Author: blass uri
Date: 16:48:28 09/16/98
Go up one level in this thread
On September 16, 1998 at 19:09:16, Amir Ban wrote: >On September 16, 1998 at 12:09:31, blass uri wrote: > >>I believe a slow searcher can be better because you must have a slow searcher to >>evaluate which lines are more important to analyze. >> >>I think that a slow searcher can be better also in test positions than fast >>searcher because of this reason. >> >>What do you think about it? > >This question is phrased badly. > >All things being equal of course you prefer a fast searcher over a slow one. > >You must define what makes a searcher "slow", or maybe you don't know, but you >are assuming that everybody else here knows what makes a slow searcher different >from a fast searcher, which is not true. I think that 2 evaluation functions may make a slow searcher different 1)evaluation function of the depth of the node you are going to search. an example to a simple rule is to decide a capture is half ply. Other rules may be for every possible legal move how many plies is the move. For example a rule that the move white bishop from g2 to h1 is 4 plies if there is no black piece that can capture g2 and is 3 plies if... This is a lot of work to write all the rules for every possible legal move. 2)evaluation funnction of the position. I think that this is the only part that programmers of slow searchers do and this is the reason that today slow searchers are not better than fast searchers. 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.