This page is about the chosen procedure of the ICGA regarding the composition and installation of the Panel, the experts who were going to judge if Rybka was an original program or a derivative of Fruit as listed in the ICGA charter.

The ICGA procedure against Rybka
Inconsistencies
PART 1

From the Charter:

2 Membership

The Panel shall be open to:

[a] Programmers who have participated in any ICGA or ICGA World Championship
    for computer chess or in any Computer Olympiad or in any other computer
    competitions deemed appropriate by the ICGA.

[b] The office-bearers of the ICGA;

[c] Persons accepted as experts by the ICGA or by a unanimous decision of the
    Secretariat of the Panel.
The main criticism on [a] is that the ICGA gave Rybka's direct competitors a vote of whom many had an interest in a guilty verdict.

Because Rybka was disqualified and robbed from its 4 continues world champion titles (2007-2010) the following programs profited:

2007
1st Zappa (World Champion)
2nd Loop
=3rd GridChess
=3rd Shredder

2008
1st Hiarcs (World Champion)
2nd Junior
3rd Cluster Toga

2009
=1st Junior (Joint World Champion)
=1st Shredder (Joint World Champion)
=1st Deep Sjeng (Joint World Champion)

2010
=1st Rondo (Joint World Champion)
=1st Thinker (Joint World Champion)
3rd Shredder
As the panel voting shows Mark Uniacke (Hiarcs) and Stefan Meyer-Kahlen (Shredder) voted Rybka guilty and both became world champion.

This kind of pretence should have been avoided by all means and as such was addressed to the ICGA Board.

-----------------------

Date: Thu, 08 Sep 2011 17:31:35 +0200
From: Ed Schroder
Subject: Rybka in retrospect (part-2)
To: David Levy
Cc: Jaap van den Herik, board@icga.org

2. The installation of the Panel containing many of his direct opponents of whom many had an interest in a guilty verdict. This kind of pretence should have been avoided by all means.

...

I understand that criticism in retrospect is always easy and that doing new things a first time usually is far from perfect most of the time. Therefore instead of criticizing only allow me to suggest an improvement for future cases,

1. Suspect engines are investigated in silence by a small panel of fully independent experts without any interests or whatsoever.

2. When found guilty the evidence is offered to the accused programmer for comments. Insights may change or not because of that.

3. After that the case is made public.

I think that in this scenario a suspect programmer in general will be more cooperative also.

Answer by David Levy:

Date: Sat, 10 Sep 2011 03:20:21 -0700 (PDT)
From: David Levy
Subject: Rybka in retrospect
To: Ed Schroder
Cc: Jaap van den Herik, board@icga.org

You say yourself that chess programmers are the best people to judge such issues, and therefore it is inevitable that any such panel will include people who have something to gain, directly or indirectly, by a finding the defendant guilty. By conducting the panel discussions transparently, so that every panel member could see what every other panel member was saying, the ICGA made it possible for any unfair bias to be spotted and commented on. I cannot recall seeing any valid accusations of this type.

It is also important to bear in mind two further points:

[a] Although the panel comprised of more than 30 people, only three people made up the secretariat and were responsible for drafting the report.. Every member of the panel had an opportunity to see that report and to comment if they felt that it did not fairly reflect the evidence and the discussion that had taken place. So I think it is fair to say that the secretariat’s report did indeed fairly reflect the evidence and the discussion, otherwise there would have been objections, before the release of the report, by those panel members who felt otherwise.

[b] The secretariat’s report and supporting evidence was read by the members of the ICGA board (excluding the Vice-President for Asia who has almost no understanding of English), any of whom could have disagreed with any of its comments and findings. These board members were unanimous in finding Vasik guilty.

...

Your suggestions here are very well founded. The ICGA would like this question of future cases to be discussed by the leading programmers before a procedure is finalized, so perhaps you could initiate a discussion (preferably on the panel Wiki) with other top programmers, so we can make a decision when we meet in Tilburg in November. Personally I like your suggestion, but whatever the ICGA decides in respect of its procedure for the future needs to have the majority support of the top programmers as a group.


Criticism: it's just wrong to install a Panel (jury) with people that have mixed interests and give them a vote and this kind of pretence should have been avoided by all means.

It's appreciated the ICGA board takes the criticism at heart nevertheless without consequences to the Rybka ruling it's void following standard justice, a court would nullifie such a case.




The ICGA procedure against Rybka
Inconsistencies
PART 2

From the Charter:

2 Membership

The Panel shall be open to:

[c] Persons accepted as experts by the ICGA or by a unanimous decision of the
    Secretariat of the Panel.
And on [c] David Levy in the same email said:

I agree with you that chess programmers and those who are experts on the techniques employed in chess programming are the only ones who can make detailed judgments on whether or not one program copies from another.

Nevertheless 10 people the Panel were not chess programmers and yet were given the power to vote.




The ICGA procedure against Rybka
Inconsistencies
PART 3

During the accepting procedure of the Panel a famous chess programmer known for being an outspoken critic towards the Rybka accusations, someone who could have made a difference during the Panel talks was denied access to the Panel for dubious reasons.

The correspondence on this:

Date: Thu, 15 Sep 2011 04:01:44 -0700 (PDT)
From: David Levy
Subject: Re: The reason why I did not show up as Panel member
To: Ed Schroder
Cc: Jaap van den Herik, board@icga.org

David Levy - Given everything you say above, and how strongly you feel about all this, I feel that it is particularly unfortunate that you did not take part in the panel, so as to make your views known. You were originally going to take part but then you decided to withdraw.

Ed Schroder - I can imagine your curiosity. It's a long and ugly story starring Bob Hyatt in dubious role as gatekeeper. To make a long story short, it's about the refusal of a critical voice to enter the panel who could have made a difference in the outcome. If you are interested in the details of that scandalous event I am happy to share. For me it was enough to withdraw my participation.

David Levy - No, I'm not interested. It was your decision and you were quite entitled to make that decision. Sadly for Vasik your decision, and possibly those of the people to whom you referred below, cost him some support. I am not saying for a moment that the verdict would necessarily have been different with you and the others on the panel, but your voices, and his, would at least have given his position some semblence of a defence.

It's bizar to see the chairman of the ICGA in the role of judge and final executioner without caring about the composition of his jury.




Articles
  • October 5, 2011
    Why Rybka is an original engine by Ed Schröder

  • October 4, 2011
    Fruitication of Rybka

  • August 17, 2011
    The ICGA/Rybka fiasco
    Another perspective, by Dann Corbit. Web version

  • News
  • October 17, 2011
    A new paradigm to ensure fair play.

  • October 3, 2011
    A conversation with David Levy chairman of the ICGA.

  • August 22, 2011
    The CSVN speaks out. Rybka is welcome again.

  • Copyright © 1984 - Ed Schröder Mail Me


    Practicle Analysis