If you are interested....

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

Stockfish_23040114_x64_bmi2 (...the latest....)
=========================================

...................Waterloo.................. :shock:

Match conditions:

-TC=4'+2"
-Hash=256 MB
-GUI: Arena 3.5.1
-Book: SuperGM_4mvs.abk (8 plies book) (only for Stockfish_23040114_x64_bmi2)
-1 thread-CPU=Intel i5-7400-3GHz (Kaby Lake)
-TBS: 6-men Syzygy bases for both engines
-NNs: defaults (Stockfish_23040114_x64_bmi2 uses the new nn-dabb1ed23026.nnue net; Dark SisTer 4.6 bmi2 uses the SF net:
nn-52471d67216a.nnue)
-Dark SisTer 4.6 bmi2 uses a special experience file
-Ponder: off
-OS: Windows 10 Home.

The final result:

Image

Image

+0 -16 =84

+56 Elo points in 100 games for Dark SisTer 4.6 bmi2

A sample (interesting):



=============================================================
Disclaimer
==========
The architecture of a chess engine is a choice of its programmer(s). If chess engines can (or can't) use:
-endgame databases (Syzygy, Gaviota, Nalimov, various bitbases...);
-evaluation based on 1-n data accumulators (NNUEs, NNs, DCNNs) or HCE;
-experience files;
-internal or external books;
-etc, etc, etc......
does not imply unfairness in matches/tournaments. Each chess engine programmer achieves what he wants, given the skills he
has.
Top
Guenther
Posts: 4718
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: If you are interested....

Post by Guenther »

Sylwy wrote: ↑Mon Apr 03, 2023 4:38 pm Stockfish_23040114_x64_bmi2 (...the latest....)
=========================================

...................Waterloo.................. :shock:



=============================================================
Disclaimer
==========
The architecture of a chess engine is a choice of its programmer(s). If chess engines can (or can't) use:
-endgame databases (Syzygy, Gaviota, Nalimov, various bitbases...);
-evaluation based on 1-n data accumulators (NNUEs, NNs, DCNNs) or HCE;
-experience files;
-internal or external books;
-etc, etc, etc......
does not imply unfairness in matches/tournaments. Each chess engine programmer achieves what he wants, given the skills he
has.
The Darkshit thingy is completely illegal, because it is 'released' with a fake source which in reality is the source of
an old unchanged Glaurung version of 08/2009. (I downloaded and checked it...ofc the MZ exp. file is not there too)
This means this thread should be moved. (moderation informed)

The result only reflects the usage of the 'experience file' which is nothing else than a kind of book file,
no matter what your 'funny' disclaimer wants to tell us.
DS is out of 'book' (exp file) in move 11, saved time 62s already vs. SF.

[Event "NN Test340"]
[Site "ISR 3"]
[Date "2023.04.01"]
[Round "14"]
[White "Dark SisTer 4.6 bmi2"]
[Black "Stockfish_23040114_x64_bmi2"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "B01"]
[Opening "Scandinavian"]
[Time "22:35:58"]
[Variation "Portuguese, 4.Nf3 Qxd5 5.Be2 Nc6"]
[WhiteElo "2200"]
[TimeControl "240+2"]
[Termination "adjudication"]
[PlyCount "120"]
[WhiteType "program"]
[BlackType "program"]

1. e4 {+0.12/60 0} d5
2. exd5 {+0.13/41 0} Nf6
3. d4 {+0.19/41 0} Bg4
4. Nf3 {+0.21/40 0} Qxd5 {-0.44/29 19}
5. Be2 {+0.21/42 0} Nc6 {-0.40/26 5}
6. c4 {+0.17/40 0} Qf5 {-0.42/28 18}
7. O-O {+0.38/27 0} O-O-O {-0.29/26 4}
8. Be3 {+0.19/40 0} e5 {-0.36/27 6}
9. d5 {+0.18/40 0} e4 {-0.32/26 4}
10. Nd4 {+0.20/40 0} Nxd4 {-0.35/26 6}
------------------------------------------------
11. Bxd4 {+0.59/24 4} h5 {-0.32/26 4}
12. Nc3 {+0.37/32 41} c5 {-0.28/26 5}
13. Be3 {+0.38/27 7} Qe5 {-0.28/25 4}
14. h4 {+0.35/28 11} Bxe2 {-0.32/27 6}
15. Qxe2 {+0.49/26 7} Ng4 {-0.18/25 5}
16. g3 {+0.45/26 6} Nxe3 {-0.25/24 5}
17. Qxe3 {+0.38/27 8} f5 {-0.26/24 6}
18. Ne2 {+0.37/27 9} Rg8 {-0.23/22 4}
19. b4 {+0.84/24 5} b6 {+0.06/27 5}
20. b5 {+0.71/28 12} Be7 {-0.91/30 67}
21. a4 {+0.78/26 5} g5 {-0.90/24 5}
22. hxg5 {+0.78/26 6} Qd6 {-0.91/26 6}
23. g6 {+0.85/24 5} Rxg6 {-1.06/31 50}
24. a5 {+0.96/25 6} Bg5 {-0.99/22 1}
25. Qb3 {+0.92/26 6} Qb8 {-0.85/25 8}
26. axb6 {+1.00/24 5} axb6 {-0.87/21 1}
27. Ra3 {+1.15/27 17} Kd7 {-1.18/30 28}
28. Rfa1 {+1.12/24 3} Ke8 {-1.19/26 2}
29. Kf1 {+1.19/26 5} Kf7 {-1.21/24 2}
30. Ra7+ {+1.77/25 7} Kg8 {-1.15/24 0}
31. Qd1 {+1.74/24 5} Rh6 {-1.93/24 7}
32. Ra8 {+2.05/27 5} Qd6 {-1.84/19 0}
33. Rxd8+ {+2.34/25 6} Bxd8 {-2.36/23 4}
34. Ra8 {+2.58/26 5} h4 {-2.27/21 0}
35. Qc1 {+2.78/26 5} Kg7 {-2.25/23 1}
36. Ra7+ {+3.07/25 5} Kg8 {-2.80/24 2}
37. Ra8 {+3.58/31 24} Kg7 {-2.67/22 2}
38. Ra7+ {+3.59/25 3} Kg8 {-3.18/25 3}
39. Rd7 {+3.60/25 6} Qxd7 {-3.09/20 0}
40. Qxh6 {+3.76/26 8} Qg7 {-3.39/22 3}
41. Qe6+ {+3.81/24 7} Qf7 {-2.29/16 0}
42. Qc6 {+3.83/25 6} Bc7 {-3.14/22 3}
43. gxh4 {+3.91/24 5} Kg7 {-2.96/18 0}
44. d6 {+4.50/23 4} Bd8 {-2.88/20 1}
45. h5 {+4.57/24 6} f4 {-4.20/23 4}
46. Qxe4 {+4.87/27 4} Qxh5 {-4.98/23 2}
47. Qxf4 {+5.03/25 7} Qh1+ {-4.50/20 1}
48. Ng1 {+4.87/21 0} Qa8 {-4.48/19 1}
49. Nf3 {+5.63/27 4} Bf6 {-5.35/21 3}
50. Qg4+ {+5.93/30 4} Kh8 {-5.61/22 2}
51. Qf5 {+6.34/28 4} Kg7 {-6.34/22 1}
52. Kg2 {+6.39/30 3} Qe8 {-6.23/19 0}
53. Ng5 {+6.87/29 3} Qa8+ {-7.02/22 3}
54. Ne4 {+8.19/24 2} Bd8 {-6.91/19 0}
55. Qd7+ {+10.14/23 4} Kg6 {-7.60/23 3}
56. Kf3 {+12.96/27 4} Qa3+ {-6.52/15 0}
57. Kg4 {+199.64/37 3} Qd3 {-8.98/22 1}
58. Qe6+ {+M14/36 1} Kg7 {-199.35/27 3}
59. Kf4 {+M14/38 1} Kf8 {-199.33/28 1}
60. Qd7 {+M12/43 1} Bf6 {-M11/42 2}
{Arena Adjudication}
1-0
https://rwbc-chess.de

[Trolls n'existent pas...]
Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

Guenther wrote: ↑Mon Apr 03, 2023 5:24 pm ..... is completely illegal, because it is 'released' with a fake source
:shock: Did I post a link here ???

:lol: The language above tells me enough....

:wink:

============
An additional information:

the "thing" above (I have no idea what and where you downloaded what you claim) does not use MZ exp. but something else !
Top
Guenther
Posts: 4718
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: If you are interested....

Post by Guenther »

Sylwy wrote: ↑Mon Apr 03, 2023 5:34 pm
Guenther wrote: ↑Mon Apr 03, 2023 5:24 pm ..... is completely illegal, because it is 'released' with a fake source
:shock: Did I post a link here ???

:lol: The language above tells me enough....

:wink:

============
An additional information:

the "thing" above (I have no idea what and where you downloaded what you claim) does not use MZ exp. but something else !
Oh boy, check your own screenshot, but I can post another startup screenshot too...
it says it uses the MZ exp.cpp source code for using experience binary files.
(I really don't care if it was learnt by MZ or by someone else, or is simply an obfuscated one taken from MZ.)

I did not post any links, because I sent them to the moderation ofc.

And of course results from illegal engines should be in the EO forum, or forbidden anyway
https://rwbc-chess.de

[Trolls n'existent pas...]
Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

Guenther wrote: ↑Mon Apr 03, 2023 7:37 pm
....it says it uses the MZ exp.cpp source code for using experience binary files.
(I really don't care if it was learnt by MZ or by someone else, or is simply an obfuscated one taken from MZ.)
It says it uses the MZ exp.cpp source code .....anyone can read this, of course. Nothing new. Who trained the experience file (a true masterpiece) is a completely different story, boy!

If the engine is legal or illegal, I care less. I'm only interested in the result. I don't post links and in general I don't have the expertise to say whether it is yes or not.
Of course, the moderators can delete the initial post (your "analyses" will also go...). I won't waste my time with a new answer!

:wink:
Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

Arasan 23.5-x64-avx2
==================

Match conditions:

-TC=4'+2"
-Hash=256 MB
-GUI: Arena 3.5.1
-Book: SuperGM_4mvs.abk (8 plies book) for both engines
-1 thread-CPU=Intel i5-7400-3GHz (Kaby Lake)
-TBS: 6-men Syzygy bases for both engines
-NNs: defaults
-Ponder: off
-OS: Windows 10 Home.

The final result:

Image

Image

+24 -14 =62

+35 Elo points in 100 games for Arasan 23.5-x64-avx2

Smallbrain 7.0 avx2 has 3450 Elo points (Stefan Pohl rating list).

Arasan 23.4 avx2 has 3470 Elo points (Stefan Pohl rating list).

A sample:

Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

The new Clover looks extremely powerful!

Image
Top
lucametehau
Posts: 125
Joined: Thu Apr 22, 2021 3:56 pm
Location: Bucharest, Romania
Full name: Metehau Luca

Re: If you are interested....

Post by lucametehau »

Sylwy wrote: ↑Thu Apr 13, 2023 9:31 pm The new Clover looks extremely powerful!

Image
:wink:
Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

Clover.4.0-avx2
=============

Match conditions:

-TC=4'+2"
-Hash=256 MB
-GUI: Arena 3.5.1
-Book: SuperGM_4mvs.abk (8 plies book) for both engines
-1 thread-CPU=Intel i5-7400-3GHz (Kaby Lake)
-TBS: 6-men Syzygy bases for both engines
-NNs: defaults
-Ponder: off
-OS: Windows 10 Home.

The final result:

Image

Image

+13 -10 =77

+14 Elo points in 100 games for Clover.4.0-avx2

Uralochka 3.39d avx2 has 3558 Elo points (Stefan Pohl rating list).

Clover 3.3.1 avx2 has 3514 Elo points (Stefan Pohl rating list).

A sample:

Top
User avatar
Sylwy
Posts: 4825
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: If you are interested....

Post by Sylwy »

RubiChess-20230410_x86-64-bmi2
=============================

Match conditions:

-TC=4'+2"
-Hash=256 MB
-GUI: Arena 3.5.1
-Book: SuperGM_4mvs.abk (8 plies book) for both engines
-1 thread-CPU=Intel i5-7400-3GHz (Kaby Lake)
-TBS: 6-men Syzygy bases for both engines
-NNs: defaults
-Ponder: off
-OS: Windows 10 Home.

The final result in my match:

Image

Image

+8 -11 =81

-7 Elo points in 100 games for RubiChess-20230410_x86-64-bmi2/ +14 Elo points for Revenge_3.0_win_bmi2

Of course, I do not comment. I feel that there are subtleties that go beyond my primitive impression that it is a stagnation..... :wink:

Image

Image

Image

A sample from this match:

Top

Return to “Computer Chess Club: Tournaments and Matches”