World Computer Chess Championship ?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
bob
Posts: 20564
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: World Computer Chess Championship ?

Post by bob » Fri Jun 15, 2012 2:01 am

Dan Honeycutt wrote:
Don wrote:I would LOVE it if there were some simple non-intrusive test (like peeing in a cup) that would have caught Vas early on. I would gladly submit to such a test.

Since the ICGA bent over backwards to avoid the "perception" of unfairness and that did not satisfy people, what sort of investigation would you propose that would work?
I'm no more than lukewarm on Graham's idea but if I were charged with implementing it a couple competitors chosen at random would have to submit their sources to a judge or two. Judges would not be computer chess programmers but rather programming experts who would familiarize themselves with the top open-source programs. If you were chosen they would look at Komodo's sources to see if they could find any pieces of Ippolit or Stockfish or whatever. They could require a compilation to verify that the sources produced the exe you were playing but they would not be allowed to make any copies of your work. I'd expect the check to take 30 minutes to an hour, not as quick and easy as pissing into a cup but also not terribly painful.
Don wrote:When a program is tested to see if it is a clone of someone else's program, which program does it have to be checked against? Suppose the next program is a clone of Komodo but they check it against Houdini? How will I get satisfaction?
You would still be able to trigger an investigation if you had credible evidence that Komodo was cloned.

Best
Dan H.
I think the problem is that non open-source programmers are not very willing to provide their source to someone they think might look inside on the sly. It is a valid concern, and a stumbling block that doesn't have an easy solution.

bob
Posts: 20564
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: World Computer Chess Championship ?

Post by bob » Fri Jun 15, 2012 2:08 am

mcostalba wrote:
bob wrote: WCC does NOT necessarily mean the best player in the world, never has... never will...
Does WCC mean a match among a sensible representatives of best players of the year ?
Not necessarily. As I mentioned already, 1975 wcc without Fischer is just one of many examples. Or when Kasparov split with FIDE.


Should something called "World Computer Chess Championship" mean a tournament among a sensible representatives of chess engines available that year ?

Suppose for the next "World Computer Chess Championship" do apply the 11 weakest (original) engines in the CEGT list (this is not the case, but just as an example). The winner is the "Word Champion" ?

Can you see the elephant in the room ?
What happens if the top-3 FIDE players CHOOSE to not compete? As Fischer chose to not compete in 1975? One can't MAKE competitors play in an event. They can be invited, no more.

bob
Posts: 20564
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: World Computer Chess Championship ?

Post by bob » Fri Jun 15, 2012 2:13 am

Laskos wrote:
Don wrote:
Laskos wrote:
Don wrote:
Laskos wrote:
Laskos wrote:
bob wrote:
WCC does NOT necessarily mean the best player in the world, never has... never will...
Could they name it differently? Otherwise, what an unaware reader can deduce from this ChessBase ads?

Code: Select all

Written by Amir Ban and Shay Bushinsky, Junior has won multiple World Computer Chess Championships, and now, in version 12, its playing strength has been further increased – by around 200 Elo points
The strength description is associated textually with WCCC (spelled and not abbreviated).

Kai
Even better on Hiarcs page:

Code: Select all

Deep Junior 13 UCI chess engine

The World Computer Chess Champion chess engine for PC MS Windows
Suitable for all players including the World Chess Champion
Platform: PC MS Windows™ 98, ME, 2000, XP, Vista and Windows 7 (all versions including 32&64 bit)
Playing ability: World Chess Championship strength (3200+ Elo)†   
Engine authors: Amir Ban and Shay Bushinsky, Opening book author: GM Alon Greenfeld
Here the association is direct, "Playing ability: World Chess Championship strength (3200+ Elo)†". Really, rename that crap called WCCC.

Kai
Keep in mind that these are not the ICGA's words, they are ChessBase advertisements.

It makes me wonder why people think Chessbase is the better organization to represent fairness and justice. I have no issue against Chessbase but they are a company for profit - nothng wrong with that, but ICGA is an association which tries to look out for the interests of computer chess - and I think succeeds.
No, I wasn't saying that ChessBase represents better "fairness and justice", nor in the Rybka affair, where they showed themselves as just biased marketers. I have not much against ICGA generally, even WCCC could be fun because of the needed book preparation and hardware differences (besides the sometimes entertaining hazard), but just rename it or specify that engines are chosen by ICGA and in no way represent all the best engines.

Kai
The ICGA isn't perfect but as has already been commented on, a world championship does not have to represent the best. The guy on steroids may be the very best weightlifter in the world, but he should not be praised and honored for artificially enhancing himself. What would be next, a bionic arm?

And what if the worlds best player just doesn't want to come? When Karpov first become world champion Fischer claimed HE was world champion because Karpov didn't have to play him. Is that your mentality here? Karpov WAS the world champion because he showed up and also because Fischer didn't prove he was better.
I don't like these sports analogies, not even Karpov-Fischer analogy, the reason WCCC must change its name and specify what competition it is is the marketing one, as I clearly showed. The main venues of selling chess engines are using this W(Crap)CC to sell their engines, associating the result in WCCC with absolute strength, which is plain cheating, encouraged by the spelling given to WCCC by ICGA.



There is one issue that I have trouble with, but I don't think I can blame it on the ICGA - and that is the huge expense and time involved in competing at these events. That is what keeps participation so low. Now I would be considered a professional which makes it more expensive and also for the same reason I can afford it even less! I don't have a general solution for this and money does not grow on tree's for the ICGA or us, they have to go out and find sponsors and in today's financial climate it's probably even more difficult than normal. But it does highlight the issues. There is very little prestige in computer chess (and only minimal prestige in chess in general) so the top people have to pay their own way into tournaments and the ICGA has to beg for money just to have a prestigious tournament.

This is why I am offended by a super critical post that makes it sound like the ICGA is incompetent or out of touch. You don't think they understand these same issues? To me the solution is certainly not to rely on the rating lists as has been proposed as the new "enlightened" modern way. That is like accepting defeat, we can all do this in our underwear sitting in front of our computer monitors. The wave of the future.
I do not follow your reasoning. Just because Rybka got caught as a derivative of Fruit and was disqualified, the WCCC is no longer the WCCC? The ICGA STARTED the WCCC. It is their event.

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: World Computer Chess Championship ?

Post by Don » Fri Jun 15, 2012 2:31 am

bob wrote:
Graham Banks wrote:
Don wrote:
Graham Banks wrote:
Don wrote:
Graham Banks wrote:
Don wrote:......There is a procedure to determine who the world champion is and there is. It should not be open the cheaters and copiers and it isn't......
How would you know that unless every engine was put under the same scrutiny as the chosen few?
Every engine? The way this works is that there has to be an accusation by one of the authors that someone is plagiarizing their work - in this case it was Fabien who made the accusation. The ICGA did NOT make the accusation.

It's ridiculously impractical for the ICGA to just launch a thorough investigation of every program in every tournament every time, taking a kind of paranoid (McCarthyism) stance that everyone must be guilty.

I don't known how things work in your part of the world but it would be pretty horrible to be around a culture like the one you suggest, that everyone should be investigated just in case they might be doing something wrong.
Isn't random drug testing like that? Perhaps they should randomly select two participating engines to scrutinise each year, with a rider that it can't be the same engine twice in a three year period.
They could do something like that, but I don't believe it would come out the way you think it would. By your implication you may have bought in to the argument that everyone is guilty but only a couple of people got caught. The computer chess community is pretty sharp and it's almost impossible to get away with this for very long. To underscore this principle the players in the online chess club can tell pretty quickly if you are using a computer to cheat and you will get flagged.

But this is a lot like life, we generally wait for an accusation before launching investigations and it's usually up the victim to take some interest in the process even though that is usually not a hard and fast requirement. In a scenario like you describe the victim could be every competitor in the tournament but I believe there should at least be a viable complaint of some kind before launching investigations, otherwise the ICGA becomes the oppressive tyrannical organization that a few extremists are accusing them of.
I don't think that all are guilty by any means Don. I'm interested in the perception of fairness, that the perception of some programmers being above scrutiny is got rid of.
By doing random drug tests, do we automatically assume that all competitors are guilty?
How about pointing out WHICH "programmers are above scrutiny?"

ME?

I was accused of cheating in 1986 and was thoroughly investigated, as most everyone knows. We were completely exonerated, after a thorough investigation that included several people looking at the CB source, running CB on the same machine we used in the 1986 WCCC event, using an executable Cray Research restored for the investigation from a regular system backup made during the event. So WHO exactly is "above scrutiny"???

Certainly does not apply to me.
I'm really shaking my head on this one with a grin on my face. There really are some people here who think of us (or at least you) almost as if you are mafia Don - you cannot be touched. Vas got too powerful and so you sent your goombah to make a hit. Or was this a grab for power, you "hit" Vas so that you could be the new boss and get the respect you deserve. I guess that makes me part of your "crew" along with Ken Thompson and several others. If I win a tournament do I have to pay you some respect money? Don't worry, I'll never rat you out - I know what happens to rats.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

User avatar
Laskos
Posts: 9458
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: World Computer Chess Championship ?

Post by Laskos » Fri Jun 15, 2012 3:09 am

bob wrote:
I do not follow your reasoning. Just because Rybka got caught as a derivative of Fruit and was disqualified, the WCCC is no longer the WCCC? The ICGA STARTED the WCCC. It is their event.
Even 10 years ago WCCC was inadequate for its pretensions, it already was mainly book preparation and hardware, only 4-5 strong engines, 3-4 games of relevance for each engine, many crap engines, but still, most of the strongest engines did participate. Nowadays there are 3-4 second-tier engines which are 200 points behind the leading bunch, the rest being some crap engines. WCCC recently became a total joke.

Kai

Adam Hair
Posts: 3201
Joined: Wed May 06, 2009 8:31 pm
Location: Fuquay-Varina, North Carolina

Re: World Computer Chess Championship ?

Post by Adam Hair » Fri Jun 15, 2012 3:50 am

Laskos wrote:
bob wrote:
I do not follow your reasoning. Just because Rybka got caught as a derivative of Fruit and was disqualified, the WCCC is no longer the WCCC? The ICGA STARTED the WCCC. It is their event.
Even 10 years ago WCCC was inadequate for its pretensions, it already was mainly book preparation and hardware, only 4-5 strong engines, 3-4 games of relevance for each engine, many crap engines, but still, most of the strongest engines did participate. Nowadays there are 3-4 second-tier engines which are 200 points behind the leading bunch, the rest being some crap engines. WCCC recently became a total joke.

Kai
Is it possible to conduct this discussion without insulting anyone? Even if an engine is not one of the strongest, that does not mean it is "crap".

User avatar
lucasart
Posts: 3041
Joined: Mon May 31, 2010 11:29 am
Full name: lucasart
Contact:

Re: World Computer Chess Championship ?

Post by lucasart » Fri Jun 15, 2012 4:23 am

bob wrote:
Dan Honeycutt wrote:
Don wrote:I would LOVE it if there were some simple non-intrusive test (like peeing in a cup) that would have caught Vas early on. I would gladly submit to such a test.

Since the ICGA bent over backwards to avoid the "perception" of unfairness and that did not satisfy people, what sort of investigation would you propose that would work?
I'm no more than lukewarm on Graham's idea but if I were charged with implementing it a couple competitors chosen at random would have to submit their sources to a judge or two. Judges would not be computer chess programmers but rather programming experts who would familiarize themselves with the top open-source programs. If you were chosen they would look at Komodo's sources to see if they could find any pieces of Ippolit or Stockfish or whatever. They could require a compilation to verify that the sources produced the exe you were playing but they would not be allowed to make any copies of your work. I'd expect the check to take 30 minutes to an hour, not as quick and easy as pissing into a cup but also not terribly painful.
Don wrote:When a program is tested to see if it is a clone of someone else's program, which program does it have to be checked against? Suppose the next program is a clone of Komodo but they check it against Houdini? How will I get satisfaction?
You would still be able to trigger an investigation if you had credible evidence that Komodo was cloned.

Best
Dan H.
I think the problem is that non open-source programmers are not very willing to provide their source to someone they think might look inside on the sly. It is a valid concern, and a stumbling block that doesn't have an easy solution.
I can see two easy solutions:
1/ the radical one: only open source can compete. That's my preference.
2/ moderate solution: to compete, you need first and foremost, to provide your full source code to the tournament orginasers. but you must tolerate no exception, or else you'll run into the same problem forever. If Rybka had compiled with that rule from the beginning, it would either have been accepted that Rybka is OK, or it would have been shown beyond reasonable doubt that it is a plagiarized work. And these discussions would have stopped there. Same for Loop and Thinker, and all closed source programms that have been accepted in the ICGA tournaments.

User avatar
Graham Banks
Posts: 33164
Joined: Sun Feb 26, 2006 9:52 am
Location: Auckland, NZ

Re: World Computer Chess Championship ?

Post by Graham Banks » Fri Jun 15, 2012 5:17 am

bob wrote:
Graham Banks wrote:
Don wrote:
Graham Banks wrote:
Don wrote:
Graham Banks wrote:
Don wrote:......There is a procedure to determine who the world champion is and there is. It should not be open the cheaters and copiers and it isn't......
How would you know that unless every engine was put under the same scrutiny as the chosen few?
Every engine? The way this works is that there has to be an accusation by one of the authors that someone is plagiarizing their work - in this case it was Fabien who made the accusation. The ICGA did NOT make the accusation.

It's ridiculously impractical for the ICGA to just launch a thorough investigation of every program in every tournament every time, taking a kind of paranoid (McCarthyism) stance that everyone must be guilty.

I don't known how things work in your part of the world but it would be pretty horrible to be around a culture like the one you suggest, that everyone should be investigated just in case they might be doing something wrong.
Isn't random drug testing like that? Perhaps they should randomly select two participating engines to scrutinise each year, with a rider that it can't be the same engine twice in a three year period.
They could do something like that, but I don't believe it would come out the way you think it would. By your implication you may have bought in to the argument that everyone is guilty but only a couple of people got caught. The computer chess community is pretty sharp and it's almost impossible to get away with this for very long. To underscore this principle the players in the online chess club can tell pretty quickly if you are using a computer to cheat and you will get flagged.

But this is a lot like life, we generally wait for an accusation before launching investigations and it's usually up the victim to take some interest in the process even though that is usually not a hard and fast requirement. In a scenario like you describe the victim could be every competitor in the tournament but I believe there should at least be a viable complaint of some kind before launching investigations, otherwise the ICGA becomes the oppressive tyrannical organization that a few extremists are accusing them of.
I don't think that all are guilty by any means Don. I'm interested in the perception of fairness, that the perception of some programmers being above scrutiny is got rid of.
By doing random drug tests, do we automatically assume that all competitors are guilty?
How about pointing out WHICH "programmers are above scrutiny?"

ME?

I was accused of cheating in 1986 and was thoroughly investigated, as most everyone knows. We were completely exonerated, after a thorough investigation that included several people looking at the CB source, running CB on the same machine we used in the 1986 WCCC event, using an executable Cray Research restored for the investigation from a regular system backup made during the event. So WHO exactly is "above scrutiny"???

Certainly does not apply to me.
No Bob - my post was not directed at anybody in particular, more at any engine that has never been "looked at" in a similar fashion to Rybka, Loop and Thinker.
My email addresses:
gbanksnz at gmail.com
gbanksnz at yahoo.co.nz

mcostalba
Posts: 2684
Joined: Sat Jun 14, 2008 7:17 pm

Re: World Computer Chess Championship ?

Post by mcostalba » Fri Jun 15, 2012 6:00 am

Don wrote: I don't think the ICGA is obligated to change the name of a tournament to thwart what other people do, that seems like a really silly idea. I agree that chessbase gets carried away with marketing, but doing away with the title is not a reasonable solution and since when is it ICGA's business to be the chessbase police?
Oh c'mon, ICGA is business related ! Did you see who are the first 3 engines of 2011 ? Do you think ICGA lives out of love ? Even you highlighted that to run these events ICGA needs money !

C'mon Don, do you think here we are all so naive to don't see that ICGA is backed up by commercial engines and their editors ?

The only one real reason why ICGA will never give up to that void name "World Computer Chess Championship" is because that (faked) "Wold Champion" title is what is needed by marketing people of organizations that support ICGA !!! This is even more than an elephat in the room, this is an entire zoo !

User avatar
hgm
Posts: 23723
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: World Computer Chess Championship ?

Post by hgm » Fri Jun 15, 2012 6:44 am

It seems you are totally off base. Do you have any proof that ICGA receives support from commercial Chess-program vendors at all, (like having seen their financial reports), or are you just shooting off your mouth?

AFAIK ICGA is funded by members, (like me), who pay a membership fee and a subscription fee for the ICGA Journal. For the events (WCCC / Olympiad, the conference) there usually are sponsors (well advertized, and usually scientific institutes / local authorities) that sponsor mostly by providing acomodations. In years where there are no sponsors the event is held locally at the university where the person that organises it works, so it can be cheap. Participants pay a significant registration fee, which in effect means the WCCC is mostly funded by participants to the Computer Olympiad. (E.g. although 80% of the event in Tilburg was Chess, people participating to both WCCC and WSCC payed only half from what I payed to participate in two one-afternoon 3-player competitions for Xiangqi and Shogi.)

This whole discussion is a non-discussion anyway. The WCCC is the real World Championship in computer Chess, because it is open to all. Like in any World Championship, willingness to participate (and motivation to win) is one of the essential traits on which a champion is selected. This thread is just a shameful display of cry-babies that are unwilling to perform, but want to be called champion anyway!

Opening theory is considered an essential part of the art of Chess, and the suggestion that it should be eliminated from the contest is outright ridiculous. If you don't like opening theory, go to a Chess960 contest... (Or better yet, play Spartan Chess! :lol: )

Post Reply