Bibs wrote on 07/08/07 at 09:52:01:
Running with Rybka, it has the position after 18... qd7 as basically equal and prefers Qe2 to Vigus' Qe3.
Well I have a suspicion that when Black is a Pawn up and with a Bishop pair and the engine is saying =, then a human in that position would have practical problems.
As for James's suggestion, can Black hold the ending after 20. Nxe5 Qe6 21. Rfe1 Bxe5 22. Qxe5 Qxe5 23. Rxe5 Be6? That a pawn looks pretty weak after Re3->a3, but maybe the opposite coloured Bishops and the extra White pawn (after the a pawn drops) being backward it's drawable, and that's all Black has to prove.
I quite like a defensive challenge, but I'm not going to be up to offering more insight than Rybka
. I don't like its suggestion of 20. ...O-O-O as the Black King looks draughty, but I'm no computer. 21. Ra3 Rhe8 22. Re3 looks normal with an edge for White, but Black's still in there.
However I reckon if I'd found myself in this position in a practical game, I'd have gone for bolstering up with 20. ...f6 21. c4 Bc6, but after Ra3, things are starting to go downhill, which demonstrates the practical problems.
I've got some notes on 15. Bh6 which was recommended by Kaufmann, but that's about it.
PS I reckon 18. ...Qc7 may not be more than just leading to a transposition.