Author: Mogens Larsen
Date: 14:56:15 05/09/00
Go up one level in this thread
On May 09, 2000 at 16:14:05, blass uri wrote: >Most of the programmers are not interested in teaching computers thing that the >computer can learn from tablebases. You don't know that. You can't extrapolate the Nimzo case to include all programmers. >I read that Nimzo7.32's programmer dropped knowledge about simple endgames >because nimzo7.32 can use tablebases. > >Doing rules that forbid using tablebases is unfair to programmers like >nimzo7.32's programmer. Yes, but it was his own decision, noone forced him to exclude development of endgame play. It isn't fair to those who develop their program without the aid of tablebases either. >programmers have some assumptions when they develop their program. >It is not fair to change the assumption they are based on. The programmers can make all the assumptions they want, but that doesn't mean that all others are forced to abide by them. Why should it be the programmers that decide the assumptions? BTW, I don't think you know anything about the assumptions made by programmers, except maybe Nimzo. >I think that teaching programs to do somethings that humans do is important in >the middle game and also in the opening when the opponent try to take the >computer out of book. You're missing the point. From a scientific point of view _every_ aspect of the computer chess game should be capable of improvement. Everything doesn't revolve around winning or losing a chess game by all means necessary. >It is not important in simple endgames when tablebases is a better solution. You are wrong, it's the simplicity of the endgames that makes them interesting. Tablebases is just a "simple" supplement capable of hiding obvious problems, that are inherent to the program itself. Tablebases are only a temporary solution IMO. Sincerely, Mogens
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.