Rybka affair: Rajlich speaks

It seemed that Vasik Rajlich, the programmer of Rybka, had been silent ever since he had been disqualified and banned by the International Computer Games Association (ICGA), at the end of June. However, as it turns out he gave a half an hour interview on July 4th already, which has been posted on YouTube. "'I had a nice little spike of sales," said a down-to-earth Rajlich.

On June 29th we broke the news about Rybka: the chess program and its programmer Vasik Rajlich had been disqualified and banned from previous and future World Computer Chess Championships. The ICGA accused Rajlich of plagiarizing two other programs, Crafty and Fruit, and demanded that he returns the trophies and prize money of the World Computer Chess Championships in 2007, 2008, 2009 and 2010.

The news made the mainstream media: it was picked up from ChessVibes by the influential ExtremeTech and then copied by hundreds if not thousands of sites, including the New York Times and Der Spiegel. The chess news site of Chessbase, who distribute Rybka, didn't cover the story at all. In their shop, Rybka 4 still has the JPEG image with the tag line 'Computer Chess World Champion'.

Shortly before putting the story online, we had sent an email to Vasik Rajlich asking him to comment on the allegations. Unfortunately we never received an answer from him, and it seemed that the programmer remained silent to other media as well. However, during the last week several readers pointed out to us that Rajlich did agree to an interview, which was posted on YouTube. It was conducted by a Nelson Hernandez, who seems to be good friends with him.



Rajlich starts with a personal statement:

Before we start I'd like to make a short statement about these ICGA allegations. First, I did not wrongly omit anything from our entry form, nor did I break any other tournament rules. I did not list Fruit on the entry form because there is no direct game playing Fruit code in Rybka. Rybka uses a different board representation than Fruit and uses a different structure of search routines, so this should be quite clear. Now it's true that I did take a number of things from Fruit which are above the level of source code. This is something that I discussed many times, something which is perfectly normal practice, and it's not something which is asked about on the entry form.

Vasik Rajlich at the Computer Chess World Championship in Turin, 2006

Second, Rybka 1 was already disassembled and published back in 2006. Everything about Rybka 1 has been completely transparent for 4 1/2 years or so. So if somebody, if one of our competitors or the ICGA had a problem with something in Rybka 1 or with information that I provided on our entry form, then that was the time to raise it, you know, before the tournament. Once our tournament entry is approved, there's no grounds at all for reversing course and applying some kind of retroactive penalties. Third, if the ICGA wants to open this Pandora's Box of what exactly is an acceptable level of similarity, why is Rybka being singled out? Why is Rybka being the only engine which is investigated?

(...)

If the ICGA wants to go down this path to establish some kind of quantitative measures of what is and is not acceptable, they need to apply those measures to every single participant. Fourth, I completely disagree with this entire business that was conducted. There was all kinds of public statements and accusations, the ICGA made an accusation. By the time I was asked to address the accusers and discuss the topic with them, which was again not at all a proper protocol, by this point they had themselves made multiple public statements. So it was pretty obvious to me that this was not going to be some sort of a fair hearing and so I did not participate.

Then Hernandez asks Rajlich whether there is any code from Fruit in any of the Rybka versions. Rajlich:

Rybka has numerous code from other programs from the public domain inside. I use that throughout the code, that's accepted practice and that's legal and that's code that's in the public domain.

Certain public domain functions may well be in Fruit and Crafty as well, so they may be shared with me and Crafty, Fruit and Rybka, I'm not really sure. And it's even entirely possible that I would have taken them from Crafty and Fruit and put them into Rybka. For example the Nalimov tablebase code, that I remember quite specifically taking from Crafty. And that's normal.

In the interview Rajlich didn't want to comment on the ICGA's findings and in general he seems to have reservations in accepting the ICGA as a legitimate governing body. Rajlich also denied that he violated the rule involving program originality. We wouldn't be surprised if the Rybka programmer was preparing a lawsuit against the ICGA. To the question what he's going to do now, the answer was: "no comment". Asked about the effect of the negative publicity, Rajlich said: "I had a nice little spike of sales."