Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Search Instabilities

Author: Robert Hyatt

Date: 07:41:50 01/22/02

Go up one level in this thread


On January 22, 2002 at 01:53:53, Uri Blass wrote:

>On January 22, 2002 at 00:44:53, Robert Hyatt wrote:
>
>>On January 21, 2002 at 14:27:34, David Rasmussen wrote:
>>
>>>On January 21, 2002 at 10:41:39, Robert Hyatt wrote:
>>>
>>>>
>>>>You have to do two things:
>>>>
>>>>1.  If you get a fail high at the root on a zero-width window (any move after
>>>>the first move should be searched with a zero-width window) you can't trust it
>>>>unless you re-search it with a bigger beta bound and make _sure_ that it doesn't
>>>>then fail low.  Such fail-high (zero window) fail-low (non-zero window) is an
>>>>artifact of null-move and if you play such a fail high move even if it fails low
>>>>on the re-search, you will die...
>>>>
>>>
>>>If you mean, do I count value >= alpha+1 from zero-window search as a fail high,
>>>then no. In that case, I research with the original alpha;beta window. Isn't
>>>that ok?
>>>
>>
>>
>>What if you run out of time?  You failed high on the null-window search.
>>You started a new search with a wider window and time expired.  Do you play
>>the fail-high move or stick with the previous best move?  I stick with the
>>last verified move.
>>
>>Unless I fail high a second time which means the original aspiration window
>>was too small and I am now going to +infinity.  I trust the second fail high
>>but not the first.
>
>Does it mean that even if the evaluation before the fail high was mate against
>youself you are not going to play the fail high move if you have not time to
>verify it?


Maybe.  I _always_ invest the necessary time to complete the _second_ re-search
so that this doesn't happen.  Of course, it is _possible_ that I don't have
enough time to do this, but it has not happened (or even come close to happening
yet).



>
>In this case it is a mistake.
>It seems better to have a rule based on the previous evaluation when to play the
>fail high move that you cannot trust and when not to play it and not to play it
>only when the evaluation is bigger than some number.
>
>I have no idea about the exact number and looking at positions when Crafty did
>not play a move that fail high may help to get a better idea.
>
>Uri


I have a program that goes thru my old logs looking for this (among other
things).  That is how I originally found that I was failing high on the null-
window search, playing that move, and getting ripped for doing so...





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.