Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: New SSDF-list

Author: Thorsten Czub

Date: 15:42:35 02/22/98

Go up one level in this thread


>You are not going to compare 11 games in Paris with 200 by SSDF, are
>you.

I compare any data I have. And all my data says: it is not stronger.
And all my data fits to the paris data.

>I do. Anybody that tested Fritz 5 without prejudices does.

Prejudices ? We are talking about programs !
They play games, or they don't play games. They win or lose.


>It is that
>strong. Nimzo is strong too.

Nimzo is a different thing.
I had no problems in reproducing ANY good results with Nimzo !
I have told Chrilly and his friends about and I posted my Nimzo results
here the day I first time got them (arround christmas time).
But Fritz5 ?
I have this program for a long time.


> And Moritz was right all along.

Sorry. I don't believe in these results.
Substract the machine-improvement. Substract the learning stuff.
And still the "progress" is too big.
And IF there would have been a progress, we would have measured it
months before. And we didn't.
No - we can be different opinion Enrique. I doubt these results. There
is something wrong in the swedish data. I don't know what it is. But the
more crazy data they produce, the more it gets obvious.
What about doubled games ?
How do they handle those duplicated games ??

>I admit I
>didn't expect so much improvement from Fritz 5. But I knew it was as
>strong as Hiarcs 6, Rebel 9 or Nimzo98.


Sorry - but Fritz5 is not as strong as Hiarcs6 or Rebel9 or Nimzo98.

>
>Enrique



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.