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  

 

 sf16 robin round

Go down 
5 posters
AuthorMessage
Admin
Admin
Admin


Posts : 2528
Join date : 2020-11-17
Location : Netherlands

sf16 robin round Empty
PostSubject: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 7:06 pm

Robin round, 10,000+ bullet games.

Code:
No. Name    Win Draw Loss Unf.  Score Games       %
---------------------------------------------------
  1 sf16   +1067 =1862 -133  *0 1998.0  3062   65.3%
  2 sf15.1 +902 =1927 -234   *0 1865.5  3063   60.9%
  3 sf15   +852 =1932 -279   *0 1818.0  3063   59.4%
  4 sf14.1 +633 =1952 -477   *0 1609.0  3062   52.5%
  5 sf14   +447 =1949 -667   *0 1421.5  3063   46.4%
  6 sf13   +284 =1748 -1030  *0 1158.0  3062   37.8%
  7 sf12   +128 =1442 -1493  *0  849.0  3063   27.7%

Total Games:   10719
White Wins:     2905 (27.1%)
Black Wins:     1408 (13.1%)
Draws:          6406 (59.8%)

Dio and Ghppn like this post

Back to top Go down
http://rebel13.nl/
Chris Whittington




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

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 7:25 pm

Admin wrote:
Robin round, 10,000+ bullet games.

Code:
No. Name    Win Draw Loss Unf.  Score Games       %
---------------------------------------------------
  1 sf16   +1067 =1862 -133  *0 1998.0  3062   65.3%
  2 sf15.1 +902 =1927 -234   *0 1865.5  3063   60.9%
  3 sf15   +852 =1932 -279   *0 1818.0  3063   59.4%
  4 sf14.1 +633 =1952 -477   *0 1609.0  3062   52.5%
  5 sf14   +447 =1949 -667   *0 1421.5  3063   46.4%
  6 sf13   +284 =1748 -1030  *0 1158.0  3062   37.8%
  7 sf12   +128 =1442 -1493  *0  849.0  3063   27.7%

Total Games:   10719
White Wins:     2905 (27.1%)
Black Wins:     1408 (13.1%)
Draws:          6406 (59.8%)

can you show (for each engine) average length of game to win, av to draw, and av to loss?

Ghppn likes this post

Back to top Go down
Admin
Admin
Admin


Posts : 2528
Join date : 2020-11-17
Location : Netherlands

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 7:50 pm

Code:
No. Engine       1     2     3     4     5     6     7  Score  Games   Perc   Moves
-----------------------------------------------------------------------------------
 1 sf16      xxxxx 282.5 278.0 318.0 339.5 373.0 407.0 1998.0 / 3062 (65.25%)  75.0  
 2 sf15.1    227.5 xxxxx 268.5 299.5 319.5 354.5 396.0 1865.5 / 3063 (60.90%)  72.9  
 3 sf15      232.0 242.5 xxxxx 288.0 321.0 353.5 381.0 1818.0 / 3063 (59.35%)  74.7  
 4 sf14.1    192.0 210.5 223.0 xxxxx 280.5 328.5 374.5 1609.0 / 3062 (52.55%)  71.9  
 5 sf14      170.5 191.5 189.0 230.5 xxxxx 294.5 345.5 1421.5 / 3063 (46.41%)  76.2  
 6 sf13      138.0 155.5 157.5 181.5 215.5 xxxxx 310.0 1158.0 / 3062 (37.82%)  73.1  
 7 sf12      104.0 115.0 129.0 135.5 165.5 200.0 xxxxx  849.0 / 3063 (27.72%)  67.8  

I don't have such a tool, this is the best I can offer, Moves is the average game length.

Ghppn likes this post

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


Posts : 2528
Join date : 2020-11-17
Location : Netherlands

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 7:59 pm

What I have in mind for a new tool, check the engines move scores on a defined value (say 3 pawns), if both sides agree (save that move number) and the game is won use that move number for shorties calculation.

Ghppn likes this post

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

mwyoung


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

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 8:32 pm

Now play the same match at 3+2 and more cores. And see the results.

That is my results also, and Fishtest. Until you use real world conditions.

Then all the nonsense evaporates.

Ghppn likes this post

Back to top Go down
Admin
Admin
Admin


Posts : 2528
Join date : 2020-11-17
Location : Netherlands

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 8:52 pm

I know, we discussed it several times.

Ghppn likes this post

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




Posts : 112
Join date : 2022-07-19

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptySun Jul 02, 2023 9:18 pm

Admin wrote:
Robin round, 10,000+ bullet games.

Code:
No. Name    Win Draw Loss Unf.  Score Games       %
---------------------------------------------------
  1 sf16   +1067 =1862 -133  *0 1998.0  3062   65.3%
  2 sf15.1 +902 =1927 -234   *0 1865.5  3063   60.9%
  3 sf15   +852 =1932 -279   *0 1818.0  3063   59.4%
  4 sf14.1 +633 =1952 -477   *0 1609.0  3062   52.5%
  5 sf14   +447 =1949 -667   *0 1421.5  3063   46.4%
  6 sf13   +284 =1748 -1030  *0 1158.0  3062   37.8%
  7 sf12   +128 =1442 -1493  *0  849.0  3063   27.7%

Total Games:   10719
White Wins:     2905 (27.1%)
Black Wins:     1408 (13.1%)
Draws:          6406 (59.8%)
thats a crazy gap from 15.1 to 16. That leela data is doing work. Is rebel solely trained on rebel data or does is use some leela too? Because iirc, ethereal and berserk and those other engine used some leela data.

Ghppn likes this post

Back to top Go down
pohl4711

pohl4711


Posts : 104
Join date : 2022-03-01
Location : Berlin

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptyMon Jul 03, 2023 10:39 am

Admin wrote:
Code:
No. Engine       1     2     3     4     5     6     7  Score  Games   Perc   Moves
-----------------------------------------------------------------------------------
 1 sf16      xxxxx 282.5 278.0 318.0 339.5 373.0 407.0 1998.0 / 3062 (65.25%)  75.0  
 2 sf15.1    227.5 xxxxx 268.5 299.5 319.5 354.5 396.0 1865.5 / 3063 (60.90%)  72.9  
 3 sf15      232.0 242.5 xxxxx 288.0 321.0 353.5 381.0 1818.0 / 3063 (59.35%)  74.7  
 4 sf14.1    192.0 210.5 223.0 xxxxx 280.5 328.5 374.5 1609.0 / 3062 (52.55%)  71.9  
 5 sf14      170.5 191.5 189.0 230.5 xxxxx 294.5 345.5 1421.5 / 3063 (46.41%)  76.2  
 6 sf13      138.0 155.5 157.5 181.5 215.5 xxxxx 310.0 1158.0 / 3062 (37.82%)  73.1  
 7 sf12      104.0 115.0 129.0 135.5 165.5 200.0 xxxxx  849.0 / 3063 (27.72%)  67.8  

I don't have such a tool, this is the best I can offer, Moves is the average game length.


I made such a tool some time ago:
https://www.sp-cc.de/files/short_games_analyzer_tool.7z

Example of the output (from the ReadMe-File):
Code:

************************************************************************************************
*** Rank 1: "Stockfish 220806 avx2"  All games: 7000 (2791 wins / 4189 draws / 20 losses)
*** Overall score: 69.79%     Overall draw-rate: 59.84%
*** Average game length: All: 75 moves (wins: 73 moves / draws: 76 moves / losses: 78 moves)
*** Avg. eng play white: All: 73 moves (wins: 72 moves / draws: 74 moves / losses: 83 moves)
*** Avg. eng play black: All: 77 moves (wins: 76 moves / draws: 77 moves / losses: 78 moves)
************************************************************************************************
*** all short wins   wins0-30   wins31-35  wins36-40  wins41-45  wins46-50  wins51-55  wins56-60
*** --------------------------------------------------------------------------------------------
***    [20.67%] =    [00.25%  +  00.43%  +  00.72%  +  01.22%  +  02.33%  +  05.52%  +  10.21%]
***    [   577] =    [     7  +      12  +      20  +      34  +      65  +     154  +     285]
*** --------------------------------------------------------------------------------------------
*** all short sacs  sacs queen   sacs 5+p   sacs 4p    sacs 3p    sacs 2p    sacs 1p(=pawn-unit)
*** --------------------------------------------------------------------------------------------
***    [06.06%] =    [00.00%  +  00.25%  +  00.64%  +  00.86%  +  01.79%  +  02.51%]
***    [   169] =    [     0  +       7  +      18  +      24  +      50  +      70]
************************************************************************************************
*** all short draws  draw0-30   draw31-35  draw36-40  draw41-45  draw46-50  draw51-55  draw56-60
*** --------------------------------------------------------------------------------------------
***    [37.07%] =    [04.01%  +  02.12%  +  03.20%  +  04.20%  +  06.33%  +  07.73%  +  09.48%]
***    [  1553] =    [   168  +      89  +     134  +     176  +     265  +     324  +     397]
************************************************************************************************
*** all short losses loss0-30   loss31-35  loss36-40  loss41-45  loss46-50  loss51-55  loss56-60
*** --------------------------------------------------------------------------------------------
***    [00.00%] =    [00.00%  +  00.00%  +  00.00%  +  00.00%  +  00.00%  +  00.00%  +  00.00%]
***    [     0] =    [     0  +       0  +       0  +       0  +       0  +       0  +       0]
************************************************************************************************

So, it is easy to see, how much short wins or losses an engine/player has (splitted into
7 move-intervals from (0-30) up tp (56-60) moves and with an overall percent score.).
And if (and how much) short wins of the engine/player contain a sacrifice.

Mention: The numbers in the 3 "Average game length:"-lines are the average length values of all
games, the engine has played in the source pgn-file, no matter how long the games were and if
the ended before endgame or not. Same for the Overall score and the Overall draw-rate values.

The engine/players in the SGA_full_statistics.txt-file are sorted by:
(percent all short wins)-(percent all short draws)-(percent all short losses)

Additionally, the tool writes 3 statistics.txt-files separately for wins/draws/losses:
SGA_wins_statistics.txt, SGA_draws_statistics.txt, SGA_losses_statistics.txt
In there, the engine/players are sorted by (percent all short wins) or (percent all short draws)
or (percent all short wins), but, because short wins are a good thing and short draws and losses
are not, the short draws and short losses are sorted reversed (smallest number at the top of the
file).
And the tool writes a pgn-file with all short wins (up to 60 moves) followed by all short draws:
shortgames.pgn (or shortgames_no_endgame.pgn)
And all short win games with a sacrifice are stored in the file short_sac_games.pgn
(or short_sac_games_no_endgame.pgn) (sorted top-down: Highest sacs first, lowest behind..) -
in this file the game-comments are deleted, because this makes the sac-search faster.


There are 2 tools: Short_Games_Analyzer and Short_Games_Analyzer_no_endgames
The only difference between these 2 tools is, that the "no_endgames" tool investigates only short
games, which ended before the endgame (for this, a material-check in each game is done. If an
endgame is found, then the game will not be part of the statistics-outputs).

Nezhman and Ghppn like this post

Back to top Go down
https://www.sp-cc.de
Chris Whittington




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

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptyMon Jul 03, 2023 10:58 am

That’s very complicated.
A simpler (Occams razor days better) could be move number when game was won (use a material trigger), because that is really the key, no?

Ghppn likes this post

Back to top Go down
Admin
Admin
Admin


Posts : 2528
Join date : 2020-11-17
Location : Netherlands

sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round EmptyMon Jul 03, 2023 5:37 pm

As for as I know pgn_extract (the base of the EAS tool) doesn't support move scores, so it can't be done that way.

Ghppn likes this post

Back to top Go down
http://rebel13.nl/
Sponsored content





sf16 robin round Empty
PostSubject: Re: sf16 robin round   sf16 robin round Empty

Back to top Go down
 
sf16 robin round
Back to top 
Page 1 of 1
 Similar topics
-
» Results of the 8 round top engines swiss
» Useful hack to get round high concurrency issues

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