The fans definitely got their money's worth today, as both the Grischuk-Aronian and Kramnik-Radjabov playoff matches were full of drama.
To recap, the tiebreaks had the following format: a four game mini-match with a time control of 25 minutes for the whole game, plus 10 second increments per move, followed, if necessary, by up to pairs of 5' + 3" blitz games followed by an Armaggedon game. Grischuk-Aronian was settled in the rapids, while Kramnik and Rajdabov needed two sets of blitz games to decide a winner.
Grischuk got off to a great start against Aronian, convincingly outplaying him on the black side of an English. The success didn't last for long, though. In game two his handling of the normally solid Exchange QGD was anything but, and soon Black was better thanks to his bishop pair. The nature of Black's advantage changed a few times, eventually transmogrifying into a pawn up rook ending that Aronian won. In game three having White finally made a positive difference and Aronian stood better. In fact, Aronian was probably winning at (at least) one moment (31.Qc3!), but Grischuk escaped and drew. Finally, Grischuk outplayed Aronian in game 4 on the white side of a Bf4 QGD. Aronian tried to create something of a fortress, but without any time on the clock it was inevitable that something would give, and it did.
With that, the rating favorites had gone three-for-three...in losing their matches - and it was nearly four-for-four! Kramnik and Radjabov had played four short, dull draws in classical chess, and they played four more forgettable games in the rapid. The first game had some play, and Kramnik pressed for a while with White in the Moscow Semi-Slav, but in games 2-4 the games were balanced, uneventful and quickly drawn.
It was time for the blitz rounds, and now it got exciting in a hurry. Radjabov had White in game 1 (game 9 overall), and produced a match novelty: 1.e4! Kramnik chose a Berlin, which Radjabov "declined" with 4.d3. The game turned into a sort of slow Giuoco, and surprisingly Radjabov handled the maneuvering phase better than 1...e5 professional Kramnik. Black was under heavy pressure, and after missing a nice chance to escape on move 37 (37...Ngf4! rather than 37...Ne7?), Radjabov hit hard with the power shot 38.Nh6! Black's position collapsed after that, forcing Kramnik to win the next game to keep the match going.
Kramnik did manage to obtain an edge in an ending with rooks and opposite-colored bishops, but although he managed to win a pawn Radjabov achieved a clearly drawn position. Desperate to keep the game going, Kramnik wound up blundering the extra pawn back, though it has to be said that it was the least important of his pawns. Anyway, Kramnik continued trying without getting anywhere, until of all things there was a clock malfunction. Play stopped for something like 10-15 minutes while everything got sorted out. When they came back, Kramnik at least managed to find an idea, and it worked. Radjabov fell into a fatal zugzwang on his 68th move, and the game was quickly over. Time for another two-game blitz mini-match.
In game 3 Kramnik again had White, and again Radjabov sharpened things up a bit by finally playing his favorite King's Indian. As usual in this match, Kramnik didn't get anything out of the opening, and the game was still equal once it reached a heavy piece ending with both sides having a queen and both rooks. It remained equal even as a rook ending, but Radjabov made the cardinal error of defending passively. Kramnik had enough positive ideas to make progress, and finally managed to win a second straight "drawn" endgame. (Carlsen would be proud.)
Finally, game 4 was another Declined Berlin-cum-Giuoco Piano, and while Radjabov might have briefly had an edge in the opening Kramnik was soon doing well and even stood better. 42...Qd6 was an error, though, and Radjabov briefly had some hopes. Kramnik defended well and found a fine idea on move 59 that wound up forcing a draw, but his implementation was inaccurate. Kramnik should have played 59...Rxd6 first rather than doing so after 59...Qe3+ 60.Kh1, because if Radjabov had chosen 60.Kf1! instead Black would have had some difficulties. After the exchange of errors, Black forced perpetual check and won the match.
Our semi-finals are set and will begin on Thursday. The pairings are Gelfand-Kamsky and Grischuk-Kramnik. I'll preview these matches tomorrow or Wednesday, and then we'll make a new set of predictions: a 2-G final, 2-K, or one of each?
As for today's games, I haven't annotated all 12 of them, but here they are, seven with notes.