M m m w mmm a

I^SA

now after 17...^.xc8 18 hxg6 fxg6 (or 18...hxg6 19 <&fl!! ®a5 20 Wei and the attack starting with JLh6 is overwhelming) 19 £>f4 White could expect to win quickly. Black's 11th move was far from best and constituted a loss of valuable time. The opponents in the other two games played much better, and I was lucky to escape with draws. This convinced me that 10 h4 is incorrect. Now, we will reveal the correct System move.

The System and winning move is lOScl (D).

I played it three times in the 5th World Correspondence Championship (1965-8) that I won, and was lucky to get two points out of the three games. A game with the Soviet GM G. Borisenko continued 10...cxd4 11 cxd4 b5? 12 ±d5 ±d7 13 h5 e6 14 ±b3 £>a5 15 ®d2 (Black would like to exchange queens; after 15...^xb3 16 axb3, the attack commencing with JLh6 and hxg6 is unstoppable) 16 Scl Sc8 17 Bxc8 and

Why is this move correct? Firstly, it passes the important System test of not allowing the drawing line 10...£>a5, which is now met by 11 ±d3 cxd4 (the immediate 1 l...£>c6 is met with 12 d5 winning a pawn) 12 cxd4 £>c6 and now the wonderful innovation 13 Sc5! (D).

This defends White's d-pawn indirectly in a most unusual way. If

mm mimi

MM,,

m,' mm&m now 13...£>xd4, 14 Sd5 wins, and 13...e6 (13...±e6 14 d5! wins) 14 JLbl, and now White has everything his way. The move 10 Scl, which makes possible 13 Sc5!, is one of the major 'theoretical' contributions of this book.

10 Scl is the only move that avoids the above drawing line. On the face of it, 10 Scl appears to be a wasted move, but it solidifies the queenside against any coming attacks and thus allows White to concentrate all his forces on the coming attack against the black king.

I discovered the above about 1976, having retired from active competition, and being able to devote time to such research. However, this move has since been introduced into tournament play by Lev Polugaevsky in 1987.1 make no claim here for being the originator of the move, since tournament practice must hold sway over unpublished analysis. However, what is important is whether 10

Scl is the System move in this position. Below, we present strong evidence that it is, and that, despite being currently in disrepute among the top players, it is in actuality a winning continuation.

What can Black play against 10 Sc1 ?

At the time I discovered 10 Scl, I based my judgement of its effectiveness primarily on the fact that it prevented the drawing manoeuvre, and thus made it possible to continue with the attack h4-h5, which I had played prematurely on move 10 in three games in the 5th World Correspondence Championship Final. I thought that White will play 1 in response to ...®a5(+) and thereafter pursue his attack with efficiency. It was not realistic to analyse much more than that. Now, however, that the move has been tested in many GM games, and, strangely, the world has a poor opinion of it, it was possible to apply System principles to the various attempts to refute 10 Scl and see what can be found.

Let us look at how Black can continue. He must act on the queenside before the attack against the king with h4 gets too strong (strangely in several games by top players with this variation, White chickens out, and plays 0-0). It follows that Black must either pressure the centre with

...Wa5 or expand on the queenside with ...£>a5.

Defences based upon ...®a5

In some people's eyes, 10...®a5 (threatening ll...cxd4) would appear to force 11 0-0 anyway. However, this is wrong! White simply answers 11 «¿>f 1!! (D)y which is based upon much of the earlier discussion.

mॠüii iii m % m f, t«±ÄAlS S 0 A ± 0 A ^A A A

on the black king, which will succeed long before the a-pawn becomes a queen) 15h5 cxd4 16cxd4 Ht\A(D).

I

I

ii.

i Jti

foüifiil!

i'r

s m sa

m db '

BAU Ii È.

im ' &a

White intends to continue with h4-h5, and with the centre secure, the attack on the black king will proceed with overwhelming energy.

A game Browne-Wolff, US Ch 1992, which is critical to an appraisal of the whole 10 Scl idea, continued correctly 11 if 1 b5 12 ±d5 ±b7 13 h4! e6 14 ±b3 Sfd8 (the idea of 14...c4 15 ±c2 ®xa2 in this and similar positions is not to be recommended; for Black to have some say in what is going to happen, he must challenge the centre, and winning the a-pawn will expedite White's attack

White's moves to this point have all clearly been System moves, pursuing his objectives simply and purposefully. However, now Black is putting the question to the white centre, and it behoves White to play ultra-correctly to demonstrate the correctness of his previous play.

This can be done by 17 hxg6! which forces 17...hxg6 as otherwise the white attack gets much too strong. Now comes the beautiful denouement: 18 d5 which threatens to win a piece with 19 Ad2. Bl^ck almost certainly had prepared the move 18...®xe4 here, but that loses to 19 £>g3!! Ife5 (there is no other move) 20 ®g4!! exd5 (what else can be done?) 21 ®h4 (D) and White's attack is very strong:

a) On 21...d4, 22 ®h7+ if 8 23 ¿Lh6 wins at least a piece.

B

I

I

#

ii.

ii.

*

fc

$ i i ;

i

It

A' / / fl

<A

A A ¥ i

B

± i : # ;

ii

m iiii

''/■¿y'/l

w

ffl

AA A A

A Af

, ¿h A A , aw a

b) On 21...£kI4!, 22 JLf4 Wf6 23 Ag5 Wb6 24 &e4! Bdc8 25 Bel leaves Black in terrible shape. He cannot play 25...dxe4 26 ®h7+ $f8 27 JLh6; nor can he capture with 25...£>xb3, because 26 £tf6+ *f8 27 Wh8+! mates in four. So, he must play 25...^f5 26 Wh7+ *f8 27 g4 dxe4 28 gxf5 gxf5 29 jLh6 with an attack that wins a piece.

c) The other alternative, 21...Bac8 22 Bel Wc3 23 &e2 Wb2 24 i.c5 f6 25 ^f4, is also hopeless for Black. These lines are exquisite and difficult enough to find so that one can safely say that not even the strongest players in the world would have found them over the board without previous home analysis. However, they are very convincing proof of the strength of White's position.

After 10...Wa5 11 *fl, Black can also try 1 l...cxd4 12 cxd4 Bd8, but then 13 h4! is strong (D):

a) 13...«fa3 14 Bc3 «Td6 15 h5 £ixd4 16 £>xd4 ±xd4 17 Bd3 e5 18

hxg6hxg6(18...Wxg6 19±xd4exd4 20 Bg3) 19 Wf3 ®e7 (19...Ae6 20 Wf6, or 19...Bf8 20 Ag5) 20 £.xf7+ &g7 (20...Wxf7 22 Bh8+) 21 ±xg6 &xg6 22 Wh5+ wins.

b) If instead 13...h5 then 14 Wb3! e6 (14...^xd4 15 l.xf7+ *h7 16 £>xd4 ±xd4 17 ±xd4 Bxd4 18 ±g8+ «¿any 19 Iff 7 and mates; also, if 14...Bf8, then 15 Ad5! is very strong) 15 e5!!, and now Black can only sit passively by while White controls the board. For instance, 15...«fb4 16 JLg5 and Black must retreat. White's hi-rook will be developed via h3, and Black is essentially helpless.

The only other meaningful defence is based upon following up on ...Wa5 with .. JLd7 and a general deployment on the queenside, hoping to gain the upper hand on the light squares before White's attack becomes too strong. Tne line 10...cxd4 11 cxd4 Wa5+ 12 *fl Ad7 was played in a game Shirov-Kozul, Biel

1991, which continued 13 h4 Sfc8 14 h5 £kI8 (to steady the kingside and put pressure on c4). White can now simply play 15 hxg6 hxg6 (D).

How is White to Proceed?

The play to here clearly conforms exactly to System principles. So White should have an effective continuation here. In fact, when I first saw this position, I could not believe that this was considered a serious defence. White's pieces are all deployed beautifully, and Black still has a long way to go. It is instructive to examine it in detail.

White appears to have a very strong attacking position. If he could only get his queen on the c 1 -h6 diagonal or on the h-file, he would get an overwhelming attack with jLh6. However, that plan is difficult to implement in the present position. Further, if Black can exchange the light-squared bishops, he will have excellent play on the queenside. So what is White to do? With all his pieces basically deployed to great effect, a System move must be one that moves forward seriously to the attack. White has many potential System moves:

a) 16 e5 or 16 d5, which both gain space and keep the black queen from the kingside.

b) 16 Wd3, which mobilizes the queen to a slightly better square and has certain ideas for moving it to the kingside for attack. This move also prevents ...JLb5, which is exactly what Black is trying to play.

c) 16 f3, which steadies the centre and may make possible a sequence such as jLd2 and Wel-h4.

d) 16 JLh6, which aims to exchange Black's best defensive piece.

16 f4 cannot be the System move as it weakens the light squares around the white king too much without f5 being much of a threat. This position has now occurred many times in GM play, but to the best of my knowledge no one found the correct System move from the selection given above. Despite the fact that Shirov, who is certainly an exceptionally strong attacking player, has played the position after Black's 14th more than once, he failed to find the correct continuation.

After becoming aware of the current 'theory' with respect to this position, I spent over 50 hours studying it. White's pieces occupy near-optimal positions, so it is not clear which of the above candidate moves is correct. 16 jLh6 appears premature as after 16...jLxh6 White must play 17 Hxh6 and the future cooperation between rook and queen seems in doubt. However, White must proceed rapidly as moves such as e5, d5, or f3 are much too slow, and the natural-appearing 16 Wd3 is answered by 16...a6! (suggested by B. Lalic in The Griinfeld for the Attacking Player), whereupon ... jLb5 can no longer be prevented. I did spend quite a bit of time investigating 16 Wd3 a6 (after 16...b5 17 ±d5 White has a very strong position) 17 Hbl! ±b5!? 18 Hxb5!! axb5 19 JLb3, whereupon White has a strong attack, but sufficient defence can be found by a capable computer.

16...i.xh6 17 Sxh6 «fg5 18 2h2. Now, Black must essentially continue with 18...b5, as 18...i.b5 19 f4 wins a piece. After 18...b5 play continues 19 f4 Wg4 (on 19...Wf6 20 e5 Wf5, 21 *f2!! with the threat of Whl wins, and 20...Wg7 leaves Black in a pitiable state) 20 JLd5 Sxcl (20 &c6 loses immediately to 21 £.xc6 Hxc6 22 Sxc6 £>xc6 23 d5) 21 Wxcl Sc8 22 Wd2 i.c6 (22...e6 23 ±b3 leaves Black totally bankrupt with a bad bishop, and queen in danger of being trapped) with the following position (D):

Black has several possible defences, the most active one being

23 f5!! gxf5 (needed to deal with the threat of Wh6) 24 Sh3! ±xd5 25 exd5!! f4! 26 £>xf4 fobl 27 £>h5! Wf5+ 28 *gl Wxd5 29 We3!!! with the amazing threat of £tf4 followed by Wg3+ winning. Neither I nor my computer could find any defence in this position.

There are other potential defences for Black after 16 JLh6, and it would be premature to pronounce Black as being lost. However, 16...JLb5 cannot be played because of 17 jLxb5 ±xh6 (17...«fxb5 18 £.xg7 &xg7 19 Wd2 is overwhelming) 18 Sxc8 Sxc8 19 Ad7 Sc7 20 Sxh6 Sxd7 21 ®b3 with overwhelming threats on the kingside, e.g. 21...Wa6 22 Wh3 Sxd4 23 Sh8+ &g7 24 Wh6+ *f6 25 e5+ &xe5 (25...*e6 26 Wh3+ comes to the same thing) 26 We3+ Se4 27 f4+ and the rook is lost.

But after 18 Sh4! (it is important to keep the option of using the rook in front of the pawns) 18...^c6 (after 18...Wg5 19 Sh2 the position is like the one analysed above, except that the king is more exposed on g7) 19 Hf4! f6 20 e5 Sh8 21 *gl and now White has a very strong initiative in the centre (e6) and on the kingside despite the fact that the h-file has changed hands. Also, Black is unable to play 18...jLb5 in view of 19 ±xb5 Wxb5 20 Wd2 with a winning attack for White. So Black is reduced to purely defensive moves and White is very strongly placed.

Lines where Black plays

The line 10...^a5 is a natural defence, but it fails to put sufficient pressure on the centre. Black will, indeed, drive the c4-bishop off its most important diagonal, but the attack h4-h5 proceeds anyway and Black has no meaningful play against d4 and must content himself with relatively meaningless gestures against the queenside.

A typical continuation would be 10...^a5 11 &d3 Ml (there is no better move: ll...e5 is met by 12 dxe5; and ll...cxd4 12 cxd4 gives White everything he could ask for in the centre while the c8-bishop still has no better square than d7). These lines make it quite clear that if Black is to survive here, he must play a line involving ...Wa5.

Other passive lines

The move 10...JLd7 has been suggested, but this is very slow. White simply plays 11 h4! Wa5 (on 1 l...h5, 12 is very strong; and on

1 l...^a5 12 Ad3 it is hard to suggest anything for Black as 12...e5 is met simply by 13 dxe5) 12 h5 b5 13 Ad5! e6 14 Ab3 cxd4 15 £>xd4 and Black's counterplay is very slow and weak.

If Black does not react strongly, White will continue with the attack h4-h5 as previously explained. With the very difficult to find introductory move 10 Sell, which defends the centre, everything else proceeds according to plan.

In Summary

We have gone to great lengths to assure that all the above is valid.

a) All the moves do qualify as System moves as set down in this book.

b) The analysis has been carefully checked by computer and other qualified players.

Although it is somewhat early to make the following claim (as one should allow time for testing in GM practice), I believe that the above analyses basically refute the main line of the Griinfeld Defence.

The Slow Variations

Therefore, the only meaningful course for Black, if he wishes to play the Griinfeld, would be to play one of the variations that does not immediately counter-attack with 6...c5. In this, White must be continuously wary of transposition into the main line where one or more of his pieces have been misplaced. Thus, the transposition rule comes into play a great deal. For instance, if after 6

bxc3, Black plays 6... JLg7 instead of 6...c5, then we certainly have enough understanding to know that we should now play 7 JLc4, as this would lead after 7...c5 to the same position as if Black had transposed his 6th and 7th moves.

However, things will not remain that simple. White must develop his position:

a) According to System principles;

b) Taking into account that Black may transpose into the counterattack variation at any time;

c) Yet developing his attack against the kingside in such a way as to take into account the two constraints above.

The observant reader may have noticed that the lines against the Griinfeld are full of examples of the Response Pairs principle. We now restate the principle, first defined in Chapter 2, p.37, and then go to examples. The Response Pairs principle states:

There may be openings in which certain moves are required as responses to Black's moves. This will result in pairings of the type: black move -» white reply.

This will almost always be because the black move prepares some counter-attack, and the response is the System way of preventing it. There may be several counter-attacks, and each will have its own response.

Let us see how this works for the analysis just concluded. According to the Options principle, White would ideally like to develop his pieces in the following way:

a) First, jLc4 as this is where the bishop belongs, and it takes no options away from any other piece.

b) Second, ¿htl as the knight belongs here to avoid the pin ...¿.g4, and because it is still not clear where the cl-bishop should go.

c) Third, the c 1-bishop when it is clear where it should go.

However, the Response Pairs principle intervenes to change this order at times. These are the issues:

a) Black's ...c5 must be answered by JLe3 (unless it is already there). This is so the cl-square becomes free for the al-rook in case of ...Wc7 threatening ...cxd4. There is an exception to this rule, and that occurs when Black plays 6...c5. Then, we can play 7 jLc4 because now 7...Wc7 is answered by 8 Wb3, forcing a major weakness with the forced reply 8...e6.

b) White must be prepared to defend the d4-pawn as many times as it is attacked. However, since it is defended by a pawn, Black's attacks are meaningless unless ...c5 is part of the attack. So as long as ...c5 has been (or is ready to be) played White must be able to counter:

c) However, if Black were to block the c7-pawn before it can advance, then the defence of the d4-pawn is securely in the hands of the c3-pawn, and the above rules need not be invoked. In a line such as 6...±g7 7 !.c4 0-0 8 £>e2 £>c6, it would appear that White should now turn his attention to where to attack, since the centre is no longer in danger. However, this is not quite true, since Black can at some future time play ...£}a5 followed by ...c5 when the white bishop moves. So the Response Pairs rules remain valid after 8...^c6. However, if Black were to choose some deployment starting with the fianchetto ...b6, which loses a move in the attack on the centre without any transposition possibilities, White would now be free to start his attack.

d) Although it will not come out in this analysis, except possibly in unmentioned sidelines, it is necessary that any time Black plays ...f5 that this be met by an immediate e5, which will hopelessly cramp Black and leave his dark squares very weak.

There is one more thing that needs to be said about the slow variations of the Griinfeld. If White castles, then his attack against the king via h4-h5 disappears. Thus, Black's main hope is that White will castle, and the game will turn into a positional struggle. However, having seen that the attack will carry the day, even in the face of aggressive counter-action in the centre, it should be clear that White should just proceed with his kingside attack and forego castling.

Now let us go into detail about the correct System lines in the slow variation (omitting 6...c5). The reason we first examined the immediate counter-attack by Black is that it forms the basis of how the play will go given the transposition rule and the set of Response Pairs. If Black does not take aggressive counter-measures, but makes a move that could fit into some future counter, then White must play in such a way as to be able to transpose to known System positions.

Thus, Black's slow development will hinge on one of two ideas. Either, he will play ...^c6 with the idea of at some point playing and after White plays .&d3, then ...c5. Or he will play ...b6 with slow pressure on the centre. In both cases the preliminary moves 6 &g7 7 Ac4 0-0 8 £>e2 will first be played. Now Black can continue 8...£}c6, when 9 .&e3! must be played in order to answer 9...£ia5 10 ±d3 c5 with 11 Scl!, when we are back in one of the main lines. So Black usually plays 9...b6, when the attack with 10 h4! (D) can begin.

Black has no long-term hope of countering this. If 10...£>a5 11 £.d3

e5, then 12 h5 We7 13 Wd2 exd4 14 cxd4 i.b7 15 f3 and White stands clearly superior. One possibility is 15...C5 16 JLh6 ±xh6 17 Wxh6cxd4 18 e5! with a very strong attack. Clearly the position after 13 ®d2 (D) is full of possibilities. However, White has played correctly to this point, and the dynamics favour him. I would feel confident to play this in a correspondence game where one had (say) a week to work on each move.

If above instead of 9...b6 Black plays 9...e5, then 10 Wd2 gets into all the correct transpositions.

Also, 8...b6 is met by the immediate 9 h4, which resulted in a convincing win for White with the further 9...e5 10 h5 exd4 11 hxg6 hxg6 12 cxd4 We7 13 Wb3! in de Carbonnel-Messere, 5th World Correspondence Ch Final, 1965-8.

il mzMkmz

One Last Variation

One other variation still needs attention. Suppose Black plays 5...^b6 instead of 5...£>xc3. Here the books in their typical insipid manner recommend 6 h3, so that the centre can be defended without allowing the pin ...jLg4. However, it would seem that White should be able to do that without wasting time with such a silly move. The problem relates to where to deploy the fl-bishop. The best square for this bishop is c4, and that has now been denied. So where is the bishop to be placed? According to the theory of options it is not at all clear where it should go. On d3 it blocks the defence of the d-pawn from behind, and on e2 it is really not accomplishing anything it does not do on f 1. Maybe, we should wait before deploying the bishop, and see if there is another way to proceed.

White can develop while defending his centre with 6 È.t3\ È.gl 7 Wd2! £>c6 8 0-0-0 (D).

Now his centre is completely secure, and he can proceed with a kingside attack starting with h4.

While it has not been possible to deal with every variation in this analysis of the slow variations, I believe the major lines have been dealt with adequately. Once White is able to play h4-h5, he has a strong attack, and unless his centre has been weakened in the process, he should be able to carry the day with the king-side attack.

The King's Indian Defence

What is the Correct Development Plan?

After the moves 1 d4 £tf6 2 c4 g6 3 £ic3 JLg7 4 e4 d6 (D), the weakest point in White's pawn structure is the pawn at d4, which cannot be defended by either of its neighbours.

With his last move Black served notice that he intends eventually to w

IJt.1 W+m 1 iii liii

challenge that pawn and particularly the square it is on with either ...c5 or ...e5. Now it is important to note that if White were to answer such a challenge by either capturing the challenging pawn or advancing d5, Black would be left in full control of the d4-square, and that is essentially the main point of Black's strategy. True, if White chooses to play d5 it is unlikely that Black will ever get a knight to d4. But that is not as important as the fact that the square will not be available to the white pieces themselves. Therefore, it would seem a superior strategy for White to be able to meet the coming attack on his d4-pawn by being prepared to defend it with pieces as often as necessary to maintain the tension.

If White can maintain his d4-pawn by adequately defending it without otherwise compromising his position, the play will turn in his favour since he already controls most of the light squares in the centre. Very importantly, move 5 is the time when White must make key decisions as to how to deploy his pieces. It's therefore not surprising that very many alternatives have been tried in this position in master play: 5 5 Ae2, 5 f3, 5 f4, 5 Ag5, 5 Ad3, 5 £}ge2, 5 g3, and 5 h3. We will first examine this list to see which of the above moves qualify as possible System moves.

a) 5 £tf3 cannot be right because it blocks the f-pawn in a situation in which it is not at all clear whether that pawn has a role to play in the fight for the centre.

b) 5 JLe2 just can't be right. Why move that bishop to that square? It is not clear where the bishop belongs, and this is very premature.

c) 5 f3 and 5 f4 are both good candidate System moves. Each commits the pawn to a special role, and we will examine the pros and cons of these moves later.

d) 5 jLg5 again prematurely commits a piece to a square of dubious distinction.

e) The same is true of 5 JLd3, which also leaves the d-pawn unprotected at a critical time.

f) 5 £ige2 could very well be a good move, but it must be premature to make it at this point before any other development decisions are taken. It could be that one wishes to play f4 followed by £>f3 in the future, and £ige2 is too committal at this point.

g) 5 g3 is ridiculous because it commits the bishop to a dubious diagonal.

h) 5 h3 is ridiculous since it is a defensive move in a situation where there is no attack.

There are many who would argue with our cavalier dismissal of 5 It has probably been played more often in this position than any other move, so how can I dismiss it out of hand? The answer is simple. It is not a System move. It may be a perfectly good move that forces Black to play accurately to maintain equality; however, we are after more than equality. We want to pursue the System ideas to get an advantage.

After 5 play normally proceeds 5...0-0 6 i.e2 e5! (D).

The fact that Black can get away with this move at such an early stage should make one suspicious of White's strategy. The point is that White cannot win a pawn by 7 dxe5 dxe5 8 Wxd8 Sxd8 9 £>xe5 £>xe4!, and if now 10^xf7?then 10..JLxc3+ wins. Therefore, White usually continues more quietly with a move such as 7 0-0. Now, 7...£>c6 (D) already begins to pose serious threats to the white centre.

For example, Black threatens to play 8...exd4 9 ^xd4 £>xe4! 10 £>xc6 £\xc3 11 ^xd8 ^xdl, when he has won a pawn. This activity is very different from what Black achieves when he plays ...e5 in the 5 f3 line. Then the e4-pawn is well defended, and White need only concern himself with making sure d4 is also well protected.

The reader should familiarize himself with the above combination as this is the heart of how to play hypermodern defences. Because of this threat, White now usually plays either:

a) 8 d5 £>e7, after which Black will, by playing ...£>e8, seize the initiative on the kingside with ...f5; or b) 8 J.e3, after which Black has little difficulty equalizing by 8...^g4. The latter move points to one of the difficulties that White frequently faces: the queen's bishop belongs on e3 if he wishes to play to maintain his centre. However, Black can frequently counter with ...£>g4, putting pressure on d4 and threatening to exchange White's best minor piece.

The 5 f4 line

5 f4 was one of the historically first moves to be tried. This came in the days before the hypermodern ideas were very much understood, and appears to yield White a terrific centre which will sweep everything before it. But appearances are deceptive. In actuality, the move also weakens the e4-pawn, which can no longer be defended by a pawn. Thus, after the moves 5...c5 6 d5 (if 6 dxc5, then 6...1ra5 is good for Black) 6...0-0 7 &f3 e6 (D), Black will be able to open up the e-file to his advantage.

White's play above certainly seems consistent, and it took me a long time to decide between 5 f4 and 5 f3. The position after 7...e6 in the above line seems to offer White many

IrnkM 3NrP SAH nil a iiiAH

hahah is

0 0

Post a comment