Use my SALC and OLIK opening-books for a lower draw-rate

[pohl4711]

Some weeks ago, I created my SALC opening book for engine-engine matches. In all lines (created out of 10000 human-games, all lines 20 plies deep (all lines checked with Komodo 10.2 (20'' per position, running on 3 cores), evaluation inside of [-0.6,+0.6])), white and black castled to opposite sides, both queens still on the board. The idea is, to get more attacks to the king and a lower draw rate. Because the draw rate in computerchess increases more and more, the stronger the engines and the faster the hardware gets. For my Stockfish bullet-testruns, I use 500 SALC-positions since 2014, which lowered the draw rate a lot.
Now, I created another opening-book, called OLIK (Open Line In front of King). In all lines (created out of 10000 human games, all lines 24 plies deep (all lines checked with Komodo 10.2 (20'' per position, running on 3 cores), evaluation inside of [-0.6,+0.6] and outside of [-0.15,+0.15])) , there is no pawn of the own color in front of one king and both queens are still on the board. Example: White played 0-0, then ((no white pawn on g-line) or (no white pawn on h-line) or (no pawn on f-line)). The idea is the same as for the SALC-book: more attacks to the king and a lower draw rate...
For more information about the two books, how to use them and their creation, check out the readme-files and the booksetting.jpg-pictures in the book-folders in the download-package. In the download, you find a third book (SALC+OLIK), which is the SALC and the OLIK book melted in one bigger book. For all three books (created for FritzGUI and ShredderGUI), there is a pgn-file, with the games, the book was made of. And an epd-file with the positions at the end of the opening-lines.

To verify, how much the draw rate is lowered by these new books / opening-positions sets, I did three testruns. 3000 games each (=9000 games). Stockfish 8 in selfplay. 70''+700ms thinkingtime, singlecore, LittleBlitzerGUI (using the 10000 positions epd-files, playing in RoundRobin-mode, in which for each game one epd-position is chosen per random).
Test 1: 34700 standard 8-move opening epd. Draw rate: 83.0%
Test 2: 10000 OLIK epd. Draw rate: 71.9%
Test 3: 10000 SALC V2 epd. Draw rate: 68.2%

I think, these results are really impressive...Or download the games of my long thinking-time tournament, in which the SALC V2 book is used...See, how spectacular modern high-level computerchess still can be !

You can download the SALC V2-book, the OLIK-book and the SALC+OLIK-book with pgn- and epd- files in one download on my website (download & links-section).

http://spcc.beepworld.de/downloads--links.htm

Enjoy! Stefan (SPCC)

[Lyudmil Tsvetkov]

Some weeks ago, I created my SALC opening book for engine-engine matches. In all lines (created out of 10000 human-games, all lines 20 plies deep (all lines checked with Komodo 10.2 (20'' per position, running on 3 cores), evaluation inside of [-0.6,+0.6])), white and black castled to opposite sides, both queens still on the board. The idea is, to get more attacks to the king and a lower draw rate. Because the draw rate in computerchess increases more and more, the stronger the engines and the faster the hardware gets. For my Stockfish bullet-testruns, I use 500 SALC-positions since 2014, which lowered the draw rate a lot.
Now, I created another opening-book, called OLIK (Open Line In front of King). In all lines (created out of 10000 human games, all lines 24 plies deep (all lines checked with Komodo 10.2 (20'' per position, running on 3 cores), evaluation inside of [-0.6,+0.6] and outside of [-0.15,+0.15])) , there is no pawn of the own color in front of one king and both queens are still on the board. Example: White played 0-0, then ((no white pawn on g-line) or (no white pawn on h-line) or (no pawn on f-line)). The idea is the same as for the SALC-book: more attacks to the king and a lower draw rate...
For more information about the two books, how to use them and their creation, check out the readme-files and the booksetting.jpg-pictures in the book-folders in the download-package. In the download, you find a third book (SALC+OLIK), which is the SALC and the OLIK book melted in one bigger book. For all three books (created for FritzGUI and ShredderGUI), there is a pgn-file, with the games, the book was made of. And an epd-file with the positions at the end of the opening-lines.

To verify, how much the draw rate is lowered by these new books / opening-positions sets, I did three testruns. 3000 games each (=9000 games). Stockfish 8 in selfplay. 70''+700ms thinkingtime, singlecore, LittleBlitzerGUI (using the 10000 positions epd-files, playing in RoundRobin-mode, in which for each game one epd-position is chosen per random).
Test 1: 34700 standard 8-move opening epd. Draw rate: 83.0%
Test 2: 10000 OLIK epd. Draw rate: 71.9%
Test 3: 10000 SALC V2 epd. Draw rate: 68.2%

I think, these results are really impressive...Or download the games of my long thinking-time tournament, in which the SALC V2 book is used...See, how spectacular modern high-level computerchess still can be !

You can download the SALC V2-book, the OLIK-book and the SALC+OLIK-book with pgn- and epd- files in one download on my website (download & links-section).

http://spcc.beepworld.de/downloads--links.htm

Enjoy! Stefan (SPCC)

one thing I do not understand is why you need lower draw rate?
if the engines are of approximately the same strength, you can expect high draw rate.
if one engine is 300-500 elo stronger than another, draw rate will be almost 0.

those are natural things, and I do not see why they should be changed.

of course, opposite castlings are very interesting, and if you want to provide for spectacular games, that is fine, but testing only with opposite castlings is, to say the least, very unrepresentative of chess.

anyway, very well, if you enjoy it, I myself prefer testing with as wider opening variety as possible, no to skew results in any possible way.

[Joerg Oster]

Thank you very much, Stefan!

I will add the openings to my opening set.
It's also great that you not only provide opening books for the GUIs but PGN and EPD sets, as well. Really appreciated!

[pohl4711]

Thank you very much, Stefan!

I will add the openings to my opening set.
It's also great that you not only provide opening books for the GUIs but PGN and EPD sets, as well. Really appreciated!

My pleasure.

The EPD sets are needed, when using the LittleBlitzerGUI. Because of the en-passant-bug of the LBG...(if an en-passant move is in the PGN-opening-set, the captured pawn is not deleted from the chessboard). If EPDs are used, this problem is gone...