Author: Andrew Wagner
Date: 15:00:27 04/19/04
Go up one level in this thread
On April 19, 2004 at 13:16:32, Martin Giepmans wrote: >On April 19, 2004 at 12:56:18, Andrew Wagner wrote: > >>>Hi Andrew, >>> >>>It's strange that you get the same score (1.8) for every move. >>>I don't think that overwriting hash entries can explain that. >>>It looks like something is wrong with the way you store or use the different >>>types of scores (exact, lower bound, upper bound). >>> >>>Martin >> >>Hmmm. I'm not sure I understand this. I'm evaluating material count only, so >>wouldn't I expect to see the same score for every move? At least until Trueno >>wins the pawn, which it apparently can't see that it can do. > >The score (1.8) in the later iterations indicates that the engine expects to win >a pawn. (although I see no capture in the PV's, perhaps the PV's are >incomplete?) >What is strange is that it expects to win a pawn after every first move. >In fact only Kb1 wins a pawn. The score for the two other moves (Ka2, Kb2) >should be 0.9 with a material only evaluation. But it is 1.8 ... > >Martin I agree, it's very strange. Somehow it's not playing the correct line for the defense or something. Ya know....just for kicks, I should run a deep perft. Wouldn't that be funny, if my move gen was wrong....
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.