Dvoretsky Mark Amp Amp Yusupov Artur - School of Future Champions-4 Secrets of Positional Play 2008-OCR Olms 242p PDF
Dvoretsky Mark Amp Amp Yusupov Artur - School of Future Champions-4 Secrets of Positional Play 2008-OCR Olms 242p PDF
Dvoretsky Mark Amp Amp Yusupov Artur - School of Future Champions-4 Secrets of Positional Play 2008-OCR Olms 242p PDF
Editorial board
GM Victor Korchnoi
GM Helmut Pfleger
GM Nigel Short
GM Rudolf Teschner
2008
EDITION OLMS
m
Mark Dvoretsky and Artur Yusupov
Secrets of Positional
Play
2008
EDITION OLMS
m
4
E-mail: [email protected]
Internet: www.edition-olms.com
All rights reserved. This book is sold subject to the condition that it shall not. by way of trade
or otherwise. be lent. re-sold, hired out or otherwise circulated in any form of binding or cover
other than that in which it is published and without a similar condition including this condition
being imposed on the subsequent purchaser.
Printed in Germany
ISBN 978-3-283-00518-4
5
Contents
PARTV
Mark Dvoretsky
P reface
modern players master typical positions, After a l l , Bareev's irony is not at all malicious,
characteristic of the openings they employ. and in add ition it is di rected not only at his
Another approach to the study of typical opponents or l isteners, but also at hi mself.
middlegame positions is employed in my Many years ago I saw a cartoon in which a
lecture, included in the same th ird part of the g rand mother was saying to a l ittle boy: 'And
book. now, g randso n , let's repeat some words
The fou rth part is devoted to the purely which you should never say. ' I remembered
practical implementation of various princi it, when I looked through the traditional
ples of positional play. In it an analysis is concluding material by Yusu pov, analysing
given of some strategically complicated fragments from games played by pupils of
games, played in top-level events. Here too it the schoo l . On this occasion the g rand mas
is interesting to compare the ways of th inking ter focused his attention on instructive
and the approaches to the taking of deci positional mistakes made by the young
sions of such outstanding g rand masters as players. In chess teach ing such a n approach
Artur Yusupov and Evgeny Bareev. is q u ite appropriate. Not without reason do
they say: ' Learn from your mistakes. '
Incidentally, I should mention that Yusupov,
who in 1 99 1 took up residence in Germany, I n conclusion I have given several opin ions
was unfortunately unable to take part in the expressed by legendary chess players,
final sessions of the school . His lectu res emphasising the exceptional importance, for
were written later - when the book was being any player, of the problems considered in the
prepared for publ ication . This factor allowed book:
Artu r to make use of games played two or A sensible plan makes heroes of us all; the
three years later, and in particular, two absence of a plan makes us faint-hearted
brilliant wi ns by Viswanathan Anand over fools. ( E m . Lasker)
Gata Kamsky in the Final Candidates Match , Contrary to general opinion, generated by
and some impressive games played by ignorance, Morphy's main strength was not
Yusu pov hi mself at a tou rnament in Switzer his combinative gift, but his positional play
land in 1 994 . and general style. After all, a combination
As for Bareev's material - this is indeed a can be carried out only when the position
lecture given at the schoo l . It made a strong permits it. (J . R . Capablanca)
impression on the pupils, not only through its The ability to evaluate a position is just as
purely chess virtues, but also its u nusual necessary as the ability to calculate varia
presentation - ironic, at times even caustic. tions. ( M . Botvinnik)
In my view, such a manner of del ivery, fully Endeavour to memorise as few variations as
reflecting the character of the g randmaster, possible! Positional feeling should become
was an embellishment to the lectu re. It could your release from the slavery of 'variations '.
be perceived as being offensive only by a And therefore: try to develop your positional
person totally lacking a sense of humour. feeling. (A. N imzowitsch )
8
PART I
Methods of Improving
in Positional Play
Mark Dvoretsky
'
H game' - this is how we usually cha
e has a subtle understanding of the defin itely be studied. And above all Aaron
N imzowitsch 's remarkable book My System.
racterise a strong positional player. It is very Recently I looked at some of the games
flattering to hear such a comment about our played in my youth , when I had fi rst category
own play, but, alas, not all of us can boast of rati ng, and I was staggered by the great
th is. And yet positional understanding is a number of crude, obvious positional mis
very important condition for our com petitive takes. But at the time they weren't obvious to
success. What should be studied by a player me. On the contrary, I was sure that I was
wishing to ach ieve serious prog ress in this playing q u ite strongly - the only th ings that
field , and what are the main directions and h i ndered me were 'accidental' oversig hts,
forms of such work? These are the questions and a lack of opening knowledge. For a time
that will he answered in this lecture. I was qu ite unable to make any prog ress .
You , of course, know that grandmaster And then , on the advice of my trainer
Yusu pov is renowned as a skilful strateg ist. I (Alexander Roshal) I stud ied My System. I
will illustrate these thoug hts with some didn't read it, but studied it, and I even copied
examples from his games. out the most importa nt ideas and exam ples.
This work qu ickly told on both the qual ity of
my play, and my resu lts - I won several fi rst
I . Some genera l recommendations category tournaments , became a candidate
A number of books have been written about master, and then ach ieved the master norm .
positional play. By no means all of them I also recommend the study of collections of
deserve attention, but some of them should games played by outstanding positional
The Improvement of Positional Mastery l[j g
players, preferably with their own com ments. ate links do not pass through the centre of
You will realise, of course, that d ifferent consciousness. '
players have their own way of playing - a The art of evaluation is the abil ity to pick out
universal positional style does not exist. the essence of a position - those and only
Some grandmasters ('strateg ists') are char those featu res of it, which should be taken
acterised by their logical manner of th inking i nto consideration when search ing for the
(for example: Rubinstei n , Botvinnik, Par strongest move. In training sessions it is
tisch ), while for others it is intu itive (Ca useful to express in words you r perception of
pablanca, Smyslov, Karpov). Also present in the essence of a position , i n order to note
their play are other disti nctions: they may and subsequently correct mistakes in posi
adhere to an attacki ng or defensive style, tional understanding,
they may aim for classical (with seizu re of After picking out the most important assess
space) or less orthodox set-ups, and so o n . ment considerations (most often - sub
Study the games o f a player w h o is closest to consciously) and checking the variations, we
you in style, or, on the contrary, one who is find the move (or a small number of moves)
especially skilled in what you are not good at. which corresponds to our perception of the
I should like to remind you of the method with situation . Usually this move pursues some
which you should record the episodes that defi nite aim, i . e . it is part of a specific
are the most interesting and usefu l for you - operation.
'positional sketches'. We recommended it in Of cou rse, often our decisions prove to be
the first session of our school (cf. the book difficult and complex, based on the calcula
Secrets of Chess Training). Draw a diagra m , tion of lengthy variations and on numerous
and write a commentary expressing the different evaluative considerations. But nev
essence of the position , the minimum neces ertheless, at the basis of any strateg ic
sary for the understanding of its moves and decision are simple positional operations,
variations. As a result, the ideas and evalua and it is very important to be able to fi nd them
tions related to this instance will be fi rmly q u ickly and confidently. You can assi milate
engraved in your memory. In exactly the them , by studying commentaries by grand
same way you can record instructive exam masters, and it is very useful also to try
ples from your own games. solving special exercises (for example, the
form of training games for the development
II . Positional operations of intuition , which was discussed in a
previous session of the school - it will be
When assessing a position , an experienced
described in the next book).
player never tries straight away to take
account of all its featu res, and does not We will pick out the basic types of positional
compile a complete list of all its pl uses and operations.
minuses (although this is what is called for in
some books). If such work is done, then it is Improving the placing of the pieces,
(see diagram)
24 .l:txe6 4Jf4?!
In the given instance the 'dubious' sign
reflects not the objective strength of the
12 � The Improvement of Positional Mastery
25 . . . l:tc5?
A vexing mistake. By playing 25 . . . l:te3! Black
would have retained his extra pawn. For
example: 26l"Llf1 l"Llf3+ ! 27 'itf2 l:.e5 28 l:tad 1
(28 l"Lle3 l:tgg5 ! ) 28 . . . l:txd5 29 l:txd5 l:lg5 ! , or
How should the position be assessed? For
26 .U.c1 'ite7 (26 .. .f6 ! ? ) 27 l"Llf1 .l:te2 28 l"Llg3
the knight Black has a sufficient equivalent
'iti>e6! 29 .U.cd 1 .l:.e 1 + ! .
three pawns. But if White should succeed i n
consolidating (l"Llf3 , 0-0 etc . ) , he w i l l have an 26 l:.xc5 bxc5
obvious positional advantage. Someth ing 27 l"Lle4 'ite7
must be urgently done. 28 l"Llxc5
19 . . . g5! And the game soon ended in a draw.
An excellent counter! If 20 g3 there follows
20 . . . gxf4 21 gxf4 l1g8. Both black rooks Spraggett - Yusu pov
become extremely active, and in addition the
Candidates Match, 9th Game,
opponent constantly has to reckon with . . . f7-
Quebec 1 989
f6.
20 0-0 gxf4
21 l:txf4 l"Llx e5
The wh ite centre has been completely
destroyed .
22 l:txd4 l:tg8
23 l"Llb4 l:!.c3
In the event of 24 .U.e 1 f6 Wh ite's position is
worse. In the endgame a knight which has
no strong points often proves to be
weaker than pawns. Therefore Jan Tim man
hurries to return the piece.
24 l"Llxd5!? exd5
25 i.ixd5 25 . . . b5!
The Improvement of Positional Mastery ctJ 13
N u n n - Yus u pov
Linares 1 988
Yusupov is plan ning a pawn offensive in the
centre. However, he sees the manoeuvre
planned by the opponent in reply to e2--e4 :
. . . l2Jf6-g4--e5 , and the knight gains control
of the d3- and c4-squares. This manoeuvre
must be prevented .
1 3 h3! tt:Je5
An important improvement for Black was
demonstrated in the game Gelfand-Kasparov
(Novgorod 1 997): 1 3 . . . b5! 1 4 e4 'ikc8 ! ! , and
the threat of 1 5 . . . b4 is highly un pleasant.
1 4 tt:Jxe5 .l:lxe5
1 5 e4 .l:te8
1 6 .i.e3 lt:Jd7 Black is a pawn u p , but the enemy pieces are
The knight again dreams of going to e5. threateningly trained on the kingside. Both
1 7 f4! c4 wh ite rooks ca n q u ickly end up there. Now
1 9 i.. h 7+ is threatened .
16 � The I mprovement of Positional Mastery
The first desire is to remove the q ueen from Engl ish g ra n d master: after the n atu ral
the danger zone by 1 8 . . . 'it'c7 . But then there 25 ... l:tfb8?! there follows 26 l:tb3! l:t2xb3
follows 1 9 'ifh5, and the white q ueen joins (forced ) 27 cxb3, and the wh ite q ueen
the attack on the king. 20 .l:tg3 or 20 .l:.g4 is obtains use of the c-file.
threatened , and 1 9 .. .f5?! is dangerous: 20 26 h3 l:tfb8
.l:tg3 iixe5 2 1 ltJf3 'i!i'c3 22 'ili'xh6. Black is Now 27 l:.b3 is pointless - Black replies
forced to retu rn the pawn: 1 9 . . . ltJf5 20 .ll xf5 27 . . . a5! . Less convincing is 27 . . . ltJgxe5 28
exf5 2 1 'ifxf5 ltJc4 , but in this case too Wh ite ltJxe5 ltJxe5 29 'it'e3.
retains the better chances.
27 'ith2! a5
18 . . . 'ith8!
28 'itg3
Of course, on c3 the queen feels uncomfort
But what does White want now? Obviously,
able, but from there, by attacking the knight
h3-h4-h5.
on d2, it restricts the mobil ity of the wh ite
queen and in general it rivets the opponent's 28 . . . .l:.c4!
attention, diverting him from the kingside. 29 h4? is no longer possible in view of
Remember: this defensive idea - restraining 29 . . . l:tbb4.
the opponent's activity with a far-advanced 29 c3
queen - was recommended in a lectu re by Another trap: the tempting 29 . . . d4 is refuted
Mikhail Shereshevsky, which he read at the by 30 ltJxd4 ltJxd4 3 1 'ii'f 1 ! . Even so, 29 ltJd2
first session of the school ( Secrets of Chess was more tenacious.
Training). 29 . . . a4!
19 g4? 1 30 ltJd2 ltJcxe5! !
1 9 ltJf3 was better. 3 1 1lla 1
19 . . . ltJac6! 3 1 ltJxc4 ltJxc4 3 2 l:te2 a 3 i s hopeless for
By attacki ng the e5-pawn , Black prevents Wh ite.
the opening of lines on the kingside by g4- 31 . . . l:!.c7
g5.
32 1t'xa4 .l:tbc8
20 ltJf3 !:tabS !
33 'ifa5 ltJc4
Again prophylaxis - this time against the
34 ltJxc4 .U.xc4
threat of 21 .ll c4 'ii'b 2 22 l:tb3. And if 21 g5
there follows 2 1 . . . l:tb4 ! . Black clearly has a g reat advantage and
subseq uently be successfully converted it.
21 ..tc4! 'ifb2
22 .ll b3 ltJg6!
Ill. Typical Positions
23 l:ta2
The black queen is nevertheless trapped , but In our games we constantly reach positions
the compensation for it will be more than with the same pawn structure , the same
sufficient. material balance and roughly the same
23 . . . l:txb3 arrangement of the pieces , as i n n umerous
games played earlier. It is useful to know
24 l:txb2 .:!.xb2
how strong players handled such positions,
25 'ii'c 1 ?! what plans they carried out and what ideas
25 .l:.b3 was better. they implemented .
25 . . . l:tb4! Many rules, relating to particular types of
Yusupov sees through the trap set by the positions, are well-known and are clearly
The Improvement of Positional Mastery lZJ 17
1 c4 c6 1 3 dxe6 l2Jxe6
2 e4
This game was played in the USSR Team
Championsh ip, and the first challenge, in
volvi ng the choice of open ing weapon , arose
before it began. When in his preparations
Yusupov looked to see how the opponent
repl ied to the English Opening, he noticed
that there was a possibil ity of transposing
into the Caro-Kan n Defence. Sergey Dolma
tov, who was playing for the same team as
Artur, had played successfu lly aga inst this
open ing, he had h is own prepared l i nes in
the Panov Attack, and he was prepared to
share them with his friend . The problem was
that Yusupov hardly ever played 1 e4 and he 1 4 .ie3?
had no experience in the resulting positions.
A normal developing move - this would
Yet he had made a deep study of chess as a
probably have been played without th inking
whole and had analysed games played with
by almost everyone. But now Black gains
all the openings. If you possess a broad
time, by attacking the bishop on b5, to force
chess erudition , you can permit you rself (and
favourable exchanges.
sometimes it is even usefu l ) to take a risk in
the open ing. Cast an unprej ud iced glance at the position
2 . . . d5 doesn't it remind you of anyth ing? Wouldn't
you agree that we have seen someth ing
3 exd5 cxd5
similar in the old games of Paul Morphy and
4 d4 l2Jf6 Adolf Anderssen? The centre is ope n , the
5 l2Jc3 l2Jc6 black king is stuck in the centre (true, at the
6 l2Jf3 .ig4 moment Wh ite's is also there), and there is a
7 cxd5 l2Jxd5 pin on the a4-e8 diagonal as in the famous
8 'ikb3 .ixf3 Morphy-AII ies game. But do you remember
what you should do i n such situations?
9 gxf3 l2Jb6
Sacrifice if necessary, i n h ibit the opponent's
Another possibility for Black is the endgame development and at the same time complete
arising after 9 . . . e6 1 0 'i!i'xb7 l2Jxd4 1 1 .ib5+ your own development as quickly as possi
l2Jxb5 1 2 'ikc6+! rl;;e 7 1 3 'ii'x b5 'i!i'd7 1 4 ble, bring you r rooks to the open files in the
l2Jxd5+ �xd5. centre and create a mating attack on the
1 0 d5 l2Jd4 enemy king .
11 .ib5+ l2Jd7 Alas, such opening strategy is someth ing
1 2 'iia 4 e5? that we have now half-forgotten , since in
In his preparations Yusupov had only reck modern set-ups the play is usually in a q u ite
oned with 1 2 . . . l2Jxf3+ and 1 2 . . . l2Jxb5 1 3 different key. I am in no doubt that without
The Im provement of Positional Mastery ctJ 21
thinking Morphy would have played 1 4 only chance consists i n the vulnerability of
� g5! ! , i n order t o place his rook on d 1 a s the enemy king , which comes under attack
quickly as possible. Black's position would by the wh ite rooks. But this is a temporary
immed iately have become hopeless: advantage, and if the opponent should
14 .. .<�J xg5 1 5 0-0-0; manage to consolidate, it will evaporate. The
flame of the i n itiative, which is about to go
14 . . . �e7 1 5 .i.xe7 �xe7 1 6 0-0-0 �d8 1 7
out at any moment, must be skilfully fanned ,
tt'ld5;
a n d this demands extremely accurate and
14 . .lL:lec5 1 5 �xd8 lL:lxa4 1 6 .i.xa4 .l::!.x d8 1 7
.
dynamic play.
0-0-0 �e 7 1 8 :I he 1 (or 1 8 llxd7);
I n the endgame one is supposed to keep
1 4 . . . '�xg5 1 5 .i.xd7+ rJ;;e 7. N ow 1 6 lL:le4 is
the king closer to the centre , and therefore
not bad , but the striving for rapid castl ing
1 7 0-0-0+ or 1 7 rJ;;e 2 suggests itself. But, as
may suggest an even more effective way: 1 6
you will already have real ised , here the
f4! lLlxf4 1 7 0-0-0! .
position should not be approached as if it is
14 . . . a6 a n endgame. The wh ite king reti res to the
1 5 .i.xd7+ kingside, in order to leave free for the rooks
If 1 5 .i.e2 there fol lows 1 5 . . . .i.c5 or 1 5 . . . b5 all th ree c-, d- and e-files, on which the
16 "it'e4 I1 c8 and 1 7 . . . �c5. opponent's king might be able to h ide.
15 . . . 'ifxd7 17 . . . .i.d6
1 6 "i*'xd7+ rJ;;x d7 The next move seems obvious: 1 8 llfd 1 . But
what will the opponent do i n reply? -
Yusupov asked h imself. After some thought
he found a strong defensive manoeuvre:
. . . .l:ta8-c8-c6 followed by . . . 'it>c8 . For exam
ple, 1 8 .l::tfd 1 ? ! I1ac8 ! 1 9 lLle4 I1c6 20 I1 ac1
.l:1d8! , and Black successfully com pletes his
development.
1 8 lLld5!
The g randmaster frustrates the suggested
arrangement of the forces. As you see, even
when fig hting for the i nitiative , prophylactic
operations are sometimes g iven preference
over attacking ones.
18 . . . .l::!. a d8
A new situation has arise n , one that is rather What to do now? In the event of 1 9 .i.b6 after
difficult to play. Yusupov worked it out 1 9 . . . .l:!.c8 Black will nevertheless play his
splendid ly. rook to c6 . And if 1 9 .l:i.fd 1 , then 1 9 . . . rJ;;c8 20
1 7 0-0 ! .i.b6 (20 �a7 b5!) 20 . . . �de8! (20 . . . l:!.d7? 2 1
I t i s wel l known that t h e favourable factors l:!.ac1 + 'it>b8 22 lLle3 ! , a n d lLlf5 o r l:Ic2-d2 is
operati ng in a position can be sub-divided threatened ) 21 .i.a7 b5! 22 lL:l b6+ 'it>b7 23
into constant (end uring) and temporary. It is l:txd6 ctixa7, and Black, at the least, stands
obvious that Wh ite has no constantly operat no worse. I n stead of 21 �a7 it is preferable
ing advantages - on the contrary, his pawn to play 2 1 lLlf6 ! ? gxf6 22 l:Ixd6 or 21 . . . .i.xh2+
structure is worse than the opponent's. His 22 'it'xh2 gxf6 23 l:Iac 1 + 'it>b8 24 nd7
'
22 � The I mprovement of Positional Mastery
retaining some pressure, but to Yusupov this The battle to retai n the i n itiative has conclud
evidently seemed insufficient. ed successfully for Wh ite. With a series of
19 l:tac1 1 precise prophylactic moves Yusupov has
Again prophylaxis: Wh ite prevents the re cramped the opponent's pieces and gained a
treat of the ki ng. Now it does not have the c8- defin ite spatial advantage.
square, while if 1 9 .. .<iot e8 the simple 20 .l::tfd 1 Now the character of the play changes -
is strong (20 �b6 is less convincing: there comes a phase of manoeuvring with
20 . . . �xh2+ 2 1 'lt>xh2 .l::tx d5 22 l:tc8+ l2Jd8). the aim of accu mulating add itional positional
19 . . . l2Jc7 pluses and g radually breaking up the enemy
defences. I should mention that for the
20 l2Jb6+ 'lt>e6
moment Wh ite's advantage is not yet decis
Perhaps at last it is time to place a rook on
ive and for success he requ i res a certai n
d 1 ? No, it is still too early - Black will then
'cooperation' on t h e p a rt o f t h e opponent.
consolidate his position with 21 . . . g5! fol
But, of course, he won't go wrong of his own
lowed by . . . f7-f6 and . . . �e5.
free will - i n this he must be hel ped .
21 llfe1 !
26 b4!
I n the event of 2 1 f4 !? Wh ite would have had
The plan is clear: a2-a4 and then at an
to reckon with 21 . . . 'lt>f5 22 .l::tfd 1 l2Je6 23
appropriate moment b4-b5-b6 . Of course,
lld5+ 'lt>g4, and the f4-pawn is attacked .
this is not fata l , but it is rather u n pleasant.
21 . . . 'iii> f5 I n cidentally, the attempt to win a pawn did
22 .l::te d1 not work: 26 ii.c5+? t2J xc5 27 t2Jxc5 l:.xd5 28
23 .U.xd6 l:txd6 24 .l:txc7 is threatened . .l::tx d5 .l::td 8 29 .Uxd8 ii.xd8 30 t2J xb7 ii.b6.
22 . . . 'lt>e6? is bad in view of 23 I:.xd6+! �xd6 26 . .. f6
24 �f4+, while after the knight moves the
27 a4 g5?!
wh ite rook invades with gain of tempo on d 5 .
By strengthening his control of the f4-point,
22 . . . l2Je6
Timoshchenko fights against the advance of
23 lld5+ 'it>f6 the b-pawn . If 28 b5 he had prepared
23 . . . 'lt>g6!? or 23 . . . ii.e5 !? 24 l2Jd7 f6 came 28 . . . axb5 29 axb5 l:lhe8 30 b6 �f4 . Howev
into consideration . er, the cure proves worse than the ail ment
24 l:tcd 1 �c7 in Black's position there is now a real
25 l2Jd7+ 'lt>e7 weakness: the f6-pawn . The restrai ned
27 . . . l:.he8 followed by . . . 'lt>f7 was better.
28 ii.c1 ! ?
The th reat o f 2 9 �b2 i s highly u npleasant,
and in some cases .ta3 is also possible.
However, 28 ii.d4!? also deserved serious
consideration , forcing an advantageous ex
change of minor pieces (28 . . . .:txd7? is not
possible on account of 29 llxd7+ 'lt> xd7 30
.txf6+ and 31 �xh8).
28 . . . l2Jf4?
Another mistake, provoked by Yusupov's last
move. 28 . . . .l::t h g8! was essential, preparing to
defend the f6-pawn with the rook from g6.
The Improvement of Positional Mastery CtJ 23
opponent's plans that are u npleasant for logical (to defend it by f2-f3 with the pawn
him. Jorge Rubinetti did not play 4 . . . e5 already on h3 is very u ndesirable). Yusu pov
immediately, in order not to al low a well would most probably have replied 1 2 'ii'c 2 ,
known (and rather dangerous) wh ite devel 1 3 l:tad 1 (with the threat of 1 4 l2Jxc6 bxc6 1 5
opment scheme: 5 �g5 �e7 6 'ifc2 followed ..ixc5), and then either f2-f4 followed by
by e2-e3 and �d3. Now if 5 ..ig5 he has the �f3 , or l2Jf5 . Again I would not ventu re to
good reply 5 . . . h6 6 ..ih4 g5 7 �g3 l2Jh5. judge which is more important: Wh ite's
5 e4 e5 spatial advantage or the opponent's control
6 ..ie2 ..ie7 of the dark squares .
7 0-0 0-0 1 1 d5
8 h3 Fully consistent (White has prepared this
The first small strategic idea (not cou nting move) and also in the style of Yusupov, who
the choice of arranging the pieces in the l i kes positions with a spatial advantage.
opening). Wh ite prepares �e3, in order then 11 . . . "ikc7
to play d4-d5 and after the reply . . . l2Jc5 to 1 2 l2Jd2 l2Jc5
defend the e4-pawn by l2Jd2 , without sh ut
ting in the bishop. It is with such operations
that any experienced competitor plans his
play. But, of course , this is not the only
possible approach to the position - accord
ing to theory, the immed iate 8 d5 is also
strong.
8 . . . a6
Black wants to play 9 . . . b5, creati ng the th reat
of 1 O . b4 . Should White forestall the oppo
. .
various possibil ities. But the idea fou nd now an obviously u nsou n d idea . The suggestion
by Yusupov was very u n u su a l , and wou l d of Vlad i m i r Kra m n i k was fa r stronger:
appear n o t t o have occu rred before in similar 14 . . . 'ii'd 7 ! , not only cleari ng the way for the
positions - it is this that makes the game bishop to g o to b6, but also preventi ng g2-g 3
special. for t h e m oment and intending to meet 1 5 g4
13 'it>h2! ..tdB with 1 5 . . . lt:Jh5! .
1 4 l:tg1 ! ! 1 5 llb1
Wh ite prepares g2-g3 and then f2-f4 . (After As the general plan is put into effect, the
the immediate 1 4 g3 there was the reply placing of the individual pieces is made more
14 . . .'ii' d 7, whereas now it will be possible to precise. 1 5 'il'c2 is weaker - from d 1 the
defend the h3-pawn with the bishop from f1 . ) q ueen threatens the d6-pawn , and in addi
I n the event o f the exchange on f4 , witho u t tion it may later be needed on the d 1 -h5
conced ing any central sq uares Wh ite i n diagonal .
creases his spatial advantage, and his r o o k 15 . . . 'iia 7
is excellently placed o n t h e newly-opened g In order to carry out his pla n , Black has had
file. And i f Black avoids t h e exchange o f to s h ut his q u een o ut of the game - a clear
pawns, there follows f4-f5 and g3-g4-g5 , sign that his idea is faulty.
and again t h e roo k stands where i t i s 1 6 g4!
needed .
Taking into account the poor placi ng of the
This is the main plan of action , but Yusupov opponent's pieces, Yusupov chooses the
also envisaged another deve l opment of more agg ressive of his two plan ned offen
events. Wh ite can also play g2-g4 , and then sive o ptions.
either o rgan ise a pawn storm on the king side 16 . . . l:tf8?
with g4-g5 and h3-h4-h5 (the rook will
Too passive (Black wants to support his d6-
support it from g1 ), or manoeuvre his knight
pawn with his knight from e8). 1 6 . . . ..tb6 1 7
via the vacated f1 -square to f5 . In the event
g 5 li:Jfd7 was more consistent. Apparently,
of the exchange on f5 the rook will end up on
Rubinetti was afraid of losing a pawn after 1 8
an open file.
li:Jf1 li:Jf8 1 9 dxc6 bxc6 20 'ii'x d6, but even by
We see that again o n e is not talking about a playing simply 20 . . . ..tb7!? followed by . . .
clear-cut plan, since depending o n his :ad8 , . . . lt:Jce6 and . . . lt:Jg6, Black would
opponent's reactio n Wh ite can choose this or have gai ned fi ne cou nter-chances, and with
that line of play. But at any event we are right 20 . . . Itd8 ! 21 'i¥xc6 (2 1 'ii'x e5 ..tc7) 21 . . . ..td7
to ad mire the versatil ity of Yusupov's idea - he would have trapped the enemy q u een .
the moves he has made will prove u sefu l Yusupov was intending to prevent the ap
whatever the development of events. pearance of the knight at g6 by 1 8 h4! li:Jf8 1 9
For h is part, Black did not manage to cou nter h 5 , and if 1 9 . . .lt:Ja6, then 20 l:.g3, preparing
his opponent's excellent play. However, the an attack on the kingside and intending, of
plan , beg u n with his last move , of playing his cou rse, to captu re on e3 with the f-pawn .
bisho p to b6 is q u ite logica l . Also after 1 7 li:Jf1 !
1 3 . . ..tf8 1 4 l:tg 1 !! or 1 3 . . . h6 1 4 l:.g 1 !! lt:Jh7 1 5
1 7 g5?! lt:Je8 was now inadvisable - Black
.
Mark Dvoretsky
P rophylactic Thi n ki n g
positions (after the exchange on e5). The attacked after . . . tt:'lb6, and it can possibly go
opponent replied 1 2 . . . ..ic7, th reate n i n g to b1 ). But the main th ing is that now in reply
1 3 . . . e4 . to 1 2 . . . �c7 he has 1 3 tt:'lb5 ! . For example,
Now Wh ite ca n try to conti nue his plan with 1 3 ... �b8 14 �b4 c5 1 5 ..ixc5 ! , winning a
1 3 ..id3 .l:te8 14 dxe5 tt:'lxe5 1 5 tt:'lxe5 'i!lxe5 pawn . At the same time Wh ite creates the
1 6 f4 , but then 1 6 . . . 'ii' h 5, and Black is alright. positional threat of gaining the advantage of
The prophylactic move 1 3 h3!? deserved the two bishops by 1 3 tLld5 and 1 4 ..ixa5. If
serious consideration, in order after 1 3 . . . e4 this is parried , say, by 1 2 . . . ..ib6, then he
1 4 tt:'lg5 to parry the th reat of 1 4 . . . ..ixh2+. strengthens his position with 1 3 l:.ae 1 , when
Black would have repl ied 14 . . . tt:'lb6! 1 5 ..ia2 what Black should do is not clear. Perhaps it
'i!Vd6 or 15 . . . �f5 with a complicated game. makes sense to nevertheless try 1 2 . . . ..ic7 ! ?
('if i t is impossible, b u t you very much want
Mikhail Botvi nnik chose 1 3 tt:'le4 tt:'lxe4 1 4
to , then you can ! ' ) 1 3 tt:'lb5 ..ib6! 1 4 ..ib4 c5.
'ii'xe4 , and after 1 4 . . . a5? ! 1 5 �a2 tt:'lf6 1 6
However, after 1 5 dxc5 ..ixc5 1 6 l:tfd 1 ! and
'ii'h 4 e4 1 7 tt:'le5! he seized the in itiative .
the u navoidable tt:'ld6 his position is unenvi
Sounder was 1 4 . . . �d6 1 5 ..ic3 exd4 1 6
able.
'ii'x d4 tt:'lf6 or 1 4 . . . tt:'lf6 1 5 'ii' h 4 e4 1 6 tt:'le5
..ie6 with roughly equal cha nces. The strong impression made by the g iven
example induced me to make a serious
Thus in neither of these two games did Wh ite
study of the problem of prophylaxis in
demonstrate a convincing way to ach ieve an
genera l . Soon I switched my attention from
advantage. The strongest continuation was
actual prophylactic moves to the process of
pointed out by Keres.
search ing for them - from the practical point
12 �a2 ! ! of view this seemed more important.
How can i t b e arrived at? Let us ask It became clea r that there was the need to
ourselves what Black wants, and how he is develop an approach to the position, which I
now intending to play. Obviously, not 1 2 . . . e4? call 'prophylactic thinking' - the habit of
1 3 tt:'lxe4 . 12 . . . �xc3? is also clearly bad : 1 3 constantly asking yourself what the oppo
..ixc3 e4 1 4 tt:'le5 with the th reats of 1 5 ..ib4 nent wants to do, what he would play if it
or 1 5 f3 . The open ing of li nes in the centre by were h i m to move, the abil ity to find an
1 2 . . . exd4?! 1 3 exd4 ! is also to Wh ite's answer to this question and to take account
advantage - he qu ickly develops dangerous of it i n the process of coming to a decisio n .
pressure by .l:tfe 1 and tt:'le5 (or tt:'lg5). After Developing t h e skill o f prophylactic th i n king
1 2 . . . .l:te8? the f7-point is weakened , while in enables a player to make an enormous step
the event of 12 . . . h6? there is the un pleasant forward , and to g reatly raise his standard of
reply 1 3 tt:'lh4, when the knight goes to f5 or play. Why? I will single out two main reasons.
g6.
1 ) The range of positions in which prophylac
The only sensible move is 12 . . . �c7 ! , which tic th inking can be used is extremely broad.
prepares 1 3 . . . e4 and thereby provokes the Any sign ificant positional decision is bound
opponent into releasing the tension i n the to combine the implementation of your own
centre. If it were possible to prevent this plans with actions against the opponent's
move, Black would encou nter serious prob (th is is how I understand the sense of
lems. N imzowitsch 's statement g iven above about
Now we can appreciate the true worth of the the role of prophylaxis). The main principle of
modest bishop retreat. It is usefu l as regards converting an advantage is the restriction of
Wh ite's subsequent plans (the bishop is not the opponent's possibilities; it is clear that
Prophylactic Thinking lLJ 29
here one can n ot manage without prophylac- This was a famous match , in which the
.
tic th inking . By developing this trait, you also Moscow team lost to Russia with a score of
become stronger in tactics and you make %-8Y:z! A certai n folklore even developed
fewer blunders. When defending a d ifficult aro u n d it. 'We will g ive u p M oscow, but save
position, you must all the time see with what Russia', the spectators q uoted M . I . Kutuzov*
you are th reatened ; and when attacking you after the end of the match . And g rand master
must reckon with the opponent's defensive Bukh uti Gurgen idze spread his hands in
resou rces. Thus a possession of the skill of astonishment: ' E ight and half, was this
prophylactic th inking exerts a favourable deliberate? After all, there is a fi l m of this
influence o n practically all aspects of your name by Fel l i n i . ' With Wh ite against Lev
play. Polugayevsky, the Moscow team captain
2) Chess is a battle between two players with Vasily S myslov made the opening moves 1
equal rights, and your opponent's ideas may d4 lLlf6 2 c4 c5 3 d 5 . After losing the game,
be no worse than yours. It is logically clear Vasily Vasil ievich lamented : ' I was wrong to
that the o ptimal strategy should harm o n i ous play 3 d 5 ; I got carried away and over-rated
ly combine the implementation of your own my positi o n . I should have played 3 lLlf3 ! ' .
ideas and the preventio n of your o pponent's. B u t a t a team meeti ng after t h e match
Of course, it is bad to remain passive and Smyslov tried to reassure everyone: ' Never
merely destroy, but the opposite tendency is m i n d , the most important thing is that
also extremely dangero u s , and liable to lead everyone is still alive . '
to constant failures.
I a l s o made my 'contribution' to the defeat of
Meanwh ile, players often forget to th i n k the M oscow tea m , but alas, it was my
about their opponent's p l a n s . T h i s is under opponent who demonstrated the strength of
standable: concentration on o ne's own feel prophylactic th inking .
ings is typical of human natu re . After a l l ,
sometimes i n life too , u nfortunately, w e d o
Tsesh kovsky - Dvoretsky
not take t o o much account of t h e tho u ghts
and feelings of others. Riga 1 975
Thus it is not a matter, of cou rse, of giving French Defence
priority to destructive actio n s over creative 1 e4 e6
ones, but simply that the important skill of 2 d4 d5
prophylactic thinking is most probably insuffi 3 lLld2 c5
ciently well developed i n us. By improving 4 lbgf3 lbc6
this aspect i n which we are backward , and
5 exd5 exd5
making our th inking more harmonious, we
will certainly raise sign ificantly our overall 6 ..tbS i.. d 6
standard of play. 7 dxc5 ..txc5
Of my own games on the theme of prophy 8 lbb3 ..td6
laxis, the o n e that left the strongest impres 9 0-0 lbge7
sion was my encounter with Vitaly Tseshko In 1 974 I spent several days at a training
vsky from the 1 975 USSR Spartakiad . session with Victor Korch n o i , who was
*Russian a rmy com m ander M i khai l Kutuzov, po pu l arl y cred ited with saving Russia a g a i nst N a poleon's
invasion i n 1 8 1 2 . (translator's n ote)
30 � Prophylactic Thinking
preparing for his final Candidates match queen on d8 Black would have captu red with
against Anatoly Karpov. I remember that we the bishop and the d5-pawn would have
were analysing a similar position from the been defended ) 1 6 tt:'lfd4 ii.xe2 1 7 Vxe2 a6
same open ing variation, and I asked Victor 1 8 'ii'f3 llad8 1 9 .l:!.ad 1 . By carrying out
Lvovich why he developed his knight at f6, advantageous excha nges, Karpov has em
and not at e7. The grandmaster looked at me phasised the weakness of the d5-pawn . He
in su rprise. now plans to i ntensify the pressu re on it, by
'Let's stop and th ink about th is. How should doubling rooks on the d-file and playing his
the pieces be arranged when you have an knight to e3.
isolated pawn? The place for the knight is at 12 h3 ii.h5
f6 , and later - at e4 . And the bishop is best 13 ii.xc6
kept on the g 1 -a7 diagonal - from there it Tseshkovsky 'takes the bull by the horns' -
puts pressure on the f2-poi nt. In the main he immediately tries to refute the opponent's
variation Black plays 8 . . . ii.d6 and 9 . . . tt:'le7 opening set-u p. Other, more restrai ned pos
simply on account of concrete factors (if sibilities , are 1 2 i.. e 2 and 1 2 i.. h 4 (without
9 . . . tt:'lf6 there is the un pleasant reply 1 0 the inclusion of 1 2 h3) .
.l:r.e 1 + , while in the event of 8 . . . ii.b6 Wh ite will
13 . . . bxc6
immediately offer the advantageous ex
change of bishops: 9 l:!.e 1 + and 1 0 ii.e3). But 1 4 tt:'lbd4 l:tcB
if you can develop the pieces on their lawful 1 5 c4 l:teB
sq uares without being pun ished , you should Black prepares . . . f7-f6 . Later, also against
do so . ' Tsesh kovsky (Sochi 1 975), Boris Gulko
Such evaluations, heard from the lips (or played 1 5 . . . h6, and after 1 6 .ih4 g5
read in the commenta ries) of top players, ( 1 6 . . . 'i!i'c7 ! ? ) 1 7 ii.g3 .ixg3 he obtai ned a
sometimes help you to sense the subtleties good position . Wh ite chose a more critical
of opening strategy far better than lengthy conti nuation in the game Peters-Ervin (Lone
articles and specialised books . Pine 1 978): 1 6 i.. x e7 !? i.. x e7 1 7 g4 ii.g6 1 8
1 0 .l:te1 0-0 tt:'le5 .
c3 ( 1 2 i.. h 4? tt:'lf5) 1 2 . . . .ig4 1 3 h 3 ii.h5 1 4 was concerned about the reply 2 1 b5.
ii.e2 h6 1 5 .ixe7! tt:'Jxe7 (the drawback of the
(see diagram)
queen 's position at c7 is felt - the knight has
to be placed in a passive position ; with the
Prophylactic Thinking t2J 31
27 'ii' b 3 a6
28 l:te1 g6
29 ..tg3 ctm
30 'iWe3
There is no defence against the i nvasion on
the e-file.
30oo .'ii'd 7 31 f4 ttJc4 32 We6+ 'ii'x e6 33 l:lxe6
lLlb2 34 1:lxc6 .Uxc6 35 ttJxc6 �c7 36 <it>f2
a5 37 bxa5 ttJd3+ 38 �e3 ttJxc5 39 <it>d4
ttJe4 40 <it>xd5 ttJxg3 41 a6 �b6 42 a4 �e8
43 a5 �f2 44 ttJd4 Black resigns.
illustrate the style of play and way of th i n king This is a l ready prophylactic th inking. Karpov
of the then world champion. It should be said i m med iately defines the opponent's main
that when I saw the game for the fi rst time it idea , on which he will keep a careful watch
did not make much of an impression , th roughout the entire game.
because Black lost without a struggle. And 10 i.. g 5
only later, when I read Karpov's comments, A standard idea - Wh ite provokes . . . h7-h6,
did I real ise how much su btle work was i n order to then develop his q ueen at d2 with
concealed beh ind outward simplicity. an attack on the h6-pawn.
10 . . . h6
Karpov - Ti mman
1 0 . . . exd4 1 1 lZ'lxd4 h6 is tempti ng, hoping for
Montreal 1 979 1 2 i.. e 3 lZ'lc5 , when it is awkward to defend
Pirc-Ufimtsev Defence the e4-pawn ( 1 3 i..f4 tZ'lh5). But Wh ite replies
1 e4 d6 1 2 .if4 ! g5 1 3 .ic1 , and the weakness of the
2 d4 tZ'lf6 f5-square will su bsequently tel l .
3 lZ'lc3 g6 'After 1 0 . . . 'i!Vb6 Wh ite could have gained a n
4 g3 .ig7 advantage b y the simple 1 1 .l:!.b1 , since
1 1 . . . exd4 1 2 lZ'lxd4 lZ'lc5 1 3 b4 lZ'lcxe4 is
5 i.. g 2 0-0
clearly unsatisfactory in view of 14 i.. xe4!
6 lZ'lge2 e5 lZ'lxe4 1 5 lZ'lxe4 i.. x d4 1 6 lZ'lf6 + . '
7 0-0 lZ'la6?! A not altogether convincing comment! Of
Theory recommends 7 . . . lZ'lc6 or 7 . . . c6 . After cou rse, the e4-pawn should not be captu red
7 . . . c6 the usual reply is 8 a4 (preventing - the normal reply is 1 3 . . . lZ'le6 . Besides,
. . . b7-b5, a useful move for Black), but then instead of 1 2 . . . lZ'lc5 Tim man could have tried
the development of the knight at a6 becomes either 1 2 . . . d5, or 1 2 . . . lZ'lg4 !? 1 3 hxg4 i.. x d4
more justified , since the b4-square is availa 1 4 i.. e 3 i.. x e3 1 5 l:txe3 d5! with unclear play.
ble to it. This variation has its i n ner logic: the bishop
8 l:te1 c6 at g5 is hardly participati ng in the battle for
9 h3 the centre , and Black exploits this factor.
'A typical prophylactic move i n such situa 1 1 i.. e 3
tions,' writes Karpov. 'Wh ite restricts the
opponent's bishop, and at the same time he
also creates a ' no-go' area on the kingside
for the remaining minor pieces . '
The remaining pieces' - this obviously
means the knight on f6 . The wh ite bishop
wants to got to e3, and it must be safeguard
ed against the attack by . . . lZ'lg4 . The
advance of the pawn to h3 is typical in such
positio n s , and it is played without even
thinking about the opponent's possibilities .
9 . . . .l::!. e 8
'All Black's hopes of obtaining counterplay
are associated with pressure on the e4-
pawn . ' Wh ite obviously wants t o complete his
Prophylactic Thinking
Timman is an active player, sometimes Karpov's next move is probably the best in
excessively active . It was hardly good to the game. How did he find it? Obviously he
weaken the q ueenside pawns. asked himself what the opponent wanted .
18 axb4 lt'lxb4 The answer is clear: to bring the knight into
1 9 lt'lce2 play via c5. How can this be prevented?
'Black's idea was to . . ' For heavens' sake,
.
22 'ii'c 2 ! 1
what does it matter about Black's idea! On a 'A subtle move, which, firstly, prevents the
first glance at the position it is clear that black knight from moving to c5 (in view of the
White has gained a solid positional advan reply b2-b4 ! ) , and secondly, continues the
tage, and we would probably have been previous strateg ic policy - the supporti ng of
thinking about how to exploit it as soon as the e4-pawn . '
possible - whether to play f4-f5, prepare g4- A s you see , here two forms o f prophylaxis
g5, and so o n . But Karpov th inks completely mentioned by N imzowitsch are combined -
differently - even in such situations he fi rst of prevention of the opponent's plans and over
all mon itors the opponent's possible inten protectio n .
tions and endeavours to forestal l them .
It is curious that Ka rpov does not even
So, once more: 'Black's idea was to some
examine 22 lt'lxc6!? 'i¥xc6 23 e5 - a
how bri ng his pieces together, by playing
continuation which players nowadays would
. . . a7-a5, . . . ..tc8-a6, . . . e5xd4 and . . . c6-c5.
certainly seriously consider. Why? - in this
But this is a lengthy process, and Wh ite
case the enemy pieces would have become
succeeds in h indering his opponent's pla n . '
active. However, after 23 . . . 'ilkc7 24 exf6 .i.xf6
N ow i f 1 9 . . . a5 there follows 20 c3 , d riving 25 lt'le4 followed by .i.d4, or 23 . . . d5 25 exf6 ,
away the knight and supporting the centre. Wh ite would still have retained t h e better
Black also has a bad position after 1 9 ... c5 20 chances.
fxe5 dxe5 2 1 d5.
22 . . . .i.d7
19 . . . exd4
Agai n Black wants to play . . . lt'lc5 (23 . . . lt'lc5
20 lt'lxd4 aS
24 b4 axb4 25 cxb4 lt'le6, and the c6-pawn is
'White's su bsequent play essentially reduc defended ), and again Wh ite prevents th is.
es to preventi ng the opponent's pieces from
23 lt'lf3 ! l:l.e7?!
coming into play. '
21 c3 lt'la6 If 23 . . . lt'lc5! ? Ka rpov had prepared 24 e5.
Timman bel ieves his opponent, but he was
apparently wrong to do so. In difficult
positions you should ca refully check any
possibil ity of sharpening the play. I do not
see how 24 . . . lt'ld5 is refuted . If 25 exd6
'ilkxd6 26 'ilkf2 (26 c4? lt'lxe3 ), then 26 . . . l:!.xe3!
27 l:l.xe3 'ii'xf4 28 l:tee 1 lt'le6, and Black has
defi n ite compensation for the sacrificed
exchange. Wh ite would probably have had to
restrict himself to the qu iet 25 .i.xc5 dxc5,
but here too it is not easy for him to
demonstrate his advantage: 26 lt'lh5 (Adian
to) 26 . . . ..th8, or 26 'i¥c 1 ( N u n n ) 26 . . . c4 27
lt'le4 f6 28 lt'ld6 l:!.f8 .
36 � Prophylactic Thinking
Karpov's recommendation 23 . . . cS was also make? Probably 24 ..tf2 - it would not be bad
preferable to the move in the game. to defend the e4-pawn once more , and also
the th reat of the e4-eS breakthrough (after
2S 'ili'd3 ) gains i n strength . '
A s you see , prophylactic th i n king is by n o
means synonymous with passivity; i t i n
volves simultaneously taking account o f both
the opponent's, and you r own resou rces,
with the accu rate calcu lation of short varia
tions. Opposing such a manner of play is not
easy - you must be not inferior to the
opponent i n abil ity to g uess the other side's
plans. I n this respect Timman was not fu lly
Karpov's equal .
24 . . . �e8?!
24 . . . .tea was more tenacious.
24 �f2 ! 25 'ii'd 3 1 'ii' b 7
'One of the last prophylactic moves. Prior to 2S . . . tt'lb8 26 eS! .
his decisive offensive, Wh ite places his 26 l:ta 1 !
pieces in the most harmonious way possible,
The modest manoeuvres of the wh ite pieces
and . . . once again reinforces his central
with i n their own territory have led to material
outpost at e4! 24 'ii'd 3 is premature in view of
gains - the aS-pawn ca nnot be defended .
24 . . . �c8 . '
26 . . . tt:J c 7
Here i t i s perhaps t h e last com ment by
Karpov that for me is the most informative. 27 .l:lxa5 .l:ldd7
Using it, I will try to reconstruct his approxi 27 . . . 'ii'x b2? 28 .l:lb1 .
mate train of thought: 28 b4 tt'le6
'The double attack 24 'ii'd 3 is tempting, but
after 24 . . . �c8 2S eS ttJdS 26 exd6 l:Ixd6 the
move 27 c4? does not work because of
27 . . . l1xe3! (but not 27 . . . tt:Jxe3? 28 'ifxd6
tt'lxd 1 29 'ii'x e7 ) 28 lhe3 tt:Jxe3 29 'ifxd6
tt'lxd 1 (poi nted out by N u n n ) . No, there's no
point in going in for such adventures!
'But what does Timman want to play; why
did he make his last move? It can hardly be
24 . . . .l:tde8 - then simply 2S 'iid 3 , and he
loses a pawn . Perhaps 24 . . . i.e8, to support
the d6-pawn , and vacate the d 7-square for
his knight or rook. But then 2S 't!fd3, and the
bishop is no longer defending the knight. If
2S ... 'ii'b 7 it is possible to attack the aS-pawn . Now, of course, 29 fS? ltJf4 is not good for
'Well, th is means that for the moment I Wh ite. He must defend his f4-pawn , but
should simply wait. What useful move can I how? Karpov is vigilant right to the end.
P rophylactic Thinking ftJ 37
The position is completely won for White, Petrosi a n . I n their manner of play Karpov
but a certain accu racy is stil l requ i red . Thus, and Petrosian have much in common, but
after 29 'iid 2 d5!? Black could have obtained even so it seems to me that Karpov is a more
someth ing resembling cou nterplay: 30 e5 aggressive player. For h i m prophylactic
ttJe4 31 ttJxe4 dxe4 32 t2Jd4 c5, with th i n king was always a reliable weapon in
complication s . ' playing for a wi n , whereas Petrosian usually
2 9 �e3 ! c5 ai med above all to safeg uard h imself against
30 f5 tiJdB defeat and his prophylaxis sometimes looked
excessive .
31 b5
Of cou rse, the knight should not be allowed
to go to c6 . ' For "complete happiness", it only Petrosian - G ufeld
remains for Wh ite to play c3-c4 , in order to 28th USSR Championship, Moscow 1 96 1
achieve domination over the entire board . ' King's Indian Defence
31 . . . 'it>h8 1 c4 g6
32 .llf2 2 d4 .ll g 7
The bishop has done its work on e3 and it 3 ttJc3 ttJf6
again retreats, in order to support the e4- 4 e4 0-0
point.
5 �g5 d6
32 . . . 'ili'c7
6 'ili'd2
33 lla4 'ii b 8
Wh ite has chosen an u n usual move order in
34 c4 the opening.
Apart from his material advantage, Wh ite
Once I was observing a joint analysis by
also has an enormous positional advantage
Petrosian and Gufeld . The former world
- the opponent's pieces are completely
champion was constantly outplayi ng his
starved of oxygen. The decisive break
opponent.
through is not far off.
' How come , ' asked Eduard Gufeld i n per
34 . . . .l:.a7
plexity, 'surely I have the better position?'
35 .l:.xa7 I:!.xa7
'Yes, but I have the better bra i n , ' Petrosian
36 e5 dxe5 explai ned .
37 ttJxe5 .l:la2 lig ran Vartanovich did not attach too g reat
38 .ll x c5 importance to opening theory and he often
Black resigned . took certa in l iberties , in order to take his
I repeat once more: on a superficial exami opponent away from familiar paths and make
nation the game does not attract attention, use of his 'better bra i n ' . I ndeed , why allow
and the impression is created (generally the young Gufeld , who was considered an
speaking, justified ) that the play was 'all at expert on the King's I ndian Defence, to
one end ' . And only after a serious study do demonstrate his knowledge? Wouldn't it be
you beg in to sense the g reat mastery behind better to test his positional understanding, in
Wh ite's seemingly modest moves, a mastery which he was su rely lacking? Especially
largely con nected with prophylactic th inking . since Petrosian himself had an excellent
feeling for such situations: as he himself put
And now we will analyse a game by another it, he 'fed his family for many years thanks to
legend in the field of prophylaxis - ligran King's I ndian set-ups. '
38 � Prophylactic Thinking
9. . . e5?
A serious positional mistake . By blocki ng his
bishop's diagonal, Black deprives h imself of
any counterplay, and now Wh ite's spatial
advantage guarantees h i m an enduring
initiative. Petrosian gave an instructive as
sessment of the situation:
'Outwardly the position appears highly prom
ising for Black. By blocking the pawn chain in
the centre he has transferred the weight of
the struggle to the wings, and the possibility
Black is prepari ng 1 3 .. .f5. Of cou rse , he is
of playing . . . b7-b5 and . . . f7-f5 would seem
not afraid of 1 3 ..te 7?! l:te8 14 ..txd6?? 'ii' b 6,
to give him the better chances. But if Black,
while 1 3 g4 allows a standard pawn sacrifice:
in reasoning this way, was attaching the
1 3 . . . tLlf4 14 tLlxf4 exf4 1 5 ..txf4 tLle5 1 6 ..te2
greatest importance to the mobil ity of his
..td7 followed by . . . b7-b5. And in genera l ,
pawn structu re on the wings, he should not
sharp moves such as t h i s are not in
have forgotten that it is the job of the pawns
Petrosian's style.
to clear the way for the pieces. Then he
would not have overlooked the fact that 12 f3 !
White's pieces are much better placed i n the 'A good prophylactic move , d irected in
event of the position being opened . ' particular against . . . f7-f5. It transpires that
Prophylactic Thinking ltJ 39
'A continuation of the same unhurried strate Vyacheslav Chebanenko, the wel l-known
gy. Before playing his pawn to b4 Wh ite trainer from Moldova). By moving his knight
prepares to double rooks on the b-file, which to the kingside, Black has weakened his
sooner or later will be opened . At the same defence of the important sq uares c5 and b6.
time another problem is solved : it is no It makes sense for Wh ite to exploit the factor
longer necessary to watch out for . . . b7-b5 . ' i mmed iately and to open up the game on the
When studying the present game (and also queenside.
the previous one) you should not attach too 19 b4! tt:'lg8
much sign ificance to individual moves, or try 20 ..ie3 f5
to establish whether or not they were The lesser evil was 20 . . . b6 2 1 bxc5 bxc5 22
objectively the strongest - this is not the .:txb8 'ilt'xb8, although then too after 23 .l:!.b 1
point. It is more important to follow and and 24 'iib 2 Wh ite has a clear advantage.
sense how the taking of decisions was
21 bxc5 dxc5
approached by these great players, with
whom prophylactic thinking was fully devel 22 l:.fb1 tt:'lf6
oped (in the case of Tigran Vartanovich , The wh ite rook should not have been
perhaps even slightly more than neces allowed to go to b6. Of cou rse , 22 . . . b6 would
sary!). have opened new possibil ities for Wh ite,
16 . . . ..ie7 involvi ng the advance of his a-pawn, but
nevertheless it should have been played . '
1 7 .l:!.ab1 <ith8
2 3 :b6 ..id6
1 8 l:lb2 tt:'lf6
23 . . . tt:'ld7? did not work in view of 24 d6!
'Black embarks on a new reg rouping , which
..ixd6 25 tt:'ld5. However, 23 . . . fxe4! ? was
in the end enables him to advance his pawn
more tenacious, trying to obtain the sq uares
to f5. Had he tried to do this a move or two
f5 or g4 for his pieces.
earlier, then after the exchange on f5 Wh ite
would have created unpleasant pressure on
the b 1 -h7 diagonal by 'ii'd 2--c2 . In combina
tion with g3-g4 and the opening (after b3-
b4) of a 'second front' on the queenside, this
would have led to a difficult situation for
Black. It would have been most sensible for
him to stick to passive tactics, and to wait for
Wh ite to clarify his intentions. Moves such as
18 . . . b6 followed by 1 9 . . . .l:tb7 would to some
extent have improved his defensive resou rc
es. '
How should Wh ite now combat a standard
plan in such positions: . . . tt:'lf6-g8 , and after
the retreat of the bishop - . . . f7-f5 ? He can ,
of course, resort to g3-g4. 24 ..ih6!
But let's remember a typical idea when ' Despite the strong pressure, Black still
seeking a solution - a question which in such retains hopes of obtaining cou nter-chances
cases it is useful to ask you rself: 'What is the by exchanging on e4 , when after the
drawback of my opponent's move?' (It would recapture with the pawn he can eliminate the
appear that this was first recommended by dark-square bishop by . . . tt:'lf6-g4 , while after
Prophylactic Thinking l2J 41
tt:lc3xe4 or ..lii. d 3xe4 he has . . . ltJg7-f5. It 'Now on the retreat of the rook there follows
should be mentioned that, with his over 28 . . . lbh5!, when the black pieces obtain
whelming positional su periority, Wh ite would convenient posts at d6, e5 and f4. But if
have retained an obvious advantage even if Black's plan could have been realised, this
the 'threats' mentioned above were to be would naturally have called into question
carried out, but there is no reason for h i m to White's play in the middlegame. After all,
part unnecessarily with even a portion of his from the opening Black stood badly. '
positional gains.' 28 J:ie6 !
Now if 24 ... fxe4 there follows 25 ltJxe4 ltJxe4
'All is in order! Black is prevented from
26 ..be4 l:!.f7 27 g4! , restricting both the
playing 28 . . . ltJh5 by the th reat of 29 Ite8+.
opponent's knight, and his bishop.
Simultaneously an attack is made on the
24 . . . �f7 bishop at e5 - the sole barrier i n the path of
25 lL'lg1 the passed pawns in the centre. By eliminat
Another purely Petrosian-l ike move. The ing this obstacle after 28 . . . lDf8 29 �xe5! ,
grandmaster does not h urry to take specific White would have won without difficulty. '
action , but prefers , just in case , to support ( I ndeed , 29 . . . 'tli'xe5 30 f4 'ii' h 5 3 1 f5 !? looks
his f3-point beforehand . If 25 . . . fxe4 he was th reatening, but nevertheless Black should
intending 26 ..lii. x g7+!? (after 26 ltJxe4 ltJf5 2 7 have gone i n for this position - most
tt:lxd6 ltJxd6 Black g a i n s cou nter-chances) probably he had noth ing better. )
26 . . . <.t>xg7 27 ltJxe4 ltJxe4 28 ..lii. xe4, followed 'The conclusion of the game, which my
by g3-g4 and lbh3. young opponent conducted i n time-trouble,
Black should probably have evicted the resembles 'g ive-away' chess . '
dangerous rook from b6 by 25 . . . ltJd7 . I n 28 . . . b5
stead o f this Gufeld tries t o i n itiate complica
29 cxb5
tions, which tu rn out to be not i n his favour.
Thanks!
25 . . . f4?
29 . . . c4
26 gxf4 lL'ld7
26 . . . exf4 27 J:ixd6! and wins. 30 �c6 'iVd8
'Here Black fi nally remembered that, apart pawn to c6? The point is that Wh ite has not
from mate , in chess it is also possible simply yet determi ned the position of his king . After
to resign a game, which he very belatedly q ueens ide castling one of Black's best plans
did.' is the preparation of . . . c7-c5 ( . . . ..te6 , . . . a7-
a6 and perhaps . . . .l:tc8). It is clear that in this
We will now examine a few examples, case . . . c7-c6 will tu rn out to be a clear loss
showing how prophylactic th inking helps in of tempo. A typical example of prophylaxis in
taking decisions in various stages of the the opening - taking account of the op
game. ponent's possible plans, in order to arrive at
the most accu rate move order.
Opening 1 0 0-0 c6
When analysing the following game we will 1 1 .Ua b1
at the same time recall certain strategic Wh ite is preparing the standard minority
ideas, typical of the Carlsbad structu re in the pawn attack on the q ueenside. Here 1 1 llae 1
Queen's Gambit. has also occurred , and later the rather
dangerous variation 1 1 f3! ? came into
Botvi n n i k - Keres fashion. For example, the th ird game of the
20th USSR Championsh i p , Moscow 1 952 lvanch u k-Yusupov Candidates match (Brus
sels 1 99 1 ) went 1 1 . . . ttJh5 1 2 ..txe7 'ifxe7 1 3
Queen 's Gambit
e4 dxe4 1 4 fxe4 Sl.g4? ! ( 1 4 . . . Sl.e6 ! ) 1 5 e5!
1 d4 ttJf6
.l:tad8 1 6 ttJe4 , reach ing a difficult position for
2 c4 e6 Black, roughly similar to that which occu rred
3 ttJc3 d5 in the game we have j ust begu n examining.
4 cxd5 exd5 Mikhail Moiseevich , who was present at the
s ..tgs ..te7 match , caustically remarked : 'Aha, Yusu pov
doesn't know the Botvinnik-Keres game.
6 e3 0-0
That's bad ! ' In fact Artur knew this game, of
7 Si.d3 ttJbd7
course, but at the board he was unable to
8 'ifc2 lieS find a way of avoiding the u nfavourable
9 ttJge2 ttJf8 development of events.
11 . . . Si.d6?
Threatening 1 2 . . . ..txh2+ 1 3 �xh2 ttJg4+, but
Keres's main idea is to play 1 2 . . . ttJg6 and
1 3 . . . h6, forcing the exchange of bishop for
knig ht, and to recapture on f6 with the queen.
Then his pieces will be actively placed and
pressing on the opponent's kingside. Alas,
Botvi nnik refutes this idea .
The correct move order was 1 1 . . . ttJg6 (of
course, there are also other conti nuations,
for example 1 1 . . . ttJg4!?) 1 2 b4 ? ! Si.d6 (threat
ening 1 3 . . . ..txh2+ or 1 3 . . . h6) 1 3 ttJf4 Sl.xf4 1 4
exf4 'ili'd6 1 5 .l:1fe 1 Si.d7 1 6 f5 ltJf4 , and Black
is excellently placed (Lazarev-Fedorenko ,
Why was Keres not in a hu rry t o play h i s Ternopol 1 964 ).
Prophylactic Thinking l2J 43
13 f3 ! !
This move practically wins the game, since it
frustrates Black's plans. If now 1 3 . . . h6, then
14 .bf6 ( 1 4 i.. x h6 gxh6 1 5 i.. x g6 fxg6 1 6
'lxg6+ �h8 1 7 'ii'x h6+ t'Dh7 is u nconvi ncing)
14 . . 'i'xf6 1 5 e4 with the terrible th reat of 1 6
.
i.e3 is unconvi ncing: 28 . . . lZJxd5 29 .i.d4 even mentioned . As a result, such ga mes
"le6. receive a one-sided coverage and a non
28 i.f4 is tempti ng, hoping for 28 . . . '�xd5? objective assessment. It is probable that at
29 l:td3 'ii'h 1 + 30 'itc2 ! Wxa 1 31 .i.g4+ (31 some stage of chess study this even has a
J:Ixd6? 'ii'g 1 ! ) 3 1 ... '1tb8 32 llxd6 llxd6 defi n ite pedagogical point. But when 'at a
33 'i'xd6+ 'ita8 34 'ii'xf6 and wins. In the matu re age' you again turn to them and look
event of 28 . . . 'ife4?! Wh ite does not play with q u ite d ifferent eyes, you easily notice a
29 l::t d 3 on account of the pretty reply certain naivety of such examples and their
29 . . . ltJg6! ! , poi nted out by Utut Ad ianto , but book interpretation .
29 'i'd3 ! , retaining the advantage. But it is
not evident how to cal l i nto q uestion the Gottschall - N i mzowitsch
simple 28 . . . 'ii' x h5 ! . Hannover 1 926
27 . . . lZJxd5?
27. . .'ii'x h5? would have lost to 28 llxf6 'ii' h 1 +
29 .if1 ltJg8 30 'ii'e 1 + , but the captu re of the
central pawn is l ittle better.
28 lld3 llxh6
28 . . . tt:\e7 29 .i.f4 .
29 l:txd5!
29 'i'xh6? �g5 and 30 . . . lZJe3+
29 . . . 'ike4
30 .l:i.d3 !
The rook's manoeuvres have d isru pted the
oppon ent's defences.
30 . . . 'ii' h 1 +?
Kasparov's suggestion 30 . . . 'ii h 7 was much
It is Black to move . How should the position
more tenacious - here Black would still have
be assessed? N i mzowitsch 's chances are
retai n ed chances of a successful outcome.
certainly better thanks to his superior pawn
31 'it>c2 'i!Vxa1 structu re - his one pawn on a4 is holding
32 'ii'x h6 �e5 back two of Wh ite's. To j udge by the
33 1i'g5 grandmaster's com ments, his win was the
Black lost on time. logical outcome. In fact, with correct play the
game shou l d , of course, end in a draw. After
a l l , material is equal with opposite-colour
bishops, and in addition Wh ite controls the
Endgame only open file.
The fol lowi ng ending is taken from N imzo Every player is obl iged at ti mes to try and
witsch's My System. Old books give nu mer 'squeeze' a microscopic advantage, or on
ous positions in which one of the players, the contrary, defend in a slightly i nferior
much weaker than the other, fails to put up a end ing. Therefore it is instructive to follow
worthy resistance. I n the com mentary all the the actions of N imzowitsch , who completely
attention is usually focused on the play of the outplayed his opponent, and to understand
w i n n e r, and defensive possibil ities are not why this happened .
48 � P rophylactic Thinking
since here Wh ite had the now familiar h4, when after the captu re by the bishop on
tactical stroke 33 g4+ ! . h4 he can eliminate the c5-pawn with gain of
33 fxe5? fxe5 tempo.
34 llh4? g5! Of cou rse, the win of a pawn when there are
The move 3 1 .. J:If8! has its effect - Wh ite opposite-colour bishops by no means guar
can not play 35 Ilxh5? �g6+ . antees a wi n . But all the same Black has
noth ing better, and after captu ring on c5 he
35 llb4 �e6+
creates the u npleasant th reat of a bishop
36 �e2 e4
check on b5.
36 . . Jif3? 37 .:txa4 .
38 . . . �e5!
37 .1f2 .l:tf3
39 ilb4 �d5
38 .l:tb6
A zugzwang position is created .
40 h4 gxh4
41 gxh4 .l:th3?!
New prospects a re opened for the black
rook. However, now the wh ite rook also
breaks free , and yet it could have been kept
at b4 by making the prophylactic move
41 . . . �e5 ! . The rook will i nevitably reach h3 a
l ittle later, whereas Wh ite can not play 42
.l:.d4? because of 42 . . J 1xf2 + .
4 2 .:td4+ �e5
43 .l:td8 .1d5
44 .l:te8+ .1e6
Threatening 45 . . . llb3 . The concluding stage
Nimzowitsch has achieved much , and now it of the game has been analysed in detail by
is indeed not easy to defend. 'The passed g randmaster Robert H u bner. I will briefly
pawn, the penetration of the rook into the acquaint you with his main conclusions.
enemy position, and a certain weakness in 45 l:td8 �f4
White 's c5-pawn slowly wrought the destruc
The noose a round the wh ite king is d rawn
tion of White 's game. '
ever tighter. Perhaps it would have been
How can Black strengthen his position? It is slightly more accu rate to fi rst d rive the king
unlikely that he will be able to prepare . . . e4- away from the e2-sq uare: 45 . . . .1g4+ ! ? 46
e3 in circumstances such that all possible �d2 �f4 47 llf8+ .1f5 48 'it>e2 llh2.
replies - llxc6 , .1xe3 and .1e1 will prove However, in H ubner's opinion , here too with
unsatisfactory. He would like to place his accu rate play Wh ite would have gai ned a
king on d5, but what next? I n some cases d raw.
. . . h5-h4 makes sense, althoug h in principle
46 llf8+ .1f5
the exchange of all the kingside pawns is to
47 l:r.f7 l:.h2
White's advantage .
Let us suppose that with the king on d 5 and Not i m mediately 47 . . . e3 48 .1g 1 .
the rook on b4 it is Wh ite to move. If .l:td4+ , 48 l:l.e7?
then simply . . . �xc5 , not fea ring a discov If 48 �e 1 ? , then 48 . . . e3 49 .1xe3+ 'it>xe3 50
ered check. And if llb6 Black can reply . . . h5- .l:r.xf5 llh 1 + 51 .:tf1 llxf1 + (51 . . . .l:!.xh4) 52
50 � Prophylactic Thinking
The exami nation of interesting examples on several exercises of d ifferent types, some
the theme of prophylaxis could have been easy and some more d ifficult. They are
continued . But for the mastery of prophylac u nited by just one factor: everywhere the key
tic thinking (as also, in genera l , any practical to the solution is one and the same q uestion:
skill) theory alone is i nsufficient - i ndepend 'What does my opponent want, and what
ent training is required . I i nvite you to solve would he play if it were h i m to move?'
Exercises
Sol utions
This move solves both problems and reta ins Wh ite probably sti ll has the right to go in for
somewhat the better chances for Wh ite . this variation , if he finds a set-up enabling
him to parry the opponent's immediate
21 .. .'ii' b4 22 'iti>g2 "ike7
th reats : 28 'it'b5! 1i'f7 29 'ii'e 2 .l:tf6 30 .l::t b 2.
Now Wh ite has to reckon with 23 . . . ii.g5(g7) But even here Black retains defi n ite counter
and 24 .. .f5. Therefore Ratmir Khol mov opens chances, by conti nuing 30 . . . ifg6 3 1 'iti>g2 (31
the centre. 'i!i'g2? 'ii'd 3 ! ; 3 1 l:td2 ! ? ; 3 1 l:te7 ! ? ) 3 1 . . . h5!?
23 d5! exd5? (23 . . . cxd5 24 cxd5 b6 was 32 'ii'x h5?! (32 h4) 32 . . . ifd3.
essential) 24 'ii'x e7 lbe7 25 g5! (25 ii.xa7?
After 25 g3!? e5!? an i nteresting idea was
was weaker: 25 . . . l:la8 26 ii.c5 l:te4 with
suggested by grandmaster Matthew Sadler:
equal ity) 25 ... ii.e5 26 ii.xa7 ii.c7 27 cxd5
26 l:!.be3! ? exd4 (in the event of 26 . . . e4 27
l:ted7 28 h4, and Wh ite gai ned the advan
l:tb3 the position favou rs Wh ite) 27 l:.e7 'i!Vb8
tage.
(27 . . . ifd8 28 �e8+) 28 'it'd 1 ! (th reatening 29
'i!Vg4 or 29 ifh5) 28 . . . 'iti>h8 29 'ii' h 5 .l::tg 8 30
2. Ti mma n-La rsen (Mar del Plata 1 982). .l:l.xg 7 ! ! . Let us conti nue the variation:
23 ... ii.xc5? 24 lt::l e4 lt::l xe4 25 ii.xe4 is 30 . . . .l::tx g7 31 .l::t e 8+ l:.g8 32 l:txb8 .l:txb8 33
unfavourable for Black. If Wh ite should 'ii'x d5 l:tbc8 34 'ii'x d4 'iti>g7. Wh ite has a clear
occupy the e4-square u n h indered , by play advantage, but is it sufficient for a win?
ing 24 lt::le4 , his chances will be much better. Vlastimil Hart preferred a prophylactic move,
Simple prophylaxis comes to Black's aid . enabling h i m to avoid complications.
23 . . . 'iti> h 8 ! 25 .:tf3 ! ?
Now i f 2 4 lt::l e4? ! t h e pin 2 4 . . . ii.f5 i s From the practical point o f view this decision
un pleasant. After 2 4 e3 .l:tbc8 25 lt::l e 2 'ii'f7 is very sensible - White maintains a position
Black seized the initiative . al advantage, without the risk of miscalcu lat
I should mention that his attack on the ing in complicated variations. For example, if
kingside could also have been begun imme 25 . . .'ii' b 8! (with the idea of 26 . . . e5!) he can
diately, without resorting to prophylaxis: reply 26 'ii' b 3!? 'ikxb3 27 :Xb3 with the better
23 . . . iff7! 24 lt::le4 lt::l g 4 followed by . . . 'ii' h 5. endgame (27 . . . .l:l.a6 28 .l:tb7 or 27 . . . .l:f.8c7 28
a4!?).
3. Hort-Mestel (London 1 982). After 25 . . .'ii' b 7?! 26 'ii'd 1 ! (with the idea of
I n choosing a way of defending his f4-pawn , .l:l.fe3 and 'ii'e 2 ) Black should have defended
White has to reckon with the freeing advance against the threatened breakthrough on the
. . . e6-e 5 ! . e-file by 26 . . .f5 ! .
25 .l:.f1 ? is wrong in view o f 25 . . . e5. 25 g3!? 2 6 . . Jla6? 2 7 l:tfe3 1i'd7
suggests itself, but in this case too the After 27 . . . l:txa2 28 l::t x e6 'ii' b 2 White would
Prophylactic Thinking lZJ 55
have won by 29 l:16e2 ! 1Wa3 30 .:te7! 'ii' b 2 3 1 1 8 l:e3 (with the threat of 1 9 lbxh5+) in view
'fg 4 'i'xd4+ 3 2 'it> h 1 f5 33 1Wxf5 .:tf8 3 4 .:tea of 1 a . . . :ha.
'ff6 35 'i'xd5+ 'it>h8 36 1Wxa2. 18 f4! exf4
28 'i'e2 lia4 (28 .. .'ii' a 4 29 ::txe6 'ilfxd4+ 30 1 8 . . . lbd8 1 9 f5 .
�h1 ; 28 .. J:tcc6 29 f5) 29 .l:!.xe6 .l:!.xd4 30 c6!
1 9 lbe2 lbd8
(31 J:te7 'i'f5 32 g4 ! was another way to the
goal) 30 . . .'i'f7 31 .Ue8+ Uxe8 32 'ii'x e8+ 'ilff8 Little better was 1 9 .. .f3 ! ? 20 gxf3 lbf6 21 lbf4
33 'i'xf8+ lbh7 22 lbd5 'ii'd 8 23 'ii'd 2 with advantage to
White.
Black resigned in view of 33 . . . 'it>xf8 34 :c1 .
20 lbxf4
4. Tai-Ribli (European Team Champion By taking control of the e6-sq uare at just the
ship, Skara 1 980). right time, Sergey Makarychev has prevent
ed the important defensive move . . . f7-f6.
23 'ii' b 5!
Now the attack develops of its own accord .
But not 23 'ii'a4?! i..f8 .
20 . . . c6 21 'iid 3 .l:.h8 (2 1 . . . lbf6 22 e5! ) 22
23 . . . a6
l:te2 ! f6 23 lbge6+ lbxe6 24 lbxe6+ 'it>h7 25
23 .if8 24 llxf6! 'ii'xf6 25 'ii'd 5+ and 26
...
.l:tf1 lbf8 26 e5! dxe5 27 dxe5 lbxe6 28 exf6
'fxa8. 'ii'c 5+ 29 'it> h 1 ltJfB 30 .l:te7+ 'itr>h6 31 'ii'e4
24 'i'd5 'ii'x d5 lbd7 32 f7 .l:tf8 33 �c2 Black resig ned .
25 cxd5 Of cou rse, it would be strange to call the
By preventing the development of the knight attacking move 1 8 f4 ! prophylactic. But at
on b8, White achieves a winning position . any event it was found with the aid of
23 'i'd 1 ? ! is far weaker in view of 23 . . . lbc6 prophylactic thinking , suggesti ng the need to
24 l::!. d7 l::!. d 8! . The conti nuation in the game prevent the opponent from strengthening his
was also unsuccessfu l : 23 lbd2? ! lbc6 24 position .
tt:le4? (24 'ii'a4 'ii'e 8 25 lbe4 f5 26 lbg5 was
stronger) 24 . . . lba5 25 'ii' b 5 lbxc4 (th ree
6. Psa khis-Speelman (Hasti ngs 1 987/88).
moves earlier could the knight have dreamed
of such a fate? ! ) 26 .l:tc6 (26 l:r.d7? lbxe3!) Black is the exchange u p , but his knight is in
2 6 . . l:l.xc6 27 'ii'x c6 .U d 8 , and Black equal
.
danger. Wh ite is th reatening �b7-a6-d3. If
ised . 29 . . . lld2 30 �xd2 l:!.xd2, then 31 'it>e1 .l:1d7 is
u nclear, but 3 1 1Ic1 ! is very strong.
29 . . . g5! !
5. Makarychev-Be l l i n (Hastings 1 979/80).
Black wants to play 1 8 . . . lbd8 and 1 9 . . . f6 , 30 �a6 f4
driving away the menacing knight on g5. 3 1 i.. d 3
White also has to reckon with the manoeuvre The waiting move 31 i.. b 5 (or 3 1 �c4) came
. . . tt:ld7-f6-h7 and with 1 8 . . . 'ii'f6 . 1 8 .l:tf1 lbd8 into consideration .
19 f4 f6 20 lbh3 (with the th reat of 21 'ii'x h5!) 31 . . . .l:lxd3
is not bad , but in this case Black can
32 exd3 llxd3
successfully defend : 20 . . . exd4 21 cxd4 f5 . It
would be desirable to find a more active way 3 3 'it>e2 .l:td5!
of playing, after which the opponent does not It becomes clear why the kingside pawns
succeed in implementing his plan ned defen were advanced : if 34 l:.xd 1 ? Black now has
sive constructio n . But noth ing is given by 34 . . .f3+ . 34 i.. d 2? also does not work in view
56 <;i? Prophylactic Thinking
of 34 . . . tt::J xf2 ! . I ncidentally, in the event of 8 . Kozu i-Ma rja novic (Yugoslav Champion
33 . . . .l:!.d4? (instead of 33 . . . l::t d 5) Wh ite would ship, Novi Sad 1 985, variation from the
have won material by 34 gxf4 gxf4 35 .id2 ! game).
tt::Jxf2 36 .ic3 l::t d 3 37 .ie1 . It is clear that Black has to play for
Now White should probably have played 34 zugzwang. White will have to sacrifice his
'iitf3. If 34 . . . .l:ld4 he retreats his king with 35 knight on f5, since if it moves anywhere else
'iot>e2 , creating the th reat of 36 gxf4 gxf4 37 the rook will invade on the 2nd rank.
i.. d 2!. And 34 . . . tLlc3 35 i.. x c3 .l:f.d3+ 36 'iot>g4 After 64 tt::Jxf5 .llxf5 Wh ite has two possibili
.l:f.xc3 37 �xg5 leads to an unclear rook ties: 65 e4 and 65 g4. Before turning to
ending. calculation , let's see whether it is possible to
Lev Psakhis chose 34 gxf4 gxf4 35 .l:tc1 ? prevent at least one of them, i n order to deny
(here too 35 'iot>f3 was necessary). After the opponent a choice.
35 . . . tLlb2 36 .ic3 tiJd3 Black went on to win . 63 . . . �g 1 ! !
I n the event of 6 3 . . . .l:f.a5? 64 tt::J xf5 Itxf5 65
7. Stei n-Keres (Moscow 1 967). g4 ! hxg4+ 66 �xg4 Wh ite would have easily
It would appear that there is noth ing to th ink gained a d raw: 66 .. J:lf8 67 f5 'ii?e 2 68 �f4!
about here, and that Wh ite must defend his 'ii? d 3 69 'ii? e 5! 'ii? x e3 70 f6.
e5-pawn with 19 f4 . But let's ask ourselves 64 tt::Jxf5 .l:f.xf5
what Black will do then . And we establish And 65 g4 is not possible because of
that he is plan ning to defend his d5-pawn by 65 . . . h4! 66 gxf5 h 3 .
19 .. Jlad8 and then play 20 . . . c4! , including Now let us analyse the conseq uences of 65
his knight via c5 and hoping at some point to e4 .
advance . . . d5-d4.
65 e4 .l:f.a5 66 f5 .l:f.a3+ 67 'ii?t4 'iiff2 68 f6
Leonid Stein finds a way of forestalling the
Other continuations also do not help:
opponent's idea .
68 'ii? g 5 'ii? e 3 (68 . . . 'ii?x g3 is also good ) 69 e5
1 9 a4! !
'ii?e4 70 e6 .Uxg3+ 7 1 'ii? h 4 l:.g4+! 72 �xh5
Now after 1 9 . . . .l:tad8 Wh ite has the strong 'ii?xf5;
reply 20 axb5 axb5 21 .Ua6. If 1 9 .. .'ii' x e5 ,
68 e5 .l:f.f3+ 69 �e4 (69 'ii? g 5 'ii? e 3 70 e6 �e4
then 20 I1e 1 ! 'ifd6 (20 . . . .ic7 2 1 i.. g 1 ) 2 1
7 1 f6 .l:f.f5+ ! , transposing into the main
axb5 axb5 2 2 l::tx a8 l::tx a8 23 i.. xf5 , a n d the
variation 68 f6) 69 . . . �xg3 (69 . . . l::t x g3 70 f6
opening of the position is clearly to the
.l:f.f3! is also possible; if 70 e6 Black has both
advantage of White, who has the two
70 . . . h4 7 1 f6 h3 72 f7 .l:f.f3 73 e7 h2 and
bishops. It is no better to play 1 9 . . .f4 20 .if2
74 . . . h 1 'ii' , and 70 . . . l::tf3 71 �e5 h4 72 e7 h3)
'ii'x e5 2 1 axb5 axb5 22 'i!kd 3 ! .
70 e6 .l:.f1 ! 71 'ii? d 4 .l:.e 1 72 'ii? d 5 'ii?f4 (72 . . . h4)
The game continued : 1 9 . . .tLl a 5 2 0 i.. f2 ! 'it>h8 73 f6 l:le5+ ! 74 'ii? d 6 'ii? f5.
(20 .. .f4 21 b4 ; 20 . . .'i1Vxe5 21 l:.e 1 ! 'ii'd 6 22
68 . . . I:If3+ 69 �g5 'ii?e 3!
axb5 axb5 23 b4 ! ; 20 . . . tt::J c4 21 b3! tt::J x e5 22
Not 69 .. .'�xg3? 70 e5! (70 'ii? x h5? �f4 ! )
axb5 axb5 23 l::tx a8 llxa8 24 Iie1 or 24
70 . . . h 4 7 1 e 6 h3 72 e7 h2 73 e8W h 1 'it' 74
i..xf5!?) 2 1 .f:.e1 .l::[ a 7 22 'ii'e 2! b4 23 cxb4
We5+ with a d raw.
cxb4 24 i.. x b6 'ii'x b6 25 l:1ad1 'i¥c5?
(25 . . . b3 26 .ib1 'iie 6 was more tenacious) 70 e5 �e4 71 e6 l:tf5+! 72 'ii? g 6 �e5 73 e7
26 .id3 ! , and soon Wh ite converted his l:.xf6+ 74 'ii? g 7 .l:!.e6 75 'ii?f7 'ii?tS , and Black
advantage. wins.
tZJ 57
Max D l u g y
A Novelty i s born
ECO. I stopped to think. Once again I need to I n the ending only Black will have win n i ng
complicate the game and make an effort to chances.
solve my open ing problems at the same 1 7 ii.e3 'ii'x d4
time. EGO g ives 1 1 . . . lt:'ld7 , 1 1 .. .'ii' c 7 and 1 8 ii.xd4 e6!
11 . 'i'c8 as the possible moves , evaluating
The simplest way - Black completes his
. .
Mark Dvoretsky
Sol utions
This variation had to be calculated by Wh ite But, u nfortu nately, most of you chose a
when he embarked on his operation. His resou rceful , but not very successful way of
position is won , for example: 20 ... .ltxc6 parrying the opponent's main idea - 1 8 . . . b5?!
(20 . .'i'h3 21 'ii'xf3) 2 1 'ii'c 2 'ii' h 3 22 'ii'x h7+
. (for it only half a point is awarded ). After 1 9
'fxh7 23 ..txh7+ <it>xh7 24 ..txf8 . cxb5! followed by 2 0 lt:Jc4 Black does not
Jozsef Pinter found a purely concrete way of have sufficient compensation for the sacri
achieving an advantage. Only by a great ficed pawn .
stretch of the imagination can this example Less convincing is 1 9 'iix b5? ! , hoping for
be called 'positional' (Wh ite carried out a 1 9 . . . .l::ta b8 20 'iig 5 'ii'x g5 2 1 fxg5 l:txb2 22
series of favourable exchanges). But equal gxf6 l:txd2 23 fxe7 .l:txe7 (or 23 . . . .l::tx a2 24
ly, White's solution can not be called combi .l::t a 1 ) 24 .l::tf2 , when in the rook ending the
native - after all, noth ing was sacrificed . It l im it of Black's dreams is a draw. It is
would perhaps be more correct to cal l it stronger to i nterpose 1 9 . . . lt:Jg4 ! , and only
tactical . after 20 h3 - 20 .. Jlab8 (but not 20 . . . l:teb8?
Tactics a re an immeasu rably broader con 21 'iig 5 'ii'x g5 22 fxg5 .l:r.xb2 23 tt:Jxe4 ).
ception than combinations. When we say
that Emanuel Lasker was a great tacticia n ,
we don't mean that he was constantly
sacrificing something. No, simply the world
champion was excellent at fi nding the strong
est resources for both sides - accurate
moves , precise variations.
Tactical skil l plays an enormous role i n
chess , and b y no means o n l y in sharp
combinative situations. With its help a player
can tenaciously hold difficult positions, con
stantly erecting new barriers i n the oppo
nent's path , or on the contrary, he can find
the quickest way to convert an advantage .
Even the solving o f purely strategic problems
in quiet positions can not be done without The reply 21 �a4? suggests itself, but it is
tactical elements - after a l l , our plans can prettily refuted by 2 1 . . .lt:Jf5! 22 hxg4 lt:Jh6!
only be carried out by means of specific with irresistible threats. Wh ite's position also
moves, which have to be seen and, if looks anxious after 21 'ii' a 5?! :xb2 22 hxg4
necessary, calculated . 'ii'x g4 23 :f2 lt:Jf5! 24 lt:Jxe4 l:tb6 followed by
25 . . . l:th6. He is forced to play 21 'ii'g 5! 'i!i'xg5
22 fxg5 l:txb2 23 tt:Jxe4 ! lt:Jg6 24 lt:Jc5 lt:Jh4 25
4. Lisitsyn-Tolush (Len ingrad 1 938).
hxg4 or 24 . . .lt:Jxe3 25 .l:r.f2 with roughly equal
What does Wh ite want? Without exceptio n , chances.
a l l the participants in t h e competition correct
The strongest is a cool-headed prophylactic
ly decided that Black should be thinking not
move.
about the defence of his c7-pawn , but about
parrying the th reat of the q ueen exchange 18 . . . h6!
('fg5). It is incorrect to play 1 8 . . . c6? (or After 19 'ii'x c7?! lt:Jf5 there is no satisfactory
18 . b6?) 1 9 'ii'g 5! lt:Jg6 ( 1 9 .. .'i!i'xg5 20 fxg5
. . defence against the threat of 20 . . . lt:Jg4 . For
lt:ld7 2 1 tt:Jxe4) 20 'ii'x h4 lt:Jxh4 2 1 f5! . example: 20 l:!.e2 lt:Jg4 21 g3 lt:Jxg3 (2 1 . . . 'ikh3
66 � Positional Exercises
'i!i'd2 ) 22 . . . lle8 23 ltJxh7+ 'it'g8 24 ltJgS .i.h6, The solution that you are looking for does not
when the position remains unclear. necessarily lead to i m med iate success. lf the
opponent also rises to the occasion and finds
6. Hort-Karpov (Amsterdam 1 98 1 ). the best response, the outcome often re
mains u n clear. That is the case here:
Black obviously wants to play . . . b6-bS ,
although 14 a4 ! was u ndoubted ly correct, if
supporting his c4-pawn and preparing . . . ltJe4
Black had replied 1 4 . . . ltJe8! he would have
or . . . ltJb6 . The same reply 1 4 . . . bS! follows
retai ned a defensible position. I ncidentally,
both after the attempt to undermine the pawn
1 4 . . . ltJe4?! was weaker in view of 1 S i.xe7
chain by 14 b3, and after the tem pti ng 1 4
'i!i'xe7 1 6 ltJxe4 dxe4 1 7 ltJd2 bS 1 8 b3!.
tt:'le5.
1 4 a4!
7 . Geller-Fischer (Curac;ao Cand idates
An important prophylactic move. Now 1 S b3 1 962).
and 1 5 ltJeS are dangerous positional th reats
Wh ite has to reckon with the threat of
for White.
29 . . .'it'xaS. He doesn't want to put his roo k
Anatoly Karpov played badly and was soon on a 1 - this is too passive. Noth ing is
in serious d ifficulties. ach ieved by 29 'ii' b 6 'i¥xb6 30 l:.xb6 (30 axb6
1 4 . . i.c6?! (in order to answer 1 S b3 with
.
'ili>f8 ) 30 .. .'.ti>f8! (of course, not 30 . . . .l::!. x aS? 3 1
15 . . . b5) 1 5 ltJe5 \!Vc7 1 6 ltJxc6 'i!i'xc6 1 7 i.f3 l:td6) 3 1 d6 .l::!.x aS 32 h3 .l:!.cS , a n d he is also
u nsuccessful with 29 l:td 1 'i!i'xaS 30 \!VxaS
l:!.xaS 31 d6 i.. d 7 32 l:tb 1 bS.
But if he were able to play his rook to b6 . . .
This would solve the problem of the aS
pawn , and Black's central blockade would
prove insecure.
29 'ii'a 4! i.d7
29 . . . 'it'f8 30 l:tb6.
30 \!Val !
31 .l:.xb7 is th reatened, and if 30 . . . i.c8 there
follows 3 1 l:tb6 with a decisive positional
advantage. Black is forced to captu re the aS
pawn with his rook, al lowing the cou nter
stroke on b7. 30 . . . 'ii'x aS 3 1 'ii'x aS .l:!.xaS 32
After the exchange of Black's bishop his dS
.l:!.xb7 is totally bad for h i m .
pawn has become weak. 1 8 ltJxdS is
30 . . . .l:!.xaS
threatened . After 1 7 . . . .l:tfe8 Wh ite has the
strong reply 1 8 i.xf6 ltJxf6 1 9 e4 , while if 31 .l:!.xb7!
1 7 . . .l::!. a e8 , then 1 8 b3! bS 1 9 axbS axbS 20
.
Diana Darchiya and Sergey Movsesian
'i'f5!. suggested 3 1 "ii' e 7? ! , which is much wea ker
1 7 ... i.b4? 1 8 ltJxd5! ltJxdS 19 �fS (but not - because of this they each lost half a point.
1 9 li'xc4? �xc4 20 l:txc4 in view of 20 . . . bS The completely correct solution was found
followed by . . . ltJ7b6) 1 9 . . .'il¥xa4 20 .i.xdS by Vova Baklan and Vad i m Zviag intsev.
.l:!.ac8 21 b3! cxb3 22 l:lxc8 llxc8 23 'i!i'xf7+ 31 . . . 'i!i'xb7
'.t>h8 24 .i.xb3 'ii' b S 25 i.e6 l:.f8 26 i.xd7! 31 . . . .l::!.x a3 is hopeless: 32 .l::!. x c7 .:a 1 + 33 .i.f1
(not 26 'ii' x d7?! 'i!VhS ! ) Black resigned. i.fS 34 g4! (34 f3 hS 3S 'ittf2 l:ta2+ is fa r less
68 � Positional Exercises
convincing) 34 . . . i.xg4 35 h3!? (35 'it>g2) 35 . . .'i!Ve4 36 i.f3 �d4?! (36 . . . �d3 was more
35 . . . .txh3 36 'it>h2 - after the exchange of tenacious) 37 'ii'x d4 exd4 38 g4! ii.c8
bishops, the con nected passed pawns in the 38 . . . a5 39 gxf5 a4 40 d6 '.tf8 41 c5 or
centre decide the outcome. 38 . . . i.. c2 39 c5 d3 40 c6 i.a4 41 d6 was no
32 'i\Vxa5 better.
White has achieved a decisive positional 39 c5 aS 40 c6 �f8 41 d6
superiority. Here the game was adjourned , and Fischer
32 . . . g6 resigned : 41 . . . 'it>e8 42 i.d 1 , 41 . . . d3 42 d7
33 h3 'iVb1 + i.xd7 43 cxd7 'it>e7 44 i.c6 d2 45 i.a4 , or
4 1 . . . . a4 42 c7 a3 43 i.c6 a2 44 d7 i.xd7 45
i.xd7 a 1 'iV 46 c8'ii' + .
queen, whereas the black b7-pawn is weak structure, by advancing his e-pawn: 1 7 . . . e6!?
and will soon be won . 1 8 dxe6 fxe6 (recommended by Alexander
42 'iti>xd2 'ii'e4 Shabalov), or 1 7 . . . b5 1 8 tt:Ja2 e5! 1 9 dxe6
43 'ii'e 3 fxe6 .
PART I I
Ma noeuvri n g
O positional
ne of the most complicated elements of thought from N imzowitsch's book g ives an
play is the sh ifting of the excellent descri ption both of the events in
struggle from one part of the board to this game, and of the essence of manoeu
another. The point is that a game is rarely vring against weaknesses:
won by breaking through the opponent's 'The process of manoeuvring against two
defences at one place. Normally one has to weaknesses can roughly be characterised
seek roundabout ways and try to give the as follows: two weaknesses, in themselves
opponent new weaknesses , in order to then quite defendable, are in turn put under fire,
began manoeuvring against them. This the attacker relying mainly on his te"itorial
procedure is closely linked with the principle superiority - his superior lines of communi
of two weaknesses , which we have men cation. The game is lost because at some
tioned many times. This is one of the most moment the defender is unable to keep pace
important components in the technique of with the opponent in speed of regrouping. '
converting an advantage.
If the following game had been played by Anand - Kamsky
Aaron Ni mzowitsch , it would certainly have
PCA Candidates Match , 9th Game, Las
been incl uded i n the chapter 'Manoeuvring
Palmas 1 995
against enemy weaknesses when possess
ing a spatial advantage' from his book Chess Ruy Lopez
Praxis. 1 e4 e5
It is usefu l to follow how Anand constantly 2 ltJf3 ltJc6
changed the direction of the attack, creating 3 i.. b 5 a6
one problem after another for the opponent, 4 i.. a 4 ltJf6
and how subtly he combi ned offensive and 5 0-0 i.. e 7
prophylactic actions. The superficial impres
6 .l::.e 1 b5
sion, that Gata Kamsky lost the game
submissively, proves deceptive : simply he 7 .ltb3 d6
was confronted with insoluble problems, 8 c3 0-0
defend ing on different fronts. The following 9 h3 .ltb7
72 � Manoeuvring
dxc5 Black has 1 9 . . . llac8 ! ( 1 9 . . . dxc5 is 22 .i.xg7 �xg7 23 axb5 axb5 24 J:.a7 l:tb8
weaker because of Anand's suggestion 20 25 tLle3 h5, although in this case too Wh ite
'lxc5 tLlxe4 21 .i.xe4 .i.xb2 22 .i.xb7 ! ) . holds the i n itiative.
However, then he would have t o reckon with 22 .i.xg7 �xg7
1 9 e5!?. 23 tLle3
19 .i.c3 c5 This knight causes Black serious anxiety.
20 d5 'ii'e 7?! The th reat is 24 tLlg4. He is forced to weaken
Anand gives a n interesti ng assessment of somewhat his castled position.
the position : 'It is not hard to see what 23 . . . h5
White's advantage comprises. He has two Anand also analyses other possibilities for
"bad pieces" - bishop at b 1 and rook at a 1 , Black:
but both conta i n considerable potential ener
1 ) 23 .. .'it'f6 24 tLlg4! 'ii'x a 1 25 'ii'x h6+ �g8 26
gy. It will be easy for me to open the a-file "at
e5! .l:.xe5 (26 .. .'ii' c3 is weaker: 27 .i.xg6 fxg6
my leisure", while the bishop is restra i n i ng
28 lLlf6+ tLlxf6 29 'ii'x g6+ �f8 30 1i'xf6+ �g8
the f7-f5 break. But Black's knight at b4 and
31 .l:te4 ! ) 27 tLlgxe5! dxe5 28 .i.xg6 1i'xe 1 +
bishop at b7 are simply bad - and this is the
29 ttJxe 1 fxg6 30 'iix g6+ with advantage;
main distinction . '
2 ) 23 . . . tLlg5 24 ttJxg5 hxg5 (or 24 .. .'ii' xg5 25
I n view of the constant threat o f a n i nvasion
axb5 axb5 26 f4 ! 'with strong prospects on
on the a-file it is extremely d ifficult for Black
both wings' - Anand) 25 axb5 axb5 26 l:r.a5!
to carry out a blockade on the dark sq uares,
'ilkc7 27 lla7 .l:ta8 (if 27 . . . 'ii' b 6, then 28 tLlf5+
which is practically his only pla n : . . . "iie 7 and
gxf5 29 1i'xg5+ 'it>f8 30 'ikh6+ 'it>e7 31 e5!) 28
. . lL'ld7 . Therefore he should have consid
.
26 lDd1 ! !
An excellent and timely manoeuvre, consoli
dating Wh ite's advantage. He prevents the
defence-relieving move . . . l2Jg5, and the
opponent now has to reckon with the central
breakthrough e4-e5. But, above a l l , it tran
spires that in the enemy position there is a
serious weakness - the b5-pawn . The same
manoeuvre, but without the preliminary
exchange on b5, would have been weaker in
view of 25 . . . bxa4 ( cf. the note to Black's 24th
move).
The following stage of the game can serve as
a textbook illustration of how to manoeuvre
agai nst enemy weaknesses. 32 'ii' h 6!
26 . . . l2Ja6 Wh ite makes use of all the space on the
Black is forced to defend passively. If board and all the resou rces of the position to
26 . . . .Ua8 Wh ite has the u npleasant 27 l:.xa8 d isrupt the coordination of the opponent's
.:.Xa8 28 l2Jc3 'ii'd 7 29 e5. forces. Now the threat is 33 e5 dxe5 34 d6
27 l2Jc3 b4 and 35 'ii'x g6+ .
28 lDb5 32 . . . 'ii'f8
28 l2Ja4 with the idea of lDb6-c4 also looks The only move. The su icidal 32 . . . l2Jxe4? is
good . meet by the simple 33 .Ua2 with the
irresistible threat of 34 l:tae2 .
28 . . . l2Jc7
33 'ii'g 5
29 .i.d3
33 1i'f4 is also good .
Even the exchange of knights does not bring
Black any rel ief. The light-square bishop 33 . . . 'ii'g 7
moves with gain of tempo to an active If 33 . . . l2Jh7 34 1i'f4 , and the black pieces are
position. A structure favourable for White has even more badly placed . 33 . . . 'ife7 is no
Manoeuvring ltJ 75
sky).
78 <;i( Manoeuvring
Alekh ine - Sterk axb4 lt:Jb3, and not 22 'ii'c 2 'ii'd 5 ! , but wins a
Budapest 1 92 1 piece with the simple 22 'ii'a 2! lt:Jd3 (22 . . . 'i!i'd5
23 axb4 lt:Jb3 24 l:.d 1 ) 23 l:.d 1 .
20 . . . i.a5
21 .l:tab1 'ii'a 6
22 .l:!.c4 lt:Ja4
If 22 . . . l:tac8 there could have followed 23 b4
lt:Ja4 24 b5 or 23 . . . lt:Jd7 24 .l:r.e4 and wins.
Black parries this th reat (if 23 b4 he has
23 . . . lt:Jc3), but now, when the opponent's
pieces are tied up on the q ueenside,
Alekh ine unexpectedly switches the play to
the kingside.
20 llca 1 !
20 axb6 was weaker in view of 20 . . l:.xb6 2 1
.
13 . . . r3ile7
14 aS li:Jd7
If 14 . . .:!.c8! ? 1 5 l:thc1 lt:Jc6 1 6 l:ta4 b6 1 7
.
Acting in accordance with the 'two weak [Black incorrectly restricts himself to passive
nesses' principle, White changes the d i rec defence. He should have tried 34 . . . e5!,
tion of the attack, shifting the emphasis to the intending the manoeuvre . ltJf�6 - Dvoret
. .
kingside. sky.]
28 . . . hxg5 34 'it>b4 'it>c7
29 hxg5 ltJd7 35 'it>a5 ltJb6
29 . . . ltJe8 30 Ith 1 ltJc7 is bad because of 3 1 If 35 . . . ltJb8 , then 36 .i.b5 ltJc6+ 37 i.xc6
.l:!.h7! 'it>f8 3 2 .l:.h8+ 'it>e7 3 3 'it>c1 ! , a n d if 'it>xc6 38 b4 is strong , with a won rook
33 . . . ltJxa6 Wh ite has 34 l:r.a8 . ending.
30 .:th1 l:ib8 36 .l:lh2
31 .l:th7 .l:lg8 After the immediate 36 b3 I did not l i ke the
reply 36 . . . ltJc8 with the idea of . . . ltJe7-c6 .
Better defensive chances were offered by
31 . . . 'it>f8 . 36 . . . l:tcB
37 b3 'it>d7
38 lth7 .l:lg8
39 f4
A useful move, since now Black has to
reckon with a possible f4-f5 .
39 . . . 'it>e7
40 .i.b5
The tempti ng 40 'it>b5 does not promise any
immediate gains, in view of the u nexpected
resou rce 40 . . . ltJa8 ! , and if 41 'it>c6??, then
after 4 1 . . . .l:tb8 Wh ite is mated.
40 . . . ffi
Fearing the manoeuvre of the bishop to b7,
Black tries to i n itiative cou nterplay on the
32 c31 kingside. But this attempt is parried by White,
Now, when the black rook has taken u p a who reverts to his i n itial plan - the advance
passive position on the kingside, Wh ite of his king i nto the opponent's position via
again changes the d i rection of the play and b5.
activates his king . I n the process the weak b 41 .i.e2 fxg5
pawn is exchanged , but this is not so 42 fxg5 ltJd7?!
important, since Black remains with a real
It is possible that the best defence was
weakness on the queen side - he has to take
42 . . . ltJa8 ! ? , trying to prevent the wh ite king's
measures against the breakthrough of the
manoeuvre.
king to the a7-pawn . 32 'it>c1 (with the idea of
playing the king to b3) was less accu rate , 43 'it>b5
since Black would have succeeded in gain After the activation of the king Black's
ing counterplay after 32 . . . ltJb6 33 'it>b1 defences begi n to creak.
ltJc4 . 43 . . . ltJf8
32 . . . bxc3+ 44 .l:lh2 ltJd7
33 'it>xc3 'it> dB 45 'it>c6 .l::r. c 8+
Manoeuvring t2J 83
Dvoretsky.]
I n stead of 50 'it>b7 the more cun n i ng move
50 'it>b6! is possible, and if 50 . . . l:ta3 51 'it>b5
followed by 52 .i.a4.
50 �c2 .l:.c4
52 �d3 .l:.b4
After gaming an important tempo, Wh ite
again changes the direction of the offensive,
strengthening his position on the kingside.
52 g6 e5
If 52 .. J1b3 there would have followed 53
Black appears to have gained cou nterplay, �c2 lhe3 54 'it>b7 .l:te2 55 l:txf8 ! and wins.
but White has a nother attacking resou rce, The move i n the game also fails to ease
which he had to foresee i n advance . Black's position.
48 .l:th8! 53 .i.f5! exd4
With my small army I create th reats to the 54 exd4 l:txd4
opponent's king! It transpires that 48 .. J::txe3
55 'it>b6 l:ta4
is bad because of 49 .i.h5 ltJf8 50 l:tg8 g6 5 1
Or 55 . . . l:tb4+ 56 'it>a5 .l:tb2 57 l:th4 and the a
l:.g7+ 'it>d8 5 2 �xg6. Black i s torn i n two: he
pawn decides matters.
has to keep watch over the dangerous
passed a-pawn and at the same time repel 56 l:tg8!
the invasion of the white pieces on the The simplest and most thematic solutio n ,
kingsi de . illustrating t h e strength o f p l a y against two
48 . . . ltJf8 weaknesses.
49 .i.d1 l:tb4 56 . . . ltJe6
49 J:txe3 was more tenacious. If 50 'it>b7
. .
57 a7
Black has an u nexpected defence: 50 . . . .U.e 1 ! Black resigned .
Alexey Kosikov
How to d raw u p a P l a n
Botv i n n i k - Zagoryansky
Sverd lovsk 1 943
Reti Opening
1 l"Llf3 d5
2 c4 e6
3 b3 l"Llf6
4 .1i.b2 .1i.e7
5 e3 0-0
6 t"Llc3 c5
After 1 'it>g7 h4 2 'it>f6 'it>b6 (2 . . . h3 3 'it>e7 ) 3 7 cxd5 t"Llxd5
'it>e5! Wh ite saves the game only because 8 t"Llxd5 exd5
he simultaneously has two threats: to stop
don't wish to dwell on the opening
the enemy pawn by 4 'it>f4 and to support his
subtleties. I will merely remark that it would
own pawn by 4 'it>d6. Black can easily parry
have been more promising for Black to
either of these threats, but not both of them .
capture on d5 with his quee n , and if with the
The idea of simultaneously creating two pawn, then a move earlier.
threats will serve as our starting point for
9 d4 cxd4
understanding the process of compiling a
plan. 1 0 ii'xd4 .1i.f6
1 1 ii'd2 l"Llc6
The plans which we make pursue the aim of 1 2 .1i.e2 .1i.e6
creating th reats , but not simple, tactical Black plays too passively. I would have
threats, as in the example we have just preferred to develop the bishop at g4.
analysed , but long-term strategic threats. 13 0-0 .1i.xb2
In trying to master the technique of convert- 1 4 ii'xb2 ifas
How to d raw up a Plan ctJ 85
22 e4 was th reatened .
22 �b2 .l::. c 8
23 'i!i'e5 l:tcd8
By repeati ng moves Wh ite has gai ned time
on the clock.
24 l:!.d4 aS
The games which we have exam ined were the other weakness should be the e6-pawn . '
played rather a long time ago. Of cou rse, (Shirov)
modern players have assi milated the les 32 . . . f5?
sons of the past and successfu lly make use This makes things easier for the opponent.
of the same strategy. 32 . . . fxe5 33 dxe5 Wf8 was stronger, but here
too after 34 .l:!.g2 Black's position remains
S h i rov - Kinsman d ifficult. For example: 34 . . . .l:!.e7 35 �d 1 ! ..tg8
Paris 1 990 (35 . . . l:lcd7 36 l:td6) 36 l:lgd2 (note that White
immed iately switches to exploiting the new
weakness which has a risen - the d-file)
36 . . .We8 37 Wf4 followed by Wg5 and f3-f4-
f5 with a n easy wi n . Or 34 . . . ..tg8 35 .l':Ibg 1 ,
intending h4-h5-h6, and the weakness of
the e6-pawn does not allow Black to play his
bishop to f5.
33 l:tg2 g6
34 l:tbg1 l:tc8
35 .llx g6 f4+
36 Wd3 licd8
37 l:tf6 I!.xd4+
38 �c3 lld1
It is Wh ite to move . He has an obvious 39 l:tg7 �c1 +
advantage. How best to exploit it? 40 Wb3 .l:i.b1 +
The first part of the standard plan has 41 '.t>c2
already been completed . In this commentary Black resig ned .
on the game Alexey S h i rov writes:
'One weakness (the b7-pawn) is securely I n positions without cou nterplay for the
fixed . It is also important that the black rooks opponent, such as those we have examined ,
are tied down . . . But back i n my child hood I each of us would feel very comfortable. But
was taught that to win you need at least one i n practice th ings are usually much more
more weakness. And it turns out to be - the complicated and it is not often that the
g7-pawn . ' principle of two weaknesses can be put i nto
3 0 h4 gxh4 practice in such pure form . To d raw up a plan
In the event of 30 . . . .ltf7 31 hxg5 fxg5 Wh ite in sharper situations one also has to be
will at some point play f3-f4 (but, of cou rse, g uided by other pri nciples. One of these
not immed iately 32 f4? gxf4+ 33 gxf4 e5!), pri nciples, wh ich , i ncidentally, is by no
when the pattern of the game remains means well known , will now be described .
roughly the same.
31 gxh4 Ji..f7 Ka l i kshte i n - Vysoc h i n
32 e5! C I S J u n ior Championship, J u rmala 1 992
The point of Wh ite's idea. When he begins Slav Defence
attacking the g7-point, the b7-pawn will no 1 c4 tt:lf6
longer need to be defended ; this means that 2 tt:lc3 c6
90 <;t> How to draw up a Plan
3 d4 d5
4 li'lf3 dxc4
5 e3 b5
6 a4 b4
7 li'lb1
The knight more often retreats to a2, in order
after the capture of the c4-pawn to conti nue
.td2 and li'lc1 -b3. The player with Wh ite is
obviously not aiming for an opening advan
tage and is intending to transfer the entire
weight of the struggle to the midd legame.
7 . . . .ta6
8 li'lbd2 e6?! - position after 1 7 . . . �b7 -
Theory recommends 8 . . . c3 with equal ity, but
the game continuation is also q u ite possible. pawn on f2 it is always possible to place a
9 li'lxc4 .te7 barrier in the path of the opponent's bishop
1 0 .i.d3 0-0 by f2-f3 , but now this becomes impossible.
1 1 0-0 li'lbd7 The e4-point is weakened .
1 2 b3? ! The c 1 -h6 diagonal is blocked , and the
A superficial move! 1 2 li'lce5 is more logica l , bishop on b2 is now altogether without
a n d after 1 2 . . .'ii'c8 Wh ite's position remains prospects .
slightly preferable. I n positions with an isolated d4-pawn the
12 . . . c5 move f2-f4 is sometimes made, but only
when there is a hope of playing f4-f5. But
1 3 ..ib2 l:tc8
here Black immediately prevents the further
14 .l:tc1 cxd4
advance of the pawn .
1 5 exd4?
18 . . . g6!
There was no need for Wh ite to g ive h i mself
19 .U.f2
an isolated pawn . However, Black would
Wh ite's active moves have come to an end
also have been excellently placed after 1 5
and he beg ins marking time, whereas the
li'lxd4 li'lc5 followed by . . . li'ld5. White feels
opponent consistently strengthens his posi
the weakness of his c3-square, which he
tio n .
incautiously weakened with his 1 2th move.
19 . . . li'le8
15 . . . li'ld5
20 l:tcf1 li'ld6
16 'ii'd 2 li'l7f6
21 'iit h 1 li'lf5
1 7 li'lfe5 .tb7
2 2 ..ixf5?
(see diagram)
The decisive strategic mistake. Now there is
1 8 f4? nothing with which to oppose the bishop on
Another positional mistake , which puts White b7.
in an extremely difficult position. What are 22 . . . exf5
the defects of the pawn advance? 23 .l:tc1
The h 1 -a8 diagonal is weakened . With the What would you have now played for Black?
How to draw up a Plan l2J 91
Let's first see what happened in the game, with an u n usual material balance. That is
and then return again to this position . also the case here: 29 tt'lxb4! .ll x e 1 + 30
23 . . . tt'lf6 .llx e 1 deserved very serious consideratio n .
Of course, a very natural move - the e4- F o r t h e q ueen Wh ite has rook, knight a n d
square itself is as though inviti ng the knight pawn - al most a sufficient equivalent. I f
to go there. 30 . . . i.xb4 3 1 �xb4 it'xd4, t h e n 3 2 i.d6 ! . I
would like to play �e5 and then exploit the d
24 'i*'e3 tt'le4
file for an attack (�d2, .Ued 1 ). Of cou rse,
25 .Ufc2 'it'd5 White has to reckon with the cou nter-stroke
26 tt'ld3 l::!. fe8 32 . . . �xc4 , but it leads only to a draw: 33
27 �e1 �e8+ 'it>g7 34 �f8+ 'it>f6 35 �e7+ 'it>g7 (if
Black seems to have played well , but his 35 . . . 'it>e6?, then 36 i.c5+! is strong) 36
actions have not been systematic. Whereas �f8+.
White, who j ust now was losing, has sudden [I do not agree with this evaluation. After 32
ly gained cou nter-chances. The b4-pawn is i.d6 f6! Black does not allow the bishop to
attacked , and 27 . . . a5 28 tt'lb6 is bad for go to e5. The e-file can always be blocked by
Black. . . . i.e4, and switching to the d-file requires
27 . . . tt'lc3 too much time. Meanwhile, the b3- and g2-
points are vulnerable. I do not see what there
Practically forced .
is to prevent Black from converting his
28 i.xc3 i.f8 material advantage. - John N u n n . ]
Black prepares to play his bishop to g7. I n t h e g a m e Wh ite missed his chance a n d
lost without a fig ht.
29 tt'lde5? bxc3
30 'ilt'xc3 i.h6
Despite Wh ite's extra pawn , his position is
d ifficult. Black has two powerfu l bishops, and
his pieces control the entire board .
31 tt'ld3
If 31 .l:!.f1 , then 31 . . . i.xf4! 32 .Uxf4 .l:!.xe5! . [32
tt'lxg6! is White's best chance in this variation
- N u n n .]
31 . . . .Ued8
32 tt'lc5?! �a8
33 b4 �xf4
A very interesting moment. Here it is 34 .l:!.d1 .l:!.e8
appropriate to remember a procedure which 35 'ir'd3 .l:!.cd8
is constantly employed by Mark Dvoretsky. 36 l:!.e2 .l:!.xe2
In unfavourable situations he recommends
37 'it'xe2 'it> g 7
looking for a way of rad ically changing the
character of the struggle. Sometimes it is Black prepa res 38 . . . l:Ie8.
possible to complicate the play by means of 38 tt'lb2? Ite8
a positional sacrifice (for example, of a pawn 39 'ir'f1 .tc7
or the exchange) or by going i nto a position 40 b5
92 � How to draw up a Plan
This loses immed iately. 40 tZ'lc4 was more In positions with strategic manoeuvring
tenacious. (when the time factor is not of decisive
40 . . . 'ii'd 6 importance) look for the piece which is
worse placed than all the others. The
41 11'g1 .i::t e2
activation of this piece is often the most
42 tZ'lc4 'i¥d5
reliable way of improving your position
White resig ned . as a whole.
25 'iff5
What would you play now?
17 .l:.xh 6 ! ? gxh6
1 8 lDfxd5!
If 1 8 lDh5 there was the good reply 18 ... 'ili'h4 .
Let us employ the principle of the worst
18 . . . exd5 piece . It is q u ite obvious that Black's worst
1 9 'ifxh6 piece is his king . If it were to be removed
Wh ite's attacking looks menacing, but don't from the centre and placed at b8, the
forget that for the sake of it he has sacrificed opponent would have to resig n . Therefore
a whole rook. In such cases the opponent 25 . . . 'it>c8! followed by 26 . . . 'it>b8 suggests
usually has an opportunity to buy his itself.
opponent off, by g iving u p part of his extra National master Telman (the trainer of
material . Seryozha Ovseevich , who was playing in the
19 . . . lDxd4! same tou rnament) suggested another, tacti
cal solution : 25 . . . l:th4 26 f3 d4! 27 .ixd4
20 'ii'x h8+
�xf3 , achieving simpl ification adva ntageous
Nothing was given by 20 lbb5 lbxb5 2 1 to Black. Also not bad , although the march of
�xb5+ c6 2 2 �xc6+ i.xc6 2 3 'ii' x c6+ 'it>f8 24 the king away from the centre appeals to me
'ii'h 6+ 'iti>g8 25 �g5 'ii'f8 . more.
[Instead of 22 �xc6+ ? White maintains the It was a pity that, as a consequence of
balance by 22 �g5!, for example: 22. . 'ii'xg5+.
tiredness and approach ing time-trouble (of
23 'ili'xg5 cxb5 24 'ili'f6! 'it'd7 25 "iVd6+. The course, this is not a j ustification , but merely
same move 20 �g5! was also not at all bad an explanation ) , Vova did not fi nd the correct
immediately, instead of capturing the knight path and m issed an almost certain win .
- Dvoretsky.]
25 . . . 'it>e7?
20 . . . 'it>d7
26 lDe2 �c8?
21 'ii' h 5 'ili'g8
A clear waste of time - the bishop stood
22 �e3 lbe6 better at b 7 .
It is apparent that the worst for Black is over, 27 lDg3 'i!kg6
and the attack should be parried . 28 'ii'f3 �b7?
23 i.f5 'ii'g 7 [ The logical continuation of Black's preced
24 �xg4 .l:th8 ing moves would have been 28 . lbg5! 29 . .
How to draw up a Plan L'iJ 95
lf6+ (29 'ikf4 �xg4 30 'ikxg4 l'De4) 29 . . . 'Wixf6 A rather passive pla n . Theory recommends
30 exf6+ �xf6 3 1 ii.d4+ cJ;g6 32 �xhB �xg4 11 . . 'ika5 .
.
25 'ii'e 3 'ii'f6
26 tt:lb6 �c7
It's done - Wh ite has switched his q ueen and
knight to more active positions. Now he
clarifies the situation i n the centre (if it were
the opponent to move, he would happily play
27 . . . ..tf4).
27 ..txe5 tt:lxe5
If 27 . . . 'i!Vxe5, then 28 c5 with the th reat of 29
tt:lc4 . And if 27 . . . dxe5 there fol lows 28 �xdB
tt:lxd8 29 tt:ld7 and 30 tt:lxe5 .
rather pitiful role for Wh ite's strongest piece. advantageous to the opponent, since the
Where would he like to play it to? Of cou rse, knight on b6 would become vulnerable. ' -
32 bxa5 .l:.c5
38 . . . 'iid 6
33 llb1
39 'ili'c3
Of cou rse, not 33 l:txd6? tt:Jxd6 34 'ili'xc5
39 'ili'e3! ? .
4Jxe4.
39 . . . l:txb6
33 . . . lba5
40 llxb6 'ili'xb6
34 llb5 .l:ta6
41 'iix e5+ f6
In the event of 34 . . . l:txb5 35 cxb5! Wh ite
acquires a menacing passed pawn on the a 42 'ii h 5
file. Black resig ned .
White resigned .
Thus a zugzwang position is one of the
situations where the time factor tells, and i n a
highly d istinctive form .
Petrosian - Sueti n
27th USSR Championsh i p , Len ingrad 1 960
The danger of Black's position begins to be
Queen 's Gambit
felt, and he must be extremely carefu l . His
1 c4 c5
king is still i n the centre and the opponent is
2 lLlf3 lZ:lf6
already taking the in itiative and creating
3 lLlc3 lZ:lc6 concrete threats. He should complete his
4 e3 e6 development as q u ickly as possible by
5 d4 d5 playing 1 1 . . . .tb4 (with gain of tempo ! ) and at
1 00 � Sensing the Tempo
the fi rst convenient opportun ity - castle. But Soloviov - Kosi kov
that which I cal l 'sensing the tempo' appar Smolensk 1 99 1
ently betrayed Alexey Sueti n . French Defence
11 . . . b5? 1 e4 e6
1 2 bxc4 bxc4
2 d4 d5
1 3 e4!
3 e5 c5
If Wh ite wants to pun ish the opponent for
4 c3 tt:'lc6
neglecting his development, he must delay
it, come into direct contact with h i m , and 5 tt:'lf3 'ili'b6
sharpen the play. 6 a3 �d7
13 . . . dxe4 7 �e2
The lesser evil was 1 3 . . . tt:Jxe4 (exchanges The combination of 6 a3 and 7 �e2 is hardly
usually favour the defending side), but even good ; the latest word i n fashion in this
here after 14 tt:Jxe4 dxe4 1 5 d5 'ii'g 6 1 6 Ji.h5! variation is 7 b4 . However, when playing
(a recommendation by Yuri Razuvaev) Wh ite one can take some liberties without
16 . . . 'ii'f5 1 7 lle 1 Wh ite has the advantage. being pun ished .
1 4 .ll g 5 �f5 7 . . . tt:'lh6
The middlegame is in full swing, but Black is 8 b4 cxd4
still not in a hu rry to develop his kingside. 9 cxd4 tt:'lf5
Good or bad , he had to play 14 . . . �e7. 1 0 �b2 �e7
15 d5 "ilc7 1 1 0-0 0-0
1 6 �xf6 gxf6 White has problems with the completion of
1 7 �g4! his development - he can not bring out his
The last accu rate move , exchanging the knight i n view of the loss of the d4-pawn . The
opponent's only developed piece - the move b4-b5 is anti-positional , since it seri
bishop which is defending the e4-pawn (if ously weakens the q ueenside.
1 7 . . . �g6 Wh ite wins by 1 8 tt:Jxe4 �xe4 1 9 12 'ii'd 2?
.l:!.e 1 'ii'e 5 20 'il'a4+ with mate). The outcome
1 2 Ji.d3 and then �xf5 was necessary, with
is now not in doubt.
roughly equal chances.
17 . . . �xg4
What would you now have played as Black?
1 8 'ifxg4 'iie 5
1 9 tt:Jxe4 f5
20 'ii' h 5 0-0-0
21 tt:'ld2 c3
22 tt:'lc4 'il'd4
23 'ii'xf5+ .Ud7
24 tt:Je5
Black resig ned .
excellently placed .
Another suggestion is 1 2 . . . g5. You know, I
like playing . . . g7--g5 in the French Defence,
but here this is really too sharp !
There is a nother possibil ity: 1 2 . . .f6 . B u t , i n
suggesti ng i t , did you m i s s 1 3 g4 ltJ h 6 1 4 The i n itiative is on my side, but the position is
exf6 followed by the fork g4--g 5 ? You missed one where every tempo is important. If Wh ite
it? But even so, that's what I played . should succeed in playing ltJd2-f3--e5 or
12 . . . f6! ltJd2-b3-c5, I will have to forget about any
The tactical idea associated with this move is hopes of an advantage. Sensing the tempo
wel l known - it is analysed i n Aaron suggests that Black should urgently 'latch on'
Nimzowitsch's famous book My System. to the opponent, using concrete threats to
divert him from the completion of his devel
1 3 g4 ltJh6
opment.
14 exf6 .l:.xf6
22 . . . a5!
1 5 g5 .:txf3
23 bxa5 l:lxa5
16 ..txf3
But not 23 . . . ltJxa5 because of 24 ltJc3. The
Black also has the advantage after 1 6 gxh6. black rook comes i nto play, and Wh ite still
16 . . . ltJf5 cannot move his knight.
1 7 l:id1 24 a4
The positional exchange sacrifice has g iven Here I stopped to th i n k and I found what I
Black a splendid position . It is amusing that believe is a good solutio n . I was helped by
at this point my opponent offered a d raw. the ' principle of the worst piece'. The bishop
17 . . . on d7 is not taking any part in the play. The
'ii'd 8!
standard route for the bishop is via e8 to h5,
The g5-pawn is lost. The next few moves are
but from there it will be fi ring i nto empty
forced .
space. It is most probably better to attack the
1 8 ..tg4 ..txg5 a4-pawn with it, i . e . move the knig ht.
1 9 f4 ..tf6 24 . . . ltJd8!
20 ..txf5 exf5 The knight goes either to e6, or via f7 to d6. If
21 'ilr'f2 Wb6 Black should captu re the a4-pawn , he will
Wh ite wanted to develop his knight, but I already have two pawns for the exchange.
prevent this by tying h i m to the defence of the 25 ltJc3 ltJe6
d4-pawn . The d4-pawn is again under fi re.
22 l:td3 26 ltJd1
1 02 � Sensing the Tempo
Here the knight is not much better placed for decisive action - it was not tempo
than on b 1 . Now I could have simply dependent. For a time Black needs to
captured with my rook on a4 with an maintain it, to manoeuvre. Such manoeu
advantage, but in such cases it is important vring, on the one hand, enables the position
not to sell yourself too cheaply. You should to be strengthened to the maxi m u m , by
check whether or not there is someth ing making all moves that will be usefu l in the
stronger. And indeed , if you see Black's next futu re, and on the other hand, it allows the
two moves , it immediately becomes clear most appropriate moment to be chosen for
that this is what should be played . switching to positive action , when the oppo
26 . . . 'it'd& nent goes wrong and makes the task easier.
The f4-pawn is attacked . 30 . . . h6!
27 �c1 b5! Now the king will feel more secu re, and
. . . g7-g5 is also a possibil ity.
The pawn is won in a version that is more
advantageous to Black. His passed pawn will 31 .:d2 'i!i'b4
be much more dangerous on the a-file than 32 l2'la2 'ii'd 6
on the b-file. 33 lt'lc3 'it>h8
28 �e3 bxa4 As you see, in non-tempo positions the
29 l2'lc3 principle 'do not hu rry' comes to the fore. I
A new question : how would you characterise very much did not want to play . . . g7-g5 (the
the resulting situation (from the standpoint of opponent is condemned to passivity, so why
our topic - 'sensing the tempo') and what sharpen the play? ), and yet now Wh ite has to
wou ld you suggest playing? reckon with this move. You see, after 34 . . . g5
35 fxg5 hxg5 36 "ii'xf5 I have the reply
36 . . . lt'lxd4, which would not have been the
case with the king on g8. And in general, the
king stands slightly better at h8 than at g8,
even if only marginal ly.
Note that my last few actions ( . . . h7-h6,
. . . 'i!i'b4 , . . . 'it>h8) have not involved the
slightest risk. But the opponent has to be
constantly on the alert, since any move of his
may tu rn out to be a serious mistake. It is
very d ifficult to defend in such situations.
34 .Ub2 'ii'c 7
Advancing the pawn to a3 is prematu re - this
move should be held over Wh ite l i ke a sword
of Damocles and made only with decisive
I thought that I had already gai ned a decisive
effect. But for the moment Black should
advantage (two pawns for the exchange,
prolong the manoeuvring, trying with minor
dangerous passed a-pawn , weaknesses on
th reats to d isrupt the coordination of the
d4 and f4) and I spent some time looking for
wh ite pieces.
concrete ways to break through the oppo
nent's defences. But gradually I realised that 35 'iid 2 'iia 7
the defensive resou rces were qu ite consid 36 .l:td1
erable, and that the position was not yet ripe 36 lt'le2 is bad in view of 36 . . . i.b5, when the
Sensing the Tempo l2J 1 03
bishop comes very strongly i nto play. But I n the previous game no particular imagina
after the rook has moved , the advance of the tion was demanded of Black - self-control
a-pawn will crack the opponent's defences. and patience were more necessary. But for
We see the principle of two weaknesses in seizing and retaining the in itiative, accurate
action - Wh ite is u nable simultaneously to and resou rceful play is requ i red , and the
defend the d4-pawn and to combat the value of each move is usually extremely
passed a-pawn . high.
36 . . . a3
37 l:ta2 �a4
After both a move of the roo k and the Roma n i s h i n - Farago
excha nge o n a4, the d4-pawn is lost. The E u ropean Tea m C h a mpionsh i p ,
game is practically decided . Skara 1 980
In such situations the opponent usually 'goes R e ti Opening
berserk' , trying at any cost to create some 1 liJf3 ltJf6
counter-chances, and here one should be 2 g3 d5
especially carefu l . 3 i.g2 c6
3 8 l:lxa3 i.xd 1 4 0-0 �g4
39 ltJb5 5 c4 e6
A clever reply. But since Black's previous 6 d4 liJbd7
strategy was correct, the tactical compl ica
7 ltJe5 i.. f5
tions should tu rn out to his advantage.
8 ltJc3 i.. d 6
39 . . . l:txa3
9 �f4 'iWb8
40 ltJxa7 l:txe3
Black is slig htly slow in castling. Oleg
41 ltJc6
Romanishin tries to exploit this factor by
41 'i'xd 1 �xd4 was totally bad . sharpening the play.
41 . . . Iie4 1 0 ltJxd7 ltJxd7
42 'i¥xd 1 ltJxf4 11 cxd5 exd5
Of cou rse, 42 . . . ltJxd4 would also have won , 1 1 . . . i.xf4? 1 2 dxc6 .
but why exchange the wh ite knight when i t is
1 2 e4! dxe4
shut out of the game?
1 3 i.. x e4 i.. xe4
43 'iVa4
1 4 �xd6 "ii'x d6
43 'iff3 .l:!.e 1 + 44 'it>f2 �h4+ .
1 5 ltJxe4 'ii'g 6
43 . . . �h4
Here is the position , for the sake of which we
White resigned .
are a nalysing this game. What would you
suggest?
We have seen what sort of approach a player
should adopt i n non-tempo positions (the
principle 'do not hu rry' etc . ) . The fol lowi ng
examples will be devoted to the problem of
the initiative . (see diagram)
1 04 � Sensing the Tempo
1 0 lDc2 .if5
Although Black is a pawn down , for the
moment he can play without particu lar
thought - so natu ral is the development of
his pieces . He is not th inking about regaining
the material , but is planning . . . lbbd 7 , . . . e5-
e4 and . . . lbe5 . From e5 the knight will exert
pressure on both wings.
11 �d3
My opponent was apparently afraid of me
regaining the pawn by 1 1 . . . .ixc2 and
12 . . . lbxd5, which did not come into my plans
at all. However, after 1 1 . . . lbbd7 and 1 2 . . . .Uc8
the threat of the exchange on c2 would If the opponent should succeed in removing
indeed have become real . his king from the centre, he will obtain an
11 . . . e4 acceptable position . Moreover, his slight
delay in castling will then acq u i re a logical
1 2 �e2 �g6
basis. In fact, if Wh ite had castled a few
This is not yet a tempo-position , and so I moves earlier, then by placing my knight on
decided to spend time on a prophylactic e5 I would have beg u n an attack on the
move. After the immed iate 1 2 . . . lbbd7 1 would kingside. But now, when the open ing of the
have had to reckon with 1 3 g4! ? . play on another part of the board has
1 3 b4? ! diverted my pieces, the king will feel com
The start of an over-sharp pla n , which in the pletely safe on the kingside.
end led to my opponent's defeat. Thus, I am obl iged to undertake something.
13 . . . .id6 1 8 . . . lDd3+ suggests itself, but how should
one assess the position arising after 1 9
1 4 a4
�xd3 exd3 2 0 lbd4 axb5 2 1 .l::tx a 8 'ii'xa8 22
A continuation of the same faulty strategy. lbdxb5 (or 22 lbcxb5)? Of cou rse , Black
With the wh ite king still stuck in the centre, retains the i n itiative , and this cannot be bad
the opening of lines on the queenside is to for him. But nevertheless it is pity that at the
Black's advantage. Wh ite would have done end of the variation the opponent has a
better to 'let sleeping dogs lie'. choice - he can captu re on b5 with either
14 . . . bxa4 knight, 22 lbdxb5 apparently being the stron
1 5 l:.xa4 'ii'c8 ger option. And I found a transposition of
moves, which denied Wh ite this possibil ity.
I did not like 1 5 . . . lDbd7 because of the
manoeuvre lDc2-d4-c6 . 18 . . . axb5
1 6 .ib2 lbbd7 1 9 llxa8 'ii'x a8
If now 1 7 lDd4 , then 1 7 . . . lDb6 followed by 20 lbxb5
1 8 . . . lbbxd5, 1 8 . . . lbc4 or 1 8 . . . �xb4. [20 0-0 was better - Dvoretsky.]
1 7 b5 lDc5 20 . . . lbd3+
1 8 lta2 21 �xd3 exd3
How should I now contin ue? 22 lbcd4
Sensing the Tempo lZJ 1 07
7 ..tf1 ! was stronger, in order to answer arranged his pawns on light sq uares, to give
7 . . . dxe4 with 8 'iig 4. scope to the dark-sq uare bishop, wh ich has
7. . . dxe4 no opponent. But the opposite has happened
8 ..txc4 "ilc7 - the pawns are fixed on dark squares and
the bishop at b2 has been transformed i nto a
9 "ile2 tLld7
'large pawn ' .
1 0 ..tb2 tLlgf6
14 . . . tLld5
11 f3 0-0
1 5 tLlf3
1 2 ..tb3
1 5 1i'xb5? ttJxc3 (the g2-pawn is under
Give some thought to the situation that has
attack) 1 6 'iff1 'ii'a 5 is clearly bad for White.
arisen.
15 . . . ttJxc3
Black has restored material equality and
seized the i nitiative. This happened because
Ragozin correctly recognised the moment
when it was necessary to delve deeply into
the position and find the correct solution
( 1 2 . . . b5! ) .
1 6 "ild3
31 h4 tt:'Jxf3+ 1 1 �b3 d6
32 l:l.xf3 �c1 1 2 l:l.fd 1 �e7
33 .Uf1 �e3+ 1 3 .l:i.ac1 l:l.ac8
34 .l:i.f2 b4! 14 e3 eS
35 a4 .l:i.d8 1 5 lL'le1
36 'it'b1 .l:i.xd4 A thematic, but com mitting move . White
Wh ite resig ned . wants to attack the b4-pawn with his knight,
in order to resolve the situation on the
q ueenside. But as a result of the exchange of
The last game that we will look at today is
bishops, the wh ite king's position is weak·
perhaps the most important one for our topic.
ened somewhat. However, there are hardly
Its leitmotif is 'change of rhyth m ' .
any black pieces there, so that this factor
Many players remark that i t is most d ifficult
seems purely academic.
for them to fi nd their beari ngs when there is a
sharp change of scene: the transition from 15 . . . �xg2
attack to defence or from defence to attack, 1 6 Wxg2 �b7+
from a position played in accordance with the 1 7 �g1 e4
principle 'do not hurry' , to tempo play and 1 8 lbc2
vice versa , and so o n . Even for lead ing
grandmasters this is sometimes a serious
problem . To make progress i n this field you
must consciously develop and train you r
'sensing the tempo ' .
Tu kmakov - Vitolins
Yerevan 1 980
Bogo-lndian Defence
1 d4 lL'lf6
2 c4 e6
3 lL'lf3 J.b4+
4 �d2 cS
S �xb4 cxb4 Let us dwell on this position a l ittle. Vladimir
Tukmakov has carried out his pla n , and after
6 g3 b6
1 8 . . . bxa3 1 9 bxa3 followed by .l:i. b 1 he will be
7 �g2 �b7 the fi rst to 'latch on' to the opponent, by
8 0-0 exploiting the defects in his pawn structure.
A slight inaccuracy! If Wh ite had played 8 a3 Alvis Vitolins was an u nusual player, excep·
immed iately, the opponent would not have tionally inventive and always seeking the
been able to obtain the position that occurred in itiative. He had a subtle feeling fo r the
in the game. rhythm of a game, and he knew how to
8. . . aS change it. Here too he emerged with honour
9 a3 tt:'Ja6 from a d ifficult position .
10 lL'lbd2 0-0 18 . . . 'it'd7 !
Sensing the Tempo CtJ 111
By giving u p a pawn , Black essentially burns i n the event of 23 . . . exf3 24 tt::l xf3 bxa5 White
his boats behind him. If he does not g ive remains with an extra pawn , for which the
mate, he will of cou rse lose. Yet for mate opponent has insufficient compensation .
there would appear to be i nsufficient force. 2 2 d5?!
After . . . ii'h3 and . . . tt::l g 4 there follows tt::lf 1 ,
I n mechan ically preventi ng the manoeuvre
and what then?
. . . tt::l c7-e6--g5 , White weakens the i mportant
19 axb4 e5-sq uare.
It is not possible to take away the h3-sq uare 22 . . . tt::l e 8!
from the q ueen - if 1 9 'it>g2 there fol lows
The knight changes cou rse and aims for g4
1 9 a4 .
(after the other knight, by moving to e5, has
. . .
I think that in this game Black attacked really much stronger 1 8 . b5! 1 9 cxb5 1kxb5 20
. .
too recklessly, too riskily. At some point he axb4 axb4. Black stands better - all the
should have captured the a5-pawn (for opponent's pieces are cramped, and in
example, on the 22nd move). And instead of addition he constantly has to reckon with the
the clever, but objectively not fully correct switching of the queen to the kingside by
queen manoeuvre ( 1 8 'iid 7?!) he had the
. . . .. ."ilb5-h5 - Dvoretsky.]
l2J 113
Mark Dvoretsky
�e3 44 l:txe3 l:txe3 45 l:txh6+ �xh6 46 u n l i kely that my opponent would play this
�xe3 'itg5 4 7 �e4 �g4 or 43 l:r.f3 i.xe3+ 44 particular system , and so I did not bother to
lkxe3 'iil h 6! with a d raw. check the book variations, but simply accept
43 . . . �h8! ed them.
43 . . . .l:.8e7 was worse : 44 l::t x e7+ :Xe7 45 6 l:!.c1 !
l:tf3. And i m mediately I ra n into a novelty, found
44 .l:!.h4 .l:!.xf5+ by Gulko at the board . The opening g u ide
Draw. only considered strange variations such as 6
'ii'c2? ! lL'lc6 7 e3 i.f5! or 6 1i'b3 1i'xd4. The
rook move to c1 is logical - Wh ite develops a
In all the examples we have exam ined the piece, defends his knight in advance in the
question to be decided was whether to event of the capture of the b2-pawn , and,
change sharply the character of the play, or i ncidentally, prevents the immediate 6 . . .
maintain the existing situation. But some 'ii'x b2? because of 7 tt:'Ja4 'ii' b 4+ 8 .id2 .
times it is possible for a player to transform a
6 . . . lL'lc6
position in several ways . Grandmaster Boris
Gulko once said to me that he considered 7 e3 iVxb2?!
such problems to be the most d ifficult i n Consistent: if Black doesn't take the pawn , it
chess, making t h e most severe demands on is not clear why his q ueen was developed at
a player's mastery, his calculating tech nique b6. Even so, 7 . . . i.f5 or 7 . . . i.g4 8 f3 i.f5
and his depth of positional evaluation . would have been more cautious.
I will show how Gulko h i mself copes with 8 i.d3 i.g4
such problems. In the followi ng game I There is no longer time for qu iet develop
remember how one of his decisions made a ment: 8 . . . e6? 9 lL'lb5 i.b4+ 1 0 �f1 0-0 1 1
strong impression on me. .l:!.c2 .
9 ltJge2 .ixe2
G u l ko - Dvoretsky 1 0 i.xe2 ! e5
Vi lnius 1 978 The transformation of the position caused by
Slav Defence this move proves clearly advantageous to
Wh ite. 1 O . . . e6 ! ? came into consideration , for
1 c4 c6
example: 1 1 lL'lb5 'iib 4+ 1 2 �f1 .l::t c8! 1 3
2 tt:'Jc3 d5
tt:'Jc7+ ( 1 3 .l:t b 1 'ifa5 1 4 ltJc7+ .l:!.xc7 1 5 .l:!.b5
3 cxd5 cxd5 iVxa2 1 6 i.xc7 tt:'Je4) 1 3 . . . �d8 14 .l:!.b1 'it'e7
4 d4 tt:'Jf6 1 5 .l:!.xb7 g 5 ! , or 1 1 0-0 i.e? 1 2 lL'lb5 0-0 1 3
5 .if4 'ii' b 6 a4 (th reatening 1 4 .l:tb1 'ii'a 2 1 5 lL'lc3 'ii'a 3 1 6
In my preparations for the game I glanced i n l:tb3) 1 3 . . .'it'b4 ( 1 3 . . . i.b4 ! ? ) 1 4 i.e??! ( 1 4
the Encyclopaedia o f Chess Openings ( i n ltJc7 followed b y l:.b 1 i s stronger) 1 4 . . . ltJe4
the first edition t h e corresponding section 1 5 l:t b 1 'ii'd 2! (but not 1 5 . . . tt:'Jc3? 1 6 tt:'Jxc3
was written by g randmaster Alexey S ueti n ) 1!i'xc3 1 7 .l::t b 3). However, in these variations
and saw there a recommendation that after Black's position looks uneasy, and I recom
the move order chosen by Wh ite, Black mend the readers to look for an improvement
should respond with 5 . . . 'ii' b 6. Generally i n Wh ite's play - I would not be at all
speaking, it is dangerous to trust S ueti n's su rprised if one should be found .
assessments - too much hack-work was 1 1 dxe5 i.b4
produced by his pen . But I considered it 1 2 0-0 ! i.xc3
1 20 � Transformation of a Position
21 .1l.. x e5!
G u l ko - Shcherbakov
Of course, not 21 .l:i.e 1 ? ! .1l.. xf4!? (2 1 .. .f6 is
Helsinki 1 992
also possible) 22 .l:i.xe6 .1l.. x c7 , and the
position becomes u nclear. Slav Defence
ltlh5.
1 6 t"Llb1 'i¥a4
Now Wh ite restores material eq ual ity and
transposes i nto a favourable end ing. G u l ko
considered the strongest reply to be 1 6 . . . 'ifa5
and after 1 7 t"Llxe7 <it>xe 7 he was intending to
play 18 i.f4 h6! 1 9 t"Lld2 <it>f8 20 t"Llf3 �g7 2 1
ltJe5. Wh ite certainly reta ins excellent com
pensation for the pawn , but the outcome still
remains u nclear - this is why Gulko had
doubts about his choice on the 1 4th move .
Later grandmaster Viorel Bologan suggest 27 f5!
ed strengthening the attack by 1 8 'ii'f3! , for The attack on the king is mai ntained even in
example: 1 8 . . . .l:r.he8 (in the hope after 1 9 the endgame. A sample variation goes
l:tfe 1 ?! of ru n n i ng away with the king by 27 . . . i.gS 28 f6+ <it>h6 29 l:.c3 i.d2 30 .l:r.h3+
1 9 . . �f8 ! ) 1 9 'ii'f4 ! , or 1 8 . . . t"Llxe5 1 9 dxeS
.
G u l ko - Kupreich i k
5 2 n d USSR Championsh i p , R i g a 1 98 5
King 's Indian Defence
1 d4 lt:Jf6
2 c4 g6
3 lL'lc3 ..tg7
4 e4 d6
5 f3 a6
6 ..ie3 0-0
- position after 1 3 . . . e5 -
7 'ii'd 2 lt:Jc6
8 lt:Jge2 l:.b8
d6-pawn are attacked . But take note: the
9 l:tb1 position has been opened u p , and White is
A rare pla n . Theory considers the strongest behind in development. In such positions
reply to be 9 . . . b5 1 0 cxb5 axb5 1 1 b4 e5! 1 2 one must be extremely cautious, especially
d5 lt:Je7. when sitting opposite you is such a resource
9 0 . . ..id7 ful tactician as Kupreich ik. He will most
10 b4 'ii'c 8 probably play 1 S . . . l:te8 ! , i ntending to meet 1 6
If now 1 O . . . b5 1 1 cxb5 axb5 1 2 dS lt:JeS 1 3 ..txd6 o r 1 6 'ii'x aS with 1 6 . . . lt:Jxe4 ! . It is
lt:Jd4 with the better chances for White . possible to defer winning material, by rein
forcing the e4-point with 1 6 lt:Jf2 . The
1 1 b5 lt:Jas
position after 1 6 . . . lt:Jxc4 1 7 ..txc4 is advanta
1 2 lt:Jf4! c6 ! ? geous to Wh ite. But the opponent finds a
After 1 2 . . . b 6 1 3 eS!? or 1 3 l:.c1 ! ? c6 1 4 bxc6 bri l l ia nt cou nterattack: 1 6 . . . d S ! ! 1 7 'i'xa5
Black would have stood worse. Therefore dxe4 , or 1 7 ..txb8 'ii'x b8 ( 1 7 . . . dxe4 1 8 i.e5
Viktor Kupreichik i nitiates risky play - he exf3 can also be considered ) 1 8 'i!i'xaS dxe4
abandons to its fate his knight on aS, which with dangerous threats.
now has no retreat square. As we will see , Let us now tu rn to the pawn exchange 1 4
this idea has a clever tactical basis and it is dxeS dxeS. Perhaps here White should
not at all easy to refute . retreat his knight to d 1 ? Let us check: 1 5
1 3 b6! lt:Jd 1 exf4 1 6 ..ixf4 l:te8 1 7 lL'lf2 ( 1 7 'ii'xa5
Stronger than 1 3 eS?! lt:Je8 1 4 b6 ..ie6 ! . lt:Jxe4 ! ) 1 7 . . . lt:Jxc4 1 8 ..ixc4 lla8 1 9 0-0 i.e6.
13 0 . . e5 The next move will be 20 . . . lt:Jd7 , and White
has nothing.
(see diagram)
Let us check 1 S lt:Ja4 (instead of 1 S ttJd 1 )
What position should Wh ite go in for? It is 1 S . . . exf4 1 6 ..txf4 . After 1 6 . . . l:e8?! 1 7 ttJc5
clear that he must attack the knight on aS, the wh ite knight is far more actively placed at
but in retu rn he will evidently have to part cS than at f2 . But on the other hand, the
with his knight on f4 - he does not want to opponent acq u i res an excellent tactical
retreat it to h3! 1 4 lt:Jd 1 exf4 1 S ..txf4 resou rce: 1 6 . . . cS! 1 7 'ifxaS ..txa4 1 8 ..txb8
suggests itself, since both the knight and the ( 1 8 'iix a4 lt:Jxe4 ! ! 1 9 fxe4 'ili'g4) 1 8 . . . lt:Jxe4!?
Transformation of a Position ttJ 1 25
Exercises
Sol utions
29 i.e3 .l:tb5 30 'ifxe6+ 'it>h8 31 'ifxa6 it dawned on h i m - he saw a way of retu rning
It is now Wh ite who is a pawn up, and in the the extra material and forcibly transposing
subsequent play he successfully converted i nto an ending with a g reat positional
it. advantage.
44 'ii' g 2! 'ii'c 7 !
4. Miles-Romanishin (Tilburg 1 985). 45 'i!i'g3 ! ! 'ii'c 1 +
White's best saving chance is to transpose 45 . . . 1!i'xg3 46 fxg3 i.b7 47 l:ta4 is hopeless
into a heavy piece ending. for Black.
36 ltJxe4! 'ifxe4 46 lle1 'ii'x e1 +
37 i.xg7 'it>xg7 47 'it>xe1 i.xg3
38 'iff6+ 'it?g8(h7) 48 fxg3
39 h3
White's queen is well placed on f6 - it
defends the b2-pawn and ties the rook to the
defence of the f7-pawn , making it hard for
Black to create an attack on the king . And in
the event of the queens being exchanged ,
the famous formula may apply: ' Rook end
ings are always drawn . '
The tempting 3 6 'ii'f4? chosen b y Tony Miles
was weaker because of the reply 36 . . .'it'd6! .
Now 37 ltJxe4? 'iVxf4 loses immediately, and
37 'ii'xe4?! i.xf6 is also bad for Wh ite, since
after 38 'ife8+ 'it>g7 39 ltJe4 Black has the
decisive 39 . . . 'ii'd 3! 40 ltJg3 h4. There only
remains 37 'ifxd6 ltJxd6 38 i.xg7 'it>xg7 , but 48 . . . i.b7
this endgame is much more difficult than the lf 48 . . . i.d7, then 49 l:ta7 is strong . This move
one with heavy pieces, si nce Black can was also not bad now, but Psakhis decided
improve without hindrance the placing of his to exchange the rooks, since he correctly
pieces (in particu lar, h is king ). judged the bishop ending to be won .
I n the game there followed 39 .l:.d1 l:!.e7 40 4 9 l:txg8+ 'it>xg8 5 0 a 4 'it>f7 51 a 5 bxa5 52
'iitg 1 (40 g4!? ltJc4 41 gxh5 g5!?) 40 . . . lLlf5 41 bxa5 'it>e7 53 'it>d2 'it>d6 54 'it'c3 'iitc6 55
.l:ic1 'it>f6 42 b4 .Ue5 43 'it>f2 'it>e6 44 g3 ltJd4, 'iti>b4 i.c8 56 'it>a4! (zugzwang) 56 . . . i.b7 57
and Black won . a6 Black resigned .
PART I l l
Typical Position s
Mark Dvoretsky
assessments interested me, although by no tive. With opposite- co/our bishops, the
means all the examples seemed convi ncing . possession of the initiative is a serious
and in particular I used the games of Simagin I n my book School of Chess Excellence 1 :
himself, who played skilfu l ly with opposite Endgame Analysis (p.64) I g ive an ending
colour bishops. As a result I was able to gain from the 4th game of the Alexand ria
an u nderstanding of the given problem. Liti nskaya Candidates match . There , apart
Opposite-Colour Bishops in the M idd legame lZJ 1 31
whether the attack will be successfu l . But i n the i m mediate th reats. The knight on f4 is
principle this is sound strategy, a n d this i s dangerous, of course, but for the moment the
how o n e should act when there are opposite other pieces are not able to help it: the queen
colour bishops. can not go to g6, and the bishop is ru nning up
23 'Yid2 against the e4-pawn . Has Black's strategy
White would have lost after 23 .l:ixa5? exd4 , proved to be a fiasco?
but it would have been more accu rate to play With opposite-colour bishops you must be
the queen to e3 (or d1 ). l ieve in the attacking possibil ities of a
position! Black's pressure on the kingside is,
23 . . . e4
after a l l , more important than the opponent's
Apart from moves that are part of a pla n , you
material gains. I n order to i ncrease it, he
should always be on the lookout for chance
needs to drive the queen from the blockading
tactical resou rces such as 23 . . . ltJe4 ! ? . After
e3-square, and for this the pawns have to be
24 'ife3 ttJxc3 25 'ifxc3 'it'g6 Black's advan
i ncluded in the attack.
tage is obvious.
27 . . . f5!
24 h3
With the impending th reat of 28 . . .<�'Jxg2! 29
White has to prevent 24 . . . ltJg4 . If 24 'iVg5
Wxg2 f4. A 'Tal-like' piece sacrifice also came
Simagin was intending 24 . . . e3! 25 fxe3 h6 26
i nto consideration : 27 . . . ltJxg2!? 28 'it>xg2 f5
'ifg3 'Yixg3 27 hxg3 .l:ixe3 or 26 'Yif4 li'xf4 27
with dangerous threats .
exf4 lle2 , retaining an advantage in the
ending. 28 lbb7?!
32 fxe3 g5.
28 . . . 'it'h6!
29 �d2
29 �c7? lbe2 + ! .
29 . . . g5?
Black could have won by 29 .. .'it'g 6 ! ! 30 �xf4
e3 ( N u n n ) .
30 ttJcs
30 �c7? 'it'b6.
30 . . . 'ii'g 7
31 l::te 1
Wh ite overlooks the knight sacrifice, which
has been in the offing for a long time. But
Wh ite has won a pawn and he has parried what was he to do? After 31 'lt>h2 h6
Opposite-Colour Bishops in the Middlegame ttJ 1 35
A serious positional mistake . By exchanging defending all their pawns and blocking a
his active rook, Black submissively con passed pawn of the opponent, he does not
cedes the in itiative to the opponent. He have to fear the penetration by the e nemy
should have exchanged not the rooks, but pieces on sq uares of the opposite colour.
the queens. Since if 33 . . .'ilkf7 there is the Of course, in the m idd legame this logic does
reply 34 l:te2 ! , he should have chosen 33 . . . not apply. The pawns should cover the
'ii'c6 o r 33 . . . 'ii'c8 . For example: 3 3 . . . ii'c8 ! ? squares which are not controlled by the
34 exd5 exd5 (34 . . . l:.xc2? 35 1Ve4) 35 'ii'xc8 bishop. It is clear that, if Kasparov's kingside
l:tfxc8 36 l:te2 l:Ic1 37 l:l.xc1 :Xc1 + 38 'itt h 2 pawns had been standing on l ight squares
.l:tc8 39 .tg6 .tf6 with an i nferior, but (g6 and h7), he wouldn't have had any
probably tenable ending . problems. I should also mention the typical
34 �xc2 'ii'c 6 regrouping of pawns on dark sq uares (g2-g3
35 �e2 'ii'c 5 and h3-h4) carried out by Karpov.
And here is another usefu l observation . The
36 l:tf1 'ii'c 3
flexibility of the pawn structure and the
37 exd5 exd5
presence of a mobile pawn chain can deci
38 Si.b1 ! sively influence the evaluation of a posi
The triumph of Wh ite's strategy - the queen tion. A pawn storm gains significantly in
inevitably reaches the b 1 -h7 diagona l . I will strength if it is supported by an active
give the rema ining part of the game with brief bishop.
notes.
Botvi n n i k-Tal
38 ... 'ii'd 2 39 'ii'e 5 .ll d B?! (39 . . . �f6!? 40 'ii'f5
'it>g8) 40 'ii'f5 'it>gB 41 'ii'e6+ 'it>h8 (4 1 . . . 'ittf8 World Championship Retu rn Match ,
42 .tg6 'i¥f4 43 .l:.e 1 ) 42 'ii'g 6 'it>g8 43 1!i'e6+ 3rd Game, Moscow 1 96 1
'it>h8 44 Si.f5! (43 .:f.e 1 l:tf8 ! ) 44 .. .'ii'c 3 45
'iig 6 'it>gB 46 Si.e6+ 'it>h8 47 .tf5 'itt g 8 48 g3!
'it>f8 49 'it>g2 'ii'f6 50 'ii' h 7 'ii'f7 51 h4 ..td2
(otherwise 52 l:te1 ) 52 l:td1 .tc3 53 l:.d3 lld6
54 .l:.f3! (54 l:te3? g 5 ! ) 54 ... 'it>e7 (54 . . . .:tf6 55
.l:!.e3 l:txf5 56 'ii' h 8+ 'iig 8 57 lieS+) 55 'i¥h8
d4 56 'ii'c 8 l:tf6 57 'ii'c 5+ 'it>e8 58 :t4 'i!i'b7+
59 l:te4+ 'it>f7 (59 . . . .l:!.e6! ? would not have
helped in view of 60 'S'c4! lbe4 61 'ii'g 8+
'it>e7 62 'iixg7+) 60 'ii'c4+ 'it>f8 61 .th7! l:tf7
62 'iie6 'S'd7 63 'ii' e 5 Black resig ned .
36 . . . h4? 37 g4 is prematu re. Simagin finds attack the weak a 3- and c4-pawns. And if the
an excellent plan: he switches his bishop to wh ite rook goes to their defence, the
c7, setting up a battery which will th reaten exchange sacrifice on f3 then becomes
the wh ite king . Then the pawn breakth roughs possible.
will become more dangerous. 47 .l:tb1 'it>h6
36 . . . j_d8 ! Already now it was possible to give up the
37 bxc5 exchange: 47 . . Jbf3! 48 'i!Vxf3 (48 'it>xf3
White cannot get by without this exchange: 'i!Vf5+ and 49 . . . 'i!Vxb 1 ; 48 .l:tb7+ l:l.f7) 48 e2
. . .
from c7 the bishop will be defending the e5- 49 .l:!.b 7 + 'it>h6 50 'ii'f8+ �g5 5 1 h4+ '.tg4 52
point, and the b4-pawn will be attacked . 'i!Vf3+ 'it>xh4 53 .Uh7+ 'it>g5 and wins. Howev
37 . . . bxc5 er, this possibility will never ru n away from
Black.
38 �b1 ?! j_c7
48 .i::f. b 3 j_d2
39 j_a4
49 .l::t b 6
Boris Spassky tries to include his bishop in
the defence. Now, when the bishop is still on
its way to the kingside and the black pieces
have already taken up ideal attacking posi
tions, it is the right time to break through the
opponent's defences.
39 . . . e4!
40 dxe4 fxe4
41 .Uxf7+ .l:txf7
42 j_d1
42 iVxe4 'i!Vxg3 43 iVg2 'i!Ve3! is also
hopeless for White - Black effectively has an
extra piece, since the enemy bishop is taking
no part in the play. For example: 44 .Ug 1 �h8
45 .Uf1 (defending against 45 .. J::tf2) 45 . . . .Uxf1 + 49 . . . .Uxf3 !
46 'i!Vxf1 'it>g7 4 7 j_d7 'i!Vd2 ! 48 'ii'g 1 'it'e2 ,
50 .Ue6
and Wh ite is completely helpless (variation
50 'it>xf3 'i!Vh5+; 50 'i!Vxf3 e2.
by Simagin).
50 . . . 'i!Vxe6
42 . . . e3
51 dxe6 .Uf2+
Th reatening 43 . . . .l:tf2 .
52 'it>xg3
43 �f3 h4!
52 'i!Vxf2 gxf2 53 e7 e2 or 53 \t>f1 'it>g7.
Black has a decisive attack. 44 g4 or 44 gxh4
is not possible because of 44 . . . .l:txf3 . 52 . . . l:l.xe2
31 . . . i.d5 suggests itself, but after 32 e6! the Attack on the long diagonal
wh ite pieces come al ive. Therefore the
We will begi n with a rather simple example.
bishop remains on its blockading square to
the end of the game. Thus the threat of
. . . i.d5 remained (according to the i ronic Perl is-Marshall
defi n ition of Bobby Fischer) 'an eventual Vien na 1 908
possibility' .
3 2 �b4?
Wh ite parries the threat of 32 . . J:txe 1 + 33
i.xe 1 �d 1 in the most unfortunate way.
Meanwhile, in the event of 32 �c3 or 32 �f1
the outcome would have remai ned u nclear.
32 . . . iVe3+
33 i.f2 �8d2 ! !
I n this hopeless position (the black king
hides from the check q ueens at g6) Wh ite
lost on time.
Black resig ned , because there is no defence better. He has been able to provoke a2-a3,
against the th reats of 43 f6 and 43 l1c1 . and now the wh ite rook has to defend the a
pawn . The knight has occupied the strong
I now wa nt to show you two of my games, e4-poi nt and will soon be supported by . . . f7-
played with one and the same opening f5 . The exchange on e4 is u nfavourable for
variation. And although the positions arising Wh ite , whereas after tLlf3-e5 he has to
were nearly identica l , the character of the reckon seriously with the exchange on e5,
play in them was diametrically opposite. since then Black can occu py the c5-square
Everything depended on which of the players with his knight. Which is what happened in
was able to seize the in itiative. the game.
1 2 lLle5? i.. x e5!
Vi kulov-Dvoretsky 1 3 dxe5 lLlxd2
Moscow Championship Semi-Final 1 97 1 14 'ilfxd2 lLlc5
Queen 's Indian Defence 1 5 �c2 dxc4
1 d4 lLJf6 1 6 bxc4 .l:tfd8
2 lLlf3 e6 1 7 �d4? !
3 c4 b6 I n the event of 1 7 'ii'e 2 ..ie4 the kn ight is
stronger than the passive bishop on b2. But
4 e3 �b7
now Black reaches a favourable position
5 �d3 i.. b4+
with opposite-colour bishops .
The idea of this check is to lure the knight to
17 . . . lLle4!
d2, so that it should not occupy the best
1 8 �xe4 �xe4
square c3. Theory recommends 5 . . . d5 or
5 . . . c5. 19 f3 �b7
6 lLlbd2 0-0 1 9 . . . �g6! ? followed by 20 . . . c5 also came
into consideration .
7 0-0 d5
20 'ilfc2
8 a3 ..id6?!
8 . . . �e7 is preferable. Where does this tell?
Firstly, after 9 'ilkc2 ! ? lLlbd7 9 e4 dxe4 1 0
lLlxe4 the bishop would be better placed at
e7 rather than d6. Secondly, after 9 b4 ! ? c5
1 0 cxd5 it is desirable to capture on d5 with
the queen.
9 'ii'e 2?
But now my open ing set-up proves com
pletely justified .
9 . . . lLle4!
1 0 b3
1 0 b4!? c5.
10 . . . lLld7
1 1 �b2 'ilie7 Black effectively has an extra pawn on the
Black can be pleased with the outcome of q ueenside, but the d ifference in the placing
the opening - he already stands slig htly of the bishops is even more important. My
Opposite-Colour Bishops in the Midd legame ltJ 1 45
bishop is pressing on the kingside, and it can 26 . . . Ii8d 7 ! . I delayed and played someth ing
also attack the c4-pawn , whereas the wh ite slightly weaker, but this did not change the
bishop is obstructed by its own pawn on e5 character of the play.
and has no prospects at a l l . 26 . . . j_a8 ? !
In the first instance Black must prevent c4- 27 h3 h5
c5 and gain control of the only open file. 28 'ii'c 2
20 . . . c5
21 .ltc3 lid7
22 a4
My opponent wants to get rid of his vulnera
ble rook's pawn and g ive me a weakness on
b6. But we know that, when there are
opposite-colour bishops, play on the q ueen
side is less effective than activity on the
opposite side of the board , which Black will
soon develop. I would have preferred 22
l:tfd1 , although after 22 . . . .U.ad8 23 .U.xd7
'ixd7 Black has a n obvious advantage.
22 . . . .l:tad8
23 a5 'ii'g 5!
Black's pressu re on t h e kingside has ena
24 .l::!.a e1
bled him to tie down the opponent, but for the
24 f4 ? would have lost i mmed iately to moment there is no direct way to wi n . To his
24 . . . .l:!.d2 ! . aid comes the principle of two weaknesses.
24 . . . �d3 He must stretch the opponent's defences by
I make on that I am attacking the pawn . But, creating d iversionary th reats on the opposite
of course, this is not so: it is not possible to side of the board . The new target is the c4-
capture on e3 in view of f3-f4 or the pin on the pawn .
c1 -h6 diagonal . 28 . . . j_b7
2 5 axb6 axb6 29 .l:.e2 �8d7
If 26 j_d4 I was intending 26 . . . 1:.8xd4! 27 30 �h1
exd4 .l:.d2 . Of course, I overlooked the 30 'ii' b 2 j_a6.
unexpected i ntermed iate move 27 h4 ! , win 30 . . . j_a6
ning the exchange, but i n overwhelming
3 1 'ii'a 4
positions such oversights are not too danger
If 31 j_b2 Black can finally capture the e3-
ous. After 27 . . . 'ii'x h4 28 exd4 'ii'x d4+ 29 l::.f2
pawn , which has been en prise for a long
ic6 ! ? (more accu rate than 29 . . . h6 30 'iWa4)
time.
White's position is unenviable.
31 . . . l:.xc3
26 'ii' b 2
32 'ifxa6
Now Black would have liked to include his h
pawn in the attack, but after 26 . . . h5?! 27 32 'ii'x d7 j_xc4 was no better for Wh ite .
'fxb6 .l:!.xc3 28 'ii'x b7 l:tc2 (or 28 . . .'i!i'xe5 ) the 32 . . . 'ii'd 8!
opponent has the good defence 29 f4! . The The q ueen defends the b6-pawn and takes
most accu rate move was the prophylactic up the 'approved' position beh ind the rook on
1 46 � Opposite-Colour Bishops in the Middlegame
the open file. Black's position is won . If 33 Of cou rse, 12 .. .f5 suggests itself. H owever I,
'ith2 I was intend ing 33 . . . h4 followed by was not sure that the advance of the f-pawn
34 . . . l:!.d 1 (or 34 . . . .l:.d2). was appropriate in the positions arising after
33 "ii'a 1 ? ! llxc4 1 3 b4 c5 or 1 3 cxd5 exd5 1 4 Ji.a6.
3 4 l:ta2 'ili'g5 On the basis of this, it was logical to play
1 2 . . . a5! , preventing both of these possibili
35 .l:!.a8+ 'ith7
ties for the opponent. But I made a less
36 'ili'b1 + 'iig 6
accu rate move, wh ich neutralises only the
37 g4 second of them. What operated , apparently,
37 'ili'xb6 l:ic2 38 l1g 1 lld 1 . were associations with the previous g a me - I
37 . . . hxg4 remembered that there my rooks had operat
ed q u ite well on the d-file, and I hurried to
38 hxg4?!
occupy it.
Of course, there were also no saving
chances after 38 'ii'x g6+. 1 3 Ji.c2 a5
38 . . . l:.c2 Again Black defers . . . f7-f5 because of 14
b4.
39 �g1 lld2
1 4 lDe5 Ji.xe5?
Wh ite resig ned .
Another move made by analogy. But where
as i n the game against Alexander Vikulov the
N isman-Dvoretsky
exchanges led to an advantage for Black,
Moscow 1 972 here the result is the opposite.
Queen 's Indian Defence 15 dxe5 ltJxd2
1 d4 ltJf6 2 c4 e6 3 lbf3 b6 4 e3 Ji.b7 5 Ji.d3 1 6 .:txd2 dxc4
ii.. b4+ 6 lbbd2 0-0 7 0-0 d5 8 a3 Ji.d6?1 9
1 7 bxc4
'i!Ve2? ltJe4! 1 0 b3 lDd7 1 1 Ji.b2 'ili'e7
1 7 'ifxc4? ttJxe5! .
17 . . . lDc5
But I simply overlooked the following strong impossible - then the terrible d iagonal for the
move by Wh ite . b2-bishop is opened . I n cidentally, precisely
1 8 .l:i.d4! such a structu re occu rred in the highly i nter
The rook takes control of the e4-sq uare, at esting game Taimanov-Averbakh , which I
the same time creati ng the threat of 1 9 insistently recommend that you look at - you
�xh7+! 'it>xh7 20 'ifh5+ 'iit g 8 2 1 l:th4 . B ut the will find it in David Bronstein's famous book,
main th ing is that now the exchange of rooks devoted to the 1 953 Candidates Tournament.
becomes practically impossible for Black, 22 . . . h5
since Wh ite will recaptu re on d4 with his 23 'ii'f4
pawn, obtaining a mobile pawn chain in the
centre , and then will soon advance d4-d 5 .
Instead o f the plausible 1 7 . . . t"Llc5 i t made
sense for Black to play 1 7 . . . t"Llb8, with the
idea of attacking the rook on d4 with the
knight. However, as Vad i m Zviagintsev point
ed out, after 1 8 Ild4! g6 ( 1 9 ii.xh7+! was
threatened ) 1 9 .l:!.g4 ! lid? 20 h4 lifd8 2 1 ii.c3
his position would have remained anxious.
For example: 21 . . . t"Llc6 22 h5 t"Llxe5!? 23
�xe5 (less accu rate is 23 hxg6 t"Llxg6!? 24
J::l.g3 'ifh4) 23 . . . .l::r.d 2 24 hxg6! hxg6 25 .l:!.xg6+!
fxg6 26 'i!Vg4 1::!. xc2 (26 . . .�f7 27 �f1 �c2 28
t'h4) 27 'it'xg6+ 'it>f8 28 'ifxc2 with advan
tage to White .
What do you think, whose bishop is better? It
18 . . . g6
may seem that the comparison is in favour of
19 'i!Vg4 Black - after a l l , his bishop is pressing on g2,
1 9 a4 !?. whereas its opposite nu mber is obstructed
19 . . . a4! by the pawn on e5. But let us look a l ittle
The only counter-chance! By placing his more deeply. Not one of my pieces is
kn ight on b3 Black will most probably supporting the bishop, so that its activity is
provoke the advantageous exchange of the purely superficial. Whereas the opponent
dangerous bishop on c2 . The far-advanced has chances of penetrati ng on the weakened
pawn on b3 will promise tactical cou nter dark squares on the kingside (after the
chances, or for a certain time will at least opening of the long diagonal, or on the c1 -h6
divert the opponent from his attack. diagonal ) , and then my king will be in trouble.
20 liad 1 t"Llb3 Black cannot passively mark time - the
21 ii.xb3 axb3 opponent will play 111 d 3 , capture the b3-
pawn , and then prepare either e3-e4 , or g2-
22 h4
g4. He must try to seize the in itiative , but
As is customary with opposite-colour bish how? He had to decide on a very risky
ops, White attacks on the kingside. It is operation.
important to note that his rook can not be
23 . . . l:l.xd4! ?
driven from the d4-square by . . . c7-c5 - it will
occupy an even more powerful position at 24 exd4 b5!
d6. Exchanging it there will be altogether 24 . . . .l:!.d8 25 1::!. d 3 .
1 48 � Opposite-Colour Bishops in the M iddlegame
25 d5!
Boris N isman correctly senses the spirit of
the position and , not paying any attention to
pawns, endeavours to open the diagonal for
his bishop. The miserable 25 cxb5?! would
have allowed me to activate my forces by
25 . . . �d7 26 a4 .:ta8 27 .:ta 1 'ifd5 28 f3 c6! ?
(but not 2 8 . . ."ii' c4? 29 .:tc1 ) 29 b6 'iic4 .
25 . . . 'ili'c5!
Activity first and foremost! I n the event of
25 . . . bxc4? 26 d6 things are bad for Black,
since the exchange on d6 is suicidal, and
after 26 ... 'i!i'd7 27 'ii'xc4 �d5 28 'ii'f4! he has
no counterplay. How should Black defend? 29 . . . �g7? 30
26 'iff6!? i.h6+ and 29 . . . �g8? 30 i.h6 .:ta8 31 'i'f6
Wh ite is intending to switch his bishop to h6. are completely bad . Only two possibilities
For the sake of this he is ready to part with remai n : 29 . . Jig8 and 29 . . . b2 30 i.xb2 (30
his pawns and even his rook. A clever idea , �xf8 'ii'xf1 +) 30 . . . �g8 .
but, as we will see, Black has a defence. It is easy to make the only possible moves,
However, all the same I do not see a direct but far more d ifficult when there is a choice.
way for the opponent to win . In the event of The price of a mistake in such a sharp
26 dxe6 fxe6 27 "ii'g 3 Black has a choice situation is extremely h i g h , and therefore a
between 27 . . . �e4 and 27 . . . �g7 28 l:.d7+ very carefu l calculation is demanded . Alas, I
.l:i.f7 . If instead 26 d6, then 26 . . . 'i!i'xc4 . In the failed to display this.
endgame it is now White who would have to When checking the variation 29 . . . b2 30
find a way to save himself: 27 'ifxc4? bxc4 i.xb2 <ittg 8 I was not afraid of the return of
28 d7 lld8 followed by 29 . . . �d5, or 28 dxc7 the bishop to the c1 -h6 d iagonal: 31 .ic1
�d5 and 29 . . . .l:!.c8 . If 27 .l:td4! Black can reply 'ii'x e5 32 i.h6 .l::f.a 8, or 31 �d4 �e4 ! . But I
27 . . . 'ii'c6!? (not 27 .. ."W/c2? in view of 28 d7! was frightened by the combi nation 31 dxe6
i.d5 29 l:!.xd5 exd5 30 e6, but 27 . . . 'ii'e 2 'i!Vxb2 32 exf7+ :Xf7 33 'iie 8+ 'it>g7 34 e6
comes into consideration ) 28 'ifg5 'i¥d7 and and wins, for example: 34 . . . .l:.f4 35 'iid 7+ (or
29 . . . c5. Here Black's position is uneasy, but 35 'W/e7+ 'it>g8 36 �d8+ llf8 37 e7) 35 . 'it>h6 . .
nevertheless his queenside pawns guaran 36 e7 ll.g4 37 'ii'x g4 ! . In fact Black can save
tee him counter-chances. himself, by playing 32 . . . �g7! (instead of
26 . . . 'i!i'xc4 32 . . . .:txf7?). There is an even simpler draw
27 �c1 ! 'it>h7 by 3 1 . . . ..txg2! (instead of 3 1 . . ."�xb2 ).
Of course, not 27 . . . 'ii'c2? 28 i.h6! 'ifxd 1 + 29 29 . . . l:.g8?
�h2 with unavoidable mate. 30 ..tg5! b2
28 'ife7 'i¥e2 3 1 i.f6 'i!i'xf1 +
29 l:[f1 32 �h2
Black is a rook up, and his pawn is on the
(see diagram) threshold of q ueening. And yet there is no
satisfactory defence against the mating
threats, created by j ust two enemy pieces.
Opposite-Colour Bishops in the M iddlegame ltJ 1 49
the e5-pawn , which at a conven ient opportu attacks the g7-poi nt: 4 6 We7 l:tb8 (46 . .'i'b8 .
nity may also be able to advance further, to 47 h6 i.. c3 48 .l:.d7 i.. d 4+ 49 l1xd4) 47 h6
open up the black king's position. 'iWc3 48 .l:id7 .l:!.b 1 + (48 . . . 1Wc 1 + 49 iJ1 ) 49
40 . . . 'it'b6 'it>h2 11h 1 + 50 'it>xh 1 Wc1 + 5 1 'it>h2 'iixf4+ 52
A waste of a tempo. The bishop is doing 'it>g 1 'iic 1 + 53 .tf1 . It is better to give up the
nothing on c7 - its place is at d4, and Black exchange: 44 . . .'ii'xc4 45 e7 'ii'e 6 46 exf8'1+
should immed iately have begu n manoeu 'it>xf8 , in the hope after 47 'ii'xe6 fxe6 of
vri ng it there: 40 . . . 'iib 8 4 1 f4 .ta5. putting up a stiff resistance in the endgame.
B ut it is not essential to excha nge queens -
41 f4 'ii' b 8
stronger is 47 it'a?! i.. b 6 (47 . . . 'ili'b6 48 'i'a8+
'itt e 7 49 l:t a 1 ! ) 48 'ili'b8+ 'it>e7 49 l:r.b1 i.d8 50
'fia7+ and 5 1 'ifxc5.
44 l:.d61
The bishop was wanting to go to d4, blocking
the d-file, and so the rook h u rriedly advances
to an active position . The threat is 45 h6. If
44 . . . h6, then 45 'iif5 , intending 46 .Uxh6, 46
l:td7 or 46 i.. d 3.
44 . . . 'ii b 1 +
45 'it>h2 h6
46 'ifxf7+! l:txf7
47 J:td8+ 'it>h7
48 i.. x f7
42 h4! Black resigned .
A typical attacking resou rce in such situa
tions! The pawn wants to advance to h6,
King's I n d i a n structu re
breaking up the enemy king's defences. If it
is met by ... h7-h6, the b 1 -h7 diagonal is Levenfish-Kan
weakened and White's bishop and q ueen Moscow 1 927
can switch to it. The pawn on h 5 will also
come in useful if the opponent plays . . . g7-
g6.
I ncidentally, in reply to 42 . . . g6 Isaak Bole
slavsky suggests the spectacular break
through 43 e6 fxe6 44 f5. But after 44 . . . gxf5
45 .txe6+ 'it>h8 46 i..xf5 Black has a
defence: 46 . . . i.. h 2+! 47 'it>h 1 'fie? . Therefore
Wh ite should prefer the simple 43 h5! or 43
e6 fxe6 44 h 5 ! .
42 . . . .ta5
43 h5 i.. c 3
43 . . .'ii' b4 should also be considered , after
which Boleslavsky was i ntending 44 e6 ! . I n
the event of 44 .. .f6 45 'ifxe6+ 'it>h8 White This a characteristic position from the King's
Opposite-Colour Bishops in the M iddlegame l2J 1 s1
hu rry, but to keep manoeuvring and await a defends them from b4 , Wh ite advances his
more convenient moment for the break pawn to h6, putting the opponent in zug
throug h . zwang.
3 9 'ii' h 5 .l:tg6 This conclusion is not altogether accu rate
40 .l:.5f2 ifd7 in fact Black saves h i mself by sacrificing two
41 i.. e4 .l::t g 5 pawns and changing the roles of his pieces:
54 .. .f5! 55 i.. xf5 e4! 56 .txe4 i.. f6 57 'it>g2
42 'ii' h 6 'ii'g 7
cJitg7 58 �f3 �f7 59 i..f5 �e7 60 �e4 'it>d6
43 'iVh3 'ii'c 7 with a d raw. But, of cou rse, Wh ite can easily
44 'ii h 6 'ili'g7 gain the tempo that he lacks i n this variation
45 'iVh3 'ili'c7 by 52 'ii'x d7 .l::t x d7 53 i..f5! .l::t g 7 54 .l::tx h 7+ (or
46 .l::tf5 ! l:lxf5 first 54 �g2).
If 46 . . . .l::t g 7 there would probably now have 50 lixe5??
followed 4 7 g4 with the threat of 48 g5 l:txg5 I n his book of selected games and reminis
49 l1xg5 fxg5 50 .l::tf8+ cJitg7 51 l:te8 .l:lh6 52 cences, Levenfish lamented a serious defi
'iif5 'ii'd 6 (or 52 . . . i..f6) 53 i.. d 5. ciency in his play. After outplaying his
4 7 l:txf5 .l::t d 6 opponent and gaining a decisive advantage,
48 g4 I:f.d7 he would often make a serious error and ruin
the fru its of his preced ing work. This was
also the case here. It is hard even to explain
why Wh ite felt the need to exchange rooks.
Surely not for the sake of winning the e
pawn? But when there are opposite-colour
bishops, pawns are of no sign ificance - you
should be thinking only of attack!
After 50 .l::tf8+ ! cJitg7 51 .l::te 8 Black would have
had to resig n , whereas the move in the game
leads only to a draw.
50 . . . 'ii'x e5 51 'iix d7 'ii'e 7 52 'ii'f5 i.. c7 53
�g2 i.. d 8 54 �f3 i.. c 7 55 h3 i.. d 8 56 'it>g4
�g8 57 i.. d 5+ �g7 58 i.. e 4 �g8 59 i.d5+
�h8 60 �f3?! 'ii'e 3+ 61 �g4 'ii'e 2+ 62 'i'f3
'ii'e 7 63 'ii'e4 h5+?! 64 �xh5 Vxe4 Draw.
49 g5! fxg5
Kan reckons that he could have held the We have mainly been studying the strategy
position by 49 . . . .l::tf7 50 g6 .l::t g 7. Romanovsky of play with opposite-colour bishops, but for
retorted that the ending arising after 51 l1h5 dessert I invite you to solve a few combina
'ti'd7 52 l:.xh7+ l:txh7 53 'ii'x h7+ 'ii'x h7 54 tive exercises. Most of them (although not
gxh7 was hopeless for Black. The wh ite king a l l ) are elementary, but nevertheless useful,
invades the enemy position via e4 and goes since they demonstrate tactical ideas typical
across to the queen side pawns. If the bishop of positions with opposite-colour bishops.
Opposite-Colour Bishops in the M iddlegame ltJ 1 53
Exercises
Sol utions
Mark Dvoretsky
Noth ing is given by 35 'it'd4 .l:!.e8 36 .tf4 (with This was the position for which N u n n aimed,
the th reat of 37 'iie 5) in view of 36 . . . 'ifd5. when he made his 29th move. He sensed
Therefore 35 'iix b4 suggests itself, but after that, despite the material equal ity, things
35 . . . .l:td8! 36 �xd8+ �xd8 Black successful would be bad for Black. A very deep
ly defends, for example: 37 .lli. g 5+ �c7 38 calculation and a completely correct assess
'ifa5+ �b8 39 'ifd8+ .lli. c8 , or 37 'iff8+ 'ife8 ! ment!
(weaker is 37 . . . �c7 3 8 .i.f4+). Why is Wh ite's advantage so appreciable?
I n a midd legame with opposite-colour bish The entire blame l ies on the awkward
ops the most important th ing, as we know, is position of the black king in the centre of the
the in itiative. Even if there are comparatively board , and , as usua l , the presence of
few pieces left on the board , in the first opposite-colour bishops g reatly strengthens
instance you should th ink not about winning the attack. N u n n 's pieces a re domi nant on
material, but about creating threats to the the dark squares. The enemy bishop is
enemy king . unable to help here in any way, and the
35 'ife5! ! q ueen also is hardly participating in the
Threatening 36 .lli.f4; i f 3 5 . . . .Uh5 there follows defence, since it is tied to its own bishop.
36 .i.g5. Thus the king remains alone against Wh ite's
superior forces .
35 . . . .l:rd8
I should also mention the good position of
36 .l:rxd8+ ..t>xd8
the f3-pawn (as stipulated by the rules - on a
37 .i.g5+ �d7 sq uare of the colour of the opponent's
37 . . . �c8? 38 'it'h8+ , or 37 . . . ..t>e8? 38 'ifh8+ bishop) - it secu res the e4-poi nt for the
�f7 39 'ii'h 7+ �f8 40 'ii'e 7+ 'itg8 41 .i.f6 . queen and takes away this square from the
38 'ii'g 7+ �d6 black quee n . The q ueenside pawns are also
You can't get by without a Combination ! l2J 1 59
ready to join the attack: b2-b3 followed by 5 1 . . . �b6 52 f5 e3, then after 53 �d4+ �c6
c2-c4+ . Wh ite is also threatening the i m me 54 'iig 6+ 'iid 6 55 'ilfe8 it will be lost.
diate 41 c4+ �d4 (4 1 . . . �e5 42 Wc3+ �d6 51 . . . 'ii'd 7
43 'it'd4+ ) 42 'ii'c 3+ �c5 43 b4+ �b6 44 52 'ifh6+ �d5
'i'd4+ with inevitable mate . I n the event of
53 Wb6 ! 'iic 6
40 . . . �e5? he g ives mate by 4 1 �f4+ �f5(f6)
42 'ii'fB . If 40 . . . 'ii'd 7 there is the strong reply 54 'ii'd 8+ �e6
41 i.f4 ! with the threat of 42 c4+ �c6 43 55 'ikf6+ �d7
'l'a4+ . 56 'it'g7+ �e6
The concl uding stage of the game convinc 56 . . . �c8 57 f5 e3 58 f6 e2 59 'ikg4+.
ingly demonstrates how helpless Black is. 57 'ii'g 4+ �
We will only examine the main variations - a 58 f5 'ifh6
detailed analysis can be found in the afore
59 f6
mentioned book.
Black resigned .
40 . . . e5!
41 'ii'e 4+ �d6
Accord ing to the comments in the book,
4 1 . . .'iti>c5 42 'ii'x e5+ 'ifd5 43 'ii'c 7+ 'ikc6 44 transposing into a position with opposite
'i'e7+ �b5 45 �e3! followed by b2-b3 and colour bishops was the only correct solution
c2-c4+. for Wh ite . However, in their analysis only two
42 'ii'x g6+ 'iti>d5?! of our pupils opted for this cou rse (and on the
42 ... 'iti>c5 43 �e3+ �d5 was a much tougher 35th move, alas, they only considered 35
defence, but even then after 44 'ii'f7 + �d6 45 'ii'x b4? instead of 35 'ike5 ! ! ) . All the rest tried
c4! 'it'd? (45 . . . 'ii'c 7? 46 c5+ �c6 47 'ii'e 6+ to exploit i m mediately the unfortunate posi
'it>b5 48 c6 ! ) 46 'ii'f8+ �e6 47 'ii' h 6+ �f7 48 tion of the black king and the power of the g7-
'lh7+ Wh ite would have retained a powerful pawn . To my su rprise they succeeded , and,
attack. what's more, i n various ways.
43 'iff7+ 'ifi>d4 Let us return to the position arising after 29
44 'ii' b 3! e4 e5 .U.xh8 30 exd6+ .i.xd6 . Apart from the
The only defence against mate by the bishop move i n the game, N u n n also examines 31
from e3. ttJxe6+ fxe6.
45 .i.e3+ �e5
46 f4+
Apart from a continuing attack, White has
also acquired a new 'trump' - a passed f
pawn .
46 . . . �f6
47 'i!Vg8 'ii'd 5
48 �d4+ ! �e7
49 'ifg7+ �d6
50 b3! �c6
51 �e5
Threatening not only a terrible check at c7,
but also 52 f5. If Black advances his pawn :
1 60 � You can't get by without a Combination!
The variation he gives is 32 'ii' b 6+ 'Ot>b8 33 After checking the variations we concluded
'ifxd6+ 'Ot>a8 , and White has no time to that Black is helpless, for example: 30 . Ir.he8 . .
captu re on h8, since his c2-pawn is attacked . 31 i.. x e 7 l:lxe7 32 exd6+ �xd6 33 tt'lb5+
Kiryakov and Baklan found the excel lent 'Ot>e5 (33 . . . '0t>c6 34 l:td6+ 'Ot>xb5 35 'iY'b6+) 34
qu iet move 32 'ii'd 4! 1 . If 32 . . . l::l. h d8 Wh ite 'ii'd 4+ 'Ot>f5 35 t2J d6+ 'Ot>g5 36 l:tg 1 + 'it>h6 37
wins by 33 g8'ii' ! . If instead 32 '0t>b8, then 33
• • •
l:.h 1 + 'it>g5 38 'i!kh4 mate . N u n n agreed with
b3! 'ii'c6 (33 . . . 1\i'ea 34 1\i'xd6+ 'Ot>a8 35 our opinion , adding the i nteresting variation
gxh81\i' 1\i'xh8 36 'ii' b 6) 34 gxh8'ii' 'ii'x c2+ 35 30 . . . l:.hg8 31 i.. x e7 dxe5 32 tbb3 (perpetual
'Ot>a1 l:.xh8 36 1\i'xhS+ i.. c 8 37 l;lc1 and wins. check results from 32 t2Jxe6+ fxe6 33 .id6+
'Ot>d7 34 i.. c 5+ ) 32 . . . 'ii' b 5! 33 tbc5 (33 .id6+
However, as N u n n pointed out in the 2nd
'Ot>c6 34 'ii'e 3 is unclear) 33 . . . l:.a8 34 If.d7+
edition of his book Secrets of Grandmaster
'Ot>c8 35 'iVxb7+ 'ii'x b7 36 l:txb7 l:.xg7 37 .id6
Chess, Black can put up a tough defence by
and wins (the refinements in brackets are
32 . . . i..d 5! 33 gxh8'i!V l:.xh8 34 'iig 7+ 'ii d 7 35
mine). However, here Black's defence can
'ii'x h8 i..xf3 or 34 'ili'xh8 b3.
be improved : 34 . . . '0t>c6! (instead of 34 . . . Wc8?).
The analysis by Kadymova was less good , Going into an ending by 35 'ifxb7+ is now
unfortu nately: 32 i.. b 6+ 'Ot>c6 33 gxh8'iV l:.xh8 u npromising, which means that White is
34 i.. d4 .Uc8 35 'ii' b 6+ 'Ot>d7 . She then obl iged to repeat moves : 35 :d6+ 'it>c7 36
considered 36 'iWxb7+ l:!.c7 37 'ii'e4, but this l:td7 + '.t>c6 ! . So that, alas, the brill iant bishop
is unconvincing in view of 37 . . . b3! 38 cxb3 move to g 5 is not good enough to win .
'ili'xb3, and Black's chances are not worse.
36 i.. c 5! wins. However, Black can defend Let us retu rn t o the position with which we
more accurately: 34 . . . 'ii' b 5 (instead of 34 . . . beg a n . Seryozha Movsesian analysed the
.:c8?) 3 5 i.. x h8 �c5 3 6 'ii'x c5 i.. x c5, and the consequences of 29 tiJb3 ! ? 'i!kb5 (29 . 'i'c6 . .
most probable outcome is a draw. Another 30 tba5 'ii' b 5 3 1 l:txg8 .l:txg8 32 J:!.d4).
way of making a draw was later suggested I ncidentally, i n the game the knight had only
by N u n n : 34 . . . l:tb8!? 35 1i'b6+ 'Ot>d7 36 i.. c 5 just been at b3, and the q ueen at b5 - the last
'ii'a 2+! 37 'Ot>xa2 i.. d 5+ 38 lixd5 l:!.xb6 . moves were 28 tbd4 'iWa4 .
Perhaps the most unexpected and spectacu
lar idea was found by g randmaster Sergey
Dolmatov during a training game which
began from the orig inal position: 30 i.. g 5!?.
I. Igor Khenkin
pawn not on d5, but on d6, preparing . . . e6- such cases - his f5-pawn is 'hanging') 1 0
e5. For Wh ite I can recommend that you .i.xb8 ! l:.xb8 1 1 lDf4 Wh ite's chances are
.
check 5 d 5 ! ? , which , I think, as yet no one better - he places his knig hts on d3 and f3,
has played . and then he beg ins an attack on the
Of course, I must also mention the l lyin queenside, by advancing his b-pawn .
Genevsky Variation (4 . . . .i.e7 , 5 . . . 0-0 and 7 b3 fie? 8 .i.b2 0-0 9 lDd2 is another
6 . . . d6). 4 . . . .i.b4+ has also been employed , promising set-up. Then the d2-kn ight goes to
with the idea after 5 .i.d2 of retreating e5, and the other knight is qu ite well placed
5 . . . .i.e7 - the bishop on d2 is not too well on h3, controlling the f4-point. (If it had come
placed . 5 lDd2 is stronger, when it is not clear out in the opening to f3 , to obtain a similar
what the black bishop is doing on b4. construction Wh ite would have had to spend
5 lDh3! a couple more tempi: lDf3-e5-d3 and lDd2-
f3-e5). One of the possible subsequent
plans is 'ifc2 , l:tad 1 , lDf4 , f2-f3 and e2-e4 .
6 0-0 0-0
7 b3
7 'ili'c2 is also possible, but I prefer fi rst to
develop my bishop - who knows, perhaps
the q ueen will also come in useful on d 1 .
7 . . . c6
Grand master N igel Short, who constantly
employs the Dutch Defence with Black,
plays 7 . . . lDc6 ! ? and then . . . a7-a5 i n such
positions. The knight presses on d4, thereby
preventing the exchange of the dark-sq uare
bishops (8 .i.a3? .i.xa3 9 tDxa3 dxc4 ). After 8
For Black this plan is the most dangerous. .i.b2 a5 Wh ite is obl iged to make the not very
When I made my first g randmaster norm , I useful move 9 e3 - otherwise he ca nnot
scored a very important win in this variation develop his q ueen's knight. However, even
against Vlad imir Tukmakov. in this case I prefer Wh ite's position.
8 i.b2
Khenkin - Tu kmakov The exchange of bishops does not bring
Metz 1 99 1 Wh ite any particular benefits - after 8 i.a3
5. . . .i.e7 .i.xa3 9 lDxa3 the black q ueen obtains a
5 ... .i.d6 6 0-0 c6 has also been played . If 7 comfortable post at e7, whereas the knight at
i.f4 Black should reply 7 . . . .i.e7 ! . I n the a3 is badly placed , and it faces a lengthy
game Bareev-Vaisser (Pula 1 988) after 8 journey via c2 and e 1 to d3 or f3 . If the king's
'ii'b 3 0-0 9 lDa3 (if 9 lDc3 there is the knight were at f3 , controlling the e5-point,
unpleasant reply 9 . . . 'ii' b 6) 9 . . . h6!? 1 0 l:lad 1 such a plan would make sense, but here,
g5 1 1 i.d2 a5! 1 2 f3 b5! Black seized the while the q uean's knight is on its way, Black
in itiative . But later Kozul improved Wh ite's will surely have time to play . . . d5xc4 and
play against Bareev (Biel 1 99 1 ): 8 lDd2 0-0 9 . . . e6-e5 .
'ii'c2 h6 (after 9 . . . lDbd7 1 0 cxd5 Black does 8 . . . lDe4
not have 1 0 . . . exd5, the essential capture in Black has problems with the development of
1 68 � Modern I nterpretation of the Dutch Defence
1 0 tt:Jxe4!
I exchange the opponent's only active piece,
and then open l ines in the centre by f2-f3 , to
exploit my lead in development. It is impor
tant that the pawn is stil l at e2 - after the
exchange on f3 Wh ite will recapture with this
pawn .
10 . . . dxe4
Modern I nterpretation of the Dutch Defence lZJ 1 69
The central idea of Black's strategy is the which it can be exchanged , or . . . c6-c5
restriction of the g2-bishop. In my view, here prepared .
it is no stronger at all than the bishop on c8 . But Black more often plays . . . b7-b6 , . . . .ib7
I have played this opening for both sides and and . . . lZ'lbd 7 , developing his queenside as in
as a result I have become aware that it is the Queen's I ndian Defence or the Catalan
even simpler to play for Black than for Wh ite . Opening. The main d ifference with these
At any event, it is usually more difficult for openings is the position of the pawn on f5.
White to choose a pla n . His actions must Wh ite's thematic e2-e4 advance is now
often vary depending on the opponent's g reatly h indered (if Wh ite prepares it with f2-
plans, i.e. he has to adapt flexibly to Black's f3, Black i m med iately cou nters with . . . c6-
play, which is never easy. I like to employ the c5). On the other hand, after Black carries
'stonewall' as Black against a player with an out his basic plan - completes his develop
attacking, combinative style, because here ment, places his rooks on c8 and d8, and
Wh ite will not give mate , and the strategic advances . . . c6-c5 - pawn exchanges occur
problems someti mes prove too difficult for in the centre and the weakness of the e5-
such players. square may become perceptible.
Wh ite usually chooses one of two continua
I n one order or another let us make the i n itial tions: 7 il.f4 or 7 b3. 7 lZ'lbd2 (or 7 'i¥c2) has
moves: no i ndependent sign ificance - all the same
1 d4 fS he cannot manage without b2-b3. The move
2 c4 lZ'lf6 7 lZ'lc3 is not very dangerous for Black, and in
3 lZ'lf3 e6 general it seems to me that the knight is not
best placed at c3 - all the time Wh ite has to
4 g3 dS
reckon with . . . d5xc4 . However, this is some
5 ..1ll. g 2 c6 times played with the idea of developing the
6 0-0 il.d6 bishop at g5 on the next move. 7 l2Je5 also
occasionally occu rs .
I think that the development of the bishop at
f4 is more logica l . With it on b2, both white
bishops lack prospects: one is obstructed by
its own pawn on d4, and the other by the
enemy pawn on d 5 . I n cidental ly, I agree with
Khenkin : it is more d ifficult for Black to
defend if the opponent's king's knight is
developed not at f3 , but at h3.
7 b3 'iie 7
If 7 . . . 0-0 there fol lows 8 ..ta3, and White's
After Black castles, he has to choose one of position is better - he gains control of the e5-
two ways of developing his q uean's bishop. sq uare. There is no point i n al lowing the
Sometimes he directs it via d7 to e8, exchange which is adva ntageous to White.
vacating the d7 -square for the knight. Later 8 ..lll. b2
the bishop can be taken further, to h5, after 8 lZ'lbd2 is also played .
Modern I nterpretation of the Dutch Defence ltJ 1 71
First let's analyse the plan with . . . .i.c8-d7--e8 . hardly doing anything here. I magine that
8 . . . 0-0 Black has not managed to exchange it: the
9 ltJbd2 knight has gone from f3 to e5, and the e2-
.i.d7
pawn has moved to e3. When Black advanc
10 ltJe5 i.e8
es . . . c�5. the g2-bishop i m med iately
1 1 lDdf3 beg ins operating on the long diagona l , and
White's moves are natu ra l , but if you ask there is nothing to oppose it.
what does he want, what is his plan, it is not 8 . . . b6
easy to obtain a sensible reply. Most This move is probably more accu rate than
probably he has to adapt to the actions of his 8 . . . 0-0 .
opponent.
9 'iVc1
11 . . . �h5
Wh ite insists on the idea of exchanging
12 'ifc2 ltJbd7 bishops. A completely harmless undertaking!
1 3 ltJd3 I!ac8 It simply wastes too many temp i . Once I
Black prepares . . . c6-c5. myself played this, but I i m mediately ended
14 .l:.ac1 u p i n a somewhat inferior position .
9 . . . �b7
1 0 i.. a 3 ltJbd7
11 �xd6 'ii'x d6
I n cidentally, after 8 . . . 0-0 (instead of 8 . . . b6) 9 b6 1 2 'ifc2 ..tb7 1 3 l:fc1 (th reatening 1 4 c5)
1Vc1 b6 (9 . . . b5!?) 1 0 ..ta3 ..tb7 1 1 ..txd6 1 3 . . . tt:'la6 1 4 cxd5 cxd5
1Vxd6 1 2 'iVa3 Black can play the same Forced : 1 4 . . . exd5 is n ow i mpossible, while if
endgame, although without his king o n e7 - 14 . . . tt:'lb4 , then 1 5 d6! 'i!Vxd6 1 6 'ii' b 2 followed
here there is nothing terrible for h i m . But he by tt:'lac4 is strong.
is not obliged to go i nto the endgame -
1 2 . . . c5 is not bad .
If Wh ite really wants to exchange the dark
square bishops on a3, he should do this by 8
a4 and 9 ..ta3.
8 a4 aS
8 . . . 0-0 9 ..ta3 ..txa3 1 0 tt:Jxa3 a5 is equally
good , but not 11 a5! with the better chances
for Wh ite, as in the game Kasparov-Short
(London 1 987).
9 i.a3 ..txa3
1 0 tt:Jxa3 0-0
1 2 tt::l d 3 b6
1 3 'ii'c 2 ii.a6
1 4 l:tfc1 l:tac8
1 5 'i¥b2 tt::l e4
times in practice.
Chiburdanidze-Agdestein (Haninge 1 988):
12 ltJd3 ltJa6 ( 1 2 . . . ttJd7 is also possible) 1 3
ltJf3 ltJb4 1 4 c5 (there is noth ing else - the
a2-pawn is attacked ) 14 . . . bxc5 1 5 dxc5 il.. c 7
16 a3
1 4 . . . b5!
Black is agreeable to the exchange of a
couple of minor pieces. After a l l , his pawn will
now advance to b4 , sh utting the opponent's
bishop and knight out of the game.
15 ttJxd6 'i¥xd6 16 ltJc3 il.. a 6 17 'ifd2 �feB
1 8 f3 b4 1 9 ltJd 1 a4 20 ltJe3
Of course, White was counting on 1 6 . . . ttJxd3
In the event of 20 bxa4 both wh ite pawns on
1 7 exd3, but Simen Agdestein unexpectedly
the a-file will soon be lost.
retreated his knight - via a6 and b8 to d7. A
strong and original manoeuvre! 20 . . . a3
Black has an obvious advantage. The result
16 ... ltJa6! 1 7 ltJfe5 ltJb8
ing position illustrates well the thoughts
Wh ite appears to stand wel l , but in fact she expressed earlier about the unenviable fate
has nothing. Black will now exchange the of the wh ite bishops in this variation. Com
knight on e5, then place his bishop on a6 and pare the light-square bishops on g2 and a6 -
rook on b8, and endeavour to play . . . e6-e5.
which of them is bad?
18 f3 ltJbd7 19 ltJxd7 ttJxd7 20 e4 fxe4 21
fxe4 .l:!.xf1 + 22 'il'xf1 il.. a 6 23 'i!fd 1 �b8
The following is a more solid continuation for
Black has seized the in itiative . Wh ite:
1 2 e3 ltJa6
Tu kmakov-Agdestein (Dortmund 1 987): 1 3 'ii'e2
12 ltJb1
Black usually repl ies 1 3 . . . ltJe4 . I n the g a me
White plays his knight to c3, in order then to Petursson-Dolmatov (Aku reyri 1 988) Black
exchange pawns on d5 and in the event of carried out a different, although strategically
. . . c6xd5 to occupy the b5-sq uare with the rather risky pla n . I have to admit that it does
knight. But this plan is slow, and now it is not not g reatly appeal to me.
obligatory for the black knight to go to a6.
13 . . . il.. xe5?! 14 dxe5 ltJd7 1 5 .Ufd 1 ttJac5 1 6
12 ... ltJbd7 1 3 cxd5 cxd5 14 ltJc4 ltJf3 l:tac8 1 7 il.. a 3 .l::t fe8 1 8 'ifb2
1 8 fxe4 axb3
1 9 axb3 �xeS
20 dxeS l:iad8
21 exdS exdS
1 8 ... gS!
At first sight a senseless decision. But i n fact,
with the centre closed it is hardly possible to
exploit the weakening of the kingside. And
Black, by advancing his pawn to g4, will
threaten the e5-pawn . Black's position is preferable. He will now
1 9 ttJe1 g4 play . . . c6-c5 and later either follow u p with
The next move plan ned is 20 . . . 'i!Vg7 . True, if . . . d 5--d4, or force the advantageous ex
the wh ite q ueen retreats to a 1 , it will be very change on d 5 . He will place one of his pieces
dangerous to captu re the e5-paw n . On the on e6, blocki ng the e5-pawn and thereby
other hand, the mobil ity of the g2-bishop is restricti ng the mobil ity of the bishop on b2 .
now still fu rther restricted , and it is u n l ikely The position of the wh ite king is somewhat
that the wh ite knight will manage to reach f4 . weakened and in the future it may come
At any event, Margeir Petu rsson hastened to u nder attack.
simpl ify the play. Here is the conti n u ation of the game, without
20 �xeS ttJxcS 21 lLld3 ttJxd3 22 l:Ixd3 any notes.
The position would appear to be roughly 22 'ii'e 3 cS 23 'ife2 �a6 24 .l::t a 1 d4 2S �c1
equal . bS 26 �f4 'i¥e6 27 'iVhS?! bxc4 28 �h3
'i¥b6 29 bxc4 �xc4 30 l::!. d b1 'i!Vc6 31 .l::ta 7
Petu rsson-Short 'i!Ve4 32 .Uba 1 �dS 33 l:xc7 'ii' h 1 + 34 �2
Reykjavi k 1 987 l:.xf4+ 3S gxf4 'ii'x h2+ 36 'lt>e1 'iig 3+ 37
Black won this game is classic style. 'it>e2 i.. c4+ 38 'it>d1 'it'd3+ Wh ite resigned .
13 . . . ttJe4 As you see , i n all the examples we have
1 4 l:.fd 1 ttJc7 examined n oth i n g special was demanded of
Black - he simply completed his develop
Later the provocative 1 4 . . . ttJb4 ! ? 1 5 a3 ttJa6
ment and fully prepared h imself for the
was also tried , when Wh ite has problems
opponent's active possibil ities. The e3-e4
with his a3-pawn.
advance did not promise any particu lar
15 f3 ttJxd2 benefits and was double-edged , but Wh ite
1 6 'i!Vxd2 a4 does n ot appear to have any other effective
1 7 e4 fxe4 pla n .
1 76 � Modern I nterpretation of the Dutch Defence
White is condemned to passive defence . continues 1 1 'ii' b 3! .tea 1 2 tDe5 ..t h 5 1 3 e3,
After 2 6 11e2 I defended t h e pawn with my and the development of the black pieces is
rook from c6 , advanced . . . b6-b5-b4 , ex h i ndered . The knight cannot be played to d7,
changed queens by . . . 'ii'd 5, and placed my and a6 is the wrong place for it.
rook on a6, my king on f7 and my pawn on
1 3 . . . 'itt h 8 1 4 .l:.c3 ( 1 4 'itt h 1 ! ? ) 1 4 . . . tDa6 1 5
h5. This led to an i nteresting bishop end i n g ,
'ii'a 3 lDb4 1 6 c5. Black's position i s worse,
which I managed t o win .
since his knight is roaming about on inappro
priate squares.
Plan with i.c1 -f4
Thus, if we want to develop our bishop at d7,
7 ..tf4 ..txf4 we must seriously reckon with 'ii' b 3. And if
Of course, White's kingside pawns must be we choose this pla n , we should do so
spoiled . Otherwise he will play 8 e3 and immediately, without 9 . . 'ii'e 7. .
Modern I nterpretation of the Dutch Defence ttJ 1 79
1 2 . . . dxe4? !
An original decision ! The standard 1 2 . . . fxe4
is sounder. Then events can develop roughly
Now Black's knight wants to go to d7, and his as fol lows: 1 3 tt::l d 2 tt::lf6 14 f3 exf3 1 5 tt::l xf3
queen , incidentally, can retreat to c7 . There ...i d 7 1 6 tt::l e 5 .ie8. Black is pla n n i ng to play
fore Wh ite h u rried to exchange q ueens. 1 7 . . . tt::l d 7 and after the exchange of knig hts
12 'ifxb6 axb6 1 3 tt::l e s ...i h5 1 4 .if3 .ixf3 1 5 to bri ng his bishop via g6 to f5 . If he should
4Jdxf3 tt::l a 6 succeed i n doing th is, it is Wh ite who will be
In the endgame even this route is not bad . having to try to equal ise.
Although White's position is slig htly more 1 3 tt::l d 2?!
pleasant, he has no real win n ing chances. The only way to cast doubts on the oppo
The game ended i n a d raw. nent's strategy was by 1 3 tt::l e 5. Now Black is
entirely successfu l .
Kal i n ichev-Giek ( U S S R 1 987): 9 tt::l b d2 1 3 . . . c5 1 4 tt::l b 3 (altogether the wrong place ! )
4Jbd7 1 4 . . . b6 1 5 dxcS ttJxcS 1 6 tt::lx cS bxcS
Along with . . . .ic8--d 7 , also not a bad pla n . Wh ite's position is worse . The . . . e6-e5
Black aims for a n y exchanges o f knights i n advance is immi nent, the g2-bishop is shut
the centre, after which h e usually does not in, and if Wh ite should try to activate it by f2-
have any problems. f3, then after the exchange of bishops the
10 .Uc1 ( 1 0 e3 and 1 1 'ii'c2 is more accu rate) weakening of his king's position may tell .
1 0 . . . tt::le4 1 1 e3 'i!Ve7
After 1 2 a3 in one game there followed The game which I now wish to show you
1 2 . . . 4Jdf6 1 3 tt::l e 5 ...i d 7 1 4 f3 tt::l d 6 1 5 'it>h 1 ended in a crushing defeat for Wh ite in j ust
...i e 8 1 6 �g 1 .ih5. Then Black played . . . 'it>h8 23 moves. And this is not surprising - on an
and . . . .Ug8, and after c4-c5 he retreated his examination of it one gains the impression
knight to f7 and prepared . . . g7-g 5 . For that the player with Wh ite simply did not
Wh ite it is simply not apparent what he can know where to place his pieces, or which
do. As usual , the g2-bishop is no better than changes i n the structu re were advantageous
its opponent on h5. to h i m , and which were not.
12 tt::lxe4
1 80 \t> Modern I nterpretation of the Dutch Defence
has to repeat moves: 24 .i.f1 (24 l:e 1 .l:r.e8) perfectly playable for Black!
tLJ 1 83
PART IV
A forced move, since now 1 1 . . . 0-0 is very Wh ite's idea : in this way he regains the
strongly met by 1 2 liJd4 ttJxe5 1 3 'ii' h 5 tiJg6 pawn.
14 f4 f5 1 5 h4. 15 . . . <t>xh7
12 .Ue1 0-0 1 6 'ifc2+ �g8
Nowadays 1 2 . . . 'ifd7 1 3 tiJf1 l:d8 , su pporti ng It is extremely dangerous to play 1 6 . . . tiJed3
the d5-pawn , is more usual. 1 7 tiJxg4 f5 ( 1 7 . . . <t>g8 ! ) , since, along with 1 8
13 tiJf1 tiJge5 ttJxe 1 1 9 'ii'xf5+ with a guaranteed
The other developing move is 1 3 tiJb3 ! ? . d raw, there is also 1 8 l:.d 1 fxg4 1 9 tiJe5, and
Wh ite stands better.
13 . . . J:te8 ! ?
1 7 ttJxe5 .i.e6
A n interesting continuatio n , the point of
which is to begin play against the e5-pawn . 1 8 tiJc6
14 . . . ttJxe5
After serious thought Black decided to go
along with his opponent's idea, rightly as
suming that the resulting position would
promise him at least equal chances. The
alternative 14 . . . .i.xf3 1 5 'ii'xf3 ttJxe5 1 6
'iVxd5 is also acceptable for Black, only now 20 . . . tiJd7!
he should not continue 1 6 . . . tiJcd3 in view of
One of the most d ifficult moves i n the game.
1 7 l:td 1 , but 1 6 . . . i..f8 , obtaining cou nterplay
It is always u npleasant to have to retreat, but
thanks to the weakness of the d3-square.
the natural 20 . . . tiJe4 is weaker, since after 2 1
15 i.. x h7+! .i. b 2 .l:r.ae8 ( i f 2 1 . . . c 5 , then 22 c4) 22 liad 1
This intermediate check is the point of the black centre is under fire: c3-c4 is
Crux of the Position l2J 1 85
which it was aiming 1 3 moves earlier! White lDa4 Wd8 1 4 Wb4 or 1 2 . . . lDxd5 1 3 lDxd5
resigned . cxd5 14 .tc7 'iWe6 1 5 'iWb4 lDb6 1 6 0-0 Ilfc8
I n this game I was able t o concentrate at 1 7 .l:lc5 with the better game. What operates
critical moments and come to successfu l here is a standard consideration in such
decisions on the 1 4th , 20th and 23rd moves. structures: White should delay the exchange
After the 23rd move the game was controlled of c-pawns as long as the knight is still on b8 ,
by Black, who found and carried out the good in order not to allow its development on the
plan of playing his knight to d3. This active c6-sq uare.
positional idea became the guiding thread 12 lDxe4 .txe4
which led him to his goa l . The captu re with the pawn is less natu ra l :
then 1 3 c 5 'ilkd8 1 4 lD e 5 is good , as is the
Yusu pov - Gavri kov i m mediate 1 3 lDe5 , since after 1 3 . . . c5 1 4
Zu rich 1 994 dxc5 'iWxc5 1 5 b4 Black cannot play 1 5 . . . 'iWc7?
1 6 lDxf7 e5 1 7 lDxh6+.
Griinfeld Defence
1 d4 ltlf6 13 0-0
1Wxb7 1i'f6 b3 a5. And in the event of 1 4 cxd5 I n stead of the move in the game, 16 ... .l::!. a d8
i.xd5 15 e4 i.xe4 16 i.xh6 i.xh6 17 'ii'x h6 was more accu rate, fig hting against 1 7 e4,
i.xf3 1S i.xf3 l"Df6 1 9 .l:.fd 1 .l:.adS the on which there would have followed 17 . . . g5
opponent begins attacking my weak pawns. 1 S .te3 t"De5 1 9 i.e2 l"Dg6, although in this
What does Black want? H is most natural case too after 20 d5 White would have
plan is to exchange on f3 and c4 and then retained the better chances. 1 7 'ii'c2 'it>h 7 1 8
attack the centre with . . . e7-e5, simpl ifying 'itt h 1 e5 1 9 dxe5 t"Dxe5 20 i.xe5 i.xe5 2 1 f4
the game. It turns out that with a simple was also not bad , with a minimal advantage
developing move I can parry this th reat and for Wh ite.
simultaneously prepare 1 5 cxd5.
1 4 .l:.fd 1 !
Such prophylactic moves are always un
pleasant for the opponent. Especially since
he cannot reply in the same style: the
careless 14 . . . l:.adS?? loses immed iately to
1 5 c5.
14 . . . dxc4
After 1 4 . . . a5 Wh ite can now play 1 5 cxd5
i.xd5 1 6 e4 i.xe4 1 7 .txh6 .txh6 1 S 'ii'x h6
i.xf3 1 9 i.xf3 , when after 1 9 . . . l"Df6 he has
the unpleasant 20 h4, while if 1 9 . . . 'ii'x b2 ,
then 20 h4 'ii'x a2 2 1 h5 'ife6 22 d5 1i'f6 23
hxg6 is good , with the possible variation 1 7 'ii'c 2
23 . . . fxg6?! 24 dxc6 t"De5 25 cxb7 .:tabS 26
I was wrong to avoid the consistent 1 7 e4,
l:tc7 l"Dxf3+ 27 gxf3 with adva ntage to Wh ite.
si nce after 1 7 . . . 'it>h7 1 S .te3 'ii' b4 1 9 'i'e2
1 4 . . . .txf3 1 5 i.xf3 dxc4 1 6 .l:.xc4 e5 is no
the thematic 1 9 . . . e5 can be strongly an
better in view of 1 7 .tg3 exd4 1 S exd4 with
swered by both 20 dxe5 t"Dxe5 2 1 i.b3 'ike?
the th reat of 1 9 l:tb4 , and if 1 S . . . a5, then 1 9
22 f4 t"Dd7 23 e5, and 20 d 5 . But now the play
d5. With the move in the game Black makes
takes on a closed , manoeuvring character,
concessions in the centre, but retains a solid
where it is harder for White to probe the weak
position.
poi nts in his opponent's defences.
15 i.xc4 i.xf3
17 . . . e6
1 6 gxf3 a5 1 8 'itt h 1 l:ifd8
The point of White's idea is revealed in the 1 9 .l:!g1
variation 1 6 . . . e5 1 7 dxe5 t"Dxe5 1 S i.xe5 With the primitive threat of 20 .txe6 , which ,
.txe5 1 9 'ii'c2 ! with advantage. For example, of course, Black easily parries.
19 . . . 'ittg 7 20 .l:.d7 'ii'x b2 (20 . . . .l:.adS 2 1 .l:.cd 1
19 . . . l"Df8
:Xd7 22 lhd7 .txb2 23 i.b3 with the better
game) 2 1 'ii'x b2 .txb2 22 lic2 ! (but not 22 20 i.b3 ! ?
l:tb 1 ? in view of 22 . . . b5!) 22 . . . .ta 1 (22 . . . b5 Prophylactic play b y Wh ite. He prevents the
23 i.b3) 23 .l:.xb7 a5 24 l:td2 .l:!.abS 25 Ita? opponent from gaining space on the q ueen
.:.as 26 .l:.dd7 l:txa7 27 .l:txa7 .tc3 , and now side with 20 . . . a4 followed by . . . .l:ta5 .
the best is 2S h4 ! , to be able to answer 20 . . . 'ii b 5
2S . . . g5 with 29 h5. 21 a4 'ii' b4
Crux of the Position t2J 1 89
35 'ii'x h7? 'ii'f8! , but 35 .i.xe6! �f8 (35 . . . tt:Jxf6 Here the knight is more passively placed
36 'ii'g 6+) 36 'ii'f5! - the 'ambush' threats than at f6 , and Wh ite gains the advantage of
created by the queen are deadly. the two bishops without any particular
28 l:tg1 compensation for Black. 4 . . . c5 is another
orig inal continuation , which after 5 cxd5 exd5
Threatening to move the bishop to c7 with a
6 dxc5 (in the game Psakhis-Korzubov,
discovered check.
U S S R Championship First League 1 983,
28 . . . lt:Jg6 Black equalised after 6 .i.b5+ tt:Jc6 7 tt:le2
28 . . . �h8 29 .i.e5+ f6 would also have lost to tt:Je 7 8 0-0 0-0 9 dxc5 .i.xc5 1 0 b3 .i.g4! 1 1 h3
30 �xe6 tt:Jxe6 31 'fig6 . .i.h5 1 2 �b2 l:tc8) 6 . . . tt:Jf6 7 .i.b5+ i.d7 8
2 9 .i.xe6! �xd7+ tt:Jbxd7 9 tt:Je2 1ed to a more pleasant
All the wh ite pieces swiftly join the attack. position for Wh ite in the game Bandza
Aieksandrov (Fru nze , 989).
29 . . . �h7
5 .i.d2 ! ? 0-0
After 29 . . .fxe5 30 'ii'x g6+ �f8 any move by
the bishop from g3 is decisive. After 29 . . . 'iti>g7 5 . . . c5 6 a3 .i.xc3 7 �xc3 cxd4 8 'ifxd4 f6
Wh ite wins by 30 .i.e5+, while if 29 . . . �f8 - came into consideration, as in the game
30 'ii'f5. Psakh is-Kupreichik ( U S S R 1 984 ). However,
after 9 l:ld 1 Wh ite's position is preferable.
30 'it'f5 l:tg8
6 a3 .i.xc3
Or 30 . . . fxe6 31 'fif7+ �h6 32 .i.f4+.
7 .i.xc3 b6
31 .i.e5 �h6
8 tt:Jf3 .i.a6
32 �f4+ �h7
9 b3 c5
It was no better to play 32 . . . �g7 33 'ii'xf7+
�h8 34 .i.e5+ tt:Jxe5 35 'ii'x h5 mate, or
32 . . . tt:Jxf4 33 'ifxf4+ .i.g5 34 hxg5+ �h7 35
'ii'xf7+ �h8 36 'ifxh5+.
33 'ifxh5+ �g7
34 �h6+
Black resigned .
Yusupov-Lautier
Amsterdam 1 994
Queen 's Gambit
1 d4 d5
2 c4 e6
3 tt:Jc3 �b4 The fi rst serious problem for Wh ite: he has to
Joel Lautier plays the opening in an original decide what pawn structu re it would be
way, employing an u nusual hybrid of the desirable to obta i n . Despite his advantage of
Queen's Gambit Decli ned and the N imzo the two bishops, he has to take serious
l ndian Defence. But if Wh ite does not object account of the opponent's cou nterplay, as
to a Nimzo-lndian with 4 e3, this does not shown by the followi ng variations:
bring Black any benefits. A) 1 0 .i.d3 cxd4 11 exd4 lt:Jd7 1 2 0-0 l:tc8 1 3
4 e3 tt:Je7 'ii'e 2 lt:Jg6! 1 4 g 3 "i/c7 1 5 lt:Jd2 e5 (or
Crux of the Position ltJ 1 91
1 5 . . . dxc4 1 6 bxc4 e5); probably not have repl ied 1 1 . . . cxb3 in view of
B) 1 0 l:tc1 dxc4 11 bxc4 cxd4 1 2 exd4 tiJd7 1 2 ..tb4, but 1 1 . . . tiJd5! ? . However, after 1 2
13 ..td3 l::tc 8 14 ..tb2 tiJg6! 1 5 'ii'd 2 ..tb7 1 6 bxa7 ttJxc3 ( 1 2 . . . .l:r.xa7 1 3 ..td4 ; 1 2 . . . tiJbc6
'ife3 'ili'f6 . 1 3 'ii'd 2) 1 3 'ii'x d8 Jbd8 1 4 axb8'ili' l:taxb8 1 5
..txc4 (or 1 5 b4 ttJa4 1 6 llc1 ) 1 5 . . . ..txc4 1 6
Usually with the two bishops you should not
bxc4 it i s not clear whether he would have
avoid hanging pawns, but in the g iven
been able to demonstrate that Black's
instance the black knight obtains good
in itiative is worth the sacrificed pawns. 1 1
prospects on g6. In view of this featu re of the
bxc4 ! ? also deserved serious consideration .
position, I preferred a sound conti nuation ,
After 1 1 . . . 'ii' x d 1 + ( 1 1 . . . bxc5 is weaker be
which secu res a slight but enduring advan
cause of 1 2 ..td3) 1 2 l::. x d 1 bxc5 1 3 ..te2
tage.
tiJbc6! 1 4 tiJd2 .l:.ab8 1 5 l:tc1 ( 1 5 0-0 is of
10 dxc5! dxc4 equal merit) 1 5 . . . tiJd8 ! ( 1 5 . . .f5 is weaker i n
This intermed iate capture was the main view o f 1 6 f4! followed b y g2-g4) 1 6 ttJe4
problem . After 1 0 . . . bxc5 Wh ite would have tiJb7 White has a slight advantage .
secu red the better pawn structu re by 1 1 11 . . . ..txc4
llc1 ! ? (the simple 1 1 ..td3 is also not bad ). If
1 2 bxc4 bxc5
now 1 1 . . . ttJbc6?, then 1 2 cxd5 ..txf1 1 3 dxc6 .
The reaction to 1 1 . . . tiJbd7 is simila r - after 1 2 1 3 'ili'xd8 llxd8
cxd5 ..txf1 1 3 dxe6 Wh ite has a g reat 1 4 'it>e2
advantage. If 1 1 . . . 'ii' b 6 there is the adequate Of cou rse, in the endgame the king should
reply 1 2 ..td3. It remains to consider the most remain i n the centre .
critical reply 1 1 . . . dxc4 . After 1 2 'ii'x d8 .l:r.xd8 14 . . . tiJd7
1 3 ..ta5 Wh ite's idea becomes clear: he
1 5 tiJd2
takes play into an endgame, where the
weakness of the opponent's pawns begins to
tell . There can follow:
A) 1 3 . . . .l:!.d5 14 bxc4 ( 1 4 ..txc4 is also not
bad ) 14 . . . .l:td6?! ( 1 4 . . . �d7 is better) 1 5 tiJe5
i.b7 1 6 ..tc7 .l:!.a6 1 7 .l:!.b1 with a decisive
advantage;
B) 1 3 . . . llc8 14 ..txc4 ..txc4 1 5 l:txc4 tiJd7 1 6
'it>e2 tiJb6 1 7 llc2 c4 1 8 ..txb6 axb6 1 9 .l:!.xc4!
(1 9 bxc4 is less accu rate because of the
reply 1 9 . . . b 5 ! ) 1 9 . . .llxc4 20 bxc4 l:txa3 2 1
.l:!.b 1 with advantage;
C) 1 3 . . . .l:!.d6 14 ..txc4 ..ixc4 1 5 .l:lxc4 tiJd7 1 6
'iti>e2 l:!.a6 1 7 l:ld 1 .:Xa5!? ( 1 7 . . .tiJd5 1 8 b4)
18 .l:!.xd7 tiJd5 1 9 .l:!.b 7 .:xa3 20 ttJe5 with
Here it was Lautier's tu rn to take an
somewhat the better game.
important decision: he had to decide on a
11 ..txc4 plan of action . Generally speaking, here
I made this natural move without any Black has little possibil ity of infl uencing the
particular hesitation , althoug h perhaps it character of the subsequent play. His only
would have been worth giving it some plan of exerting some pressu re on my
thought. After 1 1 cxb6 !? the opponent would position is by active play with his knights.
1 92 � Crux of the Position
Therefore here the placing of the knights is of tage with 1 6 g4 ltJb6 1 7 .l:.ab 1 f6 18 h4 ( 1 8
great importance. My opponent did not .l:tb5 ltJa4) 1 8 . . . �f7 1 9 h 5 lid7 20 lihc1 liad8
sense the importance of the moment, and 2 1 llc2 .
without much thought he made an outwardly 16 . . . l:!.ab8
natural and 'pretty' move . . . Now if 1 6 . . . ltJb6 there would natu rally have
15 . . . ttJc6? followed 1 7 a4 .
The c6-square is only apparently a good one 1 7 lib5!
for the knight - in fact here it is restricted by I n principle it i s advantageous for White
the opponent's bishop, it is not attacking to exchange one pair of rooks, si nce he
anything, it is not preventing anyth ing and it holds the in itiative and it will be easier for him
lacks any clear prospects. When I discussed to i nvade the opponent's rear, while at the
and analysed the game with Mark Dvoretsky, same time it will more d ifficult for Black to
we came to the conclusion that this su per gain cou nterplay. However, every trifle is
ficial and aimless move was the main cause important, and before exchanging rooks it is
of Black's defeat. not bad to provoke a weakening of the b6-
The knights should have been deployed square.
more actively, aiming for pressure on the c4- 17 . . . a6
pawn. 1 5 . . . ttJc8!? was correct, in order to 1 8 lixb8 .l:.xb8
play one knight to d6, and the other to b6. As
the variations given below demonstrate, at
best Wh ite would have retained only a slight
advantage:
A) 1 6 a4 ltJd6 1 7 a5 f6 1 8 g4 .Uab8 1 9 h4 e5
(with the idea of . . . e�4. but 1 9 .. .<lt>f7 20 f4
h5!? is also not bad) 20 f3 �f7 and by playing
his knight via f8 to e6, Black equal ises ;
B) 1 6 g4 ltJd6 1 7 .:tac1 f6 (if 1 7 . . . ltJb6 1 8
i.e5 f6, then 1 9 i.xd6 .l:.xd6 20 ltJe4 with the
better chances for Wh ite) 1 8 h4 ltJb6 1 9 i.a 1
e5 20 g5 � 2 1 lihg 1 h5 22 gxh6 gxh6 23
h5 l::t g 8 24 f4 exf4 25 exf4 l:tae8+ 26 �d3 (if
26 �f3 there follows 26 . . . ttJbxc4! 27 ltJxc4
ltJxc4 28 Itxg8 liJd2+) 26 . . . l:Id8 27 .l:.xg8 What to do now? The d i rect 1 9 .l:tb1 does not
.l:.xg8 28 ltJe4 ttJxe4 29 �xe4 .l:.e8+ promise anyth ing, since the exchange of all
(29 . . . <ii?e 6!? 30 f5+ �f7) 30 �f5 ltJc8 3 1 l:id 1 the rooks merely eases Black's defence: he
ltJe7+ 32 �g4 lig8+ 33 �f3 �e6 with an will not have to worry about the invasion of
eq ual position; the enemy rook. I need another target to
attack. I n this position the superiority of
C) 16 l:.hb1 ltJd6 17 i.a5 .l:.dc8 1 8 l:.b3 <ii?f8
bishop over knight is somewhat camou
1 9 f4 (less is promised by 1 9 .l:.d3 �e7 20
flaged , but it only req uires Wh ite to begin
lld 1 in view of 20 . . . ltJb6 21 .l:.c1 ltJa4)
exploiting his qual itative advantage on the
1 9 . . . �e7 20 e4 f6! 2 1 .U.ab 1 l:.ab8 with a
kingside (the fact that he can advance his
minimal advantage for Wh ite .
pawns there, whereas Black is forced to
1 6 .l:thb1 remain passive), when the long-range power
White would have secured a slight advan- of the bishop will tell .
Crux of the Position ctJ 1 93
Bareev - Salov
- position after 1 0 b4 -
Linares 1 992
Bogo-lndian Defence
1 d4 ttJf6 Three suggestions have been received : 1 )
. . .'i!i'e7 and . . . e6-e5; 2) . . . c7-c5 ; 3): . . a7-a6
2 c4 e6
and . . . b6-b5 . It is evident that you lack
3 ttJf3 .ib4+ experience in such situations, otherwise you
Today the Queen's I ndian Defence and the would have known that fi rst you should place
Bogo-lndian Defence 3 . . . .i b4+ are to be your bishop on e4 . Not your knight, but your
found in the repertoire of nearly every player. bishop! Then you can play . . . c7-c5 . This is
Without a deep study of Queen's I ndian set how Black acted in most of the games I
ups it is impossible to play the Closed know, and Salov made his next move
Games, j ust as a knowledge of the Ruy practically without th inking.
Lopez is necessary for the Open Games. 10 . . . .ie4
4 ttJbd2 b6 1 1 l:!.c1
Positions of qu ite different types arise after For the moment the play is of a qu iet,
4 . d5 or 4 . . 0-0 .
. . .
manoeuvring character, and the players are
5 a3 .ixd2+ solving purely positional problems. Ti me for
6 .ixd2 .ib7 thought is restricted - only two hours. It is not
7 g3 known when the maximum output of time
7 �g5 is more often played , but the move and effort will be demanded , but in the future
made by me is no weaker. such a moment is bound to come. And for the
present you must comparatively qu ickly -
7 . . . d6
with in one or two m i n utes - make sensible
8 .tg2 ttJbd7 moves. If on each move you spend , say,
9 0-0 0-0 three, five or ten minutes , then later, when
1 0 b4 the crisis is reached and it becomes impor
ta nt to think perhaps for half an hour, you will
(see diagram) not have a reserve of time. And even if you
outplay your opponent and obtain the better
Do you know the typical ways of playing such position, because of approach ing time-trou
positions? What plans does Black usually ble, accu mulated fatigue and lack of fresh
carry out? ness, mistakes will become simply inevitable
1 96 � Strategy in Grandmaster Games
defences as strongly as possible. From this Salov has deployed his pieces very skilfully.
point of view it is worth considering the 30 . . . tt:Jf7
attempt to latch on to the g5- and h6-pawns I was expecting only 30 . . . tt:Jd7, so that the
by 26 h4 . If 26 . . . gxh4, then 27 gxh4, and the knight should guard the f6-pawn . It turns out
rook ca n be switched to the kingside along that this is by no means obligatory. But which
the third rank. is more correct, 30 . . . tt:Jd7 or 30 . . . tt:Jf7 ? I
It would not be bad first to include the q ueen don't know - I fear that this is an irresolvable
in the attack, by placing it, say, at c1 . But then problem.
it will cease to oppose the . . . e6-e5 break. 3 1 i.. c 3 �e8
I did not in fact manage to fi nd a move which This was Salov's idea. I am forced to lau nch
completely satisfied me, but i n a practical into tactical compl ications.
game this is often a very d ifficult, al most 32 i.. xf6 gxf4
unrealistic task. Time is limited , and a
33 i.. x g7
reserve of it is needed for the solving of
purely tactical problems after the inevitable If 33 gxf4 there is the u npleasant reply
opening of the position . Thus sometimes you 33 . . .'�Jh5.
must quickly make some logical move, 33 . . . 'it>xg7
realising that it may not be the strongest. 34 gxf4
26 "iVc1 ! ? f6 I am playi ng on the knight at f7 and the rook
Black decides against playing . . . f7-f5 be at a6, which are not too well placed , and I
cause of h2-h4 , but now the plan of f2-f3 , hope to be the first to switch my pieces to the
g3--g4 and i.. g 3 gains in strength . kingside.
27 f3 tt:Jg7 34 . . . �xe2
35 �g3
28 �g2
A natural move, but 35 "iVf1 !, intending f4-f5
The bishop is no longer needed at f1 , and
and �g3+, was probably stronger. After
from g2 it will support the e2-e4 advance. If
some five m i nutes' thought I did not find this
Wh ite should succeed in playing th is, the
possibility, but I saw another idea.
opponent will be deprived of . . . e6-e5, his
only real cou nter-chance, and therefore he 35 . . . 'it>h7
must exploit it immediately.
28 . . . es
29 dxe5
Strategy g ives way to tactics. I exploit the
fact that 29 .. .fxe5? is bad in view of 30 h4 ! ,
when the g5-pawn can not b e held. If
necessary, the pressure on the c1 -h6 d iago
nal can be fu rther intensified by i.. d 2 (now it
will be understandable why the q ueen went
to c1 ).
29 . . . tt:JxeS
30 f4
Strange: White's actions look logica l , but the
opponent's position is still holding out - 36 i.. f1
200 � Strategy in Grandmaster Games
A question: why did I play this? Was my flag the powers of perhaps only our expert
about to fall? Wel l , we each did indeed have Volodya Kramnik. Besides, Black's defence
only about a minute left. Perhaps you don't can be improved . I n stead of 38 . . . l:Ia8? he
like my move, but you realise that I did have can choose 38 . . . �e4 ! .
some idea? That's right, the switching of my It is good i f you have a t least half an hour in
rook to g2 - I very much wanted to g ive mate reserve - you can sit calmly and calculate
on the g-file. variations. But for some reason time is
The black rook has to retreat on the e-file invariably lacking when you most need it. If
(36 . . . Iic2 37 'ii'e 3 is obviously not worth only one could learn to play chess, so that at
considering). But to which square? To e4 , e7 the required moment there is always suffi
or e8 , or perhaps even e6? cient time!
The correct reply is e 7, in order to defend the 36 . . . .:te8?
7th rank. And from a8 the second rook will 37 .l:!.a2
defend the 8th rank. But defi n itely not the
It is probable that the position is now lost - it
other way round!
is hard to defend against the mate on g7.
Let's look concretely at 36 ... l:.e7 . Sacrifice
37 . . . �e4
on g6? This gives noth ing - the knight on f7
secu rely defends the king . That means 37 38 'ii'c 3 ltJhB
l:!.a2 ltJh8 38 l:!.ag2 l:!.a8. In the event of 38 . . . .l:!.g8 the simple 39 'ii'f6 is
How should White conti nue the attack? strong . It transpires that Salov had his own
That's right, 39 f5! �xf5 40 �d3, and if idea - the switching of his knight to g6. The
40 . . . �xd3 White has the decisive 41 'ifxh6+ ! . ideas of g rand masters sometimes tu rn out to
But Black fi nds the defence 4 0 . . . .:tf8 . After 4 1 be bad , but it is very rarely that they play
l:th3 there i s 4 1 . . Jie6, s o 4 1 lif3! must be without ideas.
tried , and if 41 . . . l:lef7?, only then 42 l:ih3, 39 �d3 ltJg6?
and mate is inevitable. But how to win the 39 . . . �xd3 40 'ii'x d3+ .l:te4 41 .l:tag2 .l:!.a7 was
game after 4 1 . . . .:te5 , do you see? There is a essential.
pretty idea suggested by Kramnik: 42 l:lg4
40 .Uag2! l:.xa5??
(renewing the threat of 43 \!Vxh6+), and if
42 . . . ltJg6, then 43 .l:!.h3 .l:!.e 1 + 44 'ifxe 1 �xg4 A typical fortieth move in time-trouble. In the
45 l:!.g3 with an attack. But Black has the tou rnament bulletin it is awarded two ques
reply 42 . . . �xd3, and if 43 .l:!.xf8 the double tion marks, but u ndeservedly - all the same
attack 43 . . . 1i'e7. There is no mate : 44 .l:!.xh8+ there was no way of saving the game.
�xh8 45 'i1Vxh6+ �h7. [It is not essential to [No, quite deservedly - since Black could
take the h6-pawn: 45 "ik c3! 'ii'e 6 46 I!g3 - the have seized the initiative with 40 . . . �xd3! 41
bishop is attacked, and the deadly 4 7 ltJxc6 �xd3 .l:!.e 1 +! 42 � f2 l:te4. Therefore White 's
is threatened. But after 42. . . ltJg6! 43 i1h3 last move should be awarded not an excla
Black parries the attack by 43. . . 'iig 7! - mation mark, but a question mark. He could
Dvoretsky.] have won by 40 �xe4! l:!.xe4 (40. . . dxe4 4 1
Instead of 42 l:!.g4 I would probably have ::1ag2) 4 1 :txg6 or 4 1 "ikf6. In a sharp time
preferred 42 'iff1 ! (or 42 "ikb 1 ! ) . scramble such mutual tactical errors are,
alas, almost inevitable - Dvoretsky.]
If m y opponent h a d played 3 6 . . . Iie7 ! , in time
trouble I would have been unlikely to see the 41 Iixg6
mating idea f4-f5 and 'iix h6+ - this is with in Black resigned .
Strategy in Grandmaster Games l2J 201
For a start, in a good position Jan Tim man for here the knight is more actively placed - I
some reason gave up his central pawn : 39 th ink it was Tarrasch who fi rst said this?
J:tc1 ?? l:txd4 40 l2Jxf6. And now, instead of Here the world champion fell asleep. I n
the easily winning 40 . . . .l:txh4+ 41 'itt g 2 �h3+, principle, o f cou rse, he knows everything
Karpov chose 40 . . .'it>xf6?? 41 l:.xe6+ �xe6 and he usually plays the open ing qu ickly. But
42 �xe6+ <;tJg7 43 'ii'f7 + 'itt h 6 44 'it'f6+ l2Jg6 if you set him some problem, he becomes
(44 . . . .U.g6 allows mate on h8) 45 'ii'x d4. the same as anyone else - he sta rts to think,
Although he held on for a fu rther ten moves, and sometimes for a long time.
it was exclusively thanks to his personal 7 . . . h6
qual ities.
After the game the champion timidly sug
gested that he should have played 7 . c5. . .
(see diagram)
202 cJ;t Strategy in Grandmaster Games
11 . . . g5 13 . . . .i.d7
fxg4 4Jcxe4 (with the idea of . . . 4Jxd5), and if 2 1 bxc3 lixd 5 , then 22 lLlf2 with a subse
20 i..f3 - 20 . . . 4Jxd2 followed by 21 . . . e4 . q uent blockade on the light squares. Black
[Apparently, after 20 t0c3!? t0xc3 2 1 bxc3 has th ree pawns for the knight, but without
t0xd5 22 c4 White would nevertheless have cou nterplay this is insufficient - he needs
stood better - Dvoretsky.] some dynamic factor, such as two connected
Each of us has his own style, his own way of passed pawns.
playi ng. In Black's place I would have agreed White would have lost (or almost lost) after
to the exchange of queens. But Kasparov 20 tLlf2? tLlg3 2 1 .l:tg 1 4Jxe2 22 'iti>xe2 e4, and
does not like positions in which he has no the b2-pawn can not be defended .
counterplay. He did not want to go i nto a Khenkin suggests 20 .i.b6 .l:l.xd5 2 1 i.f3. I
qu iet and slightly inferior (as it seemed to don't l i ke this - I wanted to make normal ,
him) endgame, and so he decided on a solid moves, whereas here, with his develop
rather dubious piece sacrifice. I n add ition he ment incomplete , White scatters his pieces
took into account my reputation of being a around too much. Apart from 2 1 . . ..l:lb5, he
player who is incapable of calculating a even has to reckon with 2 1 . . . 4Jf6 22 i.xd5
single variation , and he wanted to exploit 4Jxd5. He is a rook up, but the g4-pawn is
this. In the end I did indeed go wrong , but lost, and in the future also, possibly, the h5-
here he was simply lucky - the resulti ng pawn . Too complicated !
position was not one of those where I often 20 . . . tLlg3
go wrong, since in fact there was hardly
21 .l:tg1
anything to calculate.
I thought for a long time about 21 tLlf2 - an
17 . . . 4Jfxe4
attempt, by also giving u p the exchange, to
1 8 fxe4 4Jxe4 play for a blockade. But a rook is neverthe
1 9 'il'xa5 llxa5 less stronger than a knight.
What move should White make? 21 . . . 4Jxe2
2 1 . . . 4Jxh5 has been suggested . Then , if
there is noth ing better, 22 4Jxg5 lLlf4 23
4Jge4 4Jxd5 24 4Jxd5 .Uxd5 25 g5 and White
has an obvious advantage. 21 . . .f5 22 l:!.xg3
f4 also did not work: 23 i.. b 6 ! .
22 '1t>xe2
What move should Black make? Of course,
he has to open the diagonal of his dark
square bishop.
22 . . . e4
And what should Wh ite play?
The first begins with the obvious 1 3 . . . lbc5, When you are playing such a complicated
but after 14 �b2 he needs a move which game, you constantly have to chose from
does not i mmed iately suggest itself: 1 4 . . . g 6 ! . continuations that are al most equ ivalent.
After e2-e3 t h e bishop goes with gain of One move is slig htly stronger, another is
tempo to f5, and then the knight to d 3 . slightly weaker - how do you sense this?
You suggest preparing t h e development of After analysing the position later, I came to
the bishop by 1 4 . . . lbg4 (instead of 1 4 . . . g6)? the conclusion that instead of 1 6 . . . g6 it would
To be honest, I didn't see this move. Let's have been more accu rate to play 1 6 . . . a5!.
have a look: 1 5 h3 �f5 1 6 'ili'd 1 lbf6 1 7 g4 The move is a sensible one, of course, but
i.g6 - everything would seem to be alrig ht. why it should be made at this precise
Perhaps this was even stronger. moment - d u ring the game it was completely
i mpossible to understand th is. I will explai n :
I chose a different plan - an interesti ng one,
after lbg5-e4 i t is important t o have t h e reply
but on another occasion I would not have
. . . a5-a4 . There is the th reat of driving away
gone in for it ( 1 3 . . . lbc5 is simpler and safer).
the bishop by . . . a4-a3, and if the wh ite rook
13 . . . .l:leB is no longer at a 1 , it will be possible to
1 4 �b2 �f8 ! exchange pawns on b3 and adva nce the
Now if 1 5 e3 there follows 1 5 . . . d3! and rook to a2.
1 6 . . . e4 . All this is rather complicated , and I cannot
1 5 lbg5 lbc5 condemn myself for the move made in the
If the e4-square is conceded to Wh ite's game, 1 6 . . . g6. But my next inaccu racy was
kn ight, he will gain an obvious advantage, altogether more serious.
and for free. My idea is first to exchange the
knights, then the light-sq uare bishops, to
leave the opponent with only a bishop on b2
and not to allow it to become active, by
meeting e2-e3 with . . . c6-c5.
It is impossible to guess Karpov's next move.
Dvoretsky's suggestion is 1 6 a3. Close! This
was the second move suggested by the ex
champion after the game.
16 h3
Typical Karpov! After th i n king what he was
defending against, I fi nally rea l ised what I
myself want - 1 6 . . . g6 and 1 7 . . . ..if5 . I am
ashamed to say that even after 1 5 . . .lbc5 I did
An important and rather complicated mo
not yet see the idea of bringing out the
ment in the game. Can you work out what is
bishop to f5. After mentally thanking my
happen ing here and what should be played?
opponent for the h i nt, I continued playi ng.
Often , for some reason , it is bad moves
16 . . . g6 which instantly come to m i n d . However, it is
1 7 g4 easier to fi nd a bad move than to make it,
Karpov has taken control of the f5-sq uare because you can not bring you rself to play it
and is continuing the fight for the light and you beg i n to have doubts .
squares. And here we have the fi rst not altogether
210 � Strategy in Grandmaster Games
successful suggestion: 1 7 . . . b6. Why weaken 1 8 . . . �h6 (instead of 1 8 . . . dxe3) hardly helps
the c6-square? In order to combat Wh ite's Black. Wh ite has a pleasant choice between
plan? But what is his plan? You th ink it is to 1 9 exd4 and 1 9 tt::lg e4 .
place the knight on e4? Noth ing of the sort! How should Black defend? 1 7 . . . �h6 is better
His main idea is the undermining move e2-e3! than 1 7 . . . �g7, but it involves a loss of time
You would have played your bishop to g7? (the bishop is doing noth ing on h6). After 1 8
You have found someth ing to boast about. tt::lg e4 tt::l xe4 1 9 tt::lxe4 tt::l xe4 2 0 'ii'xe4
Many would have played this, and that's Wh ite's position is preferable i n view of the
what I played . But I shouldn't have done. same idea of the underm i n i ng move e2-e3.
We will fi nd the correct solution if we The correct solution is 1 7 . . . tt::l e 6 ! . Now if e2-
understand why the seemingly rather natu ral e3 Black always has the reply . . . c6-c5. I n
move 17 . . . �g7 is bad . No, it is not a matter of the event of 1 8 tt::l g e4 tt::lxe4 1 9 tt::lxe4 i.g7
1 8 b4 tt::le 6 1 9 tt::l g e4 - after all, here I have we reach a position which occu rred in the
the excellent f4-point. game.
I ncidentally, the problem also proved to be I will once again emphasise: sometimes i n a
too difficult for Karpov. He is a mig hty game a critical moment is reached . If you
prophylactic player, and is especially strong miss it, if you fail to fi nd an accu rate move,
in moves such as 16 h3, but here there is the game changes cou rse and may begin to
noth ing for Wh ite to defend against - he go down h i l l .
simply needs to fi nd an active pla n . T h e problems which Karpov and I were
After 1 7 . . . �g7? the reply 1 8 e 3 ! is very trying to solve were seemingly purely strate
strong . If 1 8 . . . d3, then 1 9 'ii'd 1 . The idea was gic, but at the same time the play was full of
to support the d3-pawn by . . . e5-e4 , but now tactics, little concrete variations, which need
1 9 . . . e4 is dangerous because of the ex ed to be discovered and calculated . Some
change sacrifice 20 �xffi �xf6 21 tt::l g xe4 , or times you have to delve i nto a position and
first 20 b4!?. If 1 9 . . . h6 20 tt::l g e4 , while after th ink for at least ten minutes. You can 't
1 9 . . . h5 there is the strong reply 20 b4 . I n spend ten minutes on every move - other
every case the d3-pawn , lacking support by wise, before you know it, you end up in time
the other pieces and pawns, is lost. trouble. But at important, critica l moments,
Black would probably have had to reject time should be spent. The q uestion is merely
1 8 . . . d3 in favour of 1 8 . . . dxe3 1 9 fxe3 , when to decide which moment is critica l . Anatoly
White has achieved exactly what he wanted . Evgenevich and I did not manage to do this,
He has deployed his forces wel l and has won and we both missed it.
the strategic battle . 17 . . . Ji.g7?
One of the experts suggested that, although 1 8 tt::l d e4? tt::lfxe4
the strategic battle has been lost, the 1 9 tt::l x e4 tt::l e 6
position can be held by tactics: 1 9 . . . �h6. Let
It is now time for Wh ite to forget about
us check: 20 .l:txffi �xg5 21 .l:l.xg6+! hxg6 22
striving for an advantage and to th ink about
'ii'xg6+, and in whose favou r are the tactics?
how to equalise. The underm i n i ng move 20
Of course , it does sometimes happen that a
e3 is now pointless because of 20 . . . c5 , and
chance tactical opportun ity changes the
the exchange on d4 will g ive me the very
evaluation of a position. But normally, if you
important f4-point.
are strateg ically lost, here everything col
lapses and all the variations turn out not in 20 b4
your favou r. Apart from 20 e3, I also considered 20 c5,
Strategy in Grandmaster Games l2J 21 1
but the move in the game did not occur to doing noth ing, whereas from e7 it covers two
me. Meanwhile it is q u ite thematic. Why was important diagonals, defends the f6-square
Karpov not in a hu rry to advance his pawn to and attacks the b4-pawn . If 24 c5, then
c5? To avoid conced ing the d 5-point to the 24 . . . .i.e6 25 lDd6 .l:lf8 . It is probable that I
bishop (20 c5 lDf4 21 e3 lbxg2 and .. .i.e6- was instinctively afraid of placing my queen
d5). at e7 because of some tactics involving
20 . . . lDf4 .l:lae 1 - but I was wrong! 24 l::ta e1 axb4 25
21 e3 lbxg2 axb4 'i!i'xb4 26 exd4 does not work in view of
26 . . . �e6 , or, even better, 26 . . Jla2 . Perhaps
22 �xg2
Wh ite should play 24 exd4 immed iately, but
How would you assess this position? More then he has to reckon with the interposition
pleasant for Black? But it is not enough to of 24 . . . axb4 ! ? .
give such an assessment - you must also
23 . . . l:tf8
demonstrate it, and this is not easy. For
example, noth ing is g iven by 22 . . . .i.e6 in The next problem is what to do as Wh ite. It is
view of 23 lDg5. Here I engaged in some like th is, when you a re playing a game, and
th ing rather pointless - I tried to breach the there is no time to relax - one problem after
opponent's defences immediately, by analys another!
ing 22 . . . c5 23 lbxc5 b6 24 lDe4 .i.b7. Finally 24 c5 is obviously illog ical - it was against
I sensed : this does not work, White's position this that Black's last move was aimed : now
is too solid. In add ition I have an extremely the knight goes to d6 without gain of tempo.
stupid bishop on g7. It was very usefu l to The correct reply was 24 f3! . White must
realise that at the very first opportun ity I consolidate. If Ka rpov had found it, he wou ld
should try to get rid of this useless piece. not have lost. But he made a move from the
22 . . . aS! ra nks of 'how not to play chess'.
23 a3 24 .l:lad 1 ?
Of course , I advanced my a-pawn . If you He wants to captu re on d4, but he plays this
don't play this, you simply lose you r self in the wrong order. For a start I open the file
respect! But what next? which has been conceded to me.
24 . . . axb4
25 axb4 dxe3 !
Now the two possi bil ities available to White
are of roughly equal merit, and neither gives
clear eq ual ity. If he captu res 26 fxe3 , then
after 26 . . . J:.xf1 27 .:f.xf1 'ii'e 7 the b4-pawn is
hanging. I develop my bishop at e6 and at
some point play . . . h7-h5.
26 f3
If Wh ite were to make another couple of
moves in succession and pick up the e3-
pawn , he would stand better. But who is
going to allow that? Can you guess what I
I still can't understand why I didn't make the played?
natu ral move 23 . . . 'i¥e7 ! . At c7 the q ueen is 26 . . . �e7
212 � Strategy in Grandmaster Games
endgame is now lost). But Karpov's flag was Here we have the first explanation : it blocks
hanging, he had no time to th ink, and for the the a2-g8 d iagonal, the king obtains the g8-
moment he did not offer the exchange of square, and after the inclusion of . . . h7-h5
queens. and h3-h4 the advance . . . e5-e4 ! becomes
39 . . . 'ii'f4 possible. This idea is correct, but what is
40 'ii'c 3?! more important is that this advance follows if
Wh ite places his q ueen on e2 (as in fact
40 'ii'c2 ! should have been played , followed
occu rred in the game). Therefore my move is
by ife2 and l:Ie3. Perhaps my opponent was
prophylaxis against the reg rouping of the
instinctively afraid of 40 . . . e4 , but he had no
wh ite pieces , about which I have already
need to be.
spoken. I sat and came to real ise that it was
40 . . . l:tc8 necessary to defend against it - this was the
41 'ii' b2 main reason why Black's move was made.
I n stead of 41 . . . 1lc4 Kramnik suggested
playing 4 1 . . . l:r.fd8, but then 42 'i!fe2 ! .l::t d4 43
l:te3 . However, now I am ready to carry out
his pla n : 42 . . . .l::td 4 followed by 43 . . . l:!.fd8.
The third idea is very primitive , but it also has
to be reckoned with - 42 . . . l:.fc8 with the
threat of . . . l:tc2 , and if 43 .l:ld2 , then 43 . . . .l:l.c3.
Karpov did not guess my idea, and he
q u ickly placed his q ueen on e2 . After my
reply he sat and thought for 45 min utes , but
he was unable to fi nd any defence.
42 'il'e2 h5!
Of course, my opponent was planning 43 h4,
Benefiting from the bitter experience of the but now he saw the refutation : 43 . . . hxg4 44
previous rou nds, after the time control I tt'lxg4 (44 fxg4 'ii'f6 ! followed by 45 . . . 'ii'c 6+)
immediately thought for 25 minutes. Of 44 ... e4 (the breakthrough has become pos
course, it will not be possible for you to sible thanks to the rook on c4) 45 fxe4 l:txe4
determine qu ickly that very strong move, 46 'ifb2+ �g8 47 'ii'a 2+ l:c4. There is no
which in the game I sought for so long . But time to exploit the pin on the rook, si nce the
nevertheless try suggesting some ideas for a wh ite king is threatened with mate . [If 48
playing for a win. How can the opponent's tt'le5!? there follows 48 . . . 1We4+ 49 �g 1
position be shaken? 'l:.f1 +! 50 <J:;xf1 "ikh 1, winning the queen -
Yes, it is tempting to play . . . h7-h5. If Black is Dvoretsky.]
able to place his pawn on h4, his winning 43 tt'le4 h4
chances will be sharply improved . Therefore A new and very serious weakness has
White is obl iged to reply h3-h4 ! . appeared in Wh ite's position - the g3-
41 . . . lic4 ! ! square. Dvoretsky l i kes to emphasise the
A very strong move . What do you th i n k the i m portance of a second weakness when
point of it is? In fact it simu ltaneously converting an advantage. It was easy to
contains three ideas (again Razuvaev would defend the f3-pawn while simultaneously
be happy). attacking e5, but now, to cover the second
Strategy in Grandmaster Games ttJ 21 5
weakness, the knight will be forced to keep order to pin the opponent's pieces and place
guard on e4 , where it is doing noth ing. If, by h i m in zugzwang.
playing his q ueen to f2 , Wh ite were able to 48 . . . Vc1 !
force the reply . . . g6--g5, from e4 the knight
49 'ii'x h4 llc2+
would at least be attacking a pawn, but this
50 liJf2 g5
hope cannot be real ised .
51 'it'g3 'ike1
44 .l:.e3 l:Ic8
52 h4
45 'ikf2 l:!.c2
To where is it more accu rate to play the king
46 lle2 lixe2
- g6 or h6?
47 ..Wxe2 �g7!
52 . . . �g6?!
In completely winning positions one should
be particularly watchfu l . After the move made
by me Wh ite has acquired a chance tactical
opportun ity: 53 f4! ? with the threat of 54
'it'd3+. Fortu nately, here Black's advantage
is too g reat and he wins by 53 . . . .:1.c3 54 'iih 2
'ii'e 2, but in other circu mstances such an
oversig ht might have cost h i m very dearly.
However, Karpov was again in severe time
trouble and was forced to make his moves
i nstantly.
53 hxg5 �xg5
54 f4+
Karpov set me many traps in this game, but I There are no other moves.
too set h i m qu ite a few. Black's last move not 54 . . . exf4
only comes into his plans (it rei nforces the 55 'itf3 l:f.d2!
g6- and f6-squares, and prevents the cap
56 �h3 'ii'f1 +!
ture of the h4-pawn with check), but it also
contains a trap, i nto which my opponent falls. An important check - I saw it earlier when I
But all the same things were bad for h i m - if played 47 . . . �g7 ! . The captu re with the
he had marked time, at some point I would q ueen on f2 would have led to stalemate ,
have jumped in with my queen at c1 , for while the captu re with the rook would have
example: 48 ..Wd3 llc7 49 'ite2 Vc1 . allowed nu merous checks. Also not possible
was 56 . . . 'ike3? 57 ltJe4+ and 58 liJxd2 .
48 ..Wf2
57 'ii'g 2 .l:lxf2
A natural move - intending to captu re on h4
or give a check on a7. What had I prepared Wh ite resigned .
here? The game concl uded , just in time for us to go
That's right, the pawn can be sacrificed , in off for d i n ner.
216 �
Mark Dvoretsky
AYusupov
t the first session o f our school Artur 6 dxc5 lbc6
showed two of his games 7 .if4 i.xc5
(against Anatoly Karpov and Jan Timman), in
8 .id3 f6
which over a nu mber of moves a tense
struggle revolved around one pivotal featu re 8 . . . 0-0? 9 i.xh7+ ! .
- some central strateg ic problem . I n this 9 exf6 "ii'xf6
struggle the outcome of the entire battle was
Opening books recommend 9 . . . tt::Jxf6. But I
effectively decided . Here it is very important
did not like playing in accordance with theory
that you skilfully maintain the tensio n , not
(partly because I did not know it very well)
conced ing an inch to the opponent, and that
and 1 used to take the first convenient
in the protracted strategic dispute you
opportun ity to diverge from the main theoret
continually find new tactical arguments.
ical paths.
1 should like to invite you to look at two of my
Objectively the capture on f6 with the knight
games of the same sort. The fi rst of them
is sounder, and the move made by me is
was exceptionally sign ificant for me in the
rather risky. But I had already tested it in two
competitive sense . It was played fou r rounds
games played not long before the Moscow
before the end of the Moscow Champion
Championsh i p , so that I had managed to
ship. Grandmaster Anatoly Lein and I had
gain some experience . Whereas my oppo
broken away from all our other rivals and
nent was probably encou ntering it for the first
were lead ing with eight and a half points out
time.
of eleven. Our individual meeting could
decide (and did indeed decide) who won the Success in the opening struggle is often
title of champion. determi ned not by objective factors, but by
you r (and you r opponent's) prepared ness for
the events which develop on the board .
Lei n - Dvoretsky
Real ising th is, in this decisive game I
Moscow Championship 1 973 un hesitatingly took a risk in the open ing.
French Defence 10 i.g3
1 e4 e6
Another possibil ity is 1 0 i.g5 ! ? . The game
2 lL'lf3 d5 Bareev-Dvoretsky (Minsk 1 972) developed
3 lL'lc3 lDf6 as fol lows : 1 O . . . "ii'f7 1 1 'ii'd 2?! (the start of an
4 e5 lDfd7 incorrect plan - 1 1 0-0 0-0 1 2 .ih4 is
5 d4 stronger) 1 1 . . . 0-0 1 2 0-0-0 tt::l d e5 1 3 tt::l xe5
This position also sometimes arises via the tt::Jx e5 1 4 f3 .id7 1 5 llhe 1 tt::l x d3+ 1 6 'ifxd3
'normal' move order: 2 d4 d5 3 lbc3 tbf6 4 e5 .l:tac8 1 7 .ie3 .ib4 1 8 .id4 :c4 1 9 <it>b1 .l:tfc8
tt::lfd7 5 tt::lf3 , althoug h 5 f4 is considered to be 20 'i¥e3 b5, and Black had the in itiative .
more dangerous for Black. 10 . . . 0-0
5... c5 1 1 0-0
Whose Strategy will triumph? ctJ 217
Al ready here one can trace the strategic At fi rst sight it seems that White has been
motifs around wh ich the struggle will revolve successful : he has seized control of the aS
over the cou rse of many moves. By advanc square and left h imself with a knight against
ing his e-pawn , Black would like to set up a a 'bad' French bishop. I ndeed , imagine if he
powerful pawn centre . But this is still a long were to play f2-f4 and switch his knight to d4
way off - fi rst he must complete his - my position would i m mediately become
development, exchange the most dangerous strategically hopeless. But the opponent
enemy pieces (the f3-kn ight and the d3- does not have time for th is, and for the
bishop) and rei nforce his dS-point. For his moment the bishop is not i n fact as bad as it
part, Wh ite d reams of establishing control appears. Black has dynamic resources at his
over the d4- and aS-sq uares, blocki ng the disposa l , i nvolving . . . d5-d4 and play on the
black pawns and making the c8-bishop 'bad ' . open c-file. To be honest, I do not even see a
11 . . . lt::l d 4 safe way for Wh ite to gain equality.
11 . . . lt::l d eS? did not work because of 1 2 1 8 . . . .l:tac8 1 9 f4
ttJxeS lt::lx eS 1 3 .txh 7 + . The followi ng variation is typical (although by
1 2 lt::l x d4 .txd4 no means forced ): 1 9 d4 l:tc4 (the immed iate
1 3 'ii' d 2 1 9 . . . bS! is more accu rate) 20 l:.d 1 bS 2 1 l:ld2
b4 22 lt::l e 2 l:tfc8 23 'it>f1 l:tc2 24 'it>e 1 cJ;;f7
Rudolf Kimelfeld (Moscow 1 972) played
(intend ing 2S . . . .tbS) 2S 'it>d 1 ? (2S lt::lf4 is
more pu rposefully against me: 1 3 'i!fe2 lt::l c 5
better) 2S . . . .ta4 ! 26 b3?! (26 lbc2 .txc2+
14 l:tae1 (nothing is g iven by 1 4 lt::l b S .txb2
27 'it>e 1 is necessary, although after 27 . . . .tfS
1 S l:tab 1 .td7) 14 ... lt::l x d3 1 5 cxd3. Here the
the initiative is with Black) 26 . . . .Uxa2! . There's
win of a pawn by 1 S . . . .txc3? 1 6 bxc3 'it'xc3
the bad bishop for you !
would be su icidal - after 1 7 .teS, relying on
his powerful bishop, White develops an Nearly th irty years later the diagram position
attack on the g7-point, whereas Black's was reached i n a bl itz game which I played
'opposite-colour' bishop is completely use against the Cuban g rand master Rivero Be
less. serra (Miami 2000). My opponent chose the
cautious 1 9 l:te2 , but he too was u nable to
There followed 1 5 ... .td7 16 .te5 .txe5 1 7
extingu ish Black's i n itiative: 1 9 . . . bS 20 l:tfe 1
'it'xe5 'it'xe5 1 8 l:txe5.
b4 2 1 lt::l d 1 l:tc1 22 f3 (White is hoping to
(see diagram) tra nsfer his knight via f2 or e3 to g4 and then
218 � Whose Strategy will triumph?
to e5) 22 . . . l:tfc8 23 lL'le3? ! �b5 24 lL'lg4 From this moment there develops an excep
�xd3 25 l:txe6 i.f5 26 l:te8+ l:txe8 27 l:txc1 tionally tense battle for control of the central
�xg4 28 fxg4 l:te2 29 l:tc7 a5 30 l:ta7 d4 3 1 squares.
h 4 (31 l:txa5 l:txb2 3 2 l:td5 l:txa2 3 3 l:txd4 b3) 1 6 lL'le2 �b6
31 . . . l:txb2 32 ltxa5 d3 33 l:td5 d2 34 'Wti>f1 1 7 lL'lxd4 'ii'x d4 1 8 �e5 was th reatened .
l:.xa2 Wh ite resigned . 1 6 . . . �xb2?! is u nfavourable, since after 1 7
1 9 ... d4! 20 lL'le2 l:tc2 21 f5?! l:tb 1 the rook penetrates onto the 7th rank.
The simple 2 1 lL'lxd4 l:.xb2 22 lL'lxe6 was Now Wh ite could play 1 7 d4, but after
preferable, when I would have had to go into 1 7 . . . �b5 the chances are roughly equal.
a sharp double-rook end ing, since 22 . . . �c6 This is not enough for Lein - he wants finally
23 llf2 or 22 . . . .l:r.c8 23 lL'lc5 followed by l:if2 is to win the battle for the centre and he finds a
hopeless for Black. subtle idea .
21 ... exf5 22 lL'lxd4 .l:.xb2 1 7 <i;h1 !
If now 23 .Ue7, then 23 .. J:tf7 24 lL'lxf5 i.xf5 A multi-purpose move! It creates a direct
(24 . . .<it>f8!? 25 l:!.xf7+ <Ji;xf7 also comes into positional threat: 1 8 i.d6 l:tf7 1 9 f4 , then
consideration) 25 l:te8+ l:tf8 26 .U.xf8+ <Ji;xf8 �e5 and at some point lL'ld4 . 1 7 . . . e5? is not
27 l:txf5+ <i;e7 with the better rook endgame possible because of 1 8 lL'lc3 , and both
for Black. central pawns are under attack.
23 l:tc1 g6! 17 . . . .U.ae8
And I remained with an appreciable advan Black parries the opponent's th reat ( 1 8 i.d6
tage. ltf7 1 9 f4 e5!) and prepares . . . e6-e5.
18 lL'lg1 !
Let us return to the game with Lei n . A logical development of the idea begun with
13 . . . ttJcs the previous move. Again I am forced to
14 .l::ta e1 lL'lxd3?! reject 1 8 . . . e5? in view of 1 9 tL'lf3 i.c7
1 4 ... �d7 was more accu rate , with an excel ( 1 9 . . . e4 20 dxe4 dxe4 2 1 'ii'x d7) 20 'ii'c 3!
lent position for Black. The prematu re ex (weaker is 20 d4 e4 ). At the same time
change affords the opponent add itional measures have to be taken against the
possibilities. seizu re of the centre by 1 9 lL'lf3 or 1 9 �e5
followed by 20 f4 (20 lL'lf3).
1 5 cxd3 �d7
18 . . . �b5!
A cou nterattack against the d3-point: 19 i.e5
'i¥f5 or 1 9 lL'lf3 'i¥f5 20 lL'le5 i.d4.
Take note: the two players are engaged in a
strateg ic war, but the means employed are
purely tactical - concrete strong moves,
short variations, th reats, double attacks . . . In
chess, tactics and strategy are inseparably
linked - deficiencies in either of these fields
will i nevitably tell on you r qual ity of play and
resu lts.
19 i.d6 l:tf7
20 f4 i.d4!
Whose Strategy will triumph? tD 21 9
B lack has defended against 2 1 �e5 (after not possible (the rook on f1 is hanging ) , and
which 21 . . . �xe5 22 l:txe5 'ii'x e5 was i ntend 22 ltJxd4 'ii'x d4 23 .l:!.xe6 iixd3 24 'ii' x d3
ed ). Less accu rate was 20 . . . 'ii'f5 21 .l:tf3 �xd3 leads to a somewhat better ending for
followed by �e5 or r!e5. B lack.
It was necessary to reckon with 22 llt'b4 . I n
the event o f 22 . . . �xd3?! 23 ltJxd4 i.xf1
(23 . . . .l:txd6? 24 'il'xd6 �xf1 25 ttJxe6 or 25
.l:.xe6) 24 �e5! �xg2+ 25 �xg2 the advan
tage is with Wh ite - his pieces are very
strongly placed . However, the cool-headed
22 . . . ..txb2 ! would have changed the pictu re ,
for example: 23 .l:t b 1 �xd3 24 l:Ixb2 :Xd6 , or
23 d4 ..txe2 24 ..te5 �e7 ! 25 'it'xe7 l:txe7 26
l:Ixe2 �a3. There is also 23 .l:tf3 , which was
later analysed by grandmaster Patrick Wolff.
In the event of 23 . . . �c6? Wh ite gains an
advantage by 24 d4! a5 25 'ii'c 5 a4 26 a 3 ! ,
b u t t h e two other possible replies, 23 . . . �a6! ?
and 23 . . . a5!?, are sufficient t o parry the
Now the culmi nati ng point of the fierce battle
opponent's aggression .
for the central squares has been reached . By
attacking the b2-pawn, I try to divert my 22 �a3 �b6
opponent from the natu ral knight move to f3 . I breathe a sigh of relief. The wh ite pieces
Even so, it deserved serious consideration . have been d riven away from the key e5-
True, after 2 1 lDf3 !? �xb2 noth ing is g iven sq uare, which means that I have won the
by 22 .l:tb 1 �c3, and 22 d4? �xf1 22 l:txf1 strategic battle. But, of cou rse, not yet the
does not work, as this can be strongly game.
answered by either 22 . . . �xd4 !? 23 ii'xd4 (23 Grandmaster Milan Matulovic used to pay
ltJxd4 e5! 24 tiJb5 a6 25 ttJc7 .l:td8) 23 .. .'iVxd4 enormous attention to opening theory. It is
24 ltJxd4 e5! (P.Wolff), or 22 . . . �d8!? 23 �b4 said that he kept a special score of the
a5! 24 �xa5 'i!ka8 . However, the simple 22 outcome of the open ing duels i n his own
�e5! �xe5 23 ttJxe5 llff8 (23 . . . .l:tc7? 24 games, and, to his delight, more often than
'ii'a 5) 24 'iib4 or 24 'ii'a 5 would have secu red not he was successfu l . B ut things were much
Wh ite excellent positional compensation for worse as regards overall successes in
the sacrificed pawn . tournaments. For me it was not enough to
21 ttJe2 win an intermediate stage - I also wanted to
Not the best square for the knig ht, but Wh ite win the race as a whole. And for this it was
has a specific idea . Here 2 1 . . . �xb2? 22 lib1 necessary, without relaxing , to carry on
is now bad for B lack, while if 21 . . . ..tb6 there worki ng.
follows 22 �e5 ii'f5 23 .l:.f3 , and the battle for 23 ttJc3 �c6
the centre concludes not in my favour. 24 'ili'e2
21 . . . l:td8! A new problem. The e-pawn is attacked , and
An important intermediate move. It would if Black defends it by 24 . . . .l:te8 , then 25 �d6
have been a mistake to attack the bishop - the bishop returns to the e5-square and
with the other rook: 21 . . . l:Id7? 22 �e5! �xe5 memories alone are all that remain of my
23 fxe5 'ili'xe5? 24 ttJc3. But now 22 �e5 is positional achievements.
220 � Whose Strategy will triumph?
40 .l:!.xd3 i.. d 6
41 h3 .l:!.a8
Here the game was adjourned , and White
sealed his next move. I am a sound pawn to
the good , and in add ition I have the
advantage of the two bishops. It appears that
the win is a matter of straightforward
tech nique. At any event, that is what I
assumed at the adjourn ment. Alas, analysis
did not confirm this evaluation - for a long
time I was unable to fi nd a convincing plan
lead ing to a wi n .
4 2 g4 .l:!.a2
- position after 36 lLle2 -
has the two bishops and more space). to l u re the pawn to e5 in order later to attack
However, Black has no real weaknesses and it by . . . d7-d6 or . . . f7-f6 .
in the end I gained a d raw, although it will be 8 tt::l g 3 tt::l c 6
realised that there is little pleasure in 9 i.d3 ..ta6
defending such a position .
10 e4 tt::l a 5
Events developed in more lively fashion in 11 'ife2
the game Vaisser-Dvoretsky (Kiev 1 970):
The same pawn structu re has been reached
6 .. .f5 7 a3 i.xc3+ 8 tt::l x c3 tt::lx c3 9 'iVxc3 i.b7
as in the Samisch Variation of the N imzo
1 0 d5! 0-0 ( 1 0 . . .'ife7 is better, intending after
l ndian Defence (4 a3). There Black is left
1 1 dxe6 dxe6 to play . . . tt::ld 7 and castle on
with his knight on f6 , whereas here he has
the queenside) 1 1 b4! (an excel lent position
his dark-sq uare bishop, which , generally
al pawn sacrifice) 1 1 . . . exd5 1 2 i.b2 .l:tf7 .
speaki ng, is a slig htly stronger piece. But on
Let's try to decide how Wh ite should develop
the other hand I have lost several tempi in
his in itiative.
the opening ( . . . tt::lf6-e4xc3, . . . i.b4-e7), and
in addition the opponent has managed
without the move a2 a3, and so Black has not
succeeded in creating cou nterplay against
the c4-pawn .
11 . . . 0-0?!
The king will not feel too comfortable on the
kingside. In the game I . Sokolov-Johansen
(Olympiad , Manila 1 992) Black retained an
acceptable position after 1 1 . . . d6 12 0-0 'ii'd 7
1 3 I:t b 1 ( 1 3 a4! ? ) 1 3 . . . h5. A recommendation
by Alexey S h i rov also deserved considera
tion : 1 1 . . . c6 ! ? (Black prepa res 1 2 . . . d5) 1 2 e5
d5 1 3 exd6 'i¥xd6 14 tt::l e4 'ii'd 7.
12 0-0 d5?!
There followed 1 3 .ll d 1 ? ! dxc4 1 4 ..txc4 d5
15 0-0 c6 16 b5 cxb5 1 7 i.xb5 tt::l d 7 18 f3 I n the event of 12 . . . c5 1 3 d 5 e5 14 f4 (or 1 4
l:i.c8 1 9 'it'd4 'iff6 with roughly equal chanc tt::lf5) Black's position i s clearly worse. 1
es. 1 3 0-0-0! was much stronger: 1 3 . . . dxc4 should probably have tried 1 2 . . . g6!? 1 3 i.h6
(otherwise 1 4 c�d5 with an obvious advan l:i.e8 , subseq uently choosing between . . . d7-
tage) 14 i.xc4 d5 1 5 b5! , and the threat of 1 6 d5 and . . . f7-f5 .
l:.xd5! i.xd5 1 7 .l:td 1 is extremely u npleas 1 3 cxd5 i.xd3
ant. 14 'ii'x d3 exd5
6. . . tt::l x c3
(see diagram)
The game Meulders-Winants (Belgian Cham
pionship 1 983) went 6 . . . ..txc3+ 7 bxc3 tt::ld 6
Wh ite's plan is obvious: e4-e5 and f3-f4-f5,
8 tt::lg 3 i.a6 9 'i!i'a4 'ii' h 4 1 0 i.d3 tt::l f5 11 i.xf5
creating a powerfu l attack on the kingside.
exf5 1 2 0-0 'ii'f6 1 3 e4 1kc6 1 4 'ika3 ! 'ifxc4 1 5
1 5 tt::l f5 also looks q u ite good . What do you
.l:te 1 , and the black king was in g reat danger.
th ink, which move is stronger? To answer
7 bxc3 ..te7 this q uestion you must also take Black's
7 . . . i.d6! ? 8 e4 i.a6 was interestin g , aiming cou nter-actions into account.
226 � Whose Strategy will triumph?
15 eS?I 'ii'd 7
1 6 lLlf5
Practically forced - since otherwise I would
have played 16 .. .f5 ! , halting the opponent's
attack. Now it is clear that he should have
begun with 1 5 lLlf5 ! , since this would have
created the add itional th reat of 1 6 lLlxe7+
'ii'x e7 1 7 exd5. I would have had to make
some not very useful move such as 1 5 . . . c6 , I n the event of 1 8 f4 f5 1 9 exf6 .i.xf6 20 f5 I
and then the e4-e5 advance would have still stand worse , but rather freer than before.
gained in strength . It is desirable to prevent . . . f7-f5, but if 1 8 g4
16 . . . g6!? there follows 1 8 . . .f6 .
Wh ite wants to play f3-f4 , then lLlxe7+ and After t h e g a m e Mikhail Tal suggested 1 8
f4-f5, and to crush me with these pawns. .i.f4 ! ? . A q uestion for you : how should the
Therefore I decided to weaken my kingside, reply 1 8 . . .f5 be evaluated?
in order to drive the knight from f5 and meet After any other move by me ( 1 8 . . . lLlc4, for
the opponent's attack with . . . f7-f5 ! . example) there follows 1 9 .l:.ae 1 , and then
We see that, a s in the previous game, a the bishop retreats, renewing the th reat of
battle of plans has developed : Wh ite d reams f3-f4-f5 . The knight on c4 is prettily placed ,
of obtaining a powerful pawn pair at e5 and but in fact it is not much use - here it does not
f5, wh ile Black tries to disrupt this plan by have the slightest influence on the defence
advancing his f-pawn , after which he will be of the kingside. I n the positional sense 1 8 . . .f5
able to breathe more easily. But whereas i n is, of cou rse, the correct reactio n , but in reply
the game with Lein I was an equal partner, Wh ite has the possibil ity of a clever combina
here my opponent has far more chances of tion .
success. However, Wh ite also has his 1 9 e 6 'ii'x e6 20 :ae 1 'ii'd 7 2 1 l:txe 7! 'iixe 7 22
problems - on every move he has several .i.e5+ l:tf6 23 lLlg4! (23 g4? lLlc4 24 g5
tempting possibilities, and it is not at all easy lLlxe5) 23 . . . fxg4 24 fxg4 - White regains his
to make the correct choice. rook and would appear to gain the advan
1 7 lLlh6+ tage.
After 1 7 lLlxe 7 + 'ii'xe 7 1 8 .i.h6 .:tea noth ing is If you r opponent has devised something,
Whose Strategy will triumph? l2J 227
knight will be stronger than the opponent's 33 'Ot>xh3 'it>h7, transposing i nto the equal
bishop. position which occurred i n the variation with
Wh ite should have simply played 22 lt:Jf2 ! 28 f4 .
(but not 22 lt:Je3? g5), and then prepared g3- 30 . . . 'ifxg4+!
g4. 3 1 'it'xg4 lt:Je3+
22 . . . �xf6 32 '>t>f3
23 exf6 lt:Jc4 32 ;t>g3 lt:Jxg4 33 f5 �xe5 (33 . . . .l::i.xf6!?) 34
At last the time has come to place the knight dxe5 �xe5 35 fxg6 fxg6 was hopeless for
on its lawful square. White.
24 �e5 c6 32 . . . lt:Jxg4
The queen must be freed from the need to 33 .l::i.a g1
guard the c7-pawn . What would you now play as Black?
25 '>t>g2 'iif5!
26 �e2
26 'iixf5 gxf5 27 'it>h3 'lt>h7 28 'it>h4 Wg6 29
.l::i.e2 .l::i. e6 30 .l::i. a e1 would have led to an
unclear ending.
26 . . . .l::i. e 6
I have already begun to take an interest in
the f6-pawn : I am i ntending 27 . . . l::i.fe8 28 f4
lt:Jxe5 and 29 . . . 'ir'xf6 (or 28 . . . Wg8 with the
th reat of 29 . . . 'iixf6). Therefore Wh ite hurries
to open lines.
27 g4 l!fg5
28 h4
The start of a forci ng variation, at the end of The exchange of minor pieces by 33 . . .
which my opponent overlooked a small lt:Jxe5+?! 3 4 dxe5 looks prematu re; also
tactical subtlety. The balance would have dubious is 33 . . . lt:Jxf6?! 34 f5! .l:!.xe5 35 dxe5
been maintai ned after 28 f4 'iVxg4+ 29 �xg4 .l:!.xe5 36 fxg6 . But after 33 . . . lt:Jh6! 34 .Ug5
hxg4 30 h3 (30 'lt>g3!?) 30 . . . gxh3+ 3 1 'it>xh3 ;t>h? Wh ite's attack would have come to an
'lt>h7! 32 l::i. h 1 M.h8 33 'lt>g3+ 'lt>g8 34 .l::i. x h8+ end and Black should have calmly begun the
'it>xh8 35 f5! gxf5 36 'it>f4 . conversion of his two extra pawns ( . . . c6-c5
28 . . . 'i!Vxh4 etc . ) . U nfortunately, here the same story
29 f4! occurred as in my game against Lei n . Having
29 .l::i. h 1 was poi ntless in view of 29 . . . 'iig 5. gained a winning position after g reat emo
.l::i.fe8 tional stress, I relaxed and quickly made a
29 . . .
superficial move with my king , underestimat
30 .Uh 1 ?
ing the exchange sacrifice which my oppo
Shamkovich is still i n the grip of a n illusion . nent had prepared .
Of course, 30 gxh5? gxh5 or 30 f5? .l::i.x e5 3 1
33 . . . ;t>g8?
dxe5 .l::i.x e5 was unfavourable for h i m , b u t it
was better to play 30 'iff3 hxg4 31 'ifg3 34 .l:!.xg4! hxg4+
'iih 3+! 32 'ifxh3+ (32 Wf2 lt:Jd2 33 'ifxh3+ 35 'it>xg4
gxh3 34 .l::i. h 1 lt:Je4+ and 35 . . . lt:Jxf6) 32 ... gxh3+ It is staggering but, deep in the endgame and
Whose Strategy will tri u m p h ? Qj 229
the exchange and a pawn down , Wh ite still .l:i.b6! are easy and pleasant to calculate.
retains an attack. He wants to prepare f4-f5 41 . . . .l:i.e1 !
and pursue my king with his rook. Of course,
42 .l:i.b3 .l:i.e4
here 36 f5 is not yet a th reat in view of
36 . . . .l:i.xe5 (it is Black's dream to return the Here the game was adjourned . Wh ite sealed
exchange at an appropriate moment), but the move 43 i.. c 5 and then resig ned without
soon such a threat will become real . It is not resuming . I was intending 43 . . . .l:i.f1 ! (43 . . . d4? !
possible for Black to strengthen his kingside is weaker in view of 44 .l:i.a3 ! ) 44 i.. d 6 .l:i.g 1 +
defences - all that remains is a cou nter 45 'it>h4 (45 'it>h6 .l:i.e2 or 45 . . . .l:i.ee 1 ) 45 . . . g5+.
attack on the queenside.
We have analysed two tense games. I n both
35 . . . c5 of them the outcome depended on the
36 �g5 correct perception by the two players of the
If now 36 . . . cxd4 37 cxd4 .l:i.c6 , then 38 f5 ! strategic problems facing them, but at the
(since Black no longer has a double captu re same time on their tactical resou rcefu l ness,
on e5) 38 . . . .l:i.c2 39 .l:i.g 1 ! with an u nclear without which it is not possible to successful
position. I found another idea - I tried to ly put your plans into effect, and later also on
undermine Wh ite's pawn chai n , to weaken their endgame tech nique. Good g rounds for
the defence of the bishop on e5. once more thinking a bout the need for every
36 . . . b5! player to rid himself of his playing weakness
37 a3 a5 es and to ach ieve harmony in his play, by
developing and perfecting all aspects of his
mastery.
What other conclusions should be drawn
from these games? Remember: in each of
them there came a moment when it should
have become clear to Wh ite that he would be
unable to ach ieve complete strategic suc
cess - the opponent had sufficient cou nter
chances. I n such situations it is important to
display a sober and flexible approach , to be
able to cal l a halt a n d , giving up your
u n realisable aims, seek a way to achieve an
acceptable, relatively safe position. Neither
of my opponents coped with this problem .
38 .l:i.b1 ? Finally, as we have see n , the loss of a
strategic battle by no means sig n ifies the
The decisive mistake , after which Black's
loss of the game. Nearly always there remain
idea triumphs. Meanwhile, after 38 �h6 or
sufficient opportu n ities to make l ife difficult
38 .l:i.e 1 followed by f4-f5 the win for me
for the opponent, by continually setting him
would still have been very much i n doubt.
new problems.
38 . . . b4
And, on the contrary, after outplayi ng your
39 axb4 axb4
opponent you must not relax, but continue
40 cxb4 cxd4 playing at full i ntensity. Otherwise you risk
41 i.. x d4 missing a deserved win , as almost occu rred
Variations such as 41 b5 d3 42 b6 d2 43 b7 with me in both of these games.
230 �
PART V
The situation has changed sharply. Black I n this example too , both players made
has been left with a 'bad' bishop. Wh ite mistakes. Here Black should have continued
should activate her knight and bring it to d4. his development with 1 4 . . ."it'a5 followed by
I n the game she chose the i ncorrect route. . . Jk8 and . . . ..tc5 or . . . ..tb4. I n stead of this
24 tL:lg3? Ji.d3 Baklan offered the exchange of q ueens,
which spoiled his pawn structu re.
The knight did not in fact reach its goal.
24 tL:ld2 ! suggested itself. If 24 . . . �e8 25 �e 1 14 . . . 'ir"b6?
f6 26 tL:lf3 ..te4 , then 27 exf6 'it'xf6 28 'i!kxf6 1 5 'ir"xb6 axb6
gxf6 29 tL:ld4 'i.t'f7 (the more tenacious 1 6 ..te2
29 . . . .l:!.e5 is also u npromising : 30 f3 ..tg6 3 1 The simple 1 6 tL:lb5 would have led to an
'.t>f2) 3 0 f3 Ji. d 3 3 1 .l:i.xe8 'i.t'xe8 3 2 tL:lxb5! and advantage for Wh ite.
wins. As Dvoretsky pointed out, instead of 16 . . . 0-0
26 . . . ..te4 Black has the stronger 26 . . . ..tg4 !
1 7 ..tf3?
27 tL:ld4 fxe5! 28 'it'g5 exd4! (after 28 . . . 'iVg6
29 'i!kxg6 hxg6 30 tL:lc6 e4 31 f4 followed by Wh ite conti nues to play inaccu rately and
'i.t'f2-e3 Wh ite retains the better chances, without a pla n . He clearly does not under
despite the pawn deficit, in view of the stand the essence of the position and simply
control over the weakened dark squares) 29 makes 'solid ' moves. The bishop is doing
.l:!.xe6 ..txe6 30 cxd4 c3 , and the passed noth ing on f3 , whereas the thematic . . . b6-b5
pawn has to be blocked by the quee n , which advance should have been. prevented .
means that the position favou rs Black. 17 . . . �a5?!
Therefore it makes sense for Wh ite to play Of course, the i m med iate 1 7 . . . b5! was
not 25 lle 1 , but 25 tL:lf3 ! ? , and after 25 . . . ..tg4 stronger.
26 tL:ld4 'i!Vxe5 27 'i!Vxe5 �xe5 28 f3 Ji.d7 29 1 8 a3 b5
�f2 the excellent position of the knight on d4
19 tL:la2 .Ua4
fu lly compensates for the sacrificed pawn,
although not more. 20 g3 f6?!
As was shown by Dvoretsky, Black should
have played 20 . . . b4 ! , since if 21 axb4??
Arrangement of the pieces
there is 21 . . . �fa8.
Morozov - Baklan ( 1 2) 21 f5? !
Kiev 1 990 T h e preparatory 2 1 tL:lc3! .l::!. a 5 was more
accu rate , and only now 22 f5 .
21 . . . d4! ?
22 exf6 ..txf6
23 fxe6 ..txe6
24 ..txb7 ? !
A typical mistake : Wh ite is tempted by
material g a i n , underestimating the oppo
nent's cou nterplay. 24 lbb4! was sounder,
with the advantage.
24 . . . �b8
25 ..tf3 b4! ?
26 tL:lxb4 .Ubxb4
232 w From Games by Pupils of the School
Baklan (1 3) - S h iyanovsky
Kiev 1 99 1
33 . . . l:tb5?!
A tech nical inaccu racy. Before changing the
structu re , it is usefu l to improve your position
as much as possible, or worsen the oppo
nent's position. As shown by Dvoretsky,
Black should first have g iven a series of
checks: 33 . . . l:tb2+ 34 'it>f1 .l:tb1 + 35 �2
l:tb2+ 36 'it>g 1 , and only now played 36 . . . l:tb5
with equal ity.
34 l:txb5 axb5
I nstead of calmly strengthening his position 35 il.f2 ltJb7
by playing 20 b4! , then a2-a3, lld3, l:tad 1 36 'it>d2 h5
and at a conven ient moment c3-c4 , White 37 il.e3 g6
hu rries to exchange his active knight and 38 'it>c2 'it>d6
loses nearly all his advantage.
39 h4 ltJc5?
20 ltJb7? l:tab8 When offering to go i nto a pawn ending , you
21 ltJc5+ il.xc5 must calculate the variations very accu rately.
22 il.xc5 ltJcB I n this case Black got it wrong! 39 . . . ltJd8 was
23 f3 ltJd6 correct.
24 ltd3 ltJb7 40 il.xc5+ 'it>xc5
25 il.f2 .:thd8 41 g3 'it>b6
26 l:tad 1 l:.d6 42 f4 c5
27 'it>f1 l:!.bd8 43 a41 exf4
28 'it>e2 c5 44 gxf4 bxa4
29 b3 c4 45 'it>b2
30 bxc4 bxc4 And Wh ite soon won .
234 � From Games by Pupils of the School
would nevertheless have retained the better Meanwh ile, Diana Darchiya could have
chances by offering the exchange of the immediately decided the outcome with an
light-square bishops: 28 ..td3 ! , and if 28 . . . ..th6 other exchange: 25 'iVb3 ! . It is important to
29 �c1 . eliminate the q ueen , the defender of the e6-
27 . . . ..tf5? pawn , and thereby completely disorganise
28 l:1g5 l:.f7? ! the opponent's defences. Black has no
defence, for example: 25 .. ."ilkxb3 26 axb3
Black recog nises his mistake and tries to
i.d7 27 l:tf7 .
prepare 29 . . . i.h6 (if 30 l:th5 he has the reply
30 .. Jih7). However, he should have recon
ciled himself to the loss of a tempo and Rasted - Kadymova (1 5)
played 28 . . . i.h7 29 .l:.h5! 'ilg6 . Now Wh ite Germany 1 992
easily prevents the exchange of the dark
sq uare bishops and q uickly decides the
game with an attack on the g- and h-fi les.
29 l::. h 5+! �g8
30 l:tg1 �f8?!
3 1 ..tg5
Black resig ned .
Pawn structu re
Chekaev (1 3) - Goldaev
USSR 1 989
I ndex of O pe n i ngs