R7 2700X 8ths x 2
Code: Select all
RR Elo +/- Games Points Score Draw
1 SugaR_AI_2.40-NNUEonly * 21 28 200 106.0 53.0% 66.0%
2 Stockfish_21091218 19 28 200 105.5 52.8% 66.5%
3 ShashChess18.1-TCP * -40 30 200 88.5 44.3% 60.5%
300 games. * with the net prior to Stockfish's current.
TCP = all 3 settings checked, sorry it did not seem to work out here.
Su Sf SC
Su ---- 49.0 57.0
Sf 51.0 ---- 54.5
SC 43.0 45.5 ----game #292, Round 146, SugaR 1-0 ShashChess , a 'Danish' vs Sicilian, the shortest win.
After 15...Bc6?? {Swordfish 15...Nf4! =} 16.Rxd5!! {exchange sac.} Bxd5 16.Nf6+!! {N sac.} gxf6 17.Nf5!! {N offer} white has a winning attack.
game #291, Round 146, ShashChess 0-1 SugaR , colors reversed: black eventually won with the gambit matterial.
_ _ _
game #62, Round 31, SugaR 1-0 Stockfish, Lolli Gambit, KGA, white sacs. a B and a N in the book.
After 5.Nd5 Nxd5?? +- {Swordfish 5...Nbc6! =} Stockfish said it was winning but its lost*: 6.Qxd5+ Qe6 7.Qg5 Qxh6 {black 3 minors up!?} 8.Qd8+ Bf8 9.Qe8! Qg7 10.Rf3! h6 11.Raf1! Kh7 12.Rxf8. *Apparently a flaw in Stockfish, SugaR's and Swordfish's evals. were correct.
game #61, Round 31, Stockfish 0-1 SugaR, colors reversed Stockfish could not find compensation for the strange opening.
_ _ _
game #283, Round 142, Stockfish 1-0 SugaR , Bishop Game with ...f5.
After 8...c5 black has made an unsound N sac., refutted by a Q sac.: 9.Qx(N)e5 Qxe5 10.Bxg6+ Kd7 11.Nf7 Qf6 12.Nxh8 and white had more than enough for its Q and went on to win.
game #282, Round 142, SugaR 1/2=1/2 Stockfish , colors rev. SugaR was only able to draw.
Games: https://www.mediafire.com/file/a1xk5grj ... u.pgn/file
