Computer Chess Club Archives




Subject: Re: The superior Rybka chess knowledge (With Corrections)

Author: Uri Blass

Date: 05:47:24 01/19/06

Go up one level in this thread

On January 19, 2006 at 08:40:30, Rolf Tueschen wrote:

>On January 19, 2006 at 08:35:23, Uri Blass wrote:
>>On January 19, 2006 at 08:27:30, Rolf Tueschen wrote:
>>>On January 19, 2006 at 03:00:07, Ernst Walet wrote:
>>>>On January 19, 2006 at 02:38:50, Tony Thomas Karippa wrote:
>>>>>On January 18, 2006 at 16:36:09, Rolf Tueschen wrote:
>>>>>>On January 18, 2006 at 16:30:58, Tony Thomas Karippa wrote:
>>>>>>>Most people would agree with me if I say you are the one that's wrong.
>>>>>>Yes, but they will also remember the voices from Paderborn which said "No, we
>>>>>>wont bust Rybka in the late endgames, we dont want to be unfair." Was that
>>>>>>promotion in Paderborn or sports? (Socrates speaking...)
>>>>>Who said that? Chilly pepper? Or Zappa?
>>>>I have stated, as operator of Rybka, that:
>>>>1 Rybka was never challenged this way (if applicable).
>>>>2 There was a very pleasant atmosphere.
>>>>3 I nevertheless would have no problem if someone had tried it.
>>>>My conclusion was that it was considered sportmanship not to push Rybka to this
>>>>point. RT has difficulties with that.
>>>Ernst, no, sorry,
>>>I had not then and I have not now a problem with your attitude, let me be not
>>>misunderstood. I see a problem in what other people made out of the Rybka win in
>>>Paderborn, in the line, as if Rybka would be the proven, unanimous, clear,
>>>single, brilliant and all-time best program of the World. All I say is that this
>>>shouldnt be said on the base of Paderborn for the given reasons and on the base
>>>of the now existing data. No matter if Rybka wins test games after testgames,
>>>Rybka is in my eyes
>>>a) not even a completed engine, it is still in making, it's in a socalled
>>>rudimentary state, its strength has by no way been proven by any single test
>>>result; simply because all test results are still NOT significant; and we all
>>>know from experience that the real test for Rybka will come the moment when
>>>other programmers have begun to tune on Rybka itself, while today Rybka is tuned
>>>on everybody else;
>>I think that you are clearly wrong here.
>>It may be possible to tune a program against another program but not against all
>> of course then the actual reports must be re-written; you can
>>>all see what will happen or what is possible in Hiarcs with these extra settings
>>>from Czub; only then we will know what Rybka can really do - and we certainly do
>>>hope that not the same thing will happen with Rybka what happened with Ruffian
>>>which once also was a new shooting star, then went xommercial and thereafter
>>>never was een again... (included its programmer BTW - on that background it must
>>>at least be allowed to remain somewhat friendly, sympathetically reserved, no?!)
>>Ruffian is clearly a different case that is not relevant here.
>>Ruffian was never clearly the best engine even when it came commercial.
>Can I record live from you, Uri, that on Jan 19th, Rybka is the best
>chessengine? Would you say such a thing?

Today it is the best chess engine as the results of CEGT show.

Note that I am talking only about the engine and not about learning and opening
book that I consider as part of the program but not as part of the engine.

When you test it in matches from predefined positions it usually beats
everything and the choice of the positions is not intentionally to help rybka.


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.