Author: Tom Kerrigan
Date: 21:23:38 03/23/00
Go up one level in this thread
On March 23, 2000 at 22:12:44, John Coffey wrote: >Say hypothetically the computer is looking 3 moves ahead. Quiescent search is >causing is to look even deeper on some lines. It might try some useless move >and then the quiescent search causes it to look deep enough to see that there is >still a win there (which could happen in some endgames.) So it has now chosen a >continuation that is 5 moves long. Now when it comes across the same win in a 3 >move continuation (without the useless move to begin with) then it will reject >it because it has the same score. > >Maybe this scenario wouldn't happen because on the 2 ply search it would have >found the win using quiescent search. I suspect that this is the answer. > >I just want to avoid a situation that I have seen many many years ago where a >program will keep making useless moves because it still sees the win but isn't >particular about how fast that win occurs. > >John This is a well known problem that is fixed in most (all?) programs. Personally, I save the last mate score returned by the search, and if the current search does not beat that mate score, I keep searching. If you're talking about a situation that is not mate, i.e., positional advantage, then I think problems like this are virtually unsolvable. My program will sometimes play e.g. Bd3 blocking its own pawn on d2. No matter what kind of penalty I have for this situation, my program will invariably play this stupid move. The reason is that it thinks it will move the bishop off of d3 later in the PV, so the penalty is never enacted. Very frustrating. -Tom
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.