Looking at Radjabov's moves in the game, perhaps part of what's involved in missing this nice resource is that he was preoccupied with the idea of bringing his king to the Q-side to defend the b2 pawn, so that he could activate the rook. But the question remains as to why he exchanged on g4 - Carlsen suggested that he was afraid of g3, with e.g. the idea of Bh3 (though at least immediately it doesn't seem to work - 50.Kd2 g3 51.Kc1 Bh3 52.Ra8 Bxg2 53.Rg8, followed by Rxg3).

People, your task now is to identify the 'old' Capablanca game that Magnus referred to in the press conference.

Kan vs Capablanca, 1936

Saying that castling was inaccurate does not point to preparation. At that point castling (for either side actually) was inaccurate because it took the rook away from the h-file and allowed the Q-exchange.

Yeah, of course... I wasn't trying to imply that it does. This just seems to be a kind of opening/position where, if you have a positional understanding of Carlsen's level, you don't really need very deep preparation.

But that seems arrogant and insane. If Carlsen works on his openings he won't need to enter inferior positions with the hope of outplaying his opponents. He was lucky not to lose to Kramnik and Morozevich. And today's game could have ended very differently had Radjabov played slightly better and less passive.

Btw, Radjabov plays this line quite often, but it's not like he tends to do great with it (or - apparently - to have such brilliant ideas in it), unless he thinks of it as no more than a drawing weapon; not a single win in all 13 previous games: http://www.chessgames.com/perl/ches...

But in my opinion, it's just the law of averages. At this level, it's not that often that Carlsen will get an advantage in the opening. Carlsen would have gone on drawing his first 8 games and i would have say it's still within the range of the law of averages.

As I mentioned, Carlsen suggested this himself after the game as a way to hold the draw (though after the exchanges on c4 Black still has a slight edge in the resulting rook endgame). He thought that Radjabov, who was already low on time at this stage, was sort of avoiding any committal decisions like that before the time control, and this caused him to play too passively.

His current approach is arrogant b/c it is based on the idea of outplaying people in top 10 from a slightly worse or equal position and it is insane because he would improve his results while suffering a lot less over the board, if he worked on his openings. Just recall how his performance when he was working with Kasparov who forced him to work on his openings.

Eyal: Position after 53.Bxd5:

click for larger view

<53...d3+!> is a very accurate intermediate check here, before the recapture of the bishop (and perhaps missed by Radjabov). In case of 53...Kxd5 54.cxd4 exd4 55.Ra5+ Ke4 56.Ra8! the black king will be exposed to checks if it tries to support the central pawns; on the other hand, after 53...d3+ 54.Kd2 Kxd5, 55.Ra5+ doesn't help since after 55...Ke4 the king is defended from checks by the e-pawn.