New engine releases & news H2 2024

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

Moderators: hgm, chrisw, Rebel

User avatar
Sylwy
Posts: 4656
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: New engine releases & news H2 2024

Post by Sylwy »

kelseyde123 wrote: Wed Nov 13, 2024 7:41 pm Hi Sylwy/Alex,

Good catch, that setting should be false when the engine is being used by a GUI... I think I will implement what I have seen quite a few other engines do, which is to turn it off as soon as the engine receives a 'uci' command (since that is always sent by GUIs, but rarely by humans).

However I am curious that you observed the engine playing weaker after disabling the Pretty option... I'm not sure how that can be, since it always outputs the same 'bestmove' regardless of whether Pretty is on or off. I will run some extra tests to confirm this.

I'll let you know when the new build is ready to download. And sorry for the inconvenience!

Dan
The intermediate result below is made with the default settings (pretty=false, but actually what the engine indicates is false because pretty=true - here is one of the problems). Version 4.3.0 behaves much better against Bagatur 5.1c (#2 in my Java list).

Image
Carbec
Posts: 153
Joined: Thu Jan 20, 2022 9:42 am
Location: France
Full name: Philippe Chevalier

Re: New engine releases & news H2 2024

Post by Carbec »

Zangdar 2.31

This version is a little stronger (50 points).
I think this will be the last HCE version. I plan to investigate the NNUE algorithm.

https://github.com/Carbecq/Zangdar/releases/tag/v2.31
kelseyde123
Posts: 17
Joined: Fri Oct 06, 2023 1:10 am
Full name: Dan Kelsey

Re: New engine releases & news H2 2024

Post by kelseyde123 »

Sylwy wrote: Wed Nov 13, 2024 8:46 pm The intermediate result below is made with the default settings (pretty=false, but actually what the engine indicates is false because pretty=true - here is one of the problems). Version 4.3.0 behaves much better against Bagatur 5.1c (#2 in my Java list).

Image
Hello Sylwy, I released a new version of Calvin addressing the problems you noticed with the previous release: https://github.com/kelseyde/calvin-ches ... /tag/5.0.2

Pretty print is now disabled as soon as the engine receives a 'uci' command, so you shouldn't need to explicitly configure it when running your tournaments.

Thanks for pointing that one out!
User avatar
Sylwy
Posts: 4656
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: New engine releases & news H2 2024

Post by Sylwy »

kelseyde123 wrote: Sat Nov 16, 2024 10:10 am
Sylwy wrote: Wed Nov 13, 2024 8:46 pm The intermediate result below is made with the default settings (pretty=false, but actually what the engine indicates is false because pretty=true - here is one of the problems). Version 4.3.0 behaves much better against Bagatur 5.1c (#2 in my Java list).

Image
Hello Sylwy, I released a new version of Calvin addressing the problems you noticed with the previous release: https://github.com/kelseyde/calvin-ches ... /tag/5.0.2

Pretty print is now disabled as soon as the engine receives a 'uci' command, so you shouldn't need to explicitly configure it when running your tournaments.

Thanks for pointing that one out!
THANK YOU ! I like a lot your engine ! The test versus Serendipyty 0.4.1 x64 is on way. Works OK !

Image
User avatar
Werner
Posts: 2921
Joined: Wed Mar 08, 2006 10:09 pm
Location: Germany
Full name: Werner Schüle

Re: New engine releases & news H2 2024

Post by Werner »

PlentyChess 3.0.0
with 100% self-generated data!

+64 STC, +55 LTC, +63 DFRC (against version 2.0.0)
<https://github.com/Yoshie2000/PlentyChe ... g/b-v3.0.0>
Werner
Wolfgang
Posts: 949
Joined: Sat May 13, 2006 1:08 am

Re: New engine releases & news H2 2024

Post by Wolfgang »

In the meantime there is version 3.0.1 with Bugfix for Fritz/Chessbase

Results are quite disappointing so far

Test is made with version 3.0.0 as I do not use any kind of Chessbase GUI 😄
Best
Wolfgang
CEGT-Team
www.cegt.net
www.cegt.forumieren.com
User avatar
Dariusz
Posts: 371
Joined: Sat Jun 13, 2015 10:08 am
Location: Poland
Full name: Dariusz Domagała

Re: New engine releases & news H2 2024

Post by Dariusz »

Wolfgang wrote: Fri Nov 22, 2024 6:42 pm In the meantime there is version 3.0.1 with Bugfix for Fritz/Chessbase

Results are quite disappointing so far

Test is made with version 3.0.0 as I do not use any kind of Chessbase GUI 😄

In MCERL the same, unfortunately :-(

# PLAYER : RATING POINTS PLAYED (%)
69 PlentyChess 2.1.0 : 3706 1323.0 1796 73.7%

70 Caissa 1.20 : 3704 2015.5 2704 74.5%
71 Caissa 1.21 : 3704 2100.0 2554 82.2%
72 Dragon 3.2 by Komodo Chess : 3701 2365.0 3945 59.9%
73 Berserk 11 : 3698 2538.0 4345 58.4%
74 Clover 7.0 : 3687 1972.5 2705 72.9%
75 Berserk 9 : 3686 1647.5 2177 75.7%
76 Alexandria 6.0.1 : 3684 1848.5 2464 75.0%
77 Caissa 1.17 : 3681 2003.0 2660 75.3%
78 Clover 6.1 : 3678 1623.5 2219 73.2%
79 RubiChess 20231202 : 3677 1266.5 1586 79.9%
80 Dragon 3.1 by Komodo Chess : 3676 1305.5 1791 72.9%
81 Dragon 2.6.1 by Komodo Chess : 3676 2238.0 3653 61.3%
82 Dragon 3.3 by Komodo Chess : 3673 1919.0 3071 62.5%
83 Clover 6.0.29 (12.2023) : 3672 1336.0 1723 77.5%
84 RubiChess 20230813 : 3671 1500.5 2490 60.3%
85 Berserk 11.1 : 3667 953.5 1559 61.2%
86 Koivisto 9.0 : 3666 2067.5 3587 57.6%
87 Caissa 1.16 : 3665 1812.0 2505 72.3%
88 Seer 2.8.0 : 3665 2136.5 2883 74.1%
89 Vafra Cfish 12.6 (Stockfish) : 3664 1286.5 1664 77.3%
90 Stockfish 13 : 3664 1307.0 2190 59.7%
91 Starzix 6.0 : 3660 1858.0 2392 77.7%
92 Berserk 10 : 3656 1832.5 3154 58.1%

93 PlentyChess 3.0.1 : 3656 1846.5 2360 78.2%

I noticed that version 3.0.1 (the same as version 3.0.0 in terms of strength) is slower than 2.1.0.

2.1.0 bench:
Total time (ms) : 2150
Nodes searched : 2477083
Nodes/second : 1152131

3.0.1 bench:
Total time (ms) : 4570
Nodes searched : 1886365
Nodes/second : 412771
Regards, Darius
https://chessengeria.eu
User avatar
Dariusz
Posts: 371
Joined: Sat Jun 13, 2015 10:08 am
Location: Poland
Full name: Dariusz Domagała

Re: New engine releases & news H2 2024

Post by Dariusz »

Dariusz wrote: Fri Nov 22, 2024 9:46 pm
Wolfgang wrote: Fri Nov 22, 2024 6:42 pm In the meantime there is version 3.0.1 with Bugfix for Fritz/Chessbase

Results are quite disappointing so far

Test is made with version 3.0.0 as I do not use any kind of Chessbase GUI 😄

In MCERL the same, unfortunately :-(

# PLAYER : RATING POINTS PLAYED (%)
69 PlentyChess 2.1.0 : 3706 1323.0 1796 73.7%

70 Caissa 1.20 : 3704 2015.5 2704 74.5%
71 Caissa 1.21 : 3704 2100.0 2554 82.2%
72 Dragon 3.2 by Komodo Chess : 3701 2365.0 3945 59.9%
73 Berserk 11 : 3698 2538.0 4345 58.4%
74 Clover 7.0 : 3687 1972.5 2705 72.9%
75 Berserk 9 : 3686 1647.5 2177 75.7%
76 Alexandria 6.0.1 : 3684 1848.5 2464 75.0%
77 Caissa 1.17 : 3681 2003.0 2660 75.3%
78 Clover 6.1 : 3678 1623.5 2219 73.2%
79 RubiChess 20231202 : 3677 1266.5 1586 79.9%
80 Dragon 3.1 by Komodo Chess : 3676 1305.5 1791 72.9%
81 Dragon 2.6.1 by Komodo Chess : 3676 2238.0 3653 61.3%
82 Dragon 3.3 by Komodo Chess : 3673 1919.0 3071 62.5%
83 Clover 6.0.29 (12.2023) : 3672 1336.0 1723 77.5%
84 RubiChess 20230813 : 3671 1500.5 2490 60.3%
85 Berserk 11.1 : 3667 953.5 1559 61.2%
86 Koivisto 9.0 : 3666 2067.5 3587 57.6%
87 Caissa 1.16 : 3665 1812.0 2505 72.3%
88 Seer 2.8.0 : 3665 2136.5 2883 74.1%
89 Vafra Cfish 12.6 (Stockfish) : 3664 1286.5 1664 77.3%
90 Stockfish 13 : 3664 1307.0 2190 59.7%
91 Starzix 6.0 : 3660 1858.0 2392 77.7%
92 Berserk 10 : 3656 1832.5 3154 58.1%

93 PlentyChess 3.0.1 : 3656 1846.5 2360 78.2%

I noticed that version 3.0.1 (the same as version 3.0.0 in terms of strength) is slower than 2.1.0.

2.1.0 bench:
Total time (ms) : 2150
Nodes searched : 2477083
Nodes/second : 1152131

3.0.1 bench:
Total time (ms) : 4570
Nodes searched : 1886365
Nodes/second : 412771
The benchmark results were created on a MacBook Pro M1.
I wrote to the author of PlentyChess to ask him to check what happened.
Regards, Darius
https://chessengeria.eu
JohnWoe
Posts: 517
Joined: Sat Mar 02, 2013 11:31 pm

Re: New engine releases & news H2 2024

Post by JohnWoe »

Released A Seirawan chess engine. MayhemSeirawan 0.1!
1000s of games w/o issues on cutechess-cli. Works great on Xboard.
Only seirawan chess variant supported.
Use basetime + inc. No cyclical tc bs.

Bench: 242391492

https://github.com/SamuraiDangyo/MayhemSeirawan

Sample game:

Code: Select all

[White "MayhemSeirawan 0.1"]
[Black "Fairy-Max 5.0b"]
[Result "1-0"]
[TimeControl "60+1"]
[Variant "seirawan"]
1. e4 {+0,77/8} Nh6/H {+0,08/6 1,3} 2. Nc3/E {+0,47/9 3} Nc6/E
{+0,12/7 2,7} 3. d4 {+0,99/12 3} d6 {-0,13/8 4} 4. Bg5/H {+1,25/13 3} f6
{+0,07/7 1,9} 5. Bxh6 {+1,62/16 3} Hxh6 {-0,17/10 9} 6. Hxh6 {+1,76/15 2,9}
gxh6 {-0,13/10 2,8} 7. Qh5+ {+1,97/13 2,8} Kd7 {-0,37/9 2,4} 8. Nf3
{+1,95/13 2,7} Qe8 {-0,13/9 3} 9. Qh4 {+1,62/14 2,7} Qg6 {+0,24/8 2,7} 10.
Nd5 {+1,93/14 2,6} f5 {+0,33/7 1,3} 11. exf5 {+2,42/14 2,5} Qxf5
{+0,12/8 1,5} 12. Ec3 {+2,47/13 2,5} Rg8 {+0,10/7 1,0} 13. Ee3
{+3,58/12 2,4} Qf7 {-0,16/8 1,0} 14. Bc4 {+4,08/12 2,4} Qg6 {-0,50/8 1,5}
15. Nf4 {+4,59/12 2,3} Qg7 {-1,60/8 1,2} 16. Be6+ {+4,53/12 2,3} Kd8
{-1,90/9 1,5} 17. Bxg8 {+4,61/14 2,2} Qxg8 {-1,76/9 1,3} 18. d5
{+4,45/14 2,2} Nb4 {-1,99/9 2,2} 19. O-O-O {+4,73/13 2,1} Nxa2+
{-0,96/8 1,7} 20. Kb1 {+4,92/15 2,1} Ea6 {-1,28/9 1,2} 21. Rhe1
{+4,99/14 2,0} Qf7 {-1,14/8 1,6} 22. Ne6+ {+5,14/13 2,0} Bxe6 {-1,39/9 2,0}
23. dxe6 {+4,98/13 1,9} Qg8 {-1,85/8 1,0} 24. Ne5 {+4,99/12 1,9} Qg7
{-2,23/8 2,0} 25. Ef5 {+6,36/13 1,9} Eb4 {-3,05/8 4} 26. Qg3 {+7,61/13 1,8}
Kc8 {-3,06/8 1,1} 27. Qxg7 {+8,98/15 1,8} Nc3+ {-7,72/11 1,7} 28. Kc1
{+12,84/17 1,8} Na2+ {-10,44/11 2,2} 29. Kd2 {+13,85/16 1,7} Ed5+
{-10,30/11 1,5} 30. Ke2 {+14,74/17 1,7} Bxg7 {-10,94/11 1,0} 31. Rxd5
{+15,22/16 1,7} Bf6 {-14,90/11 2,2} 32. Exf6 {+15,81/15 1,7} exf6
{-1000,07/12 1,0} 33. e7 {+104,85/14 0,3} Nc3+ {-1000,06/13 1,3} 34. Kf3
{+104,85/9} c5 {-1000,05/15 1,2} 35. e8=E# {+104,85/2}
{Xboard adjudication: Checkmate} 1-0
JohnWoe
Posts: 517
Joined: Sat Mar 02, 2013 11:31 pm

Re: New engine releases & news H2 2024

Post by JohnWoe »

MayhemSeirawan 0.2 released: https://github.com/SamuraiDangyo/MayhemSeirawan

+2 more variant engines:

3 and 5 checks engine: Mayhem3Checks: https://github.com/SamuraiDangyo/Mayhem3Checks
Sample game:

Code: Select all

[White "Mayhem3Checks 0.1"]
[Black "Mayhem3Checks 0.1"]
[Result "1-0"]
[TimeControl "60+1"]
[Variant "3check"]
1. e4 {+0,25/16} Nc6 {-0,47/16 3} 2. Nc3 {+0,27/16 3} e6 {-0,43/16 3} 3.
Nf3 {+0,59/17 3} Nf6 {-0,46/16 3} 4. a3 {+0,49/16 3} Be7 {-0,48/15 3} 5. d4
{+0,53/16 3} Ng4 {-0,48/15 3} 6. Bf4 {+0,53/14 2,9} O-O {-0,50/14 2,9} 7.
Qd2 {+0,83/15 2,8} d5 {-0,73/15 2,8} 8. e5 {+0,67/15 2,7} f6 {-0,66/14 2,7}
9. exf6 {+0,76/15 2,7} Bxf6 {-0,75/14 2,7} 10. O-O-O {+1,05/14 2,6} a6
{-0,82/14 2,6} 11. h3 {+1,63/15 2,5} Nh6 {-1,37/15 2,5} 12. Bd3
{+1,41/15 2,5} Kh8 {-1,42/16 2,5} 13. Bxh6 {+1,43/15 2,4} gxh6
{-1,32/15 2,4} 14. Qxh6 {+1,12/14 2,4} Rf7 {-1,22/15 2,4} 15. Rhe1
{+1,50/13 2,3} Bd7 {-1,60/12 2,3} 16. Kb1 {+2,01/13 2,3} Rg7 {-2,36/13 2,3}
17. Nxd5 {+4,48/15 2,2} Bg5 {-4,13/15 2,2} 18. Qxg7+ {+4,16/14 2,2} Kxg7
{-4,13/1 0,1} 19. Nxg5 {+4,59/15 2,1} Qxg5 {-4,60/15 2,2} 20. Nxc7
{+4,57/15 2,1} Kh8 {-4,83/15 2,1} 21. Nxa8 {+4,95/15 2,0} Qd8
{-5,09/15 2,1} 22. Bxh7 {+5,50/15 2,0} Qxa8 {-5,73/16 2,1} 23. d5
{+6,17/17 1,9} Ne7 {-6,30/17 2,0} 24. Re5 {+9,55/16 1,9} Be8 {-9,99/16 2,0}
25. Rd4 {+104,85/13 0,5} Qa7 {-104,85/12 0,4} 26. Rh4 {+104,85/9 0,1} Bh5
{-104,85/5 0,1} 27. Rhxh5 {+104,85/4 0,1} Qxf2 {-104,85/3 0,1} 28. Bd3+
{+104,85/2 0,1} Kg7 {-104,85/2 0,1} 29. Rh7+ {+104,85/1 0,1}
{Xboard adjudication: 3rd check} 1-0
Atomic chess engine. MayhemAtomic: https://github.com/SamuraiDangyo/MayhemAtomic

No illegal moves in cutechess-cli

Code: Select all

Finished game 1000 (MayhemAtomic 0.1 vs MayhemAtomic 0.1): 1-0 {Black's king exploded}
Score of MayhemAtomic 0.1 vs MayhemAtomic 0.1: 457 - 454 - 89  [0.501] 1000
...      MayhemAtomic 0.1 playing White: 235 - 216 - 49  [0.519] 500
...      MayhemAtomic 0.1 playing Black: 222 - 238 - 40  [0.484] 500
...      White vs Black: 473 - 438 - 89  [0.517] 1000
Elo difference: 1.0 +/- 20.5, LOS: 54.0 %, DrawRatio: 8.9 %
SPRT: llr 0 (0.0%), lbound -inf, ubound inf

Player: MayhemAtomic 0.1
   "Draw by 3-fold repetition": 76
   "Draw by fifty moves rule": 6
   "Draw by stalemate": 7
   "Loss: Black mates": 81
   "Loss: Black's king exploded": 141
   "Loss: White mates": 97
   "Loss: White's king exploded": 135
   "Win: Black mates": 64
   "Win: Black's king exploded": 124
   "Win: White mates": 111
   "Win: White's king exploded": 158
Player: MayhemAtomic 0.1
   "Draw by 3-fold repetition": 76
   "Draw by fifty moves rule": 6
   "Draw by stalemate": 7
   "Loss: Black mates": 64
   "Loss: Black's king exploded": 124
   "Loss: White mates": 111
   "Loss: White's king exploded": 158
   "Win: Black mates": 81
   "Win: Black's king exploded": 141
   "Win: White mates": 97
   "Win: White's king exploded": 135
Finished match
Shuffle game by MayhemAtomic:

Code: Select all

[White "MayhemAtomic 0.1"]
[Black "MayhemAtomic 0.1"]
[Result "1-0"]
[TimeControl "60+1"]
[Variant "atomic"]
[FEN "rnbbknqr/pppppppp/8/8/8/8/PPPPPPPP/RNBBKNQR w KQkq - 0 1"]
1. Ng3 {+0,60/15} d6 {-0,54/16 3} 2. Nc3 {+0,69/16 3} c6 {-0,58/16 3} 3.
Nf5 {+0,67/15 3} Bxf5 {-0,45/17 3} 4. d3 {+0,52/16 3} Bb6 {-0,50/16 3} 5.
e3 {+0,46/18 3} Ng6 {-0,34/17 3} 6. Bh5 {+0,54/16 2,9} a5 {-0,65/16 2,9} 7.
Na4 {+0,68/17 2,8} Nd7 {-0,51/16 2,8} 8. Nc5 {+0,54/15 2,7} Bxc5
{-0,67/17 2,7} 9. e4 {+0,45/15 2,7} h6 {-0,61/15 2,7} 10. f4 {+0,69/16 2,6}
c5 {-0,59/16 2,6} 11. Qf2 {+0,67/16 2,5} O-O-O {-0,55/15 2,5} 12. Bg4
{+0,93/16 2,5} e6 {-0,89/18 2,5} 13. f5 {+0,86/17 2,4} Nh4 {-0,73/18 2,4}
14. fxe6 {+0,71/17 2,4} f5 {-0,65/18 2,4} 15. Qd2 {+0,71/16 2,3} Qf7
{-0,67/16 2,3} 16. b3 {+0,75/15 2,3} Qc7 {-0,83/16 2,3} 17. c4
{+0,85/15 2,2} Nxg2 {-0,80/15 2,2} 18. Bb2 {+0,87/16 2,2} d5 {-1,10/16 2,2}
19. Bxg7 {+0,90/16 2,1} fxg4 {-1,02/16 2,1} 20. O-O-O {+1,04/17 2,1} Rg8
{-1,15/16 2,1} 21. Rf1 {+1,17/17 2,0} Rg2 {-0,87/15 2,0} 22. Kb1
{+1,17/15 2,0} Rf2 {-1,20/16 2,0} 23. Rxf2 {+1,11/18 1,9} Qf7
{-1,16/18 1,9} 24. e5 {+1,58/17 1,9} Qf1 {-1,36/15 1,9} 25. Qd1
{+1,50/19 1,9} Qg2 {-1,59/18 1,9} 26. Qe2 {+1,56/19 1,8} b5 {-1,42/17 1,8}
27. Qxg2 {+1,58/18 1,8} Kd8 {-2,19/21 1,8} 28. e6 {+2,79/19 1,8} Ke7
{-2,79/20 1,8} 29. a4 {+3,17/18 1,7} h5 {-3,40/17 1,7} 30. h4
{+9,97/17 1,7} Kf8 {-10,12/16 1,7} 31. e7 {+104,85/20 0,1} Ke8
{-104,85/19 0,2} 32. cxb5 {+104,85/14} d4 {-104,85/11} 33. Kc2 {+104,85/10}
c4 {-104,85/9} 34. bxc4 {+104,85/8} Kd7 {-104,85/7} 35. e8=Q {+104,85/6}
Kc7 {-104,85/5} 36. Qc6 {+104,85/4} Kd8 {-104,85/3} 37. Qd7 {+104,85/2}
{ White Wins } 1-0