Computer Chess Club Archives




Subject: Re: Performance of MTD(f) versus eval granularity?

Author: Robert Hyatt

Date: 14:52:53 11/01/01

Go up one level in this thread

On November 01, 2001 at 16:31:30, Pat King wrote:

>A related question: How do lazy eval and MTD interact? It seems to me an initial
>eval far from beta and based on, perhaps, materiel only, could be saved in the
>hash table and later in the search become considered the PV. For that matter,
>PVS should suffer the same difficulty. Am I missing something, or is getting to
>the next ply expected to take care of this the majority of the time?

You just use the "initial alpha/beta values" for lazy eval, rather than
the current alpha/beta values which will always be of the form X,X+1.

Of course, that hurts efficiency of the lazy evals...

This page took 0.03 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.