Author: Roland Pfister
Date: 05:59:14 10/17/98
Go up one level in this thread
On October 16, 1998 at 21:39:21, Bruce Moreland wrote: > >On October 16, 1998 at 15:33:45, Roland Pfister wrote: > >>Obviously Marty is doing something wrong. I have no problem with >>dynamic TB access in my program, I do it similiar to Crafty. > >I think it may very well be true that some endings would be played more weakly. >You start getting tons of hits on a queen and pawn ending and you can get >severely outsearched. You end up losing the 7-piece ending rather than winning >or drawing the 5-piece ending. > >I've seen node rates under a thousand nodes per second in some endings. > >bruce At a friend's computer who owns Mchess we compared MChess to Patzer in a position where 4 man TB access occurs. MChess was extremely slow (10% IIRC), Patzer was at about half of its usual speed. So Marty does something "suboptimal", that is my conclusion. Roland
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.