25,161
edits
Changes
Created page with "'''Home * Search * Path-Dependency''' '''Path-Dependency''' describes a situation when the same position gets a different value depending..."
'''[[Main Page|Home]] * [[Search]] * Path-Dependency'''
'''Path-Dependency''' describes a situation when the same [[Chess Position|position]] gets a different value depending on a sequence of [[Moves|moves]] by which it has been reached. Whilst a few programs do this by design <ref>[http://www.talkchess.com/forum/viewtopic.php?topic_view=threads&p=115849&t=13559 Re: Methods to stably evaluate nodes?] by [[Uri Blass]], [[CCC]], May 05, 2007</ref>, within the standard [[Alpha-Beta|alpha-beta]] framework path-dependency is considered undesirable, if impossible to avoid.
=Typical Cases =
* The position can be treated differently if there is a chance of [[Repetitions|repetition]] or hitting into a [[Fifty-move Rule|fifty-move rule]].
* Different [[Move Ordering|move ordering]] can cause different result (different moves causing a cutoff), when it comes from the [[Transposition Table|hash table]] <ref>[[Tony Warnock]], [[Burton Wendroff]] ('''1988'''). ''Search Tables in Computer Chess''. [[ICGA Journal|ICCA Journal]], Vol. 11, No. 1, pp. 10-13. </ref>
* Hf a program uses [[Extensions|search extensions]], search depth may be different (i.e. 1. d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 vs 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Nc3, where only the latter triggers a [[Check Extensions|check extension]])
* History-dependent heuristics, like some versions of [[Late move reductions|LMR]] may prune different sub-trees.
=See also=
* [[Graph History Interaction]] (GHI)
* [[Transposition]]
* [[Transposition Table]]
* [[Repetitions]]
* [[Fifty-move Rule]]
=Forum Posts=
==2000 ...==
* [https://www.stmintz.com/ccc/index.php?id=182927 "Don't trust draw score" <=Is it true?] by [[Teerapong Tovirat]], [[CCC]], August 08, 2001 » [[Repetitions]], [[Graph History Interaction]]
==2005 ...==
* [http://www.open-aurec.com/wbforum/viewtopic.php?t=2320 path dependent evaluation] by [[Daniel Shawul]], [[Computer Chess Forums|Winboard Forum]], April 20, 2005
* [https://www.stmintz.com/ccc/index.php?id=430487 Re: And a still unsolved test position] by [[Uri Blass]], [[CCC]], June 09, 2005 » [[Movei]]
* [http://talkchess.com/forum/viewtopic.php?t=21343 Semi-Path Dependent Hashing: a semi-useless idea] by [[Zach Wegner]], [[CCC]], May 24, 2008 » [[Transposition Table]]
==2010 ...==
* [http://www.talkchess.com/forum/viewtopic.php?t=40388 Repetitions/50 moves and TT] by [[Sergei Markoff]], [[CCC]], September 13, 2011
* [http://www.talkchess.com/forum/viewtopic.php?t=44167 Texel recipe to fix TT draws scores] by [[Marco Costalba]], [[CCC]], June 23, 2012 » [[Texel]]
=External Links=
* [https://en.wikipedia.org/wiki/Path_dependence Path dependence from Wikipedia]
=References=
<references />
'''[[Search|Up one Level]]'''
'''Path-Dependency''' describes a situation when the same [[Chess Position|position]] gets a different value depending on a sequence of [[Moves|moves]] by which it has been reached. Whilst a few programs do this by design <ref>[http://www.talkchess.com/forum/viewtopic.php?topic_view=threads&p=115849&t=13559 Re: Methods to stably evaluate nodes?] by [[Uri Blass]], [[CCC]], May 05, 2007</ref>, within the standard [[Alpha-Beta|alpha-beta]] framework path-dependency is considered undesirable, if impossible to avoid.
=Typical Cases =
* The position can be treated differently if there is a chance of [[Repetitions|repetition]] or hitting into a [[Fifty-move Rule|fifty-move rule]].
* Different [[Move Ordering|move ordering]] can cause different result (different moves causing a cutoff), when it comes from the [[Transposition Table|hash table]] <ref>[[Tony Warnock]], [[Burton Wendroff]] ('''1988'''). ''Search Tables in Computer Chess''. [[ICGA Journal|ICCA Journal]], Vol. 11, No. 1, pp. 10-13. </ref>
* Hf a program uses [[Extensions|search extensions]], search depth may be different (i.e. 1. d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 vs 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Nc3, where only the latter triggers a [[Check Extensions|check extension]])
* History-dependent heuristics, like some versions of [[Late move reductions|LMR]] may prune different sub-trees.
=See also=
* [[Graph History Interaction]] (GHI)
* [[Transposition]]
* [[Transposition Table]]
* [[Repetitions]]
* [[Fifty-move Rule]]
=Forum Posts=
==2000 ...==
* [https://www.stmintz.com/ccc/index.php?id=182927 "Don't trust draw score" <=Is it true?] by [[Teerapong Tovirat]], [[CCC]], August 08, 2001 » [[Repetitions]], [[Graph History Interaction]]
==2005 ...==
* [http://www.open-aurec.com/wbforum/viewtopic.php?t=2320 path dependent evaluation] by [[Daniel Shawul]], [[Computer Chess Forums|Winboard Forum]], April 20, 2005
* [https://www.stmintz.com/ccc/index.php?id=430487 Re: And a still unsolved test position] by [[Uri Blass]], [[CCC]], June 09, 2005 » [[Movei]]
* [http://talkchess.com/forum/viewtopic.php?t=21343 Semi-Path Dependent Hashing: a semi-useless idea] by [[Zach Wegner]], [[CCC]], May 24, 2008 » [[Transposition Table]]
==2010 ...==
* [http://www.talkchess.com/forum/viewtopic.php?t=40388 Repetitions/50 moves and TT] by [[Sergei Markoff]], [[CCC]], September 13, 2011
* [http://www.talkchess.com/forum/viewtopic.php?t=44167 Texel recipe to fix TT draws scores] by [[Marco Costalba]], [[CCC]], June 23, 2012 » [[Texel]]
=External Links=
* [https://en.wikipedia.org/wiki/Path_dependence Path dependence from Wikipedia]
=References=
<references />
'''[[Search|Up one Level]]'''
