Biel 2012: Wang Hao Wins (Thanks, Fake Scoring)
Magnus Carlsen came close to winning the tournament and threatened Garry Kasparov's all-time rating record for a while, but in the end he was second on both counts. Wang Hao quickly defeated Anish Giri in a Gruenfeld, and then waited while Carlsen tried to grind Bacrot down in a Spanish torture. Bacrot held firm, and the result was that although Carlsen outscored Wang Hao on normal scoring (7 vs. 6.5 [out of 10]) and beat him 2-0 in their head-to-head battles, Wang Hao won Biel thanks to the ridiculous 3-1-0 scoring system. The difference was that although Carlsen went +4, he had six draws, while Wang Hao only drew one game while winning six (and losing three). (Mind you, I'm not complaining about the result. Wang Hao seems like a nice, humble person who plays very exciting chess, and I think it's good that Carlsen not win every tournament he plays in. It's just that this seems a ridiculous outcome.)
A little more about Wang Hao - Giri. The line of the Russian System Gruenfeld they chose had previously seemed like an instant draw. In particular, the position after 17.0-0 had been played in 13 GM games since 2009 (and only in GM games!), with an overall score of +1 -1 =11. Yawn, right? Except it didn't work out that way. Giri's 19th move was new, but he didn't follow it up in the right way. After White's slightly inaccurate 22.Bc5, Giri would have equalized (with chances for more) with 22...Rfd8. Then if 23.Rfd1 e6 looks slightly better for Black after 24.dxe6 Rxd1+ (24...Qxc6 is fine too) 25.Qxd1 Qxc6 26.exf7+ Kh8.
Giri first played 22...Bh6, and after 23.Rc2 continued 23...Rfd8 24.Rc2 e6. This time it doesn't work: after 25.dxe6 Rxd1+ 26.Qxd1 Black cannot play 26...Qxc6 on account of 27.exf7+ Kh8 28.Bd4 (or 28.Be7), when it's time for him to resign. 22...Bh6 was an okay move, but not in conjunction with the (here) mistaken 24...e6(?). Giri played 26...fxe6, but the damage had been done. After 27.Ba4 material was even, but White's bishops were monsters and the e6 pawn was a big weakness. Giri was in trouble, and further errors on moves 27 and especially 29 led to a speedy finish. Down the exchange with the worse position, Giri gave up on move 32.
Finally, just as it was Bacrot's lot to repeatedly lose with the King's Indian, Viktor Bologan's bane was the Benko. In fact he wasn't in too much trouble in the middlegame and early endgame against Hikaru Nakamura, but couldn't quite manage to hold the rook and knight ending a pawn down. One of the curious aspects of the game was the seeming "immortality" of White's a-pawn. Starting from around move 22 it looked for all the world like the pawn would be rounded up, but there was always some trick that kept it alive. Eventually it perished as a b-pawn, but by that point Black was suffering in a clearly lost knight ending. Maybe Bologan could have kept some drawing chances with 33...e4 or some other 34th move, though it would have been difficult. Once the rooks came off, it was a "mathematical" forced win.
With the win, Nakamura finished tied for third-fourth with Giri (both on normal and fake scoring) and has come within 1.6 rating points of Bobby Fischer's American record of 2785. (If you think there has been rating inflation, then he still has a ways to go to "really" catch him, but it's still a very impressive figure in any case.)
Final Standings (3-1-0 scoring first, 1-.5-0 scoring in parenthesis):
1. Wang Hao 19 (6.5)
2. Carlsen 18 (7)
3-4. Giri, Nakamura 16 (6)
5. Bacrot 7 (3)
6. Bologan 4 (1.5 out of 8 games, with Morozevich 0 for 2 before that)