Author: Robert Hyatt
Date: 21:22:51 02/12/01
Go up one level in this thread
On February 12, 2001 at 22:56:56, Uri Blass wrote: >On February 12, 2001 at 17:36:11, Robert Hyatt wrote: > >>On February 12, 2001 at 16:59:47, Christophe Theron wrote: >> >>>On February 12, 2001 at 16:48:10, Christophe Theron wrote: >>> >>>> >>>>And I remember that Chess Tiger (was it version 11.2?) has won a KRB vs KR >>>>endgame against Zugzwang in Paderborn 2 or 3 years ago. >>>> >>>>The programs were not using tablebases, and as far as I remember Zugzwang was >>>>using a bunch of processors. But it lost. >>>> >>>> >>>> Christophe >>> >>> >>> >>>Well... Actually it was maybe P.Conners, and not Zugzwang. Maybe Thorsten >>>remembers better than me. Anyway the opponent was using a LOT of processors. >>> >>> >>> Christophe >> >> >>I'd be willing to play you KRB vs KR with my program having the rook and >>running on a 486. :) > >It is easy to draw when you have tablebases. >The point is that part of the opponents(humans and part of the programs of >today) do not have tablebases and there is a practical chance to win against >them with the bishop. > >Uri My point was this: If a program is going to compete with Crafty, Ferret, Hiarcs, Fritz, and other programs that probe the TBs in the search, then it is going to have _grave_ difficulties if it makes the silly assumption that KRPP vs KRB is winning for the KRB side. That side will _never_ win. And it will have to fight like the devil to even draw, all the while thinking it is winning. Assuming less than perfect play by the opponent is _not_ the way to make a program stronger. Closing each and every hole that is exposed is better. It is _not_ hard to solve this problem in the right way, rather than hoping for an opponent to blunder.
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.