Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Winning line for Fritz in game 6 against Kramnik is (NOT) here:

Author: Mig Greengard

Date: 01:43:26 10/18/02

Go up one level in this thread


I know this is the CCC and not a CC, but we already know that engines don't find
a win or the best defense because their evaluation of several of the endgames is
completely wrong. In what you post white has many chances to simply capture on
b5, giving up a rook for the black knight and b-pawn for a dead drawn rook vs
queen fortress position. The engines will never do this because they think the
queen vs rook position is wonderful for black, with evals of +4 or higher. But
this is the simplest draw of the many possible.

The only variations of interest are those than can refute these defensive
concepts:

1. Giving up a rook for knight and pawn for a fortress.
2. Giving up both rooks for queen and pawn for a drawn f+g pawns vs N+h-pawn
endgame.
3. Giving up a rook for both b-pawns for a drawn R vs R+N endgame.

Number three is the only one that seems to offer any chances for a black win
right now. But several GMs have looked at that ending in detail and have had no
success in winning it. According to Keene, he played an almost identical ending
with Mednis over three days (adjournments) and never won it.

There may exist some fantastical lines with an attack on the white king using
the b-pawn as bait, or other possibilities with the queen still on the board,
but this seems unlikely at this point.

Saludos, Mig



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.