Author: Mridul Muralidharan
Date: 21:31:37 04/13/04
Go up one level in this thread
On April 08, 2004 at 05:23:33, Tord Romstad wrote: >On April 08, 2004 at 01:02:42, Mridul Muralidharan wrote: > >>Hi, >> >> Importance of endgame knowledge has been time and again discussed here. >>But it is quiet frustrating for a chess patzer like me to implement this in the >>engine - much more if I attempt to do this effectively. > >I agree entirely. Adding endgame knowledge is very difficult, especially if you >want to >keep the code general rather than writing huge amounts of code for dozens of >different >special cases. And if you give up and start writing code for all sorts of >special cases, you >face the ugly problem of evaluation discontinuities when pieces are exchanged. > >Not only the eval, but also the search is difficult in the endgame. All the >well-known >selective search techniques seem to fail miserably in the endgame. Chess Genius >does >something which seems to work rather well, but I haven't been able to figure out >what >it is. Hi Tord, Same problem that I am facing ... but there are so many things that I can add for endgame evaluation that it is not much of a problem for me and still have not yet hit special case endings problems. Lots of things that I had in messchess are still not in this engine ! I have not seen Chess Genius play and so have not much of an idea of how it works - for me double null move works beautifully and so I dont have much problem with nullmove. Usually I disable singular extensions in endgame : have not found it much of a use in endgame. Another important plus for endgame I found was pushing passed pawn in qsearch if distance_to_queen <= 3 : Have not yet added this to new engine though I had this along with a host of other threatening moves in mess's qsearch. > >>[White "Borer ver 2.01 by Mridul Muraliaharan"] > >Is Borer a new name for MessChess, or an entirely new engine? > >Tord Entirely new engine - I busted mess's iterative search very very badly and introduced some very bad eval inconsistencies some time before cct6 : combined with the threading inconsistencies it had (which had major problems with my mp code) I decided to rewrite. It is much cleaner now ... the other engine evolved over 2 years dating back to antiquity when I had next to no knowledge about chess programming :) Mridul PS : Still in vacation - so more replies only after Sunday :)
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.