csmath: We are following the game:
Kramnik vs Anand, 2002
which ended in a draw.
is a new move, clearly Giri has been preparing this. In some recent games Kramnik have been surprised in the openings (by Svidler, Andreikin). Giri is banking on something like that.
[Kramnik sees no reason to deviate from his plan against Anand.]
and now we can say that black has not equalized.
[27. ...Rd8?!, 28. Qa4 Rxd1?, 29. Qxd1 Qf7, 30. b3 and black will not be able to defend f and h pawns thus white has significant advantage. Whatever Giri was preparing he is simply outplayed in the opening. The 27. ...Qb3 blocking white pawn on dark square was a better try.]
[Allows the following sacrifice. It is hard to know whether Kramnik wanted to lure Giri into that or did he genuinely overlook the sacrifice.]
[This is best hope for black since it is clear he will have a hard time defending f and h pawns.]
[This is a proper decision, Kramnik does not want to give away h-pawn as in 32. Qxg4?! Rxg4, 33. Bf3 Rxh4 and black has excellent chances for a draw.]
[This is an excellent counter-sacrifice. It has no danger for white while makes black to re-calculate the position with little time left.]
[This move loses the game. Proper response was:
34. Rg2 [bishop is lost due to weak first rank, say 34. Bg2?? Rc1 with mate or material loss]
[35. Be4 f3!, 36. Qe5+ Ka7, 37 Qc5+ with repetition.
This all is not so easy to calculate.]
34. Bd7 Qg3
35. Qd6+ Qxd6
36. Rxd6 Qc7
and now black is lost in ending.
Clean win by Kramnik. Better opening despite Giri's obvious preparation, practical counter-sacrifice at the right moment (when Giri was low on time). This is a practical win, in very important time for team Russia.
Giri might think twice the next time when he gets "prepared" against Kramnik.