There were some real surprises today, and they ran from one end of the pairing table to the other.
On board 1, the top seed is out. Sergey Karjakin lost to Judit Polgar yesterday, and was unable to make any headway on the white side of an Open Ruy today. Polgar played well, drew comfortably, and advanced to the round of 16.
On board 2, Emil Sutovsky reminded us of two things: first, why we love watching his chess; second, why he hasn't gone further in the chess world. With the white pieces and needing only a draw against Vassily Ivanchuk, whom he upset yesterday, you'd expect Sutovsky to choose something solid. Not passive, but solid. So what does he do? 1.e4 g6 2.d4 Bg7 3.Nc3 d6 4.f4 Nf6 5.Nf3 0-0 and now not 6.Bd3 but 6.e5 Nfd7 7.h4. It's not that this is a bad line, but it's absurdly committal. Stranger still, a bit further on he went into a line that scores well for Black and that the computer likes for Black as well. The resulting position was an utter mess - just what Sutovsky would have wanted if he had been the one in the must-win situation, but not helpful in his actual circumstances. Sutovsky erred early on, and Ivanchuk played very accurately (though not quite perfectly) and won an impressive game.
Mamedyarov - Zherebukh and Efimenko - Ponomariov were short draws by players who would apparently prefer to settle things in tomorrow's tiebreaks.
Gashimov - Tomashevsky saw White get nothing from the opening or the middlegame, but Gashimov ground him down in the ending to win the game and the match.
Back to the bizarro world. Alexander Morozevich lost a very exciting game to Alexander Grischuk yesterday, so with White you just know he's spoiling for revenge. Win, or at least come home on your shield. Get knocked down six times, get up seven. Eye of the tiger, rage against the dying of the light, etc., etc., etc. Okay, let's see what happened:
Morozevich-Grischuk: 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 d5 5.cxd5 exd5 6.Bg5 h6 7.Bxf6 Qxf6 8.a3 Bxc3+ 9.Qxc3 0-0 10.Nf3 Bf5 11.e3 Rc8 12.Rc1 (still theory) 1/2-1/2.
Somewhere, the caricature of Peter Leko is spinning in its grave.
Radjabov - Bacrot saw Radjabov win to advance to the next round. It's interesting that Radjabov plays such sharp and principled openings with Black, but with White is playing the slow systems of the Italian Game. Then again, he's succeeding with it, so why not? He came out of the opening with an edge thanks to his extra space, and Bacrot's attempt to neutralize it by sacrificing the exchange didn't succeed, as Radjabov showed good technique to win the endgame.
Nepomniachtchi - Kamsky: Nepomniachtchi equalized the match, winning a Gruenfeld-like line of the Symmetrical English. White obtained two advantages out of the opening: the bishop pair and one giant pawn island against Black's two islands. From there, Nepo won a textbook game, restricting Black's position more and more, tying Kamsky up and breaking through on the queenside. They're off to playoffs tomorrow.
Svidler - Caruana was a short draw; they too will continue to tiebreaks tomorrow.
Jobava - Jakovenko went into a sideline of the Botvinnik System, and while Jobava never had an advantage, the game was pretty close until 33.c4? As Jakovenko had won the first day, maybe Jobava felt he had to do something radical there to avoid a draw. In any case, the pawn sac didn't help him, and Jakovenko won, sweeping their mini-match 2-0.
Vitiugov - Potkin was a draw, but unlike many of the other drawn games in 1-1 matches, they played a real game. That's not to say that either player was ever in trouble though: Potkin comfortably equalized with Capablanca's system against the Reti, and had the slightly more comfortable half of the draw. Off to the tiebreaks.
Nielsen - Parligras: An 18-move draw: playoffs for them.
Bruzon - Le Quang Liem: A 6.Be3 e5 7.Nf3 Najdorf where Bruzon obtained a serious, at times winning advantage. There seem to be small inaccuracies here and there by both players, but White was clearly better (or more) from the opening through move 33. Here, perhaps in time trouble, Bruzon played two very strange moves: 33.h3? (the Ng4 wasn't in trouble and there were no back rank issues, so this is just an important loss of a tempo) Rxc2 34.b3? (allowing Black to eliminate the dangerous pawn on b7). The best move was 33.Rfe1 (33...Rxc2 34.Rxe4), with the subtle idea of meeting 33...Nd6 with 34.Red1! However, even the more obvious and human moves 33.b3 and 33.c3 would have maintained good winning chances as well. After the moves in the game, Le escaped and drew, and he and Bruzon will also play tomorrow.
Moiseenko - Navara: I'd love to tell you what happened in this game, but as of this writing there's still some question. Navara failed to convert a winning position yesterday, and it seems he has done so once again today - though not the way you might think. (I doubt any of my readers could even invent such a story.) On move 55, Navara promoted a pawn to a queen, leaving him with just that queen against Moiseenko's rook and e3-pawn. A pawn on that square does not give its possessor a fortress, and on move 73 Navara captured the pawn. Queen vs. rook is a theoretical win for the strong side starting from any normal position, but against strong defense the side with the queen had better know what he's doing, or the 50-move rule will become a factor. And so it was here - it took Navara 32 moves to make four moves' worth of progress. From moves 106 on, however, Navara played perfectly and Moiseenko didn't, and on move 114 (after the move 114...Kd6-c6), with an imminent and obvious mate coming (in four moves), White resigned.
Or did he? The arbiter put the White king on d4 and now a Black king placed on e5 would indicate that Black won. (For those who wonder why game scores sometimes have bizarre final moves, it's because of the idiotic design of the DGT boards, compounded by incompetent/inadequately trained arbiters. I would love to fire the people at DGT who refuse to create a design that sidesteps this issue and all the ruined and ambiguous game scores this design flaw creates...alas. Obviously the idea of putting a small switch on the side of the board with results for a White win, a Black win and a draw, with a confirmation button, is far too complicated. [No doubt other methods are possible, but isn't that simple enough?] Rant over, until next time.) Anyway, rather than putting the Black king on e5, the arbiter put it on d5. (The way it works is this: the kings go on the central squares. Put both kings on white squares: White wins. Both on black squares: Black wins. Opposite colors: draw.) So a draw?
I assumed this was just a tired or inattentive arbiter inadvertently showing once again why the DGT design is so hopelessly stupid, but the real story seems to be that it really was a draw. But how? One hypothesis that was kicked around was that Navara misread the scoresheet and offered a draw, thinking that the 50-move rule had come into effect even though only 40 moves (actually 42) had been played. But apparently that isn't it either. Rather, according to Pavel Eljanov's tweets, Navara had accidentally touched a piece in the middlegame and asked his opponent what to do. His opponent just said to make any move he wanted, and he did. But according to those tweets, this occurred in the middlegame, so why did he wait until he had mate in four to offer a draw out of guilt? And in any case, why would he have felt any guilt at all? The touch-move rule covers only cases where one intends to move a piece; when a piece is touched by (physical, not mental!) accident there's no obligation to move it at all. (Maybe this happened in time trouble, and felt that his action caused an improper distraction to his opponent?) Crazy. Anyway, they're off to tiebreaks tomorrow. (HT to Mark Crowther for supplying me with some sources on the controversy.)
Bu Xiangzhi - Abhijeet Gupta was also absurd, but in a purely chess-related way. Their first game was drawn, and today Gupta enjoyed the advantage in an a quasi-endgame with queens and opposite-colored bishops. I don't think he was ever winning (though 46...Qc1+ 47.Ke2 Qh1 gives him excellent chances, and later 50...Bf4, aiming to put the bishop on e3, would put White on the ropes), but he had enduring pressure in a position he couldn't possibly lose. (Which means, of course, that he did lose it.) The climax of the game came after Bu's ingenious 58.g3! Whether this is best or not is unclear, but it confused Gupta. After 58...hxg3 59.Kg2 Qxb4 (59...Bxb4 was better) 60.Qd8+ it's a draw after either 60...Kf4 or 60...Kh6, but after Gupta's 60...Kh5?? 61.f4! it was time to resign. White threatens both 62.Qg5# and 62.Qh8#, and neither capture on f4 saves Black: 62...Bxf4 63.Bd1+ Kh6 64.Qh8+ Kg5 65.f4#, or 62...Qxf4 63.Bd1+ Kh6 64.Qh8+ Kg5 65.h4+ Qxh4 66.Qd8+ wins the queen. Bu advances to the round of 16.
Lysyj - Dominguez: White avenged his loss on the previous day, grinding out a win in an ending with knights and opposite-colored bishops. It looks as if Dominguez might have been able to save the game with 54...Ne4+, though it's hard to believe. After 55.Bxe4 fxe4 Black seems to have a kind of fortress: White's knight can't move without dropping the c-pawn (which doesn't make the a-pawn's life a safe one either) and if his king goes to the fifth rank Black has ...e3. So maybe 55.Bxe4 isn't good, but if he doesn't take the knight Black plays 56...Nc5 and White's queenside pawns will disappear. Anyway, even if White has a study-like win after 54...Ne4+, it was Black's best chance. Dominguez preferred to bring his king to f6, to prevent White from playing Kg5, but this simply lost a piece: after 56.Ng3 Black has no defense to the idea of Be6/Bb7 followed by c8Q. They'll have tiebreaks tomorrow.