Author: Tord Romstad
Date: 09:02:00 03/15/04
Go up one level in this thread
On March 14, 2004 at 15:00:44, Christophe Theron wrote: >However I would like to point out that using a classical alpha-beta approach >improved by state-of-the-art selective search techniques, current top chess >programs have a branching factor that is clearly below 3. Even weak amateur engines like mine have a branching factor clearly below 3 in most positions, at least as long as the PV remains relatively stable. >And current programs based on alpha-beta and selective search will miss very >rarely a tactical shot, be it an "obvious" move (a fork) or not. My experience is that it is relatively easy to write a selective search which very rarely misses a tactical shot. The difficulty is to avoid pruning important positional lines. I find this very counter-intuitive. Before I started coding, I expected the situation to be exactly the opposite. Good post, by the way. I agree with most of your points. Tord
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.