Author: Robert Hyatt
Date: 20:44:29 01/28/04
Go up one level in this thread
On January 28, 2004 at 23:27:03, Russell Reagan wrote: >On January 28, 2004 at 22:50:05, Robert Hyatt wrote: > >>Bruce's case is a pathological problem that will happen. But it is caused by an >>extreme happening. In a normal search this won't/can't happen (assuming you are >>not using null-move). But in reality it can. However, 99.9% of the time, >>re-searching with beta,+infinity after a fail high on alpha,beta will produce a >>score as expected... > >Okay, let me see if I understand what's going on. > >In your newsgroup post, you were speaking from a theoretical point of view, >while Bruce was speaking from a practical view where search instability has >already crept in (via null-move, or whatever). > >As long as you're doing pure alpha-beta with aspiration search, it is safe to >assume that the real score of a failed aspiration search will be (-infinity, >alpha+1) for a fail low, and (beta-1, +infinity) for a fail high. No. If you eliminate the transposition table, this is guaranteed to be true, but with the transposition table, a potential fail-high followed by a fail-low situation can occur at any time. > >This kind of search instability problem is introduced by the same position being >handled differently at different parts of the tree due to their different paths >(ex. use of a transposition table, null-move or any forward pruning done based >upon alpha or beta, 3-fold-repetition, 50 move rule, etc.). > >Is that all correct? That is correct, yes...
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.