4. CONTENTS
List of Symbols 5
Bibliography 6
Foreword 7
1 The Chigorin Variation - General Aspects 9
2 The Rubinstein System 25
3 The Petrosian System 97
4 The Yates Variation 133
5 The Yates Variation - 10.d5 137
6 The Yates Variation - 10..te3 149
7 The d3-System 169
8 The Worrall Attack 189
9 The Delayed Exchange Variation 197
10 The a4-System 209
11 The System Based on ctJc3 217
12 The Central Attack 225
Index of Main Lines 232
5.
6. List of symbols
t Check
A strong move
!! A brilliant move
!? An interesting move
?! A dubious move
? A mistake
?? A blunder
D The only move
1-0 White won
V2-V2 The game was drawn
0-1 Black won
(ch) Championship
(izt) Interzonal
(01) Olympiad
(n) nrh match game
+- With a winning advantage for White
± With a large advantage for White
! With a small advantage for White
With equal play
<Xl
With unclear play
+ With a small advantage for Black
=+= With a large advantage for Black
-+ With a winning advantage for Black
00 With compensation for the sacrificed material
� With an attack
t With an initiative
+! With counterplay
o Better is
5 Worse is
i:::J. With the idea
C With a development advantage
X Weakness
<=> File
,;;< Diagonal
7. Bibliography
Die Moderne Schachpartie, Siegbert Tarrasch, Leipzig 1 924
Teoria Deschiderilor in sah (Chess Openings Theory) Voluille 1 -2, Paul Keres, Editura Tinetretului,
Bucharest 1 952
300 partide alese ale lui Alehin, Vasily Panov, Editura Tineretului, Bucharest 1 957
Matchi pretendentov na pervenstvo mira, B. Kajic, M. Yudovich, Belgrade 1 969
Vierspringerspiel bis Spanisch (Four Knights Game to Spanish), Paul Keres, Sportverlag Berlin, 1 976
Grossmeister Geller, Yakov Damsky, Fizkultura i Sport, Moscow 1 976
V ogoni ataki, Mikhail Tal and Yakov Damsky, Fizkultura i Sport, Moscow 1 978 (English Edition,
Attack with Mikhail Tal, Everyman Chess, London 1 994)
Korchnoi's 400 Best Games, Viktor Korchnoi, Robert Wade, L.S. Blackstock, Batsford, London
1 978
Akiba Rubinstein, Yuri Razuvaev and Valerii Murakhvery, Fizkultura i Sport, Moscow 1 980
Ispanskaia Partiia (The Spanish Game), Alexey Suetin, Fizkultura i Sport, Moscow 1 982
Ratmir Kholmov, Ratmir Kholmov, Fizkultura i Sport, Moscow 1 982
My System, Aron Nimzowitsch, Quality Chess 2007
Strategia nadejhnosti, Tigran Petrosian, Fizkultura i Sport, Moscow 1 985
Sto pobednikh partii, Arlatoly Karpov, Fizkultura i Sport, Moscow 1 984
Arlaliticheskie i kriticheskie rabotii, 1 -4, Mikhail Botvinnik, Fizkultura i Sport, Moscow 1 984-87
Strategiia Nadejhnosti, Tigran Petrosian, Fizkultura i Sport, Moscow 1 985
744 partii Bobby Fischera, (Two volumes), Arldrey Golubev and Leonid Gutzait, Roleg Limited,
Moscow 1 993
Akiba Rubinstein: Uncrowned King, John Donaldson and Nikolay Minev, International Chess
Enterprises, Seattle 1 994
Akiba Rubinstein: The Later Years, John Donaldson and Nikolay Minev, International Chess
Enterprises, Seattle 1 995
Mikhail Tal, Games 1 -4, Sergey Soloviov, Chess Stars, St Petersburg 1 994-96
The Road to the Top, Paul Keres andJohn Nunn, Batsford, London 1 996
The Quest for Perfection, Paul Keres andJohn Nunn, Batsford, London 1 997
Boris Spassky, (Two volumes), Nikolay Krogius, Andrey Golubev and Leonid Gutzait, Moscow
2000
My Great Predecessors, Volumes I-V, Garry Kasparov, Everyman Chess, London 2003-2006
Chess Informant, Belgrade 1 965-2006
New in Chess Yearbook, Alkmaar 1 987-2006
Schaknytt, Gothenburg 1 999-2006
Chess Extrapress, Bucharest 1 998-2001
64 Shakhmatnoe obozrenie, Moscow 1 976-81
8. Foreword
In Beating the Open Games I started building a complete repertoire for Black based on 1 ...e5 as an
answer to 1.e4, but space limitations forced me to leave the story unfinished halfway. The present book
is a logical sequel ofmy previous work by providing Blackwith a coherent system ofdevelopment after
l.e4 e5 2.lLlf3 lLlc6 3.ib5 a6 4.ia4 lLlf6.
Since these two books can be considered as complementary parts of the same whole, my general
approach has been the same in manyways, but there are some significant differences, too.
I have structured each chapter in the same way as in the previous book. The introductory part deals
with the general aspects as well as with the historical evolution ofthe variation. The theoretical section
contains detailed analysis, which aims to be one or two steps ahead ofthe current stage oftheory.
Throughout the book, I have put a strong emphasis on the notion ofdevelopment, which offers the
most reliable guidelines in any open game. It is easier and more constructive to followjust a few general
recommendations about development than remember a large mess of variations. At the same time,
rules tend to have many exceptions, which frequently require combining common sense with concrete
thinking. I have aimed to highlight and explain the most important situations where this latter aspect
becomes vital.
The main difference compared to the general structure of most other opening books consists of
reversing the natural order ofthe chapters. Usually, booksstartwith the sidelines and examine the main
systems at the end. To a certain extent, this is the approach I had chosen for Beating the Open Games,
too. I have examined White's alternatives to 2.lLlf3 first, then the deviations from 3.ib5 and only in
the end the Exchange Variation. The need for a formal line-up was justified by the fact that each ofthe
openings examined there had its own individuality, with more or less equal rights with the others. Once
we have entered the territory of the Spanish Game, the situation becomes different. Each of White's
attempts to deviate from the main stream implies a small concession and simplifies Black's defensive
task. The main line is by far the most complex from a strategic point ofview. Once its basic ideas have
been assimilated, the other variations become easier to understand. This aspect has convinced me to
arrange the systems in what I consider an order ofdecreasing strategic complexity. Admittedly, a certain
amount of subjectivity was involved, but I believe that the general structure is quite practical: when
examining certain "side-lines" I have frequently made reference to positions from previous chapters. For
the same reasons, the first chapters are also the longest. In addition, certain variations are unpopular
and inoffensive at the same time, which makes detailed analysis unnecessary.
Although such openings as the Scotch, the Italian Game, and the Four Knights Game have become
increasingly popular over the past few decades and require from Black a relatively high degree of
accuracy, the Ruy Lopez remains unchallenged in its privileged position ofqueen of the open games. .
This is not just a consequence of respect for tradition, but instead is due to the rich strategic content
of the fight that characterizes the Ruy Lopez. Mastering the subtleties of this eternally young opening
is synonymous with being a strong positional player. I hope that this book will help the reader make
progress on this territory.
9. 8 A Spanish Repertoire for Black
Finally, I want to express my gratitude to:
Valentin Stoica, who helped me to prepare the theoretical material for the Yates system, the Worrall
Attack, the Delayed Exchange variation and the d3-systems;
Jacob Aagaard, who suggested possible ways ofimproving the initial versions ofthe nrst chapters;
My wife, Luiza, for her patient and kind support.
Bucharest, 1" ofJanuary 2007.
10. Chapter 1
The Chigorin Variation - General
Aspects
l.e4 e5 2.liJa liJc6 3..tb5 a6 4..ta4
This is where the story ofour book starts. We
have just stepped into the terricory of the Ruy
Lopez, the handling of which is uaditionally
considered co be a cornerstone of positional
understanding.
The next series of moves until 9.h3 has
occurred in thousands of games for many
decades, and is now played almost automatically
in most cases. However, I will make some short
comments on each move, as if it were my first
time seeing the position. This will hopefully offer
an overview ofthe significance ofeach chapter.
4. .. liJf6
Black takes advantage ofthe first opportunity
co counterattack the enemy centre.
5.0-0
White does not need to defend his e-pawn
yet, because after 5...ct:lxe4 6.d4 he would win
the pawn back, due the vulnerability of the
enemy king. Actually, this would be just the
start of a completely different scory titled the
Open Variation, which is beyond the scope of
this book.
Each move that defends the pawn has some
drawbacks.
5.ct:lc3 blocks the c-pawn and prevents the
natural plan of occupying the centre with c3
and d4.
5.�e2 develops the queen at a moment when
most of the other pieces find themselves on
their initial squares. This seriously contradicts
the general rules of development and is very
unladylike.
5.d3 is more flexible, but might result in a
loss oftime ifWhite later decides co occupy the
centre with c3 and d4.
5.d4 is a premature display ofcentral activity.
The placement of the bishop on the a4-e8
diagonal serves long-term purposes, by putting
the e5-pawn under pressure. IfWhite intended
to open the centre at such an early stage, he
should have developed his bishop on c4, in
order to get tactical threats against the f7-pawn.
This would have led to a completely different
opening, though.
5...i.e7
11. 10 A Spanish Repertoire for Black
By covering the e-file, Black renews the threat
of ...ctJxe4.
6.l3el
All the moves mentioned in the previous
comment are possible here, too, but they
would have the same drawbacks. Defending the
pawn with the rook is more efficient. After the
planned c3 and d4, White's major pieces would
keep the central files under permanent pressure,
preventing an early black counterattack.
6..ixc6 is White's last chance to avoid the
natural course of the game. After 6...dxc6 the
loss of tempo is justified by the fact that Black's
pieces are not optimally placed to defend the
e5-pawn, which will demand some accurate
play from him. However, giving up the light
squared bishop reduces White's strategic
potential considerably.
6...h5
Black parries the threat of txc6 followed by
lUxe5 that was created by White's last move.
7..ih3
Finally, the bishop takes the f7-pawn under
observation, but Black is well enough developed
to avoid any major trouble.
7...d6
Black over-defends his e-pawn and creates the
threat of...lUa5. At the same time, he opens the
c8-h3 diagonal for the bishop.
The alternative is 7...0-0, when play may just
transpose after 8.c3 d6. The sharp Marshall
Attack (8...d5) is not part of this book's subject
and, from the point of view of our main line,
the move order starting with 7...d6 is more
accurate. Castling on the 7rh move would leave
Black with some minor problems after 8.a4.
8.d
White clears the c2-square for his bishop and
prepares the occupation of the centre.
8.a4 would be less effective because of8...tg4,
creating the threats ...liJd4 or ...txf3 followed
by ...lUd4. After 9.c3 0-0 Black is ready to
question White's strategy with . . .lUa5 followed
by ...b4, when the weaknesses induced by the
early advance of the a-pawn leaves the Spanish
bishop rather exposed.
8...0-0
Now, the rushed 9.d4 would allow Black to
complete his development in a natural way with
9...tg4, putting the enemy centre under strong
pressure. "ilierefore, the prophylactic
9.h3
should be regarded as a more consistent
continuation.
Thus, we finally reach what can safely be
considered the main tabiya ofthe open games.
There are several reasons to claim this. First of
all, an examination ofall the possible deviations
clearly shows that Black's first move cannot be
easily challenged by an early and unprepared
action in the centre. (Some of these lines have
been examined in Beating the Open Games, while
the rest of them were reviewed above and will
be examined in this book at a later stage.) More
than one century of practice supports this point
ofview. Finally, we can note that contrary to the
situation in the so-called sidelines where Black
usually chooses between two or three reasonable
variations, in the diagrammed position countless
numbers of systems have been tried for Black:
the Breyer, the Smyslov, the Zaitsev, and then a
whole series ofChigorin set-ups connected with
names such as Rauzer, Panov, Keres, Romanishin,
Graf, as well as manyother lines without specific
names, but which have been part of the main
repertoire of great players, including world
champions, throughout chess history.
12. The Chigorin Variation - Gen�ral Aspects 1 1
This is hardly a sign that Black can equalize
however he wishes against the main line of
the Ruy Lopez; instead it is proof that against
each of these systems White has continually
found new ways ofkeeping his opponent under
positional pressure, regularly forcing Black to
come up with a new set-up. In fact, the wisest
approach for a player who mainly relies on the
closed lines of the Ruy Lopez with Black is to
periodically switch from one system to another
in order not only to avoid specific preparation
by one's opponents, but also to gain a wider
understanding ofchess in general.
In the diagrammed position, White is just
one step away from achieving his primary goal,
the stable occupation of the centre, but this
does not necessarily mean that he has won the
strategic battle yet. Since it is quite obvious that
Black cannot physically prevent 1 0.d4, he has
to look for an optimal way of meeting it. It is
hard to claim that any of the variations listed
above is better than another, which means that
the word "optimal" needs further explanation.
A player should choose Black's further system
of development in accordance with his general
level of understanding, style of play, personal
taste and, why not, with his general mood on
the particular day.
It goes without saying that my choice of the
repertoire systems contains a high degree of
subjectivity.
I intentionally avoided fashionable systems.
Experience has taught me that fashion is an
unpredictable and capricious lady; after certain
variations havebeen well-enough forgotten, they
might come back into the limelight. Secondly
(and more importantly I would say), the task
of catching the very essence of the position in
lines where theory advances with big steps (not
necessarily in the correct direction) is rather
difficult. It is much easier to take a photo or
sketch a portrait of a virtually immobile image
than to describe a highly animated scene.
Instead, I have preferred to choose variations
with a very long past, involving the names of
great players including world champions. This
will give us the opportunity of following the
evolution of thought processes through the
years. It is also supposed to lend some stability
to the theoretical conclusions given in the
following pages. Truths that have required
years or even decades to unfold completely
to human understanding, and involve names
like Rubinstein, Botvinnik, Keres, Smyslov,
Petrosian or Karpov will hardly ever be shaken
by practice or with the help of a computer.
I must confess that, apart from some rare
moments of fear that I would not manage to
make these "antiquities" viable, I have never
regretted my choice during the whole working
process. I rather felt as though I was drinking a
very old wine, discovered in a hidden corner of
my cellar.
Another aim of mine has been to make the
information useful in general, and not just
relevant to the specific variations. Ifsome ofthe
readers would like to make a choice of their own
against the main variation ofthe Ruy Lopez, the
strategic explanation given below should help
their orientation.
All these self-imposed restrictions left mewith
a relatively narrow domain. After some further
pondering and hesitations, I picked two of
the oldest sub-lines of the Chigorin Variation,
which, in its turn, is the oldest way of reacting
to the main line ofthe Ruy Lopez.
Strictly speaking, the move
9...�a5
which defines the Chigorin system, looks
like a small deviation from the logical course
of development. Indeed, Black moves for the
second time with an already developed piece,
while the c8-bishop is still on its initial square.
From this perspective, the more natural move
is 9....ib7, which in fact leads to the highly
fashionable Zaitsev Variation, a system that
endured a thorough examination during the
matches between Kasparov and Karpov.
However, there are certain elements of the
position that explain why the generally good
and logical move 9....ib7 is not necessarily
the only correct or at least the very best one.
13. 12 A Spanish Repertoire for Black
After 1 O.d4 Black cannot easily question White's
supremacy in the centre in the near future. The
main reason is that the c6-knight is tied to
the defence of the e5-pawn, thus blocking his
c-pawn, which under different circumstances
could be used to undermine the d4-pawn. At
the same time, it is uncertain yet whether the
development of the bishop to b7 is useful, in
view ofWhite's possibility of closing the centre
with d5 whenever he wishes.
In fact, virtually all Black's possible
continuations on the 9th move have minor
drawbacks, which, I repeat, makes the choice at
this stage a mere matter oftaste. I do not intend
to question the correctness of 9...ib7 or prove
the superiority of 9. . .lLla5 in any way, but aim
to explain that in this last phase ofdevelopment
concrete thinking can and should be tightly
connected with the appliance ofgeneral rules.
Let us now return to the Chigorin Variation.
lo.ic2 c5
Anticipating White's next move, Black
prepares to put up strong resistance in the
centre.
Il.d4flc7
This is the main tabiya of the Chigorin
variation. Othermoveshave been played (mainly
1 1 . ..ib7 and 1 1 ...lLld7), but in thevast majority
of games (about three-quarters) Black prefers to
defend the e5-pawn with the queen, maintaining
maximum flexibility for his position.
The term "flexibility" mainly refers to the fate
of the queenside minor pieces. Developing the
bishop to b7 now or slightly later would put the
white centre under immediate pressure, but the
simple advance of the d-pawn would solve this
problem, leaving the bishop terribly passive on
b7. Black can correct the bishop's placement
in a rather simple way, with a further ...ic8.
Although thiswould mean theloss oftwo whole
tempi, it would at least avoid irreparable damage
to the general harmony of the position.
The problem ofthe a5-knight is slightly more
demanding. If Black does not manage to find
a comfortable location, or at least a useful job
for it, he will most likely face insurmountable
strategic problems. Dr. Tarrasch's warning, "If
one piece stands badly, the whole position is
bad" applies perfectly here. And yet, it is not
easy for White to cut the knight out of play
completely. The hidden interactions between
pieces placed on different areas ofthe board can
lead to surprising results. To a certain extent; the
situation is similar to that arising in the Yugoslav
variation of the fianchetto King's Indian.
I believe that this latter aspect deserves a
small digression. We, modern chess players, are
accustomed to using the generic terminology
of a pawn structure typical of the King's Indian
Defence whenever White blocks the centre with
the strategically dreaded triangle c4-d5-e4.
But this structure can arise from several other
openings as well, including the Closed Ruy
Lopez. Historically speaking, the choice of
name is not entirely correct, because the Ruy
Lopez acquired coherent theoretical contours
decades earlier than the King's Indian. However,
for practical reasons I find it perfectly adequate.
There is no such typical Ruy Lopez-structure,
since the opening is much too complex and
flexible, while the aforementioned blocked
position almost defines the King's Indian.
Let us return to the issue ofthe a5-knight.
The next examples will illustrate the typical
problems facing Black if he fails to solve this
delicate matter in an adequate way. They are not
intended to discourage the reader from playing
14. The Chigorin Variation - General Aspects 1 3
the Chigorin line, but to offer a clearer image of
the kind ofpositions that should be avoided.
All fragments are taken from games where the
Ruy Lopez was played (although I was tempted
to insert some games with theYugoslavvariation
of the King's Indian Defence as well). In some
of them Black chose set-ups other than the
Chigorin variation, but play soon took a course
that is relevant for our central subject.
The critical situation arises when White
blocks the centre with d4-d5.
Having been deprived of the natural retreat
to c6, the a5-knight has only fWO ways of
regrouping.
From one point of view, the retreat to b7
is the most natural. The knight physically
approaches the rest of Black's army, although
this does not necessarily mean that the overall
coordination is improved yet. In order to solve
the problem, Black has to install his knight on
(5 (after a preliminary . . .c4 if the structure is
as in the previous diagram or immediately if
the c-pawns are missing as a consequence of an
earlier exchange on d4).
However, White has a strong remedy at his
disposal. By playing b4 (or b3 and if . . .c4,
then b4) he can prevent Black's plan with a
minimum of effort. The evaluation of the
position depends greatly on Black's ability to
use the relative weakness induced by b4 in order
to generate queenside counterplay. For instance,
if the aI-rook is undefended, after . . .a5 White
cannot maintain a pawn on b4, being forced
to give up the control of the c5-square. If
such rapid counterplay is not available, Black's
situation could become critical from a long
term perspective.
The following game fragment illustrates the
nature of Black's problems in its pure form. In
spite ofthe fact that almost all the other pieces
were exchanged, the knight's bad position on b7
was the main cause ofhis defeat.
Spassky - Kholmov
Soviet Championship, Yerevan 1 962
At a superficial glance it might look as ifBlack
had no problems at all. The queenside is safely
blocked, while on the other wing only Black can
start active operations. The knight's awkward
position seems to be temporary, because after
the standard ...f5 it could go to f7.
Unfortunately for Black, things are not that
simple. A deeper look reveals a remarkable
detail: the solidity of the queenside is
undermined precisely by the dreadfully placed
knight. The threat tDa3xb5 is very unpleasant,
because after ...axb5, a6 Black could not parry
.
the threats a7 and axb7 simultaneously. Black
has no time to regroup properly. If 30...tDd8,
then after 3 1 .tDa3 he is not in time to transfer
the bishop to the queenside in order to stop the
a-pawn. In fact, this is one of the indirect but
15. 1 4 A Spanish Repertoire for Black
very important consequences of the knight's
misplacement. Adequate regrouping of the
knight is possible in principle, but it requires
the temporary obstruction of the other pieces.
Since this aspect of poor coordination makes
itself felt in a position with just two pieces left,
you can imagine how much worse the situation
could be in a complex middlegame.
30...�d8
Not having much of a choice, Black starts
regrouping in the reversed move order.
31.lLld2! ?
One natural question is why Spassky refrained
from 3 1 .ltJa3. Apparently, Black has nothing
better than to continue his plan with 31...�c7?!,
hoping that after 32.ltJxb5 axb5 33.a6 �bS
34.axb7 �fS the king's rapid transfer to the
queenside would generate adequate counterplay
in spite of the predictable loss of the b5-pawn.
However, after 35.ltJd2 �e7 36.ltJb 1 �d7
37.ltJc3 �c7 3S.ltJxb5t �xb7 39.�fl @a6
40.ltJa3 Black cannot make any progress and
has to wait for the approach of the enemy king
and a slow execution. If he tries to activate
the bishop via a7-d4, the knight jump to c4
would force the king retreat to c7, with similar
consequences.
I assume that Spassky was more concerned
about the possibility of 3 1 ...ltJxa5 32.bxa5
ha5, solving the problem of the knight
radically. One important aspect is that the
f3-knight could not approach the queenside
easily, allowing Black to put his pawns in
motion with ...ib4, ...a5, etc. For instance,
33.�fl f5 34.@e2 i.b4 35.ltJc2 ic5 36.ltJd2
a5 and White has problems keeping both
wings under control. Black would combine
the threat of advancing the passed pawns with
a king infiltration on the opposite wing, with
reasonable drawing chances.
In a general sense, this is the kind of
counterplay Black should be striving for when
White tries to restrict the b7-knight with b4.
The sacrifice on a5 gives Black's pieces a chance
to work together instead of standing in each
other's way.
Spassky's last move, and next two, are clearly
aimed at inhibiting the knight sacrifice by
transferring his forces closer to the queenside. I
believe that his approach is very instructive. He
trusted his positional advantage, based on the
passivity of the b7-knight, more than the small
material gain which could have been obtained
by force after a hurried ltJa3.
31...f5! ?
A brave decision. With the knight isolated on
the queenside, this thematic pawn break risks
only weakening Black's kingside, offering White
additional targets.
3 1 ...ltJxa5?! 32.bxa5 i.xa5 would be less
effective now because the pair ofknights would
cooperate harmoniously in their fight against
the passed pawns, for instance: 33.ltJb3 i.b6
(Better than the pseudo-active 33...i.c3. After
34.@fl White would be threatening to almost
trap the bishop with @e2-d3. 34...a5 35.ltJa3
and Black has nothing better than transposing to
a sad ending with 35...a4 36.ltJxb5 ib2 37.ltJa5
a3 38.ltJxd6 a2 39.ltJb3 a1=�t 40.ltJxal i.xal,
when White should win.) 34.�fl f5 35.f3 �f6
36.@e2 h5 37.@d3 @g5 3S.ltJb4 a5 39.ltJc6 a4
40.ltJcl ic5 4 1 .ltJe2 White has stopped Black's
initiative on both wings and is ready to question
the stability ofhis fortress with ltJdS-b7 or -e6.
Continuing the regrouping initiated by ...i.d8
would also fail to save the game after 3 1 ...�c7?!
32.ltJa3 ibS 33.ltJdbl @fS 34.ltJxb5 axb5
16. The Chigorin Variation - General Aspects 1 5
35.a6 tLld8 36.0c3 followed by tLlxb5 and a7.
A more solid defence was 3 1 ...�f8. The king
will not be in time to stop the a-pawn in case of
the immediate 32.tLla3. However, the fact that
32...�e7 defends the d-pawn.would be essential
in the variation 33.tLlxb5 tLlxa5 when 34.tLlxd6
�xd6 35.bxa5 �c5 would not offer White any
winning chances because of the active position
ofthe black king. White could try 34.tLla7 tLlb7
35.tLlc6t c;t>e8 (avoiding a fork on b8) 36.tLlc4
but Black has chances to survive in a passive
position after 36...�c7 37.g3 tLld8 38.c;t>g2 �d7
finally exchanging the bad knight.
32.�fl �f6 33.�e2
33•..fxe4?
But this is unjustified panic. Black probably
feared the threat 34.exf5 gxf5 35.g4 followed
by tLle3, taking control of the light squares.
However, giving up the e4-square without any
fight is definitely worse.
Blocking the position with 33...f4 would
have deprived Black of the planned kingside
counterplay, which would make the knight
sacrifice on a5 less efficient after 34.tLla3. At
the same time, in the line 34...c;t>e7 35.tLlxb5
ILlxa5 36.lLlxd6 �xd6 37.bxa5 �c5 38.lLlf3 the
weakness of Black's kingside makes itself felt,
for instance 38...�d6 39.�d3 g5 (Necessary
prophylaxis. If 39...�xa5, then 40.tLlh4 winning
a pawn.) 40.tLld2 �c5 4 1 .tLlc4 �c7 42.g4 �b8
43.tLlb6 �c7 44.f3. Black cannot transpose to a
pawn ending because after the exchange ofthe d
pawn for the a-pawn, the e5-pawn would remain
decisively weak, while otherwise the knight
would get to f5 somehow, winning the h6-pawn.
The correct defence was 33...h5!, preventing the
undermining of the light squares with g4. Now,
consolidating the a5-pawn with 34.lLlb3 would
offer Black a vital tempo for his regrouping with
34...�c7 35.tLla3 �b8, preventing the sacrifice on
b5 and planning to improve the position of his
pieces furtherwith ...�a7 and ...tLlf7. Apparently,
White has nothing better than 34.tLla3 (instead of
34.lLlb3) when Black could react with 34...tLlxa5
35.bxa5 �xa5, since the knight is poorly placed
on a3 and will need another tempo to regroup,
while the kingside counterplay has reached a
relatively advanced stage.
34.tLlxe4t �e7 35.tLla3
It is all over now. With such astable knight on
e4, White can parry any sacrificial counterplay
rather easily.
35....L:a5
One concrete consequence of the knight's
presence on e4 can be seen after 35 ...�d7
36.tLlxb5 tLlxa5 when White has the simple
37.tLlbxd6+-.
Sacrificing the bishop marks a slight change
from the initial plans. 35...tLlxa5 would have
given up any hope ofexchanging the dominant
e4-knight.
36.bxa5 tLlxa5 37.tLlc2 tLlb3 38.tLlb4 a5
39.tLlc6t �d7 40.�d3 a4 41.�c3 tLlc5 42.8
and White soon won.
The analysis ofthis game has offered us an image
of the trouble that can be caused by a passive
b7-knight even in simplified positions, but has
also shown us the hidden defensive methods
available. We should also remember that both
sides' play was guided by the aim of restricting/
activating this knight, even if the latter would
imply sacrificing it.
The next example features a complex
middlegame, where Black has a good position
with just one exception: his knight is hopelessly
passive on b7.
17. 1 6 A Spanish Repertoire for Black
Geller- Meeking
Palma de Mallorca 1 970
Black has the better pawn structure and ifhis
queen's knight was placed on any reasonable
square (preferably d7) his chances would
probably be better. The way it is, he cannot take
advantage of the relative weakness of the d5-
pawn and of the a4-square and will not be able
to replace the f6-knight if it is exchanged by
lLlg4. Unlike in the previous game, there is no
way to activate or at least sacrifice the knight.
The further course of the game strongly
suggests that Black is simply lost, in spite of
the material equality and the complete lack of
weaknesses.
22.•.c3
Black faced the unpleasant threat of g4-g5
followed by a crushing attack. With his last
move, he hoped to get some activitywith .. .Wc4,
attacking the d5- and b4-pawns, but in the long
run this brave pawn will just be doomed.
23.lLlg4
By exchanging the f6-knight, White insures
his central pawn against any possible problems
and takes control of the e4-square.
23....ie7 24.lLlxf6t .ixf6 25.ge4
Now all Black can do is wait. White will
slowly improve his position, without any need
to hurry. The knight has just two passive squares
at its disposal.
25...�d7 26.�f3 ge7 27.h4 �e7 28.g3 lLld8
29.a3 gee8 30.gbl ge7 31.�e2 gb8
32.gb3 �d7 33.�f3 J.e7 34.ge3
White could have won the pawn now or on
the next move, but he probably wanted to avoid
any shadow of counterplay along the c-file,
eventually combined with a discovered attack
by means of ...e4.
34....if6 35J3e4 .ie7
36.g4! f6
The threat of a massive kingside attack forces
Black to restrict the activity of his bishop even
more. True, the knight will get an extra square
now, but it will remain just as passive as before.
36....ixh4? loses the bishop to 37.g5.
37.ge3 lLlf7 38.gbxc3
In the absence ofthe threat ...e4, the pressure
along the c-file will be easier to stand after the
capture of this pawn.
38...gbc8 39.ie4 .id8 40..id2 gc4 41.gxc4
gxc4 42.gc3 .ib6 43.gxc4 bxc4 44.g5
White has an extra pawn and a kingside
attack. Black resigned a few moves later.
This game clearlyproves that the bad location
of the queen's knight can be a decisive factor in
what would otherwise be a normal position:
a perfect illustration of Dr. Tarrasch's remark
about "bad" pieces.
We have examined two examples where the
knight was practically caged by the white b4- and
d5-pawns and, for a while, by its own f7-pawn.
18. The Chigorin Variation - General Aspects 1 7
The following game will prove that the knight is
unfavourably placed on b7 even in the absence
of the d5-pawn. Although this latter detail
would ensure some higher mobility for the
knight in the long run, the risk of interfering
with the other pieces' activity is high.
Spassky- Keres
Candidates Quarter-final, Riga 1 965
Black had exchanged twice on d4 in order to
activate his minor pieces, but failed to obtain
adequate counterplay. Then he moved his a5-
knight to b7, aiming to increase the pressure
against the e4-pawn, but probably missed the
simple b4.
All White's pieces are well placed, which
allows him to keep control of the position.
23...§'e7
Black clears the d8-square for the knight,
increasing the pressure against the e4-pawn at
(he same time.
The immediate 23...ltlxe4 24.�xe4 E1xe4
would have been dangerous with the knight
isolated on b7. For instance 25.ltld5 �xb2
(Alternatively, Black can maintain control of the
e-file with 25...E1cc4 but this would leave the
c-file available for White after 26.ixg7 @xg7
27.Ehe4 Elxe4 28.1!t1c3t @h6 29.1!t1c6, winning
back the pawn and retaining a positional
advantage.Blackcannotdefendthea6-pawnwith
29... 1!t1a8? because of 30.ltlc7 winning the rook.)
26.1!t1xb2 Ele6 27.1!t1d4 E1c6 28.l"lad1 and Black is
almost paralysed if not in some sort of zugzwang
already. With the knight far from the kingside, a
plan based on h4-h5 looks appropriate, although
rapid queenside incursions such as 1!t1a7 should
not be underestimated. These two possibilities
underline the factthat the black knight is passive
and vulnerable at the same time.
24.0 �f8?!
After this hesitant move the knight will never
get a chance to reach an adequate square.
Black should have initiated his regrouping
with 24...ltld8 immediately. One possible
continuation: 25.Elad1 ltlc6!? (Keres might have
overlooked this indirect defence of the d6-pawn,
noticing only that 25...l"lc6? loses material to
26.ltld5 ltlxd5 27.exd5 1!t1alt 28.@hl E1xelt
29.Elxel when the attempt to defend his hanging
pieces with 29...E1c8 would lead to mate after
30.E1e8t �f8 31.1!t1h6.) 26.ib3 (after the rushed
26.�xd6?! �xd6 27.l"lxd6 Black can ignore the
threat against the f6-knight and counterattack
with 27...ltlxb4!) 26...l"led8 27.ltld5 ltlxd5
28.hg7 @xg7 29.�xd5ltle5 30.f4ltlc4 3 1 .1!t1d4t
@g8 32.l"ld3 White's advantage is indisputable.
He has more space, better centralization and a
more compact pawn structure. However, Black
is in no immediate danger of losing, and what
is more important from the point of view of our
main theme, his knight has relative stability.
25.,ib3 etJd8 26J'!adl
White activates his pieces with natural moves,
while Black is clearly struggling.
19. IS A Spanish Repertoire for Black
26..Jk6
Now, 26...tLlc6 would come one move too
late. White could play 27.'&xd6, because after
27...'&xd6 2S.Elxd6 the light-squared bishop
would not be hanging anymore. Black can try
to take advantage of the relative vulnerability of
theother bishopwith 2S...tLlxe4, but this would
leave White a pleasant choice between 29.Elxc6
and 29.i.xf7t @xf7 30.Eld7t, with a material
advantage either way.
27J�c1!
Very elegant play. If tried one move earlier,
the occupation of the c-file could have been
adequately met by ...tLle6. Now the knight
prevents the natural connection of the rooks
by means of ...ElecS, which will result in White
taking control of the c-file.
27•••'Wie7
The queen intends to join the battle for the
c-file, but the presence of Her Majesty in the
centre will give White the opportunity to create
one-move threats.
28.<it>h2
By removing the king from gl, White
threatens 29.lLld5 tLlxd5 30.exd5 without
allowing the saving check on a7.
28•..'Wid7 29.llJdS llJxdS 30.hdS 13xc1
31.�xc1 'Wie7
The queen has to return, causing additional
loss of time, because the generally desirable
3 1 ...ltJe6? would lose an exchange to 32.i.c6.
32..hg7 <it>xg7
In spite of the simplifications, White's
advantage has not been diminished a single bit.
His control over the only open file will most
probablylead to the capture ofBlack's queenside
pawns. Having been obstructed by the knight
over the past few moves, Black's major pieces
are doubled on a half-open file, without any
prospects ofbreaking through, as the e4-pawn is
verywell defended. The knight finds itselfunder
the authoritarian domination of the bishop,
although there is some ray of hope in view of
the possibility of jumping to e6.
33.'Wic3t?!
Slightly inaccurate. It would have been better
to double the major pieces in the other order,
starting with 33.ECS. White threatens ElaS
(eventually combined with a check on c3).
33...tLle6 can be met by either 34.i.xe6 '&xe6
35.ElxeS '&xeS 36.'&xd6 with a more or less
winning queen ending, or 34.13c6, attacking the
d6-and a6-pawns simultaneously.
33...<it>g8 34.f4 llJe6 3S.g3
The knightseems tohaveimproved its position
quite a lot, but White's last move restrains its
activity within very narrow limits.
3S...llJg7?!
Keres' desire to activate his "bad" piece, any
way he can, is understandable; but White's
infiltration to c7 will prove decisive. Anywaiting
move would have been better, but Black's
position would have remained very passive.
36.'Wic7 'Wif6 37.13c2 13£8 38.'Wib6 gS 39.fxgS
'Wixg5 40.'Wixa6 'WieS 41.'WixbS
White collects two pawns before Black's
counterattack has even started.
41...tiJe6 42.'Wif1 �g7 43.'Wif5
1-0
We have formed a general picture of the
problems that can arise if the knight stays on
the a5-b7-dS path. The alternate route is tiJa5-
c4-b6-d7, which is more time-consuming, but
offers chances for a more natural re-integration
of the knight. However, problems could arise
here, too, because there are other black pieces
"competing" for the d7-square.
20. The Chigorin Variation - General Aspects 1 9
This is a typical example:
Kasparov - Grisehuk
World Cup of Rapid Chess, Cannes 2001
12...ltle4
From here, the knight hindersWhite's natural
development. Exchanging one pair of knights
with 1 3.ctJbd2 should be avoided because it
would ease Black's task of finding adequate
squares for his pieces.
13.a4 ltlb6
Black intends to transfer his queen's bishop to
d7 in order to consolidate the b5-pawn, which
requires that the a8-rook be defended. Since the
knight was not too stable on c4 anyway, this
willing retreat cannot be regarded as a significant
concession.
14.b3 .le8 15.VMe2 .ld7
We can now see the organic defect of Black's
regrouping. For concrete purposes (the defence
of the b5-pawn) the bishop is required on d7,
bur this cuts offthe natural retreat of the knight.
However, it might seem that Black has chances
for counterplay with ...c4, undermining White's
queenside structure.
16.e4!
Before starting active kingside operations,
Kasparov decides to clarifY matters on the other
wing. If Black exchangeson c4 or a4, the queen's
knight could go to the active c3-square later,
iustifYing the slight delay in its development.
16...b417.ltlbd2 g6
The ambitious plan initiated by this move
will fail precisely because of the unavailability
of the d7-square. Since the bishop is not doing
much on d7 anymore, 17...J.c8 deserved serious
attention.
18.ltlflltlh5 19.94 ltlf4?!
Consistent but bad. The more restrained
1 9...ctJg7 followed by ...J.c8 was better.
20..hf4 exf4 21.VMd2
Since 2 1 ...g5 would fatally weaken the bl
h7 diagonal, we can state that the knight's
jump to f4 is equivalent to sacrificing a pawn.
Under normal circumstances White's slightly
over-extended kingside structure, as well as the
relative weakness of the dark squares, should
offer Black sufficient compensation. If the d7-
square was available, he could play 2 1 ...ctJd7
22.Wxf4 �f6, intending ...ctJe5. 23."lWxd6?
would lose a piece to 23...ctJe5!.
However, the d7-square is occupied and Black
does not have sufficient means to block the e4-
pawn.
21....lf6 22.e5! dxe5 23.ltlxe5 oig5
Having failed to create counterplay on the
dark squares, Black decides to maintain material
equality.
24.ltlh2 a5 25.ltlhf3 .lh6 26J�adl
White has completed his development and
has a dangerous passed pawn in the centre.
Kasparov went on to win convincingly.
21. 20 A Spanish Repertoire for Black
I would like to extend the discussion about
the availability of the d7-square. The presence
of the bishop on d7 can interfere with the f6-
knight's activity as well. The next position is a
famous example from the practice ofKasparov's
predecessor to the supreme throne.
Karpov - Unzicker
Milan 1 975
Black has managed to gain the c5-square for
his knight, but has spent two whole tempi on
relatively useless moves on the other wing (...g6
and ...l"1e8) in the process.
17.f4exf4 18.hf4
From a strategic point ofview, the opening of
the f-file is a double-edged decision as it makes
the e5-square available to Black's minor pieces.
In this specific case the presence of the bishop
on d7 makes the quick transfer ofeither knight
to e5 impossible.
18....if8?
Black intends to transfer his bishop to g7,
but he obviously missed White's next move.
His problems of coordination caused by the
unavailability of the d7-square would also have
made themselves felt after the more consistent
lS...ttJc5. White could play 1 9.'lWd4 (PlanQing
to submit the f6-knight to strong pressure by
means of �fl and �e3. The attempt to take
advantage ofthe lack ofdefence ofthe e5-square
with 1 9.e5 would fail to 1 9...dxe5 20.�xe5 i.d6!
and suddenly all Black's pieces are in the right
places.) Black is not in time to regroup with
1 9...ic8 (?) followed by ...ttJc(f)d7 because the
temporarv lack of communication between his
rooks would allow 20.e5 dxe5 2 1 .ixe5 when
after 2 1 ...i.d6 22.ixf6 the capture on g3 would
be impossible. As in the previous game, the only
chance for retaining a playable position was the
immediate ls...ic8, planning ...ttJd7-e5, ...ttJc5
and ...if8-g7. With the d6-square defended by
several black pieces, the central break e5 is not
dangerous. Moreover, the immediate 19.e5? just
loses a pawn to 19 ...dxe5 because 20.ixe5 can
be answered by 20...i.c5t when White would
lose either his e5-bishop or the g3-knight.
19..ig5!
Wecan see now that ...g6 has weakened the f6-
square, while ... �e8 has blocked the emergency
retreat for the king's knight.
19...ie7
A sad necessity. 1 9...i.g7 would lose the
knight after 20.l"1fl 'lWdS 2 1.�f3.
20.'lWd2
The point behind Black's last move is that
after 20.�fl he would survive by means of
20...�c5t 2 1 .@h1 ttJxd5 when the g5-bishop
would be hanging.
20...ic8
After a loss oftwo tempi, this move can hardly
help. White's lead in development is too big.
20...ttJc5 would not have helped either
because of2 1.'lWf4 @g7 22.�fl �dS 23.�f2 with
the threat of �afl . If 23...�f8, defending the
f7-pawn in order to enable the knight's retreat,
then simply 24.ih6t wins the exchange.
21.�fl lLld7 22.lLlg4
White has obtained a massive attack against
the weakly defended black king. Threats such
as ttJh6t or �f2 are impossible to meet in an
adequate way. Obviously disgusted with his
position, Unzicker decided to call it a day.
1-0
In all the examples examined so far, Black has
been constantly striving to reroute his as-knight
via b7-c5 or c4-b6. The only open question is
22. The Chigorin Variation - General Aspects 2 1
whether the knight really stands badly on the
edge ofthe board. The following game will offer
some answers.
Tal - Gligoric
Leningrad Interzonal 1 973
Just as a cunoSlty, I will mention that the
knight reached a5 through what we could call
a reversed process compared to the Chigorin
variation. The Breyersystemwas played (9...'Llb8
1O.d4 'Llbd7) and Black's queenside position
started resembling that from the Kasparov
- Grischuk game, when the knight landed on
b6 and the bishop on d7. Gligoric reacted to a4
with ...bxa4 and after White's bxa4 moved his
knight to c4 then a5. In doing this, he probably
aimed to take advantage ofthe relativeweakness
of the c4- and b3-squares, and hoped that the
knight would have absolute stability on a5 in
the absence ofthe b-pawns.
Generally speaking, the whole plan is quite
coherent, but Black's development is not
sufficient to justify it concretely. With his queen
on c7 and the rook on a8, his position would be
entirely viable, but the way it is, he will not be
able to get his forces fully coordinated.
Tal's next move, which is quite typical of his
:endency to solve strategic problems by tactical
:neans, came as a shattering blow to Black's
dreams.
25.c4!
During his previous planning, Gligoric must
have considered this move "impossible". White
threatens Wc3 followed by �d2, which would
mark the start ofendless misery for the knight.
25...'Llxc4
25 ...h6 has been recommended, in order to
answer26.�xh6 with 26...'Llxc4. However, after
26.hf6 hf6 27.Wxh6 'Llxc4 28.�d3 'Lla5
29.h4 the weakness of Black's kingside offers
White excellent attacking prospects. Although
the black knight has avoided immediate
"personal" problems, it can hardly contribute to
the defence ofthe king in any way.
26.�d3 'Lla5 27.�xa6 .ic8 28.�a7
Black whole queenside is under enemy
occupation. The knight is in great danger and
restricts the mobility of Black's other pieces.
White's immediate threat is Elb8 and l"!xc8.
28...ge7
The only way to survive. After 28...'Llc4
29.�d3 the knight would not get anywhere.
29.�a8 �c7
Exchanging rooks with 29...l"1b7 30.l"1xb7
'Llxb7 would not help. Whitewouldplay3 1 .ctJd2,
threatening 'Llc4 followed bythe advance ofthea
pawn, practically forcing the rerum ofthe knight
with 3 1 ...ctJa5. Play could continue 32.�d3! .
(threatening ia6) 32...h6. This weakening move
is necessary in order to unpin the knight. 33.�e3
ctJd7 Now, 34.ia6 can be met by 34...'Llb6, but
34.�b5 creates the equally strong threat ofi.xd7.
23. 22 A Spanish Repertoire for Black
Black has to defend his queen with 34...if6 but
then 3S.!xh6 wins a pawn, maintaining the
advantage on the other wing.
30.tiJd2
Threatening 3 U '1bS tLlb7 32.tLlc4 followed
by aS-a6.
30...c4
Black prevents the blockade of his c-pawn,
but weakens the gl-a7 diagonal.
31..ie3 tiJd7 32J�b5 tiJb7 33.a5
A familiar picture. The advance ofthe a-pawn
is extremely unpleasant for Black if he has a
knight on b7, as we could see in some lines of
the game Spassky - Kholmov.
Tal won easily.
I hope that this series ofexamples has not been
too depressing for the reader. I also believe that
becoming acquainted with the worst that can
happen will help the reader understand Black's
general strategy in the variations examined in
the next two chapters.
However, I would like to end this chapterwith
a ray of optimism. The next game will reveal a
hidden facet of our main subject. The passive
placement ofthe knight can be compensated by
other positional factors, especiallyifall the pieces
are still on the board. If White concentrates all
his efforts against the as-knight, he may suffer
irreparable damage on other parts of the board,
tilting the balance in Black's favour.
Nijboer - Marin
Spanish team championship, Sanxenxo 2004
The queenside situation looks disastrous for
Black. The a5-knight's retreat is cut offby the b7-
bishop, which cannot regroup to d7 easilybecause
of the rook's presence on c8. However, Black has
some attacking chances on the other wing.
19...i.d8!
An important move. Black clears the seventh
rank in view of the queen's transfer to the
kingside and prepares the activation of the
bishop via b6. It is worth remembering that once
the centre is blocked, this bishop often becomes
Black's worst piece in the King's Indian, while in
this game it will play a decisive part in the final
phase of the attack.
20.b3
During our post-mortem analysis, Friso
Nijboer called this natural move the "decisive
mistake", probably having in mind that White
not only loses an important tempo but also
weakens the c3-square. The further course of
the game supports his point ofview to a certain
extent, but I would extend the criticism from
this specific move (which is probably not so bad
in itself) to the fact thatWhite did not resist the
hypnotic attraction of the knight placed on the
edge of the board.
20.ifl g5 2 1.tiJd3, exchanging the active
cS-knight and preparing the occupation of the
c-file, looks like a better plan.
20...g5
As we can see, the bishop is useful even from
the apparently passive dB-square.
21.i.f1 VfJg7
24. The Chigorin Variation - General Aspects 23
22.tLlc2
Consistently playing against the a5-knight,
White places his own knight on a rather passive
and exposed square. Even here, after having lost
a tempo, 22.lLld3 deserved serious attention.
22..J1l�g6!
This strong move, ignoring the threat b4
that required so much time from White to
be prepared, marks a turning point in the
game. Black attacks the e4-pawn (something
unthinkable in the previous game) and prepares
the advance ofthe h- and g-pawns.
23.tLld2
Unpleasant but necessary. White has to
defend his pawn although this means retreating
another developed piece. If23.b4?! lLlxe4 Black's
activity would be hard to keep under control.
For instance: 24.�d3 (Or 24.bxaS when apart
from the simple 24...�xaS, Black has 24...0xf2!
25.'�xf2l'hc2twhenWhite'spoorlycoordinated
army cannot survive the concentrated attack of
Black's pieces and pawns.) 24...lLlc3! 25.ixg6
(2S.�d2 loses to 25...e4) 2S ...lLlxdl 26.id3
Now the elegant 26...e4! provokes complete
chaos in White's territory, for instance 27.l'i:xe4
(27.ixe4 would be answered by 27...lLlc3, when
the c2-knight would be hanging) 27...lLlc4 with
complete domination. Ironically, neither of the
forked knights has been eliminated. On the
contrary, they feel quite comfortable in White's
territory.
23...h5 24..ie2
Trying to stop Black's attack at any cost.
Once again, 24.b4 would lead to tremendous
compensation after 24...lLla4 25.bxaS ixa5
when White's pieces are helplessly massed on
the first two ranks, or even 2S...lLlc3 26.�f3 g4
27.�d3 ha5 with a similar verdict.
24...g4!
Black cannot afford to block the kingside.
25.hxg4hxg426.i.xg4 tLld3
After four moves, the Sword of Damocles
(consisting of the threat b4) has been removed
and Black can carry out his attack under normal
circumstances.
27.'11l�f3! ?
The best practical chance. Accepting the
exchange sacrifice with 27.ixcs would lead
to a lethal attack after 27...ixcs. For instance
28.�f3 (the attempt to retain a material
advantage with 2S.l'i:fl would lose instantly to
2s...ig4 29.lLlf3 ih3) 2S...lLlxel 29.lLlxel ib6.
Friso rightly understood that his light-squared
bishop is one of the main defenders and should
nor to be given away for a rook. With his last
move he tries to block the f-pawn, preventing
an eventual sacrifice ...f3 which would clear the
f4-square for the knight.
27. •J3c3! ?
Finally, Black occupies the weak c3-square,
creating threats from both sides.
28J�fl ?!
After this move it is unlikely that White
can save the game. 2S.ie6t would also be
bad because of 2S...�g7, enabling the rook's
transfer to the h-file. The best practical chance
consisted of 28.�fl !, avoiding the dangerous
gl-a7 diagonal, as suggested by Friso the next
day.
28....ib6
With every move Black brings new forces
into the attack. The only piece that does not
have realistic prospects of joining its colleagues
remains the as-knight. However, we can easily'
notice now that White's efforts to isolate
this piece have cost him very dearly: several
queenside pieces are just as passive and do not
contribute to the defence in any way.
25. 24 A Spanish Repertoire for Black
29.i.f5 �g5 ?!
Blackcould havewon instantlywith 29...Wfg3!
which would enable not only the threat ...tLlxf2
but also the simple ...:1hc2, by defending the
c3-rook and the d3-knight in X-ray mode,
something not easy to spot during the game.
30.ctJbl
This awkward knight retreat is a desperate
attempt to complicate matters in time trouble.
30.Wfh3 would lose to 30...l"lxf5 3 1 .Wfxf5 (or
3 1 .exf5 tLlxf2 and White cannot save his queen)
31...Wfxf5 32.exf5 l"lxc2 with a decisive material
and positional advantage.
30...ctJe1!
Simplest. After winning the queen, Black's
attack will develop by itself.
31.ctJxe1 l"lxfJ 32.ctJxfJ �h5 33.ctJc3 i.c8
Exchanging the last possible defender of the
h-file.
34.hc8 :gxc8 35.i.d2 �f7 36.:gfdl :gh8
37.�f1 �hlt 38.�e2
38.tLlgl is no better because of 38...l"lh2-+.
38...�xg2 39.i.el :gh3 40.:gdJ
White managed to escape his time trouble,
and I still had some time to make my last move
before the time control. I very much like to
consider chess a logical and harmonious game,
which made me rather unenthusiastic about the
possibility ofwinning the game with the knight
still on a5. Therefore, I tried to make 40...tLlc4
work, but soon understood that it would just
prolong the fight in view of4 1 .tLlg5t.
Finally, I resigned myselfto the bitter thought
that I might never understand the hidden logic
of chess, and decided to call it a day by playing
the trivial
40...i.e3!
White's last point of resistance (f3) has been
undermined, which caused my opponent's
immediate resignation.
0-1
26. Chapter 2
The Rubinstein System
Though the group of systems of development
introduced by 9...ltJa5 bears Chigorin's name,
the first great player who seems to have settled
on a coherent middlegame plan and regularly
employed it in practice was Akiba Rubinstein.
There is a small paradox regarding the place
occupied by Rubinstein in opening theory. In
his career he invented so many major opening
'ariations (most of them still fashionable
nowadays) that it would be simply too
monotonous (and confusing) to name all of
them after him.
Nevertheless, considering the limited part of
the chess universe that is covered within this
book, I believe that it will do no harm to baptize
,he variation examined in this chapter after its
inventor.
The first recorded game where Rubinstein
-:xperimented with it was
Bernstein - Rubinstein
Ostend 1 907
The game took place in the sixth round ofa 28-
�ound marathon, which ended in a shared win
�-or... Rubinstein and Bernstein! We can see that
Rubinstein chose to rehearse his invention in a
§:ame of great sporting importance. Mter the
:!1uoductory moves
l.e4 e5 2.tLla tLlc6 3.i.b5 a6 4.i..a4 tLlf6
5.0-0 i.e7 6.:Sel b5 7.ib3 d6 8.d 0-0 9.h3
tLla5 1O.i.c2 c5 1l.d4 V!!!c7 12.tLlbd2
Rubinstein retreated his knight with
12...tLlc6
By putting additional pressure on the enemy
centre, Black invites his opponent to release the
tension as soon as possible. Besides, he declares
his intention to immediately correct the position
ofhis knight, although the concreteway he will
achieve that might seem not entirely clear yet.
13.d5
The most natural reply, which has preserved
its relevance up to the present. Apparently, the
black knight has no favourable squares at his
disposal.
13...tLld8! ?
27. 26 A Spanish Repertoire for Black
Bur of all the possible retreats, this certainly
looks like the most awkward! I wonder what
Bernstein thought at this stage ofthe game...
It should be said that the seemingly most
flexible answer 1 3...tt:lbS, with the intention of
re-developing to the natural d7-square, has the
drawback of denying the bS-square to the rook
for just one move, which can lead to problems
ofcoordination after the accurate 1 4.a4.
14.tt:lfl �e8
Only this new knight retreat to the back rank
casts some light on Black's further intentions.
By clearing the way for the f-pawn, he prepares
the transfer of the problem knight to f7, while
its colleague wiU land on g7.
Commenting on a later game in which
Rubinstein reached the diagrammed position,
Lasker made the following clever remark:
"According to Philidor: the pawns in front,
the pieces backing them up!" I would add:
According to Rubinstein's inclination towards
long strategic plans, too.
If compared with the King's Indian, we can
easily understand that the dark-squared bishop
is better placed on e7, by not interfering with
the knight's intended regrouping in any way.
Digressing a bit, we can also understand the
creation ofthe so-called Czech (or Old) Benoni
( l.d4 tt:lf6 2.c4 c5 3.d5 e5 4.tt:lc3 d6 5.e4 i.e7)
as an intended improved version of the closed
variations of the King's Indian. One of Black's
main ideas behind developing the bishop on e7
is to regroup with ...tt:le8, ...g6 and ...tt:lg7 in
order to prepare ...f5. IfWhite places his bishop
on h6 in the Benoni, in order to put indirect
pressure on f5, then the other knight could
come to help by means of ...tt:ld7-f6, ...i>hS
and ...tt:lgS. Similarly, once the queen's knight is
transferred to f7 in the Chigorin Ruy Lopez, the
h6-square is taken under control, too.
I must confess that some pleasant
experimentation with the Old Benoni is one of
the indirect factors that recently increased my
interest in the Rubinstein system. The following
is my favourite game, which, as we shall see as
the chapter continues, bears strong similarities
with our examined variation:
Moskalenko - Marin
Barcelona 2005
l.d4 �f6 2.c4 c5 3.d5 e5 4.�c3 d6 5.e4fie7
6.fid3 �bd7 7.h3 a6 8.�f3 0-0 9.fie3 �h5!
This way of transferring the knight to g7 was
enabled by the development of the white light
squared bishop to d3. The typical combination
10.tt:lxe5 tt:lxe5 1 1 .Wxh5 does not work now
because the bishop is hanging. From h5, the
knight slows down White's intended kingside
attack by means of g4, in view of the knight's
jump to f4.
10.�e2 g6 1l.g4 �g7 12.�g3
28. The Rubinstein System 27
White has managed to advance his g-pawn
only by transferring his queen's knight to the
kingside, which has left the other wing slightly
unattended. From the perspective ofourchapter,
the situation starts becoming increasingly
interesting, because in the Ruy Lopez White
frequently adopts an identical kingside set-up.
The fact that the queen's knight reaches g3 via
a different route is completely irrelevant, of
'
course.
It should be mentioned that White had no
way of keeping both wings under control,
because 1 2.a4 would have allowed 1 2...f5.
12...bS 13.b3 l3b8
Now, the queenside situation also strongly
resembles that from the Ruy Lopez. In fact, the
only significant difference is the placement of
Black's queen's knight and its further route to
[he kingside.
It should be said that Black cannot claim an
advantage on the queenside yet, but the gain of
space is a relatively safe guarantee that he will
not be submitted to pressure on this part of
[he board in the near future. This also applies
tor the Ruy Lopez, as we shall later have the
opportunity to convince ourselves.
For the moment, neither side can undertake
anything concrete, which explains the slow
manoeuvring phase that follows.
14.1t>f1
The king is heading for the only relatively safe
zone. After Black's progress on the queenside,
-:astling long is out ofquestion, ofcourse.
14...'lWe8 lS.lt>g2 �h8
Not only removing the king from the g-file in
-iew ofthe planned advance ofthe f-pawn, but
also clearing the gS-square for the knight.
16.'lWe2
Possibly not the best square for the queen.
-iktor probably hoped to force me to clarifY the
queenside situation, failing to understand that
:he threat against the b5-pawn is illusory.
16...i.d8!
We are familiar with this move from the game
'ijboer - Marin. Black intends to gradually
-:lear the seventh rank in order to enable the
transfer of the queen's rook to the other wing if
the position opens. Besides, White will have to
permanently watch out for the activation ofthis
bishop via a5-c3-d4.
17.i.d2
Not only radically preventing the latter idea
behind Black's last move but also preparing
the transfer of White's own bishop to a more
active position. 1 7.cxb5 axb5 l S.�xb5 would
have released the pressure against the f5-square,
allowing dangerous counterplay after l S...f5�.
17...�f6
In order to make the threat ...f5 real, Black
has to open the diagonal the way for his queen's
bishop. Even though there is no enemy bishop
on h6, gS is still the best square for the knight.
18.Ad
From here, the bishop puts some sort of
X-ray pressure against the enemy king. In case
of the opening of the kingside, Black will have
to permanently consider piece sacrifices on e5.
During the game, I considered that a logical
continuation would be ls.lLlh2 (threatening f4)
1 s...lLld7! (taking the e5-square under control
in order to inhibit White's plan and threatening
�h4xg3 to enable f5) 1 9.1Llf3 lLlf6= with a draw
by repetition. Apparently, Viktor hoped for
more.
18...�g8
Black has completed his regrouping and can
look forward with optimism.
19.'lWc2
29. 28 A Spanish Repertoire for Black
White puts additional pressure on the f5-
square and refrains from concrete action yet. I
considered that White's last chance for active
play consisted of 1 9.cxb5!? axb5 20.a4 which
is a logical consequence of Black's exodus
to the kingside. Actually, we shall meet this
situation quite frequently in the Ruy Lopez:
Black's apparently active queenside chain of
pawns can easily become a target for White, as a
consequence ofhis higher mobility in the centre.
Simple moves like a4 can be veryannoying under
certain circumstances.
19...f5!
Blackdecidestostarthisattackatonce,ignoring
the bishops' pressure against his kingside. Mter
the more cautious 19...b4 20.i.b2 a5 followed
by ...a4, the most probable result would have
been a draw, since there is no way for White to
prepare the opening of the position by means of
f4 without allowing ...f5 in a favourable form.
20.exfS gxf5 21.gxf5
Another continuation that would have led to
complications was 2U'lael e4 22.lt:Jh5 and now
the long awaited rook lift 22...:Sb7! followed by
Wg6 gives Black the initiative.
21...lLlh6 22.:!3ae1
Creating the strong threat of tLlxe5, when the
black king and the b8-rookwould find themselves
on dangerous diagonals, and the black queen on
a dangerous file. 22.f6 would just give Black time
to regroup, for insrance: 22...tLlh5! 23.�7
lLlf4t 24.<;!;>h2 Wd7+
22.•.gb7
After removing the rook from its exposed
position and preparing to safely defend the
seventh rank and the e-file, the threat of
capturing on f5 with a positional advantage
becomes real. White has to react energetically.
23.lLlxe5!
Finally, I could feel that my generally
aggressive opponent was in his element, in spite
ofhis severe time trouble.
23...dxe5 24..be5 Vlifl 25.f4
With three pawns for a knight and his strong
central domination, White has sufficient
compensation. However, the exposed position
ofhis king will enable certain tactical nuances.
25....ih4
Black has to choose the moment ofcapturing
on f5 with care. After 25 ...bxc4 26.bxc4 tLlhxf5
27..txf5 ixf5 28.tLlxf5 Wxf5 29.Wxf5 :Sxf5
30.<;!;>f3� followed by <;!;>e4 he can only be
worse.
26.:!3e3 bxc4
Once again, I disliked 26...ixg3 27.:Sxg3
tLlhxf5 28.Lf5 Lf5 29.Wf2!?� because of the
permanent pin along the dark diagonal and the
weakness of the c5-pawn.
27.bxc4.if6 28.:!3fl
Apparently, Black has just lost a tempo with
his bishop manoeuvre, but it will soon become
clear that the enemy queens rook has in fact
been attracted to an exposed square.
28...lLlhxf5!
Nowor never. White's pressure was starting to
become almost unbearable...
29.lLlxfS
29.Lf5? is worse because of 29...i,xf5
30.tLlxf5 Wg6tFF followed by ...Wxf5, winning
thepiece backwithout thenecessityofpreviously
exchanging on e5.
29. ...ixf5 30..ixf5 .L:e5 31.fxe5 lLlxfS
32.:!3xfS ?
The decisive mistake in time trouble. Correct
was 32.Wxf5 Wxf5 (Viktor feared that after
32.. J'!b2t Black would get attacking chances
but after 33.<;!;>g1 the king enjoys absolute safety)
33.:Sxf5 :Sxf5 34.e6 and, facing the awesome
pair of passed pawns, Black has nothing
30. The Rubinstein System 29
better than giving a perpetual with 34...l"lg7t
35.@h1 l"lflt 36.@h2 l"lf2t.
32...'!Wg6t!
Winning a decisive tempo compared to the
previous line.
33J:!g3
Now, 33.@h1 would lead to a mating attack
after 33...!!xfS, leaving White no time to
advance his pawns.
33...'!Wns 34.'!Wns gns 35.e6 gb2t 36.@gl
l'!e5
The pawns have been stopped and the white
king is still in danger. The rest is easy.
37.Wfl gee2 38.gd3 gee2 39.gdl Wg7
40.d6 B:h2
0-1
Let us draw some general conclusions that will
be useful for understanding the similar structure
in the Ruy Lopez.
In spite of his apparently aggressive kingside
set-up, White cannot easily start an attack on
this wing. On the contrary, it is mainly Black
who can start active operations here. The pawn
:,reak ...f5 requires thorough preparation but
is hard to prevent in the long run. However,
opening the position on this wing has a double
edged character because the central shield of
pawns does not offer a 100% warranty against
''hite's well coordinated minor pieces.
The main thing Black should be worried
.;.bout on the kingside is a possible pawn break
:.. means of f4, at a moment when his knights
.;.re too far from the e5-square. This is an eternal
:heme in the Ruy Lopez, but in a slightly
different context and its essence needed decades
:0 reveal itself completely.
On the other wing, the situation is to a certain
�xtent reversed. It is Black who apparently has
§":ood prospects of developing an initiative here,
:'ur in fact his good-looking tandem of pawns
�an easily become a target as well lsee White's
:nissed opportunity 19.cxb5 followed by
20.a4).
We can state that the strategic situation is
quite complex and requires from hoth players
the ability to maintain the balance berween the
tensions created on both wings.
It is also worth mentioning that the regrouping
typical of the Rubinstein Variation had been
employed just once before (in a completely
different line though) in the following game
g!.ayed 8 years before Akiba was born. I guess
we shall never know whether the great Pole was
aware of this fact.
MaeDonnell -Wisker
London match 1 874
l.e4e52.<!iJf3 tile6 3.i.b5 a64.ia4tilf6 5.0-0
ie7 6.tilc3 b5 7.ib3 d6 8.h3 0-0 9.d3
A completely inoffensive system of
development by modern standards oftheory, of
course. White does not manifest any ambitions
in the centre and leaves the Spanish bishop
exposed to the attack ofthe enemy knight.
9...tila5 10.ie3 B:b8 Il.tile2
The knight is heading for g3 via the same
roure as in the Old Benoni.
1l...e5 12.c3 '!We7
Of course, Black would have obtained
comfortable play by capturing the bishop now
or anywhere in the previous phase.
13.i.c2 b4
Obviously premature. Black usually advances
his b-pawn only as an answer to a4.
14.i.d2 tile6 15.d4a5 16.d5 tild8 17.e4 tile8
31. 30 A Spanish Repertoire for Black
Here we are! One cannot help wondering
whether it is possible to invent anything really
new in chess.
18.g4 g6 19.1Llg3 lLlg7 20.�h6 f6 21.lLlh2
ttlf'7 22.hg7 s!?xg7 23.h4 �d7 24.b3 lLlh6
25.0 gaS 26.gn Wlc827.gg2 a4
and Black had a comfortable position which
he eventually managed to win.
It is time to return to the game Bernstein -
Rubinstein.
15.a4
This can be considered a sign that White
finally understood the depth and soundness
of Black's plan. Up to this moment he might
have hoped that by the massive concentration
of forces on the kingside he would soon obtain
a decisive attack. The clumsily placed dS-knight
would supposedly complicate Black's defensive
task. Then, all of a sudden, Black's play started
proving its coherence: the intended regrouping
will not only solve the problem ofthe knight but
also put up worthy resistance on the kingside.
Hence, the desire to question Black's theoretical
superiority on the other wing, before his pawns
would become threatening.
On the other hand, in spite of the fact that
Black temporarily has to give up the control ofthe
a-file due to the lack ofcommunication between
his rooks, the whole issue is less one-sided than it
might seem at this stage of the game.
15...gb8 16.axb5 axb5 17.g4
In view of the transfer of the knight to £7,
the direct opening of the kingside position by
means of tLl3h2 and f4 would only weaken
the e5-square. Therefore, White puts the f5-
square under strict observation in order to
prevent Black's further expansion on this
wing.
17...g6 18.lLlg3 lLlg7 19.s!?hl f6 20.gg1 lLlf'7
21.Ae3 �d7
This move marks the completion of
development. Remarkably, Black needed
only eight moves in order to transform an
apparently chaotic position into a harmonious
one. It should be mentioned that he was helped
by the closed character of the position, which
prevented White from launching a direct
attack.
We can now draw a parallel with the Old
Benoni. The placement of the knight on £7 has
both advantages and drawbacks. By controlling
the e5-square it protects Black against the
opening ofthe kingside by means off4 without
obstructing Black's light-squared bishop as it
would on d7, but also makes carrying out the
plan based on ...f5 more difficult by interfering
with the action of the f8-rook. Although there
is no direct pawn contact on the queenside, the
situation is more or less similar to the Benoni:
it is not yet clear whether the black pawns are a
strength or a weakness.
32. The Rubinstein System 3 1
However, a transposltlon to an almost
identical situation as in my game against
Moskalenko is still possible if White plays b3
and c4 followed by the transfer ofthe bishop to
the long diagonal, allowing Black play ...tLlh6
followed by ..J5. This is purely hypothetical
at this stage of the game, but clearly shows
how similar positions can be obtained from
completely different openings. In fact, Dorian
Rogozenko explained to me that this is not so
surprising: "They start the games with the same
pieces" he said. Quite correct.
22.tve2 �a8
Once the rooks have been connected, Black
can challenge White for control of the a-file.
This apparently insignificant diversion will in
fact prevent White from building up a kingside
attack at his leisure.
23. .!LJd2
This is not just a sign that White intends to
operate on the queenside. By clearing the dl
h5 diagonal, he creates the potential threat of
�f5 and if . . .gxf5 then gxf5 followed by Wffg4
0r Wffh5. True, this would work only under
�ertain circumstances, but given the fact that
play takes place on both wings now, Black has
,0 be careful.
23...Wh8
The simplest way to remove the danger in
order to be able to concentrate on the queenside
..Jone. By quitting the dangerous g-file, Black
o-ins a crucial tempo in the event of a later
;:'f5.
24.b3
By preventing ...c4, White intends to keep
:,l.e b5-pawn under pressure. However, Black
�as sufficient resources to defend it and keep
:'ghting for control of the a-file.
24...tvb7 25.id3 �a6
Black's space advantage on the queenside
.;Jaws him to win the strategic battle on this
,.-ing by simple means. His position i.., not worse
i: all now.
26.�gbl �fa8
and a draw was agreed.
A rather premature finish, as any chess fan
would say. A brief glance reveals the fact that,
with his minor pieces massed on the kingside,
Black could not build on the partial success of
conquering the a-file so easily. However, with so
many pieces on the board, there certainly was a
lot of play left. On the other hand, we should
not forget that both player hadfound themselves
on unexplored territory from a relatively early
stage ofthe game.
Be that as it may--, here are two games where
play continued for a longer while from similar
positions.
Alapin - Rubinstein
Lodz 1 907/08
33. 32 A Spanish Repertoire for Black
29.'l;Ybl
Understandably, White prevents the
infiltration of the enemy queen to a2.
29...ttldS?!
For once, Rubinstein's "long plan" seems to
have been not entirely adequate. The migration
of his minor pieces ro the other wing initiated
by his last move will prove rather pointless.
There is simply not sufficient space available for
manoeuvres on the queenside.
Things would change ifBlackcould manage to
gain the c5-square for his knights by means ofa
well-timed ...c4. However, ifplayed at this stage
ofthe game, 29...c4 would allow White ro open
the queenside under favourable circumstances
by means of 30.b3 cxb3 3 1 .Wixb3 followed by
.td3, liJd2 and (eventually after liJbl-a3) c4.
Another drawback ofRubinstein's plan is that,
by giving up the potential threat ofopening the
kingside with ...5, he will leave White a free
hand for manoeuvres.
Black's next strategic aim should be the
exchange or (in the case of refusal by White)
activation of the dark-squared bishop. True,
this was part of Rubinstein's plan, too, but a
more economical way ro achieve it consisted
of 29...liJe8 planning ...if8-h6. For instance,
30.b4 c4 This move is not really forced yet,
since the exchange on c5 would free the d6-
square for Black's minor pieces as compensation
for the deterioration of the pawn structure. We
shall have a longer debate about this matter
later in this chapter. For the moment I would
like to simplify things in order to follow the
main idea. 3 1 .Wib2 White prepares to neutralize
Black's pressure along the a-file. Preventing
the above mentioned plan by means of Wid
would have led to slight problems after ...Wia2.
3l ....tf8 32.ibl ih6 33.Wia2 An interesting
moment. Both sides have offered the exchange
they were interested in, but neither of them is
interested in improving the opponent's position
by carrying them out. In blocked positions,
one important possibility is the sacrifice of a
piece for two pawns, in order to create a mobile
mass of pawns. We can clearly see that Black
cannot easily carry out such an operation, bur
his b5- and c4-pawns are in some danger. From
this perspective, 33....tc8 looks quite natural,
leading to equality and a probable draw after
34.Wixa6 ixa6 35.h4 he3 36.liJxe3 @f8.
30.b3
White anticipates the intentions of his
opponent and prevents ...c4 followed by ...liJb7-
c5. 30.b4 would have been equally good.
30...tiJb7
It was not too late for Black to return to the
aforementioned plan with 30...liJf7.
31.'i'b2 idS 32.tiJd2 ttleS 33.i.fl tiJc7
34.tiJbl Wa5 35.b4 Wa6 36.id3 c4 37.i.e2
tiJaS 3S.Wa3 Wxa3 39.tiJxa3
Black's regrouping hasn't yielded the expected
results. His pieces are rather standing in each
other's way. Meanwhile, White has managed
to neutralize the pressure along the a-file and ,
concentrate his forces around the c4-pawn. The
game ended in a draw after ten more uneventful
moves. Things could have become livelier if'
White had decided ro sacrifice a piece on c4 at
some moment.
In the next game, we shall see an opposite
picture: it will be the unnecessary transfer of
one ofWhite's knights ro the queenside that will
cause him trouble.
Thomas - Rubinstein
Baden-Baden 1925
34. The Rubinstein System 33
White has placed his pieces slightly less
harmoniously than in the previous games, but
he is by no means worse; the margin ofequality
is wide enough. He apparently connected his
hopes with a sacrificial attack along the g-file,
but will not be able to carry it out. Such radical
operations can succeed only as a consequence
of a strong concentration of forces, but part of
White's forces will be inevitably distracted to
the other wing.
22•.ie3
Quite deceptively, the attack would get
nowhere after 22.�g2 E1a8 23.E1xa8 (In the case
of 23.E1bl the least Black could do is 23...�h8
followed by a gradual increase ofthe queenside
pressure.) 23...E1xa8 24.ct:lf5 gxf5 25.gxf5 i,f8. If
the queenside were closed, White could possibly
increase his pressure (for instance by means of
ct:lh2-g4). The way it is, Black can harass him
with ...E1al, pinning the bishop and eventually
threatening to give back part of the material
with ...E1xcl if things get too dangerous on the
kingside.
22••.1'l:a8 23.!;lfg2 :1l:xal 24.:1l:xal !;lfb7
Black has carried out the exchange ofone pair
of rooks at a slightly earlier stage than in the
other games (meaning before placing the queen
on b7) in order to reduce White's attacking
potential as soon as possible. With his queen far
away, White lacks just one tempo to take over
the control of the a-file by means of E1a2 and if
...E1a8 then �al .
25.cJ;1h2 :1l:a8 26.!;lffl :1l:a6
27.tLld2
The start of a mistaken regrouping. The
simplest way to keep Black's queenside activity
under control consisted of 27.b4 when after
27...c4 we would reach a familiar pattern.
27•••!;lfa8 28.:1l:xa6 !;lfxa6 29.tLlb3 ?!
This is a significant inaccuracy. White intends
to neutralize Black's pressure along the a-file
as soon as possible, but employs exaggerated
means to achieve his aim. Once transferred to
the queenside, the knight cannot rapidly return
to the other wing, leaving the king's position
without sufficient defenders. 29.�bl followed
by b4 would have maintained approximate
equality.
29•••tLlg5
The lack of defence of the f3-square allows
Black to activate his position with gain oftempo
before launching the attack.
30.cJ;1g2
Actually, the immediate return of the knight
with 30.ct:ld2 would have been a better answer.
Things would work out perfectly well for White
ifBlack would play as in the game. For instance
30...h5 3 1 .h4 ct:lf7 32.gxh5 gxh5 33.�g2 �f8
34.i,dl f5 35.ct:lxh5 ct:lxh5 36.i,xh5 f4 37.�g6
i,e8 and now the fact that the queen didn't
have time to move to c8 or a2 allows 38.i,xc5!.
However, the simple 30...�a2 (instead of
30...h5?!) would give Black comfortable play. If
3 1 .�bl then 3l...�xbl followed by 32...h5.
30•••h5 31oh4 tLlf7 32.gxh5 gxh5 33.cJ;1h2
White's coordination is far from impressive.
The king, which had been used to defend the
f3-square, clears the g-file for the queen, in
order to create some counterplay.
33•••!;lfc8
The incursion of Her Majesty to the enemy
queenside with 33...�a2 would have been an
equally effective way to prepare .. .f5 as the
regrouping to the opposite wing. In case of
34.�bl , the exchange of queens would allow
Black to open the kingside without the undesired
dose of risk regarding the king's position, while
after 34.�g2 �f8 35.i,cl f5 36.exf5 i,xh4 the
queen would keep captive no fewer than three
35. 34 A Spanish Repertoire for Black
minor pieces: quite a satisfactory equivalent to
compensate for her absence from the kingside.
Moreover, after ....ixg3t followed by ....ixf5,
White's pieces would be hanging rather badly.
. 34.'&g2 �f8
35.tild2
This admission of the erroneous character
of the previous plan comes slightly too late as
White does not have sufficient means to counter
Black's plan. For instance, the attack against
the relatively weak h5-pawn with 35..idl
would not inhibit Black from · playing 35...f5
(35....ig4 would be playable but too modest)
36.ltlxh5 ltlxh5 37..ixh5 f4 and the strong
concentration offorces against the white king is
rather annoying. As in certain lines ofthe King's
Indian, the black pawn chain severely restricts
White's pieces. Now, the seemingly active
38.'&g6 does not spoil Black's coordination in
view of the calm 38....ie8 39..id2 hh4. For
instance, 40.iel ltlg5 41 .'&xe8t (if 41 ."Mfxd6t
�g8 the white queen would remain out of play,
while the intrusion of her colleague to h3 will
have decisive consequences) 41 ..."Mfxe8 42..ixe8
�xe8 43.ltld2 f3 and White is completely
paralysed. There is nothing he can do against the
transfer of the black king to f4, leading to the
gradual elimination of all White's pawns.
35. . .£5 36.exfS .hh4
Winning a pawn, but also undermining
White's control over the f5- and e4-squares.
We are familiar with this kind ofincursion ofthe
bishop from the game Moskalenko - Marin.
37.f6!?
Giving up the already doomed pawn diverts
the enemy bishop...
37....ixf6 38.Wif3 i.h4
...but not for very long.
39•.ig6 hg3t 40.fxg3 �J5 4L.ixf7 �xf7
42.tile4 Wid7
Black has won a pawn while retaining the
better structure. White's attacking chances
against the relatively weakened kingside are not
too realistic and Rubinstein went on to win a
nice technical game.
0-1
Bogoljubow's challenge
One natural question raised by the experience
from the previous games is whether the opening
of the a-file really favours White. As mentioned
above, one of the basic purposes of a4 is to
radically prevent a gradual black expansion on
the queenside, but is the capture on b5 really
necessary?
Before giving an answer, I will allow myself a
short digression.
In the previous games White played in a
relatively inoffensive way and Rubinstein did
not face any concrete dangers (apart from those
that he created for himself against A1apin). The
real test for Akiba's variation consisted ofseveral
tournament games against Bogoljubow, a strong
and enterprising player with a natural talent for
attack. Bogoljubow understood perfectly well
that, although separated from each other by
the blocked centre, the opposite wings could
still influence each other, albeit indirectly.
Therefore, he did not only try to find optimal
kingside regroupings but looked for queenside
solutions, too. From one game to another, he
constantly aimed to change plans on both wings,
in the hope that he would find the "winning
combination" of patterns. Anticipating matters
rather, I can say he did not manage to achieve
this aim, but he did unearth many interesting
36. The Rubinstein System 35
ideas and posed 'his mighty opponent serious
practical problems.
In the following game, he decided to block the
queenside completely.
Bogoljubow - Rubinstein
Breslau 1925
13.dS otd8 14.a4
The significance of the fact that White plays
this move at an earlier stage than in the previous
games will be revealed a little bit later.
14. . JrlbS lS.c4!? b4
Black is best advised to release the tension at
once. White's pieces are optimally placed for a
fight on the light squares. Ifallowed to exchange
twice on b5 he would place his knight on c4 and
the bishop on a4. Later, he could transfer his
king's rook to the a-file using the third rank as a
line ofcommunication. We can understand now
why it was so important to start the queenside
operations before transferring the knight to the
other wing.
16.�h2 �e8 17.g4 g6 18.B:gl f6 19.�fI
�f'7
We shall see that down the years Rubinstein
had refined his treatment of the position, too.
By refraining from the otherwise desirable
...tLlg7 yet, he leaves the h6-fS diagonal available
in view of the possible transfer of his bishop to
h6. This would become impossible if White
develops with i.e3 and lMfd2, but for the time
being Black can just make all the other useful
movesand then choose the furtherplan ofaction
in accordance with the concrete circumstances.
20.ote3 mb8 21.b3 B:g8 22.i.b2
By placing the bishop on the long diagonal,
Bogoljubow probably intended to inhibit
an eventual opening of the kingside by . . .f5.
However, if Black avoids that, the bishop
stands rather passively on b2, being restricted
by the over-defended e5-pawn. Besides, Black's
kingside potential consists of more than just
advancing the f-pawn.
22••.i.f8!
It looks as if Rubinstein had guessed
Bogoljubow's intentions. Once the white bishop
has been developed to b2, the h6-c1 diagonal
remains available for the black bishop.
23.M Be? 24.B:g2 .ih6 2S.Be2 .ti4t
26.mbl
Apparently, Bogoljubow had not formulated
his further plan of action yet. Otherwise he
would have played 26.@gl at once.
26.. :i!Jf8
The regrouping initiated with this move
again reveals Akiba's tendency to overestimate
the effectiveness of the concentration of forces
regardless of the concrete circumstances, as in
the game against Alapin. The simple 26....id7
followed by ...tLlg7 and ...h5 would have offered
Black very promising play.
37. 36 A Spanish Reperroire for Black
27.'it>gl Wh6 28.i.cl .td7 29.'it>fl lLlg7
30.WdJ �bf8 31.�dl
Black has reached the apogee ofhis intended
regrouping but cannot make any furrher
progress. In the meantime, White has starred to
gradually improve his position. Once his king is
evacuated, he will be ready for active operations.
Rubinstein was wise enough to admit his error
in judgment and immediately starts the reverse
process.
31...lLle8 32.'it>el Wg7 33..tb2 lLlh6 34.lLlgl
We7 35.lLlh3 lLlf7 36..tf3 .th6 37.'it>e2
White intends to increase his pressure with
�gl or �hl, then hide his king on a2 and look
for a favourable way to open the position. Black
cannot afford to wait.
37...f5!?
Now or never! What makes these positions
so difficult to play over the board is the fact
that the possible consequences of this thematic
pawn break must be calculated virtually on
every move. White can count on an advantage
only ifhe permanently retains sufficient tactical
resources to meet ...f5 in an efficient way.
38.exfS gxf5
Possibly underestimating White's 40th move,
but the alternative looked pretty dangerous for
Black: 38...1Mfxh4 39.fxg6 l:!xg6 40.�hl with
strong threats along the h-file.
39.lLlxfS
39.g5 ig7 40.CLlxf5 doesn't work because of
40...e4! when too many white pieces would be
hanging, for instance 41 .i.xg7t CLlxg7 42.ttlxe7
exd3t followed by ...�xh3.
39.. .hf5 40.WxfS!
Very resourcefully played. In the case of the
modest 40.gxf5 Black would obtain excellent
play with 40...�xg2 41 .1i.xg2 1Mfxh4 followed
by ...CLlf6 and ...0g5, with a firm blockade on
the dark squares. We can notice the relative
uselessness of the b2-bishop.
40...Wxh4
A wise decision. Black refrains from "winning"
the queen with 40...CLlg5 4 1 .hxg5 �xf5 42.gxf5
hg5 which would have led to a very powerful
attack for White after 43.�agl h6 44.icl when
White's pieces cooperate perfectly. Instead of
exposing himself to such dangers, he builds up
a reliable fortress on the dark squares, for which
the elimination of the h4-pawn is essential.
41.�hl lLlg5 42.<�:lxg5
42..J�xfS!
In view ofWhite's 44th move, this might look
like a blunder. However, if 42...1Mfxg5 White
would avoid the exchange of queens with
43.1Mfd3 leaving Black's pieces exposed to such
threats as �cl , �e4 or �gh2.
43.�xh4 hg5 44.l;xh7t 'it>xh7 45..te4
White has won a pawn but has not managed
to challenge Black's darksquare stabiliry. Besides.
the position has been considerably simplified.
depriving White of further tactical possibilities.
38. The Rubinstein System 37
45...a5!
A last accurate move. Black removes his last
pawn from a light square in order to avoid its
loss after a5 and an eventual infiltration of the
enemy bishop to cB.
46.�g1 lLlf6 47 !;lg7 48.J.c1
White cannot make any further progress
without exchanging this bishop, which has
remained rather passive on b2. Carrying out
mis operation now will allow the black king
(0 replace the bishop in blocking the g-pawn,
but White could not avoid this anyway. For
instance, 4B.B:h1 (aiming to deprive the king
of the h6-square) could be met by 4B...B:hB
49.B:xhB �xhB when the bishop would lack
support on d . White could transfer the king
to d1 , but meanwhile Black would activate his
king, remaining clear of any danger.
48.. .,hc1 49.gxc1 !;lh6 50.gh1t !;lg5
White unsuccessfully tried to break Black's
fOrtress for 1 6 more moves before agreeing to
• draw. . .
�1j2
Decades later, the outstanding Ruy Lopez
.specialist Paul Keres stated that White has
· no real chances for success if he plays on just
GIle wing. If this is true (and I tend to trust
Xeres) then White's plan in the previous game
. ofcompletely blocking the queenside is slightly
questionable. Bogoljubow must have come to
the same conclusion since one year after the
previous game, he introduced some significant
changes in his general set-up.
Bogoljubow- Rubinstein
Berlin 1 926
First ofall, he exchanged on b5 and only after
15.axb5 axb5
provoked the blocking of the position wim
16.c4 b4
Let us draw some parallels with the game
against Thomas. White enjoys slightly more
space on the queenside, which would allow
him to defend against an intrusion to a2 in a
very natural way and without the necessity of
losing coordination, namely with the bishop
on d3 and the queen somewhere on the second
rank.
A less obvious aspect relates to the d1-a4
diagonal. The placement of the bishop on a4
would not disturb Black too much since the
significance of the eB-square is relatively low
with the e-file safely closed. However, this
diagonal could be successfully used by the king
in the event of mass simplifications: an aspect
that Black must always guard against. At the •
present stage of the game this might sound like
pure fiction, but it is precisely what will happen
some 40 moves later.
17.lLlfl lLle8 18.g4 g6 19.J.h6
•
39. •
3B A Spanish Repertoire for Black
Bogoljubow intends to place his fl-knight
on e3 as in the previous game, but the problem
of the dark-squared bishop is not easy to solve.
Placing it on h6 and cutting off its retreat with
ctJe3 means giving up the bishop pair. Since the
d1-a4 diagonal is meant to be left open, the
development of the bishop to b2 is impossible.
The only reasonable alternative is 19.ctJg3
leaving the e3-square available for the bishop.
However, Black is not without chances here, too.
A logical continuation is 1 9...f6 20.'it>h2 'it>hB
2LE1g1 ctJf7 22..ie3 ctJg7. Now that White has
defined his intentions regarding his bishop, the
plan based on ....ifB-h6 does not look realistic
anymore and Black should just connect rooks as
soon as possible. 23.Wd2 .id7 Both sides have
completed their development and White finds
himselfat a crossroads.
24.b3 (This looks like a concession but
after 24.E1g2 Black can get counterplay with
24...b3 25..id3 E1aB. For instance, 26.E1b1 E1fbB
27.Wc3 Wa5 2B..id2 .idB with a harmonious
arrangement of the black forces.) 24...E1aB It is
useful to play this move before White gets his
king's rook to g2, clearing the gl-square for
the other rook. 25.E1afl (Black would also win
the battle for the a-file in the case of a neutral
answer such as 25..id3, by means of 25..J::!a3!)
25...E1gB 26.E1g2 WeB 27.E1fg1 E1a2 Black has
achieved a very rational distribution of forces
and has every chance of holding the balance.
The pressure exerted by the a2-rook prevents
White from concentrating all his efforts on the
kingside.
19...ttlg7 20.ttle3 f6 21.@g2 ttlf7 22..ixg7
'it>xg7 23.ttld2
This is the position White had been aiming
for.
Generally speaking, his strategy is quite
coherent: after giving up the dark-squared
bishop for a knight he has placed his remaining
minor pieces in such a way that they control
several light squares.
However, there are certain concrete
circumstances that cast some doubt on White's
concept. First of all, the c4-square, on which
both knights concentrate their action, is
occupied by White's own pawn. Had the pawn
been placed on c3, White would have had a
significant advantage on the queenside, bur
the way it is, the control of the b3- and a4-
squares is not so important yet. This is also a
consequence ofthe fact that Black's pawn chain
offers him an entirely adequate defence against
minor piece pressure on this part of the board.
If, for instance, the d6-pawn were missing, the
pressure exerted by a knight placed on b3 on the
c5-pawn could be quite unpleasant.
On the other wing, White's position is also
not without drawbacks. The most obvious of
these is the weakness of the f4-square: a direct
consequence ofWhite's previous play.
40. The Rubinstein System 39
All these aspects make the position rather
complicated from a strategic point of view and
quite difficult to assess.
23• •J�h8
From the previous discussion, the only
question that seems to remain open is whether
Black can effectively occupy the f4-square or
take advantage of its weakness in some other
way. Rubinstein decided to open the h-file in
order to take over the control of the h3-square
and enable the spectacular transfer ofthe knight
to f4. However, the long-term consequences of
these operations are far from clear as we shall
see.
Although from an aesthetic point of view it
might seem that the knight is the best suited
piece to occupy the f4-square, there are strong
arguments for the bishop's activation, too.
Opening a file on the wing where one's own
king is located can be a double-edged decision.
It is generally better to choose an alternate plan
if one is available. Secondly, it is the bishop
that truly needs a change, since on e7 it does
absolutely nothing and risks becoming a typical
"bad bishop" in the case of simplifications. On
the other hand, the knight (which, by the way,
has consumed five of Black's tempi already) is
relatively well placed. It can either drive the
enemy bishop back if it infiltrates to c6 or jump
to g5 to support its own bishop if it reaches f4.
Moreover, it keeps the e5-square under control,
inhibiting the opening ofthe position with f4 or
an eventual exchange on f4.
All these aspects strongly suggest that a plan
of regrouping initiated by 23...c;!?hS is safer
(rom a general point of view, but can lead to
equally sharp positions after, for instance,
24.lLldfl (White needs to increase his pressure
against the f5-square. After the straightforward
24.�a4 l"lgS 25.�c6 Black would get strong
counterplaywith 25...f5.) 24...l"lgS 25.lLlg3 �f8
26.�a4 �h6. White has problems keeping the
kingside under control. The immediate threat is
. ...�e3 followed by ...f5. White needs to spend
a tempo with 27.'lWd3 to prevent this, when a
race between the plans on opposite wings could
start. For instance, 27....if4 2S..ic6 lLlh6 29.l"la4
'lWg7 (such a rapid and effective regrouping has
been made possible by the fact that the dark
squared bishop has cleared the seventh rank)
30.l"leal f5 3 1.l"la7 (the immediate capture of
the pawn with 3 1 .exf5 gxf5 32.lLlexf5 leaves
White in deadly pins after 32...'lWg6!) 3 L..'lWf6
32.l"lc7 (now the f5-square is under sufficient
control by Black and 32.exf5 gxf5 would only
open the g-file for Black's attack) 32...fxg4
33.l"laa7 (We can see that the description of
23...c;!?hS as "safer" is relative. The immediate
doubling of rooks is stronger than 33.hxg4
lLlxg4 34.lLlxg4 Lg4 35.l"laa7 when Black
has the simple defence 35...'lWh4. In the case
of the exchanges on h7, Black would soon
double rooks on the f-file with some initiative.)
33...gxh3t 34.�hl Black cannot parry the
threat against the h7-pawn, but it appears that
he can live with that, since other important
squares along the seventh rank arewell defended.
34...l"lfS 35.l"lxh7t �gS 36.lLlef5!? The only way
to maintain the initiative. Otherwise, White's
pieces would remain hanging. 36...lLlf7 (only not
36...gxf5? because of37.lLlh5 followed byl"lag7t)
37.l"lg7t �hS 3S.i.eS! A strong move, bringing
this bishop from its offside position into play
and proving that White has not exhausted his
resources yet. 3S...�xf5 Not the only possible
answer, but the most logical one: Black attacks
the intruding bishop and aims to bring his
queen's rook into play. 39.l"lgxf7 l"lxf7 40.l"lxf7
'lWg5 41 .exf5 l"lxeS 42.fxgG 'lWg4 It is a sharp
position, where Black's chances are not worse.
24.{)dfl h5
Just in time, before White takes the h5-square
under control with lLlg3.
25.{)g3 hxg4
The opening of the h-file is more ambitious
than 25...h4 which would have been equivalent
to starting a cheerless fight for a draw. After
26.lLle2 Black would have serious problems
of coordination. If ...i.d7 in order to connect
rooks, then the exchange of the light-squared
bishops with �a4 would make the opening of
the kingside by means of f4 quite unpleasant,
41. 40 A Spanish Repertoire for Black
since ...exf4 would allow White to transfer his
knight to e6.
26.hxg4 c!LJg5 27.i.a4
White hurries to make use ofhis main trump.
27.:Elhl would only delay but not prevent
the knight's manoeuvre in view of 27...:Elxhl
28.lMfxhl .id7 29.i.a4 :Elh8 30.lMfdl . This
situation would be even more advantageous for
Black than in thegame because he couldpreserve
his bishop from exchange with 30...ic8 without
obstructing the communication between his
rooks.
27• • • c!LJh3 28.i.c6 c!LJf4t 29.<j;lf3
Both sides have been consistently carrying
out their partial plans. They each control one
ofthe open files and have each installed a minor
piece on an advanced outpost. However, their
achievements are not entirely equivalent.
The stability of the far advanced minor
pieces is only relative, but we can notice
certain nuances of difference. The exchange
of the c6-bishop by means of ...�d7 favours
White from a strategic point of view, because
of the number of black pawns blocked on dark
squares.
In contrast, an eventual exchange of the f4-
knight by means ofCLle2 or CLlg2 has a "neutral"
character and cannot in any way be considered
a strategic concession from White's point of
•
view.
In certain cases, White can even sacrifice a
pawn on c6 (by refusing to take on d7) in order
to get the d5-square for his e3-knight. Black can
hardly do the same with his knight since there
is no piece that can make effective use ofthe e5-
square after CLlxf4 exf4.
The significance of the control of the open
files is strictly related to the aspects revealed
in the previous paragraphs. It might seem that
Black cannot fight for the a-file unless he trades
the light-squared bishops, while the influence
of the f4-knight over the h-file is practically
•
nonexistent.
These aspects strongly suggest that the
modifications of the position induced by both
sides' manoeuvres over the past few moves tend
to favourWhite, but the plan initiated by Black's
next move introduces new elements.
29• • J�b6!
Rubinstein decides to neutralize White's
pressure along the a-file by side-stepping the
enemy bishop. The drawback of this plan is
that Black loses all coordination between his
rooks, thus allowing White to take control of
the other open file, situated in the immediate
neighbourhood of the kings. This will lead to
very complicated play, the consequences of
which are quite hard to evaluate.
It should be mentioned that the main
alternative, 29...i.d7, does not offer Black
easy play after 30.1Mfa4 (More ambitious than
30.ixd7 lMfxd7 31 .1Mfa4 lMfxa4 32.l"1xa4 l':!.a8
33.l"1eal when after 33...l"1xa4 34.:Elxa4 st>f7
35.:Ela7 :1l:h2 Black has chances to resist in view
of the closed character of the position and the
activity of his rook, which makes the exchange
of the f4-knight quite difficult.) Now, Black
should refrain from 30...l"1a8? 3 1.1Mfxa8! l"1xa8
32.l"1xa8 �xc6 33.dxc6 1Mfxc6 34.!!eal with
an indisputable advantage for White. Instead
of 30...:Ela8, Black should defend his bishop
with 30...l"1hd8 but this clearly shows that the
initiative is all White's.
30.:Elhl :Elxhl 31.Wixhl ga6 32.gg1
White gladly abandons the a-file for the sake
ofconcentrating his forces on the kingside.
32• • • <j;lg8