Since you’re here...

We hope you will consider supporting us today. We need your support to continue to exist, because good entries are more and more work time. Every reader contribution, however big or small, is so valuable. Support "Chess Engines Diary" even a small amount– and it only takes a minute. Thank you.
============================== My email: jotes@go2.pl



Chess engine: Nectar 0.0.5


Nectar - UCI chess engine
Author: Dragjon

The engine was created recently and has not played enough games yet to be on the CEDR ranking list. Tests of this version soon.

v 0.0.5:
Changes from V0.0.1

2x More data (0.0.5)
Trained with AdamW optimiser (0.0.5)
Tweaked Futility pruning (0.0.4)
Changed 2 NNs to 1 NN (for both colours) (0.0.4)
Tuned with weather factory (0.0.3)
Changed some implementations (0.0.3)
SCReLU nets (0.0.2)
SPRT STC vs V0.0.4

Score of NectarV0.0.5 vs NectarV0.0.4: 430 - 334 - 189  [0.550] 953
...      NectarV0.0.5 playing White: 220 - 162 - 96  [0.561] 478
...      NectarV0.0.5 playing Black: 210 - 172 - 93  [0.540] 475
...      White vs Black: 392 - 372 - 189  [0.510] 953
Elo difference: 35.1 +/- 19.8, LOS: 100.0 %, DrawRatio: 19.8 %
SPRT: llr 2.96 (100.6%), lbound -2.94, ubound 2.94 - H1 was accepted
STC vs V0.0.1

Score of NectarV0.0.5 vs NectarV0.0.1: 59 - 29 - 22  [0.636] 110
...      NectarV0.0.5 playing White: 35 - 9 - 11  [0.736] 55
...      NectarV0.0.5 playing Black: 24 - 20 - 11  [0.536] 55
...      White vs Black: 55 - 33 - 22  [0.600] 110
Elo difference: 97.2 +/- 60.4, LOS: 99.9 %, DrawRatio: 20.0 %
SPRT: llr 0 (0.0%), lbound -inf, ubound inf
Elo Estimate

Rank Name                          Elo     +/-   Games   Score    Draw
   1 snowy-v0.2                    191     103      64   75.0%    0.0%
   2 CDrill_1800_Build_4           138      93      61   68.9%    6.6%
   3 Napolean-v1.4.0                81      91      61   61.5%    1.6%
   4 shallowblue-v2.0.0           -191     101      62   25.0%    4.8%
   5 NectarV0.0.5                 -248     111      62   19.4%    6.5%

SPRT: llr 0 (0.0%), lbound -inf, ubound inf
Approx: 1560 - 1722


Comments