Author: Anson T J
Date: 03:02:55 07/01/03
Go up one level in this thread
On July 01, 2003 at 05:41:31, Graham Laight wrote: >When a GM is contemplating a move, he doesn't say to himself, "Hmmmmm. I would >give the resulting position a score of 1.723". > >Such an evaluation is nonsense anyway. There should properly be only 3 >evaluations: > >1. Winning position > >2. Drawing position > >3. Losing position > >It would be nice if a program could work as follows: > >"nb5. This position contains a possible bishop trap". > >"nd5. This puts more pressure on the opponent's king" > >"Opponent classification: bishop trap success rate = 25%" > >"Opponent classification: king attack success rate = 15%" > >"Choice = nb5". > >-g Hi, I think your thinking is very logical but unfortunately not practical. Sometimes before I write a programme I think in my mind a logical way of implementing the task then it turns out that its completely impractical. Unfortunately the method you described above seems to be one of those impractical implementations. Putting more pressure on the opponent king and trapping a bishop might still be lines which lead to draws where giving up material and putting your king in the middle of the board might be part of a winning line. As the programme needs to "decide" which factors are more important like "how much" pressure is on the king etc, numerical evaluation will creep in. I'm not one to kill ideas without giving them much thought but I think that many people here will stick with alpha-beta, interative deepening of ply method of searching which needs some sort of evaluation or score function. Perhaps you can try to implement your idea using fuzzy linguistic variables. One problem with just "winning, losing or drawn" evaluation means that the program will not know how to converge to a mate because a mate in 100 is a win and a mate in 1 is a win. Even here some evaluation is needed to inform the program that a mate in 1 is better than a mate in 100. Its a nice idea and I think that we have all thought of this at some stage :) Good luck with your implementation :) Anson
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.