Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Why can't Kasparov play Shredder 7.04 or 8.0?

Author: Terry McCracken

Date: 06:27:29 11/07/03

Go up one level in this thread


On November 07, 2003 at 08:45:11, Sune Fischer wrote:

>On November 07, 2003 at 08:13:40, Terry McCracken wrote:
>
>
>>Sune your asking the wrong question, as this statistical analysis really means
>>very little.
>
>Hey great.
>
>I was trying to provoke a proper argument from you, your last one didn't suffice
>IMO :)
>
>>I'm annoyed so much _faith_ can be attributed to such such small numbers as
>>proof of Shredder 7.04s strength.
>
>But here is the thing, if Shedder beats Fritz, then Shedder has *proven*
>it has a better understanding of the game.

I didn't know that a programme could have understanding?;-)
>
>The "understanding" includes everything from tactics to positional play, what
>you lack in one place you can make up for somewhere else.

With humans, I agree.
>
>The old "I would have had a won endgame if you hadn't mated me" argument
>falls to the floor here, winning is all that counts :)

Very true.
>
>>What would be interesting is the quality of the games, and how much book
>>influence contributed to the wins as well.
>
>I don't think quality is that easy to compare, honestly I think all engines look
>pretty stupid now and then.

You're correct. To really answer the question which programme is best, best
engine, best book, (which plays a large part as they are tuned for beating other
computers in tourneys) is to have a panel of top Grandmasters study the output,
to study the actual games. Of course this isn't practical, and it's not going to
happen.

I'd like to have this data, as it would be telling, but it would be very time
consuming, and isn't forthcoming. I'm certain in the realm of human play this
data is revealed more than in the field of computer chess.

I'll concede that under the condition the SSDF plays these games under
autoplayers, Shredder has a very small edge.
>
>Whether Fritz is better than Shredder (against Garry) is impossible to know, but
>the circumstantial evidence (SSDF) suggests it probably isn't.

I agree.
>
>>It could be with a very small change
>>in parameters that Fritz 8 would be sitting at top spot.
>
>I'm not so sure, it looks like a pretty solid lead right now for Shredder.
>
>In any case you could make the same argument the other way around, that a small
>change in Shredder would make it even stronger.

Yes, that's true.
>
>>Of course there is some chance at work here as well. I thought my point would be
>>understood. I guess not..
>
>I don't see a good argument in favor of Fritz being the strongest.

I don't think I was trying to _prove_ it is!;-)

Terry

P.S.

Forgive my lazy response, I'm in need of sleep at the moment....but I doubt I'll
get any...sigh..
>
>-S.
>>Terry



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.