Author: Bas Hamstra
Date: 07:18:43 10/26/00
Go up one level in this thread
On October 26, 2000 at 07:45:10, Leen Ammeraal wrote: >On October 26, 2000 at 02:37:29, Bruce Moreland wrote: > >>On October 26, 2000 at 01:17:44, Leen Ammeraal wrote: >> >>>On October 26, 2000 at 00:46:23, Bruce Moreland wrote: >>> >>>> >>>>On October 25, 2000 at 23:26:15, Pham Minh Tri wrote: >>>> >>>>>Hi, >>>>> >>>>>Could someone explain to me how SEE (static exchange evaluator) function works >>>>>or show me books, old posts (when) or links to read about it? >>>>> >>>>>Thanks in advance. >>>>>Pham >>>> >>>>It is not mentioned in any books that I know of. >>>>....... >>>>bruce >>> >>>It is discussed in detail, under the heading 'Swapoff', in >>>'The Chess Computer Handbook' by David Levy, 1984, >>>ISBN 0 7134 4220 4. >>>Leen Ammeraal >>>http://home.wxs.nl/~ammeraal/ >> >>Having no reason to dispute this, I stand corrected. Is the description any >>good? >> >>bruce > >Levy's description is is not particularly easy reading, >and it is not presented as a ready-to-use algorithm, >but it shows what swapoff is about. I borrowed the book >some time ago and photocopied those 3.5 pages about >this subject. I still have to implement it, and I plan >to do this one of these days. >Leen Actually I think there are 2 "schools". One school thinks the qsearch should be kept as small as possible, tries to do as much as possible in the full width search, because the qsearch "is unreliable anyway", to quote Bob Hyatt. SEE pruning in the qsearch and futility pruning is more or less standard. The other camp tries to get an accurate score from the qsearch and is horrified by the idea of a). I am not sure they even want to do SEE pruning, probably not. Regards, Bas.
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.