Computer Chess Club Archives




Subject: Re: Artificial Intelligence in Computer Chess - *DETAILS* as promised

Author: martin fierz

Date: 07:31:53 03/29/04

Go up one level in this thread

On March 29, 2004 at 10:15:03, Dan Andersson wrote:

> I would if I could. But I'm afraid your question isn't sufficiently well
>defined for a non simulation answer. I.e.: x% doesn't tell us how many moves
>were needed to get a cutoff.
> I believe have previously posted some crude approximations in this forum
>concerning the minimum tree. Or at least some formulas that can be used for that
>purpose. I recommend a search.
>MvH Dan Andersson

right, the question is ill-defined; however, i was using x% to mean found on the
first move (which is what "everybody" always uses). the problem is that the
remaining (100-x)% are not defined, but any computation would be nice, for any
model you can think of for when you fail high if you don't fail high on the
first move (e.g. say you fail high on the N/2 move on average if you don't do it
on the first or on the sqrt(N)th move.

i posted this on the top level again in the hope that somebody did this


This page took 0.01 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.