GRL - test runs

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

GRL - test runs

Post by Rebel »

From now on I will post ongoing matches and results in this thread instead of the new engines section in the main section.

Currently Berserk 4.4.0

http://rebel13.nl/a/grl.htm
90% of coding is debugging, the other 10% is writing bugs.
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: GRL - test runs

Post by Rebel »

Code: Select all

Gambit Rating List : Sun Jul 11 04:24:47 2021
Running      : Gauntlet Berserk 4.4.0
Time Control : Time control 40/120
Games        : 1000

Results from file gauntlet-berserk-440.pgn:

No. Name            Win Draw Loss Unf.  Score Games       %
-----------------------------------------------------------
  1 Berserk 4.4.0  +265 =461 -274   *0  495.5  1000   49.5%
  2 Ethereal 12.00  +65  =97  -38   *0  113.5   200   56.8%
  3 Pedone 3.0      +64  =79  -57   *0  103.5   200   51.8%
  4 Komodo 10       +58  =89  -53   *0  102.5   200   51.2%
  5 rofChade 2.3    +49  =96  -55   *0   97.0   200   48.5%
  6 Booot 6.5       +38 =100  -62   *0   88.0   200   44.0%

Total Games:    1000
White Wins:      271 (27.1%)
Black Wins:      268 (26.8%)
Draws:           461 (46.1%)
Unfinished:        0 (0.0%)

Estimated elo gain for Berserk_4.4.0
Elo pool : 3271
Berserk 4.3.0 : 3231.0
Berserk_4.4.0 : 3268.4
Difference : 37.4

Berserk 4.4.0 : + 37
90% of coding is debugging, the other 10% is writing bugs.
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: GRL - test runs

Post by Rebel »

Provisional Rating List

Code: Select all

Gambit Rating List         : Provisional

   1 PLAYER                :  RATING  ERROR  POINTS  PLAYED   (%)  CFS(%)     W     D     L  D(%)
   2 Stockfish 14          :  3683.0   11.9  2635.0    3300    80      99  2040  1190    70    36
   3 Komodo-Dragon         :  3582.4   16.1  1998.5    3000    67      61  1317  1363   320    45
   4 Lc0-v27               :  3530.6   27.9   501.0     800    63      97   307   388   105    49
   5 SlowChess 2.6         :  3419.9   15.4  1229.5    2600    47      92   632  1195   773    46
   6 RubiChess 2.2         :  3408.2   15.1   880.0    2000    44      84   407   946   647    47
   7 Pedone 3.1            :  3360.0   10.5  1389.5    3100    45      74   767  1245  1088    40
   8 Igel 3.0.5            :  3356.0   15.1  1150.0    3100    37      53   478  1344  1278    43
   9 Ethereal 12.75        :  3355.1   21.3  1025.0    2900    35      66   428  1194  1278    41
  10 Nemorino 6.00         :  3309.4   16.7  1129.0    2900    39      60   620  1018  1262    35
  11 Berserk 4.4.0         :  3267.6   27.0   495.5    1000    50      78   265   461   274    46
New entries
Bitgenie 8
RubiChess 2.2
Koivisto 4.83
Berserk 4.4.0
Marvin 5.1

Full list - http://rebel13.nl/a/grl-provisional.txt
90% of coding is debugging, the other 10% is writing bugs.
jhonnold
Posts: 117
Joined: Wed Feb 17, 2021 3:16 pm
Full name: Jay Honnold

Re: GRL - test runs

Post by jhonnold »

Thanks for the testing, I really appreciate it!
Madeleine Birchfield
Posts: 512
Joined: Tue Sep 29, 2020 4:29 pm
Location: Dublin, Ireland
Full name: Madeleine Birchfield

Re: GRL - test runs

Post by Madeleine Birchfield »

Ed, do you plan on also maintaining a list with all engine versions on it? It is fairly common in other rating lists to have two separate lists, one for all engines versions and another for the strongest engine version.
jhonnold
Posts: 117
Joined: Wed Feb 17, 2021 3:16 pm
Full name: Jay Honnold

Re: GRL - test runs

Post by jhonnold »

Madeleine Birchfield wrote: Sun Jul 11, 2021 4:15 pm Ed, do you plan on also maintaining a list with all engine versions on it? It is fairly common in other rating lists to have two separate lists, one for all engines versions and another for the strongest engine version.
I see both a "Best Version": https://rebel13.nl/grl-best-40-2.html
and a "Full Version": https://rebel13.nl/grl-40-2.html
Madeleine Birchfield
Posts: 512
Joined: Tue Sep 29, 2020 4:29 pm
Location: Dublin, Ireland
Full name: Madeleine Birchfield

Re: GRL - test runs

Post by Madeleine Birchfield »

jhonnold wrote: Sun Jul 11, 2021 4:33 pm
Madeleine Birchfield wrote: Sun Jul 11, 2021 4:15 pm Ed, do you plan on also maintaining a list with all engine versions on it? It is fairly common in other rating lists to have two separate lists, one for all engines versions and another for the strongest engine version.
I see both a "Best Version": https://rebel13.nl/grl-best-40-2.html
and a "Full Version": https://rebel13.nl/grl-40-2.html
Huh, must have missed it. Thanks.
User avatar
j.t.
Posts: 239
Joined: Wed Jun 16, 2021 2:08 am
Location: Berlin
Full name: Jost Triller

Re: GRL - test runs

Post by j.t. »

I am not a native English speaker, but maybe it would help to rename "Full Version" to something like "All Versions" or "Complete List".
Another thing: What does the teal color highlight of some engines mean?
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: GRL - test runs

Post by Rebel »

Code: Select all

Blue                : New Entries
Text above the buttons.

Complete List sound fine to me also.
90% of coding is debugging, the other 10% is writing bugs.
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: GRL - test runs

Post by Rebel »

Code: Select all

Gambit Rating List
Running      : Gauntlet Seer 2.1.0
Time Control : Time control 40/120
Games        : 1800
http://rebel13.nl/a/grl.htm

Maybe a new 3200 engine....
90% of coding is debugging, the other 10% is writing bugs.