Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Is there program, which evaluates this as draw?

Author: Torstein Hall

Date: 05:28:07 11/08/01

Go up one level in this thread


On November 08, 2001 at 08:20:27, Vincent Diepeveen wrote:

>On November 08, 2001 at 07:30:45, Hans van der Zijden wrote:
>
>>
>>>>if there is such a program, it will probably be not very strong, because the
>>>>extra chessknowledge will slow it down too much.
>>>>
>>>>Hans.
>>>
>>>I disagree that the extra knowledge is going to slow it too much.
>>>I do not think that the knowledge that is needed it too compliacated.
>>>
>>>Uri
>>
>>You are right about this one situation. Unfortunately there are millions of
>>other positions where the program will not have a clue. If you implement
>>chessknowledge to cope with all these positions, I garantee you that the program
>>will be very slow.
>
>define slow please. I hear so many amazing opinions about knowledge inside
>chess programs. Yes diep is dead slow, but that's because it scans
>a lot. However when you talk about patterns then you can see all the
>knowledge in Theta(log2 n).
>
>So 100000 patterns, just take the 2 log out. that's 16 times slower than
>a normal evaluation (with the same datastructure that is).
>
>Best regards,
>Vincent


How deap can Diep search in normal games? (On normal hardware, say AThlon TBird
1.4)
Can a program with more knowledge search faster then "expected" because it can
stop many bad lines earlier?

Torstein




This page took 0.01 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.