OpenChess

Dynamic Openings Test Suite (DOTS) is a set of selected positions for running engine matches. The purpose of this set is not to measure Elo differences but to provide dynamic and interesting games.

DOTS3 has 200 positions.

https://www.mediafire.com/file/r47xjimg ... 1.zip/file

G/10'+5", R7 2700X | RTX 2060
7 random openings reversed from DOTS4, 14 interesting games!
Stockfish_21122211 9.5-4.5 Lc0.28_611040_211225: 7 - 2 - 5 [0.679] 14
Elo difference: 129.8 +/- 165.5, LOS: 95.2 %, DrawRatio: 35.7 %

Game #6, Lc0 0-1 Sf , 32 moves (rev. #5 Sf-Lc0 was a draw)
Black played 18... R(h5)xg5(B)! 19.Rxg5 f5! and is winning the exchange for pawn down even though it did not castle.
Game #13, Sf 1-0 Lc0 , 22 moves (rev. #14 Lc0-Sf was a draw)
White outplayed and then after 15...Nx(B)d7 16.Rd5 did not recapture on d7 until black's king position was completely wrecked. Games:

Here is another from my last post here.

Game #9, Sf 1-0 Lc0, 42 moves (rev. #10 Lc0 1/2-1/2 Sf)
White had already sacrificed a rook for bishop and pawn, then with 23.Rxb7! it gave up its other rook for black's other bishop, then 23...Rxb7 24.g3! and white's B-pair outplayed black's R-pair.

The tactical advantage of Stockfish over Lc0 is very big.

Don't you test Blue Marlin anymore?

dorsz wrote: ↑

Tue Dec 28, 2021 7:44 pm

The tactical advantage of Stockfish over Lc0 is very big.

Don't you test Blue Marlin anymore?

Sorry, I've been too busy, but I would like to get back to your Swordfish/Blue Marlin - very interesting ideas!

I've used DOTS3 on Lc0's discord a few times with a link to their posts here. I am thinking of a DOTS4 there and will always include the link.

Agree about tactics. Lc0 finished 3rd, about -105 to Stockfish at CCC 16 Blitz Semifinal 2|1, they say the TC is not good for it.

I enjoy the Romantic Era openings in your DOTS -unique to books lately-, but of course it has more than that also.