New engine releases & news H2 2024

[Redux]

Hi guys! I'd like to announce here the release of my chess engine. It's called Donut.

First of all I NEED to thank Maksim Korzh for his YouTube video series, he did help me a lot and I learned a lot of things from his videos. My chess engine has a backbone of BBC (Maksim Korzh's engine) with some modifications (I've changed the NNUE to a newer version, including new probing functions for example).

My objective would be to 'register' my engine to participate in a CCRL tournament or similar in order to give it an estimated ELO, because I currently have no idea how strong it could be evaluated. Hopefully it will be possible!

If you guys have some advice, I'd be glad to read them. Thank you!

https://github.com/th3-riddler/donut

Is it able to be used without the SF net?

Unfortunately that's not possible, it's a problem, isn't it?

I can't speak for others, but I prefer it when engine authors train their own net rather than borrow one from elsewhere.

I understand. Is there a chance that I can make it in a CCRL tournament anyway?

[Graham Banks]

Hi guys! I'd like to announce here the release of my chess engine. It's called Donut.

First of all I NEED to thank Maksim Korzh for his YouTube video series, he did help me a lot and I learned a lot of things from his videos. My chess engine has a backbone of BBC (Maksim Korzh's engine) with some modifications (I've changed the NNUE to a newer version, including new probing functions for example).

My objective would be to 'register' my engine to participate in a CCRL tournament or similar in order to give it an estimated ELO, because I currently have no idea how strong it could be evaluated. Hopefully it will be possible!

If you guys have some advice, I'd be glad to read them. Thank you!

https://github.com/th3-riddler/donut

Is it able to be used without the SF net?

Unfortunately that's not possible, it's a problem, isn't it?

I can't speak for others, but I prefer it when engine authors train their own net rather than borrow one from elsewhere.

I understand. Is there a chance that I can make it in a CCRL tournament anyway?

Probably, but the games would not be rated.

[JoAnnP38]

New Pedantic 2.1.0 is available.

• +119.0 +/- 6.9 Elo over Pedantic 2.0.0 in self-play at 20+0.2 time controls.
• +92.0 +/- 10.2 Elo over Pedantic 2.0.0 in self-play at 2:00+1 time controls.
• +61.8 +/- 14.5 Elo over Pedantic 2.0.0 in self-play at 12:00+8 time controls.

[eduherminio]

Lynx 1.7.0
+175.7 (+/- 7.0) vs 1.6.0
Estimated 3101 CCRL blitz

[Guenther]

Frolic (XB+UCI) for chess was released two months ago (still under the name SSGC13 Chess Engine at that time - meanwhile changed to Frolic)
(initially started as a plain shatranj chess engine)
author: Chris Bao

https://github.com/sscg13/sscg13-chess- ... s/tag/nnue

Code: Select all

uci
id name sscg13 chess engine
id author sscg13
option name EvalFile type string default <empty>
uciok
isready
readyok
position startpos
go wtime 120000 btime 120000 winc 2000 binc 2000
info depth 1 nodes 20 time 0 score cp 25 pv
info depth 2 nodes 175 time 0 score cp 11 pv g1f3 g8f6
info depth 3 nodes 512 time 0 score cp 24 pv g1f3 g8f6 d2d4
info depth 4 nodes 1261 time 0 score cp 11 pv g1f3 g8f6 d2d4 d7d5
info depth 5 nodes 2855 time 1 score cp 22 pv g1f3 g8f6 d2d4 d7d5 b1c3
info depth 6 nodes 6856 time 2 score cp 11 pv g1f3 g8f6 d2d4 d7d5 b1c3 b8c6
info depth 7 nodes 13361 time 3 score cp 17 pv g1f3 g8f6 d2d4 d7d5 b1c3 b8c6 c1e3
info depth 8 nodes 36261 time 7 score cp 11 pv g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 c1e3 c8e6
info depth 9 nodes 129147 time 21 score cp 9 pv g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 c1e3 c8e6 a1c1
info depth 10 nodes 188055 time 30 score cp 11 pv g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 e2e3 e7e6 f1d3 f8d6
info depth 11 nodes 386015 time 61 score cp 13 pv d2d4 d7d5 b1c3 b8c6 e2e4 d5e4 d4d5 c6e5 c3e4 g8f6 c1f4
info depth 12 nodes 853792 time 130 score cp 1 pv d2d4 g8f6 b1c3 e7e6 e2e4 f8b4 f1d3 d7d5 e4d5 f6d5 g1e2 e8g8
info depth 13 nodes 1910554 time 284 score cp -7 pv d2d4 g8f6 b1c3 d7d5 c1g5 b8c6 g1f3 h7h6 g5f4 e7e6 a1c1 f8e7 c3b5
info depth 14 nodes 4537824 time 658 score cp 5 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8f6 b1c3 f8e7 e4e5 c7c5
info depth 15 nodes 10250620 time 1465 score cp -7 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8f6 b1c3 f8e7 e4e5 c7c5 d4d1
info depth 16 nodes 17224534 time 2451 score cp 4 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8e7 c1g5 e7c6 d4d2 f8e7 g5e7 d8e7
info depth 17 nodes 31761418 time 4493 score cp 9 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 f8c5 d4c6 b7c6 b1c3 g8e7 f1d3 e8g8
info nodes 31761418 nps 7067000
bestmove e2e4

Code: Select all

xboard
protover
feature ping=1 setboard=1 analyze=0 sigint=0 sigterm=0 myname="sscg13 engine" variants="normal" done=1
new
time 12000
go
1 25 0 20
2 11 0 175 g1f3 g8f6
3 24 0 512 g1f3 g8f6 d2d4
4 11 0 1261 g1f3 g8f6 d2d4 d7d5
5 22 0 2855 g1f3 g8f6 d2d4 d7d5 b1c3
6 11 0 6856 g1f3 g8f6 d2d4 d7d5 b1c3 b8c6
7 17 0 13361 g1f3 g8f6 d2d4 d7d5 b1c3 b8c6 c1e3
8 11 0 36261 g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 c1e3 c8e6
9 9 2 129147 g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 c1e3 c8e6 a1c1
10 11 3 188055 g1f3 g8f6 b1c3 b8c6 d2d4 d7d5 e2e3 e7e6 f1d3 f8d6
11 13 6 386015 d2d4 d7d5 b1c3 b8c6 e2e4 d5e4 d4d5 c6e5 c3e4 g8f6 c1f4
12 1 13 853792 d2d4 g8f6 b1c3 e7e6 e2e4 f8b4 f1d3 d7d5 e4d5 f6d5 g1e2 e8g8
13 -7 28 1910554 d2d4 g8f6 b1c3 d7d5 c1g5 b8c6 g1f3 h7h6 g5f4 e7e6 a1c1 f8e7 c3b5
14 5 65 4537824 e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8f6 b1c3 f8e7 e4e5 c7c5
15 -7 146 10250620 e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8f6 b1c3 f8e7 e4e5 c7c5 d4d1
16 4 245 17224534 e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 c6d4 d1d4 g8e7 c1g5 e7c6 d4d2 f8e7 g5e7 d8e7
17 9 449 31761418 e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 f8c5 d4c6 b7c6 b1c3 g8e7 f1d3 e8g8
move e2e4

In a quick test > 2500 rating points at CCRL scale
tc 0:10+0.2 start positions: UHO_2024_6mvs_+100_+109 concurrency: 4 (0 time losses etc.)

Code: Select all

   # PLAYER               :   RATING  ERROR  POINTS  PLAYED    (%)
   1 Frolic_Current-64    :  2577.34  36.72   184.5     298  61.91
   2 Tcheran_30-64        :  2485.00   ----   113.5     298  38.09

White advantage = 101.72 +/- 17.98
Draw rate (equal opponents) = 29.30 % +/- 2.96

Code: Select all

Head to head statistics:

1) Frolic_Current-64 2577.34 :    298 (+145,=79,-74),  61.9 %

   vs.                       :  games (   +,  =,  -),   (%) :     Diff,     SD, CFS (%)
   Tcheran_30-64             :    298 ( 145, 79, 74),  61.9 :   +92.34,  18.73,  100.0

2) Tcheran_30-64     2485.00 :    298 (+74,=79,-145),  38.1 %

   vs.                       :  games (  +,  =,   -),   (%) :     Diff,     SD, CFS (%)
   Frolic_Current-64         :    298 ( 74, 79, 145),  38.1 :   -92.34,  18.73,    0.0

[op12no2]

Lozza 3 - https://github.com/op12no2/lozza/releases

[Andres Valverde]

New EveAnn 2.0 released, details at :

https://sites.google.com/view/dirtychess