Author: Frank Schneider
Date: 09:49:53 02/10/03
Go up one level in this thread
On February 09, 2003 at 02:57:48, Frank Schneider wrote: >On February 08, 2003 at 02:42:34, Vincent Diepeveen wrote: > >>cross posted at rec.games.chess.computer: >> >>Subject: An interesting forward pruning experiment - with pseudo description >>Date: Fri, 07 Feb 2003 23:51:19 GMT > >Hi, > >I tried a similar idea: >when there were no extensions and enough depth left, do a shallow nullmovesearch >around beta-65. If it fails low (threat) increase depth by 1/4 ply, otherwise >reduce depth by 3/4 ply. Because of the extension when the nullmove fails >high, you don't get more depth here. >This was OK for testsuites but (as expected) very bad in games. >I like your idea to research when the expected cut node fails low, maybe >this helps. After first tests with Vincents idea (C.) it seems it works very well for testsuits, but didn't improve strength in games (at least not for Anaconda/Gromit). Shit, it really looked good at first :-( Frank > >Frank
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.