Author: Uri Blass
Date: 06:54:21 03/16/04
Go up one level in this thread
On March 16, 2004 at 09:28:46, Bas Hamstra wrote: >On March 16, 2004 at 07:26:09, Uri Blass wrote: > >>On March 16, 2004 at 06:02:22, Bas Hamstra wrote: >> >>>After Rxe6 Qxe6 Re1 Qf5 Qxf5 gxf5 this is the position that should justify the >>>sac: >>> >>>[D] 2r1r3/p3bk1p/1pn2p1B/3n1p2/3P4/PB3N2/1P3PPP/4R1K1 w - - 0 4 >>> >>>White has sacced the exchange but wins a knight back, plus white has an extra >>>pawn for it, being a freepawn, plus it has the bishoppair, plus better >>>pawnstructure. >>> >>>Anyway, Tao evaluates this as +1.00 for white or so. And it will not *ever* play >>>that sac because it evaluates the original position as slightly better, giving >>>about 1.30 or so for attacking chances in the original position. >>> >>>I don't think there is a big crush, just a good endgame, that's all. Other >>>opinions? >> >>Yes >> >>I think that the +1.30 for the original position is too optimistic. >> >>The endgame also is evaluated by movei as more than +1 >>White has a pair of bishops in the position that you posted and black has >>weaker pawn structure. >>White has a passed pawn that gives another bonus for white and white has also >>better mobility . >> >>Movei cannot find Rxe6 in a reasonable time but the problem is that the tactics >>is simply too deep for it to see it at tournament time control and it has not >>big king safety evaluation to see it by evaluation like another program that >>give positive score for white inspite of the fact that black has a rook for 2 >>pawns. >> >>Uri > >I think big king-safety penalties are not going to help here, on the contrary. >They can easily make the engine evaluate the original position as too >optimistic, like you said, and *not* choose a promising endgame. > > > >Bas. It is dependent on the question when you have big king safety evaluation. I guess that Cheetah 0.73 has not big king safety evaluation in the original position but has big king safety evaluation in the position when white sacrificed a rook. In the case of movei it has not big king safety evaluation in the position that white sacrificed a rook because my king safety evaluation evaluate only attacks in square after the king. If the king is at f6 then attacks of g5 f5 e5 are evaluated but attacks of g7 f7 e7 g6 e6 are not evaluated so it cannot see big king safety evaluation. There are also pawns after the king and it is another reason that it does not see big king safety evaluation. Uri
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.