Author: Russell Reagan
Date: 12:51:50 04/13/02
Go up one level in this thread
I have given this idea some thought before. I think it would be possible to create some method using pattern recognition to create a database of much smaller size, but in my thinking, I have thought that you probably wouldn't gain any giant size difference if you still want the program to play "perfect". I think it's probably possible to develop some general rules or "steps" for each specific endgame and store those patterns in a much smaller file, but the down side to this is that it doesn't play absolutely perfect, it might only play 95-99% perfect (that's just a wild guess). The idea has some potential, but as you pointed out, Eugene Nalimov is quite a talented programmer and I think it would take a talent like him (or someone of that talent level) to implement a working program. I'm still waiting for Eugene Nalimov and Ken Thompson to release chess programs. That would be interesting IMO. Russell
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.