Author: Robert Hyatt
Date: 12:56:44 09/22/98
Go up one level in this thread
On September 21, 1998 at 17:25:38, Inmann Werner wrote: >On September 20, 1998 at 20:58:08, Robert Hyatt wrote: > >>On September 20, 1998 at 17:49:17, Inmann Werner wrote: >> >>>On September 20, 1998 at 17:30:31, Robert Hyatt wrote: >>> >>>>On September 20, 1998 at 14:21:15, John Coffey wrote: >>>> >>>>>I think that the subject says it all. >>>>> >>>>>John Coffey >>>> >>>> >>>>Yes. You store the best move when you store a position, assuming you >>>>have a best move (you don't if you searched all moves and failed low, >>>>of course.) You should try this move first each time you encounter this >>>>position... >>> >>>Whats the thing with the fail low? >>>Why can I not use the Hash table for move ordering in this case? >>>(I do now!) >>> >>>Werner >> >> >>You can't... if you fail low, all you know is that for every move at that >>ply, the opponent has a move that "refutes" it. You have *no* idea about which >>move is best, which is worst, which are "in the middle". All you know, thanks >>to alpha/beta, is that *all* are bad. So which one would you try first? In >>this position, at this depth, or at a deeper depth? No way to know because >>*all* moves looked bad... > >OK. You are right. But in this occasion, there is no chance at all to get a >good first move to try. So why not use the hash entry and spare on "if"? >Its a little bit faster and there is no harm and no use at all. > >Werner The question is still the same... "Where do you get this move from?" When I store an exact score or a fail-high indication, I know that for the exact score, the best move is in the PV array, and for fail-high, the best move is the move I just searched. But *where* do you get one for the fail-low case? because *every* move failed low, which do you stick in the table? A random one? That would be bad because it is likely that trying good captures would be better than a move chosen at random... or trying killers, or trying history moves, or trying anything but a random one..
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.