Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Ruffian 0.76 is still playing incredible strong!

Author: Uri Blass

Date: 23:17:19 09/19/02

Go up one level in this thread


On September 19, 2002 at 20:16:29, pavel wrote:

>On September 19, 2002 at 20:02:44, Robert Hyatt wrote:
>
>>On September 19, 2002 at 18:21:10, pavel wrote:
>>
>>>On September 19, 2002 at 18:06:32, Uri Blass wrote:
>>>
>>>>>>>>
>>>>>>>>That's the point.  I can't imagine someone working on an engine, in a vacuum,
>>>>>>>>for a decade, before anyone finds out about it.  I can't imagine anyone writing
>>>>>>>>a program in 6 months that would come close to beating _any_ top program.
>>>>>>>>
>>>>>>>>Anything is possible, I will agree.  But the probability is so low that it
>>>>>>>>will make people wonder...
>>>>>>>
>>>>>>>I hope you realize that you have essentially *publicly* leveled an accusation
>>>>>>>without presenting any evidence. I would prefer in such cases that either an
>>>>>>>investigation be conducted discretely or a wait and see policy be followed.
>>>>>>
>>>>>>
>>>>>>I haven't leveled _any_ accusation whatsoever.  I asked a direct question,
>>>>>>pointing out what has happened in the past.  I have _not_ said that this
>>>>>>program is a "clone" or anything else, and you can feel free to find some-
>>>>>>thing I wrote that does make that claim...
>>>>>>
>>>>>>As I said previously, I find it _hard_ to believe that a newcomer would be
>>>>>>that strong.  _not_ impossible, just _unlikely_.  I stand by that...  If some-
>>>>>>one takes that the wrong way, fine by me...
>>>>>
>>>>>No offense.
>>>>>But you had the same impression when YACE came out.
>>>>
>>>>Yace first release was clearly wekaer than yace of today.
>>>
>>>Yes I know that.
>>>But the time between the first released version of Yace and the last version of
>>>Yace was very short, if you consider the increase in strenth.
>>>It was when Yace starting doing as good (or even better) that Bob thought it was
>>>a crafty clone.
>>>Later Dann verified himself and convinced Bob.
>>>
>>>
>>>
>>
>>I don't remember that happening, but if it did, Dann can refresh my
>>memory.  I have never claimed a program _was_ a clone without proof.  For
>>Le Petite I posted the proof here in CCC.  As I did for Voyager.  As I did
>>for Bionic Impact.  As I did for ...
>>
>>
>>
>>
>>>>
>>>>>It is a fact that Yace came out of blue and became as strong as Crafty in a very
>>>>>short time.
>>>>>Same can be said about Aristarch, which also improved in a very short time.
>>>>
>>>>Aristarch first release was clearly weaker than aristarch of today.
>>>>
>>>>Aristarch also played before it was released in one of the CCT tournaments if I
>>>>remember correctly.
>>>
>>>
>>>Yes, true.
>>>But the time differance between the 2 versions were short, consider the playing
>>>strength.
>>>Which goes on to say that it is very much possible to come up with a strong
>>>version in a short time.
>>>
>>>Nothing impossible, nothing hard to believe.
>>>
>>>cheers,
>>>pavs
>>
>>
>>Again, this is _not_ the same thing.  Yace improved a lot, quickly.  But the
>>author was here participating in discussions, present on ICC, etc.  So he
>>didn't just "drop out of the sky".  When _that_ happens, suspicions are bound
>>to be tweaked...
>
>Actually, Dieter started posting in here and log in to ICC much later IIRC.
>At first he only used to post at winboard forum IIRC.

It is not important for this discussion.

The point is that the author of Ruffian does not post in the winboard forum and
I do not know that he post in another forum.

I never read him discussing about ideas to improve Ruffian or asking questions
about what other programmers know.
Dieter got a fidback after releasing Yace that helped him to improve yace.

The author of Ruffian never did it.

If the first version of some programmer is so strong it is natural to suspect
that it is a clone(it does not have to be only crafty clone and it may be a
program that choose if to use Crafty Comet or another winboard engine based on
the root position).

If Russel is going to release a strong chess program I will be less suspicious
because he asked a lot of questions here.

Uri



This page took 0.02 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.