Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Null-Move: Difference between R = 2 and R = 3 in action

Author: Uri Blass

Date: 12:45:16 07/15/02

Go up one level in this thread


On July 15, 2002 at 14:42:52, Robert Hyatt wrote:

>On July 15, 2002 at 13:30:54, Gian-Carlo Pascutto wrote:
>
>>On July 15, 2002 at 13:11:09, Christophe Theron wrote:
>>
>>>Why does double null move prove that null move is a correct search method????
>>>
>>>Doing two null moves in a row means going back to standard search (a search not
>>>involving an illegal move like null move is).
>>>
>>>I fail to see how it legitimates null move.
>>
>>I think the idea is simply to show that you can nullmove and still
>>have a correct search. (but only with double nullmove, not single)
>>
>>--
>>GCP
>
>
>I think the discussion is pointless, myself.  Null-move search is _not_
>equlvalent to non-null-move search.  It demonstrably has more errors due to
>reducing the depth on many sub-trees.
>
>The concept of "correct search" is therefore so abstract as to mean absolutely
>nothing.  double-null eliminates _some_ zugzwang problems, but not _all_.  It
>is obviously impossible to eliminate all of them, particularly when you are
>whacking 6 plies off the tree depth and then drawing conclusiong about the
>resulting sub-trees..
>
>The only useful definition of "correct search" would be to somehow prove
>that a null-move search to depth=D would produce the _same_ move as a
>non-null-move search to depth=D, for all positions.  This is simply not
>possible.

No
The definition of correct search based on my understanding is  a search that
enable you to solve every position if you search deep enough.

Double null move with no bugs is a correct search based on that definition(you
also need not to use hash tables to prune the tree because it is clear that
scores based on hash tables may be wrong because of hash collision).

A search that is using normal null move and no double null move is a wrong
search because it may fail in zunzwang positions even if you give it to search
100000000000000000 plies forward.

Uri



This page took 0.02 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.