which engine have the best own book??..

[ArmyBridge]

Hi friends
I wonder which engine has the best own book, Rybka? Junior?, I know Hiarcs has one very good.. few days ago
M ANSARI ran a match between Zanzibar x64 on Quad vs Rybka 2.3.2a x64 Quad, and Rybka crushed Zap.. in 100
games Zap only got 3 wins!! both engine played with own Book, now CEGT and CCRL have tested Rybka 2.3.2a and
get very diferent result on Rybka-Zanzibar, here Zanzibar get close to Rybka, problably if someone test Rybka with
own book wolud get 3200 elo?? I´m sure that Junior is stronger with own book , how much?, I think 30-50 elo (maybe more??) than
it have in CCRL or CEGT (both hanlde their own book), I know their test the "pure" engine and this is the reason
that both tester used their own books (that have a balanced opening selection).But I Think Kasparov performance
would have decreese if forced him to play the french defence or caro kan etc, because he is Atack player so why not let play Junior
Fritz or Hiarcs under their own book´s??, at leats it would be very fun to see, dont you think so?

[Tony Thomas]

It would be nice to see them testing engines with own books. Almost all the top engines except Naum, and Tiger got highly tuned books. May be they should let some of Rybka's top 5 rivals to play with own books to see if it indeed help them get closer in rating to rybka.

[Mike S.]

With general, not engine tuned books, the information value for the normal chess player and customer is better. For him, it is not important which repertoire the engine is good in, but how good it is in a competition with wide generic repertoires.

[cornetmike]

Now I'd like to see which engines are the strongest WITHOUT using opening books! This might be an interesting test.

regards
Mikey

[Mike S.]

Now I'd like to see which engines are the strongest WITHOUT using opening books!

But by that, only 2 games per pairing make sense, X-Y and Y-X, the rest would be a test of learning functions many don't have and not be representative for the general engine strength. In other words, a second game X-Y without book will basically be a copy of the first game X-Y, etc.

Other alternatives are Chess960, or short generic books.

http://chesstigers.de/ccm7/computer_wm. ... 6&subkat=1
(see also the events from 2005 and 2006 there)

http://www.utzingerk.com/FT_2006.htm
http://www.utzingerk.com/5moves.htm

from: http://www.utzingerk.com/index.htm

(K.Utzinger and R.Bühler have ended their series of tournaments)

[Dr.Wael Deeb]

Testing with generic opening books doesn't mean that you are testing the'"pure" engine's strength,you are actualy testing the different opening books....
What a "pure" engine strength we are talking about when you force the engine to get into positions not familiar with it at all
Junior and the Rybka\Zap!Chess Zanzibar is a brilliant demonstration of what I am talking about....

[AdminX]

Testing with generic opening books doesn't mean that you are testing the'"pure" engine's strength,you are actualy testing the different opening books....
What a "pure" engine strength we are talking about when you force the engine to get into positions not familiar with it at all
Junior and the Rybka\Zap!Chess Zanzibar is a brilliant demonstration of what I am talking about....

I agree!

[Mike S.]

Do you ignore the normal chessplayer, a user who wants to study the games and openings HE is interested in, and he wants to practise HIS repertoire and if he is a club player, most probably practise against the repertoire of HIS opponents.

It is a WRONG CONCEPT of engine strength if you see it only in combination with it's specific tuned book. For most users and buyers, an engine is a tool for analysis and practise, and it is relevant how good it is in ALL openings, across a big variety (wide repertoire), not only in those the engine is especially good in. Engine vs. engine sport for server rankings etc. is something entirely different and only a small niche of computer chess in general. It can be interesting, exiting, entertaining, but it is not a relevant criteria.

(Actually, rankings with own or generic book are not that different anyway, Junior being the only big exception. Didn't notice such a big effect with Zappa, yet.)

Thanks god that the rating list testers of CCRL,CEGT,CSS know what they are doing! Their decision against engine-specific books was right.

[Rubinus]

1.RybkaII.ctg! Splendid work by Jeroen. And 2.Hiarcs 11.1, but own format (team H. no sleep ...). 3.? (probably Shredder - but little elderly or Fritz).
Other:
Zap!chess - some odd variants, e.g. Blackmar-Diemer, limited Volga gambit. Sizilian close variant is logical choice (more knowledge, little speed).
Sjeng - original perforated, e.g. King Indian
Junior - little variability
Loop - newllbook.ctg and LOOPandLISTBook2006.ctg? Humorous.
Gandalf - solid
Ktulu - limited
Smarthink - original, interesting, but some odd variants, limited. Sizilian 2.Nc3 and 3.f4 or 2.c3 logical choice for little speed.
Naum - passable

[Rubinus]

If not will use original or limited (e.g. 9moves.ctg) books then use Takker TourBookII.ctg. Example, for Loop my guess cca >50ELO+.