Author: Uri Blass
Date: 22:41:29 02/26/03
Go up one level in this thread
On February 26, 2003 at 18:48:53, Stephen Ham wrote: >Thanks, > >Dann, Uri, and Robin! I think I understood all of that. That was very helpful. >Again, I'm know next to nothing about chess engine programing, so I assumed that >the evaluation function guided the search function. You are not exactly wrong. In part of the cases search decisions are based on the evaluation. For example null move pruning tells the computer to prune moves that threat nothing. The point is that it is not always the case and there are search decisions that are not based on the evaluation. I don't know why I assumend >that...I just did. > >Yes, Shredder's evaluations are very "optimistic", Robin. I was surprised to see >how it enjoyed the White side of the Smith-Morra Gambit, while other engines >naturally prefer Black's extra pawn. Perhaps the latest Junior is similarly >"optimistic." I hope that SMK prepares Shredder 8 with less optimistic >evaluations in order to try to match reality. > >It seems too from my review of chess engines in positions resulting from >Ham-Nimzo 7.32 (please see my review), that certain closed pawn formations >absolutely baffle the latest chess engines. I've read that Hiarcs 8 is one of >the most heavily "knowledge" laden chess engines. So I hope that somebody would >test Hiarcs 8 on those same positions from my correspondence games. I do not know about an engine that understands the draw there except maybe David Omid's second engine that is not public(I say maybe because I only know that unlike other engines it understands a lot of fortress position but it does not understand every fortress position). I also do not think that Hiarcs8 has more knowledge in the evaluation then other engines. I remember for example that it lost an advanced passed pawn in the match against smirin thanks to lack of knowledge when passed pawns are weak. Smirin did not win the endgame with one pawn advantage and I am not sure if he could win but I expect a knowledge based program not to give smirin the opportunity in the first place. Uri
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.