Author: José Carlos
Date: 17:45:28 05/24/02
Go up one level in this thread
On May 24, 2002 at 20:09:39, Christophe Theron wrote: >On May 24, 2002 at 19:41:52, José Carlos wrote: > >>On May 24, 2002 at 18:14:57, J. Wesley Cleveland wrote: >> >>> I was looking at the source of crafty and saw that on a fail low, the value >>>returned is alpha, rather than Max of all the values from the level below. Is >>>there some disadvantage in returning the real upper bound? I think it causes >>>some problems in Crafty making it much slower resolving fail-highs and fail-lows >>>at the root as these branches all need to be re-searched. >> >> I use fail soft because I think (not tested) that it generates smaller trees >>generally. >> The drawback is that what you call "real upper bound" is no always so. Null >>move and hashing make the search return "not totally correct bounds" some times. >>It is very common that you fail high in the root, research with a different >>window and get a fail low. >> So fail hard makes search more stable (it's less probable to get such >>instabilities if you always return the search bounds). >> >> José C. > > > >In this case you achieve stability by ignoring some information. > >That's an arbitrary choice anyway, and only testing can tell you which works >best. Logical reasonning will not. > > > > Christophe Yes, I admit my choice is arbitrary. When you have very little time for a chess program (my case) you can't test everything, so you spend your time in the bottlenecks of the program. But AFAIK, there's no conclusive test in fail soft vs fail hard debate and it seems to depend a lot on the rest of the choices you make about your search, just to try to answer the original question of the thread. José C.
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.