ICGA rule #2 / opening books / Diep-Crafty, Turino 2006

[Terry McCracken]

You refer the letter to the CSVN?

Since the signees were former participants the number was restricted anyway.

As already mentioned by Don it was fair towards the CSVN to declare the boycott rather than only to don't silently participiate. To make their current direction aware of their decision to don't ban Rybka from their tournaments. Their reasoning was flawed and wrong. They wondered why programmers took 5 years to change their mind, and on the other side did not mention Ed's change of mind within five weeks, using him as primary advocat for their pro Rybka decision.

I guess your mail statistics from all your fans does not reflect a meaningfull measure

One thing we've not been told is how many other programmers were invited to sign the "Leiden letter", but chose not to.

You starting it again are you? You've proven my point in the other thread you started. You know damn well what you're doing and it's not a neutral position you're taking. In fact you're almost as bad as George with the language toned down a few notches. You're trolling Graham, it's that simple!

The best of the best looked at the evidence and all came to the same conclusion. If you have no idea who Ken Thompson is or Tony Marshland et al than look them up instead of defaming their characters!

AFAIC you're a liar and a hypocrite! It's not Keith who's the tosser, that's for certain!

I believe that it's important that we know the full story behind such matters Terry, that's all. Sometimes we only get presented with half the story.
For example, if it was known that 50+ programmers chose not to "sign" the "Leiden letter", whereas 14 did, it does put a rather more interesting perspective on the issue.
I don't know how many programmers were invited to sign the letter, but it would have been interesting to know.

You really are dense. You're spewing nonsense. You're the same mule today as you were in '07 when this began.

Thanks for your continued insults Terry. Seems like the charter has gone out the window lately.

You're welcome....the charter isn't the only thing that's out the window!

Not that I expect you to understand.

[Peter Berger]

I think the main thing to avoid is a wholesale copying of any particular work. And this wouldn't probably produce a really strong book, in any case. Many current opening books, especially the professional ones, contain strong novelties arrived at by computer analysis. So the best books are far from just being copies, but have quite a bit of originality.

I am not convinced that this is true, unless there have been massive changes during the last 4 or 5 years.

Of course opening books will contain some strong novelties in common lines arrived at by computer analysis, but mathematically they have to be pretty irrelevant. Even if an opening book of a chess program had 100 strong and somehow relevant novelties ( which is an absolutely impressive number), there will be many more lines without these novelties.

What really matters is the design of an opening book For a tournament a good book will be pretty deterministic. This would be horror for a release version for the public but in a tournament with just a few games learning or variety are mostly irrelevant ( e.g. after a lost or problematic game you will fix the holes manually as no learning procedure will be better than that).

Starting with some database of games will most likely lead to desaster in at least one tournament game,and I never saw this approach really work for a competitor. IMHO it is impossible to avoid that some random game(s) in it contain losing blunders the engine could have easily avoided on its own, which is the absolute horror scenario.

So the good books are designed like the „opening books“ of a grandmaster. They form some kind of repertoire for a given engine for an event.

Now let's assume that the book author himself has an ELO of 1700-2300 ( this should cover nearly all of them).

This clearly makes him less competent than any good common chess book author for humans ( most of them are strong or very strong grandmasters) – and of course mostly also less competent than the engines themselves .

So the main task and quality of the book author is to make choices IMHO ( a bit like a trainer in other sports who doesn't have to be more competent than his pupil in the given sport either)..
But what will he choose from? I am convinced that all good ones borrow heavily from copyrighted work done by stronger and more original and competent players than themselves.

Speaking for myself and Crafty 2004 I came to the conclusion that it sucked heavily ( compaired to the absolute top) when it is about being speedy at tactics and the understanding of king safety. It was competitive in endgames though and not much worse in positional understanding. In fact a bit like an older good human chessplayer.

So I had some idea on what kind of positions I wanted to reach ( this could be called original thought work). I also had some idea of what I wanted to avoid ( any line not blunder-checked by a computer and those going against its strengths and playing to its weak sides instead).

But given the above assumptions of playing strengths of book authors and engines the actual implementation was mostly theft and tedious work to crosscheck existing ideas by others.

I was lucky to know a concise opening book for humans ( I still think it is great btw!) that met my basic needs. „The Chess Advantage in Black and White: Opening Moves of the Grandmasters“ by IM Larry Kaufman, published this very year by Random House ( yeah, I know this is funny in itself, but he was not active or connected with any computerchess project at that time for all I know)

So the basic framework of Crafty's opening repertoire for Ramat-Gan in the main event was simply a blunderchecked typing work of this book ( of course a bit more than that, but that was the basic design).

I know that other book „authors“ did extremely similar things, they just don't talk about it.

Peter

[jdart]

See http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=19575 for an example of what I am talking about.

[Peter Berger]

See http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=19575 for an example of what I am talking about.

That's actually one of the *very* few messages I have read about computerchess in recent years ( big coincidence).

I feel that it absolutely confirms my point of view how good books are done( you just have to read between lines just a little bit).

Btw - don't get me wrong: Crafty's books in these events contained loads of original work and as a whole could have been somehow copyrighted by me if they had been published ( or have they? I don't know - I would definitely have no problem with it at all).

I think that some version of the 2006 book e.g. when Marc Lacrosse and me joined forces was most probably the second best book of all competitors ( Rybka's was still ahead ..).

But everything really new was borrowed from a creative point of view IMHO. And I think the reason for this is simple logic - it is about chess understanding and knowledge.

Peter

[jdart]

I still don't think borrowing is exactly the bulk of it.

This site is also interesting: http://chessok.com/?page_id=352. This is an online chess database (Chess Assistant) but if you look near the leaf nodes, you will often find moves that have computer evaluations but no game reference - I believe this means they haven't yet been played in a game, or at least not one in the db. So this db contains both known moves from games and potential novelties. I think this is pretty common now, in opening books too. So the source material is not just games and books anymore.

[geots]

You refer the letter to the CSVN?

Since the signees were former participants the number was restricted anyway.

As already mentioned by Don it was fair towards the CSVN to declare the boycott rather than only to don't silently participiate. To make their current direction aware of their decision to don't ban Rybka from their tournaments. Their reasoning was flawed and wrong. They wondered why programmers took 5 years to change their mind, and on the other side did not mention Ed's change of mind within five weeks, using him as primary advocat for their pro Rybka decision.

I guess your mail statistics from all your fans does not reflect a meaningfull measure

One thing we've not been told is how many other programmers were invited to sign the "Leiden letter", but chose not to.

You starting it again are you? You've proven my point in the other thread you started. You know damn well what you're doing and it's not a neutral position you're taking. In fact you're almost as bad as George with the language toned down a few notches. You're trolling Graham, it's that simple!

The best of the best looked at the evidence and all came to the same conclusion. If you have no idea who Ken Thompson is or Tony Marshland et al than look them up instead of defaming their characters!

AFAIC you're a liar and a hypocrite! It's not Keith who's the tosser, that's for certain!

I believe that it's important that we know the full story behind such matters Terry, that's all. Sometimes we only get presented with half the story.
For example, if it was known that 50+ programmers chose not to "sign" the "Leiden letter", whereas 14 did, it does put a rather more interesting perspective on the issue.
I don't know how many programmers were invited to sign the letter, but it would have been interesting to know.

It is nice to be able to stay calm and try to explain your point. It is useless here and a waste of time. Why not try the truth without any dessert. The open letter signers to Cock think they are so goddam important they can influence what does and doesn't happen. Kudos to Cock. And kudos to the ones who refused to sign the stupid letter.