Author: Tony Werten
Date: 12:35:06 01/19/02
Go up one level in this thread
On January 19, 2002 at 08:23:04, Bas Hamstra wrote: >On January 19, 2002 at 03:09:56, Tony Werten wrote: > >>On January 18, 2002 at 14:47:28, Bas Hamstra wrote: >> >>>On January 18, 2002 at 12:36:55, Gian-Carlo Pascutto wrote: >>> >>>>On January 18, 2002 at 12:26:00, Bas Hamstra wrote: >>>> >>>>>You can do only "evasions" or you can do checkgiving moves, in which case you >>>>>must also do evasions. I do checkgiving moves. >>>> >>>>My biggest question is always how do you manage this without >>>>blowing the search totally out of proportions :) >>>> >>>>-- >>>>GCP >>> >>>I don't know how others do it, but it helps a lot if you don't allow too many >>>consecutive checks that are not captures. Many do only checks at the first ply, >>>and that already helps a lot against the blind spots. Others do it more extreme >>>and are not even using SEE pruning. Did you ever play SOS? He does it pretty >>>extreme, but still has a very strong program (ex amateur world champ). >>> >>>Question for me is how good it "scales". In blitz it is no question: it is >>>better. To prove this, see below a blitz tournament with 119 engines after round >>>25 :-) >> >>You could have a point. I also use checkgivers and at the speedchess tournement >>at the WMCC I did a lot better than one would expect from my modest hardware. > >Tony why don't you release a winboard version? I get very interesting data from >german CC enthousiasts, who play a lot of matches between dozens of programs. I >just found out Tao-Pepito 0-6 at 60/0. Painful, but probably that's the stuff >you can learn the most from. There already is a winboard version if XiniX. I use it to test at home. Just didn't have the time to completely implement all features. Have to finish Bert and Ernie first. Tony > >Best, >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.