Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The Limits of Positional Knowledge

Author: Robert Hyatt

Date: 08:49:12 11/12/99

Go up one level in this thread


On November 12, 1999 at 02:31:16, blass uri wrote:

>On November 11, 1999 at 12:53:35, Ratko V Tomic wrote:
>
><snipped>
>>This error propagation is the underlying reason why the searchers with imperfect
>>evaluation as they increase the depth eventually hit the region of diminishing
>>returns.
>
>I do not see this deminishing return in the ssdf list.
>
>I do not see that the difference between K6-450 and p200 in ssdf rating is
>smaller relative to the difference between p200 and p90 in ssdf rating.
>
>Uri


I believe that 'diminishing returns' does happen, _if_ you take a static engine
and run it on faster and faster hardware.  But if you continually update the
engine/evaluation/search as it is tested on faster hardware, I don't see
anything (yet) that says that once we reach depth D, going deeper will not
make the program play any better...



This page took 0 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.