Author: Brian Richardson
Date: 13:24:40 10/09/04
Go up one level in this thread
On October 09, 2004 at 07:42:56, David B Weller wrote: >Uri, > >I too use wacnew [mostly for bug/error detection] > >I think that, since so many are solving wac141 without trouble, it implies there >may be some particular deficiency in ga that needs addressing. > >GES has difficulty with finding Qg2+ in WAC.163. I dont know why. > >[D] 5rk1/2p4p/2p4r/3P4/4p1b1/1Q2NqPp/PP3P1K/R4R2 b - - > >It may be, in both GES and ga, that we have an imbalance or disproportionate >extensions applied??? > >Or, maybe it is something else. But the fact that these positions are so readily >solved by others seems to suggest that possibly there is a bug/error someplace. > >The closest thing to a 'magic bullet' in chess programming is bug elimination - >IMHO > >-David I tried Crafty 19.17 with and without null move mate threats. For WAC.163 it does not matter and finds Qg2+ in 3-4 seconds (Athlon 1.6GHz). For WAC.141 it does make a difference (under 5 seconds v more than 30 without). Brian
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.