Saturday is a rest day for the participants in the FIDE Grand Prix tournament in Zug (we're between playing days there, so is this a zwischenzug?), while the players at the Alekhine Memorial have Friday and Saturday off as their tournament moves from Paris, France to St. Petersburg, Russia. Let's take the opportunity then for a quick scoreboard update:
Zug: After eight rounds of 11, the standings are as follows:
1. Topalov 5.5
2. Ponomariov 5
3-4. Caruana, Karjakin 4.5
5-7. Kamsky, Morozevich, Nakamura 4
8-10. Giri, Leko, Mamedyarov 3.5
11-12. Kasimdzhanov, Radjabov 3
Two quick comments about ratings. After his loss to Viswanathan Anand in the 2010 World Championship match, Veselin Topalov's play and his rating took a bit of a nosedive, and his results right after a fairly long break weren't especially good either, as I recall. Lately though, he has been playing very well again, and he has worked his way back up to #4 in the world. Going in the opposite direction is Teimour Radjabov, who entered the Candidates' as the world's #4 player. He finished last there and his tied for last here, and has lost a staggering 46 rating points between the two events. (A note to readers only familiar with USCF or ICC ratings: FIDE ratings, especially for top players, are far more stable. Losing a game to a peer doesn't cost 16 points or so, but only around five points. So if you want to "translate" his last month into USCF "language", imagine your rating going down 150 points or so. Yikes!)
Round 9 Pairings (Sunday):
- Giri - Kasimdzhanov
- Leko - Karjakin
- Kamsky - Caruana
- Topalov - Mamedyarov
- Nakamura - Morozevich
- Radjabov - Ponomariov
Alekhine Memorial:
Here the standings are a bit of a shock. The leader is not world champion Viswanathan Anand, nor is it world #2 Levon Aronian. Vladimir Kramnik started with an impressive win in round 1, but since then it has been a bit of a nightmare for the ex-champion. One won't find the leader in the ranks of near-champions, whether of the absolute sort (Boris Gelfand) or the FIDE variety (Michael Adams). No, the leader after the French segment, five rounds into this nine-round tournament, is Maxime Vachier-Lagrave! Can he keep his lead to the end of the tournament? I doubt it, but with only Aronian left on his schedule among the super-big guns, he might not be as big an underdog as one might otherwise suspect.
Standings After Round 5:
1. Vachier-Lagrave 3.5
2-5. Adams, Aronian, Fressinet, Gelfand 3
6. Anand 2.5
7-9. Kramnik, Vitiugov, Ding Liren 2
10. Svidler 1
You might be surprised to see Fressinet up there, but he earned it by butchering Kramnik in round 5 - with Black, no less. Among the leader's games, Vachier-Lagrave's win over Ding Liren (which certainly involved home prep, as he acknowledged in his post-game comments) was a positional classic in which he managed to keep Black's king's bishop and king's rook out of the game from start to finish. (In this it was an interesting reversed echo of a famous win of his against Morozevich from Biel 2009. There it was Vachier-Lagrave whose rook was frozen for a long time, but in that game he finally managed to disentangle everything and pull out the win, despite having been lost early on. Another echo of Vachier-Lagrave's win over Ding Liren pertained to the concluding mating attack, which somehow reminded me of Kramnik's win over Topalov from their blindfold game in the 2003 Melody Amber tournament.)
Speaking of Kramnik and tournament leader Vachier-Lagrave, their game in round 3 had some surprising moments. Kramnik seemed on his way to a typical smooth win: a strong opening idea led to enduring pressure and then an extra pawn. But then a funny thing happened on the way to the textbook victory, and it happened, as Dr. Tarrasch intimated long ago, in a rook endgame. Where exactly, I'll leave to you to discover; my interest right now is in pointing out where it didn't happen.
Here's the position after 38...Kd5. Kramnik continued 39.Rb1, which allowed Black to get in 39...h5 right away. White has two possible winning strategies here in general. One is to use zugzwang to force Black back. For instance, if Black's rook retreats, White may push b4-b5, and if Black continues with ...Kc5 White's idea is to follow up with e3-e4, and after everything is traded White's king will devour Black's kingside. The other idea is to play e4 with check; that is, while Black's king is still on d5 (and White's pawn is still on b4). After ...fxe4+ Ke3 White will meet a "pass" move with Rd2+ followed by Rd4 and taking the e-pawn.
For the first strategy to succeed, White must put Black in zugzwang, and that will only happen once Black's h-pawn has gone to h5. For the other strategy, however, there's no reason for White to wait, and there may be good reason for him to hurry. Which strategy would you choose? What should Kramnik do?
In the game, Kramnik played 39.Rb1, opting for the first strategy. After 39...h5 40.Rb2 Rb6 the consistent move would be 41.b5, but by this point he realized that after 41...Kc5 42.e4 Rxb5! 43.Rxb5+ Kxb5 White isn't winning. Strange, but true! Here's a variation to illustrate the basic point: 44.exf5 Kc6! 45.f6 Kd7! 46.Ke4 Ke6 47.f7 Kxf7 48.Kf5 and now 48...h4! 49.Kxg4 hxg3 leads to an elementary draw.
Perhaps Kramnik lacked the time to calculate that variation before the time control, though my recollection was that he had a fair amount of time left. By now he understood that it didn't work and reverted to the second plan: 41.Rb1 Rb5 42.Rb2 Rb6 43.e4+ fxe4 44.Ke3, but here Vachier-Lagrave obtained sufficient counterplay to hold the balance with 44...h4! 45.gxh4 g3.
But what if Kramnik had gone for e4+ immediately, while Black's pawn was still on h7? Dutch chess legend Jan Timman was doing the English-language commentary for the tournament website, and proclaimed that the idea won, and Mark Crowther cites him approvingly. Specifically, Crowther (citing Timman) claims that from the diagrammed position the variation 39.e4+ fxe4+ 40.Ke3 h5 41.Rd2+ Ke6 42.Rd4 results in a winning position for White. Quoth Crowther: "It's hard to understand why Kramnik didn't do this."
Well, I can think of two reasons why he didn't. The first is that he probably believed he had calculated the variation 39.Rb1 h5 40.Rb2 Rb6 41.b5 out to a win, which reminds me of a joke I heard a couple of decades ago from a comedy tape called something like "How to be a Jewish Mother". One of the "techniques" went like this:
Step 1: Buy your son two sweaters.
Step 2: When he wears one of them, ask "What's the matter, don't you like the other sweater I bought you?"
As Crowther seems to recognize that Kramnik didn't realize until it was too late that 41.b5 (in the 39.Rb1 line) didn't win, I'm not sure what he expected him to do - especially as there's no way for Kramnik to wear both sweaters.
Now for reason #2: the Timman/Crowther line doesn't win! Timman was a really great player for many years and is still a very good grandmaster, but his online commentary was more him having a fun time with his friend Dirk Jan ten Geuzendam than an exercise in hardcore analysis, and he wasn't using an engine. This was a case where Ronald Reagan's slightly oxymoronic adage "trust, but verify" should have been employed: Timman is a legend, but the computer will give you the truth about such a position. (Not that Timman or even Crowther couldn't have worked things out with a bit of elbow grease.) Or to go another way with the trust angle, would it make more sense to trust the computer-less analysis of a fully motivated, fully concentrated Kramnik (rating: 2800+) or that of a very relaxed, informal Timman (rating: 2576)?
Let's get down to business: after 39.e4+ fxe4+ 40.Ke3 h5 41.Rd2+ Ke6 42.Rd4 White is not winning, because 42...h4! draws. The idea is to undermine the f-pawn, and the ensuing Black counterplay will let him hold the balance. White has plenty of options, but to at least give the most obvious one we have 43.Rxe4+ Kf5 44.gxh4 g3 45.Kf3 (45.Rc4 Kg4) 45...g2 46.Re1 Rxb4 with an elementary draw, as Black will win White's f-pawn. (You can replay these lines and a little more besides, here. [Using the software I don't want to. Sometimes a guy has to pick his battles.])
Finally, before sending this post into the world and myself to the land of Nod, here are the pairings for Sunday's round 6 games:
- Vachier-Lagrave - Gelfand
- Aronian - Adams
- Fressinet - Vitiugov
- Kramnik - Anand
- Ding Liren - Svidler