I stumbled across this while searching for something else. It is a collection of test sets from 1993. I converted them to epd (when they were created, epd was a new format Steven Edwards was trying to promote). I tried to not change anything, but there were a number of positions with both a best move and an avoid move. I converted those to avoid move. There were many interesting comment lines in the original .fin files that I had to strip out.
Here they are. http://www.mediafire.com/file/uodibzjtyw4/chesstest.ZIP
Blast from the past
Moderator: Ras
-
- Posts: 4663
- Joined: Sun Mar 12, 2006 2:40 am
- Full name: Eelco de Groot
Re: Blast from the past
Thanks for the testsets John! There is some nice material there, and also useful for learning endgames for instance. I'm not sure if Speelman's testset would not still be copyrightable as a collection, but I guess for personal use it's okay. I would like to show some results for those positions, but Rainbow Serpent went into a stall when it saw a Mate in 6 in one of the very last added positions, when it had done 100 iterations. So I had no end output. That is something to do with one of the last Stockfish bugfixes, that is not yet implemented. It's a nice endgame test! I have "Analysing The Endgame" from Jon Speelman but not his other book mentioned, and there are also positions added from Averbakh, Horowitz and some other standard books. If you test without tablebases the fundamental endings are also a good test. And that is just one testsuite
Eelco

Eelco
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan