No. 14, 31 July 1999

Dear Chess Friends!

   August is going to be a very busy month for chess fans with the World Championship starting in Las Vegas today. It will be also a busy time for me as I will play in the British Championship (Scarborough, 1-14 August) and in Mind Sports Olympiad (London, 21-29 August). So, in this issue I decided to take it easy and to review some of the previously covered material. But first I want to give a couple of tactical positions and provide some useful links.

Sharpen your tactical vision!

   Consider these 2 positions - this may help you to spot tactics in the games, played in Las Vegas! :) You will find the answers in the end of this issue:

White to play

Black to play

Links

   More sites post CBC now and one of them - by Ruben Casafus - might be particularly interesting for those whose first language is Spanish. You can find it at: bigfoot.com/~ruben-casafus. Chess is getting more popular in India and this reflects on the Net - to learn about chess in that country visit the following interesting site www.chennaionline.com/chess. Much as I liked India and its people during my visit there in April, I am very unhappy about the way All India Chess Federation and its Secretary (he is also a Vice-President of FIDE) Mr. Ummer Koya treated me during the Commonwealth Chess Championship. I had a written contract with the organisers, but they still managed not to pay my appearance money. Another victim was GM Miles, who has written about that at Chess Cafe, of which he recently became a columnist. See his story (all true!) at: chesscafe.com/miles/miles.htm. I am going to cover that story in chess press too, so we are not going to forget about that (considerable!) money. The Indian Chess Federation is badly damaging the image of Indian chess, which is really a shame ... Recently I came across yet another interesting site: www.freespeech.org/chessdate. Hope these links will entertain you this weekend! :) Now let's come back to chess!

Kramnik-Timman, Wijk aan Zee 1999

   After I showed Kramnik's analysis of his endgame vs. Timman and suggested a few possible improvements, I heard from Vladimir, who reviewed my suggestions (that was later published in 64 magazine). I also heard from other people, so I would like to come back to that fascinating endgame in order to share with you some new findings.

First of all, I'd like to show the game and the moment when the rook endgame arose. For those who did not see the analysis in CBC No.6, I recommend to have a look at that issue first.

V. Kramnik - J. Timman
Hoogovens Wijk aan Zee, 1999

1 Nf3 c5 2 c4 Nf6 3 Nc3 Nc6 4 g3 d5 5 d4 cxd4 6 Nxd4 dxc4 7 Nxc6 Qxd1+ 8 Nxd1 bxc6 9 Bg2 Nd5 10 Ne3 e6 11 Nxc4 Ba6 12 Na5 Bc5 13 Bd2 0-0 14 Rc1 Bd4 15 b4 Bb5 16 Nxc6 Bb2 17 Rc5 Bxc6 18 Rxc6 Rac8 19 Rxc8 Rxc8 20 Bxd5 exd5 21 f4 Ba3 22 Rf1 Rc4 23 Rf3 Bxb4 24 Bxb4 Rxb4 25 Ra3!

25...h5 26 Rxa7 Rb2 27 a4 Ra2 28 f5! Ra1+ 29 Kf2 d4 30 a5 f6 31 Kf3 Kh7 32 a6 Kh6 33 h4 g6 34 fxg6 Kxg6 35 Ra8 Ra2 36 Kf4 Kf7 37 a7 Kg7 38 Kf5 Ra5+ 39 Ke4 Ra4 40 Kd5 Ra1 41 Kxd4 1-0

First of all, Vladimir agreed that instead of 35...Ra2? Black could survive by playing 35...Kf5! 36 Kf2! Ra2 37 Ke1 Kg4 38 a7 Kh3 39 Kd1 f5! 40 Kc1 Kg2 41 Kb1 Ra4 42 Kb2, as my recommendation - 42...Rb4+! 43 Ka3 Rb7! - works fine and Black draws after 44 Ka4 Rg7! 45 Kb5 Kxg3 46 Kb6 Rg6+ 47 Kc5 Rg7 48 Kxd4 Kxh4 49 Ke5 Kg5.

Now Kramnik considers 34 fxg6 to be a serious mistake and recommends 34 Ke4!. Indeed, this move is strong and seems to win: 34...gxf5+ 35 Kxf5 (Vladimir also mentions 35 Kxd4!?, which is even more convincing). From a practical point of view here Black should try 35...Ra2!? (Kramnik's line 35...Rf1+ 36 Ke4 Rf2 37 Kxd4 Rxe2 38 Rc7 Ra2 39 a7 Kg6 40 Kc5+- does not offer Black any chance). Then the hasty 36 Kxf6 allows Black counterplay after 36...Rxe2, as now he has a passer too. After 37 Ra8 (or 37 Rd7 Ra2 38 a7 d3=) 37...Kh7! 38 Kg5 Re5+ 39 Kf4 Re6! 40 a7 Ra6 41 Ke4 Ra4 42 Kd5 Kg7 43 Kc5 Ra1 44 Kxd4 Kh7 45 Ke5 Kg7 we reach the following position:

Here two extra pawns do not mean a winning advantage, as White cannot win the h5-pawn or utilise his a-pawn. This all looks sweet for Black, but after the correct 36 Ke4! Rxe2+ 37 Kxd4 White still wins, for example: 37...Re6 38 Ra8 Kg7:

The same position arose in one of the lines, which I considered in CBC No.6. There I wrongly claimed that Black was OK in the diagrammed position. In fact, Black loses here after 39 Kd5 Rb6 40 Kc5 Re6 41 Kb5 Re5+ 42 Kc6 Re6+ 43 Kd7 Rb6 (or 43...Re3 44 Rb8 Ra3 44 Rb6 Rxg3 45 Kc7, winning) 44 Kc7 Re6 45 a7. Now both 45...Ra6 46 Kb7 and 45...Re7+ 46 Kd6 Rb7 47 Rg8+ fail. Kramnik showed this line, but even before I heard from him, one of my students - 16-years old Sam Collins from Dublin - pointed out to me the same variation.

Instead of 33 h4 GM Philipp Schlosser suggested 33 Ke4!?. Indeed, this move seems to win easily (e.g. 33...Re1 34 Kd3 Rd1+ 35 Kc4 Re1 36 Re7!) and therefore it's better than 33 h4. However, 33 h4 was not really a mistake, as only 34 fxg6? was a bad slip, while the correct 34 Ke4 would still have kept White in control, as our previous analysis proved.

During analytical work I often caught myself trying to convince myself that some line is the 'final truth', etc. I guess this is in human nature, as we need to have some solid ground. Yet, chess often proves to be like an ocean - variations are too numerous and often cannot be calculated 'till the end'. This ending is a good example, as even after such a long analysis it's hard to claim that White had a win after 25 Ra3!. Kramnik gives 26...h4 and 27...g6 as possible improvements for Black, while I want to suggest yet another one: instead of 25...h5 Black should consider 25...g6!?, which gives him an escape square and prevents f4-f5. After 26 Rxa7 Rb2 we reach the following position:

Here Black might be able to play ...h5 later on and then ...Kg8-g7-f6 - something he could not do in the game. Of course, 25...h5 looks active (and obviously that's why Kramnik gave it '!' mark), but it later forced Black to damage his pawn structure on the kingside in order to get his king out of the back rank. Note that White cannot really block the enemy king here, as 27 g4?! will be met with 27...Rb4!.

On this note I'd like to finish this review of this great endgame, which showed us one more time how deep chess is! I bet computers would not play such endings well enough yet in many years to come! Well, that's my secret hope, anyway... :)

Solutions

1. Solovjov-Akhmadeev, St. Petersburg 1999: 31 Ra6! This is better than 31 Qh8+ Ke7 32 Rxb7+ Rxb7 33 Qxc8 Qb6+ 34 Kh1 Rb8 35 f6+ gxf6 36 Qh3. 31...Qxa6 32 f6!+- Ke8 33 f7+ Kd7 34 f8Q Re7 35 Qh3+ Kd6 36 Qg3+ 1-0

2. Kiselev-Gubajdullin, St. Petersburg 1999: 30...Rc1! 0-1

Hope you enjoyed this issue - I will be back in mid August with CBC-15. Meanwhile enjoy the fight in Las Vegas!