Author: Eric Oldre
Date: 20:14:45 07/21/04
Go up one level in this thread
On July 21, 2004 at 17:51:44, Ed Trice wrote: >I think crafty keeps either two killer moves or two history heurstic moves to >try, I forget which. It has been a while since I looked at that code, but I am >sure Dr. Hyatt would know the answer right away. > >The benefit is found in the move ordering, which can reduce the tree size if >implemented properly. Usually your first strong move gives you the greatest >benefit in terms of the cutoffs, but having a second one around just in case >will help get you very close to the theoretical minmal skeleton tree size >possible with alpha-beta. I just got a basic SEE routine implemented last night. I couldn't believe what a huge difference it made in my move ordering over the MVV/LVA i was using before. I'm hoping that once i get killers, a history heuristic, and the 2 tier trans table, that it improves greater still. that won't come for a while though, my eval function needs some attention!!!
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.