Botvinnik and the WCCC/ICCA

[BB+]

From Computers in Chess: Solving Inexact Search Problems by M. M. Botvinnik translated by A. A. Brown [I also have access to the German translation, if anyone is interested]. Original from 1978, translation from 1984.

From Chapter 6, The Second World Championship, where MMB first gives some background:

The First World Championship contest among chess playing programs took place in Stockholm in 1974, under the aegis of the IFIP (the International Federation for Information Processing). The winner was the Soviet program KAISSA.

The history of chess tournaments among computers begins in 1970, when the first U.S. Championship was held in new York, timed to coincide with the annual meeting of the ACM (Association for Computing Machinery). Since then, these championships have been conducted annually.

World championships for computers are held triennially, as they are for humans. The second was held in August 1977 in Toronto, Canada.

These contests demand a substantial financial outlay, in machine time and communication channels. Therefore they are usually conducted on the Swiss system, in four rounds. The time limit is set at 20 moves per hour.

After giving the 1974 and 1977 results, and the demonstration game between KAISSA and Chess 4.6, Botvinnik mentions OSTRICH as being unique for being a portable microcomputer [which crashed in Round 3], and then turns to the "spectacle" of the WCCC. I emphasise his concluding comment here:

A championship contest among computers is an engrossing spectacle. The authors of the programs sit at chess boards. Using video terminals, they inform their own distant computers of the moves made by the opponent. The computer's answer is displayed on the terminal and echoed on the chess board. While waiting for the answer the scientists visit amicably among themselves, analyze positions, argue, joke, and often criticize the performance of their own programs. This is all quite understandable: a computer tournament is only formally a sporting event---in essence, it pursues scientific aims.

Botvinnik then mentions a bit more about PIONEER, and then turns to the WCCC and ICCA formation.

The third World Computer Championship will be held in 1980. [...] At the end of the Toronto championship, the competitors---the authors of the programs---took part in a conference in which the Dutch programmer B. Swets called for the formation of an ICCA (International Computer Chess Association). This was approved in principle. A sign of the times!

[ernest]

Hi BB+,

Usually (I think) I understand (and mostly appreciate) what you mean to say in your posts, but this time I quite didn't make it...

So what's your point here?

Besides, I'm always very cautious when I read about Botvinnik in Computer Chess: his Pioneer project and demos have been shown to be dishonest fakes

[Jeremy Bernstein]

Hi BB+,

Usually (I think) I understand (and mostly appreciate) what you mean to say in your posts, but this time I quite didn't make it...

So what's your point here?

Besides, I'm always very cautious when I read about Botvinnik in Computer Chess: his Pioneer project and demos have been shown to be dishonest fakes

If I read correctly, I think he was pointing out the difference between "a chess tournament" (common public perception of computer chess events) and "a programming competition/exhibition" (the actual nature of those events). In light of the current Rybka situation, one of the complaints is "you can't have a WCCC without the #1 [sic] chess engine!", which is clearly fallacious once it's understood that the WCCC is a programmer's event based around original, self-programmed chess software.

Jeremy

[BB+]

Besides, I'm always very cautious when I read about Botvinnik in Computer Chess: his Pioneer project and demos have been shown to be dishonest fakes.

Yes, I was careful not to quote from those parts of the book. I see no reason to see why Botvinnik's opinion about the purpose of the WCCC/ICCA should not bear some weight, though of course it is subject to ridicule just as anyone else's.

Personally, I'd say that "scientific aims" is a bit strong for the "essence" of the WCCC, though interchange of ideas between researchers is indeed one of the goals. The German translation is not split into paragraphs the same, but the last line of the second paragraph there says: Die Weltmeisterschaften im Computerschach sind nur der Form nach ein sportlicher Wettkampf, im Grunde genommen dienen sie jedoch wissenschaftlichen Zwecken.

The first and third quotations were somewhat interesting from a historical point of view (I didn't know that B. Swets got credit for proposing the ICCA, or that the IFIP organised the first WCCC). With the now-perennial debate on whether 11 games/9 days is too many, the first quotation about resource outlay vis-a-vis the later science/sporting aspects might be of interest.

[BB+]

Another (similar) opinion is noted in the Postscript in the Second Edition of More Chess and Computers by Levy and Newborn. On page 101, in recapping the tenth ACM computer chess championship (Detroit 1979), we find: John McCarthy awarded the trophies at a conference luncheon following the tournament and stressed the importance of viewing the tournament as both a scientific experiment and a competitive event.

[hyatt]

I believe there was a comment in an ACM Journal several years back (90's) that said "the ACM computer chess championship is the longest-running experiment on record," having covered 25 years by the time the last tournament ended.