ProDeo
Would you like to react to this message? Create an account in a few clicks or log in to continue.
ProDeo

Computer Chess
 
HomeHome  CalendarCalendar  Latest imagesLatest images  FAQFAQ  SearchSearch  MemberlistMemberlist  UsergroupsUsergroups  RegisterRegister  Log in  

 

 Rebel 16.2 release

Go down 
+12
Ipmanchess
Damir Desevac
texium
Dann Corbit
mwyoung
Peter Berger
stevemaughan
Mclane
Dio
pohl4711
Chris Whittington
Admin
16 posters
Go to page : Previous  1, 2, 3, 4  Next
AuthorMessage
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 3:42 am

Peter Berger wrote:
Welcome Rebel 16.2
Rebel ( with the Noomen book) introduced itself with a great game on my computers, which was basically a book victory though, in a line that was initially worked out by an otherwise pretty unknown German correspondence player in like 2001 ( kudos to Mr Winckelmann).
I wonder whether there is a bug in the book utility though. Rebel had to play moves 3 and 9 on its own, that I am pretty sure should have been in Jeroen‘s book.
[Event "Lang 120min+10sek"]
[Site "Berlin"]
[Date "2023.02.27"]
[Round "?"]
[White "Crafty 25.6"]
[Black "Rebel-16.2"]
[Result "0-1"]
[ECO "B33"]
[PlyCount "56"]
[TimeControl "7200+10"]

{4096MB, LAPTOP-NCDN8BTK} 1. e4 {[%emt 0:00:00]} c5 {[%eval 1,0] [%emt 0:00:01]
} 2. Nf3 {[%emt 0:00:05]} Nc6 {[%eval 1,0] [%emt 0:00:00]} 3. d4 {[%emt 0:00:
05]} cxd4 {[%eval 76,33] [%emt 0:03:42]} 4. Nxd4 {[%emt 0:00:06]} Nf6 {[%eval
1,0] [%emt 0:00:01]} 5. Nc3 {[%emt 0:00:06]} e5 {[%eval 1,0] [%emt 0:00:01]} 6.
Ndb5 {[%emt 0:00:05]} d6 {[%eval 1,0] [%emt 0:00:00]} 7. Bg5 {[%emt 0:00:05]}
a6 {[%eval 1,0] [%emt 0:00:00]} 8. Na3 {[%emt 0:00:06]} b5 {[%eval 1,0] [%emt
0:00:00]} 9. Bxf6 {[%emt 0:00:05]} gxf6 {[%eval 48,34] [%emt 0:03:38]} 10. Nd5
{[%emt 0:00:08]} f5 {[%eval 1,0] [%emt 0:00:01]} 11. Bd3 {[%emt 0:00:07]} Be6 {
[%eval 1,0] [%emt 0:00:01]} 12. Qh5 {[%emt 0:00:08]} Rg8 {[%eval 1,0] [%emt 0:
00:01]} 13. g3 {[%emt 0:00:06]} Rg5 {[%eval -24,34] [%emt 0:05:55]} 14. Qxh7 {
[%emt 0:00:08] (Dd1)} Nd4 {[%eval -251,31] [%emt 0:05:04]} 15. c3 {[%emt 0:00:
06]} Nf3+ {[%eval -271,33] [%emt 0:03:05]} 16. Ke2 {[%emt 0:00:06] (Kd1)} Bxd5
{[%eval -460,31] [%emt 0:04:03]} 17. exd5 {[%emt 0:00:07]} e4 {[%eval -599,31]
[%emt 0:07:01]} 18. Bc2 {[%emt 0:03:54] (Sc2)} Qf6 {[%eval -766,28] [%emt 0:01:
59]} 19. Qh3 {[%emt 0:02:53]} b4 {[%eval -754,31] [%emt 0:04:07]} 20. Nb1 {
[%emt 0:00:06]} Ke7 {[%eval -864,29] [%emt 0:02:53]} 21. Qf1 {[%emt 0:10:25]
(a3)} Bh6 {[%eval -900,30] [%emt 0:04:03]} 22. Qd1 {[%emt 0:02:41]} Rh5 {
[%eval -1042,32] [%emt 0:05:44]} 23. h3 {[%emt 0:03:36] (h4)} Rg8 {[%eval
-1261,26] [%emt 0:01:11]} 24. Kf1 {[%emt 0:10:59] (Df1)} Rhg5 {[%eval -1474,30]
[%emt 0:01:37]} 25. Nd2 {[%emt 0:08:07] (g4)} Rxg3 {[%eval -1790,27] [%emt 0:
01:14]} 26. Nxf3 {[%emt 0:01:55]} Rxf3 {[%eval -2086,29] [%emt 0:01:37]} 27.
Qd4 {[%emt 0:00:55]} Qg6 {[%eval -2157,28] [%emt 0:01:22]} 28. Qa7+ {[%emt 0:
12:41]} Kf8 {[%eval -2445,33] [%emt 0:06:21]} 0-1



I show it as 2010, unless another ICCF player fell into the same line almost a decade later. And that is possible.

Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

[pgn][Event "Lang 120min+10sek"]
[Site "Berlin"]
[Date "2023.02.27"]
[Round "?"]
[White "Crafty 25.6"]
[Black "Rebel-16.2"]
[Result "0-1"]
[ECO "B33"]
[Annotator "Tactical Analysis 6.4"]
[PlyCount "56"]
[EventDate "2023.??.??"]

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8. Na3 b5 9. Bxf6 gxf6 10. Nd5 f5 11. Bd3 Be6 12. Qh5 Rg8 13. g3 Rg5 14. Qxh7 {-1.80/33 [#] B33: Sicilian: Pelikan and Sveshnikov Variations.} (14. Qd1 $1 $11 {-0.19/34}) 14... Nd4 $1 $19 {Threatens to win with ...Nf3+.} 15. c3 {[#]} Nf3+ 16. Ke2 {-3.14/31} (16. Kd1 $142 {-1.84/32}) 16... Bxd5 17. exd5 {Black must now prevent h4.} e4 18. Bc2 {-4.31/25} ({But not} 18. h4 Rg6 19. h5 Rg5 $19) (18. Nc2 $142 {-3.05/29} Qe7 19. Rae1 {Decoy}) 18... Qf6 19. Qh3 {[#]} b4 $1 20. Nb1 (20. Nc4 bxc3) 20... Ke7 {White is weak on the dark squares. Black is clearly winning.} 21. Qf1 {[#]} Bh6 $146 ({Predecessor:} 21... bxc3 22. bxc3 Rg8 23. Qd1 Qh8 24. h4 Bh6 25. Kf1 Qe5 26. Qe2 Rab8 {0-1 Skalicky,F (2251)-Silin,V (2234) W-ch WS/M/258 email ICCF email 2010}) 22. Qd1 Rh5 23. h3 Rg8 24. Kf1 Rhg5 25. Nd2 {[#]} Rxg3 $1 26. Nxf3 Rxf3 27. Qd4 Qg6 {( -> ...Be3)} 28. Qa7+ Kf8 {White got outplayed after the opening. Excellently played by Rebel-16.2! Weighted Error Value: White=0.55/Black=0.00 (flawless) . Loses game: --- --- Missed win: --- --- Mistake: White=2 --- Inaccurate: --- --- OK: White=1 Black=9 Best: --- Black=1 Strong: --- Black=1 Brilliant: --- ---} 0-1

[/pgn]

Admin likes this post

Back to top Go down
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 8:18 am

mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.
Back to top Go down
http://rebel13.nl/
Peter Berger




Posts : 123
Join date : 2020-11-20

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 10:52 am

mwyoung wrote:

I show it as 2010, unless another ICCF player fell into the same line almost a decade later. And that is possible.


The relevant games were Satici-Winckelmann and Petraitis-Winckelmann, 2001 (I know this because Mr. Winckelmann’s analysis made the ECO, quite an honour for an amateur player 😉 ).
Another great very positional game today (see below). This also means that is pretty obvious that the book utility won’t play any captures from the book – no biggy, but it is conceivable to lead to problems.

[pgn]
[Event "Lang 120min+10sek"]
[Site "Berlin"]
[Date "2023.02.28"]
[Round "?"]
[White "Rebel-16.2"]
[Black "Crafty 25.6"]
[Result "1-0"]
[ECO "B42"]
[PlyCount "89"]
[TimeControl "7200+10"]

{4096MB, LAPTOP-NCDN8BTK} 1. e4 {[%eval 1,0] [%emt 0:00:00]} c5 {[%emt 0:00:06]
} 2. Nf3 {[%eval 1,0] [%emt 0:00:00]} e6 {[%emt 0:00:06]} 3. d4 {[%eval 1,0]
[%emt 0:00:00]} cxd4 {[%emt 0:00:05]} 4. Nxd4 {[%eval 70,32] [%emt 0:03:42]} a6
{[%emt 0:00:05] (Sc6)} 5. Bd3 {[%eval 1,0] [%emt 0:00:00]} g6 {[%emt 0:00:06]}
6. O-O {[%eval 115,36] [%emt 0:05:57]} Bg7 {[%emt 0:00:06]} 7. Be3 {[%eval 124,
35] [%emt 0:05:56]} Nc6 {[%emt 0:00:06] (Se7)} 8. Nxc6 {[%eval 220,32] [%emt 0:
02:58]} bxc6 {[%emt 0:00:06]} 9. Nd2 {[%eval 122,39] [%emt 0:07:17]} Ne7 {
[%emt 0:00:11] (Sf6)} 10. b4 {[%eval 210,31] [%emt 0:02:08]} O-O {[%emt 0:03:
16] (Dc7)} 11. Rb1 {[%eval 206,34] [%emt 0:02:22]} f5 {[%emt 0:04:25] (Dc7)}
12. f3 {[%eval 232,31] [%emt 0:05:39]} d6 {[%emt 0:00:06] (Dc7)} 13. a4 {
[%eval 256,33] [%emt 0:02:30]} d5 {[%emt 0:07:38] (Dc7)} 14. Nb3 {[%eval 301,
29] [%emt 0:01:28]} Bc3 {[%emt 0:06:10]} 15. Bc5 {[%eval 334,33] [%emt 0:03:08]
} Re8 {[%emt 0:02:11] (Tb8)} 16. Kh1 {[%eval 366,29] [%emt 0:02:11]} Rb8 {
[%emt 0:03:57] (e5)} 17. Na5 {[%eval 405,32] [%emt 0:06:07]} Qc7 {[%emt 0:00:
07]} 18. Qe2 {[%eval 409,31] [%emt 0:02:53]} Ra8 {[%emt 0:15:25]} 19. Qe3 {
[%eval 393,37] [%emt 0:11:18]} Bf6 {[%emt 0:03:55] (Lg7)} 20. c4 {[%eval 451,
31] [%emt 0:09:11]} Bd7 {[%emt 0:02:13] (e5)} 21. Nb3 {[%eval 460,30] [%emt 0:
01:56]} e5 {[%emt 0:05:19] (g5)} 22. Bb6 {[%eval 491,35] [%emt 0:03:04]} Qb8 {
[%emt 0:03:07] (Dd6)} 23. cxd5 {[%eval 586,31] [%emt 0:01:14]} cxd5 {[%emt 0:
03:30]} 24. a5 {[%eval 637,32] [%emt 0:04:14]} Nc8 {[%emt 0:02:33]} 25. exd5 {
[%eval 646,33] [%emt 0:01:11]} Ba4 {[%emt 0:05:05] (Kg7)} 26. Bc5 {[%eval 640,
31] [%emt 0:02:21]} Qb7 {[%emt 0:03:06] (Sd6)} 27. d6 {[%eval 737,34] [%emt 0:
02:40]} Kh8 {[%emt 0:01:04] (Kg7)} 28. Bc4 {[%eval 814,26] [%emt 0:01:12]} Qd7
{[%emt 0:02:01] (Td8)} 29. Rfd1 {[%eval 876,27] [%emt 0:00:49]} Rd8 {[%emt 0:
00:13] (e4)} 30. Qe2 {[%eval 992,28] [%emt 0:02:03]} Bb5 {[%emt 0:01:24] (h5)}
31. Bxb5 {[%eval 1080,28] [%emt 0:00:57]} axb5 {[%emt 0:00:28]} 32. Rd5 {
[%eval 1134,30] [%emt 0:02:46]} Kg8 {[%emt 0:03:36] (e4)} 33. Rbd1 {[%eval
1244,25] [%emt 0:00:59]} Ra6 {[%emt 0:00:47]} 34. Nc1 {[%eval 1313,26] [%emt 0:
01:24]} Re8 {[%emt 0:01:41] (Tc6)} 35. Nd3 {[%eval 1376,26] [%emt 0:00:42]} e4
{[%emt 0:00:07]} 36. fxe4 {[%eval 1371,27] [%emt 0:01:21]} fxe4 {[%emt 0:03:34]
} 37. Nf2 {[%eval 1540,30] [%emt 0:00:41]} e3 {[%emt 0:00:20]} 38. Ng4 {
[%eval 1543,28] [%emt 0:00:41]} Bg7 {[%emt 0:01:08]} 39. Nxe3 {[%eval 1547,28]
[%emt 0:00:55]} Ra8 {[%emt 0:00:06] (Te6)} 40. Qg4 {[%eval 1638,29] [%emt 0:01:
07]} Rd8 {[%emt 0:01:00]} 41. h3 {[%eval 1732,29] [%emt 0:01:50]} Ra6 {[%emt 0:
01:18] (Sa7)} 42. Qe4 {[%eval 1815,24] [%emt 0:00:36]} Rf8 {[%emt 0:01:26]} 43.
Ng4 {[%eval 1948,27] [%emt 0:01:01]} Rd8 {[%emt 0:00:14] (Tc6)} 44. Qd3 {
[%eval 2026,25] [%emt 0:00:30]} Nxd6 {[%emt 0:02:11] (Te8)} 45. Rxd6 {[%eval
2371,27] [%emt 0:00:51]} 1-0[/pgn]
Back to top Go down
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 11:02 am

Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

Admin and Brendan like this post

Back to top Go down
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 11:42 am

mwyoung wrote:
Mclane wrote:
mwyoung wrote:
Chris Whittington wrote:
Admin wrote:
Thanks Thorsten, can you mail me the pgn ?

Those results don’t make a lot of sense

Why?

Maybe he dislikes that some engines have problems with the time controls.
IMO it is not my job judge why some engines, e.g. Stockfish 15 or 15.1 have time control problems no matter if you test in 40/3‘ or 40/10 or even 40/20.

An engine that makes zillions of NPS and goes into the 40-50 ply should be capable to play 40/10‘ games.
If not it’s obviously a bug.

Rebel16.2 is doing a very good job and plays very strong.
Thats what counts.



I agree.

There were a lot of forfeits, also for Rebel, and the total favored Rebel and so the outcome is not fully trustworthy. Maybe Thorsten should consider to drop Arena.
Back to top Go down
http://rebel13.nl/
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 11:45 am

mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

Interesting tool.... is it available for download or must I buy Fat-Fritz? lol!
Back to top Go down
http://rebel13.nl/
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 11:54 am

mwyoung wrote:
Admin wrote:
Just for the fun of it.

Code:
RANK   ENGINE                                    GAMES  POINTS   1  2  3  4  5  6  7  8  9  10
-------------------------------------------------------------------------------------------------
1.   REBEL 16.2 64-BIT                            9     5.0    *  =  =  =  =  =  =  =  =  1
2.   STOCKFISH 15.1 64-BIT                        8     4.5    =  *  1  =  =  =  =  .  =  =
3.   SLOWCHESS BLITZ 2.9 64-BIT                   9     4.5    =  0  *  =  =  =  =  1  =  =
4.   BERSERK 11.1 64-BIT                          8     4.5    =  =  =  *  =  =  .  =  1  =
5.   DRAGON 3.2 BY KOMODO 64-BIT                  8     4.0    =  =  =  =  *  =  =  =  =  .
6.   ETHEREAL 14.00 64-BIT                        8     4.0    =  =  =  =  =  *  =  =  .  =
7.   REVENGE 3.0 64-BIT                           8     4.0    =  =  =  .  =  =  *  =  =  =
8.   IGEL 3.4.0 64-BIT                            8     3.5    =  .  0  =  =  =  =  *  =  =
9.   KOIVISTO 9.0 64-BIT                          8     3.5    =  =  =  0  =  .  =  =  *  =
10.   RUBICHESS 20221120 64-BIT                    8     3.5    0  =  =  =  .  =  =  =  =  *
 
Total games = 41

Graham's current tournament.

Great results for Rebel 16.2

You know what this means!

You are now on the enemies list of the Stockfish fanboys. They do not like it when another chess engine challenges Stockfish. As that is a hard and fast rule for testers. As the programmer of Rebel 16.2. You will need to hire body guards. lol!

Code:
RANK   ENGINE                                    GAMES  POINTS   1    2    3    4    5    6    7    8    9    10
---------------------------------------------------------------------------------------------------------------------
1.   STOCKFISH 15.1 64-BIT                       30    18.0    **** ===. ===. ==1= =1=. ===1 ===. ===. 11== ==1.
2.   DRAGON 3.2 BY KOMODO 64-BIT                 30    16.5    ===. **** ==== ===. ==== ===. ===. ==1. =1== =1=.
3.   REBEL 16.2 64-BIT                           30    16.0    ===. ==== **** ===. ==== ===. 1==. =1== ===. ===.
4.   BERSERK 11.1 64-BIT                         30    15.5    ==0= ===. ===. **** 1==. =0=. ===. ===. ===1 ==1=
5.   KOIVISTO 9.0 64-BIT                         30    14.5    =0=. ==== ==== 0==. **** =1=. ==== ===. ===. ===.
6.   ETHEREAL 14.00 64-BIT                       30    14.5    ===0 ===. ===. =1=. =0=. **** ===. ==== ===. ====
7.   RUBICHESS 20221120 64-BIT                   29    14.0    ===. ===. 0==. ===. ==== ===. **** ==== ===. ===.
8.   REVENGE 3.0 64-BIT                          30    14.0    ===. ==0. =0== ===. ===. ==== ==== **** ===. ===.
9.   SLOWCHESS BLITZ 2.9 64-BIT                  30    13.5    00== =0== ===. ===0 ===. ===. ===. ===. **** 1==.
10.   IGEL 3.4.0 64-BIT                           29    12.5    ==0. =0=. ===. ==0= ===. ==== ===. ===. 0==. ****

Total games = 149

It's indeed remarkable, 30 games and no single loss, well not at this moment, I don't expect it will stay this way.

Mclane and Dio like this post

Back to top Go down
http://rebel13.nl/
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 12:08 pm

mwyoung wrote:
You are now on the enemies list of the Stockfish fanboys. They do not like it when another chess engine challenges Stockfish. As that is a hard and fast rule for testers. As the programmer of Rebel 16.2. You will need to hire body guards. lol!

If that's the case Chris will need more body guards Very Happy
Back to top Go down
http://rebel13.nl/
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 12:10 pm

Admin wrote:
mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

Interesting tool.... is it available for download or must I buy Fat-Fritz? lol!

You will need to buy Chessbase 17.
Back to top Go down
Mclane

Mclane


Posts : 2944
Join date : 2020-11-17
Age : 57
Location : United States of Europe, Germany, Ruhr area

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 1:08 pm

Admin wrote:
mwyoung wrote:
Mclane wrote:
mwyoung wrote:
Chris Whittington wrote:
Admin wrote:
Thanks Thorsten, can you mail me the pgn ?

Those results don’t make a lot of sense

Why?

Maybe he dislikes that some engines have problems with the time controls.
IMO it is not my job judge why some engines, e.g. Stockfish 15 or 15.1 have time control problems no matter if you test in 40/3‘ or 40/10 or even 40/20.

An engine that makes zillions of NPS and goes into the 40-50 ply should be capable to play 40/10‘ games.
If not it’s obviously a bug.

Rebel16.2 is doing a very good job and plays very strong.
Thats what counts.



I agree.

There were a lot of forfeits, also for Rebel, and the total favored Rebel and so the outcome is not fully trustworthy. Maybe Thorsten should consider to drop Arena.


There is nothing wrong with Arena. Older Stockfish versions do not have these time forfeits.
Back to top Go down
http://www.thorstenczub.de
Mclane

Mclane


Posts : 2944
Join date : 2020-11-17
Age : 57
Location : United States of Europe, Germany, Ruhr area

Rebel 16.2 release - Page 2 Empty
PostSubject: The religious sects in computerchess    Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 1:13 pm

Admin wrote:
mwyoung wrote:
You are now on the enemies list of the Stockfish fanboys. They do not like it when another chess engine challenges Stockfish. As that is a hard and fast rule for testers. As the programmer of Rebel 16.2. You will need to hire body guards. lol!

If that's the case Chris will need more body guards Very Happy


They treat Stockfish as god like.
I have no idea why they do it.
There is a tendency for religious behaviour for technical things.
Apple products comes to my mind. The fans of this brand do also behave like a religious sect. We have a couple of apple fans in the company i work. And almost all of them have to reload their mobile phone during the day because the battery is nearly empty.
All the android colleagues, we never reload our batteries at work. We do it at home before or after work. When you refer to this weakness they react very upset or even deny the fact that this is a weakness of this brand. They simply deny the technical fact that the apple batteries are smaller then the batteries of the competitors.

And similar it is wit the „best“ chess engine.

In the moment you critisize apple products or stockfish or in older days it was chessbase, the religious fans attack you.

If you play fast games the game is too fast (from their point of view).
If you play slow games you do not have “enough games” to make a point (for them). If you present a PGN noone answers. Because these guys do not play chess or do not replay chess games as it seems.
They are mainly interested in RESULTS not in chess content.

They are used to show and discuss elo list tables instead of pgns.


Last edited by Mclane on Tue Feb 28, 2023 1:22 pm; edited 1 time in total

Brendan likes this post

Back to top Go down
http://www.thorstenczub.de
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 1:20 pm

Mclane wrote:
Admin wrote:
mwyoung wrote:
You are now on the enemies list of the Stockfish fanboys. They do not like it when another chess engine challenges Stockfish. As that is a hard and fast rule for testers. As the programmer of Rebel 16.2. You will need to hire body guards. lol!

If that's the case Chris will need more body guards Very Happy


They treat Stockfish as god like.
I have no idea why they do it.
There is a tendency for religious behaviour for technical things.
Apple products comes to my mind. The fans of this brand do also behave like a religious sect.
And similar it is wit the „best“ chess engine.

In the moment you critisize apple products or stockfish or in older days it was chessbase, the religious fans attack you.

If you play fast games the game i too fast.
If you play slow games you do not have “enough games” to make a point. If you present a PGN noone answers. Because these guys do not play chess as it seems.
They are mainly interested in RESULTS not in chess content.

They are used to show elo list tables instead of pgns.

Correct. The die hard Stockfish fanboys take their devotion very seriously.

In other words they are nuts!

Mclane likes this post

Back to top Go down
Chris Whittington




Posts : 1254
Join date : 2020-11-17
Location : France

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 1:47 pm

mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

And which is the 'god' engine reference?
Back to top Go down
Chris Whittington




Posts : 1254
Join date : 2020-11-17
Location : France

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 1:55 pm

mwyoung wrote:
Admin wrote:
mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

Interesting tool.... is it available for download or must I buy Fat-Fritz? lol!

You will need to buy Chessbase 17.

Can't we write one ourselves?

Algorithm:
Take PGN game.
For each move, ask SF for best move+eval, second best move+eval etc.
If PGN move = SF move, score cp-loss=0
else score cp-loss=bestmoveEval minus PGNmoveEval.

calculate the average cp-loss.
that's it.

If you tell me how to ask SF for Nth move eval, I'll Python it in a few minutes.

Brendan likes this post

Back to top Go down
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 2:18 pm

Chris Whittington wrote:
mwyoung wrote:
Admin wrote:
mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

Interesting tool.... is it available for download or must I buy Fat-Fritz? lol!

You will need to buy Chessbase 17.

Can't we write one ourselves?

Algorithm:
Take PGN game.
For each move, ask SF for best move+eval, second best move+eval etc.
If PGN move = SF move, score cp-loss=0
else score cp-loss=bestmoveEval minus PGNmoveEval.

calculate the average cp-loss.
that's it.

If you tell me how to ask SF for Nth move eval, I'll Python it in a few minutes.

Yes except Chessbase 17 weights their score by difficulty. And Chessbase 17 does not count opening moves or captures and other forced moves.
Back to top Go down
Mclane

Mclane


Posts : 2944
Join date : 2020-11-17
Age : 57
Location : United States of Europe, Germany, Ruhr area

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 2:39 pm

mwyoung wrote:
Mclane wrote:
Admin wrote:
mwyoung wrote:
You are now on the enemies list of the Stockfish fanboys. They do not like it when another chess engine challenges Stockfish. As that is a hard and fast rule for testers. As the programmer of Rebel 16.2. You will need to hire body guards. lol!

If that's the case Chris will need more body guards Very Happy


They treat Stockfish as god like.
I have no idea why they do it.
There is a tendency for religious behaviour for technical things.
Apple products comes to my mind. The fans of this brand do also behave like a religious sect.
And similar it is wit the „best“ chess engine.

In the moment you critisize apple products or stockfish or in older days it was chessbase, the religious fans attack you.

If you play fast games the game i too fast.
If you play slow games you do not have “enough games” to make a point. If you present a PGN noone answers. Because these guys do not play chess as it seems.
They are mainly interested in RESULTS not in chess content.

They are used to show elo list tables instead of pgns.

Correct. The die hard Stockfish fanboys take their devotion very seriously.

In other words they are nuts!

I have posted 2 sacrificial games of CSTAL in PGN format and there is NO comment. Nothing.

https://forum.computerschach.de/cgi-bin/mwf/topic_show.pl?tid=13629

They do not replay chess games.
Do they play chess at all ??
Are they fascinated by a game of chess ?
I highly doubt that.


Last edited by Mclane on Tue Feb 28, 2023 2:48 pm; edited 1 time in total
Back to top Go down
http://www.thorstenczub.de
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 2:43 pm

Mclane wrote:
Admin wrote:
mwyoung wrote:
Mclane wrote:
mwyoung wrote:
Chris Whittington wrote:
Admin wrote:
Thanks Thorsten, can you mail me the pgn ?

Those results don’t make a lot of sense

Why?

Maybe he dislikes that some engines have problems with the time controls.
IMO it is not my job judge why some engines, e.g. Stockfish 15 or 15.1 have time control problems no matter if you test in 40/3‘ or 40/10 or even 40/20.

An engine that makes zillions of NPS and goes into the 40-50 ply should be capable to play 40/10‘ games.
If not it’s obviously a bug.

Rebel16.2 is doing a very good job and plays very strong.
Thats what counts.



I agree.

There were a lot of forfeits, also for Rebel, and the total favored Rebel and so the outcome is not fully trustworthy. Maybe Thorsten should consider to drop Arena.


There is nothing wrong with Arena. Older Stockfish versions do not have these time forfeits.

Rebel had 3 forfeits, impossible.
Back to top Go down
http://rebel13.nl/
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 2:46 pm

Mclane wrote:
They are used to show and discuss elo list tables instead of pgns.

Rebel 16.2 release - Page 2 Like
Back to top Go down
http://rebel13.nl/
mwyoung

mwyoung


Posts : 880
Join date : 2020-11-25
Location : USA

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Feb 28, 2023 2:50 pm

Chris Whittington wrote:
mwyoung wrote:
Admin wrote:
mwyoung wrote:
Flawless victory by Rebel 16.2! CP loss = 0.00. bounce

What is CP loss and why is that important?

I assume CP stands for centi pawn.

Centipawn loss is just another analysis tool to show accuracy in the game play. Or more important in a serries of games.

It means in this case no detectable errors were found in the game play using retrograde analysis, and measured in centipawn loss per move average.

For reference the best human players GM Carlsen, GM Fischer, GM Kasparov measures around .12 CP playing at peak form.  

On average a typical or just average Grandmaster player scores below .40 CP for the average Grandmaster player.  

Meaning playing Stockfish GM Carlsen, GM Fischer, GM Kasparov would be down 1.2 pawns ever 10 moves playing Stockfish on average.

Using this measurement it very easy to see why no human player has a chance playing the best chess engines.

And which is the 'god' engine reference?

You can use any engine as the reference chess engine. But it is best to use a top chess engine as reference. In the game posted I used Stockfish DEV.
Back to top Go down
Admin
Admin
Admin


Posts : 2536
Join date : 2020-11-17
Location : Netherlands

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptySun Mar 05, 2023 10:02 am

CCRL 40/15 is out

Code:
Rebel 16 64-bit         3395
Rebel 16.1 64-bit       3407
Rebel 16.2 64-bit       3431

Only changes time control.

Mclane, Dio and Eelco like this post

Back to top Go down
http://rebel13.nl/
Mclane

Mclane


Posts : 2944
Join date : 2020-11-17
Age : 57
Location : United States of Europe, Germany, Ruhr area

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyThu Mar 09, 2023 10:01 pm

Ed Schroeders comeback as grandpa programmer Wink :

Rebel 16.2 release - Page 2 F18f8310

Admin and Brendan like this post

Back to top Go down
http://www.thorstenczub.de
Peter Berger




Posts : 123
Join date : 2020-11-20

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptySat Mar 11, 2023 2:05 pm

A short update. I am still playing these slow Rebel 16.2-Crafty games , when I have the time to let them play and follow the games live (which doesn’t happen too often because of real life). Current score is 8-0.
Not only is Rebel completely destroying its opponent, the games are also very nice and impressive – they are looking way better than Stockfish’s to my personal taste. I am getting the impression that Rebel profits from the long time control, but this is guesswork of course as I am testing nothing else 😉 . But nothing really has to be posted, nothing strange.
I can also give a feedback on the Noomen book now: I really like – high quality work.
Much as expected the book utility bug doesn’t matter much, other than stealing a little time from Rebel’s bank, that doesn’t matter at all. Only thing is that a few lines Jeroen might well have chosen will never get played (e.g. QGA).

Admin, Mclane and matejst like this post

Back to top Go down
Dann Corbit




Posts : 188
Join date : 2020-11-26

Rebel 16.2 release - Page 2 Empty
PostSubject: I added Leptir to my contest   Rebel 16.2 release - Page 2 EmptyTue Mar 21, 2023 4:27 am

Engine             Score                                        St                                     Le                                     Lc                                     Lc                                     Te                                     Re    S-B
1: Stockfish-x64-avx2 108.5/189 ······································ =====================================  ======1===================1=========== ====11=1==1==========1============1=== ===1=======1=======1===1===1=1=1=1=1== =1111=11===1=1=====1===========1=1====  9851.0
2: Leptir-x64-avx2    103.0/189 =====================================  ······································ ====================================== ==============1==================1==== =========1===============1===1=1====== ===1===1=1=1=1===1=1===1=1=1=1========  9456.0
3: Lc0-T78            100.0/189 ======0===================0=========== ====================================== ······································ =====================================  =========1=======1====1===========1==1 =====1===1=1===========1=1=====1=1110=  9246.5
4: Lc0-T60            91.5/189  ====00=0==0==========0============0=== ==============0==================0==== =====================================  ······································ ================================1===== ====0==========1=============1========  8638.0
5: Test-62t           85.5/189  ===0=======0=======0===0===0=0=0=0=0== =========0===============0===0=0====== =========0=======0====0===========0==0 ================================0===== ······································ ====================================1   8158.5
6: Rebel-16.2         78.5/189  =0000=00===0=0=====0===========0=0==== ===0===0=0=0=0===0=0===0=0=0=0======== =====0===0=0===========0=0=====0=0001= ====1==========0=============0======== ====================================0  ······································  7587.0

567 of 900 games played
Name of the tournament: Tdog
Site/ Country: DESKTOP-426Q3JG, United States
Level: Blitz 5/2
Hardware: AMD Ryzen Threadripper 3970X 32-Core Processor with 63.9 GB Memory
Operating system: Windows 10 Home Home Edition (Build 9200) 64 bit
PGN-File: C:\Program Files (x86)\Arena\Tournaments\Tdog.pgn
Back to top Go down
Dann Corbit




Posts : 188
Join date : 2020-11-26

Rebel 16.2 release - Page 2 Empty
PostSubject: Threading test, with Leptir added for kicks   Rebel 16.2 release - Page 2 EmptyTue Mar 21, 2023 2:45 pm

  Program              Elo    +   -   Games   Score   Av.Op.  Draws
1 Stockfish-x64-avx2 : 3445   20  16   202    57.7 %   3391   84.7 %
2 Leptir-x64-avx2    : 3424   16  12   202    54.2 %   3395   91.6 %
3 Lc0-T78            : 3414   15  13   202    52.5 %   3397   91.1 %
4 Lc0-T60            : 3390   11  13   202    48.3 %   3402   93.6 %
5 Test-62t           : 3374   13  17   202    45.5 %   3405   90.1 %
6 Rebel-16.2         : 3352   18  21   202    41.8 %   3410   81.7 %



I added Leptir, since there is a great hullabaloo about that engine.
Here are the games and statistics:
https://drive.google.com/file/d/1EGdcE8DgWPFWK2ofQx0hbLWwSogOsNBx/view?usp=share_link
Back to top Go down
Dann Corbit




Posts : 188
Join date : 2020-11-26

Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 EmptyTue Mar 21, 2023 3:09 pm

select sum(acd)/count(acd) as aveDepth, Engine from Tdog group by Engine order by 1 desc
Gives this result:
Code:
aveDepth, Engine
55,Stockfish-x64-avx2
48,Leptir-x64-avx2
34,Test-62t
33,Rebel-16.2
9,Lc0-T78
8,Lc0-T60
Back to top Go down
Sponsored content





Rebel 16.2 release - Page 2 Empty
PostSubject: Re: Rebel 16.2 release   Rebel 16.2 release - Page 2 Empty

Back to top Go down
 
Rebel 16.2 release
Back to top 
Page 2 of 4Go to page : Previous  1, 2, 3, 4  Next
 Similar topics
-
» Release Rebel 15
» Release Rebel 15x2
» New Release Announcement Rebel-14-Chess-System-Tal [NNUE]
» Texel 1.08 release
» First release ADRL Blitz List

Permissions in this forum:You cannot reply to topics in this forum
ProDeo :: Computer Chess-
Jump to: