In Answer To Why Zappa Is So Far Down In CEGT

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

Moderator: Ras

User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by geots »

AGove wrote:Okay George, You've explained yourself very well and I accept what you say.

There is one caveat to what i have said that i failed to mention. Zappa is scaling very well on 8, but we will not see this play on a single cpu, or even a dual. You would just about have to go to a quad to start seeing the benefits. And how many of us can afford that? Thats the only drawback i see to the program that is playing and what we will see. But i bought it and will be starting a gauntlet in a day or 2 with Zappa Mexico playing 20 games each ag. all the top engines- 1 CPU, for CCRL at 40/40/ repeating.
Time will tell.

The Best to You
Tony Thomas

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Tony Thomas »

George, for a change you guys should start using own book for the engine. I dont agree with your speculations, it is true that you might not get the special book that Erdo is using, but other than that I doubt that Anthony is trying to misguide the customers.
User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by geots »

Tony Thomas wrote:George, for a change you guys should start using own book for the engine. I dont agree with your speculations, it is true that you might not get the special book that Erdo is using, but other than that I doubt that Anthony is trying to misguide the customers.

As i stated above, if Anthony says it, then i have to believe it, as i believe his word is good. As for CCRL using own books, neither CCRL nor CETG use own books. I have been thru this before, but i will repeat it once again for your benefit: These are the 2 basic ways of testing- when you use own book you are testing the complete program. When you use the same generic book for both, you get a much better indication of pure engine strength- as neither engine has the advantage of a better book. Neither way is wrong- you are just looking for 2 entirely different things. SSDF does use own books. They are not wrong to do this, and we are not wrong NOT to do this. Its purely a matter of what you are looking for. There has been some speculation that there are a few programs that will play better with own book, because that engine has been progammed to get better results with its own book. So far as i can tell, no one has offered any concrete proof that is true. Just as no can offer any concrete proof that ponder on is better than ponder off. The important thing is that both are doing it the same way. But there are people out there chomping at the bits to get in an argument about that. But they will have to find someone other than me to argue with. I dont have the time or the inclination. One thing that i find extremely amusing is that there are people ready to write Rybka's eulogy. But in my Rybka 2.3.2a Gauntlet, Fritz actually started off with a 3 GAME lead over Rybka. Of course Rybka overtook him and won. But if i had conluded at the time Fritz had the 3 game lead- after 7 games, that because of this Rybka was weaker than Fritz- i would have been laughed off the forum. But the same people who would have laughed at me there are preparing to attend Rybka's funeral. You run tourn. yourself, Tony. You tell me what you can conclude from 6 or 10 games. ABSOLUTELY NOTHING. Like Fischer said, "One day you give me a lesson- the next day i give you one" I have to naturally pull for Rybka, because Vas is a very good friend of mine- and we talk frequently. I was actually talking to him today about this match that's going on. If Zappa wins, which it looks more like it will with each passing day- you will hear no excuses from Vas. He knows it's just the nature of the beast. And the world will keep turning.

Best To You Tony,

George
Tony Thomas

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Tony Thomas »

If you want cocrete proof as to weather or not own books help the engine, why not test two or three engines with own books? Zap chess Zanzibar, Junior and Hiarcs comes to mind. Yes, I know that nothing can be said from 10 games, I have had instances were a new version of an engine didnt lose in the first 25 games, and all of a sudden it had a hard time winning even one game. My response was to your initial post, in it you hypothesized that it is possible that the version that plays against Rybka is much different. My argument was, in a tournament like this the book makes a lot of difference.
User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by geots »

Tony Thomas wrote:If you want cocrete proof as to weather or not own books help the engine, why not test two or three engines with own books? Zap chess Zanzibar, Junior and Hiarcs comes to mind. Yes, I know that nothing can be said from 10 games, I have had instances were a new version of an engine didnt lose in the first 25 games, and all of a sudden it had a hard time winning even one game. My response was to your initial post, in it you hypothesized that it is possible that the version that plays against Rybka is much different. My argument was, in a tournament like this the book makes a lot of difference.

I agree with everything you have said here, Tony. Just understand that when i made that statement about Zappa, Anthony had not come out with his statement that it was the same. And i take him at his word- he is an honorable man. As for us testing some with own books- that is something you will have to take up with Kirill, Graham and Ray. I am just a lowly peon in the scheme of things.

Nite Now and Regards,
User avatar
Graham Banks
Posts: 44868
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Graham Banks »

geots wrote:As for us testing some with own books- that is something you will have to take up with Kirill, Graham and Ray. I am just a lowly peon in the scheme of things.

Nite Now and Regards,
Now now George - we're nice people. We don't peon anybody! :lol:
gbanksnz at gmail.com
Uri Blass
Posts: 10970
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Uri Blass »

geots wrote:
AGove wrote:Okay George, You've explained yourself very well and I accept what you say.

There is one caveat to what i have said that i failed to mention. Zappa is scaling very well on 8, but we will not see this play on a single cpu, or even a dual. You would just about have to go to a quad to start seeing the benefits. And how many of us can afford that? Thats the only drawback i see to the program that is playing and what we will see. But i bought it and will be starting a gauntlet in a day or 2 with Zappa Mexico playing 20 games each ag. all the top engines- 1 CPU, for CCRL at 40/40/ repeating.
Time will tell.

The Best to You
I guess that it is wrong that you will not see this play on a single cpu.

You will see it if you wait more time.

You may not see the same results if you play equal time control but I see no reason to play equal time control and you can give zappa more time as compensation for the fact that zappa is not using 8 cpu's.

Uri
Tony Thomas

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Tony Thomas »

Graham Banks wrote:
geots wrote:As for us testing some with own books- that is something you will have to take up with Kirill, Graham and Ray. I am just a lowly peon in the scheme of things.

Nite Now and Regards,
Now now George - we're nice people. We don't peon anybody! :lol:
It would be nice though if you were to add two or three top engines, not Rybka, with own book, just for experimental reasons.
Spock

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Spock »

Tony Thomas wrote: It would be nice though if you were to add two or three top engines, not
Rybka, with own book, just for experimental reasons.
We certainly could not do that on our main list. But some separate tests might be interesting.
Tony Thomas

Re: In Answer To Why Zappa Is So Far Down In CETG

Post by Tony Thomas »

Spock wrote:
Tony Thomas wrote: It would be nice though if you were to add two or three top engines, not
Rybka, with own book, just for experimental reasons.
We certainly could not do that on our main list. But some separate tests might be interesting.
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.