More on Karjakin-Topalov

Chessbase have now published their GM analysis: very brief indeed, and not mentioning Topalov's comment about 'not 29…Ke7 30 Nd4!’. So, curious, I decided to check my analysis of 29…Rg8 30 Qe1 Ke7 31 Nd4, with startling, or humbling, results. In brief, two major mistakes.

Firstly, after 31…ed, Houdini instantly the position after 34…Bc1 as +4: indeed plays 35…Be3, but then improves considerably on my play, with the mating tactic 36 Bd3! Nd5 37 Bb5, when the threat of the same mate that I spotted in a similar position, forces the passive 37…Nc7, when it is game over: the h pawn will fall.

However, black has a vast improvement, 31…Bd2!!, exploiting the LPDO Rg4. White is then suffering: best seems to be 32 Nc6+ Rc6 33 Rd2 Rg4 34 dc d5, but even here Houdini keeps finding tactics. 35 Bf3 Rf4 36 c7! Qc7 (Kd5 37 Bd5!) 37 Bd5 Nd5 38 Rd5 Qc4 39 Rd1 with a queen and rook ending in which probably black is better, but it really would be anyone's game.

So, my 31 Nd4 was, in short, a poor move. At least I have enjoyed the thinking about the position.

Posted with Blogsy