KASPAROV,G (2815) -
KRAMNIK,V (2770) [C67]
WCCwch London (3),12.10.2000
[Kramnik created a whole new match situation by winning the 2nd game convincingly, having drawn the 1st without much difficulty with Black. Kasparov hasn't lost an official game with regular time controls since January 1999, so, naturally, his immediate reaction in the 3rd game will be crucial. On such occasions Karpov used to calm down with a safe draw, even if this meant failure to take advantage of the White pieces. He knew fully well that there was enough time to take revenge and he kept working at it. Of course, this was in the days of 24-game matches.
The World Champion is a different species and has often won decisively immediately after a defeat, thus redressing the balance. Therefore, the choice of opening will be crucial: it shall demonstrate both his psychological state and his overall creative approach to the match.
Kasparov may continue as if nothing has happened and apply a strategy similar to the one he used against Anand: knowing it will be a long match, he may test all his opponent’s basic defenses and then strike against what he perceives as the weakest one. In that case, it makes sense to prefer 1.d4.
Another option is to go for a theoretical duel with 1.e4, which Kramnik should accept.]
[This is all according to my predictions on the Hellas Chess Club site: the Sicilian must be kept for a later date and would suit Kasparov fine at this moment. The challenger has already indicated that he is ready for the Scotch, but if Kasparov persists with the Ruy Lopez, then Kramnik should enter some main line.
The purpose of such a change is twofold: first he must keep his opponent in shifting ground (psychological reason) and second he must demonstrate mastery of the classical Ruy (chessic reason). Nobody ever became World Champion without deep understanding of openings like the Ruy Lopez and the Queen’s Gambit (in later days, the Sicilian as well, while even later the King's Indian was included in the list).
In any case, it will be interesting to see if Kramnik will be able to succeed in maintaining two trends I identified in games 1 and 2: disallowing fruitful pawn breaks and transposing quickly to endgames.]
2.Nf3 Nc6 3.Bb5 Nf6
[I have to admit this came as a surprise to me. Although I do not consider the variation at all bad, it does not seem to have sufficient stature to be played repeatedly in a World Championship match. Either Kramnik really believes in it and has incorporated it as a crucial weapon in his opening repertoire, or he considers its subjective merits worthy of the risk.
Time will tell, of course, but for now the ball is in Kasparov's court!]
4.0-0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8+ Kxd8
[The general ideas in this variation, for both White and Black, are presented extensively in my commentary to first game of the match.]
9.Nc3 Bd7 10.b3 h6 11.Bb2 Kc8 12.Rad1
[The first new move in this championship, but not in grandmaster practice. As mentioned in the comments to Game 1, this was previously played in Shirov-Krasenkow, Polanica Zdroj 2000 (August 20), which is given there in full. For some reason which may become clear later, Kasparov refrains altogether from the idea of advancing his kinside pawn majority, at least for the time being.]
[Kramnik is the first to vary from that game, although what counts here is not so much individual moves, but ideas. Transpositions occur quite easily, because the two armies have little immediate contact and pieces of both sides may reach the same squares in different move orders.
Krasenkow chose 12...a5 and only after 13.h3 did he play 13...b6 , so that in reality it is Kasparov who makes a new choice with his plan in the next few moves.]
[So here it is: the World Champion refrains from moving the h-pawn at all. He will concentrate on piece play, hoping for domination of space and exchanges that will eventually reveal the power of his pawn majority deep into the endgame.
Also, the manouvre ...Bf8-b4 now becomes completely meaningless.]
[As in the first game, the Ra8 cannot be developed without moving the King to b7, which necessitates the removal of the B from the d-file. My comments there show why its transference to e6 is inferior, so Black has no alternative but to deploy it on the long h1-a8 diagonal. At the same time Kramnik prevents the use of d4 by the enemy Ns, but in return temporarily surrenders d5.]
[Although Kasparov invested a few minutes on this move, it is practically forced. After 14.Nf4 there would follow 14...a5 , threatening to develop the R economically through the a-file, and on 15.a4 Black would acquire the breakthrough 15...c4 .]
[Kramnik could also essay 14...a5 , once again provoking 15.a4 . At this point in the game, it is hard to tell whether blockage of the queenside and weakening of the Pb3 counterbalance each other completely or favour one side more than the other.
Perhaps the challenger prefers to wait for a more opportune moment, when a2-a4 would be a clear disadvantage for White. On the other hand, at that point Kasparov may simply refrain from it and the opportunity to provoke it would be gone for ever.]
[Kasparov declares once and for all that he does not mind doubling his Ps and crippling his majority, if he gets the white-square B in return. In that case the superior development of his pieces is bound to tell, for example Black would have to keep the King on c8 in order to guard the invasion square d7.
Naturally, Kramnik is not at all interested in relinguishing early on the main advantage he obtained from the opening. The result of this behind-the-lines conversation, however, is that a N will soon appear on d5 and Kasparov will save considerable time and energy in comparison to Game 1.]
[The 1-point question, of course, is whether this apparently dominant N is accomplishing anything productive. For the time being it prevents the completion of Black's development, as ...Bf8-e7 is now quite dangerous. On the other hand, the N blocks the d-file and no longer supports the advance e5-e6.]
[Eyewitness reports reveal that Kramnik had used up to now almost half an hour more than Kasparov. This is a first for the match, and a psychologically important first at that.
Kasparov took great care not to fall behind on the clock in his previous match against Anand, as the Indian is well known for his speedy play. This concern should not be equally important against Kramnik, but it is a measure of control of the proceedings, anyway.
Besides, the time spent on a position often reveals the extent of difficultie faced by each player, either to avoid disadvantage or to capitalize on a minute superiority.
Returning to the board, the N retreat follows Kramnik's ideas from the 1st game. It seems the challenger has investigated the position in depth and concluded that this particular rearrangement of forces is the most productive for Black. Nevertheless, I will venture to propose that it is not the only one, especially given the fact that Kasparov has done absolutely nothing to use his kingside pawn majority.
Another very reasonable idea is 16...a5 (now is a good time to execute this advance, since the R will soon leave the a-file) 17.a4 Re8 , simply continuing the process of development. The fundamental premise of such a plan is that Kasparov will soon reach the point of maximum mobilization and will find it hard to improve his position using only piece manouvres. Gradually Black will catch up and establish complete equality.
A typical sample continuation would be 18.Rfe1 Be7 19.Nxe7 (otherwise the B may even retreat to d8, when manouvring behind the lines becomes much easier for Black, but this would be the normal course of events in this variation) 19...Rxe7 20.Rd3 Rhe8 (one idea now is ...f7-f6) 21.g4 (for which reason White could also try 21.Kf1 ) 21...Bxf3 22.gxf5 (worse is 22.Rxf3 Nd4) 22...Bg4 23.Rg3 Bxf5 24.Rxg7 Bg6 and ...Re8-d8. Naturally, this variation is far from forced, but I think it indicates the resilience of Black's position when White restricts the play on piece manouvres.]
[Kasparov has completed the mobilization of his pieces and may soon attempt the breakthrough e5-e6. This would cede the square d6 to the enemy black-squared B, but the removal of the Pe5 from the board will open wide new possibilities for 3(!) of the Champion's pieces.]
[A rather unexpected move at this point, although typical in this variation. Kramnik obviously wishes to avoid any unpleasant surprises on the a1-h8 diagonal and the Pg7 in particular, but it is unclear whether this fits in with his previous move. A more consistent option is 17...Rd8 , adding more pressure on the dominant Nd5, but it runs into 18.e6 when Black faces a difficult choice: comparatively best is 18...Bxd5 (18...fxe6 19.Nf4 Rxd1 20.Rxd1 when invasion on d8 is impossible to prevent without fatal concessions 20...Nf5 21.Rd8; 18...Nxd5 19.exf7 and White should win without much difficulty; 18...f6 (trying to keep lines closed) 19.Nf4 Rxd1 20.Rxd1 Kc8 21.h4 Rg8 22.Kf1) 19.cxd5 (but not 19.exf7 Bxf3 20.Rxd8 Nc6) and now practically forced is 19...f6 (of course not 19...fxe6 20.dxe6; or 19...Nxd5 20.exf7 c6 21.Ne5 etc.) 20.Nh4 Rg8 transposing to the line analysed immediately below.;
Since all this could be predicted by Kramnik, he must have also examined the variation 17...Bxd5 18.cxd5 Rd8 . Now White has an imposing centre with apparently sufficient piece support, but cannot maintain it. Black's problem is that Kasparov would not want to, in any case, since its primary function is to advance at the appropriate moment, which seems to have arrived: 19.e6 f6 (the variations mentioned above demonstrate the necessity to keep closed as many lines as possible) 20.Nh4 A) much better than the immediate 20...Rxd5 , when the forced sequence 21.Rxd5 Nxd5 22.Ng6 Rg8 23.e7 Bxe7 24.Rd1 (of course, not 24.Nxe7 Re8) 24...c6 25.Nxe7 Re8 26.Nxd5 cxd5 27.Kf1 Kc6 28.Re1 leads to a technical win for White; B) 20...Rg8 (a surprisingly productive move for White) 21.g4 Rxd5 (after 21...Nxd5 22.Ng6 there is no adequate defense to 23.e7) 22.Nf5 (now 22.Rxd5 Nxd5 23.Ng6 Bd6 proves insufficient) 22...c6 23.Ne3 (an attacking retreat reminiscent of Kasparov's manouvres in the 1st game) and White infiltrates the enemy camp with disicive consequences, e.g. 23...Rxd1 24.Rxd1 Nd5 25.Nxd5 cxd5 26.Rxd5 Kc8 27.Rd7 etc.
All these variations demonstrate the problems Black faces every time the position opens up before development is complete. If 17...Rg8 is the best move in the position, I believe the alternative I suggested earlier deserves closer scrutiny.]
18.Nf4 g5 19.Nh5 Rg6 20.Nf6 Bg7 21.Rd3 Bxf3 22.Rxf3 Bxf6 23.exf6 Nc6 24.Rd3 Rf8 25.Re4 Kc8 26.f4 gxf4 27.Rxf4 Re8 28.Bc3 Re2 29.Rf2 Re4 30.Rh3 a5 31.Rh5 a4 32.bxa4 Rxc4 33.Bd2 Rxa4 34.Rxh6 Rg8 35.Rh7 Rxa2 36.Rxf7 Ne5 37.Rg7 Rf8 38.h3 c4 39.Re7 Nd3 40.f7 Nxf2 41.Re8+ Kd7 42.Rxf8 Ke7 43.Rc8 Kxf7 44.Rxc7+ Ke6 45.Be3 Nd1 46.Bxb6 c3 47.h4 Ra6 48.Bd4 Ra4 49.Bxc3 Nxc3 50.Rxc3 Rxh4 51.Rf3 Rh5 52.Kf2 Rg5 53.Rf8 Ke5 ˝-˝
[A N retreat reminiscing the Champion's manouvres of the 1st game, reportedly after nearly 30 minutes of thinking by Kasparov. The obvious move is 18.e6 , but then 18...fxe6 (18...f6 19.Nh4 Bxd5 20.cxd5 Rd8 21.g4 transposes to the analysis of the previous note) 19.Rxe6 Nxd5 (the capture 19...Bxd5 allows White to nurse his development advantage into the endgame after 20.cxd5 Rd8 21.d6 cxd6 22.Rdxd6 Rxd6 23.Rxd6 etc.) 20.cxd5 Bd7 allows Black to solve most problems in a very satisfactory way.]
[Kramnik could not imitate the N retreat with 18...Nf5 , as then he would have played the useless move ...Rh8-g8 against Rf1-e1 by Kasparov. The pawn advance is clearly risky, but is now the only reasonable way to activate the black-squared B and its neighbouring R.]
19.Nh5 Rg6 20.Nf6
[Otherwise Black will play ...Rg6-e6 and feel much more comfortable. A big part of the battle to come will undoubtedly revolve around this R's ability to participate actively in the fight.]
[Another forced move, to prevent 21.e6. The game is clearly in a transition phase, provoked by Kramnik's radical 18...g5 advance, at the end of which the position will have very different characteristics.]
[A sensible, multi-faceted move that more or less blackmails the following exchange. Bringing the R on the 3rd rank serves various functions, aside from the obvious intent to double on the d-file: in the first place it protects the N, while in addition it may be used on the kingside, should lines open there all of a sudden.
An equally "Soviet" preparatory action is 21.Kf1 , removing the K from the g-file and bringing it closer to the centre.]
[In changing the nature of the position and bringing it closer to a real ending, Kramnik relinguishes his main advantage: the white-squared B. The only way of keeping the tension is 21...Rf8 22.Red1 Kc8 , not an attractive possibility when he has only about 30' minutes to reach move 40. In such a situation it is actually Black who has maximized the effectiveness of his pieces and only White can make serious improvements, but at least there would exist the threat 23...Bxf6 24.exf6 Ng8.
Although the challenger's practical choice appears prudent under the specific sporting circumstances, it is probably worth it to go a little deeper in this position. Such "gray" areas in important matches are not few and far between, but there is never enough time to examine all of them in sufficient depth. Therefore, I will limit myself in pointing some of them out.;
Nevertheless, it is worth adding that the immediate 21...Bxf6 22.exf6 Ng8 fails against 23.Ne5 Rxf6 24.Nxc6 Rxc6 25.Rd7 , when the extra pawn in meaningless but the domination of open lines of paramount importance, e.g. 25...f5 26.Rf7 f4 27.h4 and the combination 2Rs+B is definitely superiori to 2Rs+N in this kind of position.]
[The alternative capture 22.gxf3 allows the forcing sequence 22...Bxf6 23.exf6 Ng8 24.Rd7 Nxf6 25.Rxf7 Ne8 and 26...Nd6, after which most of Black's serious problems are solved.]
[Now that the Nf6 has twin protection the positional threat 23.e6 has been revived, in addition to 23.Nh5 which opens both the f-file and the a1-h8 diagonal. Therefore, yet another exchange of B for N is forced and the game soon enters a completely new stage.]
[The time has come to reevaluate the position, with White seemingly enjoying a distinct superiority. Every one of Kasparov's pieces is far more active than its enemy counterpart, the only queston being whether Black will be able to play at some point ...Nc6-d4 in order to force further liquidation. Otherwise, it will be much easier for the White Rs and King to improve their positions even more and initiate a decisive breakthrough.
Kasparov should concentrate his attention on the kingside, because that's where his King is. A timely h2-h4 or f2-f4 will allow it to attack some enemy pawn, but this should be done carefully, since at the same time the enemy R will also be freed.
This endgame presents special interest in the way small asymmetries add up to destroy the equilibrium.]
[Necessary, otherwise Black will play 24...Rd8, ...Rd8-d6 and very soon ...Nc6-d4.]
[The R must make room for the King, who in turn will prevent occupation of the 7th rank.]
[Once again, it is clear that White cannot make much headway without a pawn break. Kasparov prepares both h2-h4 and f2-f4, but which is really the better of the two?]
[White cannot delay action any longer, since Black is by now fully ready for ...Nc6-d4, e.g. 26.Kf1 Nd4 27.Bxd4 cxd4 28.Rdxd4 Rxf6 29.Re7 Re6 30.Rxe6 (30.Rdd7 Rxe7 31.Rxe7 Kd8 is completely useless; , while 30.Red7 Rfe8 and ...Re7 is equally unappealing) 30...fxe6 and the pawn weaknesses are not enough for a decisive advantage. It is tru White could try a different initial move like 26.g3 in order to use the sub-variation with 30.Red7, but then a counterattack with ...Re6-e2 would be possible.;
Naturally, the other break 26.h4 needs to be considered carefully. After 26...Nd4 27.Bxd4 cxd4 28.Rdxd4 Rxf6 29.hxg5 hxg5 30.Rd5 (not 30.Re5 Rg6 with a minimal advantage, e.g. 31.Re7 Re6) the only correct reply is 30...Rd8 (the small but important difference is that after 30...Rg6 now follows 31.Re7 and Black cannot play 31...Re6 32.Rxe6 fxe6 33.Rxg5) , when 31.Rxg5 Rd1+ 32.Kh2 Rxf2 ensures full equality.]
[Apparently Kramnik feels it is necessary to aim for active piece counterplay. The variation 26...Nd4 27.Bxd4 cxd4 28.Rdxd4 Rxf6 would bring the game close to equality once again. Could it be that he was hoping for more at this stage?]
[The real question in this position is the advanced Pf6. It can certainly be dangerous as part of White's majority, but it also blocks the B and may become a target. Its value will probably be determined by the activity dispalyed by the Kings in the near future.
For the time being, Kasparov must take care of his 2nd rank.]
[A rare defensive move by the Champion. After 28.Kf2 Re6 (but not 28...Ne5 29.Re3 , since 29...Ng4+ is unplayable because of 30.Rxg4 Rxe3 31.Rxg6 etc.) 29.Rg3 Rxg3 30.hxg3 Black is just in time for 30...Nd4 to prevent the idea Rf4-h4.]
28...Re2 29.Rf2 Re4
[Kramnik's only problem seems to be that he has 18 minutes left to reach move 40 (inclusive), against 33 for Kasparov. On the board, the alternative 29...Rxf2 30.Kxf2 Nd4 31.Bxd4 cxd4 32.Rxd4 Rxf6+ 33.Kg3 could give him some extra headache, because of the idea Rd4-f4.]
[As a matter of fact, Kasparov cannot afford to remain idle, either. About 12 minutes of thought must have convinced him that the idea ...Nc6-d4 and ...Kc8-d7-e6 could make life dangerous even for him.
This is why he is redirecting his forces against the weakling Ph6.]
[The counterargument is that meanwhile Black has taken over the centre. Kramnik essays a counterattack that begs the question whether it would have proved much more useful to have the moves ...a7-a5 and a2-a4 inserted at an earlier moment.
With new complication coming up, it was reported that both players had 15 minutes available for the next 10 moves.
It was also worth considering 30...Ne5 ;
, as well as 30...Kd7 . After the latter, for example, White cannot play 31.Bd2 Ne5 32.Rxh6 Rxh6 33.Bxh6 Ng4 34.Rd2+ Ke6 and the mate threat wins the B.]
[Kasparov decides to prevent ...Nc6-e5, probably in anticipation of his next suprising move.]
[An astoundingly unorthodox decision, with which Kasparov practically abandons the queenside at the mercy of the enemy pieces in order to promote his chances on the opposite side of the board. He is surely hoping that in such situations the long-range B most often proves superior to the short-stepping N. On the other hand, every tempo in advancing the Ps may prove crucial.
Meanwhile, it is tru that both 32.Bd2 Ne5 and;
32.h3 axb3 33.axb3 Re3 34.Bd2 Rxb3 35.Rxh6 Rxh6 36.Bxh6 Ne5 also contained serious losing chances for White.]
32...Rxc4 33.Bd2 Rxa4 34.Rxh6 Rg8
[It is practically impossible to calculate all the consqeunces of an exchange like 34...Rxh6 35.Bxh6 in time trouble, but it seems quite clear that Kramnikfeared then the rapid advance of the h-pawn.]
[Of course, not 35...Rf8 36.Bh6 when White wins the Pf7 anyway, but without surrendering the Pa2.]
[Similar positions with so many passed Ps from both sides arise very rarely in World Championship matches. One that readily comes to mind is the 13th Spassky-Fischer battle (1972), which must have caused equal excitement.]
[I have to admit that I am unable to calculate all the intricacies of this endgame in the limited time available to me, so I will stick to what comments I am most sure about. I suspect the verdict on the tactical details of what follows shall be finalised after the match.]
37.Rg7 Rf8 38.h3
[For example, this is a move I do not understand. Kasparov probably wishes to avoid any unpleasantness of the ...Ne5-g4 type, but this is unnecessary when g4 is guarded by a R. 38.h4 looks much preferable to me, time trouble or not. One only has to calculate that 38...Nd3 is unplayble because of 39.f7 Nxf2 (or 39...Kd7 40.Rg8 Ke7 41.Rxf8 Kxf8 42.Bh6+) 40.Rg8 etc.]
[This manouvre justifies Kasparov's previous move, but I still think the rapid advance of the h-pawn was preferable.]
39...Nd3 40.f7 Nxf2
[The last move of the first time control is an accurate one. After 40...Kd8 41.Re8+ Rxe8 42.Bg5+ Kd7 43.fxe8Q+ Kxe8 44.Rxa2 Black would lose immediately.]
41.Re8+ Kd7 42.Rxf8 Ke7
[More flashy, but actually worse, was 43.Rd8 Kxf7 44.Kxf2 c3 45.Ke3 Rxd2 46.Rxd2 cxd2 47.Kxd2 with a draw in the offing. Kasparov is naturally interested in testing his opponent much more than that.]
43...Kxf7 44.Rxc7+ Ke6 45.Be3 Nd1 46.Bxb6
[The trap 46.Rxc4 , hoping for 46...Nxe3 47.Re4+, almost backfires after 46...Rxg2+ . Fortunately for White, he can still draw in that case with 47.Kh1 .]
[I would definitely prefer 47.Kh2 , since the Ps almost always need K support. Now Kramnik forces simplification to a theoretically drawn R endgame.]
47...Ra6 48.Bd4 Ra4 49.Bxc3 Nxc3 50.Rxc3 Rxh4 51.Rf3 Rh5 52.Kf2 Rg5 53.Rf8 Ke5
[53...Ke5 A likely continuation would be 54.Kf3 Rf5+ 55.Rxf5+ Kxf5 56.Kg3 Kg5 Definitely the most exciting game so far, so I am quite unhappy not to have been able to follow it live. In addition, it is the most difficult to analyse among the three, which probably means the tension is building up. This should favor Kasparov, who has vast experience in matches at this level, but so far Kramnik has demonstrated he can play at an equal level.]