ICGA WCCC prospects ...

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

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 11:35 am

Re: ICGA WCCC prospects ...

Post by George Tsavdaris » Tue Jan 06, 2009 12:13 am

Mike S. wrote: Fairness requires more than "same rules for all", especially if a lot of money may be involved. Some years ago, it was a very wise decision of the sheik who financed Hydra, not to participate in WCCCs.
Yes it is very wise not to participate because you have very strong hardware. :shock:
In fact only the one with the weakest hardware should participate to make it totally fair, right?

It's like asking a football team not to participate in a tournament because their players are better than the others.

Sorry but that kind of arguments are utter crap. We will say no to Hydra if it wanted to participate because it uses very fast hardware? Are we serious here?


This decade is much different from the 1970s or the 1980s, in terms of top computer chess and the hardware used for it. Levy's decision is an update for ICGA which was required (and comes very late, but at least it comes now).
Different? How it's different?
Then and now there are participants with big and small hardware. Then and now there are big books and small books, good books and not so good books. Then and now are people with money and with more money to have different hardware. Then and now there are good programs and even better programs.
Were are the differences?
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....

User avatar
Bill Rogers
Posts: 3562
Joined: Thu Mar 09, 2006 2:54 am
Location: San Jose, California

Re: ICGA WCCC prospects ...

Post by Bill Rogers » Tue Jan 06, 2009 12:30 am

Vincent
You are right as far as it goes but I think and am pretty sure that a great many others agree with me that times and opinions have changed.
I think what people are interested today is which program plays chess the best on normal hardware, and by that I mean the kind that 90 percet of the public owns. Of course it is easy to say that if you won an 8 core machine you really don't give a hoot, but believe me the majority of the masses can not afford such a machine. Maybe in a few years when that is all that the public can buy but right now it is not the case.
Another thing is that there are appearantely dozens of books available to the public and I have even tried to download a few but to me they were all junk as I don't know how they were incoded so all I see is junk charactors. Good old TSCP books are printed in plain ascii and redable by all but then again they don't compare to the rest of the computer giants.
Please don't construe this as an attack on your posting, it is just the opinion of another chess player/author only.
Bill

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: ICGA WCCC prospects ...

Post by bob » Tue Jan 06, 2009 1:04 am

Mike S. wrote:
bob wrote:Do we also limit what positional structures a program can evaluate? What about the search space? Do we limit the kinds of extensions so that one program can't see that much deeper than another?

(...) Why don't you find a group of _programmers_ that want to see such restrictions. I don't care what _users_ want or what _spectators_ want, in this regard.
Of course, none of the other elements you mention should be limited.

But I think, actually the chess programmers should be the first and the loudest, to demand uniform platform hardware and book limits for a WCCC (!; not generally, of course all of us would like to use the best available hardware "at home" and that is normal). Differences in hardware and books distort the competition between the chess software (codes) themselves. Opening books are supporting data.
"I thnik...". First mistake. How about asking. _this_ programmer wants an open event with no restrictions of any kind to limit playing skill. I would like a restriction to enforce automatic interfaces so that no human intervention is possible, but that is a different issue. If I am the fastest, fine. If not. Fine. I have played on both sides of this fence and it didn't deter me from participating or continuing development. Again, don't "think". Just ask the programmers, as they are the ones involved in the process, not the spectators or end-users.


Again, an opening isn't 20 or 30 moves but much less. As long as book moves are taken, there is no computer chess research happening.
Please provide a citaton that supports that statement. First, GM players know some openings to 50 moves and beyond. Some opening variations are almost forced for 30+ moves in fact. Second, there is a _lot_ of computer chess research happening within the opening book. Ever heard of learning? Automatically adding to a book? How many different things can one use to choose which opening move to play (frequency, score, learned result, hints from the book creator, etc.) There is far more to this than you might suspect.

I cannot believe that chess programmers are happy with opening theory battles and hardware races, at a WCCC. A clearly better engine may lose an important game against a weaker one, due to a super clever opening trap. Is that funny? You will probably say yes, very entertaining. Backslapping for the book author. - But in fact, it is a tragedy if World Champion titles are decided by such BS, or by hardware one has and the other one has not.

A book and hardware competition is useless.

Hardware means money (or good relations/privileges).

For my taste, with current technology also the 6-piece tablebases are absurd. They are useful for endgame research and for correspondence chess analysis, but in comp vs. comp games they are in the wrong place IMO. That certainly is again something I won't convince many people about. But maximum is not optimum.
Actually, it is. Using EGTBs in a program is non-trivial, and is a subject all unto itself..

Also, as Levy mentiones there ARE chess programmers who dislike such unlimited tournaments. Maybe they prefer not to talk much about it, but simply keep away. It is called "voting with the feet."
Funny. They also "voted with their feet" at the uniform-platform event that used to be held (notice the "used to be held" part of that).

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: ICGA WCCC prospects ...

Post by bob » Tue Jan 06, 2009 1:08 am

Mike S. wrote:
Uri Blass wrote: World championship are not designed to find which program is best.
There are not enough games for that purpose.
I know, but it should be a fair computer chess competition which makes sense, not an opening theory battle and not a hardware test. I want that engine code wins world champion titles, not super books or super hardware.

Fairness requires more than "same rules for all", especially if a lot of money may be involved. Some years ago, it was a very wise decision of the sheik who financed Hydra, not to participate in WCCCs. This decade is much different from the 1970s or the 1980s, in terms of top computer chess and the hardware used for it. Levy's decision is an update for ICGA which was required (and comes very late, but at least it comes now).
You are arguing from an absolutely untenable position.

First, to find the "best program" (apparently your goal) requires a _lot_ of games. Not 11. Not 1100. More. Can't do that in a tournament. With just 9-10-11 rounds, it is _already_ a crap-shoot as to who wins. So that is moot.

Second, if you want equal hardware, why not just take the program at the top of the CCRL or SSDF list and crown it the champion? Equal hardware? Done. Plenty of games. Been there. Normal opening books? Got the T-shirt. That would seem to be exactly what you want.

Leave the WCCC alone. The ideas you propose would kill computer chess research completely. Until you are an author, you won't understand why.

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: ICGA WCCC prospects ...

Post by bob » Tue Jan 06, 2009 1:14 am

Mike S. wrote:There never was any interest of big scale anyway, in a WCCC, except in computer chess circles. Even in the chess world, a WCCC doesn't raise much interest.

Reasonable limitations make such a tournament certainly more interesting for customers, users, and normal chess players, than something which is lightyears away from anything they have at home.
The problem is, we don't care about customers, users or normal chess players. The purpose of the WCCC is to encourage computer chess program development. Not to please the spectators. Not to satisfy the users. You can design whatever kind of event you want. Some have done that. See CCRL, SSDF, Etc.


It may be that some participants are absolutely not interested in any public interest and do chess programming only to impress other chess programmers. That is their choice and needs to be accepted. - Once, I have suggested to play tournaments where that approach prevails, in a secret bunker. Then, I wouldn't want to know anything about it and I wouldn't want World Champion badges on a box.

There, they could even standardize the engine :mrgreen: with books and hardware remaining the differences. (Playchess' engine room actually almost works like that!)

If some programmers and others didn't get it yet, maybe they will get it now: If exclusive components, like absurd super hardware and top-secret "wonderbooks" (I don't need those), have big influence in the WCCC competition, it degrades both the title and the winning engine. That is what makes people lose interest in a WCCC.
People? Perhaps. Programmers? Hardly. And _that_ is the purpose of the WCCC. Not pleasing spectators, users or customers...


Nothing against Junior, which is a good engine, but the big number of titles it collected while it never was the strongest engine, is the perfect illustration of that problem. (Shredder which won many titles too, was at least really the strongest engine, for some period of time).

User avatar
Mike S.
Posts: 1480
Joined: Thu Mar 09, 2006 4:33 am

Re: ICGA WCCC prospects ...

Post by Mike S. » Tue Jan 06, 2009 1:51 am

I think everyone should do his computer chess activities in the way he enjoys the most. That includes many ways to create, test and use chess engines, and different groups of people. Since viewpoints and focuses of interest can differ (a lot sometimes), different opinions emerge, and disagreements. But I am sure that for the bigger part, most of us would agree about (other) computer chess topics.

There are more important problems than how to regulate a WCCC. Nevertheless, it is the "flagship" of important computer chess tournaments, and usually is the only one (if any) which makes it into the general chess media. This showcase function is very important for computer chess.

(Btw. I believe that it is not about demonstrating chess strength to the "wider public" anymore, since a couple of years. Does it matter if an event is played at 3000 Elo level, or at 3500 Elo level? I don't think so. That matters only within the competition itself.)

You are probably the one who knows best the differences between the 1980s with Cray Blitz, Belle, Bebe etc., and today where the hardware consists of general purpose personal computers.

In 1986, I had a homecomputer with 8 bit cpu, none of my chess comps was over 2000 Elo, and I read about Cray Blitz and thought it was the 8th wonder of the world. Now, close to me is a supermarket where I could buy a cheap computer which that company offers sometimes, with quite good components. They have it there between the vegetables and the freezers. Currently their offer contains an Intel quad cpu I think.

Also, up to the mid-eighties computer chess existed in private homes as nothing much better than a toy. - Nowadays, computer chess provides essential tools for the serious chess player, and the number of customers and users has grown. Chess software has become important beyond the computer chess "scene" of programmers and engine fans.

Due to all these changes, I think it is entirely normal that also the WCCC has to change. It's just about time.
Regards, Mike

User avatar
Mike S.
Posts: 1480
Joined: Thu Mar 09, 2006 4:33 am

Re: ICGA WCCC prospects ...

Post by Mike S. » Tue Jan 06, 2009 2:55 am

bob wrote: The problem is, we don't care about customers, users or normal chess players.
Yes, that is in fact a problem. I think you tell me your opinion, but would for example SMK, who lives from programming and selling chess software, sign this? On the first page of his online shop, he shows the list of Shredder's world champion titles. I'm not sure if you get my point.

You cannot isolate a WCCC from it's audience (that is why I made the satirical "secret bunker" remark). In the very moment when customers, users or normal chess players are interested in a WCCC, it is NOT only about computer chess research anymore.

Btw. what would that research be for, if the people doing it would not care for these groups mentioned? That seems like saying, I do medicine research but I don't care for diseased patients. Fortunately, computer chess advances have achieved practical goals and have led to useful products.

Do you dislike it, that computer chess is not restricted to the academic world anymore?

Also, I don't want the WCCC to try to find the strongest engine. I know that the few games of a WCCC are not statistically meaningful. But I don't see a contradiction here, with my other suggestions. A title is not important because he would be statistically meaningful, which it is not, but a title has a "real world" value, ideational and sometimes material.

Please don't get me wrong, but some of your opinions would make more sense to me, if all chess programmers of the world would be professors, and none of them would need to sell anything, and access to big hardware would not be a problem for any of them.

P.S. When Anand became world champion in the tournament in Mexico 2007, the number of games was fourteen.
Regards, Mike

bnemias
Posts: 373
Joined: Thu Aug 14, 2008 1:21 am
Location: Albuquerque, NM

Re: ICGA WCCC prospects ...

Post by bnemias » Tue Jan 06, 2009 3:14 am

With hardware limits, nothing noteworthy can ever happen.

I remember watching the ACM in 1991 here in ABQ and being most impressed with MCHESS. Here it was running on a 486 @ 50 Mhz, and it came in 2nd, ahead of Cray Blitz and Hitech. It looked like it was going to draw Deep Thought II for a long time until it made an error in the endgame.

You can argue that the situation is different today. You might even be right. But if you limit hardware, you absolutely will never watch a home computer beat a supercomputer.

User avatar
Harvey Williamson
Posts: 1949
Joined: Sun May 25, 2008 9:12 pm
Location: Newport. Shropshire, UK.
Full name: Harvey Williamson
Contact:

Re: ICGA WCCC prospects ...

Post by Harvey Williamson » Tue Jan 06, 2009 8:42 am

Mike S. wrote:
bob wrote: The problem is, we don't care about customers, users or normal chess players.
Yes, that is in fact a problem. I think you tell me your opinion, but would for example SMK, who lives from programming and selling chess software, sign this? On the first page of his online shop, he shows the list of Shredder's world champion titles. I'm not sure if you get my point.

You cannot isolate a WCCC from it's audience (that is why I made the satirical "secret bunker" remark). In the very moment when customers, users or normal chess players are interested in a WCCC, it is NOT only about computer chess research anymore.

Btw. what would that research be for, if the people doing it would not care for these groups mentioned? That seems like saying, I do medicine research but I don't care for diseased patients. Fortunately, computer chess advances have achieved practical goals and have led to useful products.

Do you dislike it, that computer chess is not restricted to the academic world anymore?

Also, I don't want the WCCC to try to find the strongest engine. I know that the few games of a WCCC are not statistically meaningful. But I don't see a contradiction here, with my other suggestions. A title is not important because he would be statistically meaningful, which it is not, but a title has a "real world" value, ideational and sometimes material.

Please don't get me wrong, but some of your opinions would make more sense to me, if all chess programmers of the world would be professors, and none of them would need to sell anything, and access to big hardware would not be a problem for any of them.

P.S. When Anand became world champion in the tournament in Mexico 2007, the number of games was fourteen.
Good Post.

User avatar
Kirill Kryukov
Posts: 492
Joined: Sun Mar 19, 2006 3:12 am

Re: ICGA WCCC prospects ...

Post by Kirill Kryukov » Tue Jan 06, 2009 9:43 am

Mike S. wrote:I support Levy's decision (which I hope it is, in fact) but as always, I see things from the user & fan viewpoint. It would not a bad decision though, if programmers would share exactly that viewpoint, except somebody works only for (or against) the other programmers, or for academic fame in a small circle. - But I know that there are many super technology freaks who have very much trouble now, due to this 8 cores limit.

Other people have explained that 8 cores are still too many, because a good 8 core computer is MUCH more expensive than a good quad, which means that money is still a big factor of success in that competition.

To put it short, money (or insider's relations to hardware labs with latest CPUs and other privileges like that) should be banned, and the hardware limit is a big step forward in this direction. Bravo Levy!

Note, that at the same time he announced that at the computer olympiads, participation of unlimited systems will be possible. But not at the WCCC which is about computer chess, not about (all purpose) hardware, and not about money. Otherwise, maybe you would not even need to play chess. Just bring all your cash, and the one who has the biggest pile, is champion.

Also, I would to mention that uniform platform was not a problem at all, at the Chess960 computer world championships in Mainz, 2005-2008. Did anyone NOT join them because the hardware was limited?! I don't think so. It's a prove that it can be done, and the software competition is more meaningful under such circumstances.

I would even standardize and/or limit the books, but I am afraid that is a too radical proposal. Nevertheless, I wonder why some seem to mistrust their chess software creations so much, that they need to put their hopes in super hardware and super tuned opening books, instead.
Just my personal 2 cents.

As a computer chess fun, I am strongly against the WCCC hardware limit. The no-limit tournament is more fun to watch, and it inspires imagination.

As a member of CCRL testing group, I however welcome the change, as it will result in WCCC losing any remaining relevance it still had, and my (and others) testing efforts will become more valued and appreciated.

So I guess I am happy in any case.

Post Reply