempting for Black as it mentions 16.. .fS 17 "d3 e4 18 "b1 'ii'xb1
leads to the main lines with a useful extra with ...c6-c5 and ... b7-b6 to follow instead of
move for White. Consequently, his most the wild 16...lDcS!? 17 dxcS 1:d8 18 1:d3
active possibility is 8...lDe4 (8 ... a6 is i..g4+! 19 f3 fS! which led to incredible
considered in Games 14-16). complications) 14.....e7 15 f4 c5 160-0 1:b8
17 fS lDf6 18 ..f4 i..d7 19 e4 exfS 20 e5
Game 13 cxd4 21 exf6 'ii'xe2, the game was a mess in
Pinter-Prandstetter Ftacnik-Franzen, Czechoslovakia 1984, as 22
Taxeo Interzonal 1985 'ii'g5 is countered by 22...'ii'g4!
Timman-Prandstetter, Taxco Interzonal
1 liJf3 d5 2 d4 liJf6 3 c4 e6 4 liJc3 i..e7 1985, saw the quieter 10 i..d3, when
5 .tg5 0-0 6 e3 liJbd7 7 ':c1 c6 8 'l'c2 10... tDxc3 11 bxc3 h6 12 cxdS exdS 13 0-0
liJe4!? lDf6 14 c4led to a type of position we saw in
the section on Lasker hybrids in Chapter 1
(P Nikolic-Andersson, Leningrad 1987)
which favours White slightly. Prandstetter
played more accurately against Smejkal at
Trencianske Teplice 1985 with 11...dxc4! 12
i..xc4 (12 i..xh7+ ~h8 13 i..e4 f5!) 12...b6
13 0-0 i..b714 e4 c5.
This is a very comfortable version of both
Semi-Tarrasch and Queen's Indian-type
positions. Although Black lost a tempo with
...c7-c6-cS, White lost two himself with i..f1-
d3xc4 and e2-e3-e4. Moreover, the white
queen is badly placed on c2 and will be
Now that White has played his queen to forced to move once a black rook comes to
e2, Black attempts to transpose into the "i'c2 c8.
variation of the Lasker system, which is Qi.estion 20. Can't White play 11 "ii'xc3?
usually less dangerous for Black. Arz.srrrer 20. Then 11...dxc4 12 i..xc4 b6 13
Qi.estion 19. Wait a minute! Isn't Black just 0-0 i..b7 compares favourably with the
losing a pawn? Lasker line 6...h6 7 i..h4 tDe4 8 i..xe7 "xe7
Answer 19. Amazingly not! After 9 i..xe7 9 "c2 lDxc3 10 'ii'xc3 dxc4 11 i..xc4 b6 12
'fIxe7 10 tDxe4 dxe4 11 1i'xe4, Black has 0-0 i..b7.
11...1Ib4+ 12 tDd2 'ii'xb2 regaining his pawn. Qi.estion 21. Why is that?
The endings after 13 'ii'e2 'ii'xc2 14 1:xc2 eS Amwer 21. White cannot prevent Black
are fme for Black, but White can try 13 1:b 1 from achieving ...c7-cS. Consequently, in
as the pawn grab 13 ...'fIxa2 14 i..d3 lDf6 order to fight for an advantage, he played 13
(14...g6 15 h4! was very strong in Ftacnik- i..e2 c5 14 dxcS 1:c8 IS b4 bxcS 16 b5
Ree, Lucerne Olympiad 1982) 15 ..h4! followed by a2-a4 hoping to exploit the
(intending g4-gS) is too dangerous for Black. weakness of the c5-pawn and to create a
34
Orthodox Variation (6 .. . li:Jbd7): Old Main Line with 7 lie 1 e6
passed pawn on the queenside. In this case, 8 .i.d3, the answer is again the same: 9 i.f4!
the white rook is of more value on al than The typical 9... a6 to expand on the queenside
el; consequently, Black's tempo ...llJb8-d7 with ...d5xc4 and ... b7-b5 is met by 10 c5! as
(supporting ... c6-cS) is much more useful with the bishop on f4, Black does not have
than White's l:tal-c1 (which does not manage the freeing break 10...eS.
to prevent ...c6-c5). Qtestian 23. Can't Black do anything else
9.i.f4! than 8...llJe4 or 8...h6?
This is the most annoying move for Black Arzsuu 23. Now it gets complicated! A
to face. White refuses to fall in with his crucial point is that this was the last time that
opponent's drive for exchanges and forces Black was guaranteed to get in the freeing
Black to seek another development scheme. move ...tDf6-e4. Once White plays both
9 ...f5 'i'dl-c2 and a2-aJ, ...lDf6-e4 is no longer
This is Black's best option, consolidating possible as after i.g5xe7 ..:ifd8xe7, llJc3xe4
the central space he gained with ...lDf6-e4. If ... d5xe4, ifc2xe4, the pawn on aJ prevents
he cannot free his position with multiple ... 'i'e7-b4+. After 8 aJ therefore, the
exchanges, then he must give his pieces more positional threat is 9 ifc2, and vice versa
room to breathe. However, White has a vety If Black is not going to play ...llJf6-e4,
good set-up against this Stonewall Dutch then he must wait, and aim instead to win
fonnation (the bishop is excellent on f4) and 'the fight for the tempo'
maintains a slight advantage, although he Qlestian 24. You mean, wait for White to
eventually went astray in this game and lost. move his light-squared bishop and then take
10 h3 ~df6 11 .i.d3 .i.d7 12 0-0 .i.e8 13 onc4.
~e5 ~d7 14 f3 ~xe5 15 Axe5 ~xc3 16 Anm.er 24. Right! Of course, Black must
bxc3 .i.d6 17 .i.xd6 1»'xd6 18 'it'b3 'it'e7 have a follow-up to ...d5xc4 ready, and for
19 %lfe1 ~h8 20 .i.f1 g5 21 :b1 b6 22 this purpose he has the waiting move ... a7-a6.
'it'b4 'it'f6 23 .i.d3 ':d8 24 :f1 :g8 25 Qiestion 25. What does it do?
cxd5 exd5 26 :be1 .i.g6 27 'it'b1 c5 28 Arzsuu 25. The move ...a7-a6 takes control
~h1 lIde8 29 'it'b5 'it'e7 30 'it'c6 c4 31 of b5j consequently, Black is primed for
.i.b1 'it'e6 32 'it'c7 f4 33 .i.xg6 1»'xg6 34 rapid queenside expansion with ...d5xc4 and
-.xa7 fxe3 35 lIe2 h5 36 lIfe1 g4 37 then ...b7-b5 and ...c6-cS. This achieves all of
fxg4 hxg4 38 lIxe3 :xe3 39 ':xe3 -'g5 Black's aims: he attacks White's centre and
40 :g3 1»'14 41 'it'xb6 %lb8 42 'it'xb8+ frees b7 for his light-squared bishop, while
1»'xb8 43 :xg4 ~h7 44 a4 'i'b3 45 :g3 opening the aB-h 1 diagonal.
'it'xa4 46 :f3 'it'd1+ 47 ~h2 'it'e1 48 ':g3 Black's other waiting move is the
~h6 49 :f3 ~g5 50 ':g3+ ~h4 51 :g4+ consolidating ...l:tf8-e8. This has little active
~h5 52 ':g3 -.c1 53 ':f3 ~h4 54 g3+ value but it is useful: the rook will support a
~h5 55 h4 'it'd1 56 %lf5+ ~g6 57 :f2 future ...e6-e5, it protects whichever black
'it'd30-1 piece comes to e7 and it frees £8 for the black
This is the best way to meet attempts to knight on d7 in case the black kingside needs
transpose into Lasker systems. It also applies some extra support.
in the 8 a3 variation: after 8... tDe4, White's Qiestion 26. So who will win the 'fight for
best reply is 9 i.f4! the tempo'?
A nslU!r 26. To let you into a secret, only
Qiestion 22. What if Black plays 8...h6 fIrst White can! Since he has more space, he
to strengthen ...lDf6-e4? inevitably has more useful waiting moves.
Amzrer 22. For both 8 'ifc2 and 8 a3 as for Qiestion 27. So why is Black bothering?
Arz.m.u 27. The game has paused for a breaks, it creates the opportunity for the
moment as both sides' stop and listen' before other. The alternatives 9 cxdS and 9 aJ are
proceeding further. Black's contention is that seen in Games 15 and 16 respectively.
moves such as ... 37-a6 benefit him more than 9 ... e5!
a2-aJ or h2-h3 does White, and that these This would also be the answer to 8 aJ a6 9
differences are enough to even up the game. cs, when 9... es 10 dxes ~e4! 11 i..xe7 fixe7
Black may ultimately lose the 'fight for the 12 lLlxe4 dxe4 13 ~d2 ~xcs 14 'i'c2 1Lld3+!
tempo', but in the course of this skinnish, is fme for Black.
certain details will have arisen which will help 10 dxe5 ttJe8
him in the overall battle. 10... ~g4 11 .if4! is Polugayevsky's
suggestion. Then 11...~xcS 12 h3 ~h6 13
Game 14 i..xh6 gxh6 looks disgusting for Black at first
Rivas Pastor-Toth sight, though I don't think it's as bad as it
Rome 1984 looks: Black intends .. .f7-f6 to remove the
strong es-pawn and he has the two bishops
1 e4 e6 2 ttJe3 d5 3 d4 ttJf6 4 ttJf3 .i.e7 and a strong centre.
5 i..g5 0-0 6 e3 ttJbd7 7 :e1 e6 8 'i'e2 11 i..xe7
a6 Perhaps 11 h4!? h6 (11...1L1xcs 12 i..xe7
If Black wishes to wait, it is safest to play 'i'xe7 13 liJxdS! wins) 12 b4!? hxgS 13 hxgS
this active move first, so that he is ready to g6 14 e6!? or 14 i..d3!? is worth considering.
start his countetplay'sequence' ...dSxc4, ...b7- 11 .. :flxe7 12 .i.d3 h6 13 0-0 ttJxe5 14
bS and ...c6-cS at a moment's notice. For lDxe5 'i'xe5 15 e4 ttJf6 16 f4
example, 8...1:te8 is nicely met by 9 i..d3 as Here White should instead play 16 exds,
after 9...dxc4 (9 ...a6!? 10 lLles!?; 10 a4!?) 10 as in the note to White's 11th move in Game
i..xc4 lLldS 11 i..xe7 fixe7 12 lLle4 lLlSf6 17.
(12 ...b6!?) 13 lLlg3 eS 14 0-0 exd4 15 lLlfS 16 ...'i'd4t 17 ~h1 dxe4 18 lDxe4 ttJxe4
'ii'd8 161L13xd41L1es 17 i..b3 i..xfs 18 ~xfS, 19 .txe4 :e8 20 .tf3 'i'f6 21 1i'b3 ':e7
it can be seen that the inclusion of the moves 22 'i'b6 .te6 23 b3 g6 24 :cd1 ~g7 25
'iWd 1-c2 and ...:f8-e8 clearly favours White. 1i'b4 :tae8 26 ~g1 .i.f5 27 1i'd4 :e3 28
9 c5!? 'i'xf6+ ~xf6 29 :d6+ :8e6 30 :xe6+
i..xe6 31 :tc1 :d3 32 :e2 ~f5 33 .i.e2
:d4 34 93 c:re4 35 ~f2 .i.d5 36 :e3
l:d2 37 a3 :b2 38 b4 ~f5 39 :te3 .tc4
40 ':e5+ ~f6 41 ~f3 :b3+ 42 :te3
.txe2+ 43 ~xe2 l:b2+ 44 ~f3 ':xh2 45
:e8 :ta2 46 :b8 :xa3+ 47 ~f2 a5 48
:xb7 axb4 49 :txb4 :le3 0-1
QIestian 28. You are putting 8 aJ and 8

'i'c2 together as if they were the same thing.
Is that really true?
A1'19UU 28. If White is intent on a waiting
plan, then there is no difference between
This ambitious move seeks to exert a 'big them. However 8 'ii'c2 is more flexible than 8
clamp' on Black's position. However, aJ as it allows him a nwnber of independent
whereas c4-cs prevents one of Black's central possibilities.
36
Orthodox Variation (6 .. .l'i~bd7): Old Main Line with 7 ::'c 1 c6
similar to the Rubinstein-Takacs game in the

Game 15 ltJf3 Exchange variation chapter (Game 76).
Alekhine-Capablanca 10 i.xe7 fixe7 11 i..e2 :e8 12 0-0
Buenos A ires (2nd matchgame) 1927 lDxc3 13 fixc3 e5 14 :fd1 exd4
14... e4! 15 lDd2 ltJf6 (Rubinstein) is less
1 d4lbf6 2 c4 e6 3 4:Jc3 d5 4 iLg5 i.e7 accommodating, though White is slightly
5 e3 0-0 6 lDf3 4:Jbd7 7 ::'c1 c6 8 ....c2 better in this typical position.
a6 9 cxd5 15 lbxd4 lDf6 16 i.f3 i.g4 17 i.xg4
White aims to transpose into a favourable 4:Jxg4 18 lDf5 "'f6 19 "'xf6 lbxf6 ~ - ~
version of the Exchange variation. Of Alekhine claims a slight edge for White
course, for this purpose, 8 'ii'c2 is more with 20 lDd6 :e7 21 e4.
useful than 8 a3. Finally, we examine White's most
Another interesting idea of Alekhine's is 9 consistent idea: to continue the fight for the
a4!? tempo with 9 a3.
Qtestion 29. This looks weird! Qtestion30. Is this White's best?
Anszeer 30. I don't think so. My feeling is
that Black's waiting moves are far more
relevant than White's: they connect with
Black's positional ideas whereas White's are
just froth, topping up his position without
adding anything concrete.
Game 16
Pirc-Tylor
Hastings 1932/33
1 d4 lDf6 2 c4 e6 3 lbf3 d5 4 iLg5 lbbd7
5 e3 i.e7 6 lDc3 0-0 7 :c1 c6 8 'i'c2 a6
An3UU 29. With S...a6, Black pcimes 9 a3 h6
himself for rapid queenside expansion with Qtestion 31. Hey, I thought that this just
...dSxc4, ...b7-bS and ... c6-cS. White's idea is allowed 10 ..tf4!
that after ...dSxc4, ..tflxc4 ...b7-bS, ..tc4-d3 Ansr.rer 31. I agree! In my opinion, 9...h6 is
White's pressure on bS prevents Black from an inaccurate move order. After 10 ..tf4,
playing the freeing ... c6-cS. White threatens 11 c5 so Black must
Alekhine-Rubinstein, Carlsbad 1923, saw a immediately play 10...dxc4 11 ..txc4 bS 12
typical and interesting positional idea: 9...:eS ..ta2 c5. In comparison with the game, for
10 ..td3 dxc4 11 ..txc4 lDd5 12 ..tf4!? lDxf4 example, White denies his opponent the
13 exf4 {White's doubled pawns clamp down opportunity to play the useful ...lUS-e8.
on Black's position} 13 ...cS 14 dxcS 'ifc7 15 Instead 9...:eS 10 h3!? (10 ..td3 h6 11
0-0 'ii'xf4 16 lDe4 ltJxcs 17 lDxcs ..txc5 IS ..txf6!? is interesting) 10...h6 (a useful move
..td3! b6 19 ..txh7+ ~hS 20 ..te4:a7 21 b4! later, when White develops pressure on the
with a clear advantage for White. Instead of b I-h7 diagonal) to meet 11 ..tf4 with
9...:e8, 9...ltJe4! is an untested suggestion of 11...dxc4 12 ..txc4 bS 13 ..ta2 ..tb7 (13 ...c5!?)
Alekhine's. followed by ...c6-c5 is the most accurate
9 ...lDxd5!? move order. Black should always be ready to
9...exdS transposes into a type of position meet ..tgS-f4 with ...d5xc4, otherwise White
,., ..,
can employ the 'clamp' with c4-cS. After 14 ~f8 20 'ili'g6 mates) 19 lhe6 (19 i.xe6 is
0-0 cS 15 dxcS tDxcs 16 lUdl 'i!i'b6 17 i.eS met not by 19...~f8 20 tDh8!, but by
l:[ac8 18 "e2 ltJce4 19 i.d4 i.cs 20 ltJxe4 19...i.c5! which seems good for Black) looks
i.xd4! 21 tDxd4 i.xe4, the position was extremely awkward for Black: so many
equal in Alekhine-Capablanca, World discovered checks are coming! Again, this is
Championship 1927. even better with an extrah2-h3.
10 i.h4!? 14...ltJxc5!
10 cxd5!? was played in Yusupov-Van cler The knight annoyingly eyes the d3-square.
Sterren, Amsterdam 1982, aiming for 14... i.xcS 150-0 i.b7 16 l:tfd1 "b6 led to a
10... hxgS 11 dxe6 fxe6 12 tDxgS with brilliant attack in Pirc-Steiner, Prague
interesting play for the piece. With the text, Olympiad 1931: 17 .ib1 i.d6 18 l:txd6!
White utilises the fact that ... tDf6-e4 is not "xd6 19 l:td1 "c7 20 :Xd7! 'i'xd7 21 tDeS
possible to keep the bishop on the h4-d8 'i'd8 22 i..xf6 "xf6 23 "h7+ ~f8 24 ttJd7+!
diagonal. winning the queen.
10 ... .:e8 11 i.d3!? 150-0
11 h3 is possible when 11...dxc4 leads to 15 i.bl prevents .....d8-d3, but after
the same position, but with a useful extra h2- 1S....i.b7, Black has nothing to fear. 16 .i.xf6
h3 for White. 11 i.g3 is the alternative, when .i.xf6 17 'i'h7+ ~f8 leads nowhere.
11...dxc4 12 .i.xc4 bS 13 i.a2 cS 14 dxcS
tDxcs 15 l:tdl 'iWb6 16 b4 ttJcd7 17 l:tdl is
assessed by Polugayevsky as slightly better
for White. Simply 17...aS looks nice for Black
however.
11 ... dxc4 12 i.xc4 b5 13 i.a2 c5 14
dxc5
I would nonnally prefer to take my
chances in the IQP; although 140-0 cxd4 15
exd4 ~b7 is supposed to be comfortable for
Black, there is always life in White's position:
16 tDeS 'ii'b6!? and now White has two
choices: 17 i.b1!? (17 l:tfd1 tDxeS 18 dxeS
'i'c6! [Alekhine] 19 f3 and now 19...'i'cS+ 15 ... i.b7?
wins a pawn) 17...'i'xd4 (risky! 17...tDf8 is the An important mistake. 1S .....d3!, as in
safer option) 18 tDxd7 'ifxh4 19 ttJxf6+ Green-Reinfeld, USA Championship, New
i.xf6 20 'iWh7+ ~f8 21 tDdS! (threatening York 1940, equalises: 16 l:tfd1 'ii'xc2 17 l:txc2
'i'h7-h8+ mate) 21.. ..i.xdS 22 l:tc7 looks .i.b718 tDeS l:tad8 19 l:[xd8 Ld8 20 f3 ~f8
decisive for White, but Black has the amazing 21 e4 ~d3 22 tDxd3 lhd3.
resource 22... 'i'xh2+!! (Fritz, of course!) 23 16 ':fd 1 .b6 17 .i.xf6 i.xf6 18 b4!
~xh2 .i.eS+ followed by 24....i.xc7 with Suddenly, the knight on cS is embarrassed.
defensive chances. Note that if White had 18 ... i.xf3 19 gxf3 lDb7 20 i.b1 lDd6 21
played 11 h3 instead of 11 .id3, Black would 'i'h7+ cJrf8 22 lDe4 ':ad8 23 lDxf6 gxf6
be lost here as ..:i'h4xh2+!! would be 24 'ifxh6+ ~e 7 25 'i'h4 lDc4 26 :xd8
impossible! The other idea is 17 %He 1 as the :xd8 27 'i'f4 e5 28 'i'e4 ltJd2 29 'i'f5
pawn grab 17...'i'xd4 18 ltJxf7 (18 i.xf6!? ltJc4 30 ':c3 :d1+ 31 ~g2 'i'd8 32 i.d3
when both 18...i.xf6 and 18...ltJxf6 are met ltJd6 33 'i'g4 ':d2 34 h4 'i'b6 35 i.f1 f5
by 19 tDxf7!) 18.....xh4 (18...lt>xf7 19 .i.xe6+ 36 'ifg7 ~e6 37 :c5 1-0
38
Orthodox Variation (6 ... tiJbd7): Old Main Line with 7 :c 1 c6
Summary
Against the Orthodox variation, 7 :ct is White's best choice in my opinion. In the old main
line with 7...c6, 8 .td3 followed by 11 0-0 as in Game 10 seems to promise White a small,
pleasant advantage, though as always Black's position remains sOWld
1 d4 d5 2 c4 e6 3lbc3lbf6 4 tiJf3 i..e7 5 ..tg5 0-0 6 e3lbbd7 7 ':c1 c6
8 i.d3
8 'ilc2 (D)
8...ltJe4 - Game 13
8... a6
9 cS - Game 14
9 cxdS - Game 15
9 a3 - Game 16
8 ...dxc49 i.xc4lbd5 10 i..xe7 'iixe7 (D) 11 0-0
11 ltJe4 ltJSf6 (D)
12 ltJg3 - Game 11
12 ltJxf6+ - Game 12
11 ...lbxc3 12 ':xc3 - Game 10
8 'ilc2 10.....xe7 11 ... ltJ5f6
39
CHAPTER THREE
Orthodox Variation (6 ... ttJbd7):

Other Systems after 7 ~c 1
1 d4 d5 2 c4 e6 3 It)c3 i..e7 4 lDf3 lDf6 Qiestion 2. And how does Black launch his
5 i..g5 0-0 6 e3 It)bd7 7 :c1 countetplay?
In this chapter we shall consider some Arzsrrer 2. Black will play ...d5xc4 and then
other possibilities for Black after 7 l:[c1, follow up with ...b7-b5 and ...c7-c5.
starting with a 'QGA-type' approach. The Qtestion 3. How does this solve all Black's
'QGA-type' systems are characterised by the problems?
move ... a7-a6, played either immediately or An..mer 3. First of all, the combination of
after a prior ...d5xc4 (see Games 21 and 22). ... a7-a6 and ...b7-b5 frees b7 for the light-
We shall begin with 7... a6. squared bishop; secondly ...d5xc4 clears the
a8-hl diagonal of pawns, giving the bishop
on b7 a clear run of the diagonal; and thirdly,
Black strikes at the white centre with ...c7-c5.
In this way, instead of solving his space
problems by exchanges, Black solves them
by gaining queenside space.
Qtestion 4. Why do you call them 'QGA-
type' systems?
An..mer 4. This approach to Black's
problems lies at the heart of the Queen's
Gambit Accepted. You only have to see the
line 1 d4 d5 2 c4 dxc4 3 tiJf3 tiJf6 4 e3 e6 5
.i.xc4 c5 6 0-0 a6 7 'iWe2 b5 8 .i.b3 .i.b7 to
Qiestion 1. This looks subtle! understand! Incidentally, this line also has
A~ 1. Black's general aims as always parallels with the Nirnzo-Indian. The Russian
are to activate his light-squared bishop and to Grandmaster Kharitonov is an expert in the
create countetplay against the white centre. 7... a6 lines, and his favourite system against
The advance ... a7-a6 is a preparatory move to the Rubinstein Nirnzo-Indian is 1 d4 tiJf6 2
ensure that once Black launches his c4 e6 3 lLJc3 .i.b4 4 e3 0-0 5 .i.d3 d5 6 tiJf3
queenside play, it flows quickly and proceeds c5 7 0-0 dxc4 8 .i.xc4 cxd4 9 exd4 a6!?
without delay. intending ...b7-b5 and ....i.c8-b7!
40
Orthodox Variation (6 .. . 0.bd7): Other Systems after 7 :'c 1
Qiestion 5. Any drawbacks? active play.

Anszw- 5. 7... a6 pursues Black's interests Qiestion 8. How does it stop Black's plans?
without attending to White's. Thus with Amrrer 8. Black was relying on the
7... a6, Black does nothing to counter the c- sequence ...dSxc4 followed by ...b7-b5 and
fIle pressure that White developed with 7 ...c7-cS to fInd a post for his light-squared
l:.c 1. Consequently, White may force bishop and to gain space on the queenside.
transposition to an Exchange variation with After 8 cS, while White increases his
8 cxdS exdS (8 ...tDxdS 9 ltJxdS exdS command of queenside space, Black can
[9...i.xgS 10 ltJxc7 l:.a7 11 dS! is good for neither play ...dSxc4 nor strike at White's
White] 10 i.xe7 ii'xe7 11 :hc7 loses the c- centre with ...c7-cS.
pawn) 9 i.d3 c6. Qlestiun 9. Can't Black just strike back
Qiestion 6. Why would White want to do with 8...e5?
that? Amrrer 9. Here we see another drawback
Ansz.e.u 6. White's contention is that the of 7...a6 compared to 7...c6. 7...a6 does not
inclusion of l:.a1-c1 and ...a7-a6 over a add protection to the dS-pawn, and thus
normal QGD Exchange is significantly in his does nothing to consolidate Black's centre.
favour. These types of positions are Consequently, 8 cS eS 9 dxeS costs Black his
discussed in the Rubinstein-Takacs game in dS-pawn. Before he can play ...e6-e5, Black
the Exchange variation chapter. must reinforce his centre.
The second drawback is seen in the The alternative 8 b3 is seen in Game 20.
following game. 8 ... c6
Qlestiun 10. Wait a minute! Haven't I seen
Gamell this position before?
Epishin-Ziatdinov Amrrer 10. Nearly! 9 ii'c2 or 9 a3 would
World Open, Philadelphia 199 1 transpose into 7...c6 8 ii'c2/8 a3 a6 9 c5. In
this move order, however, White can play a
1 d4 d5 2 c4 e6 3 lLlf3 lLlf6 4 lLlc3 i.e 7 more useful move than either ii'd1-c2 or a2-
5 i.g5 0-0 6 e3 lLlbd7 7 :c1 a6 8 c5!? a3.
Instead 8...ltJe4 9 i..xe7 'fIxe7 10 ltJxe4!
dxe4 11 1tJd2 ltJf6 (11...fS 12 c6! breaks up
Black's queenside) 12ltJc4! (preventing ...e6-
e5) was very pleasant for White in Karpov-
Jakobsen, Malta Olympiad 1980.
9 i.d3!
1bis position can also be reached via 7...c6
8 i.d3 a6!? 9 cS.
Qlestiun 11. I don't understand If ...e6-eS
is coming, isn't 9 b4 better to hold the cS-
pawn after d4xe5?
A nm:er 11. This was also my fIrst reaction:
it is natural to wish to maintain the structure
Qiest:ion 7. What is the point of 8 c5? that seems to suffocate Black's position.
A1'7..9UU 7. The c4-c5 advance is an However, White's slow development offers
extremely ambitious positional idea that cuts Black an unusual way to create counterplay
out all of Black's queenside plans, forcing and solve his opening problems: 9... aS 10 a3
Black to search elsewhere in order to obtain axb4! (White now regrets l:Ia1-c 1 which
allows Black to take over the a-fUe) 11 axb4 Qlestion 13. This looks like a very
b6! 12 .i.f4 (to prevent ...e6-eS; 12 .i.d3 bxcS aggressive plan! What is the basis for it?
13 bxcS eS! 14 dxeS lDe8 leads to the main AnJZreY 13. First of all, activity. While
game, except that the exchange of all the Black still has to fmd active posts for the
queenside pawns increases the activity of rook on as, the bishop on c8 and the knight
Black's pieces enonnously) 12... bxcS 13 bxcS on e8, all of White's pieces combine against
.d2 lDe4! winning) 14 .d2 .as "as

l:ta3! (threatening 14...l1xc3 15 l:txc3 16
15 i.e2
Black's position. For example, White's
current threat is 13 tiJxdS! cxd5 14 l:txcS!
.i.a6! (polugayevsky) when Black exchanges wmnmg a pawn.
his light-squared bishop while at the same Secondly, Black's kingside weakness. 10
time developing queenside counterplay! dxeS perfonned the key function of
QIestion 12. 9 i.d3 is better? depriving Black's king's knight of its excellent
Anszw- 12. First of all, White activates his defensive post on f6.
last minor piece and prepares to castle his Qation 14. So what?
king to safety; secondly, White confiscates An.mer 14. This inevitably weakens Black's
more central territOIY by stopping ...lDf6-e4. defence of h7, a square for which White is
9 ... e5 perfectly placed to attack: White can set up a
The break-out! This is Black's most battexy on the b1-h7 diagonal with .d1-c2,
aggressive attempt. The other move, 9...b6, is while the pawn on h4, knight on f3, rook on
seen in Game 19. h 1 structure gives rise to many 'Greek gift'
10 dxe5 ltJe8 11 h4!? combinations starting with i.gSxe7,
The quieter 11 i.xe7 seems to promise an .ib1xh7+ and tiJf3-gS+.
edge for White: 11...•xe7 12 .c2 h6 13 e4! 12 ...f6
tiJxeS 14 lDxeS .xeS 15 0-0, as in
Kotronias-Goldin, Sochi 1989, was vel}'"
pleasant for White after 1S... tiJf6 16 exdS
tiJg4! (16...tiJxdS 17 l:tfe1 'iihs 18 tiJxdS
.xdS 19 i.c4 is clearly better for White
according to Ftacnik) 17 g3 'iihs 18 h4 cxdS
19 'ii'e2 i.e6 20 .f3 l:tad8 21 tiJe2, so
Ftacnik suggests 18... gS!? 19 tiJe4 fS to stir
up some counterplay.
11 ...ltJxc5 12 i.b1
Black's most logical continuation, ridding

himself of one of White's most dangerous
attacking units: the eS-pawn. However, it
loosens the pawn cover around the black
king, particularly along the b 1-h7 diagonal.
The alternative 12...tiJe6 is the subject of the
next mam game.
13 'ifc2! g6
Forced.
14 i.h6! .tf5
Orthodox Variation (6 . .. li:Jbd7): Other Systems after 7 'I:c 1
14... ttJg7 15 h5 .i.f5 16 'iVe2 .i.xbl 17 I don't like this move. Although it
:xb 1 transposes to the game. anticipates White's threat of 13 ttJxd5, and
15 'iie2 iLxb1 16 'I:xb1 li:Jg7 17 h5 f5?! keeps Black's kingside solid, 12... ttJe6 cramps
Ageichenko-Gavrilov, Moscow 1989, Black's set-up and pulls back the one black
continued instead 17... We8 18 l:h3!? (18 piece that succeeded in interfering with
hxg6 Wxg6!?, intending ...ttJc5-d3+, would White's ideas.
have offered Black some counterplay. The 13 "c2!
text intends 19 hxg6 Wxg6 20 i.xg7!; 18 l:d1 The ineffective 13 ttJd4 g6 14 i.h6 108g7
intending :dl-d4 is another interesting idea.) 15 h5 i.g5 16 hxg6 hxg6 17 i.xg5 Wxg5 18
18... fxeS 19 hxg6 hxg6 20 ttJxeS i.f6 21 ttJf3 lOf3 was agreed drawn in Korchnoi-
(21 ttJg4!?) 21...ttJe4 22 ttJxe4 Wxe4 23 l:dl Agdestein, Tilburg 1989.
with unclear play. The game continuation is 13 ... lDxg5?!
much worse as it leaves White with his This is really asking for it! 13 ...g6 14 i.h6
strong e5-pawn. 106g7 (14...ttJ8g7 15 h5 Wa5+ 16 Wd2 :d8
18 hxg6 hxg6 19lDd4 'ild7 20 f4:f7 21 is recommended by Kharitonov, but
g4! fxg4 22 i..xg7 :xg7 23 b4 lDe6 24 something like 16 ~f1 looks very appealing
'i'xg4 lDf8 25 'ilxd7 lDxd7 26 ~f2 lDf8 for White) 15 h5 i.f5 16 e4 dxe4 17 lLlxe4
27 a4 :d8 28 a5 :h7 29 :xh7 ~xh7 30 Wa5+ 18 i.d2 WdS 19 hxg6 i.xg6 20 i.c3
ll'la4 lDd7 31 lDe6 :b8 32 ~e2 ~g8 33 lOc7 21 :d1 'ife6 22 lLld4 led to an
ll'lac5 lDf8 34 lDd4 ~f7 35 :h 1 :c8 36 advantage for White in lzeta-Sulskis, Yerevan
~f3 :c7 37 ~g4 ~g8 38 :h2 ~f7 39 Olympiad 1996.
:h1 ~g8 40 lDde6 :c8 41 :h6lDxe6 42 14lDxg5 g6 15 lDxh7 ~xh7
ll'lxe6 i..xb4 43 :xg6+ ~f7 44 f5 i..f8 45 15...i.f5 16 e4 i.xe4 17 ttJxe4 ~xh7 18
lDxf8 ~xf8 46 e6 1-0 hS dxe4 19 hxg6+ followed by 20 Wxe4 is
. . . . - - - - - - - - - - - - - -... very powerful according to Petursson.
Game 18 16 h5 ~g7?
Romanishin-Ehlvest The decisive mistake according to
Biel SKA 1996 Petursson, who suggests that 16...f5! was the
only way to keep going. For example, 17 exf6
1 lDf3 d5 2 d4 lDf6 3 c4 e6 4 lDc3 iLe 7 (17 hxg6+ ~g6! 18 ttJe2 [18 We2 ttJg7 19
5 i..g5 0-0 6 e3 lDbd7 7 :c1 c6 8 i..d3 g4!? looks very dangerous for Black]
a6 9 c5 e5 10 dxe5 lDe8 11 h4 lDxc5 12 18... ttJg7 19 g4 ~f7 20 ttJg3 ~g8 holds
i..b1 lDe6!? according to Petursson) 17. Jhf6! 18 hxg6+
~g8 when it is not easy for White to bring
his queen into the act, e.g. 19 'i'e2 ttJg7.
17 hxg6 f5 18 'ile2! iLh4 19 g3 :h8 20
gxh4 :xh4 21 'ilf3 lDc7 22 ~e2lDe6 23
:xh4 'ilxh4 24 :h 1 1-0
Black can also attack his opponent's bind
with c4-cS in a different way.
Game 19
Eingorn-Balashov
Riga 1985
1 d4 li:Jf6 2 c4 e6 3 li:Jf3 d5 4 i..g5 i..e7
4.'?
5 liJc3 0-0 6 e3 liJbd7 7 ':'c1 a6 8 c5 c6 The alternative 14.. .1Ic8 15 :e1 tZJbd7 16
9 i.d3 b6 10 cxb6! 1i'b3 .i.a8 17 1i'a4 l:tb8 18 l:te2 was more
10 b4 as 11 b5 JLb7 12 bxc6 .i.xc6 13 pleasant for White in Gavrikov-Balashov,
cxb6 1i'xb6 14 t2Je5 offered White a small USSR Championship 1985.
edge in Vaganian-Anikaev, USSR 1979, but 15 i.xe7 'ifxe7 16 tLla4 1:.ab8 17 :c7
the text is better. 17 t2Jxb6 tZJxb6 18 :tc3! was even
10 ... c5 stronger according to Polugayevsky.
Black plays the ...c6-c5 break while he can. 17 .. :i'd6 18 ~xb6 ltJxe5 19 'ifc2 g6 20
The routine 10...1i'xb6 11 O-O! 1i'xb2 12 t2Ja4 'ifc5 ltJc4 21 ltJd7 :fc8 22 tLlf6+ <it>f8 23
1i'b7 13 t2Je5 c5 14 t2Jxd7 i..xd7 15 t2Jxc5 :xb7 :xc5 24 dxc5 'iff4 25 tLld7+ <it>g7
i..xc5 16 i..xf6 gxf6 17 1i'g4+ ~h8 18 'iih4 26 ltJxb8 ltJd2 27 :d1 'ifc4 28 c6 d4 29
f5 19 1i'f6+ followed by d4xc5 gave White a c7 ltJxb1 30 ltJxa6 d3 31 1:.b8 'ifc2 32
huge initiative in Hort-Portisch, Madrid 1:.f1 d2 33 1:.d8 1-0
1973.
11 0-0 i..b7 12 .i.b1! cxd4 13 exd4 Game 20
liJxb6 Zviaginsev-Kharitonov
Russia 1995
1 d4 ltJf6 2 c4 e6 3 ltJf3 d5 4 ltJc3 .i.e 7
5 .i.g5 0-0 6 e3ltJbd7 7 1:.c1 a68 b3!?
Qiestion 15. This just looks nice for Black!

A129lrer 15. Although Black's position is
optically attractive, he has difficulty finding
an active plan.
Qiestion 16. What do you mean? With this move, rather than try to refute
A129lrer 16. Create a normal IQP position 7... a6, White seeks to prevent his opponent
by moving the black pawn from d5 to b5. from implementing his plan without
The benefits are obvious: Black's light- undertaking any positional commitments. 8
squared bishop on b7 has an open diagonal a4 preventing ...b7-b5 after ...d5xc4 is a
and Black can use the semi-open d-flle to similar idea
attack the IQP. With the pawn on dS, Black's Qiestion 17. Doesn't this just weaken the
position is too rigid: White's centre is b4-square?
impervious to attack which gives him a free A 129lrer 17. It is always annoying to
hand to pursue his interests on the wings. concede a square in your territory like b4.
14ltJe5 However, it is really an aesthetic complaint
With the future idea of f2-f4-f5. rather than a real problem: what can Black
14... ltJfd7 do with this square? Meanwhile, ...a7-a6 also
44
Orthodox Variation (6 .. .tijbd7): Other Systems after 7 :'c 1
has drawbacks: it weakens the queenside dark 17... ~xe4!? 18 'ifxe4 'ii'xa2 19 .i.e7!? lUe8
squares, giving White the positional option of 20 d5! .i.a3! 21 d6 .i.xc1 22 :Xc1 'ii'a3 23
a2-a4-aS, tying down the black pawns on a6, lie3 was given as unclear by Zviaginsev, but
b7 andc6. 22 ...lidS! 23 'iixdS exdS 24 c6 1:txe7! 25 dxe7
QHstion 18. How should Black react? tiJf6 was very good for Black in Kragely-
A173tW' 18. The standard response is to Lazovic, Ljubljana 1996.
play 8...dxc4 9 .i.xc4 c5, transposing back 18 l:.b1 "a3 19 tDd6 :fb8 20 i..g3 a5
into the 7...dxc4 8 .i.xc4 a6 9 a4 c5 line, but 21 :b2 'iia4! 22 :fb1 'iixc2 23 :xc2
having avoided Zviaginsev's 9 .i.d3!? idea i..c6 24 tDe5 It)xe5 25 i..xe5 f6 26 i..g3
Black could also tty 8...c5 which is l:.d8 27 f3 l:.d7 28 l:.b3 i..d5 29 l:.e3 eS
considered at the end of this chapter. 30 tDxb5 i..c6 31 It)d6 exd4 32 l:.d3 i..c3
8 ...h6 9 i..h4 33 i..f2 l:.b8 34 i..xd4 l:.b1+ 35 ~f2
As always, 9 .i.f4!? is interesting. I quite i..e1+ 36 ~e2 1-0
like White after 9...i.b4 10 .i.d3 tiJe4!? 11 We now tum our attention to 7...dxc4 8
i.xe4 dxe4 12 tiJd2. .i.xc4 a6.
9 ... i..b4 10 i..d3 c5 11 0-0 cxd4
11.....aS 12 'iic2 .i.a3 13 :b1 (13 1:tcd1 Game 21
.i.b4 14 liJb1!?, intending tiJbd2 and e3-e4 Zviaginsev-Kharitonov
seems better) 13 ...i.b4 14 1:tfc1 .ta3 is Russian Team Ch, Kazan 1995
suggested by Kharitonov as an annoying line!
12 exd4 'iia5 13 'iic2 dxc4 14 bxc4 b5!? 1 d4 d5 2 c4 e6 3 It)c3 i..e7 4 It)f3 It)f6
15 c5! i..b7 5 i..g5 0-0 6 e3 It)bd7 7 btc1 dxc4 8
i..xc4 a6
16 i..e4!?
I would prefer a move like 16 tiJb1!? (or Qtestion 19. What is the point of this move
16 tiJe2!?) leaving the dark-squared bishop order?
hitting thin air and looking to snare it with a Anszrer 19. Black wants the advantages of
later a2-a3. 16... .i.dS (16 ... .i.xf3 gives up the the 7... a6 system without allowing White the
bishop pair and too many light squares, while opportunity to prevent ...dSxc4 with either 8
there is no obvious way to exploit the c5 or 8 b3. However, Black loses both the
weakened white kingside) 17 'iib2!? 'fight for the tempo' and a lot of his
(intending a2-a3) 17... .i.xa2 18 c6 followed flexibility: he is now fully committed to the
by .i.h4xf6 gives interesting chances. plan of queenside expansion.
16 ...lt)xe4 17 It)xe4 i..d5 9 i..d3!?
46
QIest:ian 20. What is the point of this? c6-square so that after 10 tiJe5 b5, 11 tiJc6!
An..w.e.er 20. This move is borrowed from gains the bishop pair with a clear advantage.
the QGA variation: 1 d4 dS 2 c4 dxc4 3 tiJf3 White thus prevents his opponent from
tiJf6 4 e3 e6 5 .i.xc4 c5 6 0-0 a6 (intending achieving the freeing sequence that he
...b7-b5) 7 .i.d3!? to meet 7...b5 with 8 a4! envisaged when he played 7...dxc4, and he
8...a6 prepares 9...b5 and then 10...cS. does so without giving conceding anything to
Obviously the white bishop on c4 will have Black in the fonn of a queenside weakness.
to move after ...b7-b5, so with 9 ..td3, White Moreover, against passive play, White will
makes the necessary move flrst. The point is cement his central presence with f2-f4.
that after 9...b5, White does not routinely 10 ... cxd4
castle, but instead plays 10 a4! and if 10 ...b4 10... tiJxe5 11 dxe5 tiJdS 12 ..txe7 'ilxe7 13
then 11 tiJe4! ~e4 is clearly better for White due to the
weakness of d6, according to Curt Hansen.
11 exd4ltJd5
Black must free his position in order to
develop.
12 i.xe7 "xe7 130-0 ltJxe5!?
13 ... tiJ7f6 was played in Kutirov-San
Segundo, European Team Championship,
Pula 1997, when 14 tiJxdS tiJxdS 15 ..te4
lId8 16 'ilf3 ~f6 17 :c4?! tiJxe4 18 ilxe4 f6
19 ttJf3 b5 20 :c6 1:[a7 was fme for Black. 16
'iib3 ~f6 17 i.f3 lhd4 18 lIfd1 is an
interesting pawn sacrifice, as Black is vexy
tied up.
'This prevents Black from playing ...c7-c5. 14 dxe5 :d8 15 "f3 :b8
If Black cannot play ... c7-c5, then he has to Black even has to be careful about his
accept a weak backward pawn on the half- king: 15...~c3?! is met by Curt Hansen's 16
open c-flie. 10... bxa4 11 tiJxa4! ..tb4+ 12 :xc3 l:tb8 (intending ... i.c8-d7-c6) 17
~e2! is the same. White's central king is quite i.xh7+ ~7 18 'ifhS+ ~g8 19 lIh3 f620
safe as Black cannot muster a central break exf6 ilxf6 21 'ifh7+ ~f7 22 lIf3 winning.
quickly enough to trouble him. 16ltJe4!
Qiestion 21. I know! Can't Black play ...c7-
c5 first, and then ...b7-b5 after?
Amr.rer 21. Yes he can, and this is where
.
m.
,
the second part of Zviaginsev' s plan comes
9 ... c5 10 ltJe5!
Qiestion 22. Well?
Amr.rer 22. With this move, White exploits
his opponent's early development of the
knight to d7 in two ways:
1. On d7, the knight does not pressure the
IQP (as it would from c6) so White is free to
move his knight from f3.
2. The knight on d7 no longer covers the White has a vexy pleasant position.
46
Orthodox Variation (6 .. . l'Dbd7): Other Systems after 7 Z:c 1
16 ...i..d7 17 -'93 .i.c6 18 :fd1 h6 9 ... c5

18...tDb4!? is an interesting idea of Curt 'This position can also arise after 7...a6 8
Hansen's. After 19liJf6+ ~h8 20 i.xh7 gxf6 a4 dxc4 9 i.xc4 c5.
21 "iWh4 ~g7 22 i.bl :xd1+ 23 :xdl :h8 10 0-0 cxd4 11 exd4!?
24 exf6+ 'iWxf6 25 "iWxb4 :114 Black has 11 .xd4 plays for a small edge due to
some compensation for the pawn. White's lead in development. The
19 a31 continuation l1.. .•aS 12 e4 (12 l:.fdl b6 13
Now Black no longer has this resource. i..f4 i..c5 14.d3 i.b7 15 tDd4 i.e7 16 .e2
19 ... l'Db6 20 l'Dd6 -.g5 21 i.e4 -'xg3 22 ended in a draw in Marin-Ubilava, Roses
hxgl l'Dd5 23 i.xd51 exd5 1992) 12...i..c5 13 .d3 i..b4 14 i..xf6 tDxf6
23 ... i.xdS 24 :c7 is veIY good for White. 15 e5 :d8 16 'ii'c2 i..xc3 17 .xc3 .xc3 18
24 l'Df5! ~f8 25 g4 :bc8 26 f3 with a :xc3 lDe4 19 l:.e3 tDc5 20 as i..d7 was
clear advantage to White which he roughly level in lzeta-San Segundo, Elgoibar
subsequently converted to vict0IY. (Sony, the 1994.
rest of the moves don't make senseO 11 ...lbb6 12 i.b3
The game Portisch-Chiburdanidze, Mon-
Carne 22 aco (Veterans-Women) 1994, saw the inter-
Spangenberg-San Segundo esting 12 ..te2, which worked out well after
Buenos Aires 1995 12...lDfdS 13 i..d2!? i..d7 14 as tDxc3 15
bxc3 i.a4 16 'ifel tDc8 17 tDe5 ttJd6 18
1 l'Df3 d5 2 d4 lbf6 3 c4 e6 4 lL\cl i.e7 i..d3 :c8 19 c4 i..f6 20 .e2.
5 i.g5 0-0 6 e3 l'Dbd7 7 :c1 dxc4 8 12 ...i.d71 13lbe5
i..xc4 a6 In Petursson-A.Sokolov, Reykjavik World
Cup 1988, White tried 13 as tDbdS 14 i..xdS
lDxdS 15 liJxdS i..xg5 16 liJxg5 'it'xg5 17
tDb6 :ad8 18 :c5 .f6 19 :el i.c6! with a
good game for Black.
13 ... i.c6!
9a4
Q/estion 23. Isn't 9 e4 possible here?
A11.Sl.a7 23. It is but it doesn't seem to
bring anything, e.g. 9...b5 10 i.d3 i.b7 11 e5
tDdS 12 i.xe7 .xe7 13 tDxdS i.xdS 14
:xc7 i..xf3 15 gxf3 :ad8 16 0-0.g5+ 17 This is Black's typical idea: due to the
'it>h 1 'ii'h4 gave Black good play for the pawn weakness of the white queenside in the wake
in IDescas-Garcia, Las Palmas 1989. of the restraining move a2-a4, Black does not
The text is the normal move, but it has mind allowing tDe5xc6 as this opens the b-
scored extremely well for Black in practice. ftle for Black to attack the white queenside
47
pawns. gives White a slight edge.

14 ild3 llJfd5! Qtestion 26. So what is interesting about
The standard exchanging manoeuvre. that?
15lLlxc6 Anszar 26. If White does not play iLf1-d3
15 iLb 1 g6 16 i.h6 iLg5! was fine for systems and wants to win the fight for the
Black in Vescovi-Hoffman, Buenos Aires tempo, he might play 9 'iic2, when 9...lLJe4!
1997. 10 i.xe7 "xe7 11 j.d3 lLJxc3 12 "xc3 dxc4
15 ... bxc6 16 i.c2 g6 17 i.xe7 ilxe7 18 13 i.xc4 b6 14 0-0 i.b7 gives Black an
a5 lLlb4 19 ile2 thd7 20 ilc4 :fb8 21 excellent version of a Lasker variation. We
i.e4 :'a7 22 :'fe1 ild8 23 :'a1 lLlf6 24 analysed this position in the note to White's
:'ed1 lLlbd5 25 ile2 :b3 26lLla4 :'b5 27 eighth move in Pinter-Prandstetter, but
i.xd5 cxd5 28lLlb6 :'c7 29 :'d3 lLle4 30 without the useful ...h7-h6 for Black! It also
f3 lLld6 31 :'c3 :'xc3 32 bxc3 lLlb7 33 transposes to I.Sokolov-Timman, Yerevan
ilxb5 axb5 34 a6 ilxb6 35 a7 b4 36 Olympiad 1996, when 15 :tfd1 c5 16 dxc5
a8i1+ ~g7 37 :'b1 b3 38 ila3 llJa5 39 lLJxc5 17 'iie5 :'acS 1S b4lLJd7 19 'iif4 :'fds
ilb4 ila6 40 :'e1 h541 h4 ~h7 42 ~h2 20 h3 lLJf6 gave Black comfortable equality.
ila7 43 ~h3 ilc7 44 :'a1 ilf4! 45 :'xa5 TIlls is yet another move order with which to
g5 46 g3 g4+ 47 fxg4 hxg4+ 48 ~h2 confuse your opponents!
ilf2+ 49 ~h1 ilf1+ 50 ~h2 ilh3+ 51 Qu;stion 27. Okay, well how about the
~g1 ilxg3+ 52 ~f1 ilf3+ 53 ~e1 g30-1 immediate 7...c5. Is it tactically bad?
Anszar 27. 7... c5 is a very natural move,
We now conclude our examination of 7 but it is virtually never played! S dxc5 seems a
:c 1 with Black's seventh move alternatives. natural reply, meeting S...dxc4 (S ...lLJxc5 9
Qiestion 24. I wanted to ask you two cxdS exdS 10 .i.xf6 i.xf6 11 lLJxd5 wins a
things! pawn) with 9 c6! bxc6 10 ~xc4 with a
A~ 24. Go ahead! First question? structural advantage. However, S cxd5 lLJxdS
Qiestion 25. Is the move 7...h6 a good 9 J..xe7 tiJxe7 10 J..e2 as in Geller-Larsen,
idea? Copenhagen 1966, is the standard
A~ 25. Interesting point! There are continuation when Black had development
several points to this move: problems after 10...b6 11 0-0 .i.b7 12 dxc5!
1. As Black has not committed his c-pawn lLJxc5 13 b4! lLJe4 14 li:Jxe4 i.xe4 15 'iia4
to c6, S i.f4 can be met by S...c5!? with :'cS 16 :'fd1lLJdS 17 bS i.g6 1SlLJe5!
unexplored play in a weird sort of i.f4 Another try is 7... a6 and only if S a4 then
system. S...cS, as the line 9 dxc5 dxc4 10 c6 bxc6 is
2. After S i.h4 c6 9 i.d3, Black plays much less effective now that White has
9...lbe4!? (not 9...dxc4 10 i.xc4 lLJdS as 11 weakened his queenside with 32-a4. 9 cxdS
i.g3! is annoying for Black) when 10 i.xe7 li:JxdS 10 i.xe7 'iixe7!? (10...li:Jxe7 is also
(10 j.g3!?) 10...lLJxc3! 11 lhc3 'iixe7 12 interesting) 11 lLJxdS exdS 12 b3 was tried in
cxdS leads to P .Nikolic-Yusupov in the Zlochevsky-Vukovic, Fonrua 1995, when
Lasker hybrids section of Chapter 1, while 11 12...cxd4 13 lLJxd4li:Jf6 14 J..e2 .i.e6 15 0-0
bxc3 'iixe7 12 cxdS! 'iixe7 13 c4 also was a little better for White.
48
Orthodox Variation (6 ... 0.bd7): Other Systems after 7 ltc 1
Summary
These are interesting lines with still many unexplored avenues. At the present tune
Zviaginsev's ideas (Games 20 and 21) seem the simplest and most promising for White.
1 d4 dS 2 c4 e6 3 ltJc3lbf6 4lbf3 i..e7 5 il..gS 0-0 6 e3lbbd7 7 :c1
7 ... a6
7...dxc4 8 i.xc4 a6 (D)
9 i.d3 - Game 21
9a4-Game22
S c5
8 b3 -Game 20
S ... c6 9 ~d3 (D) eS
9...b6 - Game 19
10 dxeSlbeS 11 h4liJxc5 12 J..b1 (D) f6
12...llJe6 - Game 18
13 'ifc2 - Game 17
8 ... a6 9 i..d3 12 i.b1
4'9
CHAPTER FOUR
Orthodox Variation (6 ...ti)bd7):

7 'iVc2 and Other
Seventh Moves
1 d4 d5 2 c4 e6 3 tDc3 i.e7 4 tDf3 tDf6 commonly an integral part of White's fIrst

5 Jtg5 0-0 6 e3 lLlbd7 incursions against Black's position. The
In this chapter we shall consider White's following scenario is typical: White lines up
alternatives to the standard 7 l:k 1 here. By his i.d3 and "c2 against the h7-pawn, and
far the most common of these is 7 'ii'c2 (for then establishes his knight on the advanced
7 'iib3, 7 cxdS and 7 i.d3 see Games 28-30 central outpost e5. Black cannot play
respectively) . ...liJd7xe5 as d4xe5 forces the knight on f6 to
Oetion 1. How does 7 'ii'c2 compare with move, when h7 hangs.
7%:.c1? 7 'ii'c2 is also a veIY flexible move.
Amzrer 1. In common with 7 %:.c1, 7 'ii'c2 Aggressive 0-0-0 options followed by a
establishes a presence on the c-file, which kingside hack are just as common as the
prevents Black from freeing his position with quieter positional options.
exchanges as 7...ttJe4 fails to 8 i.xe7 ii'xe7 9 Qiestion3. But I bet there are drawbacks!
cxdS liJxc3 10 "xc3 (here 10 dxe6 also does Ansrrer 3. Afraid so! Although the queen
the trick) 10...exdS 11 'ii'xc7. However, 7 was not developed on d1, it was influential.
'ii'c2 also starts White's play along the QIestion 4. What do you mean?
important b1-h7 diagonal. A nsrrer 4. In the opening, the battleground
QIestion 2. Why is this diagonal important? revolves around the c- and d-files.
Amzrer 2. For two reasons: With 7 %:.c1, White brought major pieces
1. Because the e4-square is on it. By to both of the important files. 7 'iVc2 by
maintaining his centre, Black provides an contrast just moves a major piece from one
advanced central outpost on e4 for his file to the other. Compared to 7 %:.c1,
knight. As we have seen, ...ttJf6-e4 is often therefore, Black will inevitably have more
used as a freeing manoeuvre to exchange a central freedom.
couple of minor pieces. If White can prevent
this option for his opponent with natural Game 23
development, then he is inevitably reducing Salov-Piket
Black's options and thus the flexibility of Madrid 1997
Black's position.
2. Early activity. The b I-h7 diagonal is 1 d4 d5 2 c4 e6 3 tDc3 i.e7 4 lLlf3 lLlf6
50
Orthodox Variation (6 .. . lDbd7): 7 1ic2 and Other Seventh Moves
5 i..g5 lDbd7 6 e3 0-0 7 'ii'c2 c5

is a very famous trap: 12...~7 13 hxgS+!
'it>g6 14 'De7 is checkmate) 11 tt:Jxe7 'ii'xe7
12 'DgS g6 13 0-0-0 offers White good
attacking chances.
10 lDxd5 exd5 11 1i.d3
White uses his presence on the b I-h7
diagonal to gain a tempo for development
while forcing Black to weaken his kingside.
11 ... g6
Qtestion 5. Why this rather than 11...h6?
Amzrer 5. With 11...g6, Black restricts his
opponent's activity to the greatest degree.
After 11...h7-h6, White maintains his breadth
of access to the b 1-h7 diagonal: for example,
Black chooses to free squares for his fS is now a particularly pleasant square for a
pieces with central liquidation, hoping that white knight (as it cannot be driven away by
this will enable him to solve the problem of ... g7-g6) or even White's queen or bishop.
his light-squared bishop. The quieter 7...c6 11...g6 takes control of the fS-square and
and 7...h6 are seen in Games 26 and 27. destroys White's avenue of activity on the
8 cxd5 b I-h7 diagonal. It also provides an outpost
Although this is White's most natural for his own light-squared bishop on fS.
move, he can also tty 8 0-0-0 (Game 25) and Qtestion 6. Anything wrong with it?
8 :d1 (Game 26). A mzrer 6. The drawback is that it puts
8 ...ltJxd5 another square on the same colour as his
8...cxd4 is featured in the next main game. bishop, which can cause problems later in the
91i.xe7 'ifxe7 endgame.
9...tjJxe71eads to positions similar to those 12 dxc5
after 8...cxd4 9 tDxdS tDxdS 10 i.xe7 'i'xe7 Black was threatening ...cS-c4, expanding
(see Game 24). If White wishes to avoid on the queenside with gain of tempo.
these lines, then 9 tDxdS exdS 10 i.xe7 12... ltJxc5 13 0-0 i.g4
'iVxe7 forces the game continuation as
9... i.xgS 10 h4! i.e7 (10....a5+?? 11 b4!
cxb412 'ii'xh7+!!
Qlestion 7. What is gomg on in this

position?
Amzrer 7. Black has reaped several benefits
51
by accepting the isolated pawn on dS: 2. It is Black's only weakness, and thus
1. The recapture ...e6xdS re-opened the easy to defend Weaknesses usually only
c8-h3 diagonal, thus solving the problem of become a problem when they are in pairs.
the light-squared bishop. Qtestion 12. Why is that?
2. By allowing his knight to be exchanged An..mu 12. The greater the number of
on dS, Black has avoided the inactive knights weaknesses, the more thinly you have to
that he gets after 9.. /lJxe7. 9...'W'xe7 spread your forces in order to defend them,
exchanged one of Black's potential problem and thus the more vulnerable your position
pieces, leaving him with free development becomes.
for the rest of his forces. Qtestinn 13. So why all the wonying about
3. Black has transposed to a position this position?
where his own pieces are more actively and Ansr.eu 13. The essence of the position is
sensibly-placed than White's: thus, the white that there is vel)' little in it, but anything that
queen is vel)' strange on c2 as it merely does exist belongs to White. Only White can
encourages Black to take the open c-fue with seriously entertain any hopes of winning.
tempo. Consequently, Black must be prepared to
Qtestion 8. So this is just fme for Black? settle for a draw here, as he has no real
A113Uer 8. Not so fast! Black has had to winning chances. The result of the game will
accept the permanent structural weakness of be decided in the psychological approach of
an IQP on dS, having already exchanged the both sides.
dark-squared bishops. Qlestion 14. What should Black's approach
Qtestion 9. Is this serious? be?
A113Uer 9. The dark-squared bishop An.flre" 14. Black must adopt the 'I'm
performs both defensive and attacking roles: annoying you' approach: 'Hah! I've solved all
it covers the weak dark squares - cS, e5 and my opening problems, and all you gave me in
d6 - arOlll1d the IQP, while catalysing black return was an IQP. I've emerged safe from
counterplay against White's kingside, the opening and you never even got the sniff
particularly against h2. Without it, Black's of an attack!'
position becomes rigid, preventing him from Qlestion 15. Hmm, I see. And White's?
exploiting the attacking features of the IQP: An..mu 15. White needs the 'We'll see in
the open lines and easy development it the end, young man' approach. 'Well, even if
provides, and the advanced knight outpost you are more active than me, and you hold
one4. the balance at the start, activity always has a
Qtestion 10. So Black isn't fme then? tendency to fade away, and then you'll just be
Anm:er 10. Not so fast again! Since Black left with one more weakness than me.
cannot develop a kingside attack, he must Consequently, I will always have something
channel his activity into another task: that of to play for. We're in for a nice long game
achieving ...d5-d4 and liquidating his here.'
weakness. Black can often have problems with his
Qtestion 11. I'm confused. What is your position on aesthetic grounds: whatever he
verdict on the position? does, his position always looks a little worse
A113Uer 11. Objectively, Black can be than White's, and it can get a little depressing
confident about his position. The weakness to look at if White hangs in and grinds.
of d5 is not so serious for two reasons: However, if you accept this, and a draw will
1. Black's pieces are more active than his really make you happy when you achieve it,
opponent's. then this is an excellent choice.
52
Orthodox Variation (6 .. ,&f:jbd7): 7 iic2 and Other Seventh Moves
Qiestion 16. I meant to ask. Couldn't Black :ac 1 were nonna!.

tty 13...ttJxd3? ~tion 19. Doesn't Black just play
AnJZ2l'Y'16. Yes, but this exchanges Black's 15... ~e6 here?
best minor piece for White's least effective Arzsrrer 19. This allows 16 'iYxc8+ winning
one. Black should aim to exchange the the two rooks for a queen.
knights. ~tion 20. Is this good for White?
Qiestion 17. Why is that? Anszm- 20. As White has no strucnu-al
AmTJX!r 17. With his knight, White can weaknesses, the queen has no clear targets.
attack squares of both colours. TIlls makes it Without these, the queen's mobility and
the ideal piece to draw out weaknesses in long-range power are of less use than the
Black's position, for example by attacking the combined action of two rooks on the open
queenside pawns and forcing them forward c-file.
The light-squared bishop is only good for the QIestion 21. So what can Black do?
one-dimensional task of attacking the d5- Arzsrrer 21. In Piket-Morovic, Wijk aan
pawn. Zee 1994, after 15 ~fc1, Black played
Qiestion 18. Well, then why doesn't White 15.....td7! preparing 16...ltJe6. After 16 'iVd2
play 13 %:tel to force 13 ...ltJxd3+? ltJe6 17 ..te2 "'f6 18 lbf3 ..tc6 19 :c3 d4!
Anszm- 18. Because the presence of 20 exd4 :d8 Black soon recaptured the d4-
White's king in the centre gives Black good pawn with equality. A little later at Dos
countetplay with 13...ltJxd3+ 14 'iYxd3 J.f5! Hennanas 1995, Piket tried the black side
(making use of 11...g6!) 15 'iYd4 (15 'iVxd5 against Shirov and after 15 :ac1 .ltd7 16
IUd8 16 'iVe5 'iVxeS 17 ltJxeS :ac8! gives "'d2 'iVf6 17 l:Hdl 'iVb6! 18 :c3 a6! 19 :dc1
Black the lnttlatlve according to ttJe6 20 :b3 'i'a7 21 ttJxe6 :xc1+ 22 'i'xel
Polugayevsky) 15...i.e4 16 0-0 .txf3 17 gxf3 ..txe6, his disadvantage was negligible.
'ii'g5+ 18 <it;hl "'f6 19 <ittg2 'i!VgS+ 20 'i!Vg4 QIestion 22. What is Salov's idea?
"'f6 and Black had equal chances in the Armrer 22. I'm not 100% certain.
game Cramling-Campora, Spanish Team Probably, it is to meet 1S.....td7 with 16
Championship 1994. 'iVb4!? preventing ...ltJc5-e6 due to the loose
14liJd4 :fc8 queen on e7.
QIestion 23. It doesn't seem that much.
Arzsrrer 23. No, but that's not the point.
Salov keeps the game going, and whilst the
game is still going, there is always the chance
of a win, especially for Salov! There is no-one
better at the 'we'll see in the end' kind of
position than him!
15 ...liJe4
Perhaps 15...ltJe6!?, although the retreat of
the bishop on g4 is then blocked
16 'iie1 i..d7 17 'iie2 l:::tc5 18 :lfc1 :lac8
19'iie1 'iif6 20 l:xc5lLlxc5 21 'ii'd2'iib6
22 b3 a5 23 l:c1 :1c7 24 h3 h5 25 i..e2
Black intends to challenge the knight on lLle4 26 'iWb2 l:xc1+ 27 "xc1 'iWb4 28 a3
d4 with ...It)c5-e6. 'iid6 29 'iib2 b6 30 i..f3 h4 31 b4 axb4
151Wd2!? 32 axb4 f5 33 'iWa2 ~g7 34'iia7 lLlf6 35
A new idea at the time. 15 :fel or 15 b5 ~h6 36 'fIa8 95?
fi.'?
Too risky. 36...if..e8 was still okay for is better here.

Black according to Tsesarsky. AmlW" 26. He is slightly better for two
37 "d8 c.1tg6 38 lDc6 g4 39 hxg4 fxg4 reasons:
40 .te2 c.1tg5 41 "xb6 i..xc6 42 "xc6 1. More active minor pieces. The contrast
"b4 43 "c7 g3 44 "e5+ ~g6 45 iLd3+ is seen clearly with the knights: White's
~f7 46 "f5 c.1te7 47 "e5+ rJitf7 48 "f5 knights from their outposts on c3 and d4
~e7 49 iLf1 "b2 50 "f4lDe4 51 "xh4+ prevent Black's from reaching their own
~e6 52 "g4+ rJite7 53 "h4+ rJite6 54 outposts on c6 ~'1d dS.
"g4+ c.1te7 %-% 2. Black's weak dark squares. The
exchange of dark-squared bishops has left
Game 24 Black with the vulnerable central squares d6
P.Cramling-Campora and cS. These are ideal squares for the white
Spanish Team Ch. 1993 knights to probe and loosen Black's
queenside structure. Since Black has dark-
1 d4 d5 2 lDf3 lDf6 3 c4 e6 4 lDc3 iLe7 square weaknesses, White will often aim to
5 i..g5 0-0 6 e3 lDbd7 7 "c2 c5 8 cxd5 exchange queens to remove his opponent's
cxd4 9lDxd4lDxd5 10 iLxe7lDxe7 best remaining defender of the dark squares.
Qlestiun 27. But White isn't going to get a
kingside attack, is he?
Anmer 27. No, but you can't get a kingside
attack every game! White's target is the black
queenside. Consequently, his light-squared
bishop is more likely to come to f3 (via e2)
than to d3!
Qlestion 28. How can Black develop?
A mlW" 28. Black has three methods:
1. ...lLld7-f6 and ....i.c8-d7.
2....b7-b6 and ....i.c8-b7.
3....lLld7-f6 and ...e6-eS.
Qearly, the last two are the most risky
QIestion 24. Black has got everything he since they both loosen Black's structure. The
wants here, hasn't he? move ...b7-b6 in particular greatly weakens
Anszrer 24. Black has achieved the fIrst of the queenside light squares. It also takes away
his aims. Through exchanges and the central the b6-square from the black queen, which is
break ...c7-cS, he has managed to free his its most comfortable post.
position slightly. Now he must attend to his Qlestiun 29. Hey, ...e6-eS looks like a good
attention to his development problems. idea!
QIestion 25. Development problems? He's Anmer 29. It is a key resource for Black.
ahead in development, isn't he? The ...e6-eS advance reopens the c8-h3
Anszrer 25. Although Black has castled and diagonal, providing the light-squared bishop
White hasn't, this is very easily remedied for on c8 with a choice of squares, and drives
White. His bishop will come to e2 or d3 and White's knight from its outpost on d4.
then his king will castle to safety. However, QIestion 30. Problems?
Black's logjam of queenside pieces will take AmlW" 30. First of all, it loosens Black's
longer to resolve. structure: the pawn on e5 will have to be
Q«stian 26. You seem to think that White defended by a piece since it is no longer part
54
Orthodox Variation (6 . .. tjjbd7): 7 "iic2 and Other Seventh Moves
of the pawn chain. Moreover, while it drives for Black in Kiselev-Arbakov, USSR 1987.
the white knight from d4, it does not create b) 12 0-0 i.d7 13 lUdl (13 'tWb3 'tWaS 14
any new outposts for Black's own knights. !tfdl !tac8 15 !tacl [15 llJdb5 i.c6 16 'iWa3
This is Black's main problem: when White's 'iVxa3 17 lDxa3 ltJedS was nice for Black in
knight is driven from d4, it can advance to b5 Peev-Ziatdinov, Belgrade 1990] is the ECO
to come to d6, or if it drops back to b3, it recommendation and seems very reasonable,
will jwnp to as or c5 to attack the b7-pawn. although 15... b6 16 lDdb5 ltJedS 17 ltJxdS
Black's knights have no prospects: they can ~xdS 18 ':xc8 :xc8 19 e4 ~f4 20 i..f1 is
occupy safe squares, but they are not actively not as large an advantage as claimed)
placed. 13...iih6 14 'i'b3 'ii'xh3 15 ~xb3 l:fd8 16
QIestian 31. How serious is this? ltJd4 ltJc6 17 ttJxc6 (17 i.f3 llJe5) was
Answer 31. These are not life-threatening agreed drawn in Kharitonov-Komarov,
factors, but once you are aware of them, you Leeuwarden 1995.
can understand why Black often has to suffer 12... e5 13 ~db5 .i.f5 14 'ifb3 'ifb6 15
a bit at the beginning of the middlegame. 0-0 a6 16 ~d4 'iWxb3 17 ~xb3 .i.d3 18
11 i..e2 :tfd1 e4 19 .i.e2 .i.e2 20 lId6 i.xb3 21
11 O-O-O!? was the enterprising choice in axb3 :tfd8 22 :tad 1 ~e6 23 g4
Ruzele-Khurtsidze, Groningen 1996, when
11...'ii'aS?! 12 g4! ~f6 13 ~b3 'iWc7?t 14 g5
ttJfd5 15 ~xdS 'ii'xc2+ 16 ~c2 ttJxd5 17
i..g2 was very pleasant for White. Black
should have played her queen to b6 when 12
i..d3 ~f6 13 g4 transposes to Khenkin-
Arbakov, USSR 1987, after which 13 ...e5! 14
g5 exd4 15 gxf6 'ii'xf6 16 exd4 i..g4 was
good for Black.
Qtestion 32. Doesn't 11 i..d3 gain a tempo
against h7?
Answer 32. Black's favoured development
scheme is ...~d7-f6 followed by either ...e6-
e5 or ....tc8-d7. Consequently, after 11 .td3 White is pressing, but Black defends well.
ltJf6 (protecting h7) is Black's intention 23 ... h6 24 h4 :'xd6 25 :'xd6 :'e8 26 g5
anyway. White's bishop is better on e2, from hxg5 27 hxg5 ~h7 28 g6 ~g5 29 i..c4
where it can move to f3, eyeing the <iPf8 30 gxf7 :'e7 31 <iPg2 ~xf7 32 i.xt7
vulnerable b7-pawn. Wxf7 33 <iPg3 ltJa5 34 b4 ltJc4 35 litd4
Novikov-Gorelov, Pavlodar 1987, ~xb2 36 liJxe4 b6 37 <it'f4 a5 38 bxa5
continued 12 0-0 i.d7 13 IUd1 (13 'ii'b3 bxa5 39 :'d2 liJe4 40 :'d4 ltJb2 41 :'d2
'ii'aS!) 13 ...'ii'b6 14 l:acl l:ac8 15 'ii'b3 'iVxh3 liJe4 42 :'d4 :'e7 43 :d8 :a7 44 :'d4
16 ~xb3 b6 (16...l:Ud8 and 16...~edS!? are ~b6 45 :d6 ~d7 46 liJe3 ~e7 47 :g6
suggested by Gorelov) 17 ttJd4 ~fd5 18 ~f7 48 :'d6 a4 49 ~b5 :b7 50 liJd4
llJxdS ~xdS 19 i..e4llJf6 20 i..b7 ':xcl 21 ~b6 51 lDe2 ~e7 52 :g6 <iPf7 53 :'e6
':xc1 :d8! with equality. lDd7 54 :a6 :'b2 55 :a7 ~e8 56 lDa3
11 ... ~f6 12 i..f3!? ~e5 57 :a5 :xf2+ 58 <ite5 ~d7+ 59
Two other moves have been tried here: ~d4 :d2+ 60 <ite3 :'e2 61 ~e4 :a2 62
a) 12 'iib3!? a6!? 13 0-0 'iWaS 14 i..f3 e5! ll)d6+ <ite 7 63 liJf5+ <itf6 64 liJxg7 <itxg7
15 ~c2 ':b8 followed by ...i.c8-e6 was fme 65 :a7 ~-%
55
The 8 cxdS variation is the main line in the An3la'Y 35. This is a good solid move,
7 'iW c2 systems. White does have an though it lacks the punch of 8 0-0-0, as
aggressive alternative, however: 8 0-0-0. White still has to castle his king to safety.
Play might continue 8...cxd4 9 li.Jxd4 (9 exd4
b6 10 ~d3 dxc4 11 hc4 .ib7 is fme for
Black) 9... ~b6!? (9 ... dxc4 10 .txc4 'i'aS!
[hitting the bishop on g5 and thus freeing the
knight on d7 with tempo] 11 .ih4 lDeS 12
.te2 lDg6 13 i.g3 eS [13 ....id7 14 0-0 was
agreed drawn in Dreev-Balashov, St.
Petersburg Zonal 1993, but Korchnoi's 14
lDb3 'ib6 15 h4! ~fc8 16 hS li.Jf8 17 h6 g6
18 0-0 looked vexy strong against Osnos,
USSR Championship 1963] 14 ~b3 'ii'b6
when 15 0-0 [15 h4 hS! 16 .id3 .tg4 17 ~c1
e4! 18 .ixe4lDxe4 19 'ii'xe4 :fe8 gave Black
QIestion33. Wow! Can White do this? good compensation in Gorelov-Arbakov,
Arl3lW' 33. While the intention of 8 0-0-0 USSR 1987] 1S ... J..e6 16 .id3 [stressing the
is clearly attacking, it also contains certain vulnerability of the eS-pawn] 16...i..d6 17
positional features. Most importantly, 8 0-0-0 .tfS c4 18 :fe 1 gave White an annoying
brings a rook to the d-ftle. This initiative in CHansen-Kveinys, Groningen
1. Ensures that White has a major piece 1990) 10 .te2 .id7! 11 .txf6 .txf6 12 cxdS
on each of the 'battleground' fues. lL\xdS 13 ~xdS exdS 14 0-0 'i'b6 15 :d2
2. Puts pressure on the black centre. ~ac8 16 'ib 1 g6, which was approximately
3. Places the rook opposite the queen on equal in Timoshchenko-Kharitonov, Frunze
d8, which may cause problems once the 1988.
centre is cleared of pawns. 8 dxcS is the other sensible move, but
Moreover, White possesses several 8...lDxcs 9 ~d1 'ii'aS 10 cxdS exdS 11 lDd4
'pressure points' on the black position: .id7 (11...lDce4 12 ~b3!) 12 .id3 .ia4 13
1. The unresolved central tension. lDb3 .txb3 14 axb3 li.Jce4 15 .txe4 dxe4 16
2. The h4-d8 diagonal, due to the bishop i..xf6 .ixf6 17 0-0 hc3 gave White nothing
ongS. in Tisdall-Ostenstad, Norwegian Champ-
3. The b1-h7 diagonal in the form of a ionship 1996.
future 'it'c2 and .id3 battexy.
Consequently, as well as being an Game 25
aggressive continuation, 8 0-0-0 is also well- Orsag-Bellini
founded positionally, which makes It a Montecatini 1997
dangerous continuation for Black
QIestion 34. Isn't it a bit risky? 1 d4 d5 2 liJf3 liJf6 3 c4 e6 4 liJc3 i.e 7
Anszrer 34. That is true. The drawback to 5 i.g5 0-0 6 e3 liJbd7 7 "ifc2 c5 8 0-0-0
castling queenside is that it places both the b6!?
queen and the king on a-fue that will soon be QIestion 36. Is this good?
opened This inevitably gives Black some Anszrer 36. I really like this move. Black
attacking and tactical ideas of his own. prepares to develop the light-squared bishop
QIestion 35. In that case, why doesn't on b7 and then to bring his queen's rook to
White just play 8 J:dl? the important c-ftle.
56
Orthodox Variation (6 .. . li:Jbd7): 7 "ilc2 and Other Seventh Moves
Alternatively: Tmava 1988, while 9 dxc5 lDxc5 10 cxdS

a) S...dxc4 (S...h6 9 h4Q 9 .i.xc4 'ii'aS 10 lDxdS 11 h4 .i.d7!? 12 tDxdS exdS 13 :xdS
dS! 4Jb6 11 d6! .i.d8 12 .i.xf6 .i.xf6 13 tDe4 ];rcS 14 ~bl .i.a4 15 'ii'c4 'iie8 16 .i.xe7
i.d7 14 lDxf6+ gxf6 15 h4! tDxc4 16 'iixc4 \ixe7 17 .i.e2 .i.c6 18 :d4 lDe4 gave Black
e5 17 h5 was very good for White in good play in Kiselev-Kveinys, Warsaw 1991.
Browne-I.Ivanov, USA 1995 9 ...lLlxd5 10 i..xe7 'iVxe7 11 ~xd5 exd5
b) S.. :i'aS 9 h4 4Jb6?! 10 .i.dJ! h6 11 12 dxc5 lbxc5 13 <Jttb 1 i..d7 14 i.d3 96
i.xf6 .i.xf6 12 g4 cxd4 13 exd4 dxc4 14 1 5 lLld4 i..a4!
iLh7+ ~h8 15 g5 i.e7 16 gxh6 gxh6 17 Weak.ening the white queenside.
l:.dg1 gave White fantastic attacking chances
in Vyzmanavin-Ruban, Sochi 1989. Mikhail
Gurevich considers that 9...cxd4 10 4Jxd4
(10 exd4 .i.b4!? offers Black counterplay)
10....i.b4 11 4Jb3 'iVb6 12 cxdS i.xc3 13
'iixc3 4JxdS 14 'i'd4 f6 15 .i.f4 'i'xd4 16
tDxd4 lLl7b6 followed by ...e6-e5 equalises
forBlack.
c) 8...cxd4 9 4Jxd4 (9 exd4!? h6 [9...b6!?]
10 i.xf6 4Jxf6 11 ~bl i.b4 12 c5 i.d7 13
lLle5 .i.xc3 14 \ixc3 as 15 .i.d3 was a little
better for White in Moskalenko-Hoffman,
Benasque 1993) 9...lLlb6!? (9 ... a6 10 h4 dxc4
11 iLxc4 'i'c7 12 i.e2 l:.e8 13 .i.f4 e5 14 16 b3 i.d7 17 h411fc8 18 'iWe2 a5 19 h5
.i.g3 4Jb6 was fme for Black in Cvitan- a420 hxg6 hxg6 21 llh2 axb3 22 axb3
Hoffman, Bern 1992) 10 ~b1 .i.d7 11 .i.xf6 lbxd3 23 llxd3 il.b5 24 lbxb5 'iVe5 25
.i.xf6 12 c5 4JcS 13 iLd3 h6 14 g4 iLxd4 15 lLld4 'iVxh2 26 g3 'i'h1+ 27 lld1 'iVe4+ 28
exd4 b6 16 :hgl 4Je7 17 g5 h5 18 g6! ~b2 'iVe7 29 ~b1 :a3 30 'iVb2 'iVe4+ 31
looked very dangerous for Black in Browne- lbc2 llxc2 0-1
Llvanov, USA Championship 1992. QIestion 37. Can Black use the queenside
plans like ...a7-a6 against 7 'i'c2?
Anscrer 37. White's simplest response is 8
cxdS, when 8...exdS 9 .id3 c6, transposes
into an Exchange QGD where Black has
made an unprovoked queenside weakness
with ... a7-a6. Also 8 c5!? c6 9 .idJ,
preventing ... e6-e5, is tempting.
QIestion 38. Okay, well how about 7...c6?
A nmer 38. This slower plan of
development has recently become
fashionable.
Game 26
9 cxd5 Garcia lIundain-Ubiiava
9 e4 dxe4 10 lLlxe4 iLb7 11 i.d3 lDxe4 12 Ampuriabrava 1997
.i.xe7 'i'xe7 13 i.xe4 .i.xe4 14 'ii'xe4 4Jf6
was flne for Black in Ftacnik-Balashov, 1 lbf3 li:Jf6 2 d4 d5 3 c4 e6 4 lbc3 i.e7
57
5 ~g5lbbd7 6 e3 0-0 7 'tie2 e6 8.l:.d1 This line is worth more tests.

White places his rook on a more natural The fmal idea is 7... h6. As we have seen,
square than in the 7 l:.c1 and 8 'iii' c2 lines, this move would be useful in many lines, but
opposite the black queen on dS. However, the following game has always been
the simple 8 ~d3 worked out well in considered the big problem!
Korchnoi-Hoffman, Salamanca 1991, after
8...dxc49 .ixc4 tiJd5 10 h4!? l:te8 11 e4lbb4 Game 27
12 'iVe2 cS 13 O-O-O! cxd4 14 lhd4 tiJc6 15 Kasparov -Portisch
J:td6!! f6 16 .ixe6+ ~h8 17 ':'d2 fxgS 18 Brussels 1986
J:thdl.
8 ... b6!? 1 d4 tDf6 2 c4 e6 3 liJf3 d5 4 liJe3 i.e 7
5 i.g5 0-0 6 e3 liJbd7 7 'ife2 h6 8 cxd5!
Black aims to develop his bishop to b7

and then later to free himself with ... c6-cS. 8 .tf4 is the other typical idea, e.g. B...c5 9
9~d3 dxc5 (9 cxd5 tiJxdS 10 tiJxdS exd5 11 .te2
This loses the fight for the tempo, but cxd4 12 exd4 .ib4+ 13 .id2 .id6 14 0-0
White needs to develop in response to ... b7- l:.e8 was level in Bezold-Lengyel, Budapest
b6. Moreover, any subsequent gain of 1993) 9...lbxc5 10 .te2 dxc4 (10 ...b6 11 J:td1
queens ide space with ...b6-bS will give the .ib7 is equal according to Ftacnik) 11 .txc4
tempo back to White. a6 12 a4 b6 13 0-0 ~b7 14 l:.fd1, as in
9 ... dxe4! Tisdall-Ostenstad, Gausdal 1993, and now
9...h6 10 ~h4 .ib7 11 0-0 c5 12 .tg3 14.. :ii'cB 15 tiJe5 J:td8 was best according to
cxd4 13 exd4 dxc4 14 ~xc4 .ixf3 15 gxf3 Ftacnik with a tiny edge for White.
lbhS 15 d5! gave White the initiative in 8 ...exd5
Kasparov-Amura, Buenos Aires simultane- Portisch gives the stunning line 8... hxgS 9
ous 1992. The ....ib7xf3 exchange ruins dxe6 fxe6 10 tiJxg5 lZJb6 11 h4! c5 12 h5
White's kingside pawns, but loses control cxd4 13 h6! dxc3 14 J:td1 'iVeB 15 hxg7 ~g7
over all Black's central light squares. 16 J:th7+ ~g8 17 l:.d4!, intending J:td4-h4
10 ~xe4liJd5 11 i.xe7 'ii'xe7 12liJxd5 with a big attack.
12 0-0 .ib7 13 tiJxdS led to a draw in 9 i.f4! e5
ASokolov-Landa, Novgorod 1997. 9...c6 10 0-0-0 with h2-h3 and g2-g4 to
12 ... exd5 13 ~d3 liJf6 14 0-0 i.d7 15 follow is clearly better for White according to
liJe5 :ae8 16 'ii'e2 :e7 17 i.a6 ~e8 18 Portisch. That ...h7-h6 move is a real
:Ie1 lbe4 19 ~d3 .i.b7 20 l:lxe7 %-% weakness for Black.
58
Orthodox Variation (6 .. . t'iJbd7): 7 1rc2 and Other Seventh Moves
10 .i.e2 b6 11 0-0.i..b7 12 :lfd1 :lea 13

dxe5 bxe5 14 a4!
7 ... a6!?
7...c5 8 cxdS tiJxdS 9 J..xe7 tiJxe7 10 dxc5
14 ...'ila5? tiJxcs 11 'ifa3 is a slight advantage to White
14... aS was absolutely necessary according according to Akopian. 7...c6 is the main
to Portisch. move, e.g. 8 J..d3 dxc4 (8 ...a6 9 cxdS cxdS 10
15 ttJh4! 1:fdS 16 ttJf5 i.fS 17 ttJb5 ltJeS 0-0 bS 11 a4 was good for White in Akopian-
1S i.d6! ttJxd6 19 liJfxd6 1:bS 20 ltJxb7 Ubilava, Manila Olympiad 1992) 9 'it'xc4
1:xb7 21 1:xd5 tiJd5 {9 ...c5 10 dxcS lLlxc5 11 J..c2 'i'aS
Now all Black can do is suffer. [11... 'iib6 looks more normal] 12 0-0 tiJcd7
21 ...1:dbS 22 'ifd2! "ii'xd2 23 1:xd2 liJf6 13 :fd1 tDe5 14 lLlxeS "xeS 15 lLle4! gave
24 :a2 ltJe4 25 :e2 1:d7 26 93 a5 27 White a clear advantage in Akopian-San
'itg2 g6 2S i.f3 ttJf6 29 lLla3 i.d6 30 Segundo, Madrid 1997] 10 J..xe7 'it'xe7 11
i.e6 l::ddS 31 1:a 1 .i.e5 32 i.b5 ttJd5 33 0-0 t2Jxc3! 12 'i'xc3 (12 bxc3!?) 12 ... c5 13
:b1 i.d6 34 1:d2 liJb6 35 1:e1 J.e7 36 l:.ac1 b6 and now 14 "c21? h6 15 i..e4 :b8
l::e2 :beS 37 liJb1 ~g7 3S liJd2 1:aS 39 16 'it' a4 transposes to the Lasker lines
ttJb3 :deS 40 :ee2 e4 41 liJd2 1:a7 42 covered in Chapter 1.
ltJxe4 ltJxc4 43 :xe4 l::xe4 44 1:xe4 f5 S exd5 ttJxd5 9 .i.xe 7
45 h3 h5 46 g4 hxg4 47 hxg4 fxg4 4S Short's intention was 9 tiJxdS .J..xg5 10
~g3 i.d6+ 49 ~xg4 :Ie7 50 .te6 1:f7 51 t2Jxg5 exdS 11 "xdS c6! 12 'ii'fs tDf6 13
f4 ~h6 52 i.d5 :f6 53 ':e1 ~g7 54 b3 'fics tDd7 14 'fif5 tDf6 with a draw by
:as 55 ':d1 i.e5 56 :d3 .i.a3 57 .i.e4 repetltIOn.
i.e1 58 :d7+ ~h6 59 :e7 i.d2 60 ~f3 9 ... ttJxe7 10 i.e2 b6!?
i.b4 61 :b7 i.e3 62 i.d3 ':f6 63 ~g4 10...c5 leads to a version of the 7 'i'c2 c5
i.d2 64 f5! 1-0 8 cxdS tDxdS 9 J..xe7 'fixe7 line where White
Finally, we take a brief look at the has his queen on b3 and Black has played the
remaining alternatives. unusual ... a?-a6. These changes should
r------------------.
Came 28
normally favour White rather than his
opponent.
Akopian-Short 11 0-0 .i.b7 12 1:fd1 lDf6 13 1:ae1 ttJg6
European Team Ch, Pula 1997 14 ttJe5 'ile7 15 il.f3 J.xf3 16lLlxf3 1:feS
17 g3?!
1 lLlf3 d5 2 d4liJf6 3 e4 e6 4 lDe3 liJbd7 17 e4! b5 18 eS lLld7 19 lLle4 is
5 J.g5 i.e7 6 e3 0-0 7 'ilb3 recommended by Tsesarsky as a slight edge
for White, which seems correct. 'ilc7 25 'ilb4 'ilc1+ 26 ~h2 'ilc7+ 27
17 ... b5 18 a4 l1ab8 19 axb5 axb5 20 ~g1 'ilc1+ 28 ~h2 'ilc7+ %-Yz
ltJe2 c6 21 l1c2 e5 22 dxe5 lLlxe5 23
ltJxe5 'ilxe5 %-Yz Game 30
This interesting line needs more tests. llinsky-Nenashev
Bishkek Zonal 1993
Game 29
Lputian-Cifuentes Parada 1 d4 d5 2 c4 e6 3 lbf3 lLlf6 4 ltJc3 i..e7
Ubeda 1996 5 i..g5 0-0 6 e3 lLlbd7 7 i..d3 dxc4 8
i..xc4 a6 9 a4 c5!
1 d4 ~f6 2 c4 e6 3 lbf3 d5 4 i..g5 i..e7 The simplest, transposing into the IQP
5 lbc3 0-0 6 e3 lbbd7 7 cxd5 position we saw in the previous chapter, but
The idea of this line is just to transpose without the useful tempo move :al-c1 for
into a normal Exchange QGD after 7...exdS, White!
but Black has a different possibility.
7 ... ~xd5!? 8 i..xe7 "ii'xe7 9 ~d3 lDxc3
10 bxc3 c5 11 0-0 b6
10 0-0 cxd4 11 exd4 lbb6 12 i..b3 i..d7

13 a5l2Jc8!? 14lDe5 i..c6 15 i..xf6 i..xf6
16 ':e 1 lbe 7 17 lbxc6 bxc6 18 ':a4 lbf5
12 a4 19 lbe2 "ii'd7 20 i..c2 ':ab8 21 "ii'd3
12 'iic2 h6 13 a4 i.b7 14 e4 :fc8 15 'iie2 ':xb2 22 g4 ':xc2 23 "ii'xc2 lbxd4 24
lLlf6 16 as 'iid8 was a little better for White lbxd4 i.xd4 25 ':a3 c5 26 "ii'e4 c4 27
in Notkin-Kharitonov, Russian Champ- ~g2 e5 28 l:b1 c3 29 ':a2 'i'e6 30 ':c2
ionship, Elista 1994 h5 31 h3 hxg4 32 hxg4 g6 33 l1b7 Wg7
12 ...i..b7 13 a5 e5 14 i..b5 exd4 15 34 ~g3 :h8 35 ~g2 l:h4 36 f3 :th8 37
cxd4 cxd4 ':c7 "ii'f6 38 ~f1 ':h1+ 39 'it>e2 "ii'd6 40
15...1L1f6 16 axb6 axb6 17 .ie2 ~fd8 was 'i'c6 "ii'd8 41 l:d7 "ii'g5 42 ~d3 :td1+ 43
agreed drawn in Itkis-Kharitonov, Moscow ~c4 ':d2 44 ':c1 ':h2 45 l:e1 "ii'f4 46
1995. "ii'e4 "ii'g5 47 ':xd4 exd4 48 "ii'xd4+ Wh7
16 lbxd4 bxa5 17 i..xd7 "ii'xd7 18 ':xa5 49 "ii'xc3 "ii'f4+ 50 l:e4 "ii'c7+ 51 Iit>d4
%Ud8 19 h3 g6 20 -.a1 a6 21 ':b1 ':ac8 "ii'd6+ 52 ~e3 f5 53 gxf5 gxf5 54 ':d4
22 :tc1 :txc1+ 23 "ii'xc1 :tc8 24 "ii'b2 We5-+- 0-1
60
Orthodox Variation (6 .. .liJbd7): 7 .c2 and Other Seventh Moves
Summary
To my mind, 7 'ii'c2 is a less promising option against the Orthodox than 7 I:.c1 (Chapters 2
and 3). In particular the middlegame after 7...cS 8 cxdS tiJxdS 9 .lixe7 'ii'xe7 10 tiJxdS exdS
seems a simple way for Black to play for a draw, whilst Ubilava's 7...c6 followed by 8...b6 has
defeated all White's attempts so far.
1 d4 d5 2 c4 e6 3 lbc3 lbf6 4 lbf3 .i.e7 5 .i.g5 0-0 6 e3 lLlbd7
7 'fic2 (D)
7 'iVb3 - Game 28
7 cxdS - Game 29
7 i.d3 - Game 30
7 ... c5
7...c6 - Game 26
7...h6 - Game 27
S cxd5 (D)
8 0-0-0 - Game 25
S ...lbxd5 (D)
8...cxd4 - Game 24
9 i.xe7 - Game 23
7'ikc2 Bcxd5 8 ... tjjxd5
61
CHAPTER FIVE
Tartakower Variation:
Fixed Centre Plans
The T artakower variation anses after the Qu::stian 2. What do you mean by
sequence 'hannonious'?
1 d4 d5 2 c4 e6 3 lbc3 i.e 7 4 lbf3 lDf6 Anszrer 2. Successful development
5 i.g5 h6 6 i.h4 0-0 7 e3 b6 depends on two general factors: your pieces
Qu::stian 1. What is happening here? should have their own space - they shouldn't
An.mu 1. The most far-reaching get in each other's way; and they should
development decision in any opening is the contribute to a common purpose.
placement of the bishops. Since their Qu::stian 3. So what are Black's pieces
development requires a preceding pawn doing here?
move, the mobilisation of the bishops alters Anszrer 3. Once Black has completed his
the pawn structure and inevitably creates minor piece development with ... ~c8-b7 and
some weakness in the position. Ironically, the ...ttJb8-d7, his next opening goal is to break
fates of the bishops are invariably entwined, in the centre with ...c7-cS. Logically his
with success for one leading to penwy for development should support both this break
the other! Thus in queen's pawn openings, and his centre which will come under greater
the dark-squared bishop is naturally activated strain once the central tension increases.
on the f8-a3 diagonal as ...e7-e6 is necessary Qiestion 4. Right! How is the dS-pawn
to Black's central control, but of course this protected?
blocks the access of the light-squared bishop Anszrer 4. The bishop on b7 and the
to the c8-h3 diagonal. With 7... b6, Black knight on f6 support the dS-pawn directly.
prepares the most hannonious fonn of The bishop on e7 and the knight on d7
development in Queen's Gambit openings, support dS indirectly by countering the
opening the long a8-hl diagonal to the light- pressure exerted by White's dark-squared
squared bishop by freeing the b7-square. We bishop on h4: the bishop on e7 breaks the
can also see this scheme of development in pin on the knight on f6, while i.h4xf6 can be
the QGA (1 d4 d5 2 c4 dxc4 3 tt.Jf3 tt.Jf6 4 e3 met by ...tt.Jd7xf6 maintaining a knight's
e6 5 ~xc4 cS 6 0-0 a6 7 ~e2 bS 8 ~b3 protection of dS .
.i.b7) and the Semi-Slav (1 d4 dS 2 c4 c6 3 Qu::stian 5. Okay! And how is ...c7-cS
tt.Jf3 tt.Jf6 4 ttJc3 e6 5 e3 tt.Jbd7 6 ~d3 dxc4 7 supported?
.i.xc4 bs 8 ~d3 ~b7). An.mu 5. Black's central break is
62
Tartakower Variation: Fixed Centre Plans
supported by the bishop on e7 and the In both these cases, at the expense of his
knight on d7. Most importantly, 7... b6, which uncastled king, Black has gained territory on
solved the problem of Black's light-squared the queenside. By playing ... a7-a6 and ...b7-
bishop, provides pawn support so that after bS, Black has created space behind his
c4xdS ...e6xdS, d4xcS, Black can recapture queenside pawns into which he can safely
with ...b6xcS and avoid the IQP! move his queen; thus Black will connect his
Somehow, all Black's development unites rooks and link up his position.
around Black's central goals which makes his
position very solid and hannonious.
QIestinn 6. So is this a miracle cure or are
there some drawbacks to 7... b6?
Ansuer 6. The most visible drawback is
that Black weakens his queenside light
squares by abandoning his pawn protection
of c6 and a6. The weakness of the c6-square
is particularly important as White has varied
means of targeting this square, for example
by opening the c-fue with c4xdS and then
playing :tal-d, or by occupying his central
outpost with tDf3-eS. The second drawback
is less obvious and concerns Black's major In the T artakower, with the pawns on a7
pieces and his queen in particular. and b6, Black's queen enjoys no such resting
QIestinn 7. What do you mean? place, and it thus becomes harder to
Ansuer 7. It's easiest to demonstrate this complete Black's development by connecting
by comparing it to other Queen's Gambit the rooks. Once his central break ... c7-c5 is in
openings: the QGA - 1 d4 dS 2 c4 dxc4 3 sight, Black will have to solve this last
~f3 a64 e3 ~f6 5 ii.xc4 e6 6 0-0 cS 7 'iWe2 development problem, for example with the
bS 8 i..b3 i..b7 Lasker manoeuvre ... ~f6-e4 to exchange the
dark-squared bishops and free a post on e7
for the black queen.
QIestinn 8. Oh dear! This sounds serious.
Doesn't it?
Anszrer 8. No! These are subtle points -
none of them are remotely fatal! However,
by appreciating them, we can better
understand the thrust of White's efforts to
gain an advantage.
In this chapter we shall deal with schemes
for White that involve fIxing the centre
pawns. The next chapter will then deal with
routine development plans. If White wants to
and the Semi-Slav - 1 d4 dS 2 c4 c6 3 ~f3 fIx the centre, he can either exchange on d5
~f6 4 ttJc3 e6 5 e3 ~bd7 6 ii.d3 dxc4 7 immediately (as in Games 31-33) or first
ii.xc4 bS 8 ii.d3 ii.b79 0-0 a6 10 e4 cS capture on f6 to prevent Black from
see fol/owing diagram recapturing on ciS with the knight (Games
34-47).
63
uueen's Gambit Declined
White's first attempt is 8 cxd5. the Orthodox systems in order to free his
Q«:stian 9. What is White aiming for with pOSItIOn.
this exchange? Q«:stian 12. How is that?
Amrar 9. This straightforward move has An.wrer 12. After 9 i..xe7 'ilxe7, Black has
several ideas: already solved one development problem: by
1. White crosses his opponent's desire to securing the e7 post for his queen, Black
activate his light-squared bishop on b7 by ensures that he will be able to connect his
forcing a black pawn to occupy dS and block rooks and complete his development.
the a8-h 1 diagonal. Moreover, after 10 liJxd5 exd5
2. White opens up the c-flie and stabilises
the centre, enabling him to target Black's c-
pawn and the c6-square with lIal-c 1 in
combination with tDf3-e5.
3. By settling the central structure at this
early stage, White reduces the number of
possible pawn structures to a minimum
which makes his choice of development
scheme a great deal easier (though of course
Black can also benefit from this).
8 ...lDxd5!
the exchange of two sets of minor pieces

removes any spatial worries that Black might
have had.
Qtestion 13. Why does White play 10
tDxdS?
Arzszrer 13. It is the consistent follow-up to
8 cxdS - White must play 10 tDxdS to force a
black pawn to the dS-square.
Q«:stian 14. All the same, White has
achieved what he wanted, despite 8...~xdS.
Arzszrer 14. Yes, but with the exchanges
Q«:stian 10. Why? What's wrong with secured by 8...~dS Black has also achieved
8... exdS? several things that White did not want! Thus
An9la'r 10. Let us first consider the point the character of the position has changed,
of 8... ~xdS. By committing himself to the but not the balance between the two sides.
early exchange on dS, White seeks to deny Qtestion 15. Whereas 8...exdS... ?
his opponent the active benefits of 7... b6. A rzszrer 15. ...is not such a bad move, but
Black's immediate opening task is to redress in comparison to 8...tDxdS, it is a very
the balance by improving his position in inefficient move: it concedes what White
another way. wants without solving any of Black's
Q«:stian 11. Aha, so he's losing a bit on the problems.
activity front so he should try and gain... ? Thus Yusupov-Kamsky, Linares 1991,
An9la'r 11. On the 'comfort' frond With went 9 i.d3 i.b7 10 0-0 tDbd7 (10 ...tDe4 11
8... tDxdS! Black borrows a manoeuvre from i.xe7 'ilxe7 12 'ii'b3 l:td8 13 l:tac 1 is slightly
64
better for White according to Kasparov;

while 10...c5?! 11 tDe5! tDbd7 12 .i.f5! tDxe5
13 dxe5 tDe8 14 .i.g3 tDc7 15 'ifg4 was very
strong for White in Kasparov-Beliavsky,
Candidates match 1983) 11 lic1 a6 (11...c5
12 dxc5 bxc5 13 idS! is unpleasant for Black
according to Yusupov) 12 'iib3 (12 tDe5!?)
12...lIe8 (12...c5 13 .i.xf6 tDxf6 14 dxc5
.i.xc5 15 lIfd1 is slightly better for White
according to Yusupov) 13 a3 lIb8?! 14 Jt.g3
lIc8 15 .i..f5 with a very strong position for
White.
Game 31 Qtestiun 16. Granted, but why put the

Fischer-Spassky bishop on e6 if you made space for it on b7
World Championship 1972 with ... b7-b6?
Answer 16. Black's decision to develop his
1 c4 e6 2 liJf3 d5 3 d4 liJf6 4 liJc3 i.e 7 bishop on b7 was taken under different
5 i.g5 0-0 6 e3 h6 7 i.h4 b6 8 cxd5 circumstances. Now that White has initiated
liJxd5 9 i.xe 7 multiple exchanges and blocked the centre,
9 .i..g3 is a rarely seen attempt to avoid the there is no reason for Black to cany on with
main lines. After 9....i..b7 followed by ... c7-c5, the same plan regardless.
Black stands well. On the other hand, 9 Qtestiun 17. But isn't ...b7-b6 just a waste
tDxdS exdS (9 ...1t.xh4 is given as equal by then?
Karpov) 10 .i..g3!? .i..g4 11 a3 c5 12 dxc5 A~ 17. Not at all! Even with the
bxc5 13 .i..e2 Jt.f6 14 lIc1 ~d7 15 'ifd2 "ib6 bishop on e6, ...b7-b6 supports Black's
16 b4led to sharp play in Agdestein-Van der central break ...c7-c5. There are several
Sterren, Wijk aan Zee 1988. important reasons for Black's choice:
9 .. :iixe7 10 liJxd5 1. On b7, the bishop is blocked behind
10 lIc1 Jt.b7 leads to the line 8 lIc1 Jt.b7 the dS-pawn as White intended; on e6 by
9 cxdS tDxdS 10 .i..xe7 .xe7 (Game 33). contrast, the bishop has prospects along the
10 ...exd5 11 ltc1 c8-h3 diagonal - it can move to £5 or g4, for
Bringing the rook to the 'battleground' fIle example.
- the semi-open c-file. Instead 11 i.d3 c5 12 2. Black's play is all going to happen on
0-0 .i..e6 13 e4!? ~d7 14 lIe1 .f6 15 dxc5 the queenside - he wants to gain a huge
tDxc5 16 e5 .f4 17 Jt.c2 Jt.g4 18 .d4! gave space advantage there by rushing his
White a pleasant initiative in Szabolcsi-Renet, queenside pawns down the board with ...c5-
French Team Championship 1998, but c4, ..b5-b4 etc. For this purpose, the bishop
1L.Jt.e6 12 0-0 tDd7! was safer. is better-placed on e6, pointing towards the
11 ...i.e6! queenside.
see fo/lo wing diagram 3. Finally, once Black plays ...c7-c5, the
bishop is much better placed to deal with
Qitstion 15. Err, wait a minute... White's play on e6 than b7.
An.w.a'I" 15. Black's ftrst task is to play ...c7- Qtestiun 18. What is White going to do? Is
c5. As lL.c5 loses a pawn to 12 dxc5 bxc5 he going to attack on the kingside?
13 'ii'xdS Black must defend his dS-pawn. Anscrer 18. An all-out kingside attack is
65
unlikely to succeed due to the number of simply put the queen on c2?
pieces that Black has managed to exchange.
White must concentrate on the main source
of tension in the position - the semi-open c-
flie and the pawn on cS.
Q«:stion 19. So how does White do that?
Arz,gzrer 19. Typically White captures on cS
to concentrate his play against Black's
'hanging pawns' on dS and cS. Bringing his
rooks to the semi-open c- and d-files, White
then tries to harass Black's central pawns
with his knight.
Q«:stion 20. Sounds scaryi So how does
Black react?
Arz,gzrer 20. Black plays hard on the Ansza'Y 21. White wishes to concentrate
queenside! For example, he brings his king's pressure against Black's c5-pawn; going by
rook to b8 to target the b-pawn, while the the old adage that it is easiest to hit a
other rook supports ... a5-a4 gaining stationary target, White must immobilise the
queenside space. Thus we understand why cS-pawn, or at least dissuade it from
the light-squared bishop is better on e6 than advancing. By pinning the c5-pawn to the
b7: on b7, it merely obstructs Black's b-flle black queen on e7, White fulfils this goal and
countetplay, while on e6 it can even be useful thus buys some time in which to finish
in attacking the a2-pawn at some stage. mobilising the rest of his pieces.
Moreover, the bishop on e6 adds to the In general the queen is well-placed on a3 -
defence of Black's king's position by it attacks c5 without getting in the way of
covering light squares such as f5 and f7. White's rooks on the c- and d-flies, whilst it
12 Wa4 also eyes other potential vulnerabilities in
12 .i.d3 c5 13 0-0 ltJd7 14 dxcS bxc5 15 Black's queenside: the a-pawn (which will be
e4 dxe4 16 i.xe4 ':ad8 17 .i.bl is given by isolated after the exchange on cS) and the a6-
Karpov as slightly better, but it does not square.
seem particularly impressive for White. 13 ... .:c8
12 ... c5 13...ltJd7 14 .i.a6 is annoying according to
12... a5!? is Nigel Short's pet idea, playing Karpov.
the useful move ...a7-as and at the same time 14 i.b5!7
introducing the idea of ...'i'e7-b4+. After 13 The most ambitious move. White tries to
ltJe5 (13 ':c3 ':c8 14 a3 as 15 .i.b5 ltJd7 16 intensify his campaign on the c-flle.
0-0 ttJf6 17 ltJd2 .i.f5 18 ':fc1 'i'd6 19 h3 Q«:stion 22. In what way?
gS!? was fine for Black in Schlosser-Short, Ansza'Y 22. Black's ideal defender for the
Calcutta 1998) 13 ...':c8 (13 ...'i'b4+ 14 'i'xb4 cS-pawn is the knight and its most natural
axb4 15 lhc7 lha2 16ltJd3 is quite murky) post is the d7-square. 14 ..tb5 gives White
14 a3 c5 15 i.b5 'ii'g5!? 16 g3 c4 17 0-0, the possibility of exchanging off the knight
Black had some development problems in when it comes to d7, thus depriving the c5-
Velikov-Short, Slavija-Solingen 1987. pawn of a valuable defender.
13 Wa3! The quieter 14 .i.e2 is considered in the
Q«:stion 21. That's an interesting way to next main game.
put pressure on the c5-pawn! Why not 14... a6?!
66
TartakOwel VdlldLlVII. , ,,,,,,,u ~u .... - .. - .. -
In his game against Timman in Hilversum 'iie4 lLlf6 38 :xf6 gxf6 39 :xf6 ~g8 40
1973, Geller demonstrated that 14...'iWb7! ~e4 ~h8 41 'iif4 1-0
equalises comfortably for Black. After 15 A really IX>werful game!
dxc5 bxc5 16 :xc5 l:lxc5 17 'iixcS tiJa6! 18
i.xa6 (18 'ifc6 'ifxc6 19 i.xc6 :b8! Game 32
intending 20 b3 :c8!) 18...'ii'xa6 (preventing Winants-Kasparov
the white king from castling) 19 'ii' a3 'ifc4 20 Brussels 1987
~d2 'iVg4 21 :gl d4! 22 tiJxd4 'iVh4 23 :el
'ii'xf2+ 24 lIe2 'iVf1, Black had a vety 1 d4 lLlf6 2 c4 e6 3 lLlf3 d5 4 lDe3 ~e 7
dangerous attack. 5 .i.g5 h6 6 .th4 0-0 7 e3 b6 8 exd5
Qtestion 23. If the pawn is too hot, why lDxd5 9 .i.xe7 "xe7 10 lDxd5 exd5 11
can't you just play 15 O-O? :te1 ~e6! 12 'fIa4 e5 13 "a3 :te8 14
Amz:a:r 23. After 14...'iVb7! the c5-pawn is .i.e2
no longer pinned so 15 0-0 is met by 15 ...c4! A safer development of the bishop than
when the bishop on b5 is precariously 14 ~b5.
placed. 14.. .c~f8
After Spassky's inaccuracy, the game loses
its theoretical significance, but not its
instructional value. It is a classic exposition
of White's desires and Black's fears!
15 dxe5 bxe5 16 0-0 :a7 17 .te2 lLld7
18 lLld4!
Qtestion 24. This looks a bit odd

Anszar 24. Black wants to activate his
queenside majority; consequently, he must
break the pin on the c5-pawn. With 14.. .r;t>f8,
Black protects his queen on e7, freeing him
to play ...c5-c4. Another IX>pular idea is
Brilliant use of White's "a4-a3 14... a5 15 0-0 "a7.
manoeuvre. From d4, the knight can either Qtestion 25. What?
retreat to b3 to join in White's pressure An.szrer 25. As soon as Black avoids the
against the c-pawn, or capture on e6 as a pin, White will open the centre by capturing
prelude to a light-square assault. on c5 in order to concentrate against Black's
18 .....f8 19 ~xe6! fxe6 20 e4! d4 21 f4 hanging pawns. Black will develop
"e7 22 e5 l:tb8 23 .te4 ~h8 24 "h3 counterplay by using the newly-opened b-ftle
lLlf8 25 b3 a5 26 f5 exf5 27 1:xf5 lLlh7 to attack White's pawn on b2. Two benefits
28 1:ef1 'iid8 29 .g3 :e7 30 h4 :tbb7 of Black's plan thus become apparent:
31 e6! 1:be7 32 'iie5 'iie8 33 a4 'iid8 34 1. The queen is well-placed on a7, ready to
1: 1f2 "e8 35 l:t2f3 .d8 36 ~d3 'iie8 37 move to the b-file.
67
uueen's (iambit Declined
2. The black a-pawn is ready to advance to exd5.

a4, making White's b-pawn backward
Play might continue 16 dxc5 (16 b3 tDd7
17 i..bS fIc7 18 fIfd1 ':ac8 19 h3 [19 i.xd7
fIxd7 20 ttJeS ':dc7 is fine for Black
according to Kharitonov] was agreed drawn
in Karpov-Kavalek, Linares 1981) 16... bxc5
17 ':c3 ~d7 18 ':fc1 ':cb8 19 ~e1 a4!?
(perhaps 19...c4!? or 19...':b7!?) 20 ~d3 c4
21 ttJf4 'iVcS! was very complicated in
CHansen-Schandorff, Danish Champion-
ship 1986.
In his annotations, Kasparov criticised
14.. .cltf8, but I have to say that I quite like
this move. The game is again very instructive. QIestion 26. So what's the difference?
15 dxc5 bxc5 16 0-0 as! 17 1:c3! lDd7 Ansrrer 26. As we saw previously, Black's
18 1:fc1 1:cb8! desired post for his light-squared bishop in
Avoiding White's idea of ttJf3-d4. the hanging pawns structure is e6; on b7 the
191:b3! bishop would both block Black's b-flle
19 b3 a4! 20 bxa4 c4! equalises according counterplay, and itself be blocked along the
to Kasparov. The text is a typical idea, as-h1 diagonal by the pawn on dS. By
dampening Black's counterplay by delaying his capture on dS, White hopes to
exchanging a pair of rooks. reach the same pawn structure with Black's
19 ... e4?! bishop already committed to the inferior b7-
19...a4! 20 ':xb8+ lhb8 21 iLd1 c4! 22 square.
i..xa4 'iVxa3 23 bxa3 ttJc5 is the way to QIestion 27. How bad is that for Black?
equalise according to Kasparov. Anszrer 27. While it is nothing heart-
20 :'xb8+ :'xb8 21 -.xaS :'xb2 22 lDd4 stopping, small advantages are built on the
~g8 23 1:a1 tLle5 24 'ii'a8+ ~h7 25 'ii'a3 accumulation of such details so Black should
1:b6 26 i.d1 g6 27 i.e2 i.d7 28 h3 'ii'd6 not readily accept this type of concession.
29 'ii'a5 il.a4 30 i.xa4 :'a6 31 'ii'b5 1:xa4
32 a3 c3 33 tLle2 -.e7 34 ~b1 :'a7 35 Game 33
tie8 1:b7 36 1:b4 ~g7 37 g3 lDe6 38 Korchnoi-Short
'ii'a4 tLlg5 39 h4 lDe4 40 ~g2 ~a7 41 Wij.1e aan Zee 1990
'ii'b5 'ii'e5! 42 'i'b6 'ii'f5 43 f3 tLlg5!!
This leads to a stunning finish! 1 c4 e6 2 tLle3 d5 3 d4 lDf6 4 i.gS i.e7
see following diagram 5 e3 0-0 6 tLlf3 h6 7 i.h4 b6 8 ':e 1 i.b 7
9 exd5 tLlxd5 10 il.xe7
44 hxg5 'ii'xe2+ 45 ~g1 'iid1+ 46 ~g2 10 ttJxdS i..xdS 11 i..xe7 'iVxe7
'iie2+ 47 ~h3 'ii'xf3!! 48 'ii'xa7 'ii'h1+ 49 transposes.
~g4 h5+ 50 ~f4 'ii'f1+ 51 ~e5 -.f5+ 52 10 ...'ii'xe7 11 tLlxd5 i.xd5!
~d6 'ii'e6+- 53 ~c7 'ii'e7+ 54 ~b6 'ii'xa7+ Black uses the move order to his
55 ~xa7 e2! 0-1 advantage as well! This recapture ensures the
Since this line promises very little, White activity of the light-squared bishop by
players have tried a subtle move order to keeping the a8-h 1 diagonal open. Moreover,
improve it: 8 ':c 1 i.b 7 and only then 9 from dS the bishop eyes White's unprotected
68
Tartako wer Variation: Fixed Centre Plans
a2-pawn! Black's task is now to liquidate his Black is slightly better according to
backward c-pawn with ...c7-cS. Ftacnik.
20 iLb3 l:eS 21 'fia6 :e7 22 g3 :d8 23
:e1 l:e1 24 'fixa7 :xe1+ 2S liJxe1 'fie2
26ltJe2 ::td1+ V2-V2
As White gains little from the early release
of central tension, modem practice has
concentrated on the nonnal developing
moves 8 it.e2 and 8 ~d3. Black now faces a
major choice - whether to take on c4, or
whether to maintain his centre with 8...it.b7
for example.
Q«5tion 28. I can't see what could be
wrong with 8...it.b7!
Anszrer 28. The issue is whether Black
12 i.e2 wishes to play the structures arising after, for
12 .td3 %lc8 13 0-0 cS 14 dxcS lIxcs 15 example, 8 ..te2 i.b7 9 ..txf6 i.xf6 10 cxdS
l:txcS "xeS 16 "a4 tiJc6! 17 e4 tiJb4! 18 exdS, or whether he is willing to forego a
exdS tiJxd3 19 dxe6 fxe6 20 b3 lId8 21 'iiVe4 little flexibility with 8...dxc4 to avoid this
'iVfs led to a draw in Uhhnann-Spassky, possibility altogether (as seen in the next
Solingen 1974, as did Vaganian-Short, Elista chapter).
Olympiad 1998, after 12 .tc4 .tb7 13 0-0 Q«5tion 29. Which is the best?
%lcS! 14 tiJeS tiJd7 15 tiJxd7 "xd7 16 it.e2 An.m.er 29. Black's most popular move is
%lab8 17 b3 "e7 18 .tf3. 8... ..tb7. When it has the faith of players such
12 ... eS! as Kramnik, Kasparov and Spassky, it is
This excellent idea of Short's has probably a good choice!
superseded the older 12...%lc8 13 0-0 c5 14 We shall fIrSt examine the main line 8 i.e2
dxcs %lxcs 15 :XcS "xcS 16 'iVa4 it.c6 i.b7 9 i.xf6!? i.xf6 10 exdS exdS 11
(unlike after 12 it.d3, here 16...tiJc6 17 e4! b4.
does win a piece) 17 "f4 with a small edge
for White.
13 dxeS :d8!
Short also tried the more committal
13 ...bxcs in a rapid game in Garmisch 1994,
against Brunner when 14 "a4 tiJd7 15 0-0
l:tfb8 16 l:tc2 as 17 %lfc1 lIb4 kept the
balance for Black.
14"a4
14 "c2 %lc8 15 "d2lhcS 16 :!xcS "xcS
17 0-0 tiJc6 18 %lel was agreed drawn in
Dautov-Yusupov, Bad Homburg 1998, while
14 0-0 it.xf3 15 .txf3 %lxd1 16 %lfxd1 tiJc6!
is also fine for Black according to Ftacnik. Q«5tion 30. Wait, hang on a minute! I
The text is a little too ambitious. don't understand a thing! What is White
14... ~d7! 1S e4 ltJxe5 16 :xeS 'iVxeS 17 doing?
exdS 'iVe1+ 18 iLd1 l:xd5 19 0-0 'iVxb2 Arl3Zre'" 30. The moves 9 i..xf6 and 10
69
cxdS fonn a disruptive manoeuvre against Black's knight on f6 weakens Black's defence
Black's development. There are three of his centre (this also means incidentally that
intended consequences: Black can no longer use the Lasker
1. The passivity of Black's light-squared manoeuvre ... ~f6-e4 to free his position by
bishop. exchanging two sets of minor pieces).
2. The weakening of Black's kingside light Moreover, since the dark-squared bishop has
squares. been dragged on to f6, some reorganisation
3. The deterrence of Black's freeing break will be needed before Black's knight on b8
...c7-cS. can replace its fallen comrade.
White has clearly realised the first Qtestion 34. Isn't the bishop just good on
objective. With 9 .ltxf6, White ensured that f6, raking along the long diagonal?
his opponent could no longer meet c4xdS Ansuu 34. I know that it sort of looks like
with ... ~f6xdS keeping the as-hI diagonal a KID bishop, but...! Currently it is just biting
open, but rather had to block the range of against White's pawn chain. In this structure,
the light-squared bishop on b7 by the bishop should be on d6, freeing f6 for
recapturing on dS with the e-pawn. the queen's knight and supporting ...c7-cS,
Qtestion 31. Granted, but why does White while at the same time eyeing the h2-square
play .ltgS-h4xf6? Couldn't he have saved and giving Black some future hope for
time by playing .ltgSxf6 as soon as Black kingside action.
played ...h7-h6? Qtestion 35. And how does White's
A1t.9'lrer 31. White's judgement is that this manoeuvre affect the cS-square?
plan became dangerous only after Black had Anszrer 35. After ....lte7xf6, the bishop no
committed himself to ... b7-b6 on the longer supports ...c7-cS. In essence, 9 .ltxf6
queenside. In other words, White contends disrupts the hannony of Black's development
that the extra move ...b7-b6 is more helpful which gives White the opportunity for 11 b4.
to White than to his opponent. lbis is a Qiestion 36. So what is the point of 11 b4?
typical example of a positional trade-off in An3tm" 36. Exploiting the fact that
the opening. White loses a little time and .lth4xf6 deflected Black's dark-squared
gives up the bishop pair, but in retwn he bishop from the f8-a3 diagonal, White brings
makes concrete gains in his fight against pawn pressure to bear upon the black
Black's plans and pieces. structure. (Of course, this idea .lth4xf6
Black's kingside light squares are followed by b2-b4 is very reminiscent of the
weakened in two ways: minority attack in the Exchange QGD.) 11
1. The e-pawn has been diverted from e6 b4 has two aims:
to dS, so that the fS-square is now available 1. White brings more pressure to bear on
to the white pieces. cS and hopes to deter Black from achieving
2. The knight on f6, which defended h7, his freeing break ...c7-c5.
has been exchanged. 2. White may follow up with b4-bS
Qtestion 32. How does White's plan help clamping down on c6. If Black were then to
against Black's freeing break? play ...c7-c5, then b5xc6 would leave Black
Ansuu 32. In order to achieve ...c7-cS with a weak isolated d-pawn.
comfortably, both the dS- and cS-squares Qtestion 37. Is that so serious?
require a certain level of support. An3tm" 37. It won't lead to an immediate
Qtestion 33. So how has this manoeuvre loss, but it is a concession you'd rather avoid
affected the dS-pawn? The most economical method of protecting a
Ansuu 33. Obviously the removal of pawn is by another pawn. If a piece is used,
70
Tartako wer Varia tion: Fixed Cen tre Plans
this defensive duty will inevitably reduce the stand. 11...c5 makes use of the bishop on f6
activity of that piece and will lead to a slight which combines with ... c7-c5 against the
reduction in the activity of your whole pinned d4-pawn.
position. This illustrates the structural Q.lestion 40. So isn't this just the logical
drawback to ...b7-b6. Were the pawn on b7, continuation?
then b4-b5 would not carty the same force. Anme- 40. It certainly is but after 11 b4, it
Qtestion 38. I have two points. Firstly, is it is not without positional risk. After 12 bxc5
terrible for Black if he doesn't achieve ...c7- bxc5
c5?
Amzrer 38. Of course not - as we shall see,
the patient 11...c6 is Kramnik's favourite
move in this position. However, it can then
be said that White has achieved something
with his manoeuvre .th4xf6. In return for
the bishop pair, he's kept the light-squared
bishop quiet and stopped Black from playing
his freeing break. 'The game goes on' as
Julian Hodgson always says, but at least
White can feel that he has achievements to
build on.
Qtestion 39. OK, now my other question.
Can't Black just play 11 ... c5 immediately? Black's pawn structure has been 'diluted'.
Now once White achieves d4xcS, Black will
have to accept an isolated d-pawn rather than
the hanging pawns we have seen until now.
Secondly, with b4xcS, White opens the b-fue.
This allows him to harass the restricted
bishop on b7 (how Black would prefer it to
be on e6!) with 13 :b1.
An..9Zt8" 39. He certainly can! This is

Kasparov's favourite move. As we saw
earlier, 9 i..xf6 disrupts the harmony of
Black's pieces so that they are unsuited to the
current pawn structure; for example, the
bishop on f6 would be better on d6 etc.
Black has two choices - to manoeuvre his
pieces to fit the structure or to change the Qtestion 41. Oh dear! Is this just good for
pawn structure altogether. 11...c5 espouses White then?
the latter approach: Black goes for his freeing Anmx!Y 41. Not so fast! Now look at the
break and makes use of his pieces where they position from Black's side! White hopes lie in
71
his fmner structure and the slight disruption according to Geller.

he can cause in Black's queen side Qtestion 43. This doesn't make any sense
development. However, Black has the two for Black!
bishops, a spacious position with room for Ansrw- 43. In fact, this is Black's main idea
all his pieces, and just the later possibility of in these positions!
an IQP. With only one weakness and an Qtestion 44. What?
othetwise pleasant position, Black should not Ansrw- 44. At the cost of the bishop pair,
fall into serious trouble. Black creates an IQP in White's position,
Qtestion 42. All the same, I get the feeling equalising the pawn structure. This greatly
that you don't really like this as much as relieves the pressure on Black's d-pawn by
11...c6! shutting the semi-open d-fue. White's only
Anstrer 42. It's true! In this line, Black is remaining (and very small) positional edge is
playing single-mindedly for a draw. It seems his slightly better bishop - Black's bishop is
a strange thing to do when the positions after on the same colour as his IQP.
11...c6 are so rich and interesting. However, Qtestion 45. So why is 13 ...'if'as dubious if
if Kasparov gives it his seal of approval in it forces the plan that Black wants?
World Championship matches, then it is Anszur 45. The problem is the time that
obviously a pretty good move! Black has taken to force this structure. As we
shall see, White will gain time on Black's
Game 34 queen as well as Black's light-squared bishop.
Azmaiparashvili-Short This gives White's knight the chance to reach
Manila Olympiad 1992 an aggressive outpost it could nonnally never
achieve.
1 d4 d5 2 c4 eS 3 4:Jc3 4:JfS 4 4:Jf3 .te7 Qtestion 46. Why is this so important?
5 .i.g5 0-0 S e3 hS 7 .i.h4 bS 8 .te2 An.9rrer 46. The absence of both b-pawns
.ib7 9 .i.xfS .i.xfS 10 cxd5 exd5 11 b4 from the queenside structure leaves both
c5 12 bxc5 bxc5 13 :b1 'iia5?! knights unsettled on their natural squares c3
The more accurate 13 ....tc6 is the subject and c6; as they lack the usual support of
of Games 35-37. pawns on b2 or b7, they are vulnerable to
14 'iid2 cxd4 15lLlxd4 .txd4!? 16 exd4! pressure along the open c-flle. The time that
White gains on his opponent's pieces allows
him to solve his problem before Black. This
factor is not a decisive one, but it makes
Black's task an unenviable one, particularly
against the kind of technique that 'Azroai'
shows!
1S ... .ta6
16... .t.c6 was nicely met by 17 lLJd1! in
Chernin-Beliavsky, Debrecen 1992. After
17...'if'xd2+ (17...'if'd8 18 0-0 with lLJd1-e3,
.te2-f3 and llfl-c1 to follow is better for
White according to Chernin) 18 ~xd2 lLJd7
(unlike his opponent, Black cannot develop
Best as 16 'if'xd4lLJc6 17 'if'd2 d4! 18 exd4 any pressure against the IQP as his bishop
iLa6! preventing 19 0-0 due to 19 ....txe2 20 prevents the knight from coming to c6) 19
fixe2 fixc3 is slightly better for Black l:lc1! l:lac8 20 .ta6 l:lc7 21 ~e3 ttJb8 22 .i.e2
72
Tartako wer Varia tion: Fixed Centre Plans
%:tb7 23 %:tb 1 lId8 24 it..f3 ltJa6 25 %:txb7 balance. After the text, Azmai takes control!
~xb7, 26 lIb1! it..c6 27 %:tel it..b7 28 lIc3 21 ':b3! ':ac8 22 h3! "g5 23 "xg5
followed by :a3-aS would have given White hxg5 24 ':g3! f6 25 ~g4 :b8 26 ':c3
a clear advantage according to Chernin. :b6?! 27 ':c5 ~xb5 28 axb5 ttJa5 29
17 ttJb5 "d8 ':e1! ~f8 30 ~h5 ':bb8 31 ':c7 ':b7 32
17... ttJc6 18 'ilixaS ttJxaS 19 ttJc7 it..xe2 20 ':ee7 ':xc7 33 litxc7 lDb3 34 litf7+ ~g8
ttJxa8 ~d3 (20...lIe8 21 ttJc7!) 21lId1lIeS+ 35 litxa7 ':b8 36 oli.g6 ttJxd4 37 b6 ttJc6
22 <it>d2 wins for White. Consequently, the 38 ':c7 ttJd8 39 ':d7 ~f8 40 litxd5 ~e7
black queen must retreat. 41 ':b5 ttJb7 42 ~h5 ttJd6 43 litb4 ~d7
180-0 ttJc6 44 .tf3 g6 45 ~h2 ttJc8 46 b7 ttJd6 47
1S... ttJd7 19 lIfel ttJf6 20 f3!? l:.e8 21 a4 h4 gxh4 48 ':xh4 lDxb7 49 ':h7+ ~c8
l:.e7 22 .i.d3 was vety pleasant for White in 50 ':f7lDd6 51 litxf6 ~d7 52 .tg4t ~c7
Vaganian-Geller, New York 1990. 53 ':xg6 :b2 54 f3 1-0
19 a4 Black players have lost faith in the forcing
A slightly unusual move order - 19 :lfd1 13 ...'i'aS, turning instead to the calmer
'i'f6 20 a4 is more common 13 ... ~c6 which auns for simple
19.....f6 development.
Azmai suggests 19...'i'g5!?
20 litfd1 Game 35
Topalov-Kasparov
Sofia (rapidplay match) 1998
1 ttJf3 d5 2 d4 lDf6 3 c4 e6 4 ttJc3 .te 7
5 .tg5 0-0 6 e3 h6 7 .th4 b6 8 .te2
.tb7 9 .txf6 .txf6 10 cxd5 exd5 11 b4
c5 12 bxc5 bxc5 13 ':b 1 .tc6 14 0-0
lDd7 15 .tb5
20 ...litfd8?!
20... l:.ab8 21 .tfl .tc8! as in Lobron-
Kir.Georgiev, Tilburg 1992, seems Black's
best tty to reactivate the bishop along the h7-
b 1 diagonal and to drive the knight from b5
with ...a7-a6. After 22 l:.bc1 a6 23 ttJc7 'i'd6
24 'i'd 'i'xc7 25 'i'xc6 'i'xc6 26 lIxc6 l:.b4
27 as %:ta4 28 l:.dc1 (28 .i.xa6 .i.e6 29 ~b7
lIxaS 30 ':cS ':a4 31 ~c6 ':b4 32 ~xd5 Qiestinn 47. This looks wrong: why is
':d8 33 it..xe6 fxe6 was equal in White swapping off his 'good' bishop for
Kir.Georgiev-Kotronias, Burgas 1992) Black's 'bad' one?
28 ... ~e6 29 ':xa6 ':bS 30 ':d1 g5 31 f3 Anm:er 47. This tenninology is sometimes
l:.bb4 32 ':a7 ':xd4 33 :.xd4 lIxd4 34 a6 misleading. Black's bishop is technically 'bad'
<it>g7 35 ':c7 lIa4 Black just about held the as it is on the same colour as Black's central
'73
d.5-pawn. However, the task it performs Championship 1988) 18 'ii'f5 ttJb6 (18 ... g6!?)
within Black's position - as the only defender 19 a4 a6 20 .txc6 'iixc6 21 as ttJa4 22 ttJe2!
of Black's d.5-pawn - is an absolutely pivotal when Black's offside knight on a4 gave cause
one. White's bishop, though technically for concern.
'good', performs no useful function on e2. 17 h3
Consequently, White should consider 15 1bis quiet move gives Black an
~b5 as the trade of an underemployed piece opportunity to implement a typical equalising
for a key defensive unit. manoeuvre. The more testing 17 ':'fct and
Qlestion 48. But if Black exchanges his 17 ':'fdl are considered in the next two main
'bad' bishop, then he has lost one positional games.
worry! 17 ... cxd4! 18 ttJxd4 iLb7! 19 .:tfc 1 ttJc5
Amz.rer 48. Absolutely, but White can 20 lid1 lie7
claim that he has also made an existing one
worse - the cIS-pawn is much more
vulnerable!
Instead 15 "i'd2 c4t 16 ltJel "i'aS 17 ttJc2
':'ab8 18 ~f3 tiJc5! 19 tiJe4 "i'xd2 20 tiJxd2
c3 21 lhb8 lhb8 22 ':'b 1 ':'c8 23 tiJb3 ttJe4
was pleasant for Black in Dokhoian-Pigusov,
USSR 1985.
15 .. Ji'c7 16 lid3!
An excellent square for the queen. It aims
for f5 (attacking cIS) as well as b5 (after a
prelimimuy ~b5xc6) while keeping the c-file
free for a white rook. White can also line up
against cIS with lUl-dl. The poSItIon is about level - Black's.
16 "i'a4 (intensifying the pressure on the activity and two bishops compensate for the
bishop) 16...ltJb6 17.aS cxd4 18 exd4 ':'fc8 IQP.
19 ~xc6 (19 .ta6 tiJc4! was flne for Black in 21 lig4 g6 22 iLn iLg7 23 lid1 l:ab8
the game Eingorn-Lputian, Dortmund 1988) 24 .:te2 iLa8 25 .:tbc 1 .:td8 26 g3 iLxd4
19...'iixc6 20 ':'b3 "i'c4 was fairly level in This typical idea again!
Salov-Hjartarson, Belgrade 1987. 27 exd4 ttJe6!
16 .. J~fe8
The most active move, preparing to place
the rooks on the a- and b-fues. Vaganian-
Kir.Georgiev, President's Cup, Elista 1998,
saw 16...ltfd8 17 ':'fct (17 ':'fdl ':'ab8 18
..txc6 .xc6 19 ':'xb8 ':'xb8 20 dxc5 ~xc3
21 "xc3 "i'xc5 22 "i'xc5 tiJxc5 23 h3 ttJe4
24 :xclS ':'b 1+ 25 ~h2 tiJxf2 26 ':'d8+ <&i?h7
27 ':'d7 as 28 ltxf7 ':'b2 29 a4 tiJdl with
sufflcient counterplay for Black in Karpov-
Kasparov, World Championship 1985)
17...c4!? (17...ltac8 18 h3 g6 19 ~xc6 "i'xc6
20 'ii'b5 cxd4 21 .xc6 l:Ixc6 was flne for
Black in Groszpeter-Vaganian, World Blitz In contrast to the 13...'iWaS line, Black's
74
knight fmds an ideal outpost on e6 from a) 1S... cxd4 19 tDxdS 1i.xb5 20 l:txc7
which to attack the d4-pawn. .i..xd3 21 l:lxbS l:lxbS 22 l:lxd7 dxe3 23
2S h4 'fIf6 29 tiJe2 %:tb4 30 %:teS iLb7 31 ~xf6+ gxf6 24 fxe3 1i.e4 25 :Xal is given as
:xdS+ tiJxdS 32 'fid2 %:ta4 33 'fixh6 %:txa2 slightly better for White by Kasparov.
34 'fie3 tiJe6 35 %:tb1 iLe6 36 'fIh6 'fidS b) 1S... g6 19 1i.xc6 l:lxbl 20 'iixb1! 'i'xc6
37 'fie3 'fif6 3S 'fib3 %:td2 39 'fie3 tiJxd4 21 dxcS "xcS 22 ttJe2 'i'fs was Kasparov-
40 lLlf4 'fie5 41 'fia3 ~g7 42 %:te1 iLb5 Karpov, World Championship 19S7, and
43 ':eS iLeS 44 'fie3 iLd7 45 ':dS ':d1 now according to Kasparov 23 tLlf4 ttJb6 24
46 ':xd7 'fie4 47 ':xd5 ':xf1+ 0-1 h4! h5 25 :XcS 'i'xcs 26 ttJxg6! fxg6 27
iixg6+ .i..g7 2S ttJgS 'i'd7 29 'iixhs is clearly
Game 36 better for White.
Khalifman-Chandler The text, gaining queenside space and
Gemzan Bundesliga 1995 creating a protected passed c-pawn, is
another Kasparov suggestion.
1 d4 tiJf6 2 e4 e6 3 tiJf3 d5 4 tiJe3 iLe 7 Q«stion 49. What do you think of this
5 iLg5 0-0 6 e3 h6 7 iLh4 b6 8 iLe2 move?
iLb7 9 iLxf6 iLxf6 10 exd5 exd5 11 b4 Anszrer 49. Instinctively it seems rather
e5 12 bxe5 bxe5 13 ':b 1 iLe6 14 0-0 repulsive to me! It goes against an opening
tiJd7 15 iLb5 'ile7 16 'ild3 ':fc8 17 ':fc1 principle that I learnt from books as a child -
White's most natural move, placing the that in such positions, releasing the central
rook opposite the queen on c7. tension with ... cS-c4 is always bad.
17 ....:ab8!? Q«stion 50. Why is that?
Kasparov suggests 17... .txbS IS ttJxbS Anszrer 50. The tension between the pawn
'iic6, when 19 dxcS tt:JxcS 20 'iifs (20 'i'a3 on c5 and White's pawn on d4 represents
.i..e7 seems fme for Black) transposes into Black's main source of influence over
Kasparov-Karpov, World Championship White's position. Inevitably therefore, its
19S4 (White had played 16 'iic2) when release offers the white pieces some extra
20...'i'e6 21 l2Jfd4 iixfs 22 tDxfS ttJe6 23 chances for activity: for example, utilising this
':xcS lhcS 24 tDxal ':c2 25 ttJb5 l:lxa2 26 pressure, Black dictates that his opponent's
h3 l:la5 was agreed drawn. knight should remain on f3 so that White can
18 h3 recapture with a piece after ...cSxd4.
Q.iestion 51. So now that Black has played
...cS-c4...
Anszrer 51. White can move this knight as
he pleases, perhaps to initiate some kingside
play. In the same way, White's e-pawn is tied
to the defence of the d4-pawn by the pawn
on cS. After ... cS-c4, White gains the option
of central playwith the e3-e4 break. It is clear
that ...cS-c4 has major positional
repercussIons.
Q«stion 52. So is it just bad then?
Anszrer 52. Let's consider it in this specific
position. White cannot use the e5-outpost
1S ... c4 for his knight due to Black's bishop on f6
Alternatively: and knight on d7. Moreover, there seems
75
UUt:t:fI ::; lJamDIr uecllnea
little immediate prospect of kingside action think!

by White - Black's kingside is solid, and the 19 'iic2
exchange of the dark-squared bishop robs 19 ~f5 g6 20 iVf4 'i/ixf4 21 exf4 iLxbS 22
White of kingside firepower. lDxbS lDf8 is nice for Black according to
~ 53. How about the central break? Khalifman.
Amuer 53. 'This is more dangerous: e3-e4 19... i.xb5
mines the dS-pawn and undermines Black's 19....i.a8 is given an exclamation mark by
c4-pawn. But on the other hand, it will not Kasparov. Khalifman recommends 20 iVf5
be a decisive blow by itself. ttJb6 21 e4 'with an initiative', but after
It is also very important to consider the 21...dxe4 22 ttJxe4 .i.xe4 23 iVxe4 iVd6 I
queenside situation. In normal positions, don't see a great deal for White.
another major argument against ...cS-c4 is 20 ttJxb5 'i'a5 21 ttJc3
that Black cannot find sufficient counterplay
on the queenside to compensate for White's
added freedom on the kingside and in the
centre. The exchange of the b-pawns
invalidates this argument here.
~ 54. You mean that Black can use
the b-fue for counterplay.
Amuer 54. Not only that - White's
queenside dark-squares are also greatly
weakened by the absence of a pawn on b2.
For example, though White's knight on c3 is
well-placed, blockading Black's protected
passed pawn and attacking the dS-pawn, it is
vulnerable to attack by ...•c7-aS and ... .i.f6- 23 ....:xb1
e7-b4. This leads to a nice endgame edge for
~ 55. I suppose that ... cS-c4 also White. Khalifman recommends 21 ...g6!?
makes the dS-pawn safer. when 22 lhb8 lhb8 23 e4 dxe4 24 ttJxe4
Ansrar 55. In a way, yes. Black removes l:tc8! 25 ttJd6l:Ic6 26 ttJxc4 ~d5 offers good
the possibility of d4xc5, opening up the d-file compensation for the pawn.
against the dS-pawn. However, the corollary 22 lbb 1 :tb8 23 ':xb8+ ttJxb8 24 e4!
to this is that if White does win the dS-pawn, dxe4 25 ttJxe4 'ii'd5 26 ttJxf6+ gxf6 27
then the c4-pawn will inevitably drop off and ttJd2! 'iixd4 28 ttJxc4 ttJc6 29 ttJe3 'ii'a1+
Black will be two pawns down and lost. 30 ~h2 'iie5+ 31 g3 ttJd4 32 'iic8+ ~g7
~ 56. But Black can lose the IQP as 33 'iid7 ttJf3+ 34 <it'g2 ttJg5 35 'i'g4 h5
well! 36 'ii'f5 ttJe6 37 h4 ttJd4 38 'ii'd3 ~g8 39
Ansrar 56. Yes, but this is one of the ~f1 ~f8 40 ~g2 ~g8 41 ~f1 f5 42 'ii'c4
beauties of IQP positions: you always have f4 43 'ii'd5 'ii'xd5 44 ttJxd5 fxg3 45 fxg3
chances as the rest of your position is sound, <it'g7 46 ~f2 ~g6 47 <it'e3 ttJf5+ 48 ~f4
and there are open fues on which your pieces f6 49 ~f3 ~g7 50 ttJf4 ~h6 51 ttJe2
can seek activity. After ... c5-c4, Black does ttJd6 52 <it'e3 ~g6 53 ttJf4+ ~h6 54 <it'd3!
not have this 'safety valve'. liJf5 55 ttJe2 <it'g6 56 <it'e4 ttJe7 57 liJf4+
~ 57. So after all that, is 18 ... c4 good <it'h6 58 a3 ttJc8 59 ~d5 ttJb6+ 60 ~c6
or not? ttJc4 61 a4 ttJd2 62 a5 liJe4 63 ~b7
A muer 57. It's risky, but it's playable I ttJxg3 64 <it'xa 7 ttJf5 65 a6 1-0
.... ,.
I i:I' L d 1\ U VI' .... I """ , , .... " • - • •• •... - .
Amzrer 58. 11 ...c6 is a 'halfway' holding

Game 37 move. While Black doesn't wish to loosen his
Timman-Kasparov position with the immediate 11...c5, he has to
Prague (match) 1998 act against the threat of b4-b5. clamping
down on c6 and isolating the dS-pawn from
1 d4 ttJf6 2 c4 e6 3 ttJf3 dS 4 ttJe3 Ji.e7 the support of the c-pawn. After 1L.c6,
S Ji.gS 0-0 6 e3 h6 7 ~h4 b6 8 Ji.e2 Black is ready to meet 12 bS with 12...cS.
Ji.b7 9 ~xf6 i.xf6 10 exdS exdS 11 0-0
ttJd7 12 b4 eS 13 bxeS bxeS 14 :b1
.te6 15 Ji.b5 'fie7 16 'fid3 :fe8 17 :fd1
Lining up on the dS-pawn.
17 ...:ab8 18 a4!?
Qtestinn 59. What are the drawbacks to

1L.c6?
A mzrer 59. Black provides a target for
White with his backward c-pawn and also
continues his cruelty to his bishop on b7,
18 ... exd4! 19 ttJxd4 ttJeS 20 ~fS Ji.xd4! which is now blocked along the aB-h 1
21 exd4 diagonal by not one but two black pawns! As
21 :xd4 ttJe6 22 :dd1 d423 exd4.txg2 1L.c6 cuts off the support of the bishop for
leads to great complications. the dS-pawn, White gains the chance to
21 ... g6! 22 'fif3 break in the centre with e3-e4.
22 1Ih3 ttJxa4! 23 .txa4 lhb1 24 ':xbl Qtestion 60. Sounds nasty!
.txa4 25 ttJxa4 11c1 + wins for Black. Anszrer 60. It can be very dangerous for
22 ... ttJe4 23 ttJxe4 dxe4 24 ~e3 ~d7 2S Black, but it is not without its risks for White
d5 i.xbS 26 axb5 :xbS 27 :xbS ~xbS as this central break inevitably opens lines for
28 h4 'ii'a4 29 :e1 :d8 30 hS gS 31 Black's two bishops.
'fieS ~a2 32 :d1 'ii'e2 33 ~d4 :b8 34 Qtestinn 61. So what is the upside?
'fid2 'fixd2 3S :xd2 :a8 36 :d4 fS 37 Anszrer 61. In contrast to the 1L.cS line,
g4 r3;f7 38 gxfS ~f6 39 :xe4 r3;xfS 40 which loosens Black's position, 1L.c6
:e7 as ~-~ concedes nothing to White. White still has to
work hard to create a real target in Black's
We now tum our attention to Kramnik's position.
preferred choice: 11 ... e6. It is a riskier move for Black, however.
see following diagram After 1L.cS Black says 'Well, I might have a
weakness or two, but I've played my freeing
QIestion 58. This looks like a strange break. I have space for all my pieces so
move. nothing too tenible can happen to me.' With
a move like 11...c6 where Black delays his 13 ... c5

freeing break, and restricts one of his pieces QIestion 63. What has 13 bS achieved?
just to hold back White's plan and avoid Arnzar 63. White reaps the following
weaknesses, the risk of being sat on after a benefits:
few inaccurate moves is much greater. On 1. He softens up Black's central and
the other hand, Black's winning chances are queenside light squares, gaining an outpost
immeasurably enhanced! for a knight on c6 while depriving Black's d-
pawn of the support of the c-pawn.
Game 38 2. He fixes the b6-pawn as a future target
Topalov-Kramnik for a knight on a4.
Linares 1998 However, by taking the b-pawn from b2
to bS, White weakens a host of queenside
1 d4 lDf6 2 c4 e6 3 lDf3 d5 4 lDc3 .te 7 dark squares which is particularly important
5 .tg5 h6 6 .th4 0-0 7 e3 b6 8 .te2 as Black holds the bishop pair. Moreover, as
.tb7 9 .txf6 i.xf6 10 cxd5 exd5 11 b4 the b6-pawn is so easy to defend, by dosing
c6 120-0 a5 the position, White leaves himself with only
one real one target: the dS-pawn. Whichever
way he tries to attack it, he cannot win it by
force; consequently, Black has a good
position since the rest of his position is fine -
he has no other structural weaknesses and
has the two bishops in hand for later.
14:e1
In this game, Topalov tries for flexible
manoeuvring, whereas in Gretarsson-
Yusupov, World Championship 1997, White
tried to isolate his opponent's d-pawn with
14 :c 1 (protecting the knight on c3 and
intending 15 dxc5 bxc5 16 ttJa4!) 14...ttJd7 15
Qation 62. I wasn't expecting that one! dxcS (forcing the isolation of Black's d-pawn)
Amzrer 62. The b4-pawn is a key part of 1S... tDxcS 16 ttJd4 :c8!
White's queenside set-up: by deterring ...c7- QIestion 64. Why is 16...:c8 important?
cS, it has acted as a brake on Black's Arnzar 64. The white pawn on bS is not
queenside ambitions while providing White an asset in IQP positions, due to its effect on
with the future break b4-bs against Black's the knight on c3. Without a pawn on b2 to
queenside structure. 12... a5 challenges the b- support it, the knight is unstable on the open
pawn before White is active enough to make c-ftle; moreover, it cannot move to its natural
use of its potential. bS-square since the pawn occupies this
The slower 12...•d6 is the subject of square. Obviously therefore, the queen's
Game 44. rook belongs on the open c-file. It must be
13 b5 played there on this move, otherwise after
The direct option: White immediately 16..:iid6, then 17 .tg4! is as in NIkolic-
loosens his opponent's centre by striking Beliavsky, Groningen 1993, is rather
immediately at the c6-pawn. The alternatives annoying. After 16...:c8 17 .tg4 l:c7 18
are 13 bxa5 (Games 39 and 40) and 13 a3 ttJa4 ttJe4 19 'ii'd3 l:c4 Black stood vel}' well.
(Games 41-43). 14....:e8
78
Kramnik is obviously not very impressed

by White's plan, recommending simply Game 39
14....~Jd7 as equal. Alterman-Pigusov
15lIc1liJd7 16 g3 Beijing 1997
1 c4 e6 2 liJc3 d5 3 d4 liJf6 4 ~g5 ~e7
5 e3 h6 6 ~h4 0-0 7 liJf3 b6 8 ~e2 ~b7
9 ~xf6 ~xf6 10 cxd5 exd5 11 b4 c6 12
0-0 a5 13 bxa5 %1xa5
16 ... liJf8
After this, Black is forced to release the
central tension, but even this seems fine for
him. 16.. ,lk8 17 i.f1 cxd4 18 tDxd4 tDc5 19
i.g2 tDe4 is another Vladimir Kramnik
suggestton. QIestian 65. 13 bxa5 seems a bit odd
17 liJa4 c4 18 ~f1 'i'd6?! somehow!
The queen is misplaced here according to An-szur 65. It is true that with b4xa5,
Kramnik - 18 ..JW c7 19 i.g2 l:tad8 20 tDc3 White removes the brake from Black's ...c6-
g6 is still equal. The battle now becomes very c5 break and also activates the black rook on
murky. a8 along the a-fue. However, the move also
19 i.g2 :ad8 20 h4 liJe6 21 liJc3 g6 22 has several benefits:
liJd2 1. White gains another semi-open fue to
Gaining a tempo with the threat of pressurise the black queenside - thus White
tDd2xc4 - this is why 18..:iVc7 was more can now target the b6-pawn with iVd1-b3
preCIse. and ~a1-b1.
22 ... ~a8 23 h5 2. Once Black achieves the ...c6-cS break,
23 f4!? i.g7 24 tDf3, intending tDe5, is White's knight on c3 will have a safe and
suggested as slightly better for White by impregnable square on b5.
Kramnik. It is clear that this move is almost the
23 ... g5 24 liJf1 ~e7 25 g4 'i'd7 26 liJg3 antithesis of 13 bs.
liJg7 27 a4 ~b4 28 i.h3 ~b7 29 'i'c2 14a4
~d6 30 liJf5 liJxf5 31 gxf5 ~b4 32 ~g2 Preventing the b-pawn from advancing at
'i'd6 33 f3 :e7 34 :e2 :de8 35 :ce1 all and thus fixing it as a target. 14 iVb3 is
'Yif6 36 i.g4 i.d6 37 'Yid1 ~b4 38 'Yic2 considered in the next main game.
:d8 39 lId1 i.c8 40 e4 ~xc3 41 e5 14... ~c8!
':xe5 42 dxe5 i.xe5 43 ~de1 ~c7 44 QIestian 66. Wow!
':e8+ c;t>g7 45 ':xd8 i.xd8 46 ':d1 i.b7 A nsr.m- 66. This is the modem method of
47 f4 d4+ 48 .t.f3 d3 0-1 playing these positions.
79
Qiestian 67. It seems familiar somehow! 15.:Ib1

Ansr.rer 67. The concept is derived from a 15 'iib3 ~e6 16 l:tfdl ttJd7 17 tlJel iLe7!
line of the T artakower that we have seen 18 ttJd3 iLd6! (stopping ttJd3-f4) 19 ttJb4
earlier: 8 cxdS ltJxdS 9 ~xe7 'iixe7 10 ~xdS 'iii'a8 20 l:tab 1 %:tc8 21 'iib2 ltJf6 (Black has
exdS when Black aims to develop his bishop achieved his ideal set-up) 22 ~ba2 h5! 23
not to the obvious b7-square, but to e6. ltJc1 h4! was nice for Black in Peter-Siegel,
Qiestian 68. So what's the point? Budapest 1997.
Ansr.rer 68. White can easily develop 15 ...i.e6 16 'ilc2 tiJd7 17 :'fc1
pressure against the b6-pawn by playing his Instead 17 iLd3 l:te8 18 l:tfel g6 19 ttJe2
queen to b3 and a rook to b 1. Obviously, c5 20 ttJf4 ~g4 21 ttJxdS iLxf3 22 gxf3 cxd4
Black will deal with this pressure by 23 ~b5ltJe5 24 'i'e4 dxe3 25 fxe3 ':e6 was
developing his knight to d7. The key vety mwky in Rychagov-Rustemov, Moscow
question is the bishop on b7. Just as in the 8 Championship 1996, while Maksimenko-
cxd5 line, the bishop is much less actively- Beliavsky, Tivat 1995, saw 17 %:tfe1 'i'c7 18
placed on b7 than on e6 - on b7, it is ~d3 %:taa8! ~ like this move - Black re-
blocked by the pawn on c6 and the pawn on establishes the connection of his rooks, and
dS. Even if Black achieves his ... c6-c5 break, makes his position safer in anticipation of the
all the bishop does on b7 is defend the pawn. coming central break) 19 e4 dxe4 20 ltJxe4
From e6, the bishop defends the dS-pawn, (20 ~xe4 %:tac8 21ltJe2 'iWd6 22 %:tbc1 ~dS
while having access to squares along the c8- is unclear according to Gagarin) 20...~d5!
h3 diagonal. Consequently, before White (see how useful the bishop is on e6 rather
forces ... ttJb8-d7, Black transfers his bishop than b7) 21 ~c4 %:tfc8 22 ~xdS cxdS 23
toe6! ltJxf6+ ttJxf6 24 'i'xc7 %:txc7 25 %:txb6 %:txa4
Qiestian 69. Isn't Black wasting a lot of 26 %:tb2 with equality.
time though?
Ansr.rer 69. In a way, yes, but because
Black has taken care of the safety of his king
already, there is no way for White really to
exploit this.
Qiestian 70. So what is \X1hite trying to do?
ArJ..!r.rer 70. White's ultimate aim is to
achieve the e3-e4 break and to blow open the
centre. In this context, the bishop is
excellently placed on e6: it covers the dS-
square directly while protecting sensitive
kingside light squares such as f7 and f5.
However, it must be said that Kramnik
was successful with the older 14... c5 against 17 ...'ila8!
Lautier at Belgrade 1997. After 15 'ilb3 tLla6 A typical way of actlvatmg the black
16 l:tfbl ttJb4 17 ltJa2 ltJxa2 18 ltxa2 ~a6 queen.
19 ~xa6 llxa6 20 'iib5 'tWa8 21 dxc5 bxc5 22 18 i.d3 c5 19 .i.h7+ ~h8 20 i.f5 Wc6
"xc5 :xa4 23 b 4 'tWxa4 24 h3, the game 21 h3 :'c8 22 i.xe6 fxe6 23 Wg6 :'aa8
was agreed drawn. I must say though, that I 24 'iih5 'ild6 25 tiJb5 'ile7 26 :a1 e5 27
don't fully understand why 24 'i'xdS was not .:Ie1 'ile6 28 tiJh2 e4 29 ':ed1 cxd4 30
possible. After 24...l:tdl+ 25 ttJel I don't see tiJxd4 ~xd4 31 ':xd4 tiJf6 Yz -Yz
a follow-up for Black. Qiestian 71. That's all vety well, but why
80
doesn't White just immediately go for the b- c6-pawn will need another piece to defend it
pawn with 'iVd1-b3 and l:ta1-b1 or lDc3-a4? and if it advances, then dS is chronically
Anmer 71. I'm glad you asked me that weak. However, in this case, due to
question! misplaced knight on a4, Black gains some
_--------------.. unexpected counterplay.
Game 40 17 %:tab1?!
Hulak-Lutz 17 ~xb6 loses to 17...:tbs of course.
Wijk aan Zee 1995 Siegel-Lutz, Germany 1994, saw 17 :tac1 but
17...cS! 18 dxcS? (18 ~xb6? c4 leaves the
1 d4 liJf6 2 e4 e6 3 lbf3 d5 4 lbe3 i.e 7 knight on b6 vexy precariously placed
5 i.g5 h6 6 .th4 0-0 7 e3 b6 8 i.e2 according to Lutz, whereas 18 ~c3 ~b4
i.b7 9 .txf6 i.xf6 10 exd5 exd5 11 b4 [eyeing dJ] 19 :'fd1 "a8 20 a4 :'d8 is
e6 12 0-0 a5 13 bxa5 %:txa5 14 'i'b3 White's safest, but is absolutely fIne for
.te8! Black) 18...bxcS 19 ~c3 c4! 20 "b1 ttJc5 21
ttJd4 .txd4 22 exd4 ttJdJ gave Black the
advantage. The safe retreat 17 ttJc3 is best
according to Lutz, when 17...b5!? 18 a3
"a8!? gives Black counterplay against the
white a-pawn.
17 ... b518lbb2
18 ttJc3 was safer when Black plays
18.....e7 19 a4 b4 20 ttJa2 :tb8 intending
.....e7-e6 and ...c6-cS with counterplay.
Still.
15lba4!?
15 .tdJ was met by 1S....te6 in I.Sokolov-
Lutz, Gannisch rapidplay 1994, when 16 a4
c5 17 .tb5 ~a6! 18 :tad1 c4!? 19 'iib1 Cfjc7
20 ~eS ~xbS 21 axbS .txe5 22 dxe5 'i'a8
23 f4 :ta3 24 "c2 .tfS 25 "xfs ltxc3 was
vexymwky.
Qg:stinn 72. So why not 15 :'abl?
Anszar 72. As Lutz points out, 15...i.f5! 18 ... e5! 19 dxe5 lDxe5 20 'i'b4 lbe4 21
gains a tempo on the rook on b 1 to develop lbd4 i.xd4 22 exd4? ':xa2 23 %:ta1 %:txa1
the bishop and after 16 :tb2 Cfjd7 Black has 24 ':xa1 'i'b6 25 f3 lbg5 26 lbd3lDxf3+!
the development set-up he wants. The text 0-1
seems to force ... ttJb8-d7 but...
15 ... i.a6! 16 i.xa6 lDxa6! Game 41
The exchange of light-squared bishops is Izeta-Asrian
always something that requires great care Ubeda 1998
from Black as it greatly weakens the central
and queenside light squares. For example, the 1 d4 d5 2 c4 e6 3 ltJf3 lDf6 4 ltJc3 i.e 7
c6 120-0 -.d6
A rather slow alternative to the nonnal
12... a5. White now develops a useful
initiative.
13 -.b3 !bd7 14 ~He1 iLe7 15 :ab1 a5
16 bxa5 l1xa5 17 a4 :e8 18 iLf1 iLf8 19
-'c2 g6 20 e4! dxe4 21 !bxe4 -.f4 22
iLc4!
Arzsrrer 74. White is actually aiming for a

sort of improved version of the 11 b4 and 13
a3 line above: the aims are certainly the same.
By placing the queen on b3, White deters his
opponent from playing ...c7-c5 as Black
cannot recapture on c5 with the pawn as his
bishop on b7 will be loose. Moreover, of
course, the rook on d1 anticipates ... c7-c5 so
The f7 -square is extremely sore in this that after d4xc5 White will have a large
line: you can see why Black players now tend amount of pressure on dS. The queen on b3
to use the 12... a5 move order to retain the is a key piece since by attacking the dS-pawn,
option of transferring the light-squared it prevents Black from developing his knight
bishop to e6! to d7 immediately. Consequently, if Black
22 ... iLg7 23 :e2 24 d5 :aa8 25 :be 1 wishes to develop his knight to d7, he must
l1ad8 26 -.b3 iLa8 27 g3 -.b8 28 d6 :f8 first play... c7-c6 to hold his dS-pawn. Then
29 iLxf7+! l:txf7 30 !beg5 hxg5 31 !bxg5 we get a similar black queenside pawn
l1df8 32 l1e8!! 1i'xd6 33 1i'xf7+ ~h8 34 structure to the line above where White
!be6 1-0 aimed to exploit the Black's pawn structure
By any standards this was a really great by playing for e3-e4. TIlls is White's most
game from Karpov. consistent plan: to play i.f1-d3, castle and
then aim for e3-e4.
11 ...:e8
Game 45 l1...c6 12 i.d3 i.c8!? was tried in
Dautov-Kir .Georgiev Karpov-Beliavsky, Belgrade 1996. After 13
Elista Olympiad 1998 0-0 i.g4 (13 ...i.e6!? seems possible and then
maybe 14 tLJe5!?) 14 ttJe2!? 'ife7 15 i.b1 :c8
1 d4 ~f6 2 !bf3 d5 3 c4 e6 4 !bc3 iLe7 (15 ... tLJd7 16 'ifc2!) 16 h3 i.xf3 17 gxf3, the
5 iLg5 h6 6 iLh4 0-0 7 e3 b6 8 1i'b3!? position is not easy to assess, but Black's
iLb7 9 iLxf6 iLxf6 10 cxd5 exd5 11 ':d1 light-square queenside weaknesses are more
annoying than White's kingside doubled
pawns.
Qiestion 74. This looks like rather an odd 12 i.d3
plan! The quieter 12 iLe2 avoids Black's next
84
Tartako wer Variation: Fixed Centre Plans
freeing idea, but gives him a little more time Ansrrer 75. White has a vel)' small
to arrange himself. After 12...c6 13 0-0 lLld7 advantage, but it is really vel)' little. Black's
14l:Ue1 (14 ~d3 ttJf8 15 e4lLle6 16 e5 ~e7 IQP is compensated by his two bishops.
17 i.f5 j..£8 was played in Barlov-Schlosser, Contrast this position with ones we saw in
Haninge 1988, and now Barlov gives 18 g3 the game Gretarsson-Yusupov and you will
followed by ~h3-g2 as a slight edge for see the value of White's pawn on b2 - his
White. I really have my doubts as to how position seems so much more solid.
good these positions are for White. Black will 16 'i'a3
play for ...c6-c5, perhaps after a preliminary 16 "c2 :tc8! 17 i.h7+ <itth8 18 i.f5 ttJe6!
advance of his queenside pawns with ...b6- 19 ttJd4 (19 'iib3 i.xc3 20 bxc3 "f6 21 i.g4
b5, ...a7-aS and ...b5-b4. White's dark-square ~b7 was fme for Black in Gulko-
pawn chain can be very vulnerable.) 14...ltJf8 Radashkovich, USSR 1971) 19...ltJxd4 20
15 e4 ttJe6 16 exd5 cxdS 17 i.f1 :te7 18 :te2 exd4 :tc7 21 "d3 g6 22 i.g4 h5 23 i.f3
:tc8, as in Malisauskas-Vander Sterren, :tce7 was very pleasant for Black m
Yerevan Olympiad 1996, Black was very Beliavsky-Kramnik, Belgrade 1997.
comfortable. 16 ... a5
12 ... c5!? Qlestion 76. Doesn't Black want to gain
Black's most dynamic approach, using two bishops versus two knights with
tactics to achieve his goal. 12...c6 13 0-0 ltJd7 16...ltJxd3?
is also possible, though obviously White is a Ar23ra'I' 76. It is a possibility, but it makes
tempo up on Barlov-Schlosser. the dS-pawn a little harder to defend. The
13 dxc5 lbd7! 14 c6 knight on c5 is a nice active piece, taking
White settles for a positional gain rather away b3 from the white queen and d3 from a
than entering into the tactical complications white rook and thus making it hard for White
of 14 cxb6 (14 lDa4 d4l 15 ~g6 i.xf3l 16 to co-ordinate his heavy pieces against the d-
i.xf7+ ~h8 17 gxf3 ltJeS! 18 i.xe8 "xe8 is pawn. Moreover, it has good outposts both
actually vel)' dangerous for White due to the on c5 and on e4 later, so it doesn't seem
exposed knight on a4 and the threat of worth it to exchange it for a bishop that is
...ltJe5xf3+ and .....e8-h5) 14... d4 15 i.b5 or doing little in this position.
15 i.g6!? 16...a5 secures the knight on c5 by
14... i..xc6 15 0-0 lbc5 preventing b2-b4.
17 lbe2 'i'd6 18 lbed4 i.b7 19 :c1 g6
20 :fd 1 :ac8 21 i..b5 :ed8 22 g3
Dautov claims a slight advantage with 22
:tc2 <ittg7 23 :tdc1 'iib8 24 b4 axb4 25
"xb4, but I don't feel that this is vel)'
frightening for Black.
22 ...'itg7 23 lIc2 lbe4!? 24 'i'xd6 lbxd6
25 i.c6 i..xc6?!
25 ...i.a6! 26 ltdc1 (26 i.xdS l:Ixc2 27
ltJxc2 ltJe4! with the threat of ...i.a6-e2 looks
horrible for White; 26 :tcc 1 ltJe4 is equal
according to Dautov) 26... ~d3 27 :tc3 ~e4
28 lDd2 lDf5! 29 ttJ2b3 ttJe7! 30 .i.b5 :txc3
QIestion 75. How good is this position for 31 ':xc3 ':c8 is equal according to Dautov.
White? Now White is a touch better again.
85
5 it.g5 h6 6 it.h4 0-0 7 e3 b6 8 it.e2

it.b7 9 it.xf6 it.xf6 10 exd5 exd5 11 b4
e6 120-0 a5 13 a3
16 ... b5!
A typical idea in normal Exchange
variation lines. Here, Black rules out any b4-
QIestion 73. This looks very solid b5 ideas to soften up his central light squares
An.wrer 73. This is the most flexible of and prepares an outpost for his d7-knight on
White's options. It makes the position very c4 via the b6-square. The drawback of course
close to the Exchange QGD where Black has is that the bishop on b7 is now extremely
played the rather unusual and weakening passive and will not be activated by ...c6-c5.
...b7-b6 (he always chooses to keep the b7- 17 it.d3!?
c6-d5 structure intact in the Exchange A suggested improvement of Speelman's
variation). White does not commit himself to over his game with Lputian, Kropotkin 1995,
one structure, but keeps open the possibility where 17 ltJel?! ~b6 18 ~dJ .tc8 19 ~c5
of all three. By maintaining the pawn on b4, .tf5 20 :tal .te7 21 :ta2 :'xa2 22 'ii'xa2
White gains several benefits, particularly .td6 was very pleasant for Black.
against the bishop on b7: The text very logically, in view of Black's
1. White keeps the ...c6-c5 central break entombed bishop on b7, aims for the
under wraps and so keeps the bishop on b7 advance e3-e4.
passive. This gives White a much better 17 ... tbf8?!
chance of achieving the e3-e4 break. A rather passive move. 17...ltJb6! seems
2. By maintaining the threat of the b4-b5 much more logical and after 18 e4 then
break, White makes sure that unlike in the 13 18...ltJc4! 19 eS :'a3! followed by ....tf6-e7 is
bxa5 system, Black cannot easily transfer his rather unclear, while after 18 exds cxds 19
bishop to the c8-h3 diagonal as there is still ~xb5 1i'b6! 20 ~c3 :'a3! Black regains the
always the possibility of b4-b5, attacking the pawn with a good position.
pawn on c6 and softening up Black's light 18 e4! dxe4 19 tiJxe4 it.e8 20 d5 exd5
squares. 21 tbxf&+ 'it'xf6 22 .txb5 :Ld8 23 tiJd4
13 .. .liJd7 14 'it'b3 :Le8 15 :tad 1 .tg4 24 :'d2 :'ab8 25 h3 .te6 26 .te6
This is White's most ambitious move, tbg6 27 b5 ttJf4 28 ':fd1 Ae8 29 'i'f3
preventing 15 ...ltJf8 due to 16 b5! when ':d6 30 ~h2 'i'e5 31 'i'e3 f6 32 tbf3
16...c5 loses a pawn to simply 17 dxc5. 15 'ii'xe3 33 fxe3 ttJe6 34 ':xd5 ':xd5 35
.tdJ is seen in the next main game and 15 b5 :'xd5 i..b7 36 tbd4 tiJxd4 37 exd4 it.xe6
in Game 43. 38 bxe6 .:te8 39 ':d6 :'e7 40 d5 c3;f7 41
15 ... axb4 16 axb4 :e6 :'a7 42 :'e3 1-0
82
Tartakower Variation: Fixed Centre J-'Ians
15 ... c5 Black must be prepared to give up the

Game 42 pawnondS.
Arkell-Short 15 ... c5 16 tiJxd5
British Ch., Torquay 1998
1 d4 ~f6 2 tiJf3 d5 3 c4 e6 4 ~c3 i..e7
5 i..g5 h6 6 i..h4 0-0 7 e3 b6 8 i..e2
i..b7 9 i..xf6 i..xf6 10 cxd5 exd5 11 b4
c6 120-0 a5 13 a3 ~d7 14 'ilb3 ':e8 15
i..d3!? axb4 16 axb4 ~f8!
16 ... i..xd4! 17 ':ad1

17 exd4 l:txe2 18 :lfe 1 (suggested as
interesting by Short) was tested in Ziiger-Van
der Sterren, Winterthur 1996, when 18... a4 19
'ii'c4 l:txe1+ 20 lhe1 'it>f8 21 CfJe7 ~xf3 22
gxf3 cxd4 23 'ii'xd4 CfJc5 was fme for Black.
17 ... ~e5! 18 liJxe5
Without the rook on d1, Black has no fear 18 CfJf4 ttJxf3+ 19 ~xf3 ~xf3 20 gxf3
of b4-b5 and thus can transfer the knight 'ii'd6 equalises according to Short.
immediately to e6, where it will eventually 18 ... i..xd5 19 ~c4 'ilg5 20 g3 'ilf5! 21
support the ...c6-cS break while attacking d4 ':fe1 'ile4! 22 f3 'i'xe3+! 23 'ii'xe3 ':xe3
and thus deterring e3-e4. 24 ~xe3 i..xe3+ 25 'ittf1 i..d4! 26 ':xd4
17 ':fd1 ~e6 18 i..f1 'ild6 19 tiJe1 h520 Necessary. In Arkell-Parker, Hastings
g371 ':xa1 21 ':xa1 i..xd4! 1995, White got into trouble after 25 'it>g2
Ouch! A typical and often fatal tactic. ~d4 26 ~d3 ~b3 27 l:tc1 ~f8 28 ~c4 a4
22 exd4 ~xd4 23 'ila3 ':xe1 24 ':xe1 29 l:te2 l:td8 30 f4 .tf6 due to his weak
~f3+ 25 'itth1 ~xe1 26 'ila7 'ile7 0-1 queenside pawns.
26 ... cxd4 27 ':d1 ':c8 28 ':xd4 ':c5 29
Game 43 f4 'ittf8 30 i..d3 'itte 7 % -%
Karpov-Short Finally, for this line, a classic that shows
Amsterdam 1991 what can go wrong if you get your
development a little tangled up.
1 d4 liJf6 2 c4 e6 3 liJf3 d5 4 ~c3 i..e 7
5 i..g5 h6 6 i..h4 0-0 7 e3 b6 8 i..e2 Game 44
i..b7 9 i..xf6 i..xf6 10 cxd5 exd5 11 b4 Karpov-Kir .Georgiev
c6 120-0 ':e8 13 'ifb3 a5 14 a3 ~d7 Tilburg 1994
Via the older move order, we have reached
the same position as the previous two games. 1 d4 ~f6 2 c4 e6 3 liJf3 d5 4 ~c3 i..e7
15 b5!? 5 i..g5 h6 6 i..h4 0-0 7 e3 b6 8 i.e2
The text is obviously crucial, since after i..b7 9 i..xf6 i..xf6 10 cxd5 exd5 11 b4
83
26 ltJxc6 lDf5 27 l:tdc 1 l:td6 28 g4 ltJe 7 18 'iVd2!

29 lDxe 7 :xc2 30 ':xc2 iLxe 7 31 ~f1 Aiming for b4.
':f6 32 ~e2 iLc5 33 h3 ~f8 34 ':d2 ':d6 18 .....e7 19 'i'a5 ':fe8 20 l:tb1! ~g7 21
35 tbd4 iLxd4 %-% l:tb6 ':c7 22 h4! h5 23 tbg5 iLxd4 24
Finally, we examine two systems that can exd4 ltJe4 25 i..f3 :ec8! 26 'i'b4 'i'xb4
tend to lead into one another: 8 :el .t b7 9 27 :xb4 tbd2 28 :d1 :c1 29 iLe2
.txf6 .txf6 10 cxds exds and 8 .td3 .tb7 9 :xd1+ 30 i.xd1 :c7 31 ~b2 tbe4 32
.txf6.txf6 10 cxds. liJxe4 dxe4 33 ~h2 iLd5 34 ~g3 ':c4 35
:d2 :c3+ 36 ~f4 f6 37 :b2 :d3 38
Game 46 iLb3 iLxb3 39 :xb3 :xd4 40 :a3 % - %
Akopian-Short
Linares 1995 Game 47
Pinter-Portisch
1 liJf3 d5 2 c4 e6 3 d4 tbf6 4 tbc3 iLe 7 Austria 1997
5 iL 9 5 0-0 6 e3 h6 7 iLh4 b6 8 :c1 iLb7
9 i..xf6 i..xf6 10 cxd5 exd5 11 b4 1 d4 ltJf6 2 c4 e6 3 tbf3 d5 4 liJc3 iLe 7
5 iLg5 0-0 6 e3 h6 7 iLh4 b6 8 iLd3
i..b7 9 iLxf6 il.xf6 10 cxd5 exd5 11 :c1
The 8 :tel move order of the previous
game and this move order can easily
transpose into one another. I suppose that if
White wants to force this fIXed-centre
position then he should play it via the 8 :tel
move order as 8 .td3 gives Black the extra
idea of 8...dxc4!?
11 ... c5 12 0-0 cxd4
12...:te8, waiting a little more flexibly, was
tried in Morovic Fernandez-Short, Pamu
1998, when after 13 .tbS J::.e6 14 b3 a6 15
11 ... c5!? .td3 cxd4!? 16 ttJxd4 .txd4 17 exd4 ttJc6 18
Obviously, there is a lot to be said for 'ii'g4 'ti'd6 19 .tfS :f6 20 ttJe2 g6 21 .tb1
playing this move if it is possible. However, :te8 22 ttJg3 .tc8 23 "d1 h5! Black's activity
11...e6 is also not stupid, aiming to meet 12 compensated for his slightly worse bishop.
'iib3 with 12...aS! 13 bxaS lhaS 14 .te2 13 exd4
.te8!, as in Korchnoi-Zviaginsev, Tilburg The most popular choice, though 13
1998. After 15 0-0.ie6 16 a4 ttJd7 17 'ii'b1 ttJxd4 is also not without venom:
.te7 Black. had a very reasonable position, as a) Yerrnolinsky-Short, Pamu 1998,
in the 11 b4 lines above. continued 13 ...tt)c6 14 ttJxc6 .txc6 15 ttJe2
12 bxc5 bxc5 13 dxc5 liJd7 14 liJb5 ::'c8 'ii'd6 16 'ii'd2 :tad8 with a small edge for
15 iLe2 ~xc5 White.
Akopian also suggest 15...lhe5!? b) 13 ...ltJd7 seems very reasonable as it
16 0-0 a6 17 liJbd4 g6?1 transposes to a very similar type of position
17.....a5! 18 ttJb3 (18 :tc2 ttJe4) to the 8 'ii'b3 lines, e.g. 14 ttJfS (14 ttJce2
18...~xb3 19 'ii'xb3 .ta8 was fairly level ttJc5 15 'ii'c2 ltJxd3 [15 ... aS!?] 16 'ii'xd3 'ii'd7
according to Akopian. Now Black has 17 :tc2 :tfc8 18:td2 g6 was fme for Black in
trouble dealing with the white queen. Yennolinsky-Vaganian, Groningen PCA
86
1993) 14...tDcS 1S ~b1, as in Zviaginsev- move h2-h4-hS would have no effect, but
Van der Sterren, Wijk aan Zee 1995, and here, with the h-pawn already committed to
now maybe 1S...aS to hold the knight on cS h6, it will force a reaction from Black
and protect it from b2-b4. Perhaps ...h6-hs when the g5-square
13 .. .tDC6 becomes available for the white knight, from
where it can attack the f7-square.
14i.b1 ':e8 15 .d3
15 'ii'd2!? worked well in Zviaginsev-Van
der Sterren, Reykjavik 1994, after 15 ....ta6
{15 ...'ii'd6} 16 .tdJ .tb7 17 'ii'f4 with a slight
edge for White.
15 ... g6 16 ':fe1 .d6 17 ':e3!?
17 a3 was the previous attempt with the
idea of following up with .tb 1-a2 attacking
the dS-pawn. However, the slight weakening
of the queenside light squares gives Black an
OppOltunity to activate his knight with
17...1:.ac8! 18 .ta2 lLlaS 19 1:.xe8+ 1:.xe8 20
Qiestion 77. Is this really good for White? b4 lLlc4! 21 lLlxdS .txdS 22 .txc4 'ii'f4 23
An3ZW' 77. This is one of those slightly 1:.dl .txf3 24 'iixf3 1:.e4! 25 'ii'xf4 1:.xf4 ~
mysterious positions and structures that ~ Ftacnik-Vander Sterren, Sydney 1991.
looks rather innocuous and yet scores 17 ....:xe3 18 fxe3!?
incredibly well for White in practice. A look Pinter also gives 18 'ii'xe3 as a slight
at the statistics shows that White is winning advantage for White.
seven or eight games to Black's one. 18 ...i.g7 19 a3 ':c8 20 i.a2 tiJe7 21
Qiestion 78. But White now has an IQP, ':e1 g5?
while Black has the two bishops! Really risky. 21...tDf5 22 'iib5 a6 23 'ii'b3
An3ZW' 78. White can make life awkward :le8 24 lLla4 1:.e6 25 .tbl l:tJe7 26 I:tJc3 is
for Black due to three factors related to the given by Pinter as a slight edge, but it isn't so
Tartakower system: the presence of the light- much.
squared bishop on b7, the presence of the 22 ':f1 :d8 23 'iith1 'ii'g6 24 'ii'e2 'ii'h5
pawn on h6 rather than h7, and the absence 25 i.b1 ':e8 26 i.d3 tiJc8 27 'ii'f2 tiJd6
of the knight from f6. 28 lbd2 'ii'h4?? 29 g3! 'ii'h5 30 lbb5! i..f8
Qiestion 79. Sounds like a case for 30...lLlxb5 31 g4! is the nice point!
Sherlock Holmes! 31 tiJxd6 i..xd6 32 g4 'i'xh2+ 33 'i'xh2
An3ZW' 79. Hmm. Let's take the fIrst two. i.xh2 34 'iitxh2 ':xe3 35 ':f3 ':e 1 36
The light-squared bishop is not on the c8-h3 lbf1 :d1 37 tiJg3 i..c8 38 'iith3 i.d7 39
diagonal any more. This means that the light i.f5 i.e6 40 ':d3 ':e1 41 :d2 :e3 42
squares around the black king are not ':e2 :b3 43 i.xe6 fxe6 44 ':xe6 ':xb2
covered by this bishop. Thus f5 is available 45 ':e7 a5 46 tiJf5 'iitf8 47 ':b7 'iite8 48
to a white piece, while f7 is less protected tiJxh6 'iitd8 49 tiJf7+ 'iitc8 50 tiJd6+ <&td8
than if the bishop were on e6. Thus consider 51 a4 ':b4 52 tiJb5 ':xa4 53 :xb6 ':b4
the situation after White plays a plan with 54 :d&+ 'iite7 55 :xd5 'iitf6 56 tiJc7 a4
.i.b1 and'iidJ threatening mate on h7. The 57 :f5+ <&tg6 58 :f&+ 'iith7 59 d5 :b3+
narural, indeed only, defence is ...g7-g6. If the 60 'iitg2 :b4 61 'iitf3 :b3+ 62 ~e4 ':b4+
h-pawn were still on h7 then the softening 63 ~e5 1-0
-87
Summary
In the main line, I really do prefer Kramnik's 11...c6 to Kasparov's 11...c5 - I think you need
to be a bit too strong to play Kasparov's line successfully. For White, Pinter's choice against
Portisch seems like an interesting and not theoretically heavy way to play.
1 d4 d5 2 c4 e6 3 ltJc3ltJf6 4ltJf3 i.e7 5 i.g5 h6 6 i.h4 0-07 e3 b6

8 cxd5
8 ~c1 iob7
9 cxdS - Game 33; 9 ioe2 - Chapter 6, Game 56; 9 .txf6 - Game 46
8 ioe2
8... iob7 9 i.xf6 ioxf6 10 cxdS exdS 11 b4 (D)
11...c5 12 bxc5 bxe5 13 :lb 1
13 .. :iiaS - Game 34
13 ... ioc6 14 0-0 ttJd7 15 iob5 'iVc7 16 'iVd3 :lfe8
17 h3 - Game 35; 17 ltfc1- Game 36; 17 :fdl- Game 37
11...c6 120-0
12 ... aS
13 b5 - Game 38
13 bxaS ~xa5
14 a4 - Game 39; 14 'i!i'b3 - Game 40
13 a3 ttJd7 14 'i!i'b3 :le8
15 ~ad1- Gt1J11e41; 15 iod3 - Gt1J11e42; 15 b5 - Glme43
12... 'iVd6 - Game 44
8...dxe49 i.xc4 iob7 - Chapter 6, Games 48·50
8... ttJbd7 - Chapter 6, Game 57
8 'ifb3 - Game 45
8 i.d3 (D)
8... i.b7
9 ioxf6 - Game 47
9 0-0 - Chapter 6, Game 56
8... dxc49 ioxc4 i.b7 - Chapter 6, Games 48·50
S ...lLlxd5 91Lxe7 'iixe7 10 lLlxd5 exd5 11 :'c11i.e6 12 'iia4 c5 13 'iia3 :cS (D) 141i.b5
14 ioe2 - Game 32
14 ... a6 - Game 31
11 b4 8 i.d3 13.. :~.c8
88
CHAPTER SIX
Tartakower Variation:
Development Plans
1 d4 d5 2 c4 e6 3 tDc3 i.e 7 4 tDf3 tDf6 Black can avoid the fixed-structure

5 i..g5 h6 6 i.h4 0-0 7 e3 b6 variations that arise after 8 i.e2 i.b7 9 i.xf6
In this chapter we shall consider systems i.xf6 10 cxds exds or 8 i.d3 i.b7 9 i.xf6
in which White takes a long-term view of his i.xf6 10 cxd5 exds. Note that Black waits
opponent's set-up. White simply develops his until his opponent has moved his light-
pieces on good squares, trusting that in the squared bishop before taking on c4 so that
ensuing struggle, his pieces will be better White's bishop reaches c4 in two moves.
placed than Black's. Obviously, since White Qtestion 1. Is there a downside to
does not fix the structure, both sides enjoy a capturing so early on c4?
great deal of flexibility. AnSlW'1. If you look at the statistics - not
We shall fll'st examine systems with a very really! Digressing slightly, this is one of the
early ...dsxc4 (Games 48-50), before most confusing things about the QGD -
considering various lines in which Black both for players seeking to take it up, and for
avoids an early exchange in the centre me preparing this book! In every conceivable
(Games 51-57). The most important point line, you fmd that reasonably strong players
about the ... dsxc4 move order, is that by have agreed short draws with each other.
playing an immediate 8... dxc4 9 i.xc4 i.b7 This can make it hard to gauge exactly what
is a good line and what is not! So sound is
Black's development in general, that an
inaccurate move order only has subtle
consequences that are not immediately
visible to the casual glance.
For example, the drawbacks to an early
...dsxc4 are neither tactical nor violent, they
merely involve questions of choice. By
committing himself to an early ...dsxc4, Black
significantly reduces the choice of pawn
structures available to him. For example,
after a subsequent ...c7-c5, we can now only
reach a symmetrical-type structure after
89
d4xc5 or ...c5xd4, lDf3xd4, or an IQP space advantage means that he has the e2-
position after ...c5xd4, e3xd4. White square at his disposal, but Black has no such
consequently has a much stronger idea of post. Consequently, by exchanging off the
what his opponent is playing for, which dark-squared bishops, Black aims to liberate
means that he can detennine the best squares e7 for the queen in order to avoid problems
for his pieces, particularly his rooks, at an once he breaks in the centre with ...c7-c5.
earlier stage. The immediate 10... c5? is a mistake: 11
Qiestion 2. I'm a bit disappointed! I was dxc5 'iixd1 12 J:Ifxd1 hcS 13 lLle5! :c8 14
hoping for some clear, concrete reason! i.e2 ltJc6 15 i.xf6 gxf6 16 lLlxf7 <it>xf7 17
A173U8" 2. I'm sony - that's it I'm afraid! J:Id7+ i.e7 18 :xb7 was vel}' good for
---------------- White in M.Gurevich-Kamsky, Linares 1991.
Game 48 11 i..xe7
Yurtaev-Beliavsky This falls in with Black's plan of fInding a
Yerevan Olympiad 1996 square for his queen. White's alternatives
- - - - - - - - - - - - - - - - . . here are considered in the next main game.
1 d4 d5 2 c4 e6 3 ltJc3 i..e7 4 lDf3 ltJf6 11 .. :iixe 7 1 2 ltJxe4 i..xe4 13 :c 1 l:.d8
5 i..g5 h6 6 i..h4 0-0 7 e3 b6 8 i..e2 14 i..d3
8 i.d3 dxc4 9 i.xc4 i.b7 is exactly the By exchanging the light-squared bishops,
same thing. White hopes to weaken his opponent's
8 ... dxc4 9 i..xc4 i..b7 100-0 queenside light squares and thus to gain
Now Black has a choice between the some profit from them. As Beliavsky points
active 10...ltJe4!? and the quieter 10...lLlbd7 out, the immediate 14 lLle5 is countered by
(Game 50). the clever 14... lLld7 15 i.d3 i.xd3! 16 lLlc6
10 ...ltJe4!? 'iie8 17 'iixd3 ttJc5! equalising.
14 ... i..xd3 15 'iixd3 c5 16 ltJe5
16 'iia3 ttJd7 17 :fd1 <it>f8 is nothing for
White according to Beliavsky.
16 ...'iib7
Intending ...ltJb8-d7 with an end to
Black's problems.
17 b4!? cxd4
17...cxb4 18 f4, intending f4-f5, gives
White reasonable attacking chances
according to Beliavsky.
18 exd4 ltJd7 19 ltJc6 :e8
Qiestion3. What exactly is the point of this White's knight on c6 provides
move? compensation for his isolated queen's pawn,
A173U8" 3. This manoeuvre is very familiar but no more than that. In trying to prove an
to us from the Lasker variation. Black's advantage, White opens lines that only his
, .
major inconvenience, as always in the QGD, opponent s pieces can use.
is fmding ways to activate his major pieces. 20 f4 ltJf6 21 f5 :xe6 23 b5 :ae8 24
Black's main central break is ...c7-c5, but 'iif5 'iid6 25 l:.cd1 1:e2 26 a4 a6 27 d5
once the d-fue is opened, both sides have to axb5 28 axb5 l:.b2 29 l:.de1 :xe1 30
fmd a spot for their queen. White's slight :xe1 g6 31 "f3 ~g7 32liJe7 'iVc5+ 0-1
90
Tartakower Variation: Development Plans
':d8 18 'iWc2 exdS 19 ~xdS ':d6 20 'iif5

.i.c8 21 'ii'f4 ':d7 22 e4 .i.d8 23 e5 turned
out in White's favour.
QIestian 5. Isn't it a rather disgusting
positional idea to take on g3 a'W'<rf from the
centre like this?
Anszar 5. It is, but it is an idea worth
remembering. As alw'<rfs in the Tartakower,
the problem can alw'<rfs tend to be the
weakness of Black's kingside light squares.
QIestian 6. Why?
A nszar 6. This is due to several factors:
1. Black alw'<rfS uses his king's knight to
QIestian 4. Wow, this looks very easy for extract certain concessions from his
Black! opponent, but by moving it from the
A173rW" 4. It seems that in order to fight kingside, he denudes the h7-square of
for a real advantage, White must not give his protection, for example, while allowing
opponent a square for his queen so easily. access to h5 and g4 to the white queen.
Thus, at some stage, he must play.th4-g3. 2. The move ...h7-h6, while giving Black
some room on the kingside, does weaken the
Game 49 kingside light squares and the g6-square in
Vyzmanavin-Zarubin particular. Consequently, for example, a
Russian Team Ch. 1995 knight that comes into e5 cannot be driven
aw'<rf by ...f7-f6 as the knight can then simply
1 d4 lLlf6 2 c4 e6 3 lLlc3 iLe7 4 lLlf3 d5 hop into g6.
5 iLg5 h6 6 .th4 0-0 7 e3 b6 8 iLd3 3. Finally, the presence of the light-
dxc49 .txc4 .tb7 100-0 lLle4 11 lLlxe4 squared bishop on the a8-h 1 long diagonal
11 .tg3!? ~xg3 12 hxg3 leads to a type of rather than the c8-h3 diagonal means that the
position that we will encounter later (Game kingside light squares again rather lack the
53). The alternative recapture 12 fxg3!? protection of the pieces.
Consequently, we see here that the
opening of the f-file has a certain basis - the
rook eyes f7 and can combine with a knight
on e5 and the bishop on c4 against the f7 -e6
pawn chain. Of course, it must also be said
that Black should not take his knight to the
queenside. The knight should go to d7,
aiming later for f6 if necessary.
11 ....txe4
11....txh4!? is worth considering.
12.tg3!?
was tried in Ojanen-Richter, Trencianske 12... iLd6
Teplice 1949, when 12... ~c6 13 a3 ~a5 14 Black has two other logical attempts in
.i.a2 c5 15 'iWd3 ~c6 16 ':ad1 'iWe8 17 dS this position:
91
dxcS winning a pawn.

15 i.d3!
By making Black ....ie7-d6xg3, and thus
forcing him to exchange the dark-squared
bishops without developing his queen to the
key e7-square, White has managed to steal a
tempo (no ...:tfS-dS for Black) on the
previous main game. This is rather crucial as
it makes Black's development, let alone his
chances of achieving the freeing ...c7-c5
break, extremely difficult.
a) 12...c5 was met by the powerful 13

dxc5! i.xc5 14 ~e5! in Groszpeter-Lein,
Saint John 1988. This is a typical idea for
White - to make use of Black's slight
development lag by heading for an endgame
where White has the open d-file and Black
still has to develop his queenside. See also
MGurevich -Kamsky in the note to Black's
tenth move in the previous game. Here, after
14.....xdl (14... ~c6 15 ~d7!) 15 l:Hxdl i.e7
16 :ac1 i.f6 17 i.e2 i.d5 18 b3 as 19 f3
i.g5 20 ~f2 a4 21 h4 .id8 22 e4 White was 15... i.b7 16 liJe5 :c8 17 i.b1 g6 18
in complete control. 'ii'g4 liJd7 19 liJc6 'ii'e8 20 'l'f4 ~g7 21
b) 12...~d7 must be Black's best try, but e4 i.xc6 22 :Ixc6 e5 23 dxe5 l2Jxe5 24
after 13 l:!c1 c5 14 d5!? White still stands a 'ii'f6+ ~g8 25 :Icc1 c5 26 :Icd1 :d8 27
little better. J..c2 :xd1 28 l!xd1 h5 29 :d5 l2Jg4 30
13 l:.c1 i.xg3 14 hxg3 'ii'c3 :d8 31 :xd8 'ii'xd8 32 f3 liJf6 33
Qiestion 7. Ie s strange, but White almost i.b3 'ii'e7 34 'l'd2l2Jd7 35 f4 ~g7 36 e5
always seems to welcome the doubling of his ~f8 37 i.a4 l2Jb8 38 e6 ~g7 39 exf7
g-pawns! 'ii'xf7 40 "d6 l2Ja6 41 i.b3 'ii'e8 42 ~h2
A tmJ.X?r 7. That's true. The key point is that c4 43 i.xc4 l2Jc5 44 'ifd4+ ~h6 45 b4
it takes control of the dark squares around l2Jb7 46 i.d3 l2Jd8 47 'ii'f6 l2Je6 48 i.c4
White's kingside when White plays e3-e4. l2Jg7 49 i.f7 'ii'e4 50 'ii'd8 1-0
With the doubled pawn on g3, White does Black can also try the ...~f6-e4 idea a
not cede control to a black queen or knight move or two later.
of the f4-square, which is a very common
source of counterplay for Black. It also gives Game 50
White the additional possibility of a g3-g4-gS Miralles-Spassky
thrust against the exposed pawn on h6. Angers 1990
14...'ii'e7
14... ~d7 15 i.b5! (mtending i.b5-c6) is 1 d4 l2Jf6 2 c4 e6 3 l2Jf3 d5 4 l2Jc3 i.e 7
awkward for Black as he cannot play the 5 J..g5 0-0 6 e3 h6 7 J..h4 b6 8 iLd3
freeing 15...c5 due to 16 i..xd7 'ii'xd7 17 J..b7 90-0 dxc4 10 J..xc4 ttJbd7
92
Tartako wer Variation: Development Plans
I have tweaked the move order a little 14 J.a6

here (it was actually 9...ttJbd7 10 'ii'e2 dxc4 14 ttJe5!? is worth a tty.
11 .J\.xc4) just to stay with our theme. 14... J.xg3 15 hxg3 e5 16 :fd1 'ile7 17
11 'ii'e2 :ae1 exd4 18 exd4 lLlf6 19 lLIe5 :ad8
11 %Ic1 is interesting here, as 11 ...ttJe4 12 20 'ile3 J.b7 21 J.b5 :ieS 22 J.e6 J.xe6
lDxe4 .J\.xe4 13 .i.g3 transposes to the note 23 lLIxe6 'ilb7 24 'ilf3 :ie7 25 b4 a6 26
to the 12th move in the above previous main a4 b5 27 axb5 'ii'xb5 28 :ie5 'ilb6 29
game, while 12....i.xh4 13 d5!? for example is :ide1 :ib7 YZ-YZ
quite promising for White. These lines show the independent side to
11 .ig3 is also not stupid as obviously ...d5xc4. Overall, this move is just a little too
11...lDe4 12 lDxe4 he4 13 lIc1 is nice for committal. We shall now examine White's
White, whereas 11...a6 (11...c5 12 d5! shows possibilities after 8 .id3 i..b7 9 0-0 ltJbd7.
up a bad side of taking on c4 too early) 12 a4
.id6 and now not 13 'ife2 as in Game 51
Yennolinsky-Vaganian, New YOlk open Yermolinsky -Beliavsky
1997, but 13lDe5! seems to give White a nice Groningen 1993
edge.
11 ...lLIe4 1 2 J.g3 1 d4 lLIf6 2 lLIf3 d5 3 c4 e6 4 lLIe3 J.e7
12 .ixe7 'ifxe7 13 %Iac1lDxc3 14 %Ixc3 c5 5 J.g5 h6 6 J.h4 0-0 7 e3 b6 S J.d3
15 i..b5 e5! 16 %Id1 (16 i..xd7 exd4! as the e- J.b 7 9 0-0 ~bd7
pawn is pinned to the queen on e2) ~ -~ was White now has a choice between the game
Tal-Spassky, Reykjavik 1989, while 12 ltJxe4 continuation, 10 'ife2 (see Games 52-55) and
.ixh4!? (12...i..xe4 13 i..g3 .id6 is possible 10 %Ic1 dxc4 11 hc4 (Game 56).
as 'ifd1-e2 is a less useful move than %Ia1-c1 10 J.g3!?
in this line) is interesting.
12 ... J.d6!?
Dokhoian-Unzicker, Gennan Bundesliga
1992, saw an unusual idea: 12...ltJxc3 13 bxc3
lDf6 14 lIac1 c5 15 %Ifd1 cxd4 16 cxd4 %Ic8
with a decent position for Black Instead
12... ttJxg3 13 hxg3 leads to the type of
positions we consider in Game 53.
13 lLlxe4 J.xe4
Qlestion 8. This looks a bit odd!

Ansr2.U 8. This interesting idea has a
specific sequence in mind The first idea is
that White removes his bishop from the h4-
d8 diagonal so that ...ltJf6-e4 will no longer
gain a tempo against the bishop by
uncovering the attack of the dark-squared
bishop on e7. This fact gives White the extra
positional threat of 11 cxd5 as after
93
l1...lDxd5 (l1...exd5 gives White a position central pawns, then in an IQP structure, he
of the type Yusupov-Kamsky which we saw will find it harder to stop the d4-d5
in the introduction to Chapter 5. For breakthrough without making further
example, Novikov-M.Gurevich, Lvov 1987, conceSSIons.
continued 12 ':c1 a6 13 .tb1 ':e8 14 ~e5 3. By avoiding the exchange of the bishop
when 14...ltJxe5 15 he5 gives White a slight on e7, White leaves his opponent with the
advantage according to Chemin.) Black no same dilemma relating to the development of
longer gains a tempo on the dark-squared his queen. Black is not out of the woods yet
bishop, so White can then reply 12 e4 with a and must still work hard to achieve hannony
nice advantage. Note that from g3, the in his position.
bishop covers the f4-square so that the Q1estion 11. Okay, but all the same, it
knight cannot advance there. seems to have been a pretty inglorious career
Qiestion 9. But although the bishop avoids for the dark-squared bishop - chased around
the dark-squared bishop on e7, Black can still and then exchanged for a knight, while
easily try to exchange it with his knight! creating doubled pawns in White's position!
AMJX!r 9. This is a vezy important point. An.sz.rer 11. True. I know what you mean!
White knows that his dark-squared bishop However, as we have discussed earlier, the
will be exchanged somehow - it has nowhere doubled pawns are not a problem for White.
to hide! However, the point is that White can In fact, they help him keep control of the
decide which piece to exchange it for. The dark squares on the kingside that can be a
natural assumption is that White must source of employment for the black pieces
exchange it for Black's bishop on e7, but as (particularly f4) when White accepts an IQP
we have seen, this completely frees Black's (after ...c5xd4, e3xd4) or when White tries to
position by giving space to his major pieces push with e3-e4. Moreover, the dark-squared
and the queen in particular. Paradoxically, bishop has perfonned one vezy important
White would much rather exchange his function.
bishop for Black's king's knight! Q1estion 12. What?
Qiestion 10. But Black just gains the Anszrer 12. It has teased ... h7-h6 out of
bishop pair! Black's kingside.
An.sz.rer 10. Yes, but as compensation, Q1estion 13. Wow! But isn't ...h7-h6 just a
White gains several factors: useful move, avoiding a tempo on the h7-
1. Black wastes a significant amount of pawn when White plays .tfl-d3 and'iWd1-
time (...lLIf6-e4/h5xg3) acquiring the two c2?
bishops and so White gains some extra time Ansrrer 13. From this point of view, yes,
for his own development in comparison to but the drawback to ...h7-h6 is that it
Black. weakens the kingside light squares by
2. By exchanging his king's knight, Black loosening Black's control of g6, and as we
weakens his defence of two important areas: shall see, this is of importance in a number of
2a. The kingside light squares - Black's different structures.
defence of h7 is weakened, while White's 10 ... c5
pieces gain access to g4 and h5. As we shall see, the main line for White at
2b. The dS-square. By swapping off his the moment is Kramnik's favourite 10 'iWe2
knight on f6, Black weakens his defence of c5 11 .tg3, and if White wants he can
his centre. Thus, if Black seeks to maintain a transpose to this line with 11 'iWe2 here.
pawn on d5, this exchange will make it much Q1estion 14. What is the point of playing
harder for him; if Black swaps off all the 'ii'dl-e2? It doesn't seem anything special.
94
Amcrer 14. Remember that Black's multipurpose move. The reason I like 10
position is vety solid and sound At this early oltg3 fIrst as a move order in this line is that
stage, White cannot do anything here the natural 10... ttJe4 actually loses a
extraOrdinaxy: there is no revolutionary pawn to 11 .txe4! dxe4 12 ttJd2 fS 13 ttJbS!
manoeuvre leading to a huge attackl What when 13 ... cS 14 ttJc7! forking e6 and a8 is
White has to do is to find good squares for fatal. So, in Zaichik-Petrosian, Moscow 1987,
his pieces so that in the middlegame, his Black had to play 13... eS (13 ...l:.c8 14 ttJxa7!)
pieces will be in the right area to cause the 14 dxeS lDcs 1S ttJb3 lDd3 16 lDc1! a6 17
opponent problems. The e2-square is lDd4 f4 18 ttJe6! fxg3 19 'iig4 with a winning
generally a good one for the queen. Anyone position for White. Obviously, 10...dxc4
used to a thoroughly modem opening like transposes to lines studied above (Game SO).
the Semi-Slav (like me, for instance) can
really start champing at the bit at this stage -
in that opening, already you're looking for
the little guy on e8. The QGD requires a
completely different mindset and a great deal
more patience - it's like heading back to a
pre-computer age: somehow your pieces
seem to move to a slower tempo. In this
opening good general moves are required to
prepare yourself for the middlegame - you
cannot win by opening preparation alone.
Q,iestion 15. Okay, but what does 11 'iie2
do?
An.9rtW 15. First of all, by moving the 11 cxd5!? lLlxd5
queen to e2, White connects his rooks and 11."exdS?! 12 ltJeS is obviously nice for
frees the central flles on which they can join White. Black, as always, really wants to keep
the action. TIlls highlights one of the his bishop on b7 active by leaving the as-h1
differences in the respective positions as diagonal open.
Black is not yet ready to do the same. 12 I:tc1
Secondly, White links up with his bishop on Here 12 e4 tDb4! 13 .te2 cxd4 is fine for
d3 along the fl-a6 diagonal. What will he do Black, as is 12 t2JxdS .txdS 13 e4 .tb7
there, it is difficult to say yet - weaken dS by according to Beliavsky.
a future .td3-a6 swapping off the light- 12... cxd4
squared bishop on b7 which helps to defend 12".lDxc3 was tried in Yermolinsky-
the d-pawn, for example? Maybe. Shapiro, World Open 1998, but after 13 bxc3
Q,iestion 16. You sound a bit vague! lDf6 14 'iie2 lDe4 1S .tf4 'iic8 16 lDeS lDf6
An.srrer 16. TIlls is something which will 17 e4 l:.d8 18 l:.fd1 White had a vetypleasant
only happen if a certain set of circumstances lIl1Uatlve.
arise, but the fact that such a possibility exists 13 exd4?!
is a reason why'iWd1-e2 is better than 'iid1- TIlls game is a cautionary tale: don't go
d2 for example. You can't know yet what into this type of IQP position, thinking that
you will use, so tty to play moves that set up 'well, in an IQP position, there are always
as many things as possible. attacking chances.' Black is superbly
Finally, the queen on e2 supports a later organised here and White is not, and if Black
central thrust with e3-e4. It is a nice is careful, his opponent should not get a sniff
95
of an attack. sensitive f7-square, while not only freeing the

d1-h5 diagonal for White's queen to get
involved in the kingside action but also the
third rank on which a white rook can be
swung over to the kingside to join in the fun!
Thus, for example, using an example from
the QGA, after 1 d4 d5 2 c4 dxc4 3 tDf3 QJf6
4 e3 e6 5 J..xc4 c5 6 0-0 a6 7 a4 tDc6 8 'iWe2
cxd49 l:ld1 J..e7 10 exd4 0-0 l1ltJc3 tDb4
12 QJe5 tDfd5 13 ltJe4 b6 14 l:la3!? White
gets to launch an extremely dangerous
offensive.
Q.testion 21. Why did Black have to let the
white knight into e5 by playing 11...tDb4?
Q.testion 17. Why? What is so great about Couldn't he just have waited and then
Black's position? exchanged off the knight when it came there,
Ansrrer 17. This is a very important say by playing 11...b6?
positional lesson that can also be used in A nsrrer 21. This is a very important point.
many other systems, most notably the Black played 11...tDb4 in order to take
Karpov system of the 4 e3 Nimw (1 d4 tDf6 control of the d5-square. For example, after
2 c4 e6 3 tDc3 J.. b4 4 e3 0-0 5 J..d3 c5 6 tDf3 11. ..b6, Black has to reckon with 12 dS!
dS 7 0-0 dxc4 8 ~xc4 cxd4 9 exd4 b6) with breaking through in the centre. Moreover,
which a large proportion of the positions in after 11...b6 12 tDe5 tDxe5 13 dxe5, Black
this line bear a very close resemblance. would wish to put the knight on d5, but since
Qlestion 18. What are the benefits of an White has three pieces attacking dS - the
IQP? bishop on c4, the knight on c3 and the rook
Ansrrer 18. Well, you get loads of attacking on d1 - and Black has only two - the queen
chances ... on d8 and the pawn on e6 - this is not
Qlestion 19. But why? possible.
Answer 19. There are several reasons: Q.testion 22. So what is the 'moral' to this
1. First of all, the side with the IQP always story?
has a choice of posts for his pieces - Ansrrer 22. The moral is that in the QGA
everywhere there are squares for your pieces. position, White's IQP is an active force,
For example, the queen can head just for e2, threatening to move fOIWards and break into
or maybe even for b3 or a4. Moreover, there the black position. Due to this threat, Black
is an open c-flle for White's queen's rook and has to take action to blockade it, which then
a semi-open e-file for White's king's rook. allows White to use the attacking e5-outpost
Thus we can say that the IQP offers a great for his knight. This is a good example of
deal of potential for activity. what the IQP is all about - it should contain
2. The second, and most important, some dynamic force of its own, tying down a
attribute of an IQP is that it offers two small portion of Black's energy so that other
outposts for a white knight - c5, and the small concessions appear. Another example
most natural and desirable e5. would be the most typical IQP trap of all that
Qlestion 20. Why is this so good? can arise from so many openings: 1 d4 d5 2
Ansrrer 20. From e5 a knight surveys the c4 dxc4 3 tDf3 ltJf6 4 e3 e6 5 Jtxc4 c5 6 0-0
world! In particular, it attacks the always cxd4 7 exd4 tDc6 8 ltJc3 ~e7 9 a3 0-0 10
96
.ltd3 b6 11 :le1 .ltb7 12 .ltc2 :le8 13 'ii'd3 'ii'xd1 22 l:Ixd1 .ltxe423 :lc3 .ltxd6 24 :lxd6
1:.c8?? (13 ...g6 is absolutely necesscuy) and :la8 led to a draw, while 18....ltcS!? 19 .ltg3
now 14 dS! exdS 15 .ltg5 with a winning .ltxe4 was also interesting according to
attack. Atalik. In general, this interesting idea does
Qiestion 23. But in this position... not quite seem to offer enough for White,
Ansuer 23. Black already has a super-fmn but it seems like White's best try in the 11
grip on the dS-square - the knight on dS is cxdS line. The text gives White problems
blockading, supported by Black's bishop on very quickly.
b7. This means that Black has no 13 ...:c8 14 i.b1 ttJ7f6
concessions to make to hold back his
opponent's activity: his development was
made for this position.
Q1estian 24. So what has this got to do
with the eS-outpost?
Ansuer 24. Well, you saw how in the QGA
line White got the eS-outpost because his
opponent had to divert pieces to hold back
the IQP. Here, Black has no need of this; he
also has a knight on d7 so that if White
immediately tries to put his knight to e5, then
Black can simply exchange it - end of
problem! TIlls means that Black has more
flexibility - he only allows a knight to eS 1S i.eS?!
when he wants, which in itself interferes with 1S ~eS .ltb4!? 16 .d3 ~xc3 17 oxc3
White's attacking ardour! if.e4!
Q1estian 25. But what can White do apart 1S...llJxc3! 16 l::xc3
from 13 exd4? - 13 ~xd4100ks just equal. 16 bxc3 .d5 17 'ifd3 'ifc4! 18 'ifc2 i.e4!
Ansuer 25. White does have one more is very nice for Black according to Beliavsky.
interesting idea which was seen in Atalik- 16 .....dS
Beliavsky, Yugoslavia 1998: 13 ~xd5 i.xdS 16..,lhc3 17 bxc3 'ifds 18 'ifd3 1:.c8 is
14 e4!? if.b7 (14 ...if.xalloses of course to 15 another good way to play.
1:.a1) 1S i.c7 'ife8 16 ~xd4. White has a 17 l::e3 ::ad8 18 l::fe1 'ifb5 19 llJd2 l2Jd7
temporaty initiative as his opponent's pieces 20 a4 'ifaS 21 llJc4 "b4 22 'it'd3 g6 23
are a little scrunched up, but with Black's b3llJxe5 24 l::xeS J..f6 25 l::xe6 l::xd4 26
position so solid, it doesn't seem that this can 'it'g3 fxe6 27 'it'xg6+ J..g7 28 ~f1 J..a6
last into anything significant. After 16...1:.c8 29 'fIxe6+ ~h8 30 'fIfS J..xc4+ 31 bxc4
(Atalik's suggestion of 16...~cS 17 i.bS .c8 'it'xe1+ 32 ~xe1 l::e8+ 0-1
18 i.eS, intending b2-b4, was tried out in
Khalifman-Asrian, World Championship, Game 52
Las Vegas 1999, when Black managed to Arencibia -Beliavsky
hold the balance after 18... a6 19 .g4 i.gS 20 Elista Olympiad 1998
f4 fS with unclear play) 17 i.bS?! (17 'ife2
if.cs 18 lDbS 'ii'e7 19 b4! .ltxb4 20 ~xa7 1 d4 llJf6 2 c4 e6 3 l2Jf3 d5 4 llJc3 J..e7
l:la8 20 lDc6 .ltxc6 21 :'xc6 is unclear 5 i.g5 h6 6 .i.h4 0-0 7 e3 b6 8 J..d3
according to Atalik) 17...a6 18 .lta4 and now i.b7 90-0 ttJbd7 10 'it'e2 cS
18... bS 19lDxbSlDc5 20 lDd6 'ii'xa4 21 :lxc5 The alternative 10...lDe4 is seen later in
97
this chapter in Game 55. Amzrer 28. I'm afraid so! First of all, let's
11 .tg3 cxd4 take Black's dark-squared bishop. In a
11...~e4 is also popular (see Games 53 Karpov system, it would usually be on b4
and 54), while 11...dxc4 12 J..xc4 t2Jh5 13 whereas here it is on e7.
:fd1 ~xg3 14 hxg3 1Wc7 15 dS exdS 16 Qtestinn 29. It looks better on e7, doesn't
lLlxdS iLxdS 17 J..xdS was a little better for it? Safer?
White in Shchekachev-Lupu, Bourbon-Laney Amzrer 29. Well, in actual fact, it is on a
1998. worse and less active square here. One of the
12 exd4 dxc4 13 i.xc4 key strategical ideas of the Karpov system of
the Nirnzo is that Black can give up the
bishop pair by playing ...J..b4xc3.
Qtestinn 30. Why? White's d-pawn is then
no longer isolated!
Amzrer 30. Black's reasoning is the
following: White's d-pawn is not in fact vety
vulnerable here; Black has not played to put
pressure on it. For example, he has played his
knight to d7 rather than to c6. Black's only
opening concern has been to negate any of
the active features associated with it - the
IQP's 'lust to expand' as Nirnzowitsch so
tastefully put it, and the e5-outpost.
Q.iestian 26. Hmm, so what is this IQP like So Black has prevented d4-dS by
then? developing the bishop to b7 and he has
AnJrrer 26. Interesting! First of all, I have neutralised the e5-outpost by placing a knight
to draw your attention to the huge similarity on d7 that can exchange a white knight
between this variation and the Karpov whenever it comes to eS. Though Black has
system of the 4 e3 Nirnzo (1 d4 ~f6 2 c4 e6 thus made himself safe from the IQP, Black
3 t2Jc3 J..b44 e3 0-0 5 J..d3 c5 6 ~f3 d5 7 is not putting any pressure on the IQP and
0-0 cxd4 8 exd4 dxc4 9 J..xc4 b6). Many of thus not drawing any white pieces to its
the positional ideas for this line are taken defence. Consequently, one of the aims of
from this variation. I actually play both so I ...J..b4xc3 is to create a weakness that Black
will tty and elucidate, but don't be surprised can attack: by drawing the white b-pawn on
by the number of cross-references. to the c-file, Black gives himself such a target.
Q.iestian 27. So what are the differences The other aim in the Karpov line is to give
between them? Black's queen a safe square. By semi-closing
AnJrrer 27. There are three: the c-flle, Black can play his queen to c7,
1. The position of Black's dark-squared which is a pivotal connecting square for
bishop. Black. Of course, in the QGD variation, this
2. The pawn on h6. is not actually possible due to the bishop on
3. The position of White's dark-squared g3, which is another point in White's favour.
bishop. Qtestinn 31. Why is c7 such an important
Strangely enough, these three are all square for the black queen?
interconnected. Amzrer 31. Well, from the Karpov
QIestion 28. Oh no, this isn't one of these variation stem add the further typical moves
subtle, yet huge differences explanations is it? 10 'ii'e2 J..b7 11 J..g5 ~bd7 12 :ac1 it.xc3
98
Tartako wer Variation. Development Plans
13 bxc3 'iWc7. Anszrer 33. Without even waiting for Black

to attack the bishop with ...h7-h6.
Qie5tion 34. Aha! So in the QGD
variation...
Amzar 34. Black has in effect wasted a
tempo with ...h7-h6, driving the bishop back
to a square where it wanted to go. Of course,
... h7-h6 is a very useful extra move in so
many positions, but in this structure, it is not
so useful. There is also one further value to
having the bishop on b4. Not the fact that
Black will take on c3, but the fact that Black
can threaten to take on c3!
Q,iestion 35. I hate it when you try to be
First of all, the general stuff - by moving clever! What does that mean?
the queen off the back rank, Black connects Amzar 35. Well, while Black still has not
his rooks and so becomes ready to involve all played ....ltb4xc3, White still has to prepare
his forces in the battle. Moreover, the queen for two structures - the current IQP
eyes the pawn on c3 which will force White structure as well as the possibility of the
to spend a tempo defending it. However, Karpov structure. And sometimes it can be
there is something even more important. very hard to combine the two effectively - to
With this development, Black activates his fmd a piece set-up that fits both structures.
unit of minor pieces. Q,iestion 36. So the conclusion is ...
Q,iestion 32. What do you mean? Amzar 36. That the bishop on b4 is much
An.9lrer 32. With the queen on c7, the more active than the bishop on e7, so from
bishop on b7 and the knight on f6 both gain this point of view, Black has an inferior
in power and influence. First of all, Black Karpov system. Moreover, White's bishop
gains the possibility of ... ~f6-g4, threatening has been chased to its best diagonal where it
....ltb7xf3 and ...'ii'h7xh2+; secondly, Black interferes with Black's best set-up. Finally,
gains the idea of ....ltb7xf3, forcing g3xf3 due the superfluous ...h7-h6 can also prove a
to the loose bishop on c4; and most weakness in this type of position.
importantly, Black also gains the idea of Qie5tion 37. I suppose that essentially,
...~f6-h5-f4. This is a very annoying idea for you're telling me that this is the problem with
White, harassing the queen on e2 (and his playing ...d5xc4 too early in the QGD lines.
light-squared bishop as well if it retwns to Ansrrer 37. Yes. It's not something
the natural d3-square), while combining with dramatic, but once you have played ...d5xc4,
the light-squared bishop on b7 against the unless some idea with ... ~f6-e4 really works,
g2-square. Moreover, with the knight on dl, then you're committing yourself to an
Black supports a later ...e6-e5 break, should it inferior version of the Karpov system of the
become possible. All this activity is possible 4 e3 Nimw. It's not bad, and it's playable for
only due to the presence of the queen on the Black, but... it's not really so nice to get an
pivotal c7-square. In fact, White's most inferior version of anything!
common plan is to voluntarily retreat the 13 ... ~b4!?
bishop to g3 via h4 in order to shift the black Q,iestion 38. Aha!
queen from the b8-h2 diagonal. Anszrer 38. Yes! Now you understand the
Qie5tion 33. By voluntarily, you mean... sort of thing Black is playing for!
99
Qtestion 39. All the same, Black won't be some time, Black has a great deal more space
able to get his queen to c7 as the bishop is on (and less development) than in the QGD.
g3. Thus, Black always has a wider range of
Armar 39. Black will also look at some choice of squares for his queen. The aim
stage to exchange off the dark-squared after all of ...i.b4xc3 in the Karpov system is
bishop on g3 with ...lDf6-hSxg3 and then to secure a post for the queen by semi-
achieve some hannony in his position by closing the c-fue. With ...b6-bS, Black hopes
putting his queen on c7. It will take a long to gain a little more space on the queenside
time though. for his pieces and thus to free some space for
Qtestion 40. Couldn't Black just play his major pieces - his queen fIrst - within the
13 ...liJhs immediately? position. Thus ... b6-bS would free b6 for the
An.9rW" 40. Yes, this is possible and it was queen, for example, or even for the knight
played in Romanishin-Portisch, Biel 1996, on d7. Thus 14 a4 is very logical - by
when 14 %lfd1 lDxg3 15 hxg3 lDf6 16 lDeS preventing ...b6-bS, White prevents his
i.b4 17 %lac1 i.xc3 18 bxc3 %le8 19 i.b3 opponent from freeing himself in this easy
:c8 20 c4 l:1e7 21 'ife3 %lec7 was quite space-gaining way and forces him to look for
unclear. White should defInitely investigate something else.
14 dS!? - see Vyzmanavin-Beliavsky later on Qtestion 44. But it concedes the b4-square!
in this note. Amr.ar 44. It's only a square! I know it
13 ...i.b4 immediately is quite interesting always feels annoying to give the opponent
as it used to be thought that 13 ...a6 was something like this for free, but remember
necessary. that the inclusion of ... al-a6 and a2-a4 is not
Qtestion 41. Why? all roses for Black. First of all, due to White's
Ansr.rer 41. As we shall see in the sub- battery along the f1-a6 diagonal, the black
sequent analysis, 14 lDbS was thought to be a rook is tied to a8 in order to defend the a6-
good reply to 13...i.b4 from a previous Bel- pawn, which obviously interferes with
iavsky game. Consequently, Black tried 13 ...a6 Black's activity. Moreover, if Black does play
fIrst and after 14 a4 only then 14...i.h4. ...i.e7-b4xc3, then the b6-pawn can become
Qiestion 42. 14 a4? But isn't it good for a liability on the semi-open b-fue as it is no
Black to have this? Why does White do this? longer protected by the pawn on a7 - the
Anszrer 42. Calm down! Just consider move ...al-a6 really does weaken Black's
Black's position for a moment. Why does queenside structure. Again, it isn't going to
Black play ...al-a6? make the pillars of Black's position crumble,
Qtestion 43. In order to play ... b6-bS? but you often fmd that these factors become
An.wrer 43. Exactly! Black's minor piece crucial later on, for example when you
development is excellent - all his minor consider whether to transpose into an ending
pieces are on excellent squares, but his one or not: 'I want to go into this knight ending -
remaining problem is the position his major if only my pawn was on al, then he couldn't
pieces. By developing so quickly and win a pawn on the queenside!' That's why
effIciently, Black has missed out on one thing playing 'good, positional' moves is always
that Black gets in the riskier queen's pawn important - your sins always have a way of
defences like the QGA or the Semi-Slav: catching up with you!
queenside space provided by his queenside After 13 ...a6 14 a4 liJhS!? (14... .i.b4 15
pawns. For example, in the QGA, after 1 d4 nac1 .i.xc3 [IS ... ttJhS 16 d5! lDxg3 17 hxg3
dS 2 c4 dxc4 3 lDf3 lDf6 4 e3 e6 5 .i.xc4 c5 6 exdS 18llJxdS .i.c5 19 b4 .i.d6 20 nfd1ltJf6
0-0 a6 7 'iWe2 bS 8 .i.b3 .i.b7, at the cost of 21 lDe3! 'ife7 22 llJh4 was rather tricky for
100
Black in V)'7Illanavin-Beliavsky, Novosibirsk what Beliavsky had in mind

1995] 16 bxc3 l2Je4 17 JLh4! 'ilic7 18 JLd3 14....li.xe3 15 bxe3 liJe4 16 .li.f4?!
'ii'c6 19 c4 was a little better for White in 16 JLh4, along the lines of Mikhalchishin-
Mikhalchishin-Ivanchuk, Lvov 1987 - the I vanchuk above, looks good enough for a
black knight is rather misplaced on e4, as it slight advantage. The problem with the text
should be on h5) 15 d5 (It seems right to take is that White never gets his bishop out from
the opportunity to play this move, though 15 in front of his c-pawn in time, and so never
:tfd1 tt)xg3 16 hxg3 l2Jf6 17 l2Je5 was also has time to play c3-c4 - you don't want these
interesting in Vyzmanavin-Li Wenliang, pawns to be blockaded on c3 and d4!
Lucerne 1993, when after 17...'ii'e8 instead of
18 'ii'c2?! l..d6 19 iVe2, perhaps 19 g4!?
intending a later f2-f4 and g4-g5) 15 ...ltJxg3
16 hxg3 exdS 17 l..xd5 (17 ltJxdS!?)
17...1..xdS 18 ltJxdS :e8 19 :fd1 l..c5 20
'ii'c2 as 21 'ii'f5 ltJf6 22 ltJxf6+ 'ii'xf6 23
'ii'xf6 gxf6 24 ~f1 is an edge for White due
to Black's horrible kingside pawns as in
Vyzmanavin-Timoshchenko, Norilsk 1987.
16 ...:te8 17 "bd2 liJdf6 18 liJxe4 liJxe4

19 1i'd3 1i'f6 20 .li.e3 liJd6 21 i.b3 i.e4
22 1i'd1 1i'f5 23 f3 i.d3 24 :te1:te6 25
i.f2 i.e4 26 :te5 1i'f6 27 1i'e1 :tfeS 2S
h4 1i'd8 29 h5 1i'e7 30 d5 i..xd5 31
i..xd5 exd5 32 :txd5 "be8 33 i.d4 1i'f4
34 :td1 :te6 35 'iid2 1i'c7 36 :te1 liJd6
37 i.f2 "bc4 38 1i'd4 :txe1+ 39 .li.xe1
:teS 40 i.f2 tiJe5 41 c4 tiJxc4 42 :td7
14 :tae1?! 1i'e6 43 :txa7 liJe5 44 1i'd1 b5 45 f4
In a previous game of Beliavsky's, against "bc4 461i'd71i'xd7 47 :txd7 :ta8 48 :tb7
Ftacnik in Vienna 1986, Ftacnik had played "bd6 49 :tb6 liJe4 50 :xb5 :xa2 51 i.h4
14 ltJb5! and it still looks convincing to me. "be3 52 :tbS+ ~h 7 53 :tb4 "bd5 54 :te4
Then 14...ltJe4 (14... a6 15 ltJc7 b5 16 ltJxa8 f5 55 :td4"be3 56 g3 "bg4 57 :td8 "bh2
bxc4 17 iVxc4; 14...ltJdS 15 i.xdS! i.xdS 16 58 :td3 :te2 59 :td5 liJf3+ ~ - ~
rtJc7 i.xf3 [16...:c8 17 ltJxdS exdS 18 'ii'b5 We shall now examine lines where Black
wins a pawn] 17 iVxf3 :c8 18 iVb7 is clearly plays more flexibly and avoids committing
better for White according to Ftacnik) 15 himself to an early ...d5xc4 at all.
l..c7! 'ii'e8 (15...'ii'f6 16 a3 i.e7 17 i.d3!
ltJg5 18 l!l)e5 leaves Black's queen very Game 53
awkward) 16 a3 JLe7 17 i.f4 'ii'd8 18 :ac1 Kramnik-Yusupov
l2Jdf6 19 I!l)c7 %lc8 20 .i.a6! gives White a Dortmund 1998
very nice position according to Ftacnik. This
looks convincing to me, so I don't know 1 liJf3 d5 2 d4 ~f6 3 e4 e6 4 liJe3 .li.e7
101
5 il.g5 h6 6 il.h4 0-0 7 e3 b6 8 il.d3 is met by 14 ~xe4! dxe4 {14...dxc3 15 ii.xdS!

il.b7 9 0-0 liJbd7 10 .e2 c5 11 il.g3 ii.xdS 16 ~xdS cxb2 17 'ifd2 winning a
liJe4 12 cxd5! exd5 piece} 15 ~xd4 'iWc8 16 ltJfS! with a clear
12... ltJxg3 13 hxg3 exdS 14 ~a6! 'iWc8 15 advantage to White in Belov-Donev, Pravec
~xb7 (IS ~xdS ~xa6 16 ~xe7+ ~h8 17 1989. Of course, after 13 :tfdl instead of 13
~xc8 .i.xe2 is good for Black) IS ...'ii'xb7 16 :tad 1, 16...cxb2 attacks a rook on al so that
dxcS bxcS 17 e4! dxe4 18 ltJxe4 with a nice after 17 'ii'xb2 'ii'e8! equalises. Note also that
advantage for White shows a typical idea in 13 ...i..f6, as in Vyzmanavin-Kotronias, Mos-
this line: White breaks up his opponent's cow 1989, is well met by 14 ii.xe4 dxe4 IS
hanging pawns with a well-timed e3-e4! ~eS cxd4 and now 16 exd4 'ii'e7 17 ttJxe4
13 :ad1 i..xeS 18 dxeS 4JxeS 19 4Jc3 :tfe8 20 :tfe 1 is
clearly better for White according to Arkh-
angelsky and Vyzmanavin. The main move is
13 ... 4Jdf6, as we shall see in the next game.
14 hxg3 a6!?
QIestion 47. Why?
An,grw- 47. One of White's major aims is
to exchange off the light-squared bishops
with ii.d3-a6 in order to weaken Black's
defence of the dS-pawn. With ... a7-a6, Black
prevents this as well as supporting a future
queenside expansion plan with ...cS-c4 and
... b6-bS. Another idea is simply to defend the
dS-pawn with 14... 4Jf6 so that after IS dxcS
QIestion 45. So what's happening here?
Ar13CW" 45. Black has taken the opposite
bxcS 16 ii.a6 (16 .i.c4 "'as 17 4JxdS 4JxdS
18 .i.xdS .i.xdS 19 ltxdS 'ii'xa2 is equal
approach to the previous lines. Rather than accordi.'1g to Vyzmanavin) Black can play
create a weakness in White's position (the 16...1i'b7 17 ii.xb7 'ii'xb7, as in Vyzmanavin-
IQP) and then tty to neutralise his initiative, Kolev, Burgas 1993. However, this allows 15
Black instead tries to hold his ground in the ttJeS!, intending f2-f4 and g3-g4-gS, making
centre. use of White's outpost on eS with an
QIestion 46. You always say that Black advantage according to Vyzmanavin.
should tty to keep the long diagonal free of Instead Vyzmanavin-Pigusov, Moscow
pawns to be kind to his bishop! 1987, saw the interesting 14....i.f6. Black's
Ar13CW" 46. I know, embarrassing isn't it! idea is to cover the eS-outpost and after IS
To be honest, I am always suspicious when- ii.a6 'ii'c8 16 hb7 'iixb7 17 dxcS to
ever Black plays systems like this where he interpose 17....i.xc3! The regrouping that
tries to match White for central occupation White used is typical of this line: 15 .i.bl :te8
right from the early opening (like the 16 :td2 a6 17 :c1 :c8 18 "'dt c4 19 4Jh2
Tarrasch). Since Black has a tempo less than g6 20 a4 ii.c6 214Jg4 hS 224Jxf6+ 4Jxf6 23
his opponent right from move one, this type b3 bS with a complicated position. Finally,
of play always seems fraught in my opinion. Vyzmanavin obtained two good positions
13 ...ltJxg3 against Geller after 14......c7. After IS .i.c2
The nice point to 13 l:tad 1 is that the :ad8 16 dxcS 4JxcS 17 4Jd4 ~f6 18 'iWg4
natural 13...cxd4, aiming for 14 4Jxd4 4Jxc3 :fe8 19 :c1 'ireS, 20 :lfdl g6 21 b4! hS 22
damaging White's queenside pawn structure, 'iWf3 ltJe4 23 ii.a4! :te7 24 ii.c6! worked well
102
at Sochi 1989, and 20 :fe 1 g6 21 b4 h5 22 ~b7 9 0-0 tL'lbd7 10 'ife2 tL'le4 11 ~g3
'ii'f3 ~e4 23 ~a4! worked even better at c5 12 cxd5 exd5 13 l:.ad1 tLldf6
Sochi 1990!
QIestion 48. So you think this is best?

15 dxc5 bxc5 16 i.b1 A~48. Yes. It seems to me that in the
Kramnik also suggests 16 e4 d4 17 lOb1 other lines, Black was vel)' half-hearted about
with a slight advantage for White. ttying to maintain his centre. If Black is really
16•.. tL'lb6 going to tl)' to match White for central space,
K.ramnik points out that 16...lOf6 17 e4! then he should go full out for it, and
d4 18 e5! is vel)' good for White. 13...ltJdf6 seems the most straightforward
17 a4! way of doing so.
A really beautiful dual-purpose positional 14 dxc5!?
mover White threatens a4-aS, driving the 14 ..te5 is an interesting suggestion of
knight on b6 from the defence of dS while Vyzmanavin's and Arkhangelsky's, but 14
allowing ~b 1-a2 increasing his pressure on lOe5 cxd4! 15 exd4 lOxc3 16 bxc3 is not
the dS-pawn. particularly good for White.
17 ...i.f6 18 1i'c2 g6 19 a5 tL'lc4 20 14...tLlxc3 15 bxc3 ~xc5 16 liJd4
lDxd5
20 e4! was stronger according to Kramnik,
who gives 20.....txc3 21 .xc3 .xaS 22 .el!
cJtg7 23 exdS! as clearly better for White.
2o ... lDxb2 21 tL'lxf6+ 1i'xf6 22 l:d2 i.xf3
23 gxf3 .::tab8 24 f4 c4 25 e4 .::tfd8 26
e5 l:xd2 27 1i'xd2 1i'd8 28 1i'e2 1i'd4 29
e6 lDd3 30 i.c2 fxe6 31 1i'xe6+ ~g7 32
~a4 c3 33 i.c2 l:b2 34 1i'e7+ ~g8 35
1i'e2 lDxf4 36 gxf4 ~f7 37 l:d1 1-0
Game 54
Vyzmanavin-Gavrilov
Novgorod 1995 16 ...1i'c8!?
A new idea, intending a quick ...lOf6-e4,
1 d4 d5 2 c4 e6 3 ~c3 ~e 7 4 tL'lf3 tL'lf6 attacking the c3-pawn. 16...'ii'e7 was the old
5 i.g5 0-0 6 e3 h6 7 i.h4 b6 8 i.d3 move and then 17 .th4 'ii'e5 18 f4 (18 .txf6
103
1Wxf6 19 ~a6 nabS was nothing special for 36 i.xf7 1-0

White in Timoshchenko-A.Petrosian, USSR There is one related idea that Black has
1990) 18 ...iid6 19 ttJb3 nfe8 20 ttJxc5 bxc5 tried when White plays 10 1We2: to install his
21 c4 'iVe6 22 .ltxf6 'iVxf6 23 cxdS exdS 24 knights on e4 and f6 before playing ... c7-cS.
e4 c4 25 1i.bl ::tad8 26 e5 'iWb6 27 ~hl 'iVc6 Of course, if White plays 10 ~g3, then this
28 f5 was very pleasant for White in line becomes impossible and Black must go
Vyzmanavin-A.Petrosian, Palma de Mallorca back into the previous examples with 10... c5
1989. 11 iie2.
17 f3
Preventing ...ttJf6-e4. Game 55
17 ...:e8 Nenashev-Vaganian
Novikov-A.Petrosian, Yerevan 1996, saw USSR Championship 1991
the rather bizarre 17.. :iVe8, when 18 ~fe1
'iVe7 19 .th4 nfe8 20 .tc2 a6 21 'iVd3 g622 1 d4 e6 2 e4 12Jf6 3 12Je3 d5 4 iLg5 i.e 7
ttJb3 ~g7 23 ttJxc5 'iVxc5 24 .tb3 ::tac8 25 5 e3 h6 6 i.h4 0-0 7 12Jf3 b6 8 i.d3 i.b7
~c 1 ttJdl 26 'ii'd4+ was very good for White. 9 0-0 12Jbd7 10 "Wie2 12Je4 11 i.g3 tbdf6
18 i.h4 12 exd5 exd5 13 :ae1
Perhaps 18 1i.bS!? No-one has yet tried 13 ::tad1, which is a
18 ... ttJd7 19 f4 iLf8 little puzzling.
Intending ... ttJd7-cS-e4. 13 ... e5
20 i.f5 "Wie7
14l:Ud1
Vyzmanavin considers the position equal Instead Yusupov-V aganian, Elista
here, but there are still plenty of tricks for Olympiad 1998, was agreed drawn after 14
White. dxc5 bxc5 15 1i.a6 1i.xa6 16 'i'xa6 'i'b6 17
21 tbb5 "Wie6 22 c4 ttJf6 23 ~d4 "Wid6 'iVe2 'iVe6 18 ::tfdl ~fd8.
Or 23 ...'iVxc4 24 i.d3 intending 1i.xf6. 14 ... ttJxe3 15 :xe3 e4
24 iLd3 dxe4?! 15 ...'iVd7!? 16 dxc5 bxc5 17 e4 is slightly
24... ttJe4 was better according to better for White according to Nenashev.
Vyzmanavin. 16 i.b1 b5 17 :ee1
25 iLxe4 ttJd5 26 ~f5 "Wi e5 27 ttJxh6+! Nenashev assesses this as slightly better
gxh6 28 'fig4+ <it'h8 29 :xd5 i.xd5 30 for White.
i.f6+ ~h7 31 "Wif5+ ~g8 32 i.xd5 :e6 17 ... l2Je4 18 12Je5 'iie8 19 f3 tbd6 20
33 i.d4 "Wie8 34 'fig4+ ~h7 35 f5 :d6 i.f4 i.g5 21 ~h1 "Wie6 22 \\Ve2 g6 23
104
h4!? .txh4 24 i.xh6 i.g3! 25 i.f4 i.xf4 rather than c 1. Consequently, this is what
26 exf4 ~g7 27 g4 b4 28 ~g2 ~h8 29 Black heads for.
fid2 a5 30 l::te1 fif6 31 g5 'i!fe6 32 ltJg4
t2Je4 33 .txe4 dxe4 34 ~g3 1:.h5 35 d5
'i!fxd5 36 fixd5 .txd5 37 ltJf6 exf3 38
t2Jxd5 ~ah8 39 ~xf3 ~h2 40 ~xc4 ~xb2
41 t2Jf6 1:.d8 42 l:tce4 ~xa2 43 ~e8 ~ad2
44 ~h 1 ~2d3+ 45 ~g4 1-0
And to wrap up the 8 J.d3 lines, a look at
:ac1 ideas.
Game 56
Portisch-Vaganian
St John, Candidates match 1988
1 t2Jf3 t2Jf6 2 c4 e6 3 t2Jc3 d5 4 d4 .ie 7 12 'ii'e2
5 .ig5 h6 6 .ih4 0-0 7 e3 b6 8 l::tc1 .ib7 12 J.g3 4Jh5 (12 ... a6?! 13 dS! exd5 14
9.te2 J.xdS 4Jxd5 15 4JxdS J.xd5 16 'ifxdS :a7
9 J.d3 leads to the same thing. 17 :fdl 'ii'a8 18 'ii'f5 4Jf6 19 4Je5 was very
9 ... dxc4! nice for White in Gheorghiu-Donev,
This is actually an important moment. Liechtenstein 1991 - Black must have been
9... 4Jbd7 would allow 10 cxdS, when wishing here that he had not weakened his
10... ttJxdS is impossible due to 11 4JxdS queenside with ... a7-a6) 13 :el 4Jxg3 14
.txdS 12 J.xe7 'ifxe7 13 ~xc7. Therefore hxg3 4Jf6 (Black avoids opening the c-file
10... exdS is necessary, transposing to the next and so leaves the rook on c1 looking rather
main game. Personally, I'm not so happy useless) 15 a3 :c8 16 4Je5 4Jd7 17 4Jxd7
with these fIxed centre lines for Black, and I 'ii'xd7 18 dS exdS 19 4JxdS J.xdS 20 'iVxdS
feel that Karpov systems where White has was agreed drawn in Piket-Vander Sterren,
already committed his rook to clare fme for Dutch Championship 1991.
Black. Therefore, I would recorrunend this 12 ... a6 13 a4
line against 9 .te2 and 9 .td3 (when fIXed 13 dxc5 4Jxc5 14 ~fdl 'iVe8 15 4Je5 b5
centre lines will be even more dangerous for 16 4Jxb5 was a game from the intenninable
Black, as in Yusupov-Kamsky in the 1984/85 Kasparov-Karpov World Cham-
introduction to Chapter 5). pionship match, when 16 .. :tWb8! 17 J.g3
10 .txc4 ltJbd7 11 0-0 axb5 18 4Jg6 fxg6 19 .txb8 bxc4 would have
Of course, this position could equally well been clearly to Black's advantage.
arise from the move order 8 .td3 .tb7 9 0-0 13 ... cxd4 14 t2Jxd4!
4Jbd7 10 ~c1 dxc4 11 J.xc4. As 14 exd4 4JhS! 15 J.g3 4Jxg3 16 hxg3
11 ... c5! 4Jf6 is quite a reasonable version of this
As we have seen earlier, ~al-c 1 is good position, White opts for the symmetrical
against ... 4Jf6-e4 ideas, but the rook's early option. Black's position is fme in all respects,
development is less precise in Karpov system except his queen. Once he solves this little
positions - very often, White must take problem, he cannot be worse.
advantage of Black's manoeuvrings (such as 14... t2Jc5 15 f3
... tDf6xh5xg3) to strike with an early d4-d5, Taking the e4-square from the black
in which case the rook should really be on d 1 knights.
105
15 ...'ii'e8! If Black wishes to avoid the disruption in

A nice move with quite a few little tactical his development caused when White plays
points. Black's fll"St threat is against the pawn .th4xf6 and drags the bishop on e7 to f6
on a4. instead of a knight, then Black can try this
16 'fie 2 move order, which has occasionally been
16 b3 ttJfe4! 17 ttJxe4 .txh4 18 ttJd6 'iVe7 adopted by Gany Kasparov.
19 ttJxb7 'ii'xb7 equalises according to 9 cxd5!
Vaganian. Obviously, if Black can play 9....tb7 then
16 ....:te8 17 i.a2 ~d5!? 18 i.xd5 he will have successfully avoided White's 9
18 .tb1 g6 seems fme for Black. .txf6 attempts so this is the most critical
18 ... i.xh4 move. If Black tries 9...ttJxd5 then after 10
With an equal position according to .i.xe7 .xe7 11 ttJxd5 exdS, White has a
Vaganian. White has to be a little careful now superior version of the 8 cxdS line as Black
- with his queenside pawn on a4, he can has played his knight to d7 early, before
easily become worse if Black's knight gets developing the light-squared bishop to e6.
amongst his queenside. Consequently, White forces a pawn to dS
19 i.xb7 ~xb7 20 'ii'b3?! lDa5! 21 'ii'd1 and thus blocks the light-squared bishop on
i.f6 22 'i'e2 'fId7 23 .:tfd1 'i'b7 24 lDe4 b7 along the long diagonal. You feel that this
i.e 7 25 'i'f1 .:tfd8 26 .:txc8 :xe8 27 :e 1 should be a bit better for White, but unlike
%-% Yusupov-Kamsky in the introduction to
In the final game of this chapter, we shall Chapter 5, White's bishop is rather passive
examine a different idea that Black can play here on e2 rather than d3, and this seems to
after 8 .te2. Note that an early ...dSxc4 will make some difference.
transpose into the 8 .td3 lines examined in 9 ... exd5 10 0-0 i.b7 11 .:te1 c5 12 'i'a4
Games 48-50. a6 13 dxe5 bxc5 14 :fd1 'i'b6 15 'i'b3
'it'a7!
Game 57 It is important for Black to keep the
Topalov-Kasparov queens on - Black's queen is the glue of his
Sofia (rapidplay match) 1998 position, covering all the little holes that
would usually be visible. 15 ...'ii'xb3 16 axb3
1 ~f3 d5 2 d4 lDf6 3 e4 e6 4 lDe3 i.e 7 l:tfd8 17 ttJe1! ttJb6 18 .i.f3 l:td7 19 ttJd3 gS
5 i.g5 h6 6 i.h4 0-0 7 e3 b6 8 i.e2 20 .i.g3 l:tc8 21 ttJeS l:tdd8 22 ttJc4 was
lDbd7!? clearly better for White in Lputian-Dorfman,
106
Tashkent 1984. 16 .li.g3 .l:.ad8 17 .l:.d2

Karpov-Kasparov, World Championship
1984, saw 17ltJe1 (Geller claims an edge for
White here). After 17...d4 18 exd4 cxd4 19
ltJa4 :c8 20 :xc8 :xc8 21 i.c4 :f8 22
\iVd3t White was indeed definitely more
comfortable.
17 .. JUe8 18 'iWd1 .li.f8 19 i..h4 1i'a8 20
liJe1 i..e7 21 .li.g3 liJf8 22 ~f3 ltJe6
Black has reorganised here and stands
well.
23 .li.h4 d4 24 exd4 exd4 25 ltJa4 liJf4
26 liJe5 i.xe5 27 .li.xf6 d3 28 i.xd3
i.xf3 29 gxf3 .:td5 30 i.h4 i.b4 31 .:te3
i.xe3 32 bxe3 .:ted8 0-1
107
Summary
These are very interesting lines. Black must be careful since there are many move-order tricks
and little traps. Personally, I prefer Beliavsky's 'Nirnzo-Indian' systems to Vaganian's attempts
to hold the centre, but this is more a matter of taste than anything concrete.
1 d4 d5 2 c4 e6 3lLic3 lLif6 4liJf3 i.e7 5 i.g5 h6 6 i.h4 0-0 7 e3 b6
8 i..d3
8 ii.e2
8...dxc4 9 .txc4 ~b7 - see 8 J..d3 dxc4 9 ii.xc4 J..b7 below
8...11Jbd7 - Game 57
8 .:lc1 ~b7 9 ~e2 - Game 56
8 .....i.b7
8...dxc4 9 ~xc4 ~b7 10 0-0 (D)
10...ltJe4
11 ~xe7 - Game 48
11ltJxe4 - Game 49
10...ltJbd7 - Game 50
9 0-0 lLibd7 (D) 10 .e2
10 ii.g3 c5 11 cxdS - Game 51
10 lk1 dxc4 11 ~xc4 - Game 56 (by transposition}
10 .•• c5
10...ltJe4 11 ii.g3 ltJdf6 - Game 55
11 i..g3 (D) lLle4
11...cxd4 - Game 52
12 cxd5 cxd5 13 .:tad1
13...ltJxg3 - Game 53
13 ...lLidf6 - Game 54
100-0 9 ... ltJbd7 11 ~g3
108
CHAPTER SEVEN
Exchange Variation:
Systems with ttJf3
1bis chapter is extremely important for those QIestion 4. So what is going on here?
wishing to play the Black side of the QGD. Arl3ZW'4. First of all, let's look at the move
QIestion 1. Oh, no! I've spent all this time in simple positional tenns. With c4xdS,
sorting out the theory and now you tell me White makes several changes to the position:
that the difficult work is still to come! 1. He releases the central tension, which
Arl3ZW' 1. Not exactly. 1bis chapter obviously reduces the breadth of his options.
doesn't involve any effort of memory. What 2. He opens up two files: the c-file for
it does require is understanding! himself and the e-file for Black.
QIestion 2. Oh dear! Why? 3. He opens the c8-h3 diagonal for the
Arl3ZW' 2. Well in this chapter, we examine black bishop on c8, consequently solving
the most typical structure in the QGD: a Black's general opening problem - that of
structure that Black offers his opponent with finding an active diagonal or post for his
his second move. light-squared bishop - without even first
QIestion 3. And this structure is ... ? waiting for a weakness on the queenside such
Arl3ZW' 3. The Exchange structure. In its as ...b7-b6 as in the Tartakower system.
simplest fonn, we can see it after 1 d4 d5 2 QIestion 5. This seems very odd. Why
c4 e6 3 cxd5 exd5. would White want to solve his opponent's
development problem in this way?
An.war 5. There are several ideas behind
White's 'madness'!
1. Firstly, c4xdS fixes the central pawn
structure - it forces a black pawn to the ds-
square extremely early.
QIestion 6. Umm, yes ... and .. ?
Arl3ZW' 6. Well, this has a very profound
effect on Black's central break: ...c7-cS.
Qtestion 7. What do you mean?
An.war 7. Well, if Black now breaks with
...c7-cS, White replies d4xcS and gives Black
an isolated IQP. Put simply, Black's typical
109
central break now inevitably leads to a central matter of move orders.

weakness. For example, in the Orthodox Qlestion 11. Oh dear!
system, Black often played ...dSxc4 before Answer 11. Well, there are many move-
either his ...c6-c5 or ... e6-e5 breaks just to order points to this structure so we shall deal
avoid this problem, but here Black has no with all of them at once.
escape from this scenario. Thus by giving up The first point is that White can only fight
some of his own central flexibility, White for an advantage if he prevents the light-
takes the joy out of his opponent's central squared bishop from coming to f5, or
break. alternatively if he extracts such concessions
2. The second point revolves around the that getting the bishop to f5 is a self-
c-fue. By opening this c-ftle, White uncovers defeating proposition for Black. Thus, when
an avenue which his heavy pieces can use to we talk about playing the Exchange variation,
attack Black's position. Thus by targeting the we have this as a prerequisite goal before
c-pawn with :la1-c1 and 'iVd1-c2, White can entering this structure.
force a reaction on the queenside from Black. If White wishes to force the Exchange
3. The restriction of Black's light-squared variation, he should do so via the 1 d4 dS 2
bishop. White's idea is to prevent his c4 e6 3 liJc3 liJf6 (or 3.....tel) 4 cxdS exdS
opponent from activating his light-squared move order. However, if he wishes to do
bishop on any useful square along the c8-h3 this, he will also have to make certain other
diagonal and thus to deny him the benefits of choices against other lines.
playing ... e6xdS at such an early stage. Qiestion 12. What do you mean?
Qiestion 8. How can White do that? Answer 12. Consider this. If Black
An,grar 8. By playing either his light- introduces the QGD via the move order 1 d4
squared bishop or his queen to the h7-b1 liJf6 2 c4 e6, how does White play?
diagonal to deprive the bishop of its only Qlestion 13. Well, 3 ltJc3 I suppose and
really active post: the f5-square. Note that c2 after 3...dS, then ...
is an excellent square for the white queen as Answer 13. Yes, but do you want to play
it both deprives the light-squared bishop of against the Nirnzo-Indian if Black plays
the f5-square and prepares to line up with the 3... if..b4? If not, then you can't play 3 liJc3.
a rook on c1 against the pawn on c7. Qiestion 14. And if I play 3liJf3?
Qiestion 9. So to swnmarise: Answer 14. Then Black plays 3...d5 and
A n,grar 9. White's three aims are: from this position, you cannot force a real
1. To deter Black from canying out his Exchange variation as we shall see.
central ...c7-c5 break by fixing the central Then there is another point. If after 1 d4
structure at an early stage. dS 2 c4 e6 3 liJc3, Black plays 3... c6, are you
2. To give his major pieces a chance to get prepared to play the Semi-Slav type positions
at Black's position along the c-file. after 4 liJf3 dxc4 or 4 e3 f5f?, or to spend a
3. To nullify the benefits to Black of an lifetime learning the 4 e4 Marshall Gambit?
early ...e6xdS by depriving the light-squared Or do you, like Kramnik, wish to play more
bishop of any access to the h7-b 1 diagonal. quietly against such lines and play something
Qtestion 10. One thing puzzles me - how like 3liJf3 c6 4 'iVc2 for example?
on earth can White really stop his opponent Qtestion 15. Aha, so what you are saying
from getting his bishop to f5 - it seems that IS ...
Black will always have time? Answer 15. If you want to play the
An,grar 10. Well, that's a vety important Exchange variation against the QGD at all
point: we now come to the all-important times, then you must be prepared to play the
110
Exchange Variation: Systems with tDf3
NlIllZD Indian and you must accept that your 7· .. 9 6 !

options against the Semi-Slav hybrid systems QIestion 18. This looks rather weakening.
are more limited and perhaps sharper than Answer 18. It does weaken the kingside
may be ideal. Of course, if you are sure that dark squares it is true, but it also forces
your opponent only plays the QGD, then White to take drastic action in order to
you can risk playing an early ltJc3 but if not.. .! prevent his opponent from achieving his
Q5tion 16. Okay, but why is it that White plan of ... ~fs with a gain of tempo on the
can't force the exchange schemes after 1 d4 white queen.
ltJf6 2 c4 e6 3 ltJf3 dS? For example if he 8 e4!?
goes 4 ttJc3 j.e7 5 cxdS exdS 6 j.g5? 8 e3 j.fS! is nothing for White, as we shall
Anszre-16. Well, let's have a look ... see in Games 73 and 74.
8 ...dxe4!
Game 58 Not 8...ttJxe4? 9 j.xe7 and now 9...'tixe7,
Beim-Korneev as in Karpov-Yusupov, USSR 1988, is forced
Frankfurt 1997 as 9.. :i'xe71oses to 10 lDxdS!
9 J.xf6 J.xf6 10 'ili'xe4+
1 d4 ltJf6 2 c4 e6 3 ltJf3 d5 4 ltJc3 i.e7
5 cxd5 exd5 6 J.g5
10 ...~f8!?
The simplest, though 10..:i'e7 is also
6 ... c6 playable. After 11 .ic4 0-0 12 0-0 (12 'i'xe7
I have tweaked the actual move order here j.xe7 13 0-0 .if5 14 !tfel j.f6 15 h3 h5 16
to fit into our theme (the players actually lDeS lDd7 17 lDe4 j.xeS 18 dxeS lDxeS 19
reached this line via a Semi-Slav hybrid). lDf6+ <3tg7 20 lDxhs+ gxh5 21 !txe5 <3tf6
Q5tion 17. Wasn't 6....if5 possible? was quite equal in Murshed-Serper, Dhaka
Anszre- 17. No, because 7 j.xf6 j.xf6 8 1995) 12....ifs 13 'i'f4 (13 'i'xe7 .ixe7
'Wb3 forks b7 and d5 and nets White a pawn. would transpose to the previous note)
The text protects the dS-pawn and intends to 13...'i'b4 144Je5 .ixe5 15 dxe5 j.e6 16lDe4
solve all Black's problems on the next move 'i'xc4 17 !tad 'i'b5 18 a4 'i'aS 19 lDf6+
with ... ~c8-f5. <3tg7 20 lDhs+ gxh5 21 ..gS+ led to a draw
7W'c2 by perpetual in Gulko-Yusupov, Munich
As 7 e3 (mtending 8 ~d3) 7... j.fS! is fme 1990.
for Black (see Game 75), White uses his 11 J.c4 ~g7 12 0-0 :e8 13 'ili'f4 J.e6 14
queen to prevent the inunediate development .i.xe6 :xe6 15 :fe1 'ii'd6 16 'ii'xd6 l::txd6
of the bishop to f5. 17 :e8!?
111
17 ttJe4 :'d8 18 ttJxf6 <it>xf6 19 :'e4 ltJa6 course it is vezy useful for Black to have the
20 :'ael :'d7 21 ttJe5 :'c7 was equal in pawn on h6 as White no longer has the
Lastin-Komeev, Russian Championship chance of gaining any tempi against the pawn
1996. onh7.
17 ... %;Id8 18 %;Ixd8 .txd8 19 :e 1 tiJd7 20 QIestion 21. And 7...exdS 8 e3 ~f5 ...
g4 h6 21 ~g2 ~f8 22 l:e2 tiJf6 23 h3 A11$lW" 21. ...Fails once again to 9 ..txf6
.tb6 24 tiJa4 ':e8 25 tiJxb6 axb6 26 ~g3 ..txf6 101Vb3!
b5 27 h4 b4 28 g5 hxg5 29 hxg5 tiJd5 8 .txe7 'ii'xe7 9 'ii'b3
30 a3 bxa3 31 bxa3 :a8 32 ':b2 b5 33 The natural 9 e4 gives Black rapid
':b3 tiJb6 34 tiJe5 tiJc4 35 tiJxc4 bxc4 36 counterplay against the d4-pawn after
':c3 ~e7 37 ':xc4 ':xa3+ 38 ~f4 ~d6 9... t'Dxc3 10 bxc3 c5 11 i.e2 :d8 followed
39 :b4 :a5 40 f3 :f5+ 41 ~g4 f6 42 by ...c5xd4 and ... t'Db8-c6. The exotic 9 g3 led
gxf6 ':xf6 43 ':b8 g5 44 ':b3 ~d5 45 to a draw in Miladinovic-Yusupov, Elista
~xg5 ':f8 46 f4 ~xd4 % -% Olympiad 1998, after 9... t'Dxc3 10 bxc3 c5 11
QIestion 19. Okay, maybe White took on i.g2 t'Dc6 12 0-0 :d8 13 'ii'a4.
dS too soon. What about after 5 i.g5 h6 6 9 ...tiJd7! 10 ':c1
i.h4 0-0 and only now 7 cxdS? Not 10 t'DxdS exdS 11 'ii'xdS 'ii'b4+!
Armrer 19. That's not a bad question! 10... tLlxc3 11 'ifxc3 b6!
Game 59
Krasenkov-Beliavsky
Yerevan Olympiad 1996
1 d4 tiJf6 2 c4 e6 3 tiJf3 d5 4 tiJc3 .te7
5 Jtg5 h6 6 .th4 0-0 7 cxd5
12 e3
12 'i'xc7 i.a6! 13 t'De5 :fc8 14 ~c6 'i'g5!
15 f4 :xc7 16 fxg5 :ac8 17 t'De7+ <i!i>f8 wins
for Black according to Mikhail Gurevich.
12 ....tb7 13 b4
13 i.b5 c6 14 i.e2 c5 15 0-0 :fc8 16
:fd1 cxd4 17 'i'xd4 J:.xc118 :xc1 :c8 was
7 ...tiJxd5! just equal in MGurevich-Marciano, French
QIestion 20. Aha! Team Championship 1995.
Armrer 20. We've seen this vel}' natural 13,..':ac8 14 Jte2 e5 15 dxe5 c5 16 b5
idea in both the T artakower and Orthodox ':fe8 17 0-0 tiJxe5 18 tiJxe5 'ifxe5 19
chapters! Black uses the opposition of the fixe5 ':xe5 20 ':fd1 ~f8 21 .tc4 ':c7 22
dark-squared bishops to exchange minor ':c3 .tc8 23 f3 .te6 24 ~f2 .txc4 ~ - ~
pieces and greatly free his position. Black can Past this point of course, it gets rather
also play this after 5...0-0 6 cxdS, though of difficult for White to force an exchange line:
112
Exchange Variation: ::>ystems Wiln '1.JTJ
for example, after S...h6 6 it.h40-0 7 e3 b6 shall move on to plans with the knight on e2.
we are in a Tartakower or after 7...ttJe4 we The positions in the rest of this chapter
have a Lasker variation. generally arise from three different openings:
Qiestion 22. But wait a minute: I'm going the Orthodox QGD, the 1 d4 d5 2 c4 e6 3
to play the 1 d4 ttJf6 2 c4 e6 move order for lDc3 ttJf6 4 cxdS move order and the
Black, play the Nunzo-Indian against 3 ttJc3 Cambridge Springs (1 d4 d5 2 c4 e6 3 lDc3
and only play the QGD if White goes 3 lDf3, lDf6 4ll.gSlDbd7 slDf3 c6).
when I go for the T artakower. Why do I
need to look at this chapter? Game 60
Amzrer 22. This is actually a very Van der Sterren-L.Hansen
important part of opening preparation. As a Wijk aan Zee 1995
young International Master, I used to devote
much of my time analysing the very sharpest 1 d4 d5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3 e6 5
variations, hoping all the time that I would J.g5 ltJbd7 6 cxd5 exd5 7 e3 J.e7 8 J.d3
get the chance to engage my opponent in 0-0 9 Wc2 :e8 10 0-0 tiJf8
sharp variations that I had prepared at home.
After a while, I began to notice something: I
was losing lots of games in 'unimportant
variations'. My opponents rarely seemed to
'take me on' but instead played quiet
variations, just aiming for a typical position. I
hadn't looked at these nonnal positions,
hadn't thought about them, and didn't
understand very much about them. This
meant that even good versions of the
theoretical line ended badly for me because I
didn't understand why they were good, what
exactly made the difference, and what I could
aim for in this position that I couldn't in Via a Semi-Slav, then a Cambridge
others. It's all part of your education in an Springs, we reach the main position of this
opening - knowing the typical endings, the line.
typical structures from an opening so that Qiestion 23. Tell me what is going on!
when your knowledge of previously played Ansrrer 23. The fIrst things to look for
games runs out, you don't lose, or just have when trying to assess a position are the pawn
to offer a draw, but you can play on to win breaks.
because you understand the simple positions Qiestion 24. Why?
better than your opponent. So, no you may Ansrw- 24. In such a position where the
never get this exactly, but you will get structures of both sides are so solid and
something like it as soon as you play the flawless, pawn breaks are the key method of
QGD, and you will play that position ten adding dynamism to the position. By
times better if your all-round education in the engaging the opponent's pawn structure in
opening is good hand-to-hand combat, you hope to soften up
We shall now consider the Exchange his position for a later assault by the big guns.
variation in some detail. This chapter deals Qiestion 25. So White has the e3-e4 break
with systems in which White places his king's and Black has the ... c6-cS break. Neither of
knight on f3, while in the next chapter we them look great though.
113
An.9lm" 25. That is quite true. Both these bishops.

breaks involve structural weakness on both An.mer 29. This is Black's general desire,
sides, in the fonn of an IQP. It is clear that of course, in the QGD from move two
for either of these to work, there will have to onwards! By exchanging off the bishop,
be a specific dynamic or tactical reason. But Black frees his queenside, while exchanging
in fact, both sides have another Wirf to the bishop, which is restricted by his central
engage the opponent's position. pawn chain.
Qiestion 26. You mean b2-b4-b5 and ...£7- QIestion 30. It doesn't seem very easy
f5-f4? though!
A n.9lm" 26. Yes. As this involves a lesser An.mer 30. Black has a typical and cunning
force of pawns attacking a greater force manoeuvre in ...g7-g6 followed by ...l£lfS-e6-
(White's three queenside pawns against g7 and then ...il.cS-f5, as we shall see in the
Black's four, for example) this is known as game.
the 'minority attack'. In this case, it is clear QIestion 31. Neat! And the dark-squared
that White's b2-b4-b5 assault is much easier bishops?
to achieve than Black's .. .f7-f5-f4; Anscrer 31. This is much less of a heartfelt
consequencly it is clear that the dynamism in desire from Black's point of view. It is simply
this position lies mainly with White. White's that the white bishop on g5 is a point of
idea is that after b5xc6, ... b7xc6, the black pressure on Black's position - for example,
structure is greacly weakened: there is a 11...l£le6 is impossible due to 12 il.xf6 il.xf6
backward pawn on the c-ftle and the a-pawn 13 il.xh7+ - and such points of pressure
has been isolated. have to be removed or they will become
thorns in Black's side sooner or later. Black
has several ways of attempting this:
1. The most obvious W3:f is to move the
knight on f6 to offer the exchange of
bishops. Neither h5 nor g4 are great squares:
the knight will have to retum to f6 which
makes these moves slight time-wasters, so
this leaves two possibilities:
a) ...l£lf6-e4. This is the most natural
move. Black uses the semi-open e-file to
establish his knight on e4 while offering the
exchange of bishops. To this White has two
main replies:
Qiestion 27. And Black must aim for .. J7- a1) il.gS-f4, avoiding the exchange of
f5-f4? pieces and putting the question to Black's
An.mer 27. No, it is just too hard to knight on e4 and
achieve. Black's energy is taken up with two a2) il.gSxe7 followed by .id3xe4 ...d5xe4,
matters: dealing with White's plan and l£lf3-d2 and then central action with either
freeing his own position by exchanges. d4-dS or f2-f3.
Qiestion 28. So, what does Black want to Note that White often plays il.g5xf6 (as in
exchange, and how? the T artakower variation) just to prevent the
Amut'r 28. Black's exchanging strategy knight from coming to e4. Although this
concentrates on both bishops. gives up the bishop pair, it drags the bishop
Qiestion 29. So first of all, the light-squared away from the fS-a3 diagonal where it
114
Exchange Variation: Systems with ti1f3
belongs (on f6, it merely bites against the First of all, Black can play ... a7-as to hold
granite on d4) and prevents Black from using back b2-b4.
the outpost on e4 as his other knight is too
far away.
b) ...t2Jf6-d7. Black prepares to transfer the
knight to b6 with tempo. From here, it
defends Black's queenside whilst eyeing the
c4-square (which White weakens when he
plays b2-b4).
While it is correct to concentrate on
White's queenside intentions, we should also
mention his other key resource - the knight
outpost on eS. Although he cannot
immediately make use of it (11 ttJe5 is well
met by 11...t2Jg4! here, exchanging pieces) it
is a recurrent motif in all of White's plans. White reacts by playing a2-a3 and then b2-
11 :ab1 b4 anyway. Black now has two possibilities:
to take it or to leave it.
Qiestion 33. So what is the idea if Black
takes it?
This is White's most direct idea His break

is b2-b4-bS so White just supports it! Other
possibilities which we will consider are 11 h3
(Games 62-65), 11 l:ae1 (Game 66) and 11 An..mu 33. By taking off the a-pawns,
a3 (Game 71). Black hopes to minimise the chance of heavy
11 ... g6 losses on the queenside. The b5xc6 exchange
This is actually a very crucial moment in will no longer create an isolated pawn on the
the game. a-file, so that is one pawn less to defend and
Qiestion 32. Why? one less to lose if things go wrong! Also
An..mu 32. With this move, Black takes a Black gains the a-fIle (which White
certain approach to White's queenside abandoned with l:a1-b1 to support his b-
operations. As I was taught by Mark pawn), along which he can hope to stir up
Dvoretsky, Black has a multitude of ways of some trouble.
dealing with White's queenside play and now Qiestion 34. And what are the drawbacks
is as good a time as any to explain them to to this idea?
you. Ansuer 34. As Black's central and
11~
queenside structure is on light squares, Black Another idea for Black is to play ...al-a6
inevitably has problems on the dark squares so that White has to play a2-a4 to force
in any minority assault. By removing the through b4-bS
pawn on al, Black removes his pawn
protection of the b6-square, which is now a
very useful attacking square for White. After
b2-b4-bSxc6
and now:
1. Black plays ...b7-bS.
White can use the b6-square for his rook,

for example, to attack the weak c6-pawn.
Note that this leads to the same structure as
after ... al-a6, a2-a4 and b2-b4-bS ...a6xbS,
a4xbS.
Qiestion 35. But why play ...al-aS and then
not even take the opportunity to capture the
h4-pawn?
An.9Zre" 35. This is based on a really
cunning idea!
This prevents White's b4-b5 break at the

cost of an exceptionally weak c6-pawn. This
idea is best implemented when a white knight
does not have access to the eS-square, or
when a black knight is ready to jump into the
c4-square. We saw a successful example of
this in the Tartakower in the notes to the
Game 41.
2. Black waits for b2-b4-bs and then
2a Black takes on b5, simply transposing
to the ...a5xb4lines.
2b. Black plays ...a6-aS.
By luring the pawn to a3, Black ensures
that he delays b4-bS as the pawn on a3 will
hang! Qiestion 36. This makes no sense! Why
116
Exchange Variation: Systems with li::Jf3
play ..:a7-a6 and then ... a6-a5? closed: White has only opened the c-file, and
his queen is there, whereas Black is ready to
bring a rook over to challenge it.
Note that Black must follow up this idea
with ...ttJb6 ideally, or ...b7-b6. If not, then
White may play bS-b6 and ttJc3-bS and
suddenly the problems are all Black's.
In particular against ... a7-a6, but also
against .. :a7-as, White has another idea apart
from a2-a4: Botvinnik's recommendation of
playing a knight to cS first before following
up with a2-a4 and b4-bs.
Answer 36. Black's idea is that by luring the

white pawn to a4, he has taken that square
away from the white knight on c3 so that the
manoeuvre ttJa4-cS is now impossible.
Typically, this idea is seen when White has
already taken the knight on f6 with his
bishop so that Black has the only dark-
squared bishop on the board Consequently,
he can even block the b-fUe sometimes with
...i..e7-b4.
c) Black plays ... c6xbS, a4xbS ... a6-aS often
followed by a quick ... ~b6. Here, it is exceptionally hard to dislodge a
knight from cS as ...b7-b6 loses a pawn to
~cSxa6.
Black's final idea is to play to meet b4-b5
with ... c6-cS.
Qiestion 37. Doesn't this just isolate the

black d-pawn?
Answer 37. It does, but there are several
factors in Black's favour to compensate:
1. The knight on b6 has a superb outpost Qlestion 38. Doesn't this just allow White
on c4 to aim for. to isolate the black d-pawn with d4xcS?
2. The queenside is kept reasonably Anmer 38. Yes, but these isolated pawn
117
positions where White has a pawn on b5 are ~e5!

not really so great for him. I refer you for One of the drawbacks of taking the
example to the game Gretarsson-Yusupov in queen's knight from d7 to g7 is that it gives
the notes to Game 38. After ...':c8xc5 the white knight the e5-square for free.
18 ...:tc8 19 l:b7 'fid6 20 ltJa4!
the knight on c3 is vel}'" uncomfortable on

the open c-flie: it has to protect the pawn on This excellent move puts Black in a great
b5, but it is unsettled on c3 - it needs the deal of trouble. Here, it is the knight that
support of a pawn on b2! heads for b6, making use of Black's weak
We will see practical examples of all of queenside dark squares.
these ideas later in this chapter, but whenever 20 ...il.xd3
you are faced by a minority attack, these are Here 20...lDe4 was a trickier defensive
your options! chance. We shall examine this position in
In this case, I feel that Black chose the some detail to tl}'" to get an idea of Black's
wrong one. 11 .:tab 1 attempts to implement defensive resources in a bad minority attack
White's plan in the fastest manner possible posmon.
without any subtlety or preventive moves. By Black's position is obviously under a great
ignoring White's idea and concentrating on deal of pressure: White's minority attack has
his slow exchanging plans, Black invites 'diluted' his opponent's structure so that all
pressure on to himself. I would prefer ll...aS that remains of Black's once solid queenside
12 a3 .i.d6! Here, as in Altennan-Gabriel, is now the sickly pawn on c6. White has
Bad Homburg 1996. Black has the idea of occupied the outpost on e5 and has a great
...lDf8-g6 and then ...h7-h6 to force White to deal of pressure along the important h4-d8
play .tg5xf6, but without diverting the black diagonal.
dark-squared bishop from the f8-a3 diagonal, 20...lDe4 seeks to gain some countetplay
and after 13 .:tfe 1 Oooking for e3-e4 to try to by giving up the sickly c6-pawn. After 21
exploit the pin on the knight on f6 by the .txe7 :Xe7 22 ':xe7 'iixe7, 23 :c1 keeps
bishop on g5) then 13 ....tg4! 14 ltJd2 i.h5 the pressure and is not pleasant for Black,
15 lDfl .tg6 16 b4 axb4 17 axb4 h6 18 .txf6 but if White ambitiously grabs the pawn with
'iixf6 19 b5 .txd3 20 'i'xd3 lDg6 21 bxc6 23lDxc6
the game was agreed drawn. We shall see
11....~Dg6!? in the next game.
12 b4 ltJe6 13 il.h4 a6 14 a4 ltJg7 15 b5 he exposes himself to a nasty pin along
axb5 16 axb5 .i.f5 17 bxc6 bxc6 18 the c-flle. Black now has two possibilities:
118
Exchange Variation: Systems with ti::Jf3
23 ..."f6 and 23 .. :ifb7. own position!

21 'iixd3 ti::Je4 22 .ixe7
22 l:d7! 'ife6 23 .txe7 l:xe7 24 'ifa6! is
given by Van der Sterren as winning, but I
think that Black can still fight with 24...l:ee8
25 :xf7 :a8 26 'iib7ltJf5!?, when the knight
on a4 has a few problems.
22 ...:xe7 23 :xe7 'iixe7 24 'ila6 'ile8
25liJb6 :c7 26 ':c1 liJe6 27 f3?
27 'iWa4! is again a good suggestion of Van
der Sterren, though after 27...'iWb8 28 "'as!
"'xaS 29 ~xa8 l:al 30 ttJb6 :a6 31 ~bd7
the game is still not completely over. The text
allows the game to fizzle out.
a) 23 ......f6 24 .tb5! (24 l:el .td7!? 27 ...liJf6 28 tDa8 .:tc8 29 tDb6 :c7 30
[24... ~xf2! 25 .txf5 ~xf5 26 ~b6 ~xe3! 27 tDa4 liJd7! 31 ti::Jxd7 'ii'xd7 32 tDc5 liJxc5
~e7+ ~g7 28 "'xc8 ~d3!? is also possible] 33 :xc5 'ile7 34 'iid3 %-%
25 .tb5 .txc6 26 .txc6 :Xc6 27 "'xc6 An interesting game that shows the
"'xf2+ 28 ~h1 "'xe3 29 "'c8+ ~e8 gives defensive resources available to Black.
Black good play, as 30 "'xe8+ ~g7 31 "'e5
~h6 is not what White is looking for! 30 h3 Game 61
~g7! gives Black reasonable chances) Dydyshko-K veinys
24... ~d7 25 ~b6! (the key move) 25...l:xc6 Moscow Olympiad 1994
(25 ....txc6 26 ~xd5 "'e6 27 .txc6 l:xc6 28
'iWxc6 'iWxc6 29 ~e7+ wins) 26 .txc6 'iWxc6 1 d4 tDf6 2 c4 e6 3 tDf3 d5 4 lZ)c3 J..e 7
27 "'xc6 ~xc6 28 l:el .tb7 29 ~xd5! gives 5 cxd5 exd5 6 .ig5 0-0 7 e3 :e8 8 .id3
White a clear advantage. ti::Jbd7 9 0-0 tDf8 10 'iic2 tDg6 11 :ab1
b) 23 ...'iib7 is logical to prevent .td3-b5 c6 12 b4 .id6
ideas. After 24 :el, 24... ttJd6! is best as
24... ~h8 (intending ....tf5-d7) is met by 25
'ib2! 'iWxb2 (25 ...:Xc6 26 "'xb7 l:xel + 27
.tfl ~d2 28 'iib8+ .tc8 29 g3 ~xfl 30
~g2! intending ~a4-c5) 26 ~xb2 .td7 27
.txe4 dxe4 28 d5! with a clear advantage.
After 24.. .l'tJd6 25 .txf5 ttJgxf5, 26 "'c5 is
met by 26...~c4! 27 ~a5 'iWa8! so 26 g4
seems best, but after 26... ~e7 (26 ... ~h4 27
"'c5! "'d7 28 h3 "'e6 29 ttJc3!) 27 ttJxe7+
"'xe7 28 "'d1 Black is clearly worse, but not
yet lost.
This is a common theme in this line, and
in chess in general in fact - by losing a sickly By a strange inversion of moves, we have
pawn, you often gain a swprising amount of transposed into the position after 11 :ab 1
activity - somehow, the very square that the ~g6 12 b4 .td6 in the previous game.
pawn stood on seems 'jinxed' and as soon as Against 11 l:ab 1, I think the ....te7-d6 idea is
White occupies it, the energy drains from his the best way, though throwing in ... a7-as first
119
seems very logical. 0-0 9 'ile2 :e8 10 0-0 t2Jf8 11 h3

13 b5!? h6! 14.txf6
14 s..xh6 gxh6 IS .txg6 fxg6 16 'iixg6+
~h8 17 "xh6+ ~7 leads to nothing
(Dydyshko).
14...'fixf6 15 e4!
15 .:Ifel .tg4 16 ~d2 ~h4 gives Black a
powerful initiative according to Dydyshko.
15 ...t2Jf4! 16 e5 'fie6 17 t2Je1
17 exd6?! is dubious (Dydyshko) due to
17.....g4 18 ~el ~xe119 f3 .:Ixf1+ and now
20 .txfl "d7! 21 l:.e1 ~e6! is safe for Black.
17 ....tf8 18 t2Je2 t2Jxd3 19 t2Jxd3 exb5!?
20 t2Jdf4!
Qiestion39. What on earth is this for?
ArJSlW' 39. This favourite move of
Karpov's has many points:
1. It takes control of the g4-square. This
has two benefits:
la It prevents any manoeuvre such as
....tc8-g4-hS-g6.
lb. It allows White to make use of his
central outpost with lLlf3-eS as Black no
longer has the riposte ...lLlf6-g4.
2. It provides a retreat square on h2 if
Black chases the white bishop from gS.
3. It keeps White's options extremely
20 .• :i'g4? flexible. For example, since White does not
20...'iic6!? was stronger according to commit the rook to b 1, he is always ready to
Dydyshko with a murky position after 21 switch to a central thrust plan with ~al-e 1
'iWb3 b4 22 ~fc11fa6 23 ~c7 .te6 24 ~bc1. and e3-e4 or lLlf3-e5. Of course, it doesn't
21 :xb5 b6 22 h3 'fig5 23 :b3 .tf5 24 have quite as much drive on the queenside as
'i'e6 .te4 25 :g3 :ee8 26 'i'b5 'fid8 27 11 ~ab1.
t2Je3 .tf5 28 t2Jexd5 'ii'h4 29 'fib3 ~h8 11 ... g6
30 t2Je3 "xf4 31 'i'xf7! 'ii'xd4 32 t2Jxf5 As White's queenside play is a move
'i'xe5 33 t2Jxh6 J.e5 34 t2Jf5 :e6 35 slower, the exchanging plan is much more
:xg7 :h6 36 t2Jxh6 'i'xg7 37 'fih5 :f8 tempting for Black than in the 11 ~ab 1 line.
38 t2Jg4+ 'ii'h7 39 'fie5+ 'i'g7 40 'fih5+ The alternatives 11.. ..te6 and 11...lLle4 are
'ii'h71-0 considered in the Games 64 and 65.
12 :ab1
Game 62 White can also try 12 .txf6, as in Game
Ruban-Panchenko 63.
Elista 1994 12... t2Je6
12...a5 13 a3 lLle6 leads to very similar
1 d4 t2Jf6 2 e4 e6 3 t2Jf3 d5 4 t2Je3 e6 5 play.
.tg5 t2Jbd7 6 exd5 exd5 7 e3 .te7 8 .td3 13 .th6 t2Jg7 14 b4 a6 15 a4 .i.f5! 16
120
Exchange Variation: Systems with t'iJf3
~xg7 l:txc3 'iib6!! 29 lDe5 [29 g3 'iib2!] 29 ...l:tal+

16ltJeS is the subject of Game 70. 30 ~h2 'ii'xc6 31 l:txc6 f6! 32 l:tb6 J.c7 33
16 ... ~xd3 17 'ifxd3 ~xg7 18 b5 axb5 l:tc6 :ta7! wins) 29...:tb2! looks very
19 axb5 :a3 20 bxc6! bxc6 constricted for White. Alternatively, 26 'ii'b3
20...'ii'aS 21 IUc1 ~b4 fails to 22 cxb7 ~d6 27 l:txc6 ltJe4 28 :tc2 :tal 29 g3 ltJg5!
:tb8 23 ltJe5 according to Ruban. is also very awkward for White.
21 'ifc2 'ifa5 22 l:fc1 It is clear that in general the black bishop
belongs on the active d6-square and Black
should always seriously consider the
possibility of placing it there. In the game,
Black started to lose the thread a little.
22 .•.i.b4
A rather routine attack on a knight which
wishes to move anyway.
23%:.b3
23 ltJe2! was better according to Ruban,
when 23 ...l:a2 24 :tb2 Jhb2 25 'iixb2 .ta3
26 'iia2 leads nowhere for Black.
23 ...l:c8 24 %:.xa3 'ii'xa3 25 4:Jb1 'ifa6 26
4:Je5! ~d6 27 4:Jd3
QiRst:ion 40. How is this for Black?
Anszrer 40. This type of position is quite
critical because you can imagine it arising
from so many different minority attacks. We
shall thus analyse it carefully.
QiRst:ion 41. Can't Black just break free
with 22 ...cS?
Anszrer 41. No, because 23 :tb5 (Ruban's
23 dxcS!? 'iixcs 24 'iid2 is also interesting,
threatening ltJc3-b5. After 24...'iiaS 25 :tb5
'iia6 26 'ib2! l:kS 27ltJd4 White stands very
nicely indeed.) 23 ...'iia6 24 'iib2 is very nice
for White and transposes above after
24...cxd4 25 liJxd4. Perhaps 24...c4!? 25 White has an unpleasant grip on the
ltJxdS 1:ta2 26 'iblltJxdS 271:txdS c3!? gives position: he has neutralised Black's a-fIle play
Black some counterplay. I prefer the simple and now has the ...c6-c5 break under wraps
22....td6, activating the bishop by covering as well. It is a slight but very persistent
the e5-square and allowing Black to challenge advantage.
the b-fUe with ...1:teS-bS. 23 1:tb7 is the most 27 ...4:Jd7 28 4:Jc3 4:Jf6 29 4:Ja4 h5 30
obvious challenge (23 ltJe2 'iia6!? as 24 4:Jac5 ~xc5 31 4:Jxc5 'ifa7 32 'ii'd1 'ife7
'iixc6 is met by 24...'iixe2 25 'iixd6 'iixf2+! 33 'ifa4 4:Je4 34 'ifa6 4:Jd6 35 4:Jd3 'ifb7
or 23 ...:ta2 24 l:tb2 :txb2 25 'iixb2 :tbS 36 'ifxb7 4:Jxb7 37 4:Jb4 c5 38 dxc5
seem fme for Black) and now 23 ...:tb8 l:xc5 39 l:xc5 4:Jxc5 40 4:Jxd5 h4 41 f4
(23 ...'ii'a6!?) 24 lhb8 ~xb8. Here the pawn 4:Je4 42 ~f1 f5 43 ~e2 g5 44 fxg5 ~g6
grab 25 ltJb 1 :ta2 26 'ii'xc6 is extremely risky 45 4:Jf4+ ~xg5 46 4:Je6+ ~f6 47 4:Jd4
due to 26 ...lDe4! as 27 l:tf1 (27ltJc3 ltJxc3 28 4:Jc3+ 48 ~d3 4:Jd5 49 4:Jf3 f4 50 e4
121
It)b4+ 51 ~c3 It)c6 52 ~c4 ~e6 53

It)d4+-! <it>e5 54 ltJxc6+ ~xe4 55 It)d4
~e3 56 ltJf3! ~f2 57 ltJxh4 ~g3 58 ~d4
<it>xh4 59 ~e4 1-0
Game 63
P.Nikolic-L.Hansen
Wijk aan Zee 1995
1 d4 ltJf6 2 c4 e6 3 ltJf3 d5 4 ltJc3 ltJbd7
5 .i.g5 c6 6 cxd5 exd5 7 e3 .i.e7 8 'ilc2
0-0 9 i..d3 1:e8 10 0-0 ltJf8 11 h3 g6 12
i..xf6!? i..xf6 13 b4
With this idea, White tries to avoid 15 bxc6 bxc6 16 :fc1 ltJe6 17 'ild1!
spending a tempo on l:a1-b 1 as well as 'ilf6 18 1:ab1 :e7
removing the need to retreat the bishop once Hansen suggests the more direct
Black attacks it with ...ltJfS-e6. lS ... <ii>g7!?, intending ...h7-h6 and ...ltJe6-g5
with dangerous kingside play. The text is also
fme however.
19 :b3 ~g7 20 ltJa4 .i.d7 21 .i.a6 ':ae8
22 1:b7 ltJg5! 23 ltJxg5 'ilxg5 24 ~h1
1:xe3! 25 1:xd7! 'ilf4 26 ':xd6 'ilxf2! 27
ltJc5! ':e1+ 28 'ii'xe1 :xe1+ 29 ':xe1
'i'xe1+ 30 Iiith2 'i'f2 31 :xc6 'i'xd4 32
a4 h5 33 i..b5 'i'e5+ 34 Iiith1 .e1+ 35
Iiith2 'i'e5+ 36 ~h1 'ile1+ 37 ~h2 g5! 38
ltJd3 'i'e3 39 :c2 g4 40 g3 h4 41 hxg4
'i'xg3+ 42 Iiith1 h3 43 ':h2 a6 44 .i.xa6
'i'f3+ 45 Iiitg1 'ifg3+ 46 Iiith1 'ilf3+ 47
Iiitg1 'ild1+ 48 ~f2 'ilxa4 49 ':xh3 'i'xa6
13 ...i..e7!? 50 ':f3! 'i'a7+ 51 Iiitg2 'i'a2+ 52 ltJf2 ~f8
This idea again! As well as supporting 53 ':e3 d4 54 :d3 'i'd5+ 55 :f3 ~e7 56
kingside play from d6, the bishop of course Iiitg3 ~e6 57 ltJd3 'i'c4 58 Iiitg2 "c2+ 59
also helps to cover the c5-square. An ltJf2 'i'e2 60 :f4 "b2 61 :f3 "b7 62
interesting idea of Ehlvest's is 13...b6!? Iiitg3 'i'c7+ 63 ~g2 ~d5 64 ':f5+ ~c4 65
followed by ....i.cs-b7 and ...ltJfS-e6, going ':f3 "c6 66 ~g3 "g6 67 ~h3! f6 68
for a T artakower set-up. Against Karpov in Iiith2! ••6+ 69 Iiitg2 'i'g5 70 ~g3 "e5+
Vienna 1996 he preferred 13 ...a6 14 a4 (14 71 Iiitg2 'i'd5 72liitg3 "e5+ %-%
ltJa4!?) 14... .i.e6 15 b5 (15 l:fc1 b5!? is
suggested by Karpov) 15...axb5 16 axb5 ltJd7 Game 64
17 bxc6 bxc6 lSltJe2 c5 19 .i.b5 with a very Karpov-Campora
small edge for White. San Nicolas (match) 1994
14 b5 .i.d6!
see folio wing diagram 1 d4 d5 2 c4 e6 3 ltJc3 ltJf6 4 cxd5 exd5
5 i..g5 .i.e 7 6 e3 0-0 7 .i.d3 lLlbd7 8 lLl13
c6 9 "c2 1:e8 10 0-0 It)f8 11 h3 i..e6 12
122
Exchange Variation: Systems with t'iJf3
.i.xf6?! position does not make too much sense after

1bis doesn't really seem to the point in that. Instead, 19...ttJg6! was called for
this particular position. 12 ttJe5! is much
more active, when 12... ttJ6d7 13 i.xe7 :txe7
14 f4! f6 15ltJf3 followed by f4-f5 and g2-g4
seems promising for White. 12 i.f4!? is also
interesting transposing to Game 68 below. It
is too late for 12 :tab 1, however. In Henley-
Klovan, Biel 1996, Black won brilliantly after
12...ltJe4! (exploiting the loose pawn on a2
after any captures on e4) 13 i.f4 i.d6! 14
i.xd6 ltJxd6 1S ~a4 'i'f6 16 ~h2 i.xh3! 17
~xh3 'ii'h6+ 18 ~g3 :e4!! 19 i.xe4 ~xe4+
20 'i'xe4 dxe4. Finally, 12 :tfd is seen in
Game 72.
12... i..xf6 13 b4 ':c8 14 ltJa4!? just looking for ...~g6-h4 even at the cost
It is very unusual to see this move when of the h7-pawn. After 20 ~h 1 (20 ltJcs i.c8
Black has not yet weakened his queenside 21 ltJa6 :ce7 22 ~b4 :tb7! wins) 20...ltJh4!
with ... a7-a6. 21 i.xh7+ ~h8 22 i.d3, then 22...i.xh3 23
gxh3 i.xh2 24 ~2 'i'f3 25 :tgl 'i'xf2+ 26
~hl 'i'e3 27 i.f1 'i'f3+ 28 ~h2 !tce7 is
killing for Black
After the text, it is Karpov who has all the
fun!
20 .if5! 'i'h5 21 .ixe6 ltJxe6 22 ltJf3 f5
23 ':c3 ltJd8 24 llJc5 .ixc5 25 ':xc5 llJe6
26 ':c3 f4 27 e4 h6 28 :e1 :ce7 29
':xc6 dxe4 30 ':xe4 'ii'd5 31 ':c3 'ii'f5 32
'ii'e1 'ii'd5 33 ~h1 'ii'd6 34 'ii'd2 llJg5 35
':xe7 'ii'xe7 36 'ii'xf4 'ii'b4 37 llJxg5 hxg5
38 'ii'd2 g4 39 hxg4 1-0
14...':c7 15 ':ac1 .ie7! Game 65

Yes, we know what Black is doing! I.Sokolov-OIi
16 'ii'b 1 .id6 17 b5 'ii'f6 18 bxc6 bxc6 Pula 1997
19 ltJh2 'i'h4?
Karpov's commentary to the game is 1 d4 d5 2 c4 c6 3 llJc3 llJf6 4 llJf3 e6 5
rather confusing. As far as I can see, he has i.g5llJbd7 6 cxd5 exd5 7 e3 i.e7 8 'i'c2
played rather badly and stands worse, but he 0-0 9 .id3 !te8 10 h3 llJf8 11 0-0 llJe4!?
gives little comment until move 22 when he 12 .if4! f5!?
assesses the position as slightly better for The only consistent follow-up to ...~f6-
White. e4, though I'm not sure I like it. 12...~gS was
To my mind, Black makes a very serious played in Durie-Pfleger, European Cup 1984,
mistake here. By allowing the exchange of when 13 i.xgS!? (perhaps 13ltJxg5 .txgS 14
light-squared bishops, Black loses all his i.h2!?) 13 ...ii.xg5 14 b4! i.e7 15 bS i.d6 16
hopes of kingside pressure and thus his bxc6 bxc6 17 i.f5! was an edge for White.
123
<itf3 :d4 55 litxb5 :a4 56 tlJf4 tlJh2+ 57

~g3 tlJf1 + 58 ~g4 tlJe3+ 59 <itf3 tlJc2 60
:b8+ ~g7 61 :b7+ <itg8 62 liJd5 lDd4+
63 ~f2 :a5 64 :d7 lDe6 65 96 1:xd5 66
':xd5 ~g7 67 l:td6 lDf4 68 ~e3 lDxg6 69
<itte4 ~f7 70 1:a6 lDe7 71 ~e5 lDg6+ 72
~d6 lLlf8 73 1:a1 ~f6 74 :f1+ YZ-YZ
Game 66
Yusupov-Kramnik
Vienna 1996
1 d4 d5 2 c4 c6 3 ltJf3 lDf6 4 lDc3 e6 5
13 lDe5 lDg6 14lDxg6 i.g5lDbd7 6 cxd5 exd5 7 e3 i.e7 8 -'c2
14 f3!? ltJxf4 15 exf4 ltJg3 is suggested as 0-0 9 i.d3 :e8 10 0-0 lDf8 11 ':ae 1
unclear by Sokolov
14... hxg6 15 f3 lDf6 16 -.f2 lDh5 17
i.e5 i.h4 18 'it'd2 i.g3
Qlestion 43. What is White's idea with this

move?
An..ww- 43. With 11 :ae1, White puts to
19 i.xg3 one side any thoughts of queenside
19 f4 would have left White with a slight expansion and looks instead to exploit his
edge according to Sokolov. central and kingside chances. Thus, by
19 ...lDxg3 20 ':fe1 'it'd6 21 ':ab1 i.d7 protecting the e3-pawn, White intends 12
22 b4 b5 23 a4 a6 24 i.c2 1:e7 25 axb5 ltJe5 as 12...ltJg4 can then be met by 13
axb5 26 e4 dxe4 27 fxe4 ':ae8! 28 J.xe7 'ii'xe7 14 f4! Moreover, in certain
i.b3+ ~h7 29 'it'f2 fxe4 30 ':e3 lDf5 31 cases, White is ready to break in the centre
g4 lDh6 32 ':d1 -.xb4 33 i.c2 ltf7 34 with e3-e4.
'it'e2 :f3 35 lDxe4 ':xe3 36 -'xe3 -.c4 11 ...ltJe4! 12 i.xe7
37 'it'd2 'it'd5 38 lDg5+ ~h8 39 i.xg6 QIestion 44. Why not the standard 12 J.f4
':f8 40 i.c2 c5 41 -.d3 'it'xd4+ 42 'it'xd4 here - it's never easy for Black to maintain
cxd4 43 ltxd4 ltc8 44 ':xd7 :xc2 45 his knight on e4, is it?
:b7 :c5 46 ltJe6 lte5 47 ':b8+ lDg8 48 An..ww- 44. Here, the position of the
lDf8 :d5 49 h4 g5 50 l:.b7 :d1+ 51 ~g2 queen's rook on el causes a few problems.
lDf6 52 lDg6+ ~g8 53 hxg5 lLlxg4 54 After 12...i..fS! (12 ...tLlgS 13 tLlxgS ~xgS 14
124
Exchange Variation: ::;ystems wlrn '1.JTJ
i..xg5 'it'xg5 15 f4! 'it'f6 16 f5! .i.d7 17 'ii'f2 Anm:er 45. This is a strange one: White has
l:tad8 18 e4! gave White good chances in the concrete structural weakness - the IQP -
Kalinichenko-Volynsky, correspondence and no obvious sign of an attack to
1986, whereas 12... .i.b4 is considered compensate. Black has not fully equalised,
dubious by Ivan Sokolov who stood worse however, due to the annoying pressure along
against Hjartarson in Reykjavik 1988 after 13 the e-file which tends to paralyse Black's
h3 It)g6 14 i..h2 It)h4?! [Sokolov suggests pIeces.
14.. .f5!? 15 It)d2 .i.d6] 15lt)xh4 'iixh4 16 f3) 20 h3!?
13 J..xe4 J..xe4! 14lt)xe4 dxe4 15lt)d2 J..b4 20 l:Ue3 was played in Rausis-
16 :d1 J..xd2 17 :xd2 'iid5! the game was Viglundsson, Reykjavik 1997, and after
equal in Sapis-Maciejewski, Poland 1991. 20....:d6 (20...'i'f7 21lt)e5 'iif5 22 !:tf3 'i'h7
12 .....xe7 13 i.xe4 dxe4 144Jd2 f5 23 'i'e2 gives White the initiative according
The tricky 14...b6!? has also been tried, to Rausis) 21 It)e2?! (21 'ii'e2 maintains
aiming for ... J..c8-a6 to trap the white rook White's pull according to Rausis) 21...'iff7 22
on f1. After 15 'iia4 (15 'iixe4 J..a6 16 -'xe7 'i'c5 :ed8! 23 ~xa7 It)g6 Black had good
':xe7 17 It)e2 :ad8 18 b3 J..xe2 19 l:txe2 counterplay due to the offside queen on a7
':xd4 was equal in Netusil-Kacirek, Prague according to Rausis.
1993) 15 ...aS!? 16 'iixc6 J..b7 (16 ...i..a6 17 20 .....d6
ttJd5! followed by It)d5-c7) 17 -'xb6 J..a6 18 20...~f7 21 l2Je5 ~f5 22 ':£1 IS more
It)dxe4 J..xfl 19 :xfl ""4 20 It)d5 'iixb6 pleasant for White according to Yusupov.
21 It)xb6 l:ta6 22 It)d6 l:te6 the game 21 A1 e3 .i.f7
Arencibia-Garcia, Cuba 1995, was agreed 21...:e7, keeping things tight, was better
drawn in. 23lt)bc4 :axd6 24lt)xd6 :xd6 25 than this slightly panicky attempt for
':c 1 is a vexy murky ending according to counterplay according to Yusupov.
Ftacnik. 22 :'xe8 l:xe8 23 :'xe8 i.xe8 24 "b3+
15 f3 exf3 16 4Jxf3 i.e6 17 e4 fxe4 18 ~h7 25 'iixb7 .i.h5 26 4Je4 'iif4 27
Axe4 h6 'iixc6
18...!:tad8 19lt)e2!? (19 !:tfe1) 19 ...'iid620 27 It)ed2 J..xf3 28 It)xf3 'ifc1 + 29 <it;h2
It)g3 ttJg6 21 !:tfel 'iidS 22 ~a4 as 23 b4 It)e6 30 'ti'd7 would have left White clearly
axh4 24 'iixb4 'i'd7 was fairly equal m better according to Yusupov.
Timman-Drazic, Koge 1997. 27 ... i.xf3 28 gxf3 4Jg6 29 l2Jf2 'iixd4 30
19 Afe1 Aad8 'iic2 'iid5 31 f4 'iif3! 32 f5 l2Jf4 33 f6+
~g8 34 'iic4+ ~h7 35 'iic2+ ~g8 36
'iic4+ ~h7 Yz-Yz
In general therefore, the variation with 10
0-0 l2Jf8 and then 11 h3 seems like a small
edge for White, though it is not too terrifying
for Black. We now turn to the lines with 10
h3.
Q1estion 45. I don't understand What's the
difference if Black plays 10...l2Jf8?
Am'Za'r 45. Apart from castling kingside,
which we have already looked at, White has
QIestion 45. What is this position? two extra possibilities: 11 i..f4 (Games 67
125
and 68) and 11 0-0-0 (Game 69). these positions. By delaying ...c7-c6, Black
ensures that if his opponent tries to castle
queenside vexy quickly - for example on
move 10 - then Black can tty to open the c-
ftle with ...c7-c5 in one move rather than
wasting a tempo with ...c7-c6-c5. For
example, in Korchnoi-Yusupov, Dortmund
1994, Black transposed to a position where
he had played 9...c6 rather than 9.)iJf8 and
after 10 0-0-0 ltJf8 11 i.xf6 i.xf6 12 h3!?
i.e6 13 ~bl l:tc8 14 g4, Yusupov states that
14...c5 would have given Black reasonable
counterplay. How much better then if Black
plays 9...tiJf8, so that after 10 0-0-0 he plays
Qtestion 46. What is the idea behind this 10... i.e6 11 i.xf6 i.xf6 12 h3 :c8! followed
11 i.f4 move? by ...c7-c5.
Anszrer 46. On the downside, White loses 11 i.f4
a tempo and abandons all his pressure on the
h7-pawn as there are now no longer any
i.g5xf6 followed by i.d3xh7+ possibilities.
However, there are two main points to this
move, which was a favourite of Reshevsky's:
1. It prevents Black from playing ...ltJf6-e4
by removing the tactical basis for this
manoeuvre. Black is consequently prevented
from making use of his central outpost, at
least for the time being.
2. It supports a future ltJf3-e5, using
White's own central outpost. It is therefore
vexy much a central plan in conception.
11 ... i.d6
Game 67 Black uses the opportunity to exchange
Krivoseja-Klovan the dark-squared bishops, even though he
Germany 1998 cannot use the e4-outpost for his knight.
Qtestion 49. Can't Black gain a tempo flI'St
1 d4 12Jf6 2 c4 e6 3 12Je3 d5 4 exd5 exd5 with 11...ltJg6 12 i.h2 and then play
5 i.g5 i.e7 6 e3 12Jbd7 7 i.d3 0-08 Ve2 12...i.d6?
:e8 9 12Jf3 12Jf8 10 h3 e6 Ansrrer 49. This is a reasonable question.
A word must be said here about Black's As Reshevsky points out, the problem is that
move order, which is vexy precise indeed. the knight does not have a great deal to do
Qtestion 47. What do you mean? on g6, while it may just prove a target for
Anszrer 47. As you can see, Black delayed White's on-rushing kingside pawns after 13
playing ...c7-c6 until White had played 10 h3. i.xd6 'i'xd6 14 0-0-0 i.e6 15 g4!? However,
Qtestion 48. Why was that? after 15...:ac8 16 g5 tiJd7 17 h4? .tg4 18
Anszrer 48. A common idea for all the .te2 ltJf4! 19 tiJg1 tiJxe2+ 20 lDgxe2 c5 21
hackers out there is to castle queenside in ~b 1 b5! White was in big trouble in
126
Exchange Variation: Systems with lDf3
Krivoseja-Asrian, Minsk 1998. Krivoseja Yusupov suggests the immediate

suggests that 15 cat>b1 l:tac8 16 l:.c1 15...i..c8!? with the twin ideas of ...c6-c5 and
consolidating ftrst would have been slightly ... tLlfS-g6 as equal.
better for White. 16 1.h2 1.e8 171i'e3 ltJe4 18 1.xe4 dxe4
12 .i.xd6 ifxd6 13 0-0 .i.e6 14 :ab1 a5 19 ltJe5 ltJxe5 20 dxe5 i.xe5 21 1i'xe5
15 :fe1 4)8d7 16 a3 h6 17 ltJa4 4)e4 18 1.e6 22 l:.bd1 l:.d7 23 l:.d6 1i'b6 24 "a3
ltJd2 a5 25 :lfd1 :xd6 26 exd6 1i'b4 271i'xb4
IS i.xe4 dxe4 19 tLld2 i..d5 20 ttJc5 is a %-%
touch better for White according to
Krivoseja. The text allows rapid liquidation. Game 69
18 ... 1.f5 19 1.xe4 l:.xe4 20 ltJxe4 1.xe4 Anastasian-Lputian
21 ifd1 1i'g6 22 iff1 %-% Yerevan 1996
Game 68 1 e4 ltJf6 2 4)e3 e6 3 d4 d5 4 exd5 exd5
Krasenkov-Yusupov 5 1.g5 i.e 7 6 e3 ltJbd7 7 ltJf3 0-0 8 1.d3
Pula 1997 l:.e8 9 "e2 4)f8 10 h3 e6 11 0-0-0
1 4)f3 d5 2 d4 4)f6 3 e4 e6 4 4)e3 4)bd7

5 i.g5 e6 6 exd5 exd5 7 e3 i.e7 8 1i'e2
0-0 9 1.d3 :e8 10 h3 4)f8 11 1.f4 i.e6!?
120-0 l:.e8
Now that Black has played ...c7-c6, White

feels that he can castle queenside. However,
whereas White has played a rather quiet
move - 11 h3 - Black has made an extra
developing move ...ttJdl-fS.
A vety calm and sensible idea, deterring 11 ... a5!? 12 ~b1 a4! 13 4)xa4 "a5 14
moves like 13 ttJe5 due to 13...c5 b3
(Krasenkov) and just developing the This weakens the queenside, but 14 ttJc3
queenside before taking any further action. i.e6 followed by ...b7-b5 and ...ttJf6-e4 is
The alternative 12... tLl6d7 is seen in Game also dangerous.
71. 14 ... b5 15 ltJe5 i.xc5 16 1.xf6
13ltJa41.d6 14ltJe5 l:.e7! 16 dxc5 was better, when 16...ltJSdl is
A typical idea to cover the c7-square unclear according to Lputian. After the game
whilst allowing the bishop on e6 to drop continuation, Black rapidly gains the upper
back to c8, enabling the rook on eS to hand.
support a later ...tLlf6-e4. 16 ...1.b4 17 .i.h4 :e6 18 :he1 .i.xe1 19
1 5 :ab 1 ll.)g6 :xe1 b4 20 lite1 i.a6 21 4)e1 e5 22
127
dxc5 It)d7 23 i.xa6 Aexa6 24 It)d3 1t'b5 20 ... cxb5! 21 axb5 a5! 22 'iib2 b6!
25 a4 bxa3 26 c6 It)f8 27 i.e7 It)e6 28
r;t>a2 A6a7 29 i.d6 Ad8 30 i.g3 Ae8 31
'i'e3 h5 32 It)f4 It)d8 33 'i'e5 'i'xe5 34
Axe5 It)xe6 35 <ita 1 Aa5 36 Axd5 ltJb4
37 Ad1 h4 38 i.h2 ':e2 0-1
We shall now conclude with a selection of
'classic' minority attack games. Although
their theoretical relevancy may vary, each of
them illustrates a certain approach or
important strategical theme, whether it is in
the notes or played in the game itself.
Game 70
Gelfand-Ivanchuk Now bS-b6 is prevented and White's
Linares 1993 pieces have no real targets, as he cannot get
at the black d-pawn. Black stands clearly
1 d4 It)f6 2 e4 e6 3 It)f3 d5 4 It)e3 ltJbd7 better.
5 cxd5 exd5 6 .ig5 i.e7 7 e3 0-0 8 .id3 23 ltJa4 :c4 24 ':a1 ltJe4 25 f3 ltJg3 26
':e8 9 'i'c2 ltJf8 10 0-0 c6 11 h3 g6 12 It)e5 i..xe5 27 dxe5 'i'c7 28 ~h2 ltJf5 29
':ab1 ltJe6 13 i.h6 ltJg7 14 b4 a6 15 a4 f4 ':e2 30 ':c3 ':xe3 31 'i'xe3 'i'xc3 32
i.f5! 16 ltJe5 ':c8 17 i.xg7 .ixd3 18 ltJxc3 ltJxe3 33 ltJa4 d4 34 ':a3 ltJc4 35
ltJxd3 ~xg7 19 ':b3 i.d6! Ad3 :d8 36 ~g3 ':d5 37 ~f2 g5 38 g3
~g6 0-1
Game 71
Beliavsky-Ivanchuk
Linares 1993
1 d4 ltJf6 2 c4 e6 3 ltJf3 d5 4 ltJc3 ltJbd7
5 cxd5 exd5 6 i..g5 i.e7 7 e3 c6 8 'i'e2
0-0 9 i.d3 ':e8 10 h3 ltJf8 11 i.f4 i.e6!?
1 2 0-0 ltJ6d7 13 ':ab 1 tLlb6
Black has achieved a nice set-up for his

pieces: the bishop is well-placed on d6,
preventing the knight from returning to eS,
while the rook on c8 annoys the queen when
the c-fue is opened
20 b5?
20 'ii'b2 ~g8! 21 bS axbS! (21...cxbS?! 22
axbS as 23 b6! is good for White according
to Ivanchuk) 22 axbS cS! with an equal
position according to Ivanchuk.
128
Exchange Variation: Systems with &Df3
Black transfers his knight to b6 in order to

eye the c4-square that will be weakened when
White plays b2-b4.
14 b4 .i.dS 15 .i.xdS 'ifxdS 1S a4 a6 17
lDd2! :e7 18 &Db3!
A typical manoeuvre, though with an
unusual knight. Once Black has put his pawn
to a6, a white knight is almost always very
well placed on cS.
18 ....:c7
IS ...'ii'xb4 19 tDcs "as 20 :b3 followed
by l:f1-b 1 would net the b-pawn according
to Beliavsky.
19lDc5 lDbd7 20 f4! fS 21 f5! 17 ... ':ac8 18 'i'c5 'i'b8
Another typical idea to gain space on the IS .. :ii'xcS?! 19 dxcs ~d7 20 b4 lLlgeS 21
kingside and squeeze Black on all fronts. lLlxeS lLlxeS 22 i.c2 is clearly better for
21 ...1&.f7 22 'i'f2? White according to Karpov.
A bad mistake according to Beliavsky. 22 19 'iia3
as! would have left White in complete Looking for as-a6.
control. 19 ... aS 20 ':c3 'iic7 21 ':bc1 ':a8 22
22 ... a5 23 lDxd7 ':xd7 24 bxa5 l:xa5 25 ltJd2
l:bS 'i'a3 2S ':fb1 ':a8 27 ':Sb3 'i'e7 28 Now that he has covered the c4-square,
'i'f4 ':e8 29 'it>f2 g5 30 "f3 h5 31 a5 h4 White is ready to play b2-b4.
32 lDa4 'i'd8 33 lDc5 :de7 34 lDxb7 22 ... a5 23 l:tb1 lDc8 24 b4 axb4 25
'i'c7 35 1&.f1 'it>g7 3S ~g1 'it>hS 37 as 'i'xb4 ltJdS 2S lDb3 1&.c8 27 a5 lDe7 28
1&.h5 38 'i'f2 lDd7 39 ':c1 1-0 ltJg3 gS 29 %;lcc 1 h5?
29 ...lLlefS 30 tDxfs i.xfs 31 i.xfS lLlxfS
Game 72 32 tDcs l:ebS was still tenable according to
Karpov-Kharitonov Karpov.
US SR Championship 1988 30 ':a1 h4 31 lDf1 .i.f5 32 1&.e2! ltJe4 33
ltJc5 lDxc5 34 'iixc5 1&.eS 35 lDd2 lDf5
1 c4 eS 2 lDc3 d5 3 d4 lDfS 4 cxd5 exd5 3S lDf3 'i'd8 37 as! bxaS 38 ':xaS ':xaS
5 1&.g5 i..e7 S e3 lDbd7 7 lDf3 cS 81&.d3 39 1&.xaS "a8 40 "xcS 'i'xcs 41 ':xcS
0-0 9 'i'c2 ':e8 10 0-0 ltJf8 11 h3 1&.eS ':a8 42 1&.d3 1-0
12 ':fc1!?
A typical Karpov move: it will always be Game 73
useful, but it is hard to say whether it is the Andersson-Kasparov
most accurate choice at this juncture. All you Belgrade (match) 1985
know is that Karpov will make you suffer
whatever the theoretical conclusion! Though this game is not remarkable in
12... ltJSd7 13 .i.f4lDbS 14 ':ab1 1&.dS 15 itself, Kasparov's comments are extremely
lDe2 lDgS 1S .i.xdS 'i'xdS 17 a4! instructive.
Karpov didn't want to allow a knight to c4 1 lDf3 d5 2 d4 lDfS 3 c4 eS 4 lDc3 cS 5
after 17 b4, and thus delays the advance of cxd5 exd5 S 1&.g5 .i.e7 7 "c2 gS 8 e3
the b-pawn in order to always have b2-b3 to .i.f5 9 .i.d3 .i.xd3 10 'i'xd3 0-0
chase away the knight. 10... tDbd7 is seen in the next main game.
129
11 .i.xf6 .i.xf6 12 b4 'ifd6 13 :b1 lLld7 e4 dxe4 36 fxe4 h4 37 g4 lLlxd4 38

140-0 :fd8 15 ~He1 liJxd4 'ifxg4+ 39 <it>f1 :d6 40 :d3 'iff4+
15 b5 is met by 15...c5! while 15 t;)a4 b5! 41 :f2 'ife5 42 :fd2 'iff4+ 43 :f2 'ife5
16 t;)c5 t;)xc5 17 bxc5 'ii'e6 18 a4 a6 is 44 :fd2 Yz -Yz
unclear according to Kasparov.
1S... 4Jb6 164Jd2 'ike7 17 'i'e2 a6 18 a4 Game 74
ii.g7 19 4Je2 .i.h6! Bobotsov-T .Petrosian
Lugano Olympiad 1968
This is a beautiful example of exploiting
passive play by White.
1 d4 ltJf6 2 e4 e6 3 ltJf3 dS 4 exd5 exd5
5 4Je3 e6 6 ii.g5 ii.e7 7 'i'e2 g6 8 e3
ii.f5 9 ii.d3 ii.xd3 10 'ilxd3 4Jbd7 11
ii.h6 ltJg4 12 ii.f4 0-0 13 0-0 :e8 14 h3
4Jgf6 15 ltJe5 4Jb6 16 .i.g5 4Je4 17
ii.xe7 'ilxe7 18 'ike2 4Jd6 19 4Ja4 ltJbe4
20 4Jxe44Jxe4 21 4Je5 4Jd6 22 :ac1
Preventing White from re-rouung his

knight to d3 via f4.
201:e1
20 b5 cxb5! 21 axbS :dc8 22 'ii'b2 as is
unclear according to Kasparov.
20 ...'i'e6 21 4Je1 lbc4! 22 'ike3 4Jd6!
Now Black starts to get going.

22 ...'i'g5 23 'ikd1 h5 24 ~h1 :e7 25
4Jd3 ltJe4 26 ltJe5 ltJd6 27 4Jd3 'i'f5 28
ltJe5 f6 29 lbf3 :g7 30 ltJh2 :e8 31
~g1 4Je4 32 'i'f3 'ike6 33 :fd1 g5 34
'i'xhS f5 35 :e1 g4 36 hxg4 fxg4 37 f3
gxf3 38 4Jxf3 1:h7 39 'i'eS 'i'c8 40 'i'f4
1:f8 41 'i'e5 :f5 0-1
The knight is vexy comfortably placed
here, defending b7 and preventing b4-b5 Game 75
while eyeing the e4 and c4 outposts. P .Nikolic-Kramnik
23 4Jd3 ii.g7 24 4Je5 'i'e7 25 :e2 :e8 Amber (blindfold) 1998
26 :be1 'i'e7 27 a5 1:e7 28 'i'd3 :ae8
29 f3 4Jf5 30 g3 h5 31 c.tg2 'ife8 32 1 d4 4Jf6 2 c4 e6 3 ltJf3 d5 4 ll)e3 e6 5
ltJdb3 :d8 33 :d1 .i.f6 34 'ile3 .i.g7 35 cxd5 exd5 6 .i.g5 .i.e7 7 e3 J.f5 8 .i.d3
130
Ex c han 9 e van arlo n.- .:> y:; , e:"',, vv I '" -~, ...
.i.xd3 9 'it'xd3 ~bd7 100-00-0 11 :tab1 With this flexible system, Black seeks to
a5 12 a3 ~e4 13 .i.xe7 'it'xe7 14 b4 b5! win the battle for the tempo. In the resulting
Exchange variation position, however, Black
has already weakened his queenside dark
squares. 1bis game is the punishment for this
'crime'!
White's knight cannot get to e5 to attack

c6, and Black's knights are heading for c4!
15 'ii'c2 axb4 16 axb4 ~d6 17 1:b3 lDb6
18 lDe5 1:fc8 19 lDd3 lDbc4 20 lDc5 :e8
21 h3 g6 22 1:c1 1:a7 23 'it'd1 h5 24 9 cxd5 exd5 10 .id3 1:e8 11 0-0 lDf8 12
~h1 "g5 25 1:bb1 1:ae7 261:a1 lDf5 27 1:fe1 i.g4 13lDd2lD6d7 14 i.f4 i.g5 15
1:a2 h3!
Now 15... ~xf4 16 exf4 ~e6 17 f5! wins a
pIece.
15...i.h5 16 i.h2 i.g6 17 i.xg6 hxg6 18
ii'b3 ii'b6 19 tDa4 ii'xb3 20 tDxb3!
27 ... lDcxe3 28 fxe3 1:xe3 29 1:f2 ii'h4

30 "d2 lDxd4 31 :'cf1 lDf5 32 :xf5
gxf5 33 lDd1 :'e1 34 ~g1 :'8e2 35 ii'c3
':xd1 0-1
Not a nice position for Black - the knights
Game 76 are looking for that c5-square, or even...
Rubinstein-Takacs 20 ... tDe6 21 tDa5!
Budapest 1926 Just sit back and enjoy the rest.
21 ....:a7 22 ~f1 i.d8 23 b4 f5 24 tDb2
1 c4 ~f6 2 d4 e6 3 ~c3 d5 4 .li.g5 lDbd7 g5 25 tDd3 ~f7 26 l:tc2 i.b6 27 i.d6
5 e3 .li.e7 6 ~f3 0-0 7 :tc1 c6 8 'it'c2 a6 tDd8 28 tDc5 ~xc5 29 i.xc5 .li.xc5 30
131
bxc5 ~e7 31 %:tb2 ~d7 32 %:teb1 ~c8 33

~e2 l:.e7 34 ~f3 l:.e4 35 g4 g6 36 llg1
!£Jf7 37 h4 gxh4 38 gxf5 gxf5 39 llg7
!£Jd8 40 l:.g8 f4 41 l:.h8 fxe3 42 fxe3
~d7 43 1:.g2 1:.e8 44 1:.xh4 %:te 7 45 %:th8
~c7 46 1:.gg8 1:.d7 47 ~b3 a5 48 !£Jc1
:a8 49 ~d3 b5 50 cxb6+ ~xb6 51 !£Jc5
:d6 52 a4 1:.c8 53 ~g4 1-0
Game 77
T .Petrosian-Beliavsky
USSR Championship 1983
1 d4 d5 2 c4 e6 3 ~c3 ~f6 4 cxd5 exd5 16...'ii'h6 17 'ii'f2 :e7 18 f5 g6 19 e4
5 i.g5 i.e7 6 e3 0-0 7 i.d3 ~bd7 8 ~f3 dxe4 20 ~xe4 gxf5 21 'ii'g3+ ~h8 22
1:.e8 9 0-0 c6 10 'ii'c2 !£Jf8 11 a3 !£Je4 12 ~d6 f4 23 1:.xe7 'ii'xd6 24 l:.xd7 'ii'xd7
i.f4 !£Jg5 13 ~xg5 i.xg5 14 i.xg5 'ii'xg5 25 'ii'xf4 :d8 26 'ii'f6+ ~g8 27 ~h1
15 1:.ae1! i.d7 16 f4! 'ii'xd4 28 'ii'xf7+ ~h8 29 'ii'e7 !£Jg6 30
We have often seen this plan. As Black's i.xg6 hxg6 31 h3 b5 32 :f6 :g8 33
knight is a long way from e4, White advances Axe6 :g7 34 'ii'g5 ~h7 35 Cit>h2 b4 36
his pawn to f5, restricting Black's light- :f6 bxa3 37 bxa3 'ii'c4 38 :f4 'ii'e7 39
squared bishop, and then breaks with e3-e4. 'ii'h4+ Cit>g8 40 'ii'g3 a5 41 a4 'ii'b6 1-0
132
Exchange Variation: Systems with ([jf3
Summary
Try to really understand the ideas in this chapter - they will serve you well in all your games in
this opening. The key idea for Black in many structures is to place the dark-squared bishop on
the b8-h2 diagonal in order to put pressure on White's kingside, while at the same time
defending his weak queenside dark squares.
1 d4 d5 2 c4 e6 3 ([jc3 liJf6 4 cxd5

44::lf3 ~e7
S cxdS exdS 6 ~gS c6 (D)
7'i1c2 g6
8 e4 - Game 58
8 e3 .ifs 9 .id3 .ixd3 10 'ilxd3
10...0-0 - Game 73; 10...lDbd7 - Game 74
7 e3 - Game 75
S i.g5
S...h6 6 i.h4 0-0 7 cxd5 - Gttme 59; S....ie7 6 e3 0-0 7 :let - Gam! 76
4 ...exd5 5 i.g5 i.e7 6 e3 c6 7 i.d3liJbd7 8 liJf3 0-0 9 'i'c2 :e8 (D) 10 0-0
10 h34::lf8
11 i.f4
11...~d6 - Game 67
11.. ..ie6 12 0-0: 12 ...:lc8 - Game 68; 12...tLl6d7 - Game 71
110-0-0 - Game 69
10 ... liJf8 11 :ab1
11 h3 (D)
11...g6
12 :lab1lDe6 13 .ih64::lg7 14 b4 a6 15 a4.if5
16 .ixg7 - Game 62; 164::le5 - Game 70
12 i.xf6 - Game 63
11...~e6: 12 .ixf6 - Game 64; 12 :lfet- Game 72
11...lDe4 - Game 65
11 :ae1- Game 66; 11 a3 - Game 77
11 ... g6
11...4::lg6 - Game 61
12 b4 - Game 60
6 ... c6 9 ... :e8 11 h3
133
CHAPTER EIGHT
Exchange Variation:
Flexible Systems
In this chapter White plays the Exchange Q.iestiun 1. This looks like a subtle move
variation, but does not commit his king's order!
knight in order to retain the option of placing Amrrer 1. This is Black's most ambitious
it on e2. This is Kasparov's favourite scheme
continuation and probably the most
of development against the QGD. annoying continuation for White.
These systems arise from two move Q.testiun 2. So what is the point exactly?
Ansrrer 2. Black is still looking to develop
orders: 1 d4 dS 2 c4 e6 3 ltJc3 ltJf6 4 cxdS
exdS 5 ~g5 (Games 78-86) and 1 d4 dS 2 c4
his light-squared bishop. 5... ~f5 on the last
e63 ltJc3 i.e7!? 4 cxdS exdS 5 ~f4 (Games
move would have lost the dS-pawn, so by
87-91). We shall fIrst examine 3... ltJf6.
protecting it Black threatens to put the
.......- - - - - - - - - - - - -.. bishop on f5 next move. White has two
Game 78 reactions - 6 e3 (as in this game and Games
Gulko-Short 79 and 80) can lead to a complicated ending,
match 1994 while 6 'ii'c2 (as in Games 81-86) involves
' - - - - - - - - - - - - - -.. different concessions.
1 d4 tt)f6 2 c4 e6 3 tt)c3 d5 4 cxd5 exd5 Q.iestiun 3. Is it bad for White to go into
5 i.g5 c6!? the ending then?
Amrrer 3. No, but it is a matter of personal
taste. Most White players hope for more
from the opening than to reach a
complicated ending - they want the chance
of a quick kill!
6 e3 i.f5
The solid but co-operative 6... J..e7 leads
to the main line after 7 ~d3 ltJbdl 8 'ii'c2
(see Games 82-86).
7 'l'f3!
This is the only way for White to make
anything of his opponent's early ambition.
7 ... i.g6 8 i.xf6 Wxf6 9 'iixf6 gxf6
134
Exchange Variation: Flexible Systems
QIestion 8. Why?
An.szar 8. From d6, the knight guards f5
and thus prevents any white piece from
installing itself there. It also eyes the e4- and
c4-outposts and supports .. .f6-f5.
10~d2
Preparing .tfl-d3 to neutralise the bishop
on g6 and to start taking control of the f5-
square. 10 ltJf3 is considered in the next
game and 10 h4 in Game 80, while White has
10 0-0-0 with a similar idea to the game
continuation. This simply led to a draw after
10... ltJd7 11 .td3 .txd3!? 12 :Xd3 :g8 13
QIestion 4. What's going on? g3 ltJd7 14 ltJf3 in Shipov-Khalifman,
An.szar 4. This position has traditionally European Club Cup 1999, but 11 h4, as in
been assessed as clearly better for White, but Shariyazdanov-Kharlov, Elista 1996, was
recent games have cast doubt on this. In more aggressive when 11...ltJb6 12 h5 .tf5
return for the bishop pair, White inflicts a 13 ltJf3 .tg4 14 .te2 .txf3 15 gxf3 h6 16
serious structural weakness on the black .td3 .td6 17 f4 ltJc8 was agreed drawn,
position: not only are the f-pawns doubled, though as King points out, 18 :hg1 does
but Black also has an isolated h-pawn. You look a little better for White. King suggests
can certainly imagine any Trompowsky 13 ...ltJc8 14 ~4 .te6 15 .td3 ltJd6 as an
player being happy with White's position! unprovement.
QIestion 5. But isn't it just bad having a 10... ltJd7 11 .td3 ltJb6! 12 b3 .ta3!
weakness like this in the endgame? Hauchard-Kharlov, Linares 1997, saw a
An.szar 5. There are two factors in Black's very similar idea, clearly based on the present
favour: the position is rather closed and game: 12....tb4 13 ltJge2 ltJc8 14 h4 ltJd6 15
Black's weakness is on the wrong side for h5 i.xd3 16 ~xd3 .txc3 17 ltJxc3 :g8 18
White. On the queenside, White would have :ag1 f5 19 f3 h6 20 ct>e2 ct>d7 21 ~f2 :g5
the semi-open c-file to exploit it, but on the 22 ltJe2 as and a draw was agreed.
kingside, his major pieces struggle to get 13 ltJge2 ~d7 14 ltJg3 ltJc8! 15 h4 ltJd6
involved in the action. Clearly, White does 16 h5 .txd3 17 ~xd3
not have enough immediate firepower to win
one of Black's kingside pawns so he must be
more restrained.
QIestion 6. So what does he attack?
An.szar 6. Black's main kingside
weaknesses are on the light squares as ...g7-
g6 is no longer possible to cover the h5- and
f5-squares. Ideally, White wishes to install a
knight on f5, cramping Black's kingside,
before he thinks about targeting the pawns
themselves.
QIestion 7. So how does Black react?
A173lW" 7. Black's key idea is to play his
queen's knight to d6. QIestion 9. What is this position?
135
Ar13Zrer 9. White has manoeuvred his

knight towards the fS-square and forced
Black to exchange his light-squared bishop.
However, Black's knight on d6 holds
everything together, while thanks to the
doubled f-pawns, Black can also use the g-ftle
to activate his major pieces.
17 ... i.b2! 18 :'ab1 i.xc3 19 <itxc3 :hg8
20 <itd3 a5!? 21 a4 f5! 22 :h2 ltJe4 23
ltJxe4 fxe4+ 24 ~e2 f5 25 f3 ~e6 26
:11 :'g3 27 fxe4 fxe4 28 :'f4 :ag8 29
~f2 h6 YZ-YZ
Game 79 23 :he1 a5 24 e4 dxe4 25 fxe4 :d8 26

Milov-Pigusov iof3 ltJe8 27 :d1 :'ed7 28 d5 cxd5 29
New York Open 1998 :xd5 :'xd5 30 ltJcxd5 ltJd6 31 e5 fxe5
' -_ _ _ _ _ _ _ _ _ _ _ _ _.. 32 :xe5 a4 33 bxa4 :c8 34 i.xh5 i.e4
1 c4 e6 2 ltJc3 d5 3 d4 ltJf6 4 cxd5 exd5 35 i.d1 :'c5 36 i.b3 i.xf4 37 gxf4 :c1
5 i.g5 c6 6 e3 i.f5 7 'iVf3 i.g6 8 i.xf6 38 :g5+ ~f8 39ltJf6 i.c6 40 liJh7+ rl;e7
'ilxf6 9 'iixf6 gxf6 10 ltJf3!? 41 :e5+ rl;d8 42 ltJg5 f6 43 :e6 liJe8 44
White prepares to attack the fS-square ltJf3 :h 1 45 a5 i.xf3 46 ~xf3 :xh4 47
from h4. The drawback is that White no i.a4liJc7 48 :d6+ ~c8 49 :xf6ltJd5 50
longer has the idea of h2-h4-hs. :f8+ <itc7 51 i.b3 liJe7 52 :f7 rl;d6 53
10 ... ltJd7 11 ltJh4 i.e7 12 g3 :f6+ ~c7 54 a6 bxa6 55 ':xa6 :h1 56
12 f4!? f5 13 lLlf3 lLlb6 14 .td3 lLlc8 15 :e6 ltJc6 57 :e4 ~d6 58 ~g4 ltJe7 59
0-0-0 lLld6 is fme for Black according to :d4+ ~c6 60 :d1 :h2 61 ~g5 liJc8 62
Piket. i.e6 ltJd6 63 :c1+ ~b5 64 ~f6 :e2 65
12 ... ltJb6 13 f3 f5 :f2 66 ~e5 ltJxf5 1-0
Piket's suggested improvement over the
13 :d1 of Piket-Van der Sterren, Wijk aan Game 80
Zee 1998, when 13 ....tb4 14 a3 .txc3+ 15 I.Sokolov-Dautov
bxc3 lLla4 16 ~d2 b5 17 f3 0-0 18 g4 lLlb2 Nussloch 1996
was very murky. Alternatively, 13 0-0-0, as in
Yennolinsky-Azmaiparashvili, Elista Olym- 1 d4 ltJf6 2 c4 e6 3 ltJc3 d5 4 cxd5 exd5
piad 1998, led to a draw after 13 ...lLlc8 14 5 i.g5 c6 6 e3 i.f5 7 'iif3 i.g6 8 i.xf6
.td3 lLld6 15 f3 .txd3 16 ltxd3 f5 17 lLlg2 'iixf6 9 'iixf6 gxf6 10 h4 h5!?
.tg5 18 ~c2. White seems to cause his Qlestian 10. This looks very natural.
opponent more problems by leaving the An~ 10. In fact it is quite a risky
bishop on g6 than by exchanging it off. decision for Black because it puts the isolated
13 ... 0-0 14 rl;f2 :fe8 15 :d1 lbc8 16 h-pawn within reach of White's pieces and of
lbg2lbd6 17 h4 h5 18 lbf4 i.f8 19 :e1 his pawns. It also makes the bishop on g6 a
i.h6 20 b3 ~g7 21 i.g2 :e7 22 :e2 lot less stable. Instead 10...lLld7 11 h5 .tfs 12
:ae8 ~d2 (Ivan Sokolov suggests 12 lLlf3,
White has a small pull, but Black should intending lLlB-h4) 12...:g8 13 g3 Jte6 14
be able to hold a draw ... though in the game i.d3 f5! 15 ttJge2 lLlf6 16 ttJf4 Jtd6 17 J::th3
he doesn't! :g5 worked out fine for Black in Agrest-
136
Nickoloff, Elista Olympiad 1998.

Game 81
Kasparov-Ivanchuk
Wijk aan Zee 1999
1 d4 d5 2 c4 e6 3 lLlc3 lLlf6 4 cxd5 exd5
5 i.g5 c6!? 6 'ii'c2
Preventing ....tc8-fs. Black's most
common response is 6....te7 (as in Games
82-86), but here we shall discuss an
interesting way for Black to attempt to cross
White's plan.
11 lLlh3 .i.d6 12 .i.e2

12 g3 is suggested as an improvement
with ttJf4 to follow. After 12... ttJd7 13 ttJf4
we have transposed into Drasko-Bellini,
Montecatini 1995, which was slightly better
for White after 13 ... ~xf4!? 14 gxf4 ~fS 1S f3
:d8 16 ~f2 ttJb6 17 .te2 ~e7 18 b4
followed by a2-a4.
12 ... lLld7 13 1:c1 tt:Jb6 14 g3 lLlc4!? 15
.i.xc4 dxc4 16lLlf4 .i.f5!? 17 f3 O-O-O?
This puts the king too far from the
kingside action. Instead 17...bs was better 6 ... lLla6!?
according to Ivan Sokolov. Qu:stion 11. Why not 6... g6 to play ... .tc8-
18 ~f2 b5 19 lLlxh5! 1:xh5 20 g4 .i.xg4 fS on the next move?
21 fxg4 1:h6 22 lLle4 ~d7 23 It'f3 a5 24 Ansrrer 11. This was tried in Vladimirov-
h5 a4 25 a3 1:g8 26 :cf1 ~e6 27 1:h4 Diaz, Havana 1986, when White was
.i.c7 28 It'e2 .i.b8 29 lLlc5+ ~e7 30 :a5 tempted into 7 ttJxdS?! 'iWxdS 8 .txf6 i.b4t
.i.g3 31 ':h3 .i.d6 32 'iiff3 :1g5 33 lLle4 9 ~d1 0-0 10 e4 ~e8 11 .td3 and now
:1xf5+ 34 gxf5 .i.c7 35 ~e2 ~f8 36lLlf2 11...'iWe6 12 i.eS cS would have been veIY
'iife 7 37 It'd2 .i.d8 38 'iifc3 .i.a5+ 39 ~c2 dicey for White. However, as Vladimirov
.i.d8 40 lLlg4 ':h8 41 h6 'iiff8 42 e4 'iifg8 points out, simply 7 e3 .tfS 8 'iWb3! forces an
43 d5 c5 44 e5 fxe5 45 lLlxe5 .i.f6 46 unpleasant queenside weakness on Black
h7+ ':xh7 47 ':xh7 It'xh7 48lLlxf7 b4 49 with 8...b6.
lLld6 c3 50 bxc3 .i.xc3 51 lLle4 i.d4 52 Qu:stion 12. So what is the point of
d6 bxa3 53 ~b1 c4 54 'iifa2 'iifg7 55 d7 6... ttJa6?
.i.b6 56 ~xa3 ~f8 57 f6 c3 58 lLlxc3 Ansrrer 12. Black's wishes to play ...ttJa6-
'iiff7 59 ~xa4 .i.d8 60 ~b5 'iifxf6 61 'iifc6 b4 to chase the queen from the b1-h7
'iife6 ~-~ diagonal, clearing the way for ... i.c8-fS
Black seems to be just about okay after 6 thereafter.
e3 ~fS, so we shall now move on to the 7 e3
systems with 6 'ii'c2 rather than the quieter 6 Qu:stion 13. How about 7 a3? Isn't the
e3. knight just silly on a6?
137
Arl3Zrer 13. After 7 a3, the black knight after 9... aS 10 a3 ttJa6 11 iLd3 in Dokhoian-
moves to e6 (via c7). Remember the number Vaganian Tilburg 1994. The subde 8...h6!? 9
of times in the previous chapter that Black .th4 g6 has been suggested, but after
played ...ttJbS-d7, ...:fS-eS and then ... ttJf8- Dokhoian's 10 "d1 (10 a3 .tf5 11 e4 dxe4
e6? Here, Black has achieved this without 12 axb4 'ii'xd4 'unclear' also doesn't look
having to play ...lIfS-eS, while extracting the great for Black) 10... .tf5 11 :c1 g5 12.t.g3,
tempo a2-a3 on the way. Se.Ivanov suggests 12...'ifaS can still be met by 13 'iff3!, while
the continuation 7 a3 ttJc7 S e3 ttJe6 9 .th4 12... ttJe4 13 a3 ~xc3 14 :'xc3 ~a2 15 IIb3
g6!? 10 .td3 llJg7 11 f3! (the difference in does have the feel of a tragedy in the making
this system - White can set up e3-e4 to for Black!
prevent his opponent from exchanging light-
squared bishops with ....tcS-fs) 11.. ..te7 12
llJge20-0 13 0-0 llJfS 14 .tf2 cS with unclear
play.
The text is the most critical, and Kasparov
naturally goes for it!
7 ...lbb4 8 'ild2!?
This is a very interesting moment. In
1994, Kasparov's current chief second, Yuri
Dokhoian, produced a beautiful idea that
built on Nigel Short's suggestion of S "bl!?
The point is that after S "dl, Nigel Short
had produced very active counterplay against
Ehlvest at the Manila Olympiad 1992, with dearly, however, Dokhoian's 'boss' saw
S....tf5 9 lIet "as! 10 i.xf6 gxf6 11 "d2 something he didn't like, and Ivanchuk
(11 "'3 is met by 11..."'6! according to evidently has something ready. Maybe S... h6
Ehlvest) 11...llJxa2! 12 lIa1 ttJxc3 13 IIxaS 9 .th4 g5 10 .tg3 ttJe4!? with the idea of 11
ttJe4 14 l:.xd5!? (14 "dl .t.b4+ 15 <it>e2 ttJxe4 (11 a3 .tf5Q 11...dxe4 12 "xe4+ (12
.txaS 16 f3 ttJd6 17 'it>f2 0-0-0 lS .td3 .tg6 a3!?) 12... .te7 with .....dS-aS and ....tcS-fS
gives Black good counterplay due to the to follow?
weakness of e3, while 14 'ii'c2 .tb4+ lS <it>dl S ... .tfS 9 :c1 as 10 a3 lbaS 11 lbge2
.txaS 16 .td3 lDxf2+ 17 "xf2 .txd3 lS Vyzmanavin-Se.Ivanov, Elista 1995, was
'ii'xf6 IIgS 19 llJf3 :xg2 20 'ifhS+ <it>e7 21 equal after 11 .td3 .txd3 12 'i'xd3 .te7 13
"xaS :xb2 wins for Black according to ttJf3 0-0 14 h4 lIeS 15 .txf6 .txf6 16 'it>f1
Short. Qmsequently White feels obliged to . ~bS 17 g3 ~d7.
return the queen.) 14...lbxd2 15 IIxf5 .tb4 11 ... hS 12 .tf4lbd7
16 <it>e2 <it>e7 17 llJf3 llJc4 lS <it>d1 llJd6 19
:f4 as with approximate equality. Dokhoian
played S "'1 first (still preventing ....tcS-f5) The first new move of the game. Barsov-
and only after S... g6 did he play 9 "dt! Se.Ivanov, Gennany 1994, saw 12....td6 13
llJg3 .t.h7 14 .txd6 "xd6 15 llJa4 O-O!? 16
'ifxas hS!? with some play for the pawn.
Qtestian 14. What is the point of this? 13 lbg3 .teS 14 e4 lbbS 1S exdS lbxdS
Arl3Zrer 14. Here Black no longer has any 1S lbxdS 'ii'xdS 17 .tc4 'ilxg2 18 'ile3
counterplay with .. :.d8-aS as the knight on 0-0-0 19 .txe6+ fxeS 20 'ii'xe6+ ~d7 21
f6 is hanging. White was pleasantly better 'ii'eS+ AdS 22 "eS+ Yz-Yz
138
Kharitonov) 19....ltd7! 20 'ii'g3 f6 21 f4 i..e8

22 fxgS fxgS 23 'iWxgS 'ii'xgS 24 lIxgS ltJf6
was soon agreed drawn in Vyzmanavin-
Kharitonov, Helsinki 1992, but this line
looks very dodgy for Black to me.
100-0-0 g6
The calmer way in which Black can try to

exploit the 6 • c2 move order is to play an
early ...ltJf6-hS system to free his position by
exchanging the dark-squared bishops.
Game 82
Ward-Parker 11 ~b1
4NCL 1997 Ward's concept of putting the knight on
f3 in this variation is not very common, but it
1 d4 d5 2 c4 e6 3 tDc3 tDf6 4 cxd5 exd5 works very effectively here. 11 ltJge2 is
5 .tg5 c6 6 'i'c2 .te7 7 e3 tDbd7 8 .td3 nonna! when 11...ltJb6 12 ltJg3 ltJg7 (Black is
liJh5!? trying to prove that the knight has few active
The standard s...o-o 9 ltJge2 lIeS 10 0-0 chances on g3) 13 <it>b 1 i..d7 14 lId 0-0-0
ltJf8 is seen in Games 83-86. IS ltJa4 ltJxa4 16 'iWxa4 ~bS 17 lIe3 b6 18
9 .txe7 'i'xe7 lIa3 (18 i..a6 to prevent Black from
Qtestion 15. Can't Black just play this regrouping with ... :d8-cS-c7 was played in
against 6 e3, 7 i..d3 and 8 ltJge2 as well? Kasparov-Andersson, Reykjavik 19S8, when
Arz..w.rer 15. It is much less good for Black 18... ltJe6 19 IDlel :beS 20 'i'b3 'ifd6 21
in that move order. After 6 e3 i..e7 7 i..d3 ltJfl <it>a8 22ltJd2 ltJc7 was reasonably okay
ltJbd7 S ltJge2 ~S 9 i..xe7 .xe7 10 g4 for Black) IS ...i..eS 19 'ifc2 ':cS 20 ':el i..d7
ltJgf6 11 ltJg3! g6 12 'iWe2 ltJb6 13 gS ltJg8 21 'iWd2 hS 22 ':b3 ltJe6 gave balanced
14 h4 h6 (14...hS IS 0-0-0 i..d7 16 e4 0-0-0 chances in Timman-Short, Linares,
17 lIhel ii'd6 18 exdS cxdS 19 'ifeS .xeS 20 Candidates Match 1993.
lIxeS was clearly better for White in Kosten- 11 ...tDb6 12 h3 tDg7 13 g4 .td7 14
Legky, France 1998) IS O-O-O!? (Kharitonov tDf3!? 0-0-0 15 'i'b3 tDe8 16 a4!
suggests IS f4 hxgS 16 fxgS!? [16 hxgS This aggressive thrust soon has Black
~1+ 17ltJxhl i..e6 lSltJg3 0-0-0 190-0-0 scrambling just m stay on the board
.d7 20 :hI ltJe7 intending ...ltJe7-fS is 16 ....te6 17 a5 liJc4 18 'i'a4 'i'f6 19
unclear according to Kharitonov] as .te2 tDc7 20 ':c1 h5 21 g5 'iff5+ 22
interesting) IS ...hxgS 16 hS gxhS 17 ltJxhS ~a 1 :he8 23 b3 tDd6 24 a6 liJxa6 25
l:th6 18 l:tdgl <if;d8 19 'ii'f3 (19 e4 i..d7 and tDe5 tDb8 26 'it'xa7 'it'xf2 27 tDa4 'it'xe2
19 f4 f6 20 e4 dxe4 21 ltJxe4 g4, intending 28 liJb6+ ~c7 29 liJa8+ ~c8 30 liJb6+
.. .f6-fS, are both unclear according to <iitc7 31 liJa8+ <iitc8 32 liJb6+ Yz -Yz
139
Having dealt with the tricky question of With 11.. ..lTl.e6, Black begins preparations
move order, we now move on to the main to achieve this break by first mobilising his
line. queenside. The older but perfectly playable
alternative 11...ltJhS!? is the subject of Game
Game 83 85.
Gelfand-Piket 12 :ad1
Wijk aan Zee 1998 12 l:Iael is seen in the next main game,
' - - - - - - - - - - - - - - . while Van Wely-Piket, Antwezp 1996, saw
1 d4 dS 2 c4 e6 3 lbc3 lbf6 4 cxdS exdS equality aher 12 .th4 l:Ic8 13 ltadl a6 14
S iJ..gS c6 6 'iWc2 iJ..e7 7 e3 0-0 8 iJ..d3 ~hlliJg6 15 .tf2 c5! 16 dxcS .txcS 17lDd4
lbbd7 9liJge2 :e8 100-0 lbf8 11 f3! .txd4 18 exd4lDf4 19 :felliJxd3 20 "xd3.
12 ...:c8 13 a3
White prepares b2-b4 in order to stop his
opponent from breaking out with ... c6-cS.
The sharper 13 e4 gave no more than
dynamic equality in Timman-Yusupov, Riga
1995, aher 13...dxe4 14 fxe4liJg4 15 .tf4 {IS
.tcl .tgS! is nice for Black according to
Timman, while 15 .txe7 "xe7 16"d2 cS 17
d5 .td7 followed by .....d6 and ...lDg6 is
equal according to Lautier} lS ...lDg6 16 eS
.tgS! 17 .txg6 hxg6 18 'ird2 .ixf4 19lDxf4
.tc4 20 l:Ifel c5!
Instead 13 ~hl was tried in Lautier-Oll,
The key difference from the lDf3 systems: Pamu 1998, ready to meet 13 ...c5 with 14
White threatens to occupy the centre with dxcS l:Ixcs 15 "a4 a6 16 .tbl with a slight
e3-e4 without accepting an IQP. Since Black edge. Joel Lautier recommends 13 ...ttJg6
cannot do the same with his ...c6-cS central instead.
break, White inevitably has a slight dynamic 13 ... a6
edge. The less dynamic 11 a3 is considered in Intending ...c6-cS.
Game 86. 14 lba4 lb6d7 1S iJ..xe 7 'iWxe 7 16 b4 as!
11 ... iJ..e6 17 :b1 axb4 18 axb4 b6!
The irrunediate lL.cS is suicidal: Vaisser-
C.Flear, French Team Championship 1998,
saw 12 .ibS i.d7 13 .txd7 "xd7 14 :adl
c4 15 i.xf6 .txf6 16 e4 dxe4 17 lDxe4 "c6
18 lDxf6+ 'irxf6 19 'irxc4 with a clear extra
pawn for White.
Nonetheless, the idea of ...c6-cS is
tempting in this type of position as White has
weakened the e3-square with f2-f3.
QIestinn 16. Why is this important?
Amzrer 16. After ... c6-cS, d4xc5 then
....ie7xcS will give Black pressure against the
e3-pawn that he would not have in the ttJf3
system. Black is extremely solid here and,
140
unsurprisingly, the game soon fizzles out to a 3S :c2 :xh4+- 39 ~g 1 ~f7 0-1
draw.
19 1:fc1 g6 20 liJf4 .i..f5 21 :'e1 ~xd3
Game 85
22 liJxd3 liJe6 23 liJc3 liJf6 24 'iif2 c5 Vaisser-Bricard
25 dxc5 bxc5 26 b5 d4 27 exd4 cxd4 French Team Championship 1998
2S liJe4 liJxe4 29 f!xe4 'iia3 30 'iid2 :c3
31 lDf4 %-% 1 d4 e6 2 c4liJf6 3 ttJc3 d5 4 cxd5 exd5
5 i..g5 .1e7 6 e3 0-0 7 .1d3 ttJbd7 8
Game 84 ttJge2 :eS 9 0-0 c6 10 'i'c2 ttJfS 11 f3
Lutz-Yusupov ttJh5!?
Tilburg 1993
1 c4 e6 2 liJc3 d5 3 d4liJf6 4 cxd5 exd5
5 i..g5 i..e7 6 e3 0-0 7 i.d3 ttJbd7 S
liJge2 1:eS 9 0-0 c6 10 'i'c2 liJfS 11 f3
.i..e6 12 1:ae 1 :'c8 13 )f;h 1 ttJ6d7 14
i..xe7 'iWxe7
1bis older line, forcing White to expose

his centre very early, is quite reasonable in
myoplOlon.
12 i..xe7 'iWxe7 13 e4 dxe4 14 fxe4 i..e6!
This allows Black to aim for ...c6-c5
without fearing ltJc3-dS. 14.....tg4?!, as in
Ivanchuk-Yusupov, Brussels Candidates
15 'i'd2 Match 1991, turned out very unpleasantly for
15 e4 is met by Black's standard counter: Black after 15 eS! :ad8 16 ltJe4 ltJg6 17
15...dxe4 16 fxe4 c5. :adl :f8 18 h3 be2 19 .ixe2 ltJhf4 20
15...ttJb6 16 b3 ':cdS ..tc4.
Intending to manoeuvre the knight to d6 151:f2
via c8. 15 e5 c5 16 d5 ..tdl and 15 :adl :ad8
17 a4 'iWb4 18lDe4 'i'xd2 19 ttJxd2 .1d7! with ... c6-c5 to follow are both fine for Black
After this move, preparing ...ltJf8-e6, according to Bareev.
Black has no problems according to 15 ... ttJf6 16 h3 ':ad8 17 :'af1 ttJg6 18
Yusupov. a3 c5!?
20 liJg3 a5 21 1:e2 ttJe6 22 ':c 1 ttJc8 23 18...:f8 was tried in Bareev-Ahlander,
f4 ttJd6 24 liJf3 f6 25 )f;g1 :'e7 26 h4 Naestved 1988, when after 19 b4 either
':de8 27 ~2 liJdS 28 ':ee1 ttJSf7 29 19... b6, intending ...c6-cS or ... a7-aS
ttJh2 liJh6 30 liJgf1 i..f5 31 i..e2 ttJe4+ immediately would have been fine for Black
32 ~g1 g5 33 g4 gxf4 34 gxf5 :g7 35 according to Bareev. The text seems even
.1g4 liJxg4 36 liJxg4 1:xg4+ 37 ~h2 f3 more conclusive, however.
14-1
stranded on h6.
19 e5 liJd5 20 liJxd5 i.xd5 21 liJf4 liJxf4

22 i.xh 7+ ~hS 23 :xf4 exd4 24 :xd4 14 liJg3 'fih4 15 :fe1 i.e6 16 b4 a6 17
i.xg2 25 :'xdS :xdS 26 'i'xg2 ~xh7 27 liJa4 l:e7 1S liJe5 :'aeS 19 i.f1 i.eS 20
'iie4+ ~gS 2S 'iif5 'iies+ 29 :f2 'ild5 30 a4 liJg6 21 b5 axb5 22 axb5 'fig5 23
Ae2 'fid1+ 31 ~f2 :eS 32 'iie4 :e6 33 bxe6 bxe6 24 :'aSliJh4 25 :b1 liJ6f5 26
'iixb7 :g6 34 'iieS+ ~h7 35 "f5 'i'd4+ liJxf5 liJxf5 27 :bbS liJd6 2S liJd3 :'e 7
36 ~f3 "d5+ 37 ~f2 'i'c5+ 38 ~f3 29 'i'e5 'i'e7 30 liJb4 g6 31 i.a6 'ile6 32
'fi e6+ 39 ~e3 ~gS 40 :'e2 :'g3+ 41 ~d4 i.xeS liJxeS 33 :a6 ~g7 34 :lxc6 :lxe6
"b6+ 42 ~d5 'iidS+ 43 ~e4 g6 44 -.f2 35 'i'xe6 'fixe6 36 liJxe6 liJd6 37 :xeS
:'xh3 45 :'d2 "a5 46 'i'd4 'ila6+ 47 liJxeS 3S g4 h6 39 f3 ~f6 40 ~f2 ~e6
~b4 'iib7+ 4S ~a4 'ile6+ 49 ~b4 'ile6 41 h4 liJd6 42 liJe5 liJb5 43 liJd3 g5 44
50 'fid5 "b6+ 51 ~a4 :h4+ 52 b4 'iia6+ h5 liJa3 45 ~e2 liJe4 46 liJf2 liJd6 47 e4
53 'i'a5 'iie6+ 54 "b5 "eS 55 'i'd5 ~g7 liJeS 4S liJd1 liJf6 49 e5 liJd7 50 liJe3 f6
56 'fid6 :h3 57 'fif6+ ~h6 5S :'dS 'ile2+ 51 exf6 liJxf6 52 ~d3 liJgS 53 liJf5 ~d7
59 ~b5 'iie2+ 60 ~a4 'i'e2+ 61 ~b5 54 ~e3 ~e6 55 f4 ~f6 56 ~f3 ~f7 57
'iie2+ Ya-Ya liJxh6+ liJxh6 5S fxg5 liJgS 59 ~f4 ~e6
.----------------. 60 g6 ~f6 61 g5+ ~e6 62 g7 liJe7 63
Game 86 h6 liJg6+ 64 ~g4 ~f7 65 ~f5 ~gS 66
Topalov-Piket ~f6liJh4 67 h7+ 1-0
Amber (blindfold) 1998 We now tum to Kasparov's favourite
move order, 3...i.e7.
1 d4 liJf6 2 c4 e6 3 liJe3 d5 4 i.g5 i.e 7
5 exd5 exd5 6 e3 e6 7 i.d3 0-0 S 'i'e2 Game 87
liJbd7 9 liJge2 :ea 10 0-0 liJfS 11 a3!? Vaisser-San Segundo
A much more solid continuation than 11 Greece 1997
f3. White tempers his ambitions and just
hints at playing b2-b4. 1 d4 d5 2 e4 e6 3 liJe3 i.e7
11 ... tiJg4 12 i.xe7 'i'xe7 13 h3 QIestion 17. What is the point of 3...i.e7?
see following diagram Ansrrer 17. Black does not allow his
opponent to follow up the exchange on d5
13 ... liJh6!? with i.c1-g5. The bishop is thus played to £4,
13 ...lD£6 followed by ...lDf8-e6-g5 seems which gives this line a distinct identity.
more reasonable as the knight gets rather QIestion 18. Is the bishop better on £4 than
142
ongS? I like this simple move. It prevents all of

Black's ... ttJc6-b4 tactics whilst maintaining
White's pressure against the b7- and dS-
pawns. The greedy 8 'iixb7 and 8 g4 are
considered in Games 88 and 89 respectively.
8 ... ltJa5!?
8.. J1b8 was seen in Kir.Georgiev-
Kotronias, Corfu 1991, when 9 ltJf3 0-0 10
.te2 h6!? (10...a6 as in Spraggett-Yusupov,
Hastings 1989, looked better for White after
11 0-0 bS and now 12 :tfet ltJaS 13 "dl
ltJc4 14 :ta2, intending .te2-d3, was best
according to Spraggett) 11 0-0 .te6! 12 ltJd2
.td6 13 .txd6 cxd6?! (13 .....xd6 is a touch
Anszrer 18. As we saw in the previous
chapter, there are many occasions where
better for White) 14.tf3 "as 15 "a2! ltJe7
16 b4 was not very inspiring for Black.
White voluntarily retreats the bishop to f4 9 'i'a4t e6 10 ltJf3 0-0
even at the cost of a tempo. However, White 10... ltJhS? 11 .te7! "xc7 12 ltJxdS "d8
exerts much less pressure on d5 in this line, 13 ltJxe7 ~xe7 14 b4 ltJe4 15 .txe4 bS 16
which gives Black an interesting possibility "dl bxe4 17 ltJeS "dS 18 "xhs all held
which Kasparov has favoured. together for White in Gavrikov-Ubilava,
4 exd5 exd5 5.i.f4ltJf6 Thilisi 1983.
See Game 91 for S... c6. 11 ~e2
6 e3 .i.f5!? 11 ltJes, as in Burmakin-Koniushkov,
Kstovo 1997, is the most accurate way for
White to play. After 11...b5 12 "dl :tc8 13
.td3 .txd3 14ltJxd3ltJe4 150-0 :te8 16 a4!
b4 17 ltJe2, White had a very pleasant
advantage.
11 ... b5 12 Vd1 ltJe4 13 'ike1 a5 14ltJe5
Ve8 15 0-0 a4 16 ltJxe4 dxe4 17 .i.g5
Ve7 18 l:e1 :ae8 19 .i.f3 .id3 20 e4
ltJd7 21 .i.f4 Vb6 22 Ve3 f6 23 .ig4
:d8 24 .i.e6+ ~h8 25 'i'h3 g5 26 .i.e3
'ike7 27 Vh5 .i.d6 28 .i.f5 ltJb6 29 d5
exd5 30 ltJxd5 Yz -Yz
However, there is also a complicated way
And here it is! With the bishop on gS, to play!
such a move would always lose a pawn to
i.gSxf6 followed by 'ii'dl-b3. Here White Game 88
has less pressure on the black centre, which Thorsteins-I. Zaitsev
offers Black some extra tactical resources. Protvino 1988
7 'iib3
The critical test of Black's idea, but 7 1 d4 d5 2 e4 e6 3ltJe3 i.e7 4 exd5 exd5
ltJge2, as in Game 90, is also possible. 5 .if4 ltJf6 6 e3 i.f5 7 Vb3 ltJe6 8 'iixb7
7 ... ltJe68 a3!? ltJb4 9 i.b5+ ~f8 1 0 ~d2!?
143
Here 10 l:Id1 is met by Dutreeuw's Recently, however, White has been

10...i..d6!!, as after 11 i.xd6 cxd6 White (unsuccessfully) trying another approach.
cannot avoid the perpetual attack on his
queen with ...l:Ia8-b8-aS. Game 89
10 ... lDe4+! Topalov-Kasparov
TIlls seems best. Instead 10...i.d6 fails to Linares 1997
11 i.xd6 cxd6 12 i..e8!, while
A2maiparashvili's recommendation of 10... a6 1 c4 e6 2liJc3 d5 3 d4 .i.e7 4 cxd5 exd5
11 i.a4liJd3 12 i.xc7 "cS 13 "xcS l:Ixc8 5 ~f4 liJf6 6 e3 .1f5 7 'i'b3 lLlc6 S g4?!
14 i.aS ltJxf2 15 :£1 liJ2e4+ 16 liJxe4+ lDxg4
liJxe4 17 ~e2liJd6 seems very odd, because
after Salov's 18 lbf3, I don't see much
compensation for Black.
11 liJxe4.i.xe4 12 .i.xc7!?
12 f3 :bS 13 "xc7 was agreed drawn in
Dorfman-Marciano, French Championship
1995, while Zaitsev analyses further
13......xc7 14 .txc7 :xb5 15 fxe4 dxe4 with
compensation for the pawn.
12 ...'i'cS 13 'i'xcS+ :'xcS 14 f3 :xc7 15
fxe4 dxe4
15...:c2+ 16 ~d1 :xb2 17 a3 liJa2! 1S
liJe2 i.xa3 19 i.c6 ~e7 20 i.xdS :cS 21
:xa2 :b1+ 22 ~d2 .tb4+ 23 <ittd3 :xh1 24 9liJxd5?!
:Xal+ is unclear according to Thorsteins. 9 "xdS is simply met by 9.....c8
16 83 llJd5 17 liJh3 95!? 1S .i.a4 g4! according to Kasparov, while 9 a3!? 0-0 10
see folio wing diagram i.g2 .th4 11 .tg3 .txg3 12 hxg3 liJe7 13
liJxdSliJxdS 14 "xdS "'xdS 15 i..xdS :ad8!
The position is complicated, but 16 i.xb7 :bS 17 .tf3 :xb2 18 liJe2 liJf6
nevertheless balanced. was nothing special for White in
19 lLlf2 f5 20 .i.b3 liJxe3! 21 ~xe3 :'b7 Aleksandrov-Azmaiparashvili, World
22 Jt.e6 .i.g5+ 23 ~e2 ~e 7 24 .i.xf5 Championship 1997.
:'xb2+ 25 ~e1 :lfS 26 lLlxe4 :'xf5 ~ - Yz 9 ... 0-0 10 .i.g2 .i.h4! 11 .i.g3 .i.e6 12
144
~f1! a5! 13 ~h3 a4 14 'iic4 tDa5 15 19 f3 tDb6 20 b3 i.a3 21 :tc2 the game was
'ii'c5b6 agreed drawn.
8 ... .i.e6
The opening has not been a success for

9 i.d3
White - he is fighting just to stay on level
terms. The interesting 9 'fIb3!? b6 10 i..e2 c5 11
0-0 ltJc6 12 :fd1 led to an unclear position
16 i.xh4 bxc5 17 .i.xd8 :axd8 18liJe7+
~h8 19 d5 i.d7 20 :c1 c4 21 liJg5 h6
in Kharlov-Komeev, Russian Championship
1998, after 12...c4!? 13 'iVc2 a6 14ltJf5 b5 15
22 liJf3 :b8 23 h3 liJf6 24 liJd4 :xb2 25
i.f3 l:xa2 26 ~g2 :e8 27 l£lec6 i.xc6
a3 'iVd7 16ltJxe7+ ltJxe7 17 f3 .i.f5.
9 ... c5 10 dxc5 .i.xc5 11 0-0 liJc6 12
28 liJxc6 liJxc6 29 dxc6 :xe3 30 :xc4
:b3 31 :d4 :bb2 32 :f1 :d2 33 :b4
:c1 d4!?
~h7! 34 :b7 liJe8 35 ~g3 g6 36 :e1This seems very reasonable. Gelfand-
liJd6 37 :xc7 ~g7 38 l:d7 liJf5+ 39 ~f4
Kasparov, Linares 1994, saw White keep a
:xf20-1 definite edge after 12... i.d6 13 ltJge2 :c8 14
r---------------. i.b1 i.xf4 15ltJxf4 i.g4 16 f3 i.e6 17 'iVd2.
Game 90 13 liJb5
Sherbakov-Koniushkov 13 ltJce4 i.e7 14 ltJc5 i.xc5 15 :xc5
Krasnodar 1997 dxe3 16 i.xe3 ltJb4 17 i.f5 ha2 was very
r-._ _ _ _ _ _ _ _ _ _ _ _ _.. comfortable for Black in Lautier-Ivanchuk,
1 d4 d5 2 c4 e6 3 liJc3 i.e7 4 cxd5 exd5 Moscow Olympiad 1994.
5 i.f4 liJf6 6 e3 i.f5 7 l£lge2!? 13 ... i.b6 14 e4 liJg4 15 h3 liJge5 16
This is another common way to avoid the i.xe5 liJxe5 17 f4 liJxd3 18 'ii'xd3 f6 19
complications. liJa3 %-%
7 ... 0-0 8 liJg3 The position is unclear but balanced.
8 ::tel, aiming to prevent Black from Black's traditional main line has been
achieving ...c7-c5, was played by Kasparov 5...c6 instead of 5...ltJf6, intending to meet 6
against Karpov in their World Championship e3 with 6...i.fs.
match at Seville 1987. After 8...c6 9 ltJg3
i.e6 (9 ... .i.g6 10 h4 h6 11 hs i.h7 12 .i.d3 is Game 91
clearly better for White according to Yusupov-Lputian
Dorfman) 10 i.d3 l:e8 11 "'3 'iWb6 12 Gennany-Annenia match 1996
'it'c2 ltJbd7 13 0-0 g6 14 h3 .i.f8 15 ltJge2
l:ac8 16 'ifd2 ~hS 17 i.h2 ~g7 18 g4 'iid8 1 d4 d5 2 c4 e6 3l£lc3 .i.e7 4 cxd5 exd5
145
5 ~f4 c6 6 e3 ~f5 White.

Recently at the Elista Olympiad 1998, 9 'ii'b3 b6 10 ttJf3 ~e7 11 ~e5 g5?!
Lputian played 6...lDf6 against me, but after 7 Alternatively, 11...ltJf6 12 g5 ttJfd7 13 g6
~d3 0-0 8 h3 lDbd7 9 lDf3 l:e8 10 'i'c2 ttJxe5 14 i.xe5 i.f6! (14 ... fxg6 15 i.xg7 :g8
lDf8, he was a tempo down on the 10 h3 and 16 %hh7 is clearly better for White according
11 i..f4 manoeuvre in the lDf3 system. This to Gulko) 15:xh7 O-O! 16 i.g3 fxg6 17:h2
is not fatal, of course, but neither is it ideal ~f7! 18 0-0-0 and now 18...:h8?! 19 :xh8
for Black. After 11 0-0 ltJg6 12 i.h2 i.d6 13 'iVxh8 20 e4! was horrible for Black in
i.xd6 'iVxd6 14 :bl 'iVe7 15 b4 lDe4 16 b5 Gulko-Lputian, Glendale 1994, but even the
lDg5 17 ltJxg5 'iVxg5 18 bxc6 bxc6 19 f4! improvement 18...ttJd7 (Gulko) 19 e4 i.e7 is
White stood extremely well. still not desirable for Black.
7 g4 i.e6 8 h4! The text seems no better, however.
12 i.g3 llJf6 13 i.e2
13 f3 h5! is annoying according to
Yusupov.
8 ...i.xh4
The most consistent continuation. Not
8...ltJd7 9 h5! and now 9...'ib6 10 :bllDgf6
11 f3 h6 12 i.d3 c5 13 ltJge2 :c8 14 'it>f1 13 ... 'ii'c8 14 :c1! ttJbd7 15 liJb5 ttJc5!
0-0 15 g5! hxg5 16 i.xg5 :fe8 17 'ii'el cxd4 16 dxc5 cxb5 17 'ii'xb5+! <it>f8 18 ttJc6
18 exd4 ltlli7 19 i.xe7 lhe7 20 1Wg3 was ttJe4 19 llJxe7 <it>xe7 20 i.d6+ ~f6 21
exceptionally unpleasant for Black in Knaak- 'ii'b4!
Geller, Moscow 1982, as was 11 ...0-0 12 i.d3 White has emerged from the complica-
c5 13 ltJge2 .:tac8 14 'it>f1 cxd4 15 exd4 i.d6 tions with a clear advantage.
16 1Wd2ltJe8 17 'it>g2 'ii'd8 18 :belltJb6 19 21 ...<it>g7! 22 c6 llJxd6 23 'ii'xd6 'i'd8 24
i.bllLlc4 20 'ii'd3 in Beliavsky-Geller, USSR 'fie5+ 'fif6 25 'i'xf6+ <it>xf6 26 ~d2 :ac8
Championship 1983. 27 f4 :c7 28 :h6+ ~g7 29 fxg5 <it>f8 30
Instead of 9...'ii'b6, 9...lLlh6 was tried by b4 -l;e7 31 b5 -l;d6 32 i.d3 1:g8 33
Karpov against Kasparov in the 1985 World i.xh7 :xg5 34 i.f5 :g8 35 a4 :e7 36
Championship in Moscow, but 10 i.e2 ttJb6 :ch1 ~c5 37 i.xe6 fxe6 38 l:.h7 :xh7
11 :c1 i.d6 (11...ttJc4 12 i.xc4 dxc4 13 39 l:.xh7 l:.xg4 40 :xa7 e5 41 l:.d7! :g8
i.xh6 gxh6 is suggested as unclear by 42 c 7 :c8 43 <it>c3 d4+ 44 exd4+ exd4+
Kasparov) 12 ltlli3 i.xf4 13 ttJxf4 i.d7 14 45 -l;d3 ~b4 46 -l;xd4 -l;xa4 47 Iit>d5
.:tgl g5 15 hxg6 hxg6 16 'it>d2 'iVe7 17 b3 g5 <it>xb5 48 <it>d6 <it>c4 49 :d8 :xc7 50
18 lDd3 0-0-0 19 :It 1 was vety nice for ~xc7 b5 51 'iifb6! 1-0
146
Summary . .
In my opinion these lines are not as good for White as therr reputatIon suggests. In the 3...tLlf6
variation, 5... c6, aiming for the endgame, seems a goo~ ch?ice so .long as Dokhoian's 8 'iVb1
and 9 'tid1 can be countered Otherwise, even the malO lines as 10 Game 85 seem perfectly
reasonable for Black.
1 d4 d5 2 c4 e6 3 lDc3 lbf6
3... iLe7 4 cxdS exdS 5 iLf4
5...ltJf6 6 e3 iLf5 (D)
7 'ilb3ltJc6
8 a3 - Game 87
8 'tixb7 - Game 88
8 g4 - Game 89
7ltJge2 - Game 90
5...c6 - Game 91
4 cxd5 exd5 5 ~g5 c6 6 'i'c2
6 e3 i..f5 7 'fif3 iLg6 8 iLxf6 'fixf6 9 'iVxf6 gxf6 (D)
10 ~d2 - Game 78
10 ltJf3 - Game 79
10 h4 - Game 80
6 ... ~e7
6...ltJa6 - Game 81
7 e3 ltJbd7 8 ~d3 0-0
8...4:Jh5 - Game 82
9 ltJge2 :e8 10 0-0 ltJf8 (D) 11 f3
11 a3 - Game 86
11 ... ~e6
11...4:Jh5 - Game 85
12 :ad1
12 ~ae1 - Game 84
12 ...:c8 - Game 83
6 ... ~f5 9 ... gxf6 10.. .tiJfB
147
CHAPTER NINE
Systems with ~xf6
In this chapter we shall examine lines in well as loosening Black's protection of the
which White replies to the attack on his dS-square.
bishop by capturing the knight on f6
immediately. 1bis can arise via two move
orders: 5 ~g5 h6 6 i.xf6
3. White disrupts Black's most

hannonious method of development - the
knight on f6 and the dark-squared bishop on
or 5...0-0 6 e3 h6 7 i..xf6. the f8-a3 diagonal - by drawing the bishop to
f6.
4. Black's main freeing idea in the QGD is
Qiestion 1. Why does White want to to play a move like ... tt)f6-e4 or ... tt)f6-d5,
concede the bishop pair like this? using the opposition of the dark-squared
Anm:er 1. White's reasons are the bishops on the h4-d8 diagonal to exchange
following: two sets of minor pieces. The following
1. By capturing on f6 immediately, White scenario is very common: White refuses the
speeds up his development - he doesn't exchange of dark-squared bishops and Black
waste time on a retreating move. wins White's dark-squared bishop for his
2. Removing the knight from f6 allows knight eventually anyway. With ~g5xf6,
White to consider the e2-e4 central break, as White takes a practical decision. By giving up
148
Systems with ~xf6
the bishop pair immediately, White releases a ~xe5 13 ~xe5 'ir'xe5+ 14 .i.e2 .i.g4 15 f3
pressure point on his opponent's position, .i.f5 .
but ensures that Black does not get the QIestion 6. So what does 7 'iVb3 do?
chance to try to free himself by playing the Am'ZW" 6. With 7 '11>3, White uses several
typical ... ~f6-e4, and gives himself a wider of the ideas behind the early exchange on f6:
choice of plans at the outset. 1. White threatens to win the pawn on d5,
We have seen the idea of i.g5xf6 many which now lacks the protection of the knight
times before in the QGD - in the onf6.
Tartakower or the Exchange variations, for 2. White uses the time saved on .tg5-h4
example - but here it is unusual because to continue the development of his
White's aim is dynamic rather than structural. queenside.
Qiestion 2. What about move order? 3. By freeing d1 for the white rook (with
Should you castle first or play ...h7-h6 first? gain of time) White hopes to bring sufficient
An.mer 2. This seems a rather uncertain cover to the d4-pawn to be able to play e2-
point! Every QGD expert has a fair e4.
sprinkling of games with both, though 5...h6
does tend to be the most popular choice.
QIestion 3. What are the differences?
An.mer 3. If you play 5...h6, you have to
reckon with aggressive plans using the option
of e2-e4 in one move. If you play 5... 0-0,
then these options are obviously not available
for White after 6 e3, but White does gain the
interesting move orders 6 cxd5 and 6 'iVc2.
Qiestion 4. So which do you recommend?
An.mu 4. I would play 5...h6 - it just
seems the least hassle!
Game 92 7 ... c6
P.Cramling-Amura The nonnal move in this position: Black
Merlo (match) 1994 defends his central pawn and maintains his
flexibility. Note, however, that as the c6-
1 d4 d5 2 c4 e6 3 ~c3 i.e 7 4 ltJf3 ~f6 square is no longer available to Black's
5 i.g5 h6 6 i.xf6 i.xf6 knight, White's chances of achieving e2-e4
Here White has a wide choice between the have gone up considerably.
game continuation, 7 'iVc2 (Game 93), 7 'iVd2 An alternative plan was tried in Yusupov-
(Game 94) and 7 e3 (Games 95-100). Lputian, European Club Cup 1997, when
7 "ifb3!? 7...dxc4 8 'iVxc4 a6 9 ttJe4 .te7 10 ':'c1 0-0
QIestion 5. Hey, why doesn't White just 11 e3 (11 'i'xc7 'iVxc7 12 lhc7 ttJc6
play 7 e4? followed by ... .te7-d8 traps the rook.~
AMa'r 5. It is just a little bit early for this 11...i.d7 12 ltJe5 .td6 13 ttJxd7 ltJxd7 14
move as the white pieces are not yet well .te2 ttJf6 15 ttJc5 i.xc5 16 'iVxcs c6 was
placed to cover the d4-pawn. Thus in 011- fairly equal.
Vaganian, Moscow Olympiad 1994, Black S:d1
stood very well after 7...dxe4 8 ~xe4 ltJc6 9 Cramling-Gurieli, Women's Candidates
ltJxf6+ 'ifxf6 10 'ifd2 0-0 11 l:d1 e5 12 dxe5 1997, saw the more aggressive 8 0-0-0 dxc49
149
'iVxc4 b5 10 'iib3 as 11 e4 a4 12 'iic2 ttJa6 .ite1 60 ':b1 J.f2 61 ~c4 J.e3 62 ':d1
13 ~b1 0-0 14 h4 a3 15 b3 tDb4 16 'iVd2 1-0
.tb7 17 'iVe3 'it'aS 18 e5 i..e7 19 l:.c1 l:.ac8
which turned out very nicely for Black. Game 93
S .. :iia5!? Bacrot-Korchnoi
An interesting idea to prevent a rapid e2- Albert (match) 1997
e4 from White. Instead 8...0-0 9 e4 dxe4 10
ttJxe4 'iVaS+ 11 l:Id2! .te7 12 .te2 ttJd7 1 d4 tiJf6 2 e4 e6 3 tiJf3 d5 4 tiJe3 .Jie 7
(12 ...i.b4 13 ttJc3 ttJd7 14 0-0 i.xc3 15 bxc3 5 .Jig5 h6 6 J.xf6 il.xf6 7 'iie2 dxe4! 8
b6 16 ~e1 i.b7 17 iLd3 was a touch better e3
for White in P.Cramling-Gueneau, French 8 lld1 a6! followed by ... b7-bs keeps the
Team Championship 1998) 13 0-0 e5 14 d5 pawn.
f5 15 ttJc3 cs 16 d6 iLf6 17 ~fd1 e4 was S ... e5!
rather murky in Lemer-Ahlander, Berlin
1995.
9 e3
Pia Cramling suggests the interesting 9
~d2!?, intending e2-e4 next move.
9 ... 0-0 10 .Jid3 dxe4 11 .Jixe4 e5 12 0-0
exd4 13 exd4 liJd7!?
13 ... ttJc6 14 dS exdS 15 ttJxdS would have
been slightly better for White according to
Cramling. Now 14 dS is met by 14...lLJcs 15
'iVc2lLxc3!
14 J.d3 :d8 15 .:[fe1 liJfS?!
The text is a touch passive. Crarnling
suggest 1s ... lLJb6, aiming for the dS-square. This seems a very precise way to equalise.
16 liJe5!? :xd4 9 dxe5 'fia5 10 .Jixe4 'fixe5 11 tiJe4
Risky. Cramling recommends 16...i.xes 'iia5+ 12 ~e2!?
17 dxes iLd7! 18 'iVxb7 l:tab8 19 'iVf3 l:txb2 12 lLJfd2 iLe7 13 iLbs+ lLJd7 is equal
which seems fme for Black. Now things get a according to Korchnoi.
little more hairy for Black. 12 ... .Jie7 13 g4 tiJd7?!
17 lDe4 'fidS 18 l2Jb5 :d7 19 .te4 a6 20 13 ... iLd7! was better according to
lDbd6 :bS 21 .Jie6 :e 7 22 tiJeS .Jid4 23 Korchnoi, when 14 gs!? (14 a3 iLc6 15 b4
'fig3 bxe6 24 tiJf6+ ~hS 25 'fixbS gxf6 'iib6 16 tLJes lLJd7 17 tLJxc6 'iVxc6 is equal
26 tiJd6 e5 27 'fixeS 'ifxeS 2S tiJxeS :e7 according to Tsesarsky) 14...hxgs 15 ~hg1
29 ttJb6 .Jixb2 30 :dS ~g7 31 tiJe4 .Jid4 iLbs 16 tLJexgs iLxc4+ 17 'iVxc4 tLJc6 18
32 :b1 llJd7 33 tiJd6! :a7 34 :b3 f5 35 ~ad1 ~d81eads to equality.
~f1 ~f6 36 !:teS lDe5 37 :b7 :xb7 38 14 a3 lDf6 15 tiJed2?!
lDxb7 e4 39 :eS e3 40 tiJa5 tiJd3 41 Passive. 15 b4 was better according to
tiJb3 .Jie5 42 tiJe5! tiJxe5 43 :xe5 .Jid4 Korchnoi.
44 :e6 a5 45 f4! ~g6 46 ~e2 ~h5 47 1 5 ... .Jid7 16 :hg 1 :e8 17 'iib3 'fib6 18
:e4 .Jif6 4S :e7! ~g6 49 ~d3 .Jih4 50 'iixb6 axb6 19 lbe5 J.a4 20 :ac1 .Jid6
:xe3 .Jie 1 51 :e8 .Jib4 52 a4 f6 53 :e6 21 tiJdf3 <it>e7 22 h4 :hdS 23 g5lDd7 24
~f7 54 :b6 J.e1 55 :b7+ <it>g6 56 :e7 lbxd7 :xd7 25 .1l.d3 :de7 26 ':xc7+
e5 57 fxe5 fxe5 5S :xe5 J.b4 59 ':b5 ':xc7 27 tiJd4 .ite5 2S %:g4 il.d7 29 lIe4
150
Systems with ~xf6
f6 30 gxh6 gxh6 31 f4 i..xd4 32 :'xd4 ...lDe5xf3++wins a piece for Black. 15 0-0-0

lDb6 16 'iid6 l::dS also doesn't cause Black
l:cS 33 l:b4 bS 34 'ittf2 'ittd6 35 :'d4+
any problems.) not 12.....ta6 13 bxc4 bxc4 14
~c7 36 i..e4 ~e6 37 ~xe6 ':e2+ 38 <it>f3
l::d1 lDc6 15 'Wc3 0-0 16 i..xc4 i..xc4 17
bxe6 39 b4 l:e3 40 f5 exfS 41 ~f4 :'xa3
'iixc4 'iia5+ IslDd2 l::fes 19 f4 :tabS 20 0-0
42 'ittxf5 :'xe3 43 'ittxf6 :'e3 44 ~eS h5
45 l:e4 eS 46 ~d5 cxb4 47 :'xb4 'ittb6which was very nice for White in Nikolic-
Yusupov, Linares 19S5, but 12 ...lDc6! 13
48 ~d4 l:eS 49 ':b 1 :'e4+ 50 'ittd3 l:xh4
bxc4 .i.g4! when suddenly White had very
S1 'ittc3 l:c4+ S2 'ittb3 h4 53 ':g1 ':c6 54
l:gS l:h6 0-1 big central problems. In the game, 14 i..e2?
_ - - - - - - - - - - - - - - - . was played when 14....i.xf3 15 .i.xf3 lDxe5
Game 94 16 lLlc7+ 'iixc7 17 ba8 0-0 followed by
I.Sokolov-Van der Sterren ...ttJe5xc4 was ovetwhe1ming for Black. 14
Dutch Ch.) Rotterdam 1998 lLlxe7 'iixe7 15 cxb5 .txf3 16 gxf3 ltJd4 17
0-0-0 0-0 also does not look so comfortable
1 d4 d5 2 e4 e6 3 ~e3 ~f6 4 ~f3 i..e7 for White, so 14 'Wc3 is the best way to play,
S i..g5 h6 6 i..xf6 i..xf6 7 "d2!? but it isn't wonderful for White.
11 exf6 dxc3 12 "e3+ ~f8 13 "xc5+
~g8
7 ... dxc4!
Here 7...c5 8 dxc5 dxc49 'i'xd8+ <it>ms
10 0-0-0+ ~e7 11 lLle4 .i.d7 12 .i.c6 13 14 i..xc4
lLlxf6 gxf6 14 .txc4 :c8 was almost equal in Instead of this, 14 'i'e7!? 'i'xe7 15 dxe7
I.Sokolov-Azmaiparashvili, Antwerp 1998. ~h7 16 bxc3 :te8 17 0-0-0 lLlc6 18 .i.xc4
However, 7...dxc4! is much more active and ':xe7 19 :he1 .i.e6 20 .txe6 fxe6 and 14
makes good use of the position of the bishop fxg7 ~g7 15 'i'e5+ <it>h7 16 'i'xc3 :e8+ 17
onf6. .te2, meeting 17...'i'd3 with 18 ttJd4 ltJc6 19
8 e4 cS! 9 dS exdS 10 eS d4!? :td1!, are both given as better attempts for
This is not bad, but in a recent (1998) the advantage by Ivan Sokolov. In the game
Gennan Bundesliga game between P Nikolic Black easily solves his problems but later
and King, Black played much more strongly blunders the game away.
with 10...~e7 11 lDxdS b5 and after 12 b3 14... cxb2 1S l:b1 ~e6 16 fxg7 'ittxg7 17
(12 a4 ~b7 13 lLlxe7 'iixe7 14 axb5 lLld7 0-0 "f6 18 "a3 i..g4 19 ~d2 l:ad8 20
gives Black good counterplay according to l:xb2 ~eS 21 "g3 'itth7 22 :b5 ~xc4
Yusupov, as after 15 ~xc4 ..txf3! 16 gxf3 23 ltJe4 "g6 24 'ii'xg4 "xg4 25 ~f6+
ltJxe5 the double threat of ...lbe5xc4+ and ~g7 26 ~xg4 hS?? 27 ':g5+! 'ittf8 28
151
lLlf6 lId6 29 lLlxh5 lIdh6 30 tbg3 l:.xh2 .ltfl-d3 before moving the bishop again to
31 :c5 tbd6 32 :d1 tbeS 33 lIcS l:.2h6 play .ltd3xc4.
34 :e1 :e6 35 :xe6 fxe6 36 tbe4 ~7 Qiestion 10. Aha, the 'fight for the tempo'
37 :b8 b6 38 :a8 :h5 39 :xeS 1-0 again!
We now examine the main lines which can Anmx:r 10. So you did read the Orthodox
arise from both the 5...h6 6 .ltxf6 .ltxf6 7 e3 chapter then! This is another example of the
0-0 and 5... 0-0 6 e3 h6 7 .ltxf6 .ltxf6 move interrelations between so many of the QGD
orders. lines. Having established these basic
principles, we shall now examine White's
Game 95 choices, starting with 8 'iWd2 (8 'ii'c2 is the
Gabriel-Bonsch subject of Game 98 and 8 :c1 of Games 99
Bad Homburg 1996 and 100). Instead 8 'ifb3 c6 9 :tdl ltJd7 10
i.d3 :tb8! 11 'ii'c2!? (11 0-0 b5! 12 cxdS cxdS
1 tLlf3 d5 2 d4 tLlf6 3 c4 e6 4 lLlc3 ~e 7 13 :tel a6 14 ttJe2 i.b7 15 i.bl e5 16 dxe5
5 ~g5 h6 6 ~xf6 ~xf6 7 e3 0-0 ltJxe5 was fme for Black in Piket-Yusupov,
QIestion 7. What should White aim to be Dortmund 1994) 11.....a5 12 ltJd2 a6 13
doing? ltJb3 'ii'd8 14 e4 dxc4 15 hc4 b5 16 i.e2 e5
Ar13lW' 7. A good question! We flrst must was quite unclear in Volkov-Asrian, Minsk
start by stating a few obvious principles: 1998.
1. Black's ultimate idea, as in all QGD S1i'd2
lines, is to organise a central break; ...c7-c5 is
nonnal, but as we have seen from Orthodox
lines, ... e6-e5 is also quite frequent. This is
particularly tempting here as Black's bishop is
well-placed on f6 to support ...e6-e5.
2. There are two scenarios for these
breaks: Black will either play ...c7-c5
immediately (though then he will have to
accept an IQP after c4xdS and d4xc5) or he
will take fIrSt on c4 before playing either ...c7-
cS or ... e6-e5.
QIestion 8. One question, which break
should Black be aiming for, ... c7-cS or ...e6-
e5? Qiestion 11. What is the idea behind this
Ar13lW' 8. It depends very much on the move?
position, of course, but ...c7-c5 is the easiest Anmx:r 11. Let's see how it fits in with our
to achieve (... e6-eS still requires some principles:
preparation) so White should concern 1. White continues the fight for the
himself with this one fIrst. tempo: that's good!
So White wants to take the joy out of ...c7- 2. The fight against Black's central breaks:
c5 for Black, but he also must be careful of with 8 "d2, White frees dl for his queen's
when to develop his light-squared bishop. rook; if Black does accept an IQP, White will
QIestion 9. What do you mean? be able pressure it with his major pieces
Ar13lW' 9. Black may play ...dSxc4 at any extremely quickly. Moreover, White defends
moment to prepare one of his central breaks; his knight on c3.
White doesn't want to waste a tempo playing QIestion 12. Umm, it seemed to be doing
152
Systems with ~xf6
fme already! 10 or 12 moves in.

An..mer 12. Yes, but after ...c7-c5, d4xc5 QIestion 17. So what can Black do here?
... i.f6xc3+ was always an option, hurting A nsr.rer 17. Black's basic aim is to gain
White's queenside structure and removing space for his pieces - this is the rationale
pressure from dS. Now this no longer has behind his desire for central breaks: by
the same effect as White can simply pushing pawns forward and opening lines,
recapture on c3 with the queen. Black's pieces gain new avenues and squares.
8 'i¥d2 has other dreams moreover... You Therefore, the following plans all have some
see that pawn on h6? logic:
Qtestion 13. What about it? 1. The T arrasch option: ...c7-cS without
An..mer 13. It is a weakness in Black's first capturing on c4. This is slightly risky as
kingside. If White were to push his h- and g- the resulting Black IQP (after c4xciS ... e6xciS,
pawns and get in g2-g4-g5 then Black will d4xcS) will lack the support of a black knight
have some problems to contend with. on f6.
QIestion 14. Aha, and White's king... 2. The Chigorin option: ...ltJb8-c6
A nszrer 14. Could go queenside, thanks to followed by ...d5xc4 and ... e6-eS. We see this
8i1d2! in the game continuation. I like this idea as it
8 ...lLlc6!? makes very active use of the bishop on f6.
3. The QGA option: ... a7-a6, intending to
gain queenside space with ... d5xc4 and ... b7-
bS before fmally breaking with ...c7-cS (see
Game 96).
4. The Slav option: ... a7-a6, ... c7-c6 and
...b7-bS.
QIestion 18. This looks very peculiar!
Ansr.rer 18. Black takes a slightly different
way of solving his central problems; he
reasons that any IQP posltIon is
unfavourable for him, but neither does he
want to concede any central ground by
playing ...dSxc4.
QIestion 15. Oh, this looks odd! Qlestion 19. That sounds like a tricky
An..mer 15. Now we should look at this dilemma: how can he break in the centre
position from Black's side to see the range of then?
his possibilities. As with the Orthodox Ansr.rer 19. Black decides that he cannot
system, Black's untouched queenside gives do anything in the centre unless he forces
him the chance to play a wide variety of White to release the central tension - the
queen's pawn structures. conflict between the pawns on c4 and dS. By
Qiestion 16. I thought this was a QGD! achieving ...b7-bS, Black challenges the c4-
An..mer 16. It is, but that is the beauty of pawn and forces White to make a decision:
this opening. Nearly all other queen's pawn 1. If White pushes c4-c5, then the
openings gain a definite character from the pressure is released from the ciS-pawn,
start because they all involve the concession making ... e6-eS easier to achieve.
of the centre with ...dSxc4 in different 2. If White takes on dS, Black recaptures
settings. As the QGD holds the centre until with the c-pawn - the exchange has freed
much later, these possibilities are still viable Black's position and we now have an
153
Exchange Slav structure where Black is quite idze, Women's World Championship 1995,
happy to have the two bishops. when now instead of 16...nc8? 17 'iVd3 c6 18
3. The Semi-Slav option: ...c7-c6, d6 ne8 19 0-0, Ftacnik recommends 16...c5!
intending ...ttJb8-d7 and an eventual ...dSxc4 17 dxc6 (17 'iie4 'ilVc8 18 d6 :a6! is unclear
and ...e6-e5. This is the main plan. The close according to Ftacnik) 17...bxc6 18 'ilVe4 i.d7
resemblance between this system and the with equal chances.
Moscow variation of the Botvinnik system (1 14 0-0 It)b5 15 It)e4 It)d6 16 .ii.d3 .ii.b5?
d4 dS 2 c4 c6 3 ttJf3 ltJf6 4 ltJc3 e6 5 .tg5 h6 A rather passive move: 16...:c8 is better
6 i.xf6 'iixf6) must also be noted according to Altennan, though he considers
4. The QGD option: ...b7-b6 and ...i.c8- 17 liJfd2! i.e7 18 'iVb3 to be slightly better
b7 with ...dSxc4 at some stage to keep the for White.
long diagonal open. 17 ltfd1 i.xd3 18 'iVxd3 liJxe4 19 'ii'xe4
Q«stion 20. And which is best? 'ile7 20 lbd2 :tfd8 21 'ilc4 c6 22 dxc6
Amut?r 20. Nobody knows! It all depends ltac8 23 'ile2 :xc6 24 ltxc6 bxc6 25
on the specific circwnstances. :ic1-.b4 26 It)c4 a5 27 'iVe1 lta8 28 b3
9 l:tc1 c5 29 'iVxb4 cxb4 30 e4 a4 31 <itf1 axb3
White waits for ...dSxc4 and places his 32 axb3 :ia2 33 :td 1 ltc2 34 ltd3 <itf8
rook on the soon to be opened c-file. Since 35 <ite1 h5 36 g3 h4 37 ttJe3 l:b2 38
Black's knight stands in front of the c7-pawn, 4Jg4 hxg3 39 hxg3 <ite 7 40 4Je3 <ite6 41
that pawn can easily become a target. liJf5 i.e7 42 It)xg7+ ~6 43 lbf5 .ii.c5 44
9 ... a6! f3 <ite6 45 <itd1 .ii.e7 46 f4 f6 47 ttJe3
Black also waits. exf4 48 gxf4 .ii.c5 49 liJf5 .ii.e7 50 l:d5
10.i.e2 .ii.f8 51 It)d4+- ~f7 52 ltd7+ ~e8 53 ltb7
Perhaps 10 b3!? .i.d6 54 e5 fxe5 55 lbf5 .ii.f8 56 fxe5
10... dxc4 11 .i.xc4 e5 12 d5! 4Je7 13 :ixb3 57 ltb8+ <itf7 58 e6+ ~f6 59
.c2! ltxf8+ ~-~
Game 96
Gabriel-Lputian
Germany-Armenia match 1996
1 d4 e6 2 liJf3 4Jf6 3 c4 d5 4 ttJc3 .i.e7
5 i.g5 h6 6 .i.xf6 .i.xf6 7 e3 0-0 8 'ild2
a6!? 9 :ic1
This is a calmer move than 9 0-0-0, when
9...dxc4 10 .ltxc4 b5 (10...liJd7 11 h4 b5 12
.td3 .ltb7 13 tDe4 .te7 14 .tc2 c5! was also
fme for Black in Finegold-Lputian, Las Vegas
1994) 11 .td3 c5 12 liJe4 cxd4 13 ltJxf6+
With these moves, White points to the 'i'xf6 14ltJxd4 .tb7 15 f3ltJd7 was perfectly
defects of ...ltJb8-c6. The dS-pawn holds okay for Black in Kutirov-Azmaiparashvili,
back the c7-pawn and White doubles on the Stnunica 1995.
c-fUe against it. 9 ... c6!? 10 e4 b5 11 e5 i.e7 12 c5 f6
13 ....i.d7 Although White's posltIOn seems
13 ...tDb5!? 14 tDxb5 axb5 15 il.b3 i.g4 16 overwhelming, Black's nibbling at the centre
tDd2 was the game Zso.Polgar-Chiburdan- does cause White some problems on the
154
Systems with ~xf6
dark squares.
Ansr.rer 21. Indeed he does, but by ridding

13 ltJe2 'fIe8 14 h4 ttld7 15 'fIe3 fxe5 16 himself of the obligation of defending the
tLlxe5 ltJxe5 17 dxe5 'fId8 18 ltJd4 'fIa5+ dS-pawn, Black gives himself much greater
19 'fIe3 'fIe7 20 ttle2 b4 21 'fIe3 'fIa5 22 flexibility in his development.
b3 .i.d7 23 1:e2 .i.e8 24 h5 d4 25 ttlxd4 Instead 8...cS!? 9 cxdS cxd4 10 ttJxd4 exdS
.i.xe5 26 'fIe4 .i.xd4 27 1i'xd4 l:d8 28 11 i.bS ttJc6!? 12 i.xc6 bxc6 13 ltJxc6 'ii'd6
'fIe5 1:d5 29 'fIxa5 1:xa5 30 .i.e4 1:xe5+ 14 ttJd4 i.a6 offers Black some
31 1:e2 1:e5 32 1:h4 i.xh5 33 :1xe6 i.f7 compensation for the pawn according to
34 1:d6 .i.xe4 35 l:xe4 1:xe4 36 bxc4 AlteIman.
1:f6 37 :1d8+ :f8 38 1:d6 :1t6 39 :1d8+ 9 i..xc4
~f7 40 l:e8 a5 41 1:e7+ ~g6 42 e5 ~f5 CzeIWonski-Krivonosov, Lubniewice
43 ~e2 g5 44 ~e3 ~e5 45 :1d7 a4 46 1994, saw the even more violent 9 0-0-0, but
f3 h5 47 1:g7 1:e6 48 ~d3 ~f4 49 g3+ after 9...cS 10 h4 cxd4 11 exd4 bS! 12 ttJxbS
~xg3 50 lbg5+ ~h4 51 f4 :e1 521:g6 i.b7 13 ttJeS ttJc6 14 f4 a6 15 ttJaJ c3! 16
:1a1 53 :1xe6 1:xa2 54 1:a6 b3 55 e6 bxc3 ltJxeS 17 fxeS i..xh4 Black had a big
:a1 56 e7 b2 57 :b6 1:e1 58 :xb2 advantage.
l:xe7 59 ~e4 a3 60 1:h2+ ~g4 61 :1g2+ 9 ...ltJd7
~h3 62 :a2 :e4+ 63 ~f5 :e5+ 64 ~e4 In Hillarp Persson-McDonald,
1:e3 65 f5 ~g3 66 f6 1:f3 67 ~e5 h4 68 Hampstead 1998, Black tried the interesting
~e6 h3 69 f7 h2 70 1:a1 %-% 9...cS, which seemed sufficient for a draw
Before moving on, let us take a quick look after 10 dxc5 ttJd7 11 ttJe4 i.e7 12 :d1 'ii'c7
at some rather offbeat eighth move 13 b4 as! 14 aJ axb4 15 axb4 b6 16 ttJd6
alternatives for Black. :d8 17 ttJb5 'ii'b7 18 ttJd6 'ii'c7 19 ttJbS
.......- - - - - - - - - - - - - . 'ii'b7 20 ltJd6.
Game 97 100-0-0!?
Sadler-Van der Sterren Several other moves have also been tried
Linares Zonal 1995 here:
a) 10 0-0 cS 11 :fd1 cxd4 12 ttJxd4 ttJb6
1 d4 d5 2 e4 e6 3 tLle3 ttlf6 4 ~g5 .i.e7 13 i.e2 i.d7 14 :ac1 i.xd4 15 'ii'xd4 i.c6
5 e3 0-0 6 tiJf3 h6 7 .i.xf6 .i.xf6 8 'iVd2 was agreed drawn in Bacrot-Dorfman,
dxc4!? French Championship, Meribel1998.
Qiestion 21. Doesn't Black just give up the b) Tukmakov-Bender, Zadar 1997, was
fight for the tempo like this? also equal after 10 :d1 cS 11 ttJe4 cxd4 12
165
exd4 ttJb6 13 ~b3 i.d7 140-0 i.c6 15 lIfe1 43 liJxh5 ~e5 44 liJf4 liJe4 45 :d5+ ~f6
i.d5. 46 l:d4 ':xa2+ 47 ~e3 liJg5 48 liJd5+
b) 10 h4 is quite a dangerous alternative. ~g7 49 lld1 liJe6 50 g4 a5 51 f4 b4 52
After 10...e5 (10... c5 11 g4 [11 0-0-0 cxd4 12 l:b1 :a3+ 53 ~e4 :a2 54 ~e5 l:e2+ 55
exd4 ttJb6 13 i.b3 i.d7 14 g4 i.c6 15 "e3 ~d6 ':d2 56 ~c6 b3 57 f5 ~d4+ 58
gives "White a dangerous attack according to ~c5 b2 59 ~c3 ~f3 60 liJa4 ~e5 61
Van der Sterren] 11...bS!? 12 i.d3 gives good ~b5 0-1
attacking chances according to Epishin) 11
0-0-0 exd4 12 exd4 ttJb6 13 .tb3 c6 14 'ifd3 Game 98
ltJdS 1SltJeS, as in Epishin-Faibisovic, USSR Van Wely-Vaganian
1985, lS ...ttJc3 16 bxc3 i.xeS 17 dxeS i.e6 Yerevan Olympiad 1996
would have kept Black's disadvantage to a
minimwn according to Epishin. Maybe Black 1 d4 d5 2 c4 e6 3 liJc3 .i.e7 4 ~f3 ~f6
should try 10... g6 as in the game, since the 5 .i.g5 h6 6 .i.xf6 i.xf6 7 e3 0-0 8 Vi'c2
inclusion of h4-hS and ...g6-gS is not clearly c5 9 dxc5 ~c6 10 O-O-O!?
to White's advantage. Very sharp. Instead 10 cxdS exdS 11 0-0-0
10 ... 96! 11 h4 .i.g7 12 ~b1 a6 13 i.b3 i.e6 12 ttJd4 lIc8 13 i.bS ltJe7! 14ltJb3 b6
1i'e7 15 i.a6 lIb8!? (lS ...:c6 16 i.bS :c8 17 i.a6
13... cS 14 dS! is slightly better for White. is just a draw by repetition) 16 ltJe4 i.eS 17
Black is organising himself very carefully f4 "c7 and now 18 ltJc3 i.xc3 19 'ifxc3
here, and it is difficult for White to get at his ..c6 20 ltJd4 "xcS is unclear according to
opponent's position. In the game, things VanWely.
soon went wrong for me.
10 ... ~b4 11 'i'a4 a5!?

14 l:c1 b6 15 ~e2 c5 16 ~f4 i.b7 17 11...hc3 12 bxc3 ltJa6 13 cxdS exdS 14
d5 exd5 18 .i.xd5 i.xd5 19 1i'xd5 ~f6 i.xa6 bxa6 15 c4 is clearly better for White
20 'i'c4 'i'e4+ 21 ':c2 l:fd8 22 ~c1 (Van Wely).
'i'xc4 23 ':xc4 l:ac8 24 l:d1 ':xd1+ 25 12 cxd5 exd5 13 ~d4 .i.g4 14 i.e2
~xd 1 b5 26 l:c2 ~e4 27 ~e2 ':d8 28 i.xe2?!
~e1 c4 29 ~f3 ~c5 30 ~d4 i.xd4 31 Black should have gone in for 14...i.xd4!
exd4 l:xd4 32 93 ~g7 33 :c1 ~f6 34 15 lhd4 (1S i.xg4!? i.xc3 16 bxc3 ~a6 17
b3 ~e5 35 bxc4 l:xc4 36 ':d1 l:c2+ 37 "d4 "c7 18 i.f3 ~cS 19 i.xdS gives
~e3 l:c3+ 38 ~e2 l:c2+ 39 ~e3 l:c3+ Black reasonable chances for the pawn
40 ~e2 ~e4 41 1:d2 1:a3 42 h5! gxh5 according to Van Wely) 15...he2 16 ~xe2
156
Systems with Ji..xf6
"e7, when both 17 'it>b1 'ifxc5 18 :tel 'YWd6 9 a3!?

19ltJf4 and 17 a3 'ilxcS+ 18ltJc3 are unclear White has a wide choice here:
according to Van Wely. a) The naive 9 i.d3, as in Pohl-Vaganian,
15 ~dxe2! 'ii'e7 16 a3 1i'xe5 17 1i'b5 Gennan Bundesliga 1993, allows simply
lLla2+ 18 ~b1 lDxc3+ 19 lDxe3 'iic6 20 9...dxc4 10 i.xc4 tDd7 11 0-0 bS 12 i.d3 cS
'ii'xd5 'ii'a6 21 'ii'd6 i.xe3 22 'ii'xa6 :'xa6 13 ttJe4 cxd4 14liJxf6+ liJxf6 lSltJxd4 i.b7
23 bxe3 with equality. In subsequent games White has
Now White stands clearly better. preferred to keep on fighting for the tempo.
23 ...:te8 24 l:d3 :'g6 25 g3 ':ge6 26 b) 9 cxd5 exdS 10 i.d3 c6 11 h3 ltJdl 12
~b2 95 27 l1c1 ~g7 28 l:d5 l1c5 29 0-0 i.e7! 13 a3 i.d6 14 :tel ttJf6 (14...:te8!?)
l:xe5 l:xc5 30 l:d 1 l:f5 31 f4 gxf4 32 was fme for Black in Bacrot-Giorgadze,
exf4 1:h5 33 l:d2 ~f6 34 ~b3 lite6 35 World Championship 1997.
c4 l:e5 36 a4 b6 37 Iitc3 :th5 38 :b2 c) LSokolov-Short, Groningen 1996, saw
~d7 39 l:xb6 l1xh2 40 e5 h5 41 lite4 9 'i'c2 c6 10 ..td3 bS! (switching to the Slav
l:g2 42 l1d6+ ~e7 43 1:h6 :xg3 44 plan) 11 c5 tDd7 12 e4 eS! 13 exdS exd4 and
l1xh5 l:a3 45 1:f5 l:xa4+ 46 ~d5 l:a1 47 now, instead of 14 liJe2? 'i'a5+ 15 'i'd2 b4!
:xf7+ Iitb8 48 l:f8+ rJi;c7 49 f5 hld1+ 50 which was clearly better for Black, Ivan
~c4 l1e1+ 51 ~b5 1:b1+ 52 rJi;xa5 lite6 Sokolov gives 14ltJe2 as unclear.
53 f6 1-0 9 ... e6 10 i.d3 ~d7 11 0-0 b5 12 exd5
~--------------------------~ exd513e4
Game 99 Oll-King, London Lloyds Bank 1994, saw
Gligoric-Vaganian the quieter 13 i.b1 g6 14 liJe2 tDb6 15 liJf4
Zonal 1998 ttJc4 16 a4 bxa4 17 'ii'xa4 i.d7 18 'i'c2 'i'e7
' - - - - - - - - - - - - - - . . with a reasonable position for Black. The text
1 d4 e6 2 e4 d5 3 lLlf3 lLlf6 4 lDe3 i.e7 also holds few fears for Black.
5 .tg5 h6 6 .txf6 i.xf6 7 e3 0-08 ':c1 13 ... dxe4 14 .txe4 1:.b8 15 'ii'e2 b4 16
axb4 l:xb4 17 ~a2 :'b8 18 :tfd1 'ii'b6 19
:'e2 ':d8 20 ~e3 a5 21 ':ed2 i.a6 22
'ii'e3 :tbe8 23 .tb1 i.b7 24 i.e4 .ia6 25
.tb1 .te4 26 h4 i.b3 27 1:.e1 lLlf8 28
.ta2 .i.xa2 29 ~xa2 lLlg6 30 g3 lLle7 31
94 h5 32 'ii'e4 hxg4 33 'ii'xg4 lLlf5 34 d5
lLlh6 35 'ii'h3 exd5 36 hlxd5 'ii'xb2 37
':xa5 :'e4 38 'ii'f1 :tg4+ 39 ~h1 'fib740
'ii'e2 ':f4 41 ':a3 lLlf5 42 ~g1 1:g4+ 43
~f1 lLld4 44 'ii'd3 ':f4 0-1
Game 100
Gelfand-Kramnik
The most natural move for White in the Dortmund 1997
fight for a tempo. White defends his knight
on c3 and brings his rook to the c-flle in 1 d4 lLlf6 2 c4 e6 3 lLlf3 d5 4 lLlc3 .te7 5
order to discourage ...c7-cS. i-g5 0-0 6 e3 h6 7 .ixf6 .txf6 8 1:e 1 e6
8 ...a6 Black defends his dS-pawn in order to
The QGA plan! The alternative 8...c6 is allow the queen's knight to be developed to
considered in the next main game. d7.
157
9.i.d3
Game 101
Kramnik-Short
Dortmund 1995
1 It::)f3 d 5 2 d4 It::)f6 3 c4 e6 4 It::)c3 i.e 7
5 i.g5 0-0 6 "'c2!?
6 :'el h6 7 .t.h4 transposes to Korchnoi-
Short, World Championship 1997, when
7... dxc4! 8 e3 c5 9 .i.xc4 cxd4 10 tLlxd4 ..td7
11 ..tg3 (11 0-0 ltJc6 12 ttJb3 a6 13 ..te2
ltJdS 14 .i.g3 tLlxc3 15 :Xc3 tLlb4 16 'i'b1
1i'b6 17 :'cel tLldS 18 e4 tLlf6 was equal in
Ivanchuk-KhU1l11an, Lvov 1988, whereas
As the ...c7-cS central break will now take Kharitonov-Beliavsky, USSR Championship
two moves, White abandons the fight for the 1988, was also level after 12 ltJf3 1i'b6! 13
tempo: he will regain it whenever Black plays tLla4 1i'c7 14 .i.e2 :'fd8 1S a3 .i.e8 161i'c2
...c6-cS. l:.ac8) 11...ltJc6 12 ltJdb5 eSt? (12 ... a6 13
9 ... ~d7 10 0-0 dxc4 11 i.xc4 e5! tLld6 bS 14 .i.e2 1i'b6 1S a4 b4 was fme for
So Black aims for ... e6-eS instead of ... c6- Black in Piket-Van der Sterren, Antwerp
cS! 1997) 13 a4 a6 14lDa3 ..txa3 15 bxa3 1i'e7
12 h3 exd4 13 exd4 It::)b6 14 i.b3 :e8 16 ..th4 g5 17 i..g3 .i.e6 18 .i.xe6 1i'xe6 was
15 ':e1 .i.f5 pleasant for Black.
Siegel successfully neutralised Bacrot in 6 ... h6 7 Axf6!
the French T earn Championships 1998 after Here 7 ..th4 b6! 8 .i.xf6 (8 cxdS tLlxdS 9
IS .. .lhe1+!? 16 'ii'xe1 ..tfS 17 g4 ..td3 18 ..txe71i'xe7 10 tLlxdS exdS 11 :'el i..a6 12
ltJeS .i.xeS 19 dxe5 c5 20 1i'e3 c4 21 .i.dl 1i'xc7ltJd7 gives Black good counterplay for
'ii'e7 22 J.e2 ..txe2 23 1i'xe2 :'e8 24 :'el the pawn) 8....i.xf6 9 e4 tLlc6! 10 0-0-0 dxe4!?
~-~. (10...dxc4 11 e5 .i.e7 12 ..tc4 .i.b7 13 a3 is
16 g4 unclear according to Hjartarson) 11 1i'xe4
Portisch-Van der Sterren, Ter Ape11994, .i.b712.i.d3 g6 13 h4 :'b8 14 1i'g4 ..tg7 was
was also fme for Black after 16 :'xe8+ 1i'xe8 unclear in Ree-Hjartarson, Reykjavik 1984.
17 'ii'd2 1i'd7 18 l:tel as 19 a3 :'e8 20 l:txe8+ 7 ...i.xf6 8 ':d 1 g6!?
'ii'xe8 21 1i'f4 .i.e6 22 .i.xe6 1i'xe6 23 1i'b8+ 8... c6 9 e4! dxe4 10 'ii'xe4 intending ..tf!-
'ii'c8 24 1i'a7liJc4. d3 and h4-h5 shows the point of White's
16 ...i.e6 17 i.xe6 1:xe6 18 ':xe6 fxe6 idea, but 8...cS!? 9 dxcS 1i'aS 10 cxdS exdS 11
19 "'e2 "'e7 1i'd2 ..txc3 121i'xc3 1i'xc3+ 13 bxc3 tLla6!?
The position is equal. (13 ....i.e6 14ltJd4 l:tc8 IS e4 dxe4 16 tLlxe6
20 ':e1 1:e8 21 "'c2 "'f7 22 It::)e4 ':d8 fxe617 ..tc4 ~f718 :'d4 :XcS 19 :'xe4 :'c6
23 It::)c5 .i.xd4 24 lLlxd4 ':xd4 25 It::)xe6 20 f4 ~f6 210-0 ltJd722 g4 tLlc5 was agreed
l:d6 26 "'e4 l:d5 27 f4 ~d7 28 It::)d8 drawn in Li Wenliang-Liang Jinrong, Beijing
"'f6 29 "'e8+ ~h7 30 "'e4+ ~g8 31 1996) 14 e3 ltJxc5 1S :xdS b6 16 l:td4 ..tb7
"'e8+ ~h7 32"'84+ YZ-YZ 17 ..tc4 l:.ac8 18 ltJe5 ..txg2 19 l:.gl ..th3
We shall now examine the S... O-O move was absolutely fine for Black in San Segundo-
order in more detail by looking at the Van der Sterren, Linares Zonal 1995. Instead
interesting attempts 6 'ii'c2!? and 6 :'cl. of 18 tLle5, 18 0-0 tLle4 19 tDe5 has been
158
recommended, but 18...nc5! looks much for White according to Georgiev.
stronger with reasonable play for Black. 5... 0-0 6 .i..g5 transposes to Game 100
The game leads to a position reminiscent above. White can try simply to transpose into
of a Semi-Slav Moscow system, but with a an Exchange variation with 6 cxdS exdS 7
few less tempi for Black. .i..g5, but 6... ttJxd5 7 i.d2 b6 8 g3 .tb7 9
iLg24.Jd7 10 lDxdS .txdS 11 e4 iLb7 12 0-0
c5 13 iLc3 cxd4 14 ttJxd4 a6 1S :tfd1 1i'c7
was equal in Oll-Liang Jinrong, Beijing 1997.
6 dxc5 ~a6
9 e3
9 e4 dxe4 10 4.Jxe4 i.g7 11 i.e2 4.Jc6! is
fme for Black according to Kramnik.
9 ... c6?! 10 .id3 dxc4
10... ttJd7 11 0-0 a6 is slightly better for 7 cxd5
White according to Kramnik. Two other moves have also been tried:
11 .ixc4 ltJd7 12 h4!? .ig7 13 a3 'i'e7 a) Kramnik-Short, Novgorod 1996,
14 .ia2 b6 15 .i.b1 h5 16 0-0 .ib7 17 continued 7 g3 0-08.tg2 dxc49 0-0 1i'aS 10
ltJg5 :fd8 18 .ia2 l2Jf6 19 e4 ltJg4 20 e5 ttJe4 ttJxcS 11 4.Jxf6+ .txf6 12 4.JgS .txgS
l:d7? 21 ltJe2! :ad8 22 ..txe6! fxe6 23 13 i.xg5 4.Ja4 14 i.d2 'iib5 15 :tfel 1i'xb2
'i'xg6 ltJxe5 24 'i'h7+ ~f8 25 ltJf4 1-0 16 1i'xb2 ttJxb2 with an unclear position.
And fmally a look at the accelerated b) Eingom-Beliavsky, $ochi 1986, saw 7
version of this idea .tg5 1i'aS 8 e3 tDxc5 9 ttJd2 dxc4 10 i.xc4
(10 ttJxc4 4.Jd3+ 11 .txd3 1i'xg5 12 0-0 0-0
Game 102 13 ttJe4 and now 13...'Wh5 would have been
Ehlvest-Lputian fme for Black in Eingom-Smyslov, $ochi
Yerevan 1996 1986) 10...0-0 11 i.h4?! (11 iLf4 ttJhS 12
----------------- .tg3, intending a2-a3 and b2-b4 is White's
1 d4 e6 2 tDf3 d5 3 c4 ltJf6 4 ltJc3 ..te7 best according to Eingom, but it doesn't
5 'WIc2 c5!? look anything special) 11...i.d7 with good
Kramnik-Kir.Georgiev, Yerevan Olympi- chances for Black.
ad 1996, saw 5...dxc4 6 e4 ltJc6 7 e5 4.Jb4 8 7 ...~xd5 8 e4 ~db4 9 'i'a4+ .id7 10
"'1 ttJfdS 9 i.xc4 c5! (9 ...4.Jb6 10 i.e2 .td7 'i'd1 ~xc5 11 a3 ~c6 12 ..te3 a5 13
11 0-0 iLc6 12 a3 ltJ4d5 13 ltJe4 with an .ib50-0 140-0 'WIc7
edge in Ehlvest-Yusupov, Vienna 1996) 10 Here Black has equalised.
dxc5 'WaS 11 0-0 'iixeS 12 llJxd5 llJxdS 13 15 ::'c1 :fd8 16 'iie2 ~e5 17 ltJxe5 'WIxe5
1i'e4 i.d7 when 14 'ii'g4! ~f8 (14...0-0-0 15 18 f4 'WIb8 19 e5 b6 20 b4 axb4 21 axb4
ltJgS!) 15 iLd2 h5 16 'iie4 is slightly better .ixb5 22 ltJxb5 lba6 23 1:c6 Yz -Yz
15-9
Summary
For the moment, the Ji.xf6 ideas do not seem anything special for White. For Black, I prefer
the immediate S...h6 move order.
1 d4 d5 2 c4 e6 3 lbc3 lLlf6 4lLlf3 i.e7 5 i.95

5 'i'c2 - Game 102
5 ... h6
5...0-0 (D)
6 e3 h6 7 ~xf6 i.xf6 - Games 95 -100 (see below)
6'iWc2 - Game 101
6 J.xf6 .axf6 (D) 7 e3
7 'iWb3 - Game 92
7 'ilc2 - Game 93
7'iWdl - Game 94
7 •.. 0-08 'ii'd2
8 'iic2 - Game 98
8:tel (D)
8... a6 - Game 99
8...c6 - Game 100
8 ...lbc6
8...a6 - Game 96
8...dxc4 - Game 97
9 :c1 - Game 95
5 ... 0-0 6... ~xf6 8 'Ac1
160
CHAPTER TEN
5 i..f4 Variation
1 d4 d5 2 c4 e6 3 ~c3 ~f6 4 ~f3 .ie7 Qiestion 3. Aha, so by putting the bishop

5.if4 onf4...
Qiestion 1. What is the point of this move? Anszrer 3....White avoids all these freeing
A~ 1. White will soon wish to play e2- ideas with ...tDf6-e4! From this point of view,
e3 in order to develop his light-squared White makes sure that his opponent will
bishop and complete his kingside have to work a lot harder to make space in
development. In contrast to Black, he wishes his position. It also becomes much harder for
to ensure that his other bishop - the bishop Black to net the bishop pair. Note also the
on the same colour as his central pawn chain number of times that Wbite retreats his
- remains active, and outside the pawn chain. bishop to f4 in so many lines - the Exchange
aearly therefore, White has a choice be- variation with 10 h3 and 11 i..f4 being the
tween only two possible squares: f4 and gS. most obvious example.
Qlestian 2. But why put the bishop on f4 Qiestion 4. Wow, that sounds perfect!
rather than gS? Anszrer 4. Well, unfortunately not.
A~ 2. On gS, the bishop had two Prevention in the opening is always like
major accomplishments: trying to squash jelly - there's always one
1. By attacking the knight on f6, it part that seems to squinn away from you!
weakened Black's protection of his dS-pawn. Qiestion 5. And in this case...
2. It gave White the opportunity to divert Anszrer 5. Well, the problem in this case is
the black bishop on e7 from the f8-a3 that from f4, the bishop does not put any
diagonal by playing J..gSxf6. pressure on Black's centre. And since this
Both these points had the effect of pressure is missing, this makes it child's play
making it harder for Black to achieve the for Black to achieve his desired ...c7-c5
desired central freeing break ...c7-cS. freeing central break.
However, the presence of the bishop on gS QIestion 6. Doesn't this make Black's
allowed Black new resources based on ex- equalising task easier?
changing pieces in order to free his position, Anszrer 6. In a way. The whole variation is
in patticular the patent ...tDf6-e4! idea In obviously a lot less complicated strategically
order to avoid the exchange of pieces, White for Black than the S .i.g5 lines - he no longer
was often forced to give up the bishop pair. has to wony exactly how he's going to get in
1'61
his break - but White has a number of 5 ~f4 0-0

ingenious resources to keep on throwing Attention should be given here to
problems at his opponent. Crouch's 5... dxc4!? with the idea of 6 e4 (6 e3
Qiestion 7. Such as... ? ttJdS! 7 .i.xc4 ttJxf4 8 exf4 ttJc6! is fIDe for
Ar1..9lU'Y 7. Two kinds of things generally: Black according to Crouch, who suggests 6
1. The symmetrical structure where Black 1Wa4+!? as White's best try for an advantage)
has played ...c7-c5, and White has taken on 6...b5! 7 ttJxb5 .i.b4+ 8 .i.d2 .i.xd2+ 9 ttJxd2
c5 and Black has taken on c4. a6 10 lDc3 1Wxd4 11l'Dxc4 1Wxd1+ 12 ':xd1
lDc6 with maybe a slightly better ending for
White according to Crouch, but Black has
counter-chances.
6 e3 c5
Black achieves his natural freeing break.
Instead 6... ttJbd7!? is a favourite of Spassky's
which he has played with some success: 7 cS
(alternatively, 71Wc2 cS 8 dxcS ttJxcS 9 .i.e2
dxc4 10 i..xc4 a6 11 a4 i..d7 12 0-0 :c8 13
as and now 13...bS 14 axb6 1Wxb6 would
have led to equality according to Sokolov in
Van Wely-LSokolov, Elista Olympiad 1998;
7 cxdS tDxdS 8lDxds exdS 9 i..d3 .i.b4+ 10
On a full board of pieces, the advantage of lDd2 lDf6 11 0-0 .i.d6 12 .i.xd6 1Wxd6 13
the bishop on f4 against the bishop on c8 is 'iVb3 :e8 14 :fc1 c6 was equal in CHansen-
often enough to guarantee White a slight pull Spas sky, Malmo 1998) 7...c6 8 .i.d3 b6 9 b4
in these symmetrical positions. as 10 a3 .i.a6 11 b5 cxb5 12 c6 1Wc8 13 :el
Qiestion 8. Why? 1Wxc6 14 ttJe2 ttJcS! 15 dxc5 bxcS was very
Ar1..9lU'Y 8. Mainly because the bishop on f4 good for Black in I.Sokolov-Spassky, Malmo
takes away the natural c7-square from the 1998.
black queen. Since White has not allowed his 7 dxc51
opponent to exchange the dark-squared
bishops and free space for his queen, the
queen does not have e7 available either, and
with a white rook coming to the open d-file,
this can prove a little troublesome for Black.
2. Wmg pawn advances. In this system,
White is always flying down the wings with
such ideas as a2-a3 and b2-b4 or g2-g4-g5
andh2-h4.
So enough talking, let's get down to some
concrete lines!
Game 103
Sakaev-Beliavsky Qiestion 9. It looks a little odd for White to
European Club Cup 1999 be giving up the centre like this.
Ansrrer 9. First of all, this isn't a bad move
1 d4 lbf6 2 lbf3 d5 3 c4 e6 4 lbc3 ~e 7 in general tempo tenns, as Black's bishop is
162
5 ~f4 Variation
forced to recapture on c5. Black has 11 'iixd8 lhd8 12 .ltc7!? is White's other
therefore spent two moves getting his bishop attempt for an advantage. After 12...l:td7 13
to the c5-square, so White has won this little i..e5 b6 (alternatively, 13...liJxe5 14 ~xe5
version of the battle for the tempo! l:td8 15 0-0 [15 .lte2 liJf6 16 i..f3 ltb8 17
Secondly, capturing on c5 has two effects: ~e2 is slightly better for White according to
1. It opens up the d-flle against Black's dS- Dautov] 15...~f6 16 ltfdl .ltd7 17 liJxdl
pawn - White would now like to bring the ltxdl 18 lhdl ~xdl 19 ltdl liJf6 20 ~f1
queen's rook to d1 to put pressure against it. was agreed drawn in Topalov-Gelfand,
2. By removing the c5-pawn, White frees Vienna 1996; while 13 ...ltd8 14 ~e2..tdl 15
any obstacle to gaining queenside space - a2- lthdl ~xe5 16 ~xe5 ..te8 is another
a3 and b2-b4 will now be possible, expanding equaliser according to Beliavsky) 14 ~e4
White's position while gaining a further .lte7 (14... ~e5 15 liJxe5 ltc7 16 .lte2 and
tempo on Black's dark-squared bishop. now 16.....tb7! 17 ~c5 ltxc5 18 ~d3 ltd5
7 .....txc5 8 a3 ~c6 19 .ltf3 l:txd3 20 .ltxb7 ltad8 is ftne for
Here White has a wide choice. 9 b4 Black according to Dautov) 15 ..tc3 ..tb7 16
(Games 104 and 105), 9 ltd (Games 106 ..te2 ltad8 17 0-0 ~a5 18 ~ed2 ~c6 19
and 107) and 9 'iic2 (Games 108-110) are all ~e4 ~a5 20 ~ed2 was agreed drawn in
vel}' popular, but fIrst we shall deal with the Lobron-Lutz, Nussloch 1996.
quiet 9 .lte2. 11 .....te7 12 h4!?
9 ~e2 dxc4! 12 'iixd8 l:txd8 13 ..txe7 ~xe7 was
A good moment to play this move, nothing for White in M.Gurevich-Marciano,
regaining the tempo in this mirror-image Belfort 1997.
battle! 12 ...f6! 13 ..tf4 ~xf4 14 exf4 'tIc7 15
10 ..txc4 ~h5! g3 ~e5!
Now Black stands a little better but White
manages to hold the game.
16 ..te2 :d8 17 ..c2 ~xf3+ 18 i.xf3
:b8 19 0-0 b5 20 :ac1 ..tf8 21 :fe1 b4
22 axb4 :xb4 23 ~a2 "xc2 24 :xc2
:b6 2S ~g2 ..td7 26 :d2 :d6 27 ltxd6
..txd6 28 :d1 ..te7 29 ~c3 ~f8 30 :d4
~e8 31 ~e4 ..tb5 32 ..th5+ ~f8 33
:xd8+ ..txd8 34 ~c3 ..td7 35 ..te2 ~e7
36 ..tc4 ~d6 37 ltJe4+ ~c6 38 ~f3 h6
39 ~e2 ..tb6 40 b4 ..tc8 41 ~c3 ~d6 42
ltJe4+ ~d7 43 ..tb5+ ~c7 44 i.c4 ..tb7
4SltJc3 ~d6 46ltJb5+ ~d7 47 ltJc3 ..tc6
Qtestion 10. This looks sneaky! 48 ltJbS ..tdS 49 ..td3 ..tb3 SO ltJc3 ~d6
An.tUe'" 10. White's dark-squared bishop is 51 ltJe4+ 'iitc6 52 ltJc3 ~d6 53 ltJe4+
never safe in the QGD! If Black can gain the ~e7 54 ~c3 f5 5SltJb5 g6 56ltJa3 ..td5
bishop pair, then he even has chances to be 57 ltJb5 ~6 58 ltJd6 eS 59 fxeS+ 'iitxeS
better in the resulting symmetrical position. 60 ltJc4+ 'iitd4 61 ~xb6 axb6 62 ~d2
This exchange is particularly desirable here in ..te4 63 ..te2 gS 64 hxg5 hxgS 65 ..tb5
view of the cramping influence that the f4 66 gxf4 gxf4 67 ..te2 ..tdS 68 ..td3
bishop on f4 has on Black's queenside. i.c4 69 .i.c2 .i.b5 70 .i.b3 ..tc6 71 ~a2
11 ..tg5 %-%
163
ttJa2 22 ~c7 :tclc8 23 'ii'g4 'ii'xg4 24 hxg4

Game 104 ttJc3 which should be fme for Black.
Krasenkov-Karpov 17 ... g6
Polanica Zdroj 1998 17....llb2 18 :tc2 i..xa3 19 bs ttJe5 20
i..xh7+ <it>xh7 21 'ilh5+ ~g8 22 i.xe5 is
1 d4 ~f6 2 e4 e6 3 ~f3 d5 4 12Je3 i.e 7 clearly better for White according to
5 .i.f4 0-0 6 e3 e5 7 dxe5 .i.xe5 8 a3 Krasenkov.
~e6 9 b4 18 i.b1 ~e5?!
Krasenkov suggests that 18... i.b2! 19 ':c2
i..xa3 20 b5 ttJe5 21 i.xe5 'iWxe5 22 i..a2
'iWe723 .llxdS ':d7 would have given White
only a negligible advantage, although White
did manage to win from this position in the
recent game Nielsen-Vander Sterren,
Gennan Bundesliga 1998.
19 i.xe5 i.xe5 20 i.a2 a5 21 i.xd5 :d7
22 :c4 axb4 23 axb4 ~g7 24 b5 lIad8
25 e4 h5 26 "e3 'i'f6 27 g3 h4 28 :f1
hxg3 29 f4 i.e7 30 'i'xg3 'i'b6+ 31 ~g2
"f6 32 'i'e3 'ifxe3 33 :xe3 .i.b6 34
:fe1 :a8 35 ~f3 :a5 36 :b1 :a4 37
A very simple and thematic system. White h4 f5 38 :d3 fxe4+ 39 i.xe4 :'xd3+ 40
plays for a small edge based on his queenside i.xd3 i.e7 41 .i.e4 :a3+ 42 ~g4 ~a4 43
space and the isolated queen's pawn that he :d1 :c4 44 :d7+ ~f8 45 i.xg6 :xf4+
will create in Black's position. 46 ~g5 :e4 47 h5 i.f4+ 48 Iit>f6 :e8 49
9 ... i.e7 10 exd5 ~xd5 h6 ~g8 50 i.f5 :e8 51 i.e6+ ~h8 52
The other recapture 10...exdS!? is the .i.e41-0
subject of the next main game.
11 ~xd5 exd5 12 .i.d3 i.f6 13 lIe 1 Game 105
i.g4! Topalov-Yusupov
13 ...a6 had been thought necessary to Elista Olympiad 1998
prevent b4-b5, when 14 0-0 .lle6 transposes
to the game MGurevich-Peelen. Holland 1 12Jf3 d5 2 d4 12Jf6 3 e4 e6 4 12Je3 i.e 7
1998, in which the very typical manoeuvre 15 5 i.f4 0-0 6 e3 e5 7 dxe5 i.xe5 8 a3
':c5! g6 16 'iWb1 'iWe7 17 ':fc1 ':fd8 18 h3 ~e6 9 b4 .i.e7 10 exd5 exd5!?
~g7 19 a4! led to a considerable advantage It is normally considered best for Black to
for White. take the opportunity to free his position by
140-0 exchanging a pair of knights while he can.
14 b5 'iWaS+ 15 'iWd2 'iWxd2+ 16 ~xd2 With this move, however, Black intends to
tiJaS gives Black sufficient counterplay target the white queenside with ...a7-aS. After
according to Krasenkov. b4-b5 in reply, we have a typical IQP
14...•e7 15 h3 i.xf3 16 'i'xf3 :fd8 17 structure where White is left with an
:fd1 unsettled knight on the semi-open c-fLle and
Instead Van Wely-Van cler Sterren, no square on b5 to go to.
Andorra Zonal 1998, saw 17 l:c5 as 18 .tbS 11 i.e2 i.e6 12 liJd4 ::te8
axb4 19 axb4 ttJxb4 20 l:c7 'ii'e6 21 l:xb7 Instead Golod-Lputian, European Club
164
5 ~f4 Variation
Cup 1999, saw Black implement his idea 'it'd4 .i.d6.

earlier with 12... aS 13 ttJxe6 fxe6 14 bS ttJb8 10 ... exd5 11 i.g5?!
IS ~g4 ttJxg4 16 'iixg4 :f6 17 0-0 ttJd7 18 Too ambitious. Instead 11 b4 ~a7 12
:fdl tDb6 19 tDe2 :g6 20 1ih3 .i.d6 21 e4 ~e2 d4 13 exd4 ttJxd4 14 ttJxd4 'iixd4 was
when a draw was agreed. Black's position equal in Kramnik-Ivanchuk, PCA rapidplay
looks very rickety to me. 1994, as was 11 ~d3 ~g4! 12 0-0 d4 13 tDe2
13 0-0 a5 14 liJxc6! :xc6 15 'ii'd4! ~a7 14 tDfxd4 ~xd4 15 exd4 draw agreed as
By maintaining the pawn on b4, White in Horvath-Lutz, Elista Olympiad 1998.
maintains a stable advantage. 11 ... d4!
15... axb4 16 axb4 i.d6 17 i.xd6 'i'xd6
18 h3 'i'c7 19liJb5 'i'e7 20 'i'f4liJe4 21
liJd4 :b6 22 b5 h6 23 'ii'e5 'i'd6 24
'ii'xd6 liJxd6 25 :fc1 :d8 26 f3 ~f8 27
'iitf2 ~e7 28 i.d3 :c8 29 'iite2 i.d7 30
:xc8 l2)xc8 31 'iitd2 g6 3 2 ~c3 l:td6 33
l2)e2 1:e6 34 'iitd4 ~d6 35 l2)c3 l2)b6 36
:c1 :e8 37 e4 i.e6 38 e5+ 'iitd7 39 f4
h5 40 :a1 :c841 h4 :c7 42 g3 :c8 43
liJd1 l2)c4 44 1:a2 ~c7 45 liJc3 1:d8 46
l2)a4 :a8 47 l2)c3 :d8 48 :c2 ~b6 49
1:c1 ~a5 50 1:a1+ 'iitb6 51 i.e2 'iitc7 52
:b1 l2)d2 53 :b2 l2)c4 54 :b1 liJd2 55
b6+ 'iitc6 56 1:b4 l2)c4 57 l2)a4 :a8 58 12l2)b5?!
i.f3 l2)d2 59 i.d1 l2)c4 60 i.f3 l2)d2 61 12 tDe4 'iia5+ 13 b4 tDxb4 14 axb4
i.d1 l2)c4 % -% i..xb4+ 15 tDed2 tDe4 16 ~f4 dxe3 17 i..xe3
:d8 and 12 ~6 gxf613 lDe4 ~b6 both
Game 106 leave Black with a powerful initiative
Dreev-Short according to Ftacnik.
Linares 1995 12... dxe3! 13 'ii'xd8
13 :Xc5 exf2+ 14 ~e2 'iie7+ wins.
1 d4 l2)f6 2 c4 e6 3 liJf3 d5 4 l2)c3 i.e 7 13 ... exf2+ 14 'iite2 :xd8 15 i.xf6 1:e8+
5 i.f4 0-0 6 e3 c5 7 dxc5 ~xc5 8 a3 16 'iitd 1 gxf6 17 1:xe5 i.g4! 18 l2)c3
lbc6 9 lie1 a6!? l2)d4 19 i.e4liJxf3 20 ~c2 i.f5+ 21 ~b3
Black's alternatives here are considered in lbd2+ 22 'iita2 i.e6! 0-1
the next main game.
10 exd5 Game 107
10 b4 seems the best try: after 10...~e7 Gelfand-Karpov
(10 ...i..a7!? was tried in the recent rapidplay Polanica Zdroj 1998
game Gelfand-Topalov, Monaco 1999) 11
cxdS exdS, we have a position very similar to 1 liJf3 l2)f6 2 c4 e6 3 liJc3 d5 4 d4 i.e 7
the 9 b4line except that White has played the 5 i.f4 0-0 6 e3 c5 7 dxc5 i.xc5 8 a3
slightly superfluous :al-cl, which makes liJc6 9 :c1 dxc4
... a6-aS plans much more tempting for Black. The same simple treatment as after 9 i..e2.
Savchenko-Sturua, Berlin 1998, was fairly The advance 9...d4 10 exd4 (10 tDxd4 eS! 11
equal after 12 ~e2 ~e6 13 0-0 ttJhs lDb3 i.xa3! 12 bxa3 exf4 13 'iixd8 :xd8 14
(13 ...a5!?) 14 ~eS ttJxeS IS ~xeS ttJf6 16 exf4 ~e6 IS f3 as in Kramnik-Beliavsky,
165
Belgrade 1993, and now 15 .. .l1ac8 16 tDc5 Alterman-Kasparov, simultaneous(!}, Tel

tDaS! is equal according to Kramnik) Aviv 1998, where 10... ~e7 11 ttJd2 e5 12
10... ttJxd4 11 ttJe5!? (11 ~d3!? is also ~g5 d4 13 ttJb3 'tid8 14 i.e2 as 15 tDa4 g6
interesting) 11...b6 12 ~d3 ~b7 13 0-0 h6 14 16 ~xf6 .ltxf6 17 c5 ~e6 18 e4 'tie8! 19
b4 i.e7 15 ltJb5 ttJc6 (15 ... ttJxb5 16 bxc5 ttJb6 a4 20 ttJd2 l:taS was fme for Black
ttJdS 17 i..g3 i.g5 18 :c4 a6 is unclear
according to M.Gurevich) 16 tDxc6 ~xc6 17
ttJd6, as in MGurevich-Barsov, Antwerp
1998, and now 17...i.a4! 18 'tixa4 ~xd6 19
i.xd6 'tixd6 20 :fdl 'tieS is unclear
according to Gurevich.
10 .txc4 l2Jh5 11 'iixd8 :xd8 12 .5tg5
.te7 13 .txe7 l2Jxe7
10... .te7 11 g4!

Another one of those wing thrusts!
Black's best here is to enter a long, forcing
line. The slower 11 h4 and ~b 1 are
considered in Games 109 and 110
respectively, while 11 ttJd2 'ilVb6! 12 tDb3 (or
12 ~d3 d4 13 4Ja4 'tiaS with unclear
chances) 12... tDaS 13 tDxaS'tixaS 14 e4 dxe4
14 g4 l2Jf6 15 g5 l2Jfd5 16 .txd5 l2Jxd5 15 tDxe4 tDxe4 16 'tixe4 i.xa3 17 bxa3
17 :d 1 .td7 18 l2Jxd5 exd5 19 l2Jd4 'tic3+ 18 'tic2 'tial+ 19 'iVb1 'tic3+ 20 'tic2
~ac8 20 ~d2 ~f8 21 ~c1 ~e7 22 ~hg1 was a draw in Gabriel-Lutz, Bad Homburg
~xc1 23 ~xc1 ~d6 24 ~g1 g6 25 h4 1997.
%-% 11 ... dxc4 1 2 .txc4 e5! 13 g5 exf4 14
White has the better minor piece, but gxf6 .txf6 1 5 l2Jd5 l2Je 7 1 6 liJxf6+ gxf6
Black has all his weaknesses covered. White 17 :hg1+ ~h8
does have a small edge though.
We now tum our attention to the
aggressive main line.
Game 108
Kramnik -Karpov
A mber (blindfold) 1998
1 l2Jf3 l2Jf6 2 c4 e6 3 l2Jc3 d5 4 d4 .te 7
5 .tf4 0-0 6 e3 c5 7 dxc5 .5txc5 8 a3
l2Jc6 9 'iic2 .a5 10 O-O-O!
TIlls has almost completely superseded
the old move 10 :dl. One recent example is
166
5 i..f4 Variation
18 e4 Instead 25 ...l:.b8! 26 %:txf4 (26 l:.xb4 l:.xb4

The latest attempt, trying to cut off the 27 'tixb4 'ii'xd5+ 28 ~a1 [28 'ifb3 .i.b 1+ ]
black queen from the defence of the kingside 28 ...'ifdl+ [28...'iixf3 29 'iff8 is mate] 29
with .i.c4-d5. Two other moves have also ~a2 .i.bl+ 30 ~al .i.c2+ 31 ~a2 'Wbl is
been tried: mate; 26 'ifxf4 b3+ 27 ~a1 'iixa3+!! 28 bxa3
a) 18 'ife4 ttJg6 19 'ifd4 'ifb6! 20 'ii'xb6 b2+ 29 ~a2 bl"if is mate) 26...b3+ 27 ~al
axb6 21 %:td6 .i.h3 22 .i.d5 fxe3 23 fxe3 %:tac8 'iixd2 28 ttJxd2 ..tc2 29 %H3 %:tb5 30 %:te3
24 ~b1 %:tcd8 25 :Xd8 :Xd8 26 .i.xf7 ..tf5 %:txdS 31 ttJxb3 %:td1+ 32 ~a2 .tbl+ 33 ~a1
27 ~a2 .i.e4 28 ttJd4 ttJe5 was complicated .i.c2+ 34 ~a2 .i.b1 with a draw is probably
but balanced m Beliavsky-Yusupov, better. The game now swings back and forth
Dortmund 1998. and eventually ends in perpetual check.
b) 18 ttJd4 fxe3 19 fxe3 .txf5 20 ttJxf5 26 'i'xb4 ~xb4 27 axb4 :xd5 28 :xd5
'ifxf5 21 .i.d3 e5 22 ~b 1 f5 gave White i..e6 29 'it>a3 i..xd5 30 ltJd4 'it>g7 31 b5
some compensation for the pawn m ~g6 32 ltJc6 'it>g5 33 ltJxa7 i..aB 34 b6
Akopian-Pigusov, Tilburg 1994. ~g4 35 ltJb5 'it>f3 36 ltJd6 'it>xf2 37 b7
18 ... b5 19 i..d5 ltJxd5 20 exd5 i..d7! i..xb7 38 ltJxb 7 'it>e3 39 ltJd6 f3 40 ltJf5+
An improvement over the very murky 'it>d3 41 ltJg3 f2 42 b4 f5 43 h4!! f4 44
20...b4 21 axb4 'ifa1+ 22 ~d2 'ifa6 23 ttJd4 ltJf1 ~e2 45 ltJh2! f1'if 46 ltJxf1 'it>xf1 47
(23 'ifc6 %:td8 24 ~c3 ..tb7 25 'ifxa6 .i.xa6 b5 f3 48 b6 f2 49 b7 'it>g2 50 b8'i' f1~
26 %:td4 was equal in Van Wely-Short, Wijk 51 'ifg8+ 'it>h3 52 ~xh7 f5 53 h5 ~d3+
aan Zee 1997) 23 ... %:td8 24 b5 'iib6 25 'iie4 54 'it>b2 'i'd2+ 55 'it>b3 'i'd3+ 56 'it>b2
.i.b7 26 'ii'xf4 :XdS 27 ~c1, as in Akopian- 'i'd2+ %-%
Short, Groningen 1996.
21 'it>b 1 b4! 22 :d4 :g8 Game 109
This is very safe for Black, but in the post- Gelfand-Karpov
mortem Karpov suggested the amazing Wijk aan Zee 1998
22... bxa3 23 lhf4 f5!? (23 ... axb2 24 'iixh7+!!
~7 25 l:lli4 is mate - watch out for this 1 ltJf3 ltJf6 2 c4 e6 3 ltJc3 d5 4 d4 i..e 7
one!). Analysing the position with John 5 i..f4 0-0 6 e3 c5 7 dxc5 i..xc5 8 a3
Nunn during the tournament, we came to ltJc6 9 'i'c2 'i'a5 100-0-0 i..e7 11 h4!?
the conclusion that Black is bener! White has
two possibilities:
a) 24 ttJe5 %:tab8 25 'iid3 l:Ixb2+ 26 'it;al
which looks tricky for Black, but 26... l:Id2!!
27 'iie3 (27 'iig3 %:ta2+ 28 ~a2 'iid2+ 29
~a3 'iia5+ 30 ~b2 l:Ib8+ gives Black a
raging attack) 27...l:Ixd5!! and the rook has
tidied up the whole mess! Now 28 %:th4,
intending l:lli4xh7+, is met by 28...'ifb6! and
Black has a lot of pawns!
b) 24 ttJg5 and now only 24.. .f6!!
(24 ... %:tab8 25 'iid3 %:txb2+ 26 ~al is awful
for Black) 25 ttJxh7 axb2!! 26 'iixb2 %:tfb8!
wins for Black. That doesn't seem fair! The latest idea. White supports a later
23 :xg8+ :xg8 24 'i'd2 i..f5+ 25 ~a2 4:Jf3-g5 without taking on the weaknesses of
:d8? 11 g4, which allow Black an immediate
167
resolution of the position. 5 ..if4 0-0 6 e3 c5 7 dxc5 iLxc5 8 a3

11 ... a6 12 liJg5 liJc6 9 tjic2 tjia5 10 0-0-0 iLe7 11 ~b1
A couple of other moves have also been a6
tried here: Again this is the most natural response for
a) Van Wely-Sharif, Linares Zonal 1995, Black. The disastrous 11...l:dS 12 ttJd2
was not a success for White after 12 cxdS? 'iib6?? 13 c5! 'i'xc5 14 ttJb3! trapping the
exdS 13 ttJg5 l:dS 14 ~b 1 h6 15 ttJf3 .i.g4 queen was an unsuccessful try of Karpov's
16 .i.e2 l:acS with a slight edge to Black. earlier in the same tournament.
b) 12 ~bl as in Chemin-Chernuschevich, 12 tLJd2 tjib6!
Osterskar 1995, also gave White nothing Black takes the opportunity to relocate the
after 12...dxc4 13 ttJg5 'i'f5 14 'i'xf5 exf5 15 queen whilst simultaneously threatening
.i.xc4 h6 16 ttJf3 .i.e6 17 .i.a2 l:fdS. ... .li.e7xa3 .
12 .. J~d8 13 cxd5 13 tLJb3 lLla5! 14 tLJxa5 'i'xa5 15 cxd5
13 iLd3 h6 14 g4!?, intending .li.d3-h7+ exd5 16 i.e5 i.e6 17 i.d3
followed by ttJg5xf7 and 'i'c2-g6+, is an
interesting idea of Crouch's. His analysis
continues 14... e5 15 .li.h7+ ~fS 16 ttJxf7
~7 17 'i'g6+ ~fS IS .li.xh6 gxh6 19
'i'xh6+ ~f7 20 ttJxdS ttJxh7 21 'i'xh7+ ~e8
22 iigs+ ~d7 23 'iif7! with a wirming attack
for White. I feel that 14... d4 is the way for
Black to play here; for example, 15 .li.h7+
~fs 16 ttJce4 hxg5 17 hxg5 ttJxe4 18 .li.xe4
.li.d6 when White's compensation IS not
completely dear.
13 ... exd5 14 e4 tLJxe4! 15 tLJgxe4
15 ttJcxe4 (15 lhdS lhdS 16 'iixe4 :xg5!)
15...dxe4 16 lhd8+ ttJxdS 17 'i'e4 'i'f5! is 17 ...tLJe4!
some more interesting Crouch analysis. A very important resource for Black.
15 ... dxe4 16 1::xd8+ 'i'xd8 17 tjixe4 g6 17... l:acS IS 'iid2 d4!? 19 .li.xd4 :fdS 20
18 ..ic4 ..it5 19 tjie3 tjid4! :c1 g6 21 IDldl (21 h4!? and 21 f3 are
Black is quite comfortable here. suggested by Van Wely) 21...ttJeS 22 f3 ttJd6
20 'i'xd4 tLJxd4 21 :d1 tLJe6 22 i.xe6 23 e4 (Van Wely recommends 23 'i'f2 as
i.xe6 23 g3 1::c8 24 ~b 1 f6 25 tLJd5 stronger) 23 ... ttJc4 24 iLxc4 :Xc4 25 'iif2
i.d8 26 i.e3 i.g4 27 1::d2 i.f5+ 28 ~a2 l:dxd4 26 lhd4 .li.c5 27 ttJe2 and now
:tc2 29 1::d1 ~f7 30 tLJc3 i.a5 31 ~b3 27... iLxd42S ttJxd4 ltxc1+ 29 ~xc1 ~c7+
i.xc3 32 bxc3 1::e2 33 1::d6 g5 34 hxg5 would have equalised according to Van Wely.
..ie6+ 35 c4 fxg5 36 1::b6 ..ic8 37 a4 All the same, it seems that Black is slightly
~g7 38 c5 h5 39 c6 bxc6 40 1::xc6 i.f5 struggling to prove full compensation for the
41 i.xg5 1::xf2 42 ~b4 % -% pawn.
18 f3 1::fc8 19 i.xg7 ~xg7 20 fxe4 dxe4
Game 110 21 i.xe4 i.f6 22 1::d4 i.xd4 23 exd4 f5
Van Wely-Karpov 24 i.f3 tjib6 25 tjid2 i.f7 26 d5 %!d8 27
Amber (blindfold) 1998 %!d1 tjih6 28 'i'd4+ 'i'f6 291*'f4 :ac8 30
l:!d4 b5 31 h4 h6 32 h5 1*'g5 33 1*'e5+
1 d4 lLlf6 2 c4 e6 3 liJf3 d5 4 liJc3 ..ie7 1*'f6 34 _f4 1*'g5 35 1*'e5+ "f6 Y2 -Y2
168
5 ~f4 Variation
Summary
At present Black seems to be coping well in the main lines, but Crouch's 5... dxc4 is well worth
attention also.
1 d4liJf6 2lt)f3 d5 3 c4 e6 4lt)c3 ~e7 5 i..f4 0-0
6 e3 c5 7 dxc5 i..xc5 8 a3lt)c6 (D) 9 b4

9 ~e2 - Game 103
9~cl
9... a6 - Game 106
9...dxc4 - Game 107
9 'ilc2 'ii'aS 100-0-0 i.e7 (D)
11 g4 - Game 108
11 h4 - Game 109
11 <ltbl- Game 110
9 ... i..e7 10 cxd5 (D) It)xd5
10...exdS - Game 105
11 It)xd5 - Game 104
B... CDc6 TO ... i.e7 TO cxd5
189
CHAPTER ELEVEN
Queen's Gambit Declined:

'General Knowledge'
This last chapter brings together an 12 ... tiJhS?? 13 ttJxdS!

assortment of tactical tricks, thoughts and White wins a pawn as 13 ...cxdS 14 i.c7
lines - the kind of general knowledge you traps the queen.
usually accumulate through practical exper- But you can do ... this!
lence.
Game 112
The Oldest Trap of Them All! I.Zaitsev-Sveshnikov
Never do this as Black .. ! Moscow 1989
Game 111 1 d4 dS 2 e4 e6 3 tiJf3 tiJf6 4 ttJe3 e6 S
Euwe-Rubinstein exdS exdS 6 il.gS il.e7 7 e3 il.fS 8 il.d3
Bad Kissingen 1928 il.xd3 9 'i'xd3 liJbd7 10 0-0 0-0 11 :fe1
:e8 12 il.f4 ttJhS
1 ttJf3 dS 2 e4 e6 3 d4 tiJf6 4 il.gS tiJbd7
S e3 il.e 7 6 ttJc3 0-0 7 :c 1 c6 8 il.d3 a6
9 cxdS exdS 10 0-0 ~e8 11 'iVb3 h6 12
i.f4
13 liJxdS!?
Forcing a draw.
13 ... cxd5 14 il.c 7 'ile8 1 5 .lta5
Black's queen cannot escape, but White
170
Queen's Gambit Declined: 'General Kno wledge'
cannot profit from it. this lightning tour: 4...i..b4, 4....~Jbd7, 4...c5
15 .....b8 16 Ji..c7 "c8 ~-~ and4...dxc4.
What else do I need to know as Black? a) 4 ....tb4

If you wish to play the QGD, all you will
need apart from the lines given in this book
is a line against the Catalan - 1 d4 d5 2 c4 e6
3 g3
Black fights for the e4-square in more

active fashion. This form of development
owes something to the Nimzo-Indian
Defence - in fact, 5 e3 transposes to a
or 1 d4 dS 2 c4 e6 3 liJf3 ttJf6 4 g3 Rubinstein Nimzo-Indian.
White's latest hot weapon against this line
is 5 'iWa4+ ttJc6 (by forcing the knight to c6,
White makes it much harder for his
opponent to achieve his ...c7-c5 break) 6 a3
i..xc3+ (6...i..e7 7 cxd5 exdS 8 J.f4 0-0 9 e3
..tf5 10 ..ta6! bxa6 11 'iixc61:.b8 12 b4 was
clearly better for White in Kramnik-Bareev,
European dub Cup 1997) 7 bxc3 ttJe4
(7....1i.d7 8 cxdS exdS 9 .1i.g5 h6 [9 ...ttJe5 10
'ib4 liJxf3+ 11 gxf3 b6 12 1:.g1 was
unpleasant for Black in Malakhatko-
Moiseyenko, Ukrainian Championship 1998]
10 ..th4 g5 11 ..tg3 is slightly better for
which I'm afraid lies too far outside the White according to Malakhatko) 8 'i'c2 0-0
scope of this book - to complete your (8.J2JaS 9 e3 b6 10 cxdS exdS 11 ..td3 ..tfS
repertoire against 1 d4. In general I would 120-00-0 13 ttJel ..tg6 14 f3 ttJd6 15 i..xg6
recommend the sound main lines after hxg6 16 e4 was nice for White in Maric-
4....1i.e75 .1i.g2 0-0 60-0 dxc4 7 'iWc2 a6. Matveeva, Belgrade 1998) 9 e3 b6 10 cxd5
exdS 11 .ltd3 .1i.f5 12 c41:.e8 13 cxdS'iWxdS
What else do I need to know as White? 14 0-0 ttJd6 15 .1i.xf5 ttJxf5, as in Dautov-
After 1 d4 dS 2 c4 e6 3 ttJc3 ttJf6 4 ttJf3, Dizdar, Dresden Zonal 1998, and now 16
apart from the 4...i..e7 systems to which this i..b2, intending 1:.f1-e 1, lLlf3-d2 and e3-e4,
book has been devoted, Black has a variety was slightly better for White according to
of 'secondary systems' to which we devote Dautov.
171
b) 4 ...ttJbd7 was awful for Black in Anand-Korchnoi,

Tilburg 1998) 16 i..xb3 i..xb3 17 l:tac1 i..d6
18 i..b6 ':a8 19 lDd4 i..a4 20 l:tc4 i..d7 21
lldl gave White huge pressure.
d) 4 ...dxc4
This move flexibly retains the option of

...i.f8-b4 or ...i.f8-e7. White's simplest reply
is 5 cxdS exdS 6 i.f4! so that after 6...c6 7 e3
i.e7 8 h3 0-0 9 i.d3 ':e8 10 0-0 tDf8 11
'ii'c2 a position from the Exchange variation
is reached where White has put his bishop The sharp Vienna system. White should
directly on f4, as in Sadler-Lputian, Elista continue 5 e4 (5 e3 a6! 6 i.xc4 b5 7 i.d3
Olympiad 1998, rather than lose a move with i..b7 leads to an unchallenging QGA for
i.c1-g5 ftrst and only later i.g5-f4. Black as White's knight has been placed on
c3 too early) 5...i.b4 6 i.g5 c5 7 i.xc4 cxd4
c) 4 ... c5 8 lDxd4, when 8...•aS (Black players recently
seem to have gone off the sharp 8...i.xc3+ 9
bxc3 .as, since 10 i..b5+ lDbd7 [10...i..d7
11 i..xf6 gxf6 12 'i'b3 a6 13 i.e2lDc6 140-0
.c7 15 \!Va3 ':c8 16 ':adllDaS 17 \!Vel ~e7
18 .h6 i.c6 19lDxe6!!, intending 19...~e6
20 e5!, was crushing for White in Piket-
Topalov, World Championship 1997] 11
i.xf6 \!Vxc3+ 12 ~f1 gxf6 13 h4.aS 14:tel
~e7 16 \!Vc2 a6 17 i..e2 lDe5 18 \!Vb2 ':d6
19 ':b3 lDd7 20 f4 lDc5 21 ':e3 ':b6 22 • c2
:tb4 23 ':c3 lDa4 24 llxc8, as in Lputian-
Gabriel, Armenia-Gennany 1996, is an
example of the dangers) 9 i.d2 .c5 10
This equalising attempt is currently under i.b5+ i.d7 11 lDb3 \!Ve7 12 i.d3 lDc6 13
a cloud as a result of Kramnik-Van Wely, 0-0 0-0 14 a3 i.d6 15 f4 e5 16 f5 lDd4 17
Amber (blindfold) 1998, when 5 cxd5 lDxdS i.g5 i.c6 18lDd2 b5 19lDdS i..xd5 20 exdS
(5 ... cxd4 6 \!Vxd4 lDxdS 7 e4 comes to the followed by lDd2-e4 was rather painful for
same thing) 6 e4 cxd4 7 'i'xd4lDxc3 8.xc3 Black in Nikolic-Lautier, Monaco (blindfold)
lDc6 9 a3! i.d7 10 i..e2 ':c8 11 0-0 lDaS 12 1998.
'ii'd3! i..a4 13 'ii'xd8+ .:xd8 14 i..e3 ltJb3 15
i..dl! a6 (IS ... bS 16 i.xb3 i..xb3 17 ':fc1 Qtestion 1. And what if I play 1 d4 d5 2 c4
172
Queen's Gambit Declined: 'General f(no wleage
e63 tbc3 tbf6 4 i..g5 rather than 4 tbf3?

Armar 1. Then the only independent line
really is 4 ...dxc4!? It's not stupid! After 5 e4!?
(5 e3 is the most sensible) 5... iLe7 (S ... cS!? 6
dS iLe7 7 i..f4 tbxe4!? 8 tbxe4 exdS 9 tLJg3
0-0 10 tbf3 iLf6 11 'YWd2 was Neverov-
Crouch, Hastings 1991/92, when 11..:iib6
would have given White problems with b2
according to Crouch) 6 tbf3 c5 7 iLxc4 cxd4
8 1fxd4 h6 9 iLf4 1faS 10 0-0 tbc6 11 1fd3
0-0 12 e5 tbh5 131fe4 tbxf4 14 1fxf4 1fb4!?
the game was vel)' complicated in
M.Gurevich-Kupreichik, Groningen 1997.
In the QGA, it is ... a7-a6!
The Main Move ... tLJf6-e4!! After 1 d4 dS 2 c4 dxc4 3 tbf3 tiJf6 4 e3
Finally, I hope to have instilled in you a sense e6 5 iLxc4 c5 6 0-0
of the QGD move - the move that gets
things going. EveI)' opening has such a
move:
In The Semi-Slav, it is the swprising
...dSxc4!
1 d4 dS 2 c4 c6 3 tbf3 tbf6 4 tbc3 e6 5 e3
tiJbdl 6 i..d3 dxc4!
Black can only develop his pieces actively

by 6...a6! 7 'i'e2 b5 8 iLb3 J.b7.
It feels odd to play this move after

'strong-pointing' the dS-pawn with the e6-
and c6-pawns, but it removes all the
obstacles to Black's queenside expansion,
which proceeds apace after 7 J.xc4 bS 8
iLd3 a6 9 e4 c5.
... and in the QGD, as we have seen so

often, it is that ...tbf6-e4 move!
173
INDEX OF COMPLETE GAMES
Akopian-Short, European Team Ch, Pula 1997 .......................... 59

Akopian-Short, Linares 1995 .................................................. 86
Alekhine-Capablanca, Buenos Aires (2nd matchgame) 1927 ................... 37
Altennan-Pigusov, Beijing 1997 .............................................. 79
Anastasian-Lputian, Yerevan 1996 .......................................... 12 7
Andersson-Kasparov, Belgrade (match) 1985 ............................. 129
Arencibia-Beliavsky, Elista Olympiad 1998 ............................... 97
Arkell-Short, British Ch., Torquay 1998 ................................... 83
Azmaiparashvili-Short, Manila Olympiad 1992 ........................... 72
Bacrot-Korchnoi, Albert (match) 1997.................................... . 150
Beim-Komeev, Frankfurt 1997.............................................. 111
Beliavsky-Ivanchuk, Linares 1993 .......................................... 128
Bobotsov-Petrosian.T, Lugano Olympiad 1968 .......................... 130
Cifuentes Parada-Komeev, Malaga 1998 ................................... 19
Cramling.P-Amura, Merlo (match) 1994 ................................... 149
Cramling.P-Campora, Spanish Team Ch. 1993 ........................... 54
Dautov-Georgiev.Kir, Elista Olympiad 1998 .............................. 84
Dreev-Short, Linares 1995 .................................................... 165
Dydyshko-Kveinys, Moscow Olympiad 1994 .............................. 119
Ehlvest-Lputian, Yerevan 1996 .............................................. 159
Eingom-Balashov,Riga 1985 ................................................. 43
Epishin-Ziatdinov, World Open, Philadelphia 1997..................... 41
Euwe-Rubinstein, Bad Kissingen 1928 .................................... 170
Fischer-Spassky, World Championship 1972 .............................. 65
Gabriel-Bonsch, Bad Homburg 1996 ....................................... 152
Gabriel-Lputian, Gemlany-A rmenia match 1996 ........................ 154
Garcia Dundain-Ubilava, Ampuriabrava 1997 ............................ 57
Gelfand-Ivanchuk, Linares 1993 ............................................ 128
174
Index of Complete Games
Gelfand-Karpov,Polanica Zdroj 1998 ...................................... 165

Gelfand-Karpov, Wijk aan Zee 1998 ....................................... 167
Gelfand-Kramnik, Dortmund 1997 .............................. ············ 157
Gelfand-Piket, Wijk aan Zee 1998 ........................................ . 140
Gligoric-Vaganian, Zonal 1998 .............................................. 157
Gulko-Short, match 1994 ...................................................... 134
Hulak-Lutz, Wijk aan Zee 1995 ............................................. 81
llinsky-Nenashev, Bishkek Zonal 1993 .................................... 60
Izeta-Asrian, Ubeda 1998 ...................................................... 81
Karpov-Campora, San Nicolas (match) 1994 ............................. 122
Karpov-Campora, Villarrobledo (rapidplay) 1997 ....................... 29
Karpov-Georgiev.Kir, Tilburg 1994 .......................................... 83
Karpov-Kharitonov, USSR Championship 1988 ......................... 129
Karpov-Short,Amsterdam 1991 .............................................. 83
Karpov-Yusupov, Dortmund 1997............................................ 11
Karpov-Yusupov, London (6th matchgame) 1989 ......................... 23
Karpov-Yusupov, London (8th matchgame) 1989 ......................... 17
Kasparov-Ivanchuk, Wijk aan Zee 1999 .................................. 137
Kasparov-Portisch, Brussels 1986 ............................................ 58
Khalifman-Chandler, German Bundesliga 1995 ........................... 75
Korchnoi-Hiibner, Biel 1986 .................................................. 32
Korchnoi-Short, Wijk aan Zee 1990 ........................................ 68
Kramnik-Karpov, Amber (blindfold) 1998 ................................ 166
Kramnik-Lutz, Ge171"lany 1994 ................................................ 22
Kramnik-Short, Dortmund 1995 ............................................. 158
Kramnik-Yusupov, Dortmund 1998 ......................................... 101
Krasenkov-Beliavsky, Yerevan Olympiad 1996 ........................... 112
Krasenkov-Karpov, Polanica Zdroj 1998 .................................. 164
Krasenkov-Yusupov, Pula 1997 ............................................. 127
Krivoseja-Klovan, Ge171"lany 1998 ............................................ 126
Lputian-Cifuentes Parada, Ubeda 1996..................................... 60
Lutz-Yusupov, Tilburg 1993 ................................................. 141
Milov-Pigusov, New York Open 1998 ..................................... 136
MiraIles-Spassky, Angers 1990 ............................................... 92
Nenashev-Vaganian, USSR Championship 1991 ........................ 104
Nikolic.P-Hansen.L, Wij k aan Zee 1995 ................................. 122
Nikolic.P-Kramnik, Amber (blindfold) 1998 ............................. 130
Nikolic.P-Lputian, Yerevan Olympiad 1996 ............................... 24
Nikolic.P-Yusupov, Horgen 1994 ............................................ 21
Orsag-Bellini, Montecatini 1997.............................................. 56
Petrosian.T-Beliavsky, USSR Championship 1983 ...................... 132
Pinter-Portisch,Austria 1997................................................. 86
175
Pinter-Prandstetter, Taxeo Interzonal 1985 ................................ 34

Pirc-TyIor, Hastings 1932/33 ................................................ 37
Portisch-Dizdar, Sarajevo 1986 ............................................... 25
Portisch-Vaganian, St John, Candidates match 1988 ................... 105
Rivas Pastor-Toth, Rome 1984 ................................................ 36
Romanishin-Ehlvest, Biel SKA 1996 ....................................... 43
Ruban-Panchenko, Elista 1994 ............................................. 120
Rubinstein-Takacs, B udapes t 1926 ......................................... 131
Sadler-Short, British Ch. playoff, Torquay 1998 ............................. 26
Sadler-Van der Sterren, Linares Zonal 1995 .............................. 15 5
Sakaev-Beliavsky, European Club Cup 1999 ............................. 162
Salov-Piket, Madrid 1997 ...................................................... 50
Sherbakov-Koniushkov,Krasnodar 1997 .................................. 145
Sokolov.I-Dautov, Nussloch 1996 ........................................... 136
Sokolov.I-Oll, Pula 1997 ...................................................... 123
Sokolov.I-Van der Sterren, Dutch Ch., Rotterdam 1998................ 151
Spangenberg-San Segundo, Buenos Aires 1995 ........................... 47
Thorsteins-Zaitsev.I, Protvino 1988 ........................................ 143
Timman-Kasparov, Prague (match) 1998 ................................... 77
Topalov-Kasparov,Linares 1997 ............................................ 144
Topalov-Kasparov, Sofia (rapidplay match) 1998 ........................ 73
Topalov-Kasparov, Sofia (rapidplay match) 1998 ....................... 106
Topalov-Krarnnik, Linares 1998 ............................................. 78
Topalov-Piket, Amber (blindfold) 1998.................................... 142
Topalov-Yermolinsky, Yerevan Olympiad 1996 ........................... 30
Topalov-Yusupov, Elista Olympiad 1998 ................................. 164
Vaisser-Bricard, French Team Championship 1998 ..................... 141
Vaisser-San Segundo, Greece 1997 ..........'...... 142
L L• • • • • • • • • • • • • • • • • • • • • • • • • • • •
VanderSterren-Hansen.L, Wijk aan Zee 1995 .......................... 113

Van Wely-Karpov, Amber (blindfold) 1998 ............................... 168
Van Wely-Vaganian, Yerevan Olympiad 1996 ............................ 156
Vyzmanavin-Gavrilov, N ovgorod 1995 ..................................... 103
Vyzmanavin-Zarubin, Russian Team Ch. 1995 ........................... 91
Ward-Parker, 4 NCL 1997 .................................................... 139
Wmants-Kasparov, Brussels 1987 ............................................ 67
Yermolinsky-Beliavsky, Groningen 1993 .................................... 93
Yurtaev-Beliavsky, Yerevan Olympiad 1996 ............................... 90
Yusupov-Krarnnik, Vienna 1996 ............................................ 124
Yusupov-Lputian, Gennany-A nnenia match 1996 ...................... 145
Zaitsev.I-Sveshnikov, Mosccyw 1989 ......................................... 170
Zviaginsev-Kharitonov, Russia 1995 ........................................ 44
Zviaginsev-Kharitonov,Russian Team Ch, Kazan 1995 ................ 45
176

Sadler - Queens Gambit Declined (2000)

Uploaded by

Copyright:

Available Formats

Sadler - Queens Gambit Declined (2000)

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

Sadler - Queens Gambit Declined (2000)

Uploaded by

Copyright:

Available Formats

What is the structure and content of the document?

What is the structure and content of the document?

What copyright information is provided?

What copyright information is provided?

~.

Copyright © 2000 Matthew Sadler

All rights reserved. No part of this publication may be reproduced, stored in a

British Library Cataloguing-in-Publication Data

ISBN 1 85744 256 3

Distributed in North America by The Globe Pequot Press, 6 Business Park

All other sales enquiries should be directed to Everyman Chess, Gloucester

To Steve Giddins, my first coach!

EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Typeset and edited by First Rank Publishing, Brighton.

Part One: Main Line Queen's Gambit Declined

1 Lasker Variation (6 ... h6 7 ~h4 tDe4) 10

Part Two: Exchange Variation

7 Exchange Variation: Systems with ttJf3 109

Part Three: Other Variations

9 Systems with ~xf6 148

QGD 'General Knowledge' 170

1 d4 d5 ... i..c8-f5 or -g4 first and only then ...e7-e6

begun his central counterplay with ...c6-c5

Q«;stion 6. What is White's aim?

By supporting dS with the e6-pawn, Black Gambit Accepted (QGA) after 1 d4 dS 2 c4

Move Order 2 d4 ciS reaching the QGD! Alternatively,

frequently, but Short and Ivanchuk have

Qiestion 11. What is the difference?

reaching a vanatlon of the Exchange

1 d4 d5 2 c4 e6 3 lDc3 J..e7 4 lDf3 tiJf6 World Champions of all time, Emmanuel

Here we see the power of the pawn

development advantage, or to extract a

~d7 37 g4 ~c6? 38 .e5! 'ii'xe5 39 dxc4 12 i..xc4 ll)d7 13 0-0 e5

In contrast to 13 ...b6, 13...e5 strikes

3. 'White captures on e5. In the game, Black chose a more

2S .....txg4 26 lhg4+ ~h8 27 tiJf7+ ':xf7

is best. Since White does not have the

17 d5 cxdS 18 'ii'xdS'ii'xd5 19 i..xd5lDf8 20 5 i.g5 h6 6 i.h4 0-0 7 e3 lDe4 8 i.xe 7

Qetion 38. What is the point of this move

As far as I can see, this is not mentioned

lila1+ 38 ~h2 :a2 39 ~g1 :a1+ 40 ~h2

Kramnik considers the position equal

This excellent defensive move is the only 120-0

Black plays ... 0-0 without ... h7-h6

1 d4 d5 2 c4 e6 3 tiJc3 tiJf64 tiJf3 iL.e7 5 iL.g5 0-0

8 ... 'iVxe7 11 :'xc3 130-0

Orthodox Variation (6 ... liJbd7):

1 d4 d5 2 c4 e6 3 lDc3 tDf6 4 tDf3 i.e7 by liquidating the centre. This includes

than h7. as g6 is covered by the h7-pawn) 18 tiJc4

ltJ6d5 18 dxcS a6 19 ~e2 bxc5 20 a3 g5!! (I

Qtestion 9. It doesn't seem very logical to

QIestian 28. You are putting 8 aJ and 8

similar to the Rubinstein-Takacs game in the

1 d4 d5 2 c4 e6 3lbc3lbf6 4 tiJf3 i..e7 5 ..tg5 0-0 6 e3lbbd7 7 ':c1 c6

8 'ilc2 10.....xe7 11 ... ltJ5f6

Orthodox Variation (6 ... ttJbd7):

Qiestion 5. Any drawbacks? active play.

.d2 lDe4! winning) 14 .d2 .as "as

Black's most logical continuation, ridding

Qiestion 15. This just looks nice for Black!