Feedback on the ICGA/Rybka disqualification scandal
In the following you will find virtually all the feedback we received to Dr Riis' four-part defence of Rybka programmer Vas Rajlich. We have only left out a few messages: some that were not immediately comprehensible, or gratuitously rude or accusatory. Branding people or institutions involved in the controversy liars, cheats or criminals without providing justification for these claims disqualifies a letter for publication. Also we have left out half a dozen letters that, like some that we are publishing, remind us that ChessBase sells the program Rybka. Indeed we do, and also sell (or try to sell) some of the programs that theoretically stand to profit from Rybka's downfall.
A few letters, like the first in this report, have been summarised, as their content was already spread over the Internet. However the full content has been made available for readers who are not daunted by the length of the rebuttal and the corroborative material supplied with it. We have also placed two of the longest letters, one for and one against, at the end, where readers who have persevered and still have the attention will find them. The rest are more or less in chronological order.
David Levy, ICGA President, London
In his four-part article on ChessBase.com about the Rybka scandal Dr Søren Riis has tried very hard to defend Vasik Ralich’s actions that led the ICGA to find him guilty of breaking ICGA Tournament Rule 2. As a historical review of progress in computer chess Riis’s article contains important and interesting information and comments. Unfortunately, however, his thesis lacks objectivity because it circles the core question and attempts to defend Rajlich by attacking the rule he was accused of breaking, attacking the investigative process in various ways and attacking some of those involved in that process.
When a defendant is brought before a court of Law, what is in question is whether or not (s)he broke the Law and not whether the Law itself is appropriate. And so it is with the ICGA rules. In considering the Rybka case the ICGA’s task was to decide the matter on the basis of its Tournament Rule 2, not to question the rule itself. This rule requires programmers to list all authors and the source of code, even “derived” code. Entering a competition such as this brings with it the obligation to abide by the competition rules.
Any readers of Riis’s articles who wish to understand the ICGA’s stance in this matter should visit the ICGA web site where they will find a link to the full report of the ICGA Investigation Panel together with its supporting evidence and the judgement made by the ICGA based on that report. A robust technical rebuttal to Riis’s article is now available. The ICGA would have welcomed an objective appraisal of its conduct of the investigation and its conclusions, but Søren Riis cannot be described as objective in this matter, not by any stretch of the imagination. The resulting article, full as it is of Riis’s bias, does nothing in my view to make the case for a miscarriage of justice to have taken place. It is, put simply, biased reporting.
The full ten-page rebuttal is available on the ICGA web site quoted above. David Levy has also agreed to do a full one-on-one exclusive interview with us on the entire ICGA/Rybka scandal. It is being currently conducted by sequential emails, but may take a while, since David is consulting experts who live in different time zones around the globe.
Cock de Gorter, Chairman of Dutch Computer Chess Association (CSVN)
I need not tell you that the ICGA made a terrible mess. On our site last August we declared we will not accept the Rybka ban. The computer chess world is split in two. At this time the CSVN board has the most serious doubts as to the rightfulness of ICGA's decision. Therefore, we have chosen not to abide by their sanctions against Rybka.
Ed Schröder (open letter to David Levy)
After seeing the reactions of my former colleagues on the ChessBase article it's now crystal clear to me nothing is going to change. As I see it the real power is in the hands of the programmers, when they go on strike there is no tournament, the CSVN felt the power of the programmers a couple of months ago, they destroyed two tournaments. And unfortunately I can not say you are blameless in the latter. You attacked the CSVN in a total unnecessary way in your notorious Attack of the Clones article on later echoed back in the programmers letter to the CSVN. Your motives for that are a mystery to me.
Looking back at the last six months and all the bad things that happened I do not want to be associated any longer with the ICGA in any way. Please be so kind to remove all my tournament results including my two world titles. Please inform me where to send the Shannon trophy, it's not longer welcome in my house.
Ed Schröder is the author of Rebel, which in 1991 won the World Microcomputer Chess Championship and in 1992 the World Computer Chess Championship, ahead of a field of mainframes, supercomputers, and custom chess hardware. He has written extensively in defence of his colleague Vasik Rajlich, although he was originally one of the programmers who asked for an investigation.
It should be noted that Søren is a moderator on the Rybka Forum and is not an objective bystander. It should also be noted that ChessBase is a Rybka distributor and not an impartial source of news in this case.
Shahar Tzafrir, Israel
Thank you for publishing this story. I felt from the get go that biased competitors that couldn't beat Vasik and Rybka at the board, changed venues to dirty politics instead. I played several matches (under the alias Highendman) vs. the Rybka Cluster on playchess, and always found Vas, Lukas and team to be great Gentlmen. I feel they were publicaly wronged and that in publishing this paper you're doing them right. Having said that - I think you should have added a small disclosure about chessbase being a Rybka reseller, although most anyone knows this anyway.
Matt Goddard, Amherst, NH
After reading all parts of your report, it seems that you make a good case. I went back to Zach Wegner and Mark Watkin's reports to review the specifics of the allegations, details that were rather compelling at the time, and I am uncertain. There are details that you do not address, but then you do raise general doubts. A more timely defense would have been helpful. It will perhaps be telling, if the principal accusers now respond to the defense you present, or rest on the judgement already attained. Suggesting that they are dunces is counter-productive to that test, since they might now harbor the same point about not bothering to respond to self-motivated attacks. If Rajlich is innocent, his name should be cleared. It will be interesting to see how this develops.
Paul Ruffle, Westgate on Sea, England
Is Chessbase still going to produce a version of Rybka 5 whatever happens? I hope so. In a chess program sense Rybka is probably the best thing that has happened to chess in the last 5-6 years. I find the controversy very sad and hope that a new version of Rybka will again be competing with the best soon.
I think the "gross miscarriage" here is that ChessBase is only printing one side of the story. Why don't we see the rebuttals from David Levy and Tom Watkins. I'm disappointed that you show extreme bias here. My opinion of ChessBase has greatly diminished.
I read through the entire, four-part story, and as a reader, it became clear to me that not only a tremendous amount of time and energy went into researching and writing it, but the author's heart was in the right place, just as his magnificent brain was: on the path of justice. "To stand up for the cause of justice, even if it be 10,000 foes!" And if that weren't enough, I learned a tremendous amount from the telling of the story, and now have a much better understanding and respect for the developers of these chess engines. I give this report five stars! Congrats to all who contributed, and let's hope that justice in this case is done, and done quickly.
Peter Jameson, Sydney
Dr Robert Hyatt whose credentials are absolutely first class asserted that some code of his had been blindly copied: the code in question was some redundant and useless code inadvertently left in his program yet somehow, mysteriously, this code ended up in one of the Rybka variations. This "smoking gun" is very, very hard to explain away with the mere wave of the hand.
Oswald Roggental, Vienna, Austria
Congratulations. I don't think there is any Rybka scandal, I think there is an ICGA scandal! You are on the right track.
I felt the process not fair. How could the ICGA use many competitors to decide the other competitor's fate? What giving me a deep impression was that the judges and plaintiffs were same people (not exactly). The chairman is the biggest accuser. Funny! So in my mind, Rybka is not a biggest scandal. ICGA looked like non-professional.
Kevin Cotreau, Merrimack, NH USA
I just read Dr. Søren Riis' four-part article regarding Rybka and International Master Vasik Rajlich on Chessbase.com, and frankly, I was appalled at how the ICGA conducted itself.
My first impression was that the ICGA allowed this to be conducted as a kangaroo court. The "investigators" seemed like cops, who "KNOW" a subject is guilty before any investigation has begun, so they only investigate things that prove their theory. A great example of this was the patently false evidence they provided, and attributed as being Rybka code, which they later admitted was not. The introduction of that as evidence is nothing less than disgusting.
My second impression was that the ICGA put a biased person in charge of the
investigation, so how can we take it seriously? Even the potential appearance
of bias should have been enough to exclude Dr. Hyatt, and was a major mistake
by the ICGA. Dr. Hyatt's behavior in public since has been abhorrent. A lead
investigator does not take to Internet forums to act in this manner. I also
suspect that he is jealous because Rybka far surpasses his own creation Crafty.
I have been using engines many years, and when it came along, it was immediately clear that there was nothing else like Rybka, so therefore, how could it have been plagiarized in any large amount? It is clear that Rajlich, admittedly, took ideas and rewrote code based on the ideas, making them better. Any other conclusion defies logic. Frankly, even if he were correct, and had access to the original source code, how does Dr. Hyatt get to decide what is considered "reusing code" and ideas, which is common practice, versus what is plagiarism? He gets to draw the line? Rajlich has clearly given credit to the author of Fruit rendering their findings bogus by the simple definition of plagiarism.
Personally, based on that four-part article, and some searches I did, Rajlich should file a lawsuit against the ICGA, Dr. Hyatt, and possibly others, for libel as they have really dragged him through the mud. I can say, with the evidence presented, they would not want me on their jury. This is clearly a case of the investigators' hubris gone wild.
For now, the ICGA has rendered itself an utter joke, where none of the top programmers dare compete in their, now meaningless, tournament. The suggestion in part four of the expose, "one line of thinking is that attribution plus added value should be sufficient to compete, instead of 100% originality as nothing is truly original, and you are just setting arbitrary boundaries", would seem to a great idea, which would promote the advancement of the ICGA's cause, and more importantly, the cause of chess computing, in general.
For the time being, I call on the ICGA to vacate their findings, do a new investigation with an impartial investigator, and go on from there. If they find in IM Rajlich's favor, a sincere and very public apology would go a long way toward mending any broken fences. They should also sever all ties with Dr. Hyatt.
Andres Hortillosa, President of Smart Learning Inc., Minneapolis
I call upon all sensible chessplayers everywhere to rally behind Vas Rajlich and call for the ICGA to retract immediately its ruling against him. Any more delay will only strengthen the gathering public outcry or invite the formation of another body. Furthermore, ICGA needs to update its qualifying "silly" rules to keep up with the times.
Attribution and added-value should be enough to qualify an entrant into the tournament. This singular change alone no doubt would see a leap in the overall development of chess engines leading to the accruing of benefits for chess fans everywhere to enjoy. The change that benefits all always trumps the change that benefits only the few. But the change the rights a wrong is best.
It will take years and a generation of players for the chess world to realize and measure how much chess has changed for the better because of one man's solitary effort and affliction. Chess lore will celebrate Rybka as the turning point. History will be affectionately kind to Vas and it will enshrine him as the pioneer of pioneers.
Mike Tendys, Ostrava, Czech republic
As we can see, all that case is about disposal of rival and slandering of his reputation. I believe that defamation is a crime. So who should stay before the court?! It looks like as great unjust case and attempt for destruction of opponent by others who signed that willingful defamation. That case is since its start about defamation of all members of Rybka team. Their reputation were highly corrupted by poor charge and their long-lasting work on Rybka is within second under the table as a fraud. I realized that Rybka played so ingeniously since its start as freeware beta program and was much stronger (and super innovative human-like playing style) than Fruit and all other, even more recent competitors. Without Rybka no computer chess world as we know it nowadays would exist. I think that Houdini, Stockfisch, Critter, Zappa, Toga, Komodo, etc. are rather poor copies of the innovative Rybka engine, because they learned a lot from Rybka during past years. Almost all programs have plagiarized each other. Rajlich is not plagiator in bigger way than any other programmer.
Edward R, Wellington, New Zealand
Your article correctly notes that in June 2011 it was widely reported that the ICGA had stripped Rajlich of his WCCC titles because due to violation of the ICGA rules. This event was not reported at all, however, at ChessBase. The decision to ignore this major story in computer chess was disgraceful. If there were any conflicts or associations with individuals preventing this, it should have been noted - just not ignored. Having missed the timely reporting of this incident, I think you would have been better off allowing others to cover this latest news.
Joachim Rang, Bochum, Germany
As the main tester and close associate to Fabien I'm probably not the most objective person to judge on that matter, but the evidence presented is imho just overwhelming. Vas certainly copied some code from Fruit and with copied I mean litterally copied not just learned from it. That being said Rybka is definitely more than a copy of Fruit since it's playing 400 Elo points stronger. Still Vas was not patient enough to improve his program step by step until it reaches the top, he wanted to build on something which was already very strong. He was allowed to do so (Fruit was GPL) but was not allowed to close the source and go commercial with the amount of Fruit in initial Rybka. I find the decision just.
Dari Dariouche, France
I think it is a good news for the whole chess engine community. I remember after pressing ICGA, the author of Rybka obtained the disqualification of Robbolito in a championship of a late year. And now, the truth exploses his engine. I want just to add that nobody escapes from justice, neither in this life, neither in the other.
Curt Hoyt, Mountain View, United States
This entire article made no reference whatsoever to the actual evidence that Rybka copied code from Fruit. I am a programmer myself, and I read quite a bit of the document. It is incredibly damning. Let's also not forget the fact that Soren Riis is a moderator on the Rybka forums, and clearly has a dog in this fight, so to speak. I am incredibly surprised that Chessbase would publish something like this. In my view, anybody who writes a refutation of the ICGA's findings must make actual reference to their evidence.
Michael Lampe, Bad Rappenau, DE
You cannot take 95% from others, put 5% cream on top of it, and then claim it all yours! Fruit was released under the GNU General Public license. This means that you can use that code base for your own programs, if you like. But only on the condition that you also offer your "+5% of cream" with everything else also under the GPL: You must offer complete source code and grant everyone else the same rights that your were granted. So: if Rybka only uses the least worthy 5% of Fruit's code it's illegal, because it violates the GPL, the very license under which Fruit was released to the public. And proof to this end semms to have been given by indeed painful reverse engeneering. It's the same problem we are seeing in the Open Source World (Linux mostly). To stand on the shoulders of giants is a great idea: Please do! And then play according to the rules!
Mark Schreiber, programmer, New York, NY
Thank you for the interesting article. I think Dr. Søren Riis" is 100% correct. Rybka competitors lost on the chessboard, so they try do sully Rybka good name. Shame on them.
Bill Campbell, Campbell, CA USA
I have zero respect for Dr. David Levy after reading your article "A Gross Miscarriage of Justice in Computer Chess."
Leonard Steele, Edmonton, Canada
Cry babies! Rybka/Rajlich got shafted.
Chris Kantack, Huntsville, AL U.S.A.
Excellent paper by Dr. Riis. I just wish Chessbase would be professional enough to also publish the rebuttals by David Levy and Mark Watkins.
Sergey Vasilyev, Moscow, Russia
Ben Johnson took doping, Vasik Rajlich took code, Samsung took design, Microsoft took patents, all were punished .... not really surprising, the list is endless.
Jimmy Liew, Puchong
I thought at first that Dr. Søren Riis is an impartial observer stepping in to defend Rajlich. Then I read in Part 4 that he is a Rybka forum adminstrator. Hardly "impartial", I would say and quite misleading of him not to admit this up front. This, for me at least, reduced dramatically the impact of his arguments.
Michael Lampe, Bad Rappenau, DE
This is all oafishness in a side show. Who cares if Rajlich typed '0.' instead of '0' or inadvertently missed something from copy & paste or whatever? People have committed worse crimes. (Fortran programmers are liable here. Did Rajlich ever program in Fortran?) "Finally, let me thank Vasik Rajlich for his clarification of various technical points and contemporaneous notes." Why not let him write his plea himself? Maybe he really did?? To ChessBase: If this is the long awaited statement, you owe the (computer) chess comunity since months -- then goodbye ChessBase!
Philipp Bouillon, Düren, Germany
Wow! After reading the first part of the article "A Gross Miscarriage of Justice...", I already thought to myself that this should clear Rybka and Rajlich of all accusations, having read the following articles only helped fortyfiing this opinion.
But: It's only my opinion. How did I reach this opinion? By believing everything that Riis has written! In order to really build a meaningful opinion, I'd have to read the original accusation, and go through Fruit Source with a fine comb (being a Software Engineer, I could probably understand it). Then again: to what avail? Rybka is a great chess program. I still use it to analyze. Leave it alone. I think, it is original. I believe in the innocene of Rajlich (because I want to believe). Isn't the most important thing that ultimately, there is a program as strong as Rybka that helps us (or rather: chess grandmasters) to understand chess better? Put this argument to rest. Reinstate Rajlich. He's earned it! (Btw.: I actually copied the name Rajlich in this post – I can't remember how to spell it, so I copied it and put it into different contexts. I hope nobody is going to sue me for copyright infringement...).
Mr. Gil, Phoenix, AZ
It is sensible that chessbase.com would run an article attempting to exonerate Vas Rajlich when it has a financial interest in the success of Rybka. The story includes details which make Vas look innocent and leaves out any evidence which might make him look guilty of claiming credit for code he didn't write. I don't expect to convince a financially interested party to be unbiased in its marketing efforts, but I did want to let you know that this attempt to convince the world of Vas Rajlich's honesty falls completely flat. At least you tried. Making money is important, and bending the truth in the service of protecting financial interests doesn't hurt anyone, unless of course they're interested in the truth.
Ed Zator, Toronto
Excellent follow-up article on the ICGA inquisition. Based on the analysis presented, I am convinced Rybka's programmer (who has has admitted his indebtedness to Fruit), is innocent of the charges, and was himself plagiarized by other programmers now sharing the top ELO ratings. The ICGA should get out of the game of pretending to be experts of reverse engineered chess code, and allow all entries willing to submit their source code to compete.
Why should I buy any commercial engine at the moment, and most probably in the future? The free ones are free, in some cases morally completly ok, and in the case of Stockfish. The difference in playing strenght is negligible. If you let them play a tournament, the result is more or less by pure chance. GUIs are what ChessBase can excel at. Lots of new ideas are thinkable here. Engines should be plugins anyway.
Jans F., Somerville, MA
This article makes a couple of interesting points. Vas Rajlich took code from Fruit, improved it, and that's good. The developers of Ippolit took code from Rybka, improved it, and that's bad. So, the point is made that when your friend does something, it's good, and when your enemy does the same thing, it's bad. In fact, let's write a long, complicated explanation about how it's not the same thing at all. The other interesting point is that the availability of free open source programs has advanced computer chess rapidly. Closed-source, paid programs have advanced computer chess more slowly. The point here is that profit motivation impedes progress. It also gets in the way of intellectual honesty by causing personal preference bias, which is obvious to the reader, but perhaps invisible to the author.
Michael Viking,Portland, USA
Hear hear! I hope your plans are to make the ICGA obsolete and irrelevant! One way to do that is to start sponsoring a well-run computer tournament to crown the new World Champion! They obviously hate competition.
A.J. Goldsby I, Pensacola, FL
It was obvious to many computer programs that the initial Rybka released owed much of its strength to the program Fruit. As a programmer, I have seen the papers, there is no doubt that some of the code was copied, line-for-line. The fact that Rajlich later improved his program or that today top programs are all highly similar do not alter the conclusion that the original edition of Rybka are based on the fact that the programmer copied a "free" (open source) program. I find the article to be nothing more than a game of semantics.
Andrei Olsen, Norway
Unless it was intended, there is just no way any programmer (not even a beginner) would ever write "if(movetime >= 0.)" or "if(movetime >= 0.0)", it just doesn't happen by accident. And it couldn't be a typo either, the "." (dot) is located at a completely different place on the keyboard (be it a US, Czech, Polish, Dvorak, etc. layout).
WA Oudheusden, The Hague
It is clear the ICGA made a judgement error and has now lost its credibility as organizers of a world computer chess championship. Who else will pick up this role? Could FIDE play a role?
Andrea Carta, Nemi
I'm an IT engineer. After carefully reading Riis' arguments as well as the original ICGA report, I'm convinced it is Riis, not ICGA, who's trying to subvert evidence against Rybka. I had some doubts before reading Riis' arguments, but now I'm sure Rajlich did copy Fruit's (and Crafty's) code, not only "ideas" and "algorithms".
Bobby Horne, Cologne, Texas, USA
I can only assume that this "Dr. Søren Riis" is a straw dog. Such a hotchpotch of complete nonsense. I would like to argue with him, but about what? That he doesn't have as many postings in some queer forum than Hyatt has? It's telling that a company who still sells this stuff says nothing for months and then comes up with such a lame joke. Credibility surely has to be gained otherwise. I really hope this will have repercussions and not stay in it's current state, where everyone can babble, on chessbase.com (specifically), in some rybka forum... It's all blah, blah, blah, ... to confuse the idiots and to prove the deliberate liar true.
John Hornibrook, Sydney, Australia
I very much enjoy the chessbase news site, but I feel I must comment about the horridly one-sided articles on the ICGA decision regarding Rybka. Yes, of course you are entitled to post a piece from someone on the Rybka side, but the articles do not address a majority of the evidence presented against Rybka and misrepresent some of the claims. This seems intentionally misleading for readers not in programming circles, and chessbase should weigh some responsibility against its financial interests. The section on Bob Hyatt in the final piece is, of course, unnecessary. However abrasive his manner, he is certainly generous with his time and assistance to the computer chess community.
Hanys, New Zealand
It is heartwarming to see a chess player, a mathematician and a computer scientist in one person stand in defence of one man. Vasik Rajlich already has a place in history of chess, Dr. Søren Riis deserves one too.
Albert Vasse (DGT, FIDE Arbiter), Enschede, the Netherlands
Dear Mr. Riis, after reading your article on the ICGA/Rybka controversy, I suggest you take the case to the FIDE Ethics Commission. The FIDE code of ethics is applicable to the ICGA, as an affiliated organisation to FIDE. The Ethics Commission can investigate and rule on any complaint that is brought to there attention.
Miguel Iniguez, New York. USA
Even when Dr. Søren Riis doesn't mentioned in his very professional investigation, I think that the chess world is expecting a compensation in favor of Vasik Rajlich, by those who abuse the power of ICGA. Great article Dr. Riis!
Arifur Rahman, Stockholm
I guess mostly people with the opposite view will complain about the content of the article. I just want to say something positive: first of all, the article itself is very well written and intelligently structured. If the whole thing was in one big article then it would be too heavy to read.
I have been involved with computer chess for many years. The whole Rybka affair has been quite ironic. The guy who has advanced computer chess the most last 4-5 years is now the only villain. Also the guys who have been the most jealous and vocal Rybka critics were in charge of carrying out this so called "ICGA research work". Note that these guys were Rybka critics long before the investigation. Also people who have been calling ICGA a joke organization and did not want to be associated with ICGA, has in this case embraced ICGA in a way that is very laughable. Well human nature I guess, nothing new there.
Rick Fadden, Barre, Vermont, USA
In the News article "A Gross Miscarriage of Justice in Computer Chess (part three), Soren Riis mentions my name and he gives a summary of my contribution. What he wrote is wrong, and this causes a problem.
Correction: The subject of my posts was "Strelka is a reverse engineered Rybka." I was demonstrating that Strelka source code exactly matched disassembled Rybka 1.0 beta. My posts could be interpreted as a defense of Vas Rajlick, because my focus was that someone else had copied from him, with Strelka matching what is in the Rybka binary. So many years later this author interprets my Stelka related comparison as making a false claim related to copying Fruit. Re-read my posts. If I mention the Fruit claims, it is just as a side comment. On that whole Fruit subject I was a spectator not an instigator.
Something else that should be reported though: in looking inside the Rybka 1.0 beta binary I found that Vas Rajlich was lying about how deep Rybka was searching, how many nodes it searched, and the Nodes per second reporting. In the code I point out that Vas includes a lot of code to "obfuscate" and to hide the fact that Rybka was a "fast, deep searcher", not a "slow searcher with max chess knowledge."
So I gave all kinds of supporting evidence that Vas obfuscated within his node reporting logic. I showed for example that when you enter a depth of search limit, Vas changes the input number to keep the limit consistent with his internal logic. (I could describe this in more detail.) So here's what I did to further prove this point: I created a binary patch of Rybka 1.0 beta to remove the math and instructions that hide the actual reported depth of search, and I posted this patch in the forum. With the patch applied to Rybka 1.0 beta the chess engine plays as before, and search proceeds exactly as before, *but* the program then properly displays it's actual search depth, nodes encountered, and nodes per second. (Frankly it's really neat seeing what Rybka is actually doing versus the false impression that Vas intends us to see.) Others used this patch to then watch Rybka play while reporting it's actual depth of search, and *they* noted that once you see the true search things look visibly very much like Fruit's reported search.
Keep in mind there was an initial legitimate question that I posed to Vas in the Rybka forum "Why did you obfuscate search reporting?" Vas reported in his own forum that he likes to think of nodes differently than other people, but everyone reading this weak excuse pretty well figured there was another reason why his code deliberately obfuscates it's search. People noted that Rybka search looked like Fruit search once the obfuscation was removed. Further investigation into possible coping of Fruit, pursued by other people, not me... may have been driven by this exposed blatant obfuscation.
My role was to send out information that I found during my effort to demonstrate that Strelka was a reverse engineered Rybka binary. The search depth obfuscation code that I found in that process bothered me, and I wanted to tell people what was in the code. In the Rybka forum I asked Vas to remove this obfuscation from future versions of Rybka, and Vas replied that he would likely keep this in Rybka. I see that even the latest versions of Rybka still include this search depth and nodes per second obfuscation.
Notice I didn't post anything after giving out my patch. I had written to Robert Hyatt (I've interacted with Bob since the 70's) explaining about this obfuscation built into Rybka, but I did not pursue the topic further. I never participated in an effort to prove that Rybka copied from Fruit.
Jeremy Stone, London
The kangaroo court procedures that the ICGA have indulged in are deplorable. No quasi-judicial body should allow itself the luxury of announcing the verdict in advance, before even drawing up the rules of procedure to be followed by the tribunal.
Further, the nature of the evidence is less than satisfactory. There are some putative smoking guns (the floating point timing code when Rybka uses integers; the material table that is much too large for the number of parameters that it needs to contain; the beta version that has error-checking code from Crafty that is checking for errors that don't arise in current end-game tables). But these are suggestive points for the prosecution, not conclusive.
The supposedly conclusive stuff all comes from reverse engineered code, which is claimed to be "semantically equivalent" to code in Fruit. This is shown in the evidence papers under the heading "Rybka", conveying the palpable (but false) suggestion that this is actual Rybka code. But reverse engineering is obviously a one-many relation, and semantic equivalence is not identity. This is a shabby attempt to generate direct evidence, where none exists.
However, the smoking guns do make it difficult to avoid the suspicion that Vas Rajlich's working methods relied on playing with existing code, in a way that is against the letter of Rule 2. Perhaps the evidence did, indeed, exist in 2005. One can understand all too easily why the ICGA establishment decided to make an example.
Rajlich's failure to offer a defence, or to produce the underlying Rybka code, does not help to dispel this impression. The evasive YouTube interview with Hernandez did not exactly help his case. A clear statement of the conceptual differences between Rybka and Fruit, and a positive outline of Rajlich's philosophy of software development, would have been effective in claiming the high ground, without him needing to get sucked into the grubby minutiae of arguments about reverse-engineered code. But no such clarifications were offered.
The question remains, whether Rule 2 is a good rule. Pretty obviously, it is a terrible rule, left over from the days when the only way to create algorithms was to write every line of code by hand. This is not how code has been written for at least the last 15 years (probably 20). Everybody knows that. Moreover, there is a manifest absurdity in accusing program A, which is 150 points stronger than Program B, of cheating, on the grounds that it has common elements with B. It is obvious that the common elements are not the source of the superiority of A. Particularly when this is sustained against all-comers for 5 years. The increase in strength was not marginal, even initially, arguing that there was a high degree of innovation relative to Fruit, even in Rybka 1. It is not an argument against this, to point to the high degree of innovation of Rybka 1 over the last beta version of Rybka; it would be a reasonable argument if this Rybka-Rybka advance had produced something that was much more similar to Fruit on the level of performance (and ponder hit correlation statistics, for that matter).
Essentially, Rule 2 is analogous to saying that someone who designs a better amplifier (or burglar alarm) can only be given credit for the improvement if he also invents, from scratch, all the individual transistors and capacitors that go in the circuit. At this stage, it seems to this outsider that the existence of things like PSTs in a chess program (whether pre-loaded or formula generated) is what you should expect to find, like airplanes generally have aerodynamically equivalent components called wings. What matters more than the existence of semantically equivalent sub-algorithms, and clearly mattered with Rybka from 2005 onwards, is how you put the bits together (the more so, if it is maintained that most of the bits are the same). Symmetrically, if Rybka's architecture is indeed the same as Fruit, some of its components must be significantly better.
Just for clarity, my outsider's conclusions are:
The ICGA disciplinary process falls well below the necessary standard. If this was a UK criminal trial, the verdict would be reversed on appeal for numerous reasons, not least clear evidence of bias;
The evidence produced by the prosecution is not the greatest. Some of it is methodologically very unappealling (reverse engineering does not convince, especially when presented as if it were actual code);
Unfortunately, Vas Rajlich does not convince either, in his failure to present any kind of defence, and in the face of a little too much circumstantial evidence that he probably infringed the letter of Rule 2. There remains a feeling that he was a victim of abuse of process, but not necessarily a miscarriage of justice, in relation to Rule 2 per se;
But the spirit of Rule 2 (as originally proposed) was probably the desire to stop "passing off" of clones, when entered in competition with original code of approximately equivalent power. It does not make sense in relation to an engine that utilises previous advances to considerably greater effect than the sources of those advances;
- Rule 2 itself is thus so out of touch with technological reality that it cannot be regarded as a basis for the action taken. So even if the evidence were better, and the judicial process clean as a whistle, it would not have been a good case to bring, for the good of the community.
Previous articles on the ICGA/Rybka scandal
|A Gross Miscarriage of Justice in Computer Chess (part
02.01.2012 – "Biggest Sporting Scandal since Ben Johnson" and "Czech Mate, Mr. Cheat" – these were headlines in newspapers around the world six months ago. The International Computer Games Association had disqualified star programmer Vasik Rajlich for plagiarism, retroactively stripped him of all titles, and banned him for life. Søren Riis, a computer scientist from London, has investigated the scandal.
|A Gross Miscarriage of Justice in Computer Chess (part
03.01.2012 – In this part Dr Søren Riis of Queen Mary University in London shows how most programs (legally) profited from Fruit, and subsequently much more so from the (illegally) reverse engineered Rybka. Yet it is Vasik Rajlich who was investigated, found guilty of plagiarism, banned for life, stripped of his titles, and vilified in the international press – for a five-year-old alleged tournament rule violation. Ironic.
|A Gross Miscarriage of Justice in Computer Chess (part
04.01.2012 – A core accusation against Vas Rajlich is that Rybka and Fruit have very similar positional evaluations, and the use of floating point numbers in Rybka’s time management code had to be copied from Fruit. Søren Riis enumerates the ten substantive evaluation differences and shows how the second accusation boils down to a single misplaced keystroke with zero impact on Rybka's play.
|A Gross Miscarriage of Justice in Computer Chess (part
05.01.2012 – As proof that Vas Rajlich had copied program code the ICGA presented pages and pages of Fruit and Rybka code side by side. But, according to Dr Søren Riis, what was labeled as Rybka code was actually fabricated to look like Fruit. He criticizes the ICGA's process failures and ruminates on the reasons behind the unprecedented vendetta launched against the star chess programmer.