Author: Uri Blass
Date: 06:04:46 01/11/03
Go up one level in this thread
On January 11, 2003 at 08:52:36, Jeroen Noomen wrote: >On January 11, 2003 at 08:31:44, Lieven Clarisse wrote: > >>This is most probably a well known position (pasted below). But ruffian resigns >>here (5 min/game PIII 450 64MB hash). I think this kind of situation can be >>easily dealt with: if the evaluation score for the initial position is equal to >>that, after 25 ply (for instance) for the best line, than the score for that >>position should set to 0.00. In other words, if the engine can't make any >>progress after a certain number of moves, it is draw; no mather how inbalanced >>the material is for both sides. > > >Might work for this position, doesn't work for millions of other positions. >Chess is a game that cannot be caught in a limited number of rules. A program >based on your proposal, will be much weaker than a program that doesn't do that. >Programming the exceptions in chess will do only harm to the general playing >strength. > >Only a clever human brain can solve positions as given below. Thanks for that! >Chess must remain interesting and not just a calculating game that can be solved >with a number of rules. > >Jeroen The fact that only humans can today see 0.00 score does not say that it is impossible to teach it a program without doing it weaker. The main reason that programmers did not do it is because it is not important and not because of a significant demage to the level of the program. Solving the resign problem is easy and a program can resign only if it see a winning score+increase in the winning score. I believe that correcting the misunderstanding of programs in fortress positions is also possible without demage and I expect David omid to prove it. Uri
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.