Author: Sune Fischer
Date: 04:31:13 02/26/04
Go up one level in this thread
On February 26, 2004 at 06:49:54, Vincent Diepeveen wrote: >On February 25, 2004 at 14:37:03, Sune Fischer wrote: > >>On February 25, 2004 at 14:23:46, Gian-Carlo Pascutto wrote: >> >>>On February 25, 2004 at 14:16:15, Sune Fischer wrote: >>> >>>>Actually the point isn't so much whether it is crushing or not, the point is >>>>that the right move may be played for the wrong reasons. >>>> >>>>The move might be good (objectively speaking forcing a win) but to be sure of >>>>that you need a fairly deep calculation, way too deep to be found in 1 second. >>>> >>>>When an engine makes the right move for the wrong reasons it is always cause >for concern, IMO. >>>> >>>>Bottom line it is a matter of "style", not tactical abilities, hence I'm not >>>>sure I'd consider it a good test position. >>> >>>I really dont care for the reasons my engine has, as long as it's playing the >>>right moves. >>> >>>If it's playing the wrong ones, then it's a time to care about reasons. >> >>I think there is a fundamental difference between "guessing" (eval) and >>"knowing" (search), at least when it comes to tactical test suites. >> >>Of course if your eval is super tuned then guessing can almost be as accurate as >>knowing :) > >So you still have no clue yet how much tactics gets solved by the commercial >software by using eval. Are you thereby insinuating that you do? :) We all have our little tricks of course, I try to only evaluate things that would otherwise require several plies of search, that's where the expected gain is the biggest IMO. -S.
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.