In Answer To Why Zappa Is So Far Down In CEGT

[Spock]

That's what you said about private engines for 40/40, but you still changed your mind it seems. We just want to see it on the blitz list, and you could indicate by changing names.

That is somewhat different - private engines play under the same conditions as every other engine

To include engines on the list that play with own books would distort and ruin the whole list, render it meaningless. Every engine on the list has to play under the defined playing conditions.

[Tony Thomas]

That's what you said about private engines for 40/40, but you still changed your mind it seems. We just want to see it on the blitz list, and you could indicate by changing names.

That is somewhat different - private engines play under the same conditions as every other engine

To include engines on the list that play with own books would distort and ruin the whole list, render it meaningless. Every engine on the list has to play under the defined playing conditions.

It wouldnt make it meaningless. It would just show you how much difference the own book makes. Also, I am pretty sure it is an easy task to just delete the games if you find out that it actually distorted the list. The only two lists that can make such a test is CCRL, and CGET. I doubt that SSDF has the computer time to make such a test, and introduce an engine with a general book.

[Graham Banks]

It wouldnt make it meaningless. It would just show you how much difference the own book makes. Also, I am pretty sure it is an easy task to just delete the games if you find out that it actually distorted the list. The only two lists that can make such a test is CCRL, and CGET. I doubt that SSDF has the computer time to make such a test, and introduce an engine with a general book.

We can't spare the computer time either to be honest and it won't happen anyway as it goes against our core testing conditions.

The value of specific tuned books to certain engines is well known and could easily be tested by interested individuals.

Regards, Graham.

[Spock]

The value of specific tuned books to certain engines is well known and could easily be tested by interested individuals.

Regards, Graham.

I would certainly be interested in running some tests, totally outside our ratings list, although what format they would take I don't know.

[Tony Thomas]

The value of specific tuned books to certain engines is well known and could easily be tested by interested individuals.

Regards, Graham.

I would certainly be interested in running some tests, totally outside our ratings list, although what format they would take I don't know.

Could you possibily integrate it in to the rating list (not on the main page) but just on the side to see how much difference it makes? Only one engine is required to prove the point with enough games to people that arent boneheads. Since you are a Zap fan, Zanzibar would be a good choice, so is the former world champion Jr.

[Spock]

The only way to do this in my view is to do it in a scientific manner - exactly the same opponents and the same number of games

I've not heard that Zap has a particulary good (or bad) book - Junior or Hiarcs is better choice. Say we do Hiarcs:

Hiarcs generic book vs 10 opponents with generic book, 50 games each
Hiarcs own book vs 10 opponents with generic book, 50 games each

This isolates the effect of Hiarcs using it's own book instead of a generic. And then run the games through ELOstat or bayeselo and compute the rating

Would have to be blitz of course.

Any other ideas ?

[Ovyron]

George, for a change you guys should start using own book for the engine.

I think Zappa Mexico doesn't come with an own book, however, turbojuice1122 said that TourBookII.ctg "seem to suit Zappa Mexico quite well". He made a match with Zappa Mexico TourBookII.ctg vs. Rybka 2.3.2a RybkaII.ctg 60'+20", on 1 CPU with book learning ON. Rybka won 13-11, Zappa had one book win and Rybka had 2 book wins. I think other matches like these would be interesting.