Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty 17-10 v Fritz 6a Nunn 1 @ 120'/40 + 60'/20 + 30'

Author: Mogens Larsen

Date: 08:18:37 04/27/00

Go up one level in this thread


On April 27, 2000 at 10:03:26, blass uri wrote:

>I disagree because tablebases are part of the engine.
>Some engines were developed under the assumption that there are tablebases.
>I read that nimzo7.32 does not know some simple endgame likr KRK because it
>assumes it can use tablebases for them.

A program should be capable of playing simple endgames without tablebases.
Otherwise it's not a strong program IMO. Just because its capable of accessing
tablebases, doesn't mean it's a part of the program. These points is also worth
considering:

1) Weaker engines would appear stronger than they are.
2) Not all programs uses tablebases.

>Tablebases are chess knowledge.
>I do not see why some rules of search to find the right move by search and
>evaluation are knowledge when tablebases are not knowledge.

There's a difference between statistical knowledge and evaluation knowledge
IMHO.

>No EGTB tries to prevent errors also in the middle game when a move is going to
>lead to the wrong endgame.

That only makes my argument stronger, because the margin of evaluation error is
reduced even more. Thereby making weaker programs appear stronger than they are.

>You test by this way knowledge about the opening position without opening book
>and if one engine has problems with this position by search then it has a
>disadvantage.

Exactly.

>The only reason to test engines is to know which program is better for analysis
>and not for games and in this case it is better to use positions out of book
>because people use engines to  analyse usually these positions

That is your opinion, not common knowledge. From a CS viewpoint the evaluation
and search routines are more important, I think. If you want to use it for
analysis, you want the entire package, which isn't the point we're discussing
here I believe. The Nunn positions usually have nothing to do with the positions
you are most likely to analyze. Let's use an example:

I want to analyze a middlegame position, which would be very common for most I
believe. I'm way past the opening and has yet to reach a position where the EGTB
can help me. What do I do? I think I would use the strongest search and
evaluation engine available or my brain. Since I'm a lousy chess player I'd
probably go with the engine :o).

>I do not know
>You assume that Crafty's book is better for crafty relative to Fritz's book.

No, I assume that the Fritz book is better for Fritz and nothing else.

Best wishes...
Mogens



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.