FIDE Rules on ICGA - Rybka controversy

[BB+]

Seriously, Levy was offered a rebuttal by the Chessbase folks which he took and basically you did most of the work

"Seriously", are you are calling the Levy "interview" by ChessBase(/Riis/Schröder/Whittington) a "rebuttal" opportunity that ChessBase "offered"? The ICGA's rebuttals of Riis were already published a month earlier (ChessBase was kind enough to allow Levy the first reply in their "reader feedback" which linked to said rebuttals, though they cheekily blamed Levy for the "sequential" nature of his up-coming "interview", saying: 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 -- I am pretty sure that Levy would have been happy to have it all be done at once, rather than drip-dropping the questions, and the latter was ChessBase's choice to try to wring him as tightly as they could). After the first 2 parts of Riis, MarkL had noticed that I was posting quite extensively in forums about it, and asked me to turn it into an official ICGA response (on technical matters). As I was already doing the work on my own interest, this was fairly reasonable to do.

If ChessBase had any sense of evenhandedness in the matter, they would (as with Riis) allowed Levy/ICGA to post a multi-part series in their own words, as opposed to an "interview" with rather leading (and often irrelevant) questions. Or maybe they should have allowed Levy/Hyatt/Lefler to ghost-conduct such an "interview" with Rajlich (rather than publish Riis)?

I realy didn't understand why you dragged me (and notable Dalke) into that discussion but it was quite revealing how sensitive you often are. it was quite revealing how sensitive you often are.

Not my recollection -- it was you who put Dalke into the debate (even though he made it clear, both to you and in his first statement to me that he couldn't give a hoot about the ICGA decision, but only possible GPL infringement). And as is typical, you seem to care more about your psychoanalytic prognoses (regarding the participants) than actual evidence.

the question is how to set up the review.

Back in my Riis rebuttal I had suggested (4.4) that:

In the hopes of reaching a resolution, a goal that endless Rybka Forum posts will certainly not approach, let alone reach, I might propose that Rybka and Rajlich agree to a binding (this is a key aspect) review by an uninterested party. I would suggest Ken Thompson, but he has already weighed in. Similarly, Jonathan Schaeffer should suffice, but perhaps they would fret that he has old-school ICGA connections. It should be possible to procure a suitable panel from eminent programming doyens, though this is likely no longer gratis, and the loser should be yoked with the associated costs – thus I doubt any such resolution will ever occur.

I could still suggest a few names of people I know (inevitably, though, the fact that I know them would be seen as "bias", if nothing else in the aftermath). However, given that the ICGA itself is considered biased by the official Rajlich representatives the only venue I can imagine is a case before a "national judge" (as in the EC document), which would obviously be less concerned with technical questions and more concerned with the ICGA's authority to apply Rule #2 in the manner they did, whether procedural aspects were regarded, etc. It would not be a matter of whether Rajlich's broke Rule #2, but rather whether the ICGA's conclusion on this met certain legal standards and guarantees. There is a relevant quote in (part of) the EC conclusion: By carrying out proceedings against Mr Rajlich for an alleged violation of ICGA tournament rules and by deciding to disqualify Mr Rajlich from the World Computer Chess Championships 2006/2010, the ICGA has not violated the FIDE Code of Ethics, nor any other FIDE rule or general principle of law.

[Incidentally, why is it difficult to figure out where the ICGA is "based" or "headquartered"? I could determine this in a couple minutes of Internet searching, corroborated by 3 separate data points.. Or was the reason simply to harangue the ICGA over its "tax status"? ].

[BB+]

I would say it's crucial to judge the 60% similarity of Fruit with the Rybka values vs Rybka. Especially in the light of Zach's statement - Simply put, Rybka's evaluation is virtually identical to Fruit's 60% looks bad. Or it must be the influence of Rybka's MIT, I can hardly imagine it would make a big difference, but did you cancel the MIT in your test?[

As I indicated previously (and in 4.4 of the paper), the exact percentages (say 60%) are intrinsically meaningless. Perhaps one could say (see Section 5): Simply put, when judged against a collection of similarly strong independently created engines, the Fruit/Rybka move similarity was approximately a 5σ event (which is a bit sloppy in stats, due to the nonnormality of the distribution, but). The main question seems to be (see the discussion of Ballicora's phylogenetics in 4.3) what exactly the background set of "independently created engines" should be taken to be. I don't think the material imbalance table should matter much (the percentage of Dailey's positions where it is significant should be small). The paper says we modified Fruit 2.1 to use the Rybka 1.0 Beta numerology in evaluation, and my interpretation [and recollection] of this is that the material imbalance table was not included, but rather every co-incident eval feature of Rybka that also appeared in Fruit would have its value transferred over. Again I will point out that this might make the eval functions less "virtually identical" than Zach might indicate (but still much more than an extraction of a couple of ideas as is common, and normal).

Sure, but like I stated the effect of search is surprisingly low. I quickly went once more to your document and couldn't find an experiment that would give an indication of that. ... Anyway, I think more research in this area is needed to come to some sort of consensus regarding the effets of search.

I don't find the effects of search "quite low" ... Which means ponder hit is certainly not so useful. Doing fixed-depth position comparisons gets a bit closer to the truth, but the search STILL has a powerful influence on moves and scores when LMR and such are factored in. Tiny changes balloon into major differences.

A discussion of search/eval emphasis with fixed-depth versus fixed-time searches appears in 6.3.1 (with two different opinions on the matter). The general philosophy of move-matching is discussed in 3.2 (and mentions Thinker, with its allegedly copied eval but presumably different search). The fact that the strength-based time-scaling (6.2.4) reduced the standard deviation of move similarity is another thing that could be mentioned here (whether the time-scaling increased the move-matching is harder to say, as longer searches already tend to produce larger move similarity). I agree that more experiments could be done.

[Chris Whittington]

Seriously, Levy was offered a rebuttal by the Chessbase folks which he took and basically you did most of the work

"Seriously", are you are calling the Levy "interview" by ChessBase(/Riis/Schröder/Whittington) a "rebuttal" opportunity that ChessBase "offered"? The ICGA's rebuttals of Riis were already published a month earlier (ChessBase was kind enough to allow Levy the first reply in their "reader feedback" which linked to said rebuttals, though they cheekily blamed Levy for the "sequential" nature of his up-coming "interview", saying: 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 -- I am pretty sure that Levy would have been happy to have it all be done at once, rather than drip-dropping the questions, and the latter was ChessBase's choice to try to wring him as tightly as they could). After the first 2 parts of Riis, MarkL had noticed that I was posting quite extensively in forums about it, and asked me to turn it into an official ICGA response (on technical matters). As I was already doing the work on my own interest, this was fairly reasonable to do.

If ChessBase had any sense of evenhandedness in the matter, they would (as with Riis) allowed Levy/ICGA to post a multi-part series in their own words, as opposed to an "interview" with rather leading (and often irrelevant) questions. Or maybe they should have allowed Levy/Hyatt/Lefler to ghost-conduct such an "interview" with Rajlich (rather than publish Riis)?

I realy didn't understand why you dragged me (and notable Dalke) into that discussion but it was quite revealing how sensitive you often are. it was quite revealing how sensitive you often are.

Not my recollection -- it was you who put Dalke into the debate (even though he made it clear, both to you and in his first statement to me that he couldn't give a hoot about the ICGA decision, but only possible GPL infringement). And as is typical, you seem to care more about your psychoanalytic prognoses (regarding the participants) than actual evidence.

the question is how to set up the review.

Back in my Riis rebuttal I had suggested (4.4) that:

In the hopes of reaching a resolution, a goal that endless Rybka Forum posts will certainly not approach, let alone reach, I might propose that Rybka and Rajlich agree to a binding (this is a key aspect) review by an uninterested party. I would suggest Ken Thompson, but he has already weighed in. Similarly, Jonathan Schaeffer should suffice, but perhaps they would fret that he has old-school ICGA connections. It should be possible to procure a suitable panel from eminent programming doyens, though this is likely no longer gratis, and the loser should be yoked with the associated costs – thus I doubt any such resolution will ever occur.

I could still suggest a few names of people I know (inevitably, though, the fact that I know them would be seen as "bias", if nothing else in the aftermath). However, given that the ICGA itself is considered biased by the official Rajlich representatives the only venue I can imagine is a case before a "national judge" (as in the EC document), which would obviously be less concerned with technical questions and more concerned with the ICGA's authority to apply Rule #2 in the manner they did, whether procedural aspects were regarded, etc. It would not be a matter of whether Rajlich's broke Rule #2, but rather whether the ICGA's conclusion on this met certain legal standards and guarantees. There is a relevant quote in (part of) the EC conclusion: By carrying out proceedings against Mr Rajlich for an alleged violation of ICGA tournament rules and by deciding to disqualify Mr Rajlich from the World Computer Chess Championships 2006/2010, the ICGA has not violated the FIDE Code of Ethics, nor any other FIDE rule or general principle of law.

[Incidentally, why is it difficult to figure out where the ICGA is "based" or "headquartered"? I could determine this in a couple minutes of Internet searching, corroborated by 3 separate data points.. Or was the reason simply to harangue the ICGA over its "tax status"? ].

on your review concept ....

well, yes, I agree, in principal. The decision process woudl need to be fully transparent and open. One lone judge is not an good idea, imo, there's no way to tell the WHY, and therefore discussion to reach a conclusion is needed and that requires a transparent judging GROUP. There is an unbalance on verdict acceptance. ICGA woudl have to agree to abide by decisions, but it is too much to ask an innocent person to "accept guilt" when he is not, verdict or no verdict. However, it is conceded that a guity verdict woudl take the wind out of the sails of a future will to keep arguing.

I'll think about this a bit more.

On ICGA headquarters. levy stated here is no obligation ti register for legal or tax purposes. i take that to mean there is no registration in any domain. In which case icga doesn't actually exist, legally it is simply the individuals that make up its "board", I guess is anyone wanted to sue it, they would sue the individuals direct, in the jurisdiction they reside or have assets. Personally, I see it as a business. For example, the Olympiad has been transferred into the iCGA. The Olympiad is a business, in the sense that it traded as a limited company. at one point it went bust and into receivership or whatever. Tne assets were available from the liquidator for sale. assets included things like use of name, stretching through to T-shirt rights (I kid you not). I bid 500 GBP for these assets from the liquidator, but was outbid by someone else subsequently the assets, presumably, were transferred to the icga, when the olympiad was somehow transferred to the icga. This means to me that the icga owns business assets, has the olympiad under its wing, and is therefore a business itself. Yet it is also apparently a private club not required to register for tax or legal purposes. Confused? I am. maybe there is some explanation.

[Harvey Williamson]

On ICGA headquarters. levy stated here is no obligation ti register for legal or tax purposes. i take that to mean there is no registration in any domain. In which case icga doesn't actually exist, legally it is simply the individuals that make up its "board", I guess is anyone wanted to sue it, they would sue the individuals direct, in the jurisdiction they reside or have assets. Personally, I see it as a business. For example, the Olympiad has been transferred into the iCGA. The Olympiad is a business, in the sense that it traded as a limited company. at one point it went bust and into receivership or whatever. Tne assets were available from the liquidator for sale. assets included things like use of name, stretching through to T-shirt rights (I kid you not). I bid 500 GBP for these assets from the liquidator, but was outbid by someone else subsequently the assets, presumably, were transferred to the icga, when the olympiad was somehow transferred to the icga. This means to me that the icga owns business assets, has the olympiad under its wing, and is therefore a business itself. Yet it is also apparently a private club not required to register for tax or legal purposes. Confused? I am. maybe there is some explanation.

This is from David's notes written on the day of the hearing he attended

They asked where and under what legal authority the ICGA is registered. I explained that it isn't registered and the chairman said it is not necessary for an organisation such as ours to be registered anywhere - some are and some are not.

[BB+]

For example, the Olympiad has been transferred into the iCGA. The Olympiad is a business, in the sense that it traded as a limited company. at one point it went bust and into receivership or whatever ... when the olympiad was somehow transferred to the icga.

Note that there was "Mind Sports Olympiad Ltd" and also the spinoff "Mind Sports Organization Worldwide Ltd". And some others. I'm not sure which one merged into the ICGA (if any), as they have the "Computer Olympiad", not the "Mind Sports Olympiad"?

03829591 		D 	MIND SPORTS LTD. 	Dissolved 06/02/2007
03307916 		D 	MIND SPORTS OLYMPIAD LIMITED 	Dissolved 21/01/2003
02400386 		D 	MIND SPORTS OLYMPIAD LIMITED 	Dissolved 02/08/1994
03785356 		D 	MIND SPORTS ORGANISATION WORLDWIDE LIMITED 	Dissolved 03/09/2003
07515072 		D 	MIND SPORTS PARTNERS LIMITED 	Dissolved 06/12/2011
FC030291 		D 	MIND SPORTS PARTNERS PTE. LTD. 	Dissolved 04/07/2014
BR015286 		D 	MIND SPORTS PARTNERS PTE. LTD. 	Dissolved 04/07/2014
02713411 		D 	MIND SPORTS PROMOTIONS LIMITED 	Dissolved 23/07/1996
03298145 		D 	MIND SPORTS PROMOTIONS LIMITED 	Dissolved 27/01/2015
08099257 			MIND SPORTS (INTERNATIONAL) LIMITED 	
02969808 		F 	MIND SPORTS OLYMPIAD LIMITED 	Change of Name 27/01/1997
 04712990 			M S O LIMITED 	[msoworld.com]

If you care about the Keene/Levy dirt, some of it is recycled here.

T-shirt rights (I kid you not)

He and Iweta are still awash in ”I ♥ Rybka” pens, wall hangings, and other marketing tchotchkes.

[Rebel]

As found on the ICGA website one year after the investigation

OK, as I think you can tell, I am trying my best to verify the details of this. What do you mean by "the ICGA website"?

Mystery solved.

I made a mistake.

The correspondence did not come from the ICGA-WIKI as previously stated, but from Vas because I asked him for it when I noticed the (marginal) ICGA-WIKI correspondence and I wanted to have it all.

My apologies for the confusion

So likely both versions are not complete.

Looking for answers I came accross this post of mine and I smelled the rat. Checking my email and there it was.

[BB+]

I bid 500 GBP for these assets from the liquidator, but was outbid by someone else

This is covered in the Bjorn Bjornsson letter to Private Eye (among other "ludicrous rantings").

This didn’t stop Mr. Levy restarting his company under a new name “Mind Sports Olympiad WorldWide Ltd.”, in which he bought back the original MSO assets (actually there no assets, only the software) from the receiver for next to nothing. All the money we’d paid for the software development went to other companies owned by or controlled by Levy. And guess who was Receiver? Why none other than Levy’s personal accountant Gendall Phillip.

[Rebel]

But personally I find that Schröder is simply wrong about so many factual matters to be quite annoying also (cf. "right to an appeal", Rajlich retracting that Strelka contains Rybka code,

What

I have said no such thing.

It's notable in this thread Hristo correcting Vas he couldn't claim ownership on Strelka. Obviously Strelka originates from Rybka 1, I gave you the 2 similiar main(); and the identical values of (for instance) the rook evaluation, from my notes:

Code: Select all

// Start ROOK eval   [ RYBKA 1.0 ]
// ===============
		
if (!(v84 & HIDWORD(qword_667BA0[0]) | v83 & qword_667BA0[0]) ) // open / semi open rook files
  {                       // when semi-open
     v271 += 256;         // endgame+256	
     v6 += 64;            // opening+64 }
 
if ( !(v84 & HIDWORD(qword_667BA8[0]) | v83 & qword_667BA8[0]) )
  {                       // when open file
    v6 += 971;            // opening+971
    v271 += 172;          // endgame+172  }

Exacly as Strelka 1.8, right?

I gave you the 256 | 64 | 971 | 172 numbers linking to this code example, scroll up.

PLEASE.

[BB+]

Secondly there are quite errors in the Levy article on Chessvibes. I only want to highlight one misconception that still is stubbornly alive while false. It's Vas statement about Strelka (Date 2008-01-11 12:26) and people not realizing Vas a few days (and pages) later retracted that in the same thread (for those are willing to look it up).

I still don't see where Rajlich clearly "retracted" anything in that thread, he only stated he would think over the question of claiming ownership (and in a later thread did clarify that he was not claiming ownership).

If someone were to say "Levy quoted Rajlich's statement about Strelka, but did not mention Vas retracted it (in the same thread!)", the contextual understanding of such a sentence would be that Rajlich's retracted his main claim about Strelka, namely that: The picture is quite clear. Vast sections of these sources started their life as a decompiled Rybka 1.0. Particularly as Levy heads the long quotation from Vas with: Vasik pitched in with his own comments, claiming that Strelka was a clone of Rybka.

I guess you wanted "retracted that [statement]" to indicate a rethinking of "I am claiming Strelka 2.0 as my own", but I don't think this was very clear, and the logical reading of your phrasing would be that "Raljich retracted his claim that Strelka is a clone of Rybka", or "Rajlich retracted his statement that Strelka source code started life as a decompiled Rybka 1.0".

If I were to say "Letouzey retracted his statement about Rybka", what would you take that to mean?

[Chris Whittington]

On ICGA headquarters. levy stated here is no obligation ti register for legal or tax purposes. i take that to mean there is no registration in any domain. In which case icga doesn't actually exist, legally it is simply the individuals that make up its "board", I guess is anyone wanted to sue it, they would sue the individuals direct, in the jurisdiction they reside or have assets. Personally, I see it as a business. For example, the Olympiad has been transferred into the iCGA. The Olympiad is a business, in the sense that it traded as a limited company. at one point it went bust and into receivership or whatever. Tne assets were available from the liquidator for sale. assets included things like use of name, stretching through to T-shirt rights (I kid you not). I bid 500 GBP for these assets from the liquidator, but was outbid by someone else subsequently the assets, presumably, were transferred to the icga, when the olympiad was somehow transferred to the icga. This means to me that the icga owns business assets, has the olympiad under its wing, and is therefore a business itself. Yet it is also apparently a private club not required to register for tax or legal purposes. Confused? I am. maybe there is some explanation.

This is from David's notes written on the day of the hearing he attended

They asked where and under what legal authority the ICGA is registered. I explained that it isn't registered and the chairman said it is not necessary for an organisation such as ours to be registered anywhere - some are and some are not.

What hearing is this? Or are you referring to the Chessbase interview?