Author: Albert Silver
Date: 11:11:01 08/19/04
Go up one level in this thread
Plus it slows the search down, so the tricky part (especially having
>>no idea what the program knows or doesn't) is in finding a balance where the
>>slowdown is worth it.
>
>
>here i cannot disagree. only that i have a different point of view. IMO the
>search is completely unimportant, for me the evaluation is the boss, not the
>search. the search is helpful, but the score is much more important IMO.
I have to disagree with you here. I'm not implying the search is necessarily
more important, but it is absolutely not less. The most important thing is to
find good moves, and whether this is achieved via search or the evaluation is
unimportant. Programs obviously use both, and the purpose is obviously to find
the most efficient balance. You've heard of the adage "Chess is 99% tactics",
well it's very true I'm afraid. You can formulate all the plans you want, but if
you have a hole in the variations supporting your move, it won't make a
difference how refined your understanding is. Of course, sometimes the limits in
calculation make it impossible to compensate properly lack of knowledge, so as I
said, the key is balance.
>when looking into your style it seems you trust on Ed's % comments in the REM
>part and you try to increase the search strength by using those switches Ed
>claims to get + x %.
Yes, of course. I'm sure he didn't pick those numbers out of a hat after all.
I should add that I'm not terribly impressed with the way it is handling the
center and plan to make some small changes again, but I'll test more before I
do.
BTW, I disagree with you on the Nunn positions. I think they are extremely
interesting as they represent typical middlegame structures and plans, covering
a full range of choices. I happen to completely agree with him when he states in
his intro on the second set that those who claim an engine or such is tuned to
these 30-35 positions for both sides and others not are probably missing the
point. These positions really do a good job of covering the general gamut of
typical chess middlegames, and if an engine is incapable of handling them on
average (it can certainly play some better than others), maybe it just doesn't
play very good chess.
Albert
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.