Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Chess Engine Just for Endgames

Author: Robert Henry Durrett

Date: 15:32:43 06/19/02

Go up one level in this thread


On June 19, 2002 at 18:21:07, Matthias Gemuh wrote:

>On June 19, 2002 at 12:17:28, Robert Henry Durrett wrote:
>
>>
>>Suppose you were an endgame aficionado or you had a good friend who was and you
>>wanted to help him/her by providing a first-class endgame engine.  What would
>>you provide?
>>
>>What are your thoughts on this?
>>
>>Here are some thoughts of mine:
>>
>>Endgames are different from the middlegame in one important way.  There is a lot
>>of accumulated knowledge about endgames.  Much of this has been expressed in
>>"rules."  For example, if you have an endgame with a rook and a passed pawn, the
>>rule is to put your rook behind the passed pawn.  Another example, in king and
>>pawn endgames, is to use the opposition.  The list goes on and on.  Of course,
>>all rules have exceptions, so concrete analysis is still required.
>>
>>I would expect an endgame engine to evaluate positions for TYPE often,
>>especially whenever an exchange takes place.  The set of applicable "rules" are
>>generally quite different for different types of endgames.
>>
>>Searches strategies might also be optimized for each different TYPE of endgame.
>>
>>Incidentally, an "endgame engine" could be packaged as a subprogram of a larger
>>"complete game" chess engine.  Whenever the game moved into the endgame phase,
>>the endgame subprogram would be kicked in and the middlegame subprogram turned
>>OFF.
>>
>>Just a few thoughts.
>>
>>Bob D.
>
>
>
>
>All strong engines already have special routines for all this.

Glad to hear it.  Not too surprised.

I guess the recent threads on "knowledge-based" chess engines made me start
thinking about how well suited endgames are to a "knowledge-based" approach, and
how poorly middlegames are.  Unfortunately, those threads did not give me a warm
feeling about the way knowledge was being used.  Sounded too simplistic and
sub-optimal.

Bob D.





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.