Author: Robert Hyatt
Date: 15:53:02 01/18/99
Go up one level in this thread
On January 18, 1999 at 17:18:12, Pauli Misikangas wrote: >Hi all! > >I'm trying to optimize the size of the alpha-beta window for my shogi program. I >understand that if I use a small window, the search terminates faster but >re-searches (with full window) are needed more often. On the other hand, >re-search is not so bad because I should get some useful values from the >transposition table, right? So, what would be a good hit/re-search balance? > >Do you use always the same window size or change it during the game? I've been >thinking that when nothing crucial is happening in the game (maybe during the >opening and early middlegame) one might get some advantage by using a very >narrow window. When the fight really begins, one could then start using a wider >window. Do you think that this might be worth of trying? Is anybody using >techniques like this? Is there some cases when it is better not to use >AB-window? > >All the best, > >Pauli Misikangas > >Pauli.Misikangas@cs.helsinki.fi >http://www.cs.helsinki.fi/~pmturune/index_e.html This is a classic question that can be answered experimentally. There is no 'holy grail' for all programs. The iteration-to-iteration swings can cause each program to have different. Try it with different windows, over a large test of positions, and just pick the window that works best (produces the fastest time-to-depth.) One trick.. watch your odd/even score swings (if you have them). And if you see them in your program, anticipate by correcting the window so that the odd/even swings don't take you over the edge too frequently...
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.