Author: Andrew Williams
Date: 15:29:53 08/24/02
Go up one level in this thread
On August 23, 2002 at 14:47:23, Steffen Basting wrote: >Hello! >My engine is currently reaching about 8 ply maximum under tournament time >control (2h/40 moves). This is due to a branching factor of at least 3 and very >often of about 10 or more. >Isn't that too much? Do I have failures in move ordering (I am using MVV/LVA and >2 killers per ply)? >When playing blitz games, my program reaches about 7 plies - this seems to be >ok, but under tournament time control, 8 plies are surely not enough. >Any suggestions how to trace a wrong move ordering? > >Thanks in advance, Steffen One thing that many people do is to measure the percentage of times that when you get a beta cutoff you get it on the *first* move you try at that node. I often find that if my program is struggling to get depth in a position, that number is < 90% Try to implement this metric, then pick a couple of positions where your program is struggling and post them here. Everyone who has ever been in your position (ie everyone who's written a chess program) will then post their results and you'll have a basis for comparison. Cheers Andrew
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.