US Championship, Round 3 Recap
Sunday, May 16, 2010 at 10:41PM You've got to love these win-to-draw ratios, especially in comparison to the Astrayawn Grand Prix. In this round there were only three draws, and in total there have been only 13 draws in 36 games. It's true that today's game between the US's ultimate big dogs, Nakamura and Kamsky, was drawn, but only after a big fight. Here's the recap:
Kamsky - Nakamura: For once Kamsky went right into a theoretical variation, a Bayonet King's Indian, more convinced by Aronian's win over Nakamura than the latter's wins over Gelfand and Beliavsky. Sure enough, he obtained the advantage and pushed all the way to the end. Nakamura held the draw, but it would have been interesting to see what would have happened after 32.Bxb7. That's a crazy move to play in time trouble, as it allows 32...f3, but it looks like White not only survives but wins with 33.g3. A key line is 33...Bxg3 34.hxg3! Rxd2 35.Rxc7!, winning - but not 35.e7?? Qd4 36.Rf1 Rg5, mating.
Kudrin - Onischuk was one game of many where the player with the early advantage eventually went down to defeat. In a Modern-turned Pirc-turned Dragon, Kudrin enjoyed the freer position until Onischuk transformed the game with an exchange sac. White still seemed fine, but apparently Onischuk had a much better grip on how to handle it, and about 10 moves later White was lost. This was an impressive performance - it's unusual to see a good GM like Kudrin so thoroughly outplayed, and so quickly.
Krush - Akobian was nothing short of a tragedy (at least in a chess context) for Krush. A combination of terrible opening play by her opponent and excellent play on her part gave her a completely winning position from the start. Her first step in the wrong direction was 19.f6 (according to IM Regan, and I'm inclined to agree with him), keeping the position more closed than it needed to be. 19.Bxc5 axb5 20.Bd6 Rxa2 21.Qg3 g6 24.fxe6 Bxe6 25.Qg5 keeps the black king a target in a breezy position, and more proned to a quick death.
All the same, her remaining advantage was still pretty big, and she had plenty of winning chances after that, too. (27.c4! dxc4 28.dxc4 Rxc4 29.Rd1+- was an early one, 63.Re2 Rxc3 64.Reb2 was a later one, and there were still more besides.) Her last opportunity to stand better was 71.Kf4; after that, the battle was between a draw and an Akobian win. Both players made loads of errors in the rook ending, but in a game lasting 6 hours and over 100 moves, with the players basically playing on the increments, that's not exactly a surprise. The final error seems to be 108.Rb1. It's not a particularly meaningful move in its own right, but more importantly, I don't think Black is winning after the purposeful 108.Re5. After 108.Re5 Ra5 109.Ke2 f4 110.Kf2 Rb5 111.Kg2 Rb2+ 112.Kg1! Kg3 113.Rg5+ Kf3 114.Rxd5 the position would be drawn even without White's d-pawn, and it is here, too.
Hopefully Krush will overcome this painful loss and keep herself in contention for a GM norm.
Christiansen - Ehlvest: White played 4.Qe2 against the Berlin and worked up an advantage after 13...Nxh4 - 13...Nf4! was probably equal. Christiansen's edge grew, but 37.Rc2(?) let Black off the hook. 37.Bxa4 isn't an easy move to make at the end of a time control, allowing Black's rook access to the back rank, to say nothing of 37...g4, but despite all that it seems to work for White. Anyway, 37.Rc2 was safe, but Ehlvest drew easily afterward.
Shulman - Shabalov was a King's Indian that started like a dream for Black but ended as a nightmare. White's pawn "sacrifice" gave him a little compensation at first, which Shabalov successfully neutralized. With 35...b6 Black would have been in fantastic shape: glorious bishops, White weaknesses and a promising passed c-pawn. Unfortunately, he started slipping in time trouble and allowed Shulman to equalize. A draw was still the appropriate result, but 58...Rd3?? lost the game. (58...Rg4 59.Kf3 Rg1! - maybe he missed this? - held.)
Hess - Khachiyan was a very important game for Schliemann theory, and Black must find a way to defang the important novelty 16.Nd2! Khachiyan didn't come close in this game, and Hess won without breaking a sweat. Nice prep!
Stripunsky - Altounian was a Philidor with both Bc4 and Bf4 (generally it's just one per customer). Black equalized, but Stripunsky outplayed his lower-rated opponent and won with a nice tactical finish.
Yermolinsky - Lenderman was a correct draw in a QGD. White reached a two bishops vs. bishop and knight ending, but it wasn't enough for even a tiny edge.
Finegold - Kaidanov was a Colle that finished in an entertaining draw. White sacrificed a couple of pieces, but there was no mate, only perpetual.
Benjamin - Shankland was a triumph of experience. White never had anything in a Moscow Sicilian, but he parlayed that nothing into a perpetually equal ending where he still had...nothing. But on move 57, Shankland felt uncomfortable about the status quo and played 57...Bxc5? 58.bxc5 Rbb4?!/? and went on to lose the single rook ending.
Kraai - Bhat was an up-and-down Semi-Slav. White was only a little better on move 37 when Black uncorked the tactical error 37...Bh6?. Kraai managed to figure everything out, and at the end of the tactics White was the exchange and a pawn ahead, and brought home the full point.
Gurevich - Robson: Gurevich produced a TL (theoretical lemon) with 10.b4. To be fair, it looked very logical, analogous to the well-known ...b5 pawn sac in the Modern Benoni with Bd3 and h3. It didn't work here, though. White regained his piece with 15.Rxe5, but was completely lost after 15...Rb8. Robson handled the tactics perfectly, and won in crushing style in just 24 moves. Hopefully he can make a comeback now, and I also hope that Gurevich can turn things around after "castling queenside".
Leading Standings After Round 3:
1-4. Nakamura, Onischuk, Akobian, Kamsky 2.5
5-9. Ehlvest, Christiansen, Shulman, Hess, Stripunsky 2
The tournament site is here.
Reader Comments (7)
Kamsky's post game dual analysis with Nakamura mentioned the Bxb7 f3 g3 line and Kamsky was shocked when Nakamura mentioned Rxc7! and Kamsky said he missed this.
On win-to-draw ratios: While I also wasn't exactly thrilled by some of the rounds in Astrakhan, I think you are comparing apples and oranges. The Astrakhan GP has a rather even field - total rating gap less than 100 points, and Elo outsider Inarkiev refuses to give presents and is actually now leading the field. The US Championship - as you hint in your report - has a rather uneven field with commonly big gaps regarding rating and experience. This increases the probability of decisive results, including upsets if the favorite pushes too hard, plays carelessly or provocatively. The latter (careless and/or provocative play by black) might have applied if Krush-Akobian had reached its logical result?
On Kudrin-Onischuk: Do you know where white went wrong? I don't, that's why I ask ... . Isn't it normal that white has a certain space advantage in a Pirc? And it looks like black immediately had at least full compensation for the sacrificed (or lost) exchange, which is often the case in a Dragon. It played a role that white had to make some concessions (give up the pawn on b2) to get his knight back into play from a8!?
US Championship 279 is the range on ratings with a 2570.3 rating average And three 2700 GMs. This is being hailed as one of the strongest US Championships ever.
GP 83 is the range on ratings with an average rating of 2729.5 with 12 GMs over 2700 currently (the other 2 have been over 2700 in the past).
200 pts is barely a single class difference.
Keep in mind the US Championship has a much wider field of 24 players to the GP's 14. More players means more games means more fight. Also keep in mind that the US Championship has better prizes. A final point, the GP is a roundrobin of 13 rounds (much longer event, need to conserve energy, pairings and color known ahead of time so opening preparation MUCH deeper because there is more time to prepare, more rest days). US Championship is a 7round swiss = BLOODY BATTLE. Every half point determines whether you get to go into the quad, whether you make another few grand. With a swiss, you fight to win, with an RR draws are okay the important part is not losing. In a swiss, you have less time to prepare because you find out your next opponent at BEST 24 hours ahead of time and even then you won't know your color (finding out your pairing 24hours ahead time requires you doing it on your own before its official hence color is unknown). Bigger fields lend to bigger rating gaps especially when you can only invite people from one federation.
I think non US folk just take pleasure in throwing as many insults as they can at the US Championship.
Daniel, you seem to be talking to or about me ... . I agree, to a certain extent, with most of your points but don't see how you arrive at your final sentence. What can be insulting about pointing out some obvious facts, which you don't deny either? And I only explained why it's logical that the US Championship has many decisive games.
The Swiss system may also play a role. But for one thing, a game may end up being drawn even if one or both players try hard for a win. Notably, Aeroflot (very strong field for an open, few if any games with a rating difference >100) had a fair number of draws, both long and fought out and short ones. With all the noise made about Sofia rules, some people tend to forget that a draw is a logical and legitimate result of the game (for the record, I don't think Dennis is one of them).
On preparation: Astrakhan had the drawing of lots before round 1. Yes, in principle a player can already prepare for round 12, but in practice they will probably spend (part of) the evening after and the morning before a game to prepare for their next opponent - same for the US Championship, just a few hours less available because you don't know the drawing of lots yet. Regarding long-term preparation: Gelfand could start preparing for Radjabov weeks or months ago (but didn't know the color distribution), Nakamura could do the same for Kamsky - again, no major difference.
BTW, prize money at the Grand Prix is overall at least comparable:
US Championship: 35,000/20,000/15,000/10,000/10,000 $
Grand Prix: 30,000/22,500/20,000/15,000/12,500 Euros [for the time being, the Euro is worth more than the US Dollar ...] - plus accumulated prize funds for the top10 of the overall series, plus invitation to the candidates event for two lucky ones.
The US Championship has a larger "winner takes it all" (or most) component, that's all.
Thomas,
You've been reading this blog long enough to know that I don't object to draws per se, and don't even mind players taking the occasional short draw. (Players get tired, are having lousy events and want to get it over with, are trying to consolidate their tournament position, etc.) But the first five rounds of Astrakhan were just ridiculous! It's clear watching players like Topalov and Carlsen that their opponent's high ratings and successful opening play won't guarantee them a draw. It's about having fighting spirit, and while those two are at a positive extreme on one end, the players in Grand Prix seemed to plant themselves on the opposite end.
Thomas,
the last sentence was not aimed at your in particular but an aside in general. I'm afraid a true one.
Notice also 3 more things. One the difference in rating average between the two events is 159.2 therefore they are comparable. Two, the only reason these two events would not be comparable would be due to sample size (lack of n=30 for statistics) see next point. Three, as Dennis mentions most of the US players are fighters while most of the GP players (Ivanchuk and Radjabov excluded) are often made fun of as 'drawmaster'. I take particular offense to anyone calling a strong player a 'drawmaster' as some insult... but for statistics sake without having 30 players in a tournament you can't run comparable stats to see said player. Still i'm under the impression that draw ratios at the high level tend to run in the range of 50%to70% per event at least according to the stats I ran last year. So right now the US Championship is indeed running an extra bloody event while GP is heading towards the high end.
Daniel, many points to address:
On "non-US insults": It may well be that some Europeans are less interested in the US Championship. It's well organized and as strong as it can get, but would Dennis have detailed coverage of the French championship with Vachier-Lagrave, Bacrot and Fressinet (about comparable to Nakamura, Kamsky and Onischuk)? Probably not, and nothing at all wrong with that - any blogmaster caters to himself and a significant but <100% part of his audience.
It became a bit of an issue for me last year when, IMO, "too much" was made of Nakamura's victory, making it sound like it's comparable to winning Linares or Corus A - when the average rating of his opponents was at most comparable to Corus B. Again, pointing out such facts is not an insult!? [This doesn't aim at anyone in particular, and generally more at bloggers than blog owners].
On rating average: I find it a bit amusing that you consider a 159 point difference "no big deal" (less than one class), when it corresponds to seven tournament categories (category 20 for Astrakhan, category 13 for the US Championship) - admittedly a bit unfair given the difference in number of participants. Many round-robin organizers strive for the highest possible category, as a consequence players like Onischuk (2699) and Kamsky (2702) get few top-level invitations - Kamsky maybe still a bit benefitting from his reputation and career record.
On fighting spirit and win-to-draw ratios: In my opinion, it makes more sense to compare Americans rated 2400-2700 with Europeans of similar strength than with the world top. Players rated below 2700 have an interest to _improve_ their ratings, once you are part of the 2700+ club you are also interested in _defending_ your rating - for reasons given above (tournament invitations). A bit simplistic, and - despite their different styles - Nakamura and Kamsky may both be an exception, but n=2 is statistically insignificant.
IF "American apples" still show more fighting spirit than "European apples" (hard to come up with objective or even fact-supported subjective assessments), there may be two reasons:
1) prize structure in American Swisses: It may be financially more rewarding to finish 1st, 22nd and 58th in three consecutive events than 3rd, 5th and 10th - in Europe it's the other way around. In US Swisses it seems also common and quite acceptable to drop out once you are out of competition for big prize money.
2) importance of team events: In Europe they are a source of stable income for sub-top players, generally a team has eight players. Depending on the match situation, it can be in the team interest to play it safe and concede a draw. You don't make friends with risky attempts to play for your own glory - certainly not if such attempts backfire, maybe even when you win in the end after making teammates and team manager suffer for quite a while.