Another tablebase test with SF14

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

Jouni
Posts: 3770
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Another tablebase test with SF14

Post by Jouni »

5+6 piece tablebases on SSD, level 10+0,1, book 8 moves

Code: Select all

Score of stockfish14 TB vs stockfish14: 163 - 168 - 1669 [0.499]
...      stockfish14 TB playing White: 129 - 44 - 827  [0.542] 1000
...      stockfish14 TB playing Black: 34 - 124 - 842  [0.455] 1000
...      White vs Black: 253 - 78 - 1669  [0.544] 2000
Elo difference: -0.9 +/- 6.2, LOS: 39.2 %, DrawRatio: 83.5 %
2000 of 2000 games finished.

Code: Select all

Score of stockfish14 classic TB vs stockfish14 classic: 311 - 219 - 1470 [0.523]
...      stockfish14 classic TB playing White: 201 - 68 - 731  [0.567] 1000
...      stockfish14 classic TB playing Black: 110 - 151 - 739  [0.479] 1000
...      White vs Black: 352 - 178 - 1470  [0.543] 2000
Elo difference: 16.0 +/- 7.8, LOS: 100.0 %, DrawRatio: 73.5 %
2000 of 2000 games finished.
Definitely syzygy bases are not important for NNUE Stockfish!
Jouni
User avatar
CMCanavessi
Posts: 1142
Joined: Thu Dec 28, 2017 4:06 pm
Location: Argentina

Re: Another tablebase test with SF14

Post by CMCanavessi »

I wonder how TC may affect that.... with very short TC, TB hits start very late in the game, while the longer the TC, the earlier you start hitting them, and the effect may be bigger.
Follow my tournament and some Leela gauntlets live at http://twitch.tv/ccls