Very Strong Free Book

[DomLeste]

Winning games with a book up to 35 moves defeats the purpose. How about do 32 men Tablebases Engine wouldnt have to think. I dread that day but its coming!

I like to see freestyle tournament limit books up to 15 moves then we see some more skill trying to get advantage from 15 move book!

[Mark Mason]

Hi,

Reading these latest comments I am guessing you guys don't spend too much time swimming in the shark infested waters of the Playchess server engine room, where any edge that can be brought to bear over the monster hardware there is eagerly snapped up - even a win straight from book !! Trust me there are many keen takers for a well tuned book with winning lines - however deep.

Mark

[Dann Corbit]

Rgarding the size, although it is 74Mb it is only made from 59,719 games. The reason it is a big file is that it is 50 moves deep.

Why would anyone want such a deep book? What is the use of even having the engine?

It's about the same size as the stock junior book:

Code: Select all

 Directory of H:\chess\chessbase\data\Books

06/10/2008  11&#58;59 PM    <DIR>          .
06/10/2008  11&#58;59 PM    <DIR>          ..
06/07/2008  01&#58;50 PM           177,674 Book options optimize.png
06/07/2008  12&#58;23 PM            16,396 CompMaster Beta1.0.ctb
06/07/2008  12&#58;32 PM       278,769,664 CompMaster Beta1.0.ctg
06/07/2008  12&#58;23 PM         1,048,564 CompMaster Beta1.0.cto
06/11/2008  12&#58;08 AM               502 CompMaster Beta1.0.ini
11/10/2002  03&#58;38 PM             8,204 Fritz8.ctb
11/10/2002  03&#58;38 PM       166,105,088 Fritz8.ctg
11/10/2002  03&#58;38 PM           524,288 Fritz8.cto
08/09/2003  02&#58;29 AM             8,203 Hiarcs9.ctb
08/09/2003  02&#58;29 AM       141,189,120 Hiarcs9.ctg
08/09/2003  02&#58;29 AM           524,256 Hiarcs9.cto
03/31/2003  02&#58;05 PM            16,372 Junior8.ctb
03/31/2003  02&#58;05 PM       269,717,504 Junior8.ctg
03/31/2003  02&#58;05 PM         1,047,800 Junior8.cto
11/13/2002  07&#58;10 PM             8,203 Shredder7.ctb
11/13/2002  07&#58;10 PM       176,332,800 Shredder7.ctg
11/13/2002  07&#58;10 PM           524,252 Shredder7.cto
              17 File&#40;s&#41;  1,036,018,890 bytes

As to why such a deep book:
Free wins.

Most of the time, you won't be in the book nearly as long as the stored lines anyway.

[M ANSARI]

You have to be very careful using these books ... they are almost 100% based on Rybka games and if Rybka is your target then they are very useful. They work almost as "hashed" best moves of certain positions. BUT I did try such a Rybka optimized engine match in a 16 0 tourney on very powerful 8 core hardware ... and was shocked to see that Rybka came in last place in that tourney. It came behind Zappa II and even behind Deep Fritz 10 ... I don't ever remember Rybka losing such a tournament and thus I attribute it to book choice. So I think many of the lines can be refuted at longer time controls and this book would probably only be useful at 3 0 or 5 0 ... and only if your opponent is Rybka.

[Dann Corbit]

You have to be very careful using these books ... they are almost 100% based on Rybka games and if Rybka is your target then they are very useful. They work almost as "hashed" best moves of certain positions. BUT I did try such a Rybka optimized engine match in a 16 0 tourney on very powerful 8 core hardware ... and was shocked to see that Rybka came in last place in that tourney. It came behind Zappa II and even behind Deep Fritz 10 ... I don't ever remember Rybka losing such a tournament and thus I attribute it to book choice. So I think many of the lines can be refuted at longer time controls and this book would probably only be useful at 3 0 or 5 0 ... and only if your opponent is Rybka.

This tournament:
http://www.sedatchess.com/scct_opening_book.html
is at 4 minutes + 2 seconds (but on very high performance hardware).

All engines are either Rybka or Zappa and so your description about narrow scope of opponents clearly applies.

So the information we have about this book shows that it does well at fast time control {albeit on QX6700 Quad 2.66 GHz (using 4 threads)} against Rybka and Zappa.

But lots of people like to play this sort of game online, so it may be very useful for them. I suspect that the winning book will become very popular and find its way into slower paced tournaments and we will find out if it is also a good book at a slower pace.

[M ANSARI]

Yes ofcourse ... if you know the limitations and plan to only use it on Playchess then that is fine. There will be a lot of strong lines even at slow time controls ... but also there will be a lot of bad lines it considers good ... and you can be sure that someone will be looking hard at these lines and refuting them and saving them for an opportunity to use.

[Mark Mason]

In addition to their use on the Playchess server, I think strong books like this - albeit with some limitations - are also useful to serve as a separate source of reference when working on one's own book. eg. in a particularly tricky position where you are trying to find a good continuation, it is always useful to have other views of what is considered a good line so that you can at least evaluate them and consider them for inclusion or not.

Mark