Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: History Heuristic

Author: Mikael Bäckman

Date: 02:17:43 03/17/04

Go up one level in this thread


On March 16, 2004 at 17:52:05, Sune Fischer wrote:

>On March 16, 2004 at 16:21:23, Robert Hyatt wrote:
>
>>  The idea is that a move that is good at ply 15 might also be good at ply 3...
>>
>>And vice-versa...
>
>However, local beats global in my tests :)
>
>Just a single test position to depth 15:


Testing on a single position is wrong.
If you take 20 positions and repeat the test, are the results still the same?

/Mikael


>
>no history:      329876472
>global history:  208352670  (0 calibrations)
>local history:   182337306  (29 calibrations)
>local history:   162851239  (222 calibrations)
>local history:   168449709  (648 calibrations)
>local history:   212759140  (4249 calibrations)
>
>Conclusion:
>No calibration smears the table in the long run.
>Very frequent calibration throws out too much information.
>Optimal for this scheme seems to be somewhere in the middle.
>
>-S.
>
>
>>
>>>-S.
>>>>But in any case, I depend on killers for local cutoff moves, and I use history
>>>>when all else has failed...



This page took 0.01 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.