Author: Robert Hyatt
Date: 21:41:11 09/19/02
Go up one level in this thread
On September 19, 2002 at 23:44:16, Jeremiah Penery wrote: >On September 19, 2002 at 19:44:22, Uri Blass wrote: > >>Here is a move of Ruffian that Crafty18.5 cannot find >> >>[D]r1b1r1k1/pp3p1p/2pp2p1/8/2PNPPn1/2P3q1/P2Q2B1/R1B1R1K1 b - - 0 18 >> >>Ruffian played Be6 when Bf5 is the move that Crafty prefers(I do not express an >>opinion which move is onjectively better. >> >>Uri > >It could be a different version of Crafty that plays Be6. Or it could be a few >simple changes to the evaluation that make it play Be6 instead of Bf5. It's >almost impossible to compare a few positions in this way, because small changes >in the evaluation (the result would still be a Crafty 'clone') can make it play >almost completely different moves. I agree. The real test is in tactics. It is not easy to make a copy of Crafty either better or worse at tactics, but particularly making it better is non-trivial. It's far easier to twiddle with a few eval terms to make it play significantly different. Not that I think this has happened in this case, since I haven't seen any games or output of any kind... but making a determination of "it is or isn't crafty" is not a few minutes work. It takes a lot of time, even if the executable is available for "poking around in."
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.