Author: Robert Hyatt
Date: 07:55:51 10/14/03
Go up one level in this thread
On October 14, 2003 at 10:31:49, Tord Romstad wrote: >On October 14, 2003 at 10:06:41, Robert Hyatt wrote: > >>On October 14, 2003 at 08:42:05, Tord Romstad wrote: > >>>I think it is a good idea to use the SEE for checking moves, too. Only search >>>checks which do not lose material. >> >>Then you admit defeat before you start. See the simple draw by stalemate >>idea where you have a king and rook, and your king can't move. You just >>make unsafe check after unsafe check because the rook can't be taken. > >I don't see how not searching losing checks in the qsearch is admitting defeat >any more than not searching losing captures is. In both cases, you will make >mistakes, but also save lots of time. That's the point. "admitting defeat" means "accepting additional errors". I wasn't talking about losing games, specifically. The q-search based only on captures if _full_ of errors. It fails to understand any tactic except those specifically depending on captures. IE no pins. No threats. No double attacks. No skewers. No checks. If you add safe checks, that is a small part of the above list of things it will overlook. A part small enough that it probably won't help much, and it might hurt if it wastes too much time. Perhaps my "admitting defeat" was too vague a descriptive term?? > >>Safe checks are probably better than no checks, assuming you believe you >>have to do checks in the q-search. But, as I have said so many times, if >>your q-search is going to have so many holes, find a way to avoid putting >>such a heavy responsibility on the q-search in the first place. > >Sure. I wish I could get good results with a minimalistic qsearch like >yours. Unfortunately it never worked very well so far. But perhaps >Omid is right that a minimalistic qsearch works better with nullmove >R=2 than with R=3 (which I use). I will run some experiments with R=2 >and no checks in qsearch when I find the time. > >Tord I use R=3 to R=2, and I really haven't seen anything bad. I think Bruce did the same q-search in ferret. I know he did in 1996 for Jakarta because he suggested I try it as well as it was working fine for him... I assume he still does that but could be wrong. you might try the R=3~2 approach which might help solve your R=3 problem without qsearch checks.
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.