Author: martin fierz
Date: 07:29:33 06/29/03
Go up one level in this thread
On June 28, 2003 at 11:20:21, Frederic Louguet wrote: >I read an interesting paper from Albert Xin Jiang and Michael Buro about >Multi-ProbCut and its implementation in Crafty. I have always been very >skeptical about this pruning technique (for chess) but the paper is rather >optimistic. However, the data presented does not seem very convincing from a >statistical point of view (too few games, not enough opponents). So could Robert >Hyatt tell us a little more about the effectiveness of this technique ? Does it >_really_ work ? i can't really tell you anything about chess, but i tried multi-probcut (MPC) in my checkers program. the general opinion about checkers is that null-move is not a good idea there, so everybody uses his own hand-crafted pruning algorithm. therefore, MPC seemed like an interesting candidate for my program. MPC performed much better than no pruning, but also clearly worse than my (highly checkers-specific) own pruning. i ran matches with 300 games/match, so the results were statistically significant (i don't remember the numbers off-head). i don't see why MPC shouldn't work in chess if it works in other games like othello and checkers. whether it's better or worse than nullmove i have no idea. as a general observation, i find it hard to believe that any generic pruning algorithm (like MPC & nullmove are) could perform better than one which incorporates game-specific knowledge. cheers martin
This page took 0.01 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.