Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Funny thing happened on the way of a Forum!

Author: Albert Silver

Date: 16:28:19 02/14/06

Go up one level in this thread


On February 14, 2006 at 17:23:16, Rolf Tueschen wrote:

>On February 14, 2006 at 17:14:09, Albert Silver wrote:
>
>>On February 14, 2006 at 16:39:50, Rolf Tueschen wrote:
>>
>>>On February 14, 2006 at 16:36:32, Albert Silver wrote:
>>>
>>>>On February 14, 2006 at 11:52:23, Robert Hyatt wrote:
>>>>
>>>>>On February 14, 2006 at 11:45:30, Duncan Roberts wrote:
>>>>>
>>>>>>On February 13, 2006 at 13:48:48, Robert Hyatt wrote:
>>>>>>
>>>>>>>You are missing a major problem.  Many professional engine authors have posted
>>>>>>>here, and answered questions, and the like.  But you have _not_ seen them
>>>>>>>discussing marketing ideas, nor beta testing results, nor any of the other
>>>>>>>"behind-the-scenes" activities that go on the the life-cycle of a professional
>>>>>>>chess program.
>>>>>>>
>>>>>>>That's the difference here.  The discussions ventured off into the "would you
>>>>>>>rather pay for a single copy or a subscription to multiple updates?"  and so
>>>>>>>forth, which _clearly_ do not belong here.
>>>>>>>
>>>>>>>That's that part of this that has gone over the top.  All the games played by
>>>>>>>Rybka are fine.  I can skip 'em if I am not interested, but they are about
>>>>>>>computer chess.
>>>>>>
>>>>>>changing the subject slightly, I would value your opinion on Rybka. Is it as
>>>>>>good as people say it is. ? any comments generally about its strength.
>>>>>>
>>>>>>
>>>>>>duncan
>>>>>
>>>>>I'm really the wrong one to ask.  Crafty has been on ICC sporadically the last
>>>>>few weeks, because of the offline testing I've been doing.  So I have not had
>>>>>the opportunity to watch very many games to develop an opinion about what Rybka
>>>>>is doing better than everyone else...
>>>>>
>>>>>Eventually I will start looking at ICC games and then perhaps will begin to see
>>>>>what sort of "trends" I notice.
>>>>
>>>>It won't take many games to spot a few glaring ones. IMHO.
>>>>
>>>
>>>
>>>Mod objection!
>>>Dont speak in secrets!
>>>Speak so that even Rolf understands what you are saying.
>>>Otherweise beheading should be appropriate.
>>>Rolf [= 5 sentences]
>>
>>:)  The secret is more in 'how' Rybka does what it does differently. Seeing part
>>of 'what' it does that is different from other programs won't take many games to
>>become visible.
>>
>
>
>Yes, I'm busy with my own program, so I had no time for Rybka, but _what_ again
>was the secret or the secrets you've seen, Chrilly had seen and Bob could see?

Elements such as the way it handles passed pawns in the middlegame and endgame,
compared to any other engine I've seen, are very obvious after one has seen a
few games. It doesn't take long to notice this, so the secret for this element
isn't 'what' it is, since I just described it, but indeed 'how' it was done.

                                 Headlessly yours,

                                         Albert

>And now the important question: would others be able to transform that secret
>into their own engines? Not that I needed your help, mine has 3500, but I ask
>out of scientific interest. If you laugh, you're already dead. Rolf is faster
>than the shadow of Luke, or than an atomic chain reaction...



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.