Rybka's disqualification and ban from computer chess last summer by the International Computer Games Association (ICGA) is being disputed. In a 31-page article, a computer scientist working at London's Queen Mary University, supported by several chess programmers, argues that "the ICGA's findings were misleading" and the decision to punish Rybka and its programmer Vasik Rajlich "lacked any sense of proportion".
A Gross Miscarriage of Justice in Computer Chess (part one)
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. Read More
A Gross Miscarriage of Justice in Computer Chess (part two)
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. Read More
A Gross Miscarriage of Justice in Computer Chess (part three)
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. Read More
A Gross Miscarriage of Justice in Computer Chess (part four)
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. Read More
Download the full article in PDF.
Søren Riis, London
My article “A gross miscarriage of justice in Computer Chess” [links below] was obviously a defense of Rajlich so in that sense David Levy is correct and my article can be said to show bias. However, while it’s acceptable that the defense and prosecution are biased, I find it wholly unacceptable if the people in charge of an investigation are.
And please note by Levy’s own admission – the people in charge of the ICGA investigation already prejudged that Rajlich was guilty before the investigation started.
Levy begins the interview with an astonishing admission and shift in argument: Rajlich did not copy (or translate) code verbatim from Fruit, instead he undertook “non-literal copying”. But what exactly is non-literal copying? Maybe Levy's notion of non-literal copying is supposed to be about using underlying structures from Fruit? But the most important underlying structure, the data structure, couldn't be more different. Levy's notion of "non-literal copying" is so vague and subject to mischievous interpretation as to have no real meaning.
A reasonable use of the term "non-literal copying" would be an activity which is covered and forbidden by copyright law. In that case if Rajlich had engaged in "non-literal copying" he would be in breach of copyright law and Rybka would be a “derivative work” of Fruit. This is clearly not the case, as was argued in my article and also subsequently shown by an independent computer programming expert, Andrew Dalke, who looked at the case.
Note also that the very fact that this input from Dalke exists at all bears witness to the fact that Levy is wrong when he says that the ICGA investigation had to be carried out by computer chess specialists and computer chess specialists alone. Per Levy's fallacious construct investigators need to be experts in “Pac-Man programming” in order to determine whether a “Pac-Man” program is original or breaching copyright laws.
Of course there is a good reason to have at least some "chess programmers" involved so they for example can assist with the "filtration" stages. As Andrew Dalke writes:
They [ICGA] claim to use the abstraction-filtration-comparison test to determine substantial similarity, but without the appropriate filtration. At each of the structural levels they fail to show that the discovery methods are not producing false positives, and they fail to demonstrate that the similarity level is greater than would be expected from a non-infringing chess program implementing the idea at the same structural level.
Chess programmers would be useful in pointing at which elements to use and which to filter in the ICGA (Watkins) comparison. However, the "chess programmers" were not involved in any filtering process, since none was done.
The correct way to state things is that Rybka implements ideas, concepts and algorithms that Rajlich learned from Fruit and other sources which gave Rajlich the tools to take chess programming to a new level of excellence. If Levy wants to call this "non-literal copying" it raises a host of thorny questions with answers that are inevitably highly subjective. What Rajlich took from Fruit complied with ICGA Rule 2 as stated, as well as satisfying copyright and license laws.
Rybka was heavily inspired by Fruit but is not a derivative work of Fruit. In addition, Fruit itself took many ideas from other sources and also stands on the shoulders of previous giants of the genre. Robert Hyatt’s program Crafty is greatly indebted to Letouzey, Heinz, Slate/Atkin, Keenan, Slate/Sherzer and Shannon to name but a few. This is how computer chess programs are developed.
When the ICGA received the complaint letter from chess programmers, a wise response that would have shown responsible leadership could have been:
"The ICGA’s view is that it is entirely acceptable to use ideas, algorithms and high level concepts from other programs. If it can be shown in a court of law that a program violates another's copyright or license agreement we will take appropriate action against the violator. In the absence of a court decision we would consider taking action if compelling evidence was submitted to the ICGA that a program contained copied (literal or translated) code. We have to maintain neutrality so in general we prefer to single out programs for investigation by random checks rather than selecting the ones suggested by competitors.
In the actual case of Rybka the alleged offense took place more than five years ago and in our judgment there are other programs from more recent ICGA tournaments we more urgently would need to examine. In the absence of a court decision or direct evidence of blatant code copying (or code translation) we will not initiate a time consuming and extensive investigation of Rybka."
Such a response would have been fair and would have saved computer chess (as well as the ICGA) from a lot of trouble. There is however nothing wrong with the ICGA conducting an investigation. However, in addition to technical expertise, they should also have applied non-technical judgement and thought about a fair, proportionate and reasonable way to proceed. Bob Hyatt and other aggressive pre-judgers should have been kept away from the investigation. A massive failure of the ICGA investigation was the structure of the process; it is almost inconceivable that the decision was left purely to technical experts who unfortunately were blind to the veracity of the evidence presented as well as to the implications of what they were doing.
My ChessBase article sparked off debates also among software specialists outside the small "club" of chess programmers. For anyone who followed these debates it seems virtually certain that a thorough investigation conducted by neutral software specialists would have found the great chess programmer and world champion innocent and cleared his name.
Søren Riis is a Computer Scientist at Queen Mary University of London. He has a PhD in Maths from the University of Oxford. He used to play competitive chess (Elo 2300). In his latest email Søren wrote: "Just for your information: I withdrew as Rybka forum moderator almost immediately after the ChessBase article, as there were far to much mud slinging, and moderation became almost impossible and time consuming. So I have no formal links with Rybka which allows me to speak without any hidden strings attached.