FCP-Tourney-2024-MA ... game in 66 min. + 6 sec. / move

[Frank Quisinsky]

Not read the readme from Dragon (honestly).
With personality "aggr" NN is not working.
Thanks to Peter Martan.

I changed from Dragon 3.3 NN (aggr.) to the same I am using for FCP-Tourney-2024 ...
Dragon 3.3 (Komodo) with Contempt = 0

Thanks Peter!

Best
Frank

[Frank Quisinsky]

Hi there,

two updates:
Seer 2.8.0 NN replaced Seer 2.7.0 NN
Andscacs 0.95 replaced Andscacs 0.95.123 (I forgot that the last release version has a better move-average).

No more engine updates possible!
Tourney can run now for a while.

I created a result-site, more or less an overview!
https://www.amateurschach.de/main/_ma.htm

Best
Frank

[Frank Quisinsky]

Hi there,

Round 01 landed: 1. Dragon, 2. Seer, 3. RubiChess ... 44. Fizbo

Move-average = 89,5
draws = 60,89

1. https://www.amateurschach.de/fling/index.html ... cockpit / replay-zone
2. https://www.amateurschach.de/main/_ma.htm ... flight plan /overview to results

Passengers (engines), prepare for takeoff ...
Estimated arrival time for round 02 = January 09th, 2024
Wind = 35-70km/h, temperature = 9,5, Air humidity = 83%
Satisfaction of the tournament organisers: 70% ... go Wasp v6.63 go, the big brother v6.50 was better in round-01.

Best
Frank

[Frank Quisinsky]

Hi there,

50 first dates ... results from date 02:
1. Dragon, 2. CSTal, 3. RubiChess ... 44. Andscacs

Move-average = 88,55 (I await 87-88)
Draw-quote = 60,78% (around 2% higher as I await)

Now round 03 is still running ...

Wind = 15km/h, temperature = -3,6, Air humidity = 57%
Satisfaction of the tournament organisers: 75,75% ... go Wasp v6.63 go, the big brother v6.50 only with a very light advantage.

Best
Frank

[Frank Quisinsky]

Hi there,

50 first dates ... results from date 03:
1. Dragon, 2. Seer, 3. Uralochka ... 44. Xiphos

Move-average = 88,66
Draw-quote = 61,10%

Now round 04 is still running ...

Wind = 20km/h, temperature = -1,6, Air humidity = 66%
Satisfaction of the tournament organisers: 80,00% ... go Wasp 6.63 go ... but it is very hard to beat the strong release version 6.50 with a longer time-control. Stockfish 200731 dev lost only one game in round 03 ... vs. Wasp!

Very interesting to see ...
Stockfish 16 NN is around 5-10 Elo stronger with the time-control 40/20 as Dragon 3.3 NN.
So, the different from Stockfish 200731 dev (without NN) to Stockfish 16 NN with the time-control 66+6 + 6-man is around 100 Elo, not 150, not 200, not 300 or 400 (what I read here often) ... the different is 100 Elo. And with longer time-controls as 66+6 the difference continues to fall. But let us wait of more games. I can test later in detail if SF17 is available, means I can start a round-robin vs. this group of 44 engines.

Again the links:

Best start point is my Replay-Zone:
https://www.amateurschach.de/fling/index.html

The new result-page:
https://www.amateurschach.de/main/_ma.htm

Best
Frank

[Frank Quisinsky]

Hi there,

results from round 04:
1. Dragon, 2. CSTal, 3. Caissa ... 44. Shredder

Move-average = 88,61
Draw-quote = 60,41

Now round 05 is still running ...

Wind = 5km/h, temperature = -3,9, Air humidity = 70%
Satisfaction of the tournament organisers: 65,00% ... the fun factor is clearly higher with a game in x minutes per game + x seconds per move. But I don't like the time control because it seems that different engines have not a good time-management. With such a long time control you can see better than with a blitz time control. So in the end ... I am more a fan of x moves in x minutes.

Stockfish 200731 dev (without NN) is playing a strong round again. After 172 games only 6 lost games. The difference to Dragon 3.3 NN with such a long time control is at the moment with Ordo calculation only 96 Elo. If available I will add Stockfish 17 NN later. I am more and more sure that the Neural-Network advantage is not more than 120-130 Elo with such a long time control.

It seems that the current Seer 2.8.0 NN version get a booster for longer time-controls ... must be observed.

Code: Select all

FCP-Tourney-2024-MA after round 04
03.784 games

   # Player                    :      Elo  Games  Score%  won  draw  lost  Points  Draw%  Error   OppAvg   OppE   OppD
   1 Dragon 3.3 NN (Komodo)    :  3492.49    172    73.5   81    91     0   126.5   52.9  32.84  3293.71  29.31   43.0
   2 Seer 2.8.0 NN             :  3470.87    172    71.2   76    93     3   122.5   54.1  32.72  3294.22  29.31   43.0
   3 CSTal 2.00 NN             :  3465.61    172    70.6   72    99     1   121.5   57.6  32.17  3294.34  29.32   43.0
   4 RubiChess 20230918 NN     :  3447.58    172    68.6   64   108     0   118.0   62.8  31.48  3294.76  29.34   43.0
   5 Caissa 1.15 NN            :  3442.53    172    68.0   63   108     1   117.0   62.8  29.49  3294.88  29.39   43.0
   6 Uralochka 3.40a NN        :  3427.63    172    66.3   60   108     4   114.0   62.8  29.46  3295.22  29.39   43.0
   7 Revenge 3.0 NN            :  3417.87    172    65.1   58   108     6   112.0   62.8  28.66  3295.45  29.41   43.0
   7 Clover 6.1 NN             :  3417.87    172    65.1   56   112     4   112.0   65.1  29.46  3295.45  29.39   43.0
   9 Igel 3.5.0 NN             :  3415.45    172    64.8   52   119     1   111.5   69.2  29.59  3295.50  29.38   43.0
  10 Rebel EAS NN              :  3413.04    172    64.5   54   114     4   111.0   66.3  29.88  3295.56  29.38   43.0
  10 Alexandria 5.1.0 NN       :  3413.04    172    64.5   57   108     7   111.0   62.8  29.44  3295.56  29.39   43.0
  12 Arasan 24.0 NN            :  3398.73    172    62.8   48   120     4   108.0   69.8  29.38  3295.89  29.39   43.0
  13 Stockfish 200731 dev      :  3396.37    172    62.5   49   117     6   107.5   68.0  28.25  3295.95  29.42   43.0
  14 SlowChess Blitz 2.9 NN    :  3380.00    172    60.5   44   120     8   104.0   69.8  28.94  3296.33  29.40   43.0
  15 Carp 3.0.1 NN             :  3370.76    172    59.3   43   118    11   102.0   68.6  28.01  3296.54  29.42   43.0
  16 Fritz 19 NN (Gingko)      :  3354.75    172    57.3   45   107    20    98.5   62.2  28.58  3296.92  29.41   43.0
  16 Minic 3.39 NN             :  3354.75    172    57.3   41   115    16    98.5   66.9  27.77  3296.92  29.43   43.0
  18 Fire 9.2 NN               :  3350.21    172    56.7   34   127    11    97.5   73.8  26.95  3297.02  29.45   43.0
  18 Altair 6.0.0 NN           :  3350.21    172    56.7   42   111    19    97.5   64.5  27.49  3297.02  29.43   43.0
  20 Velvet 6.0.0 NN           :  3327.65    172    53.8   37   111    24    92.5   64.5  27.05  3297.55  29.44   43.0
  21 Wasp 6.50 NN              :  3323.16    172    53.2   33   117    22    91.5   68.0  26.60  3297.65  29.45   43.0
  22 Wasp 6.63 NN dev          :  3318.68    172    52.6   36   109    27    90.5   63.4  27.44  3297.76  29.43   43.0
  23 Nemorino 6.11 NN dev      :  3300.79    172    50.3   34   105    33    86.5   61.0  26.47  3298.17  29.46   43.0
  24 BlackCore 6.0 NN          :  3291.85    172    49.1   29   111    32    84.5   64.5  26.67  3298.38  29.45   43.0
  25 Texel 1.10 NN             :  3287.37    172    48.5   32   103    37    83.5   59.9  27.14  3298.48  29.44   43.0
  26 Devre 4.0 NN              :  3285.14    172    48.3   22   122    28    83.0   70.9  27.91  3298.54  29.42   43.0
  27 Chess.cpp 4.0 NN          :  3267.19    172    45.9   30    98    44    79.0   57.0  27.38  3298.95  29.44   43.0
  28 Pawn 2.0 NN               :  3255.92    172    44.5   21   111    40    76.5   64.5  28.02  3299.21  29.42   43.0
  29 Marvin 6.2.0 NN           :  3249.13    172    43.6   13   124    35    75.0   72.1  27.23  3299.37  29.44   43.0
  30 Tucano 11.00.1 NN         :  3242.31    172    42.7   16   115    41    73.5   66.9  27.52  3299.53  29.43   43.0
  31 Hakkapeliitta 3.0 NNSV    :  3200.61    172    37.5   12   105    55    64.5   61.0  30.47  3300.50  29.36   43.0
  32 Mantissa 3.7.2 NN         :  3179.05    172    34.9   12    96    64    60.0   55.8  29.25  3301.00  29.39   43.0
  33 Midnight 8 NN             :  3176.61    172    34.6   15    89    68    59.5   51.7  29.21  3301.06  29.39   43.0
  34 Booot 6.4                 :  3171.72    172    34.0   13    91    68    58.5   52.9  29.63  3301.17  29.38   43.0
  35 Xiphos 0.6                :  3166.79    172    33.4   14    87    71    57.5   50.6  30.68  3301.29  29.36   43.0
  35 DanaSah 9.1 NN            :  3166.79    172    33.4   12    91    69    57.5   52.9  30.69  3301.29  29.36   43.0
  37 Winter 2.0 NN             :  3156.83    172    32.3    7    97    68    55.5   56.4  31.02  3301.52  29.35   43.0
  38 Nalwald 18 NN             :  3154.31    172    32.0    7    96    69    55.0   55.8  30.84  3301.58  29.36   43.0
  39 Andscacs 0.95             :  3144.14    172    30.8   10    86    76    53.0   50.0  31.74  3301.81  29.33   43.0
  40 Shredder 13               :  3141.57    172    30.5   13    79    80    52.5   45.9  29.91  3301.87  29.38   43.0
  40 Laser 1.7                 :  3141.57    172    30.5   11    83    78    52.5   48.3  32.07  3301.87  29.33   43.0
  42 Chiron 5.01               :  3138.98    172    30.2   12    80    80    52.0   46.5  32.25  3301.93  29.32   43.0
  43 Hiarcs 15.2 (aggr.)       :  3136.39    172    29.9   11    81    80    51.5   47.1  31.54  3301.99  29.34   43.0
  44 Fizbo 2.0 NN              :  3117.85    172    27.9    7    82    83    48.0   47.7  33.83  3302.43  29.29   43.0

White advantage = 53.34 +/- 3.36
Draw rate (equal opponents) = 83.21 % +/- 1.11

Replay-Zone in live mode, games, link to the current tournament-table and so on ...
https://www.amateurschach.de/fling/index.html

Best
Frank

[Frank Quisinsky]

Hi there

first time I looked at the move average stats after round 04:
It looks much better than in my 40/20 tournament.

I am adding engines with very different styles to the tournament.
Interesting is the "superstar" from FCP-Tourney-2020 ... Booot 6.4.

Booot 6.4 can beat the move average for wins of Velvet and Texel!
The style of Booot 7.2 NN is very boring for me, longest move-average in my 40/20 tournament and a completely different program with neural network to classical eval. These statistics could not be more different.

All in all... the move average stats look normal.

Three engines will be disqualified, the first after round 6, the second after round 8 and the third after round 10.
In danger at the moment are Andscacs 0.95 and Clover 6.1 NN (engines with highest move average for wins and draws). End of the day the tournament ended with 41 engines.

Have a look at the move average for Revenge, SlowChess.
Here I made configuration mistakes in my 40/20 tournament, both are not playing with contempt=0. Now everything looks wonderful.

But most interesting in my opinion is Seer.
It seems that the engine gets a big boost in playing strength with longer time controls. Of course, not enough games for such statements.

The most fun I have is watching games:
When I have time, I watch and have seen a lot of nice games. Honestly, I have the most fun with the DanaSah games and I really like seeing Hakkapeliitta again. From Revenge I saw three of the fast wins live. Wasp is stronger in the endgames, can see more with the long time control.

Enough, it's a really nice tournament, but I often think that 30% of the engines have bad time management. Start playing Blitz with the +6 after about 60-70 moves. Blitz should be start 10-15 moves later.

Best
Frank

Code: Select all

          Name                       Games     Win     Draw     Lose       Pts         S-B         %    wins-55m  lost-55m  AV-wins  AV-draws  AV-all

01. Dragon 3.3 NN (Komodo)       :   172  :   81+  :   91=  :    0-  :   126.5  :  10481.00  :  73.55%      3        0        81        91       86
02. Seer 2.8.0 NN                :   172  :   76+  :   93=  :    3-  :   122.5  :   9808.75  :  71.22%      2        0        89        81       85
03. CSTal 2.00 NN                :   172  :   72+  :   99=  :    1-  :   121.5  :   9748.50  :  70.64%      1        0        89        78       82
04. RubiChess 20230918 NN        :   172  :   64+  :  108=  :    0-  :   118.0  :   9582.00  :  68.60%      1        0        85        82       83
05. Caissa 1.15 NN               :   172  :   63+  :  108=  :    1-  :   117.0  :   9509.25  :  68.02%      2        0        86        92       90
06. Uralochka 3.40a NN           :   172  :   60+  :  108=  :    4-  :   114.0  :   9129.75  :  66.28%      4        0        85        95       91
07. Clover 6.1 NN                :   172  :   56+  :  112=  :    4-  :   112.0  :   8941.00  :  65.12%      0        1        94       102       99
08. Revenge 3.0 NN               :   172  :   58+  :  108=  :    6-  :   112.0  :   8906.50  :  65.12%      7        0        80        85       84
09. Igel 3.5.0 NN                :   172  :   52+  :  119=  :    1-  :   111.5  :   9110.50  :  64.83%      5        0        88        80       82
10. Alexandria 5.1.0 NN          :   172  :   57+  :  108=  :    7-  :   111.0  :   8911.25  :  64.53%      0        2        90        99       96
-----------------------------------------------------------------------------------------------------------------------------------------------------
11. Rebel EAS NN                 :   172  :   54+  :  114=  :    4-  :   111.0  :   8863.00  :  64.53%      2        0        82        80       81
12. Arasan 24.0 NN               :   172  :   48+  :  120=  :    4-  :   108.0  :   8746.50  :  62.79%      1        0        92        98       96
13. Stockfish 200731 dev         :   172  :   49+  :  117=  :    6-  :   107.5  :   8679.00  :  62.50%      4        0        79        95       91
14. SlowChess Blitz 2.9 NN       :   172  :   44+  :  120=  :    8-  :   104.0  :   8421.25  :  60.47%      1        0        83        89       88
15. Carp 3.0.1 NN                :   172  :   43+  :  118=  :   11-  :   102.0  :   8127.25  :  59.30%      3        0        83        90       88
16. Minic 3.39 NN                :   172  :   41+  :  115=  :   16-  :    98.5  :   7716.50  :  57.27%      0        0       101        96       97
17. Fritz 19 NN (Gingko)         :   172  :   45+  :  107=  :   20-  :    98.5  :   7698.75  :  57.27%      0        0        83        88       87
18. Fire 9.2 NN                  :   172  :   34+  :  127=  :   11-  :    97.5  :   7822.00  :  56.69%      0        0        84        89       89
19. Altair 6.0.0 NN              :   172  :   42+  :  111=  :   19-  :    97.5  :   7669.00  :  56.69%      0        1        98       100       98
20. Velvet 6.0.0 NN              :   172  :   37+  :  111=  :   24-  :    92.5  :   7283.50  :  53.78%      3        0        79        84       85
-----------------------------------------------------------------------------------------------------------------------------------------------------
21. Wasp 6.50 NN                 :   172  :   33+  :  117=  :   22-  :    91.5  :   7201.75  :  53.20%      1        0        89        76       82
22. Wasp 6.63 NN dev             :   172  :   36+  :  109=  :   27-  :    90.5  :   7033.00  :  52.62%      2        0        92        74       82
23. Nemorino 6.11 NN dev         :   172  :   34+  :  105=  :   33-  :    86.5  :   6778.50  :  50.29%      0        1        94        97       93
24. BlackCore 6.0 NN             :   172  :   29+  :  111=  :   32-  :    84.5  :   6625.50  :  49.13%      0        3       100        95       93
25. Texel 1.10 NN                :   172  :   32+  :  103=  :   37-  :    83.5  :   6400.50  :  48.55%      3        0        78        99       94
26. Devre 4.0 NN                 :   172  :   22+  :  122=  :   28-  :    83.0  :   6516.25  :  48.26%      0        0        89        95       94
27. Chess.cpp 4.0 NN             :   172  :   30+  :   98=  :   44-  :    79.0  :   6008.75  :  45.93%      0        3        95        78       81
28. Pawn 2.0 NN                  :   172  :   21+  :  111=  :   40-  :    76.5  :   5972.25  :  44.48%      0        1       101        93       91
29. Marvin 6.2.0 NN              :   172  :   13+  :  124=  :   35-  :    75.0  :   5968.00  :  43.60%      1        1        95        86       87
30. Tucano 11.00.1 NN            :   172  :   16+  :  115=  :   41-  :    73.5  :   5741.00  :  42.73%      0        3        92        89       88
-----------------------------------------------------------------------------------------------------------------------------------------------------
31. Hakkapeliitta 3.0 NNSV       :   172  :   12+  :  105=  :   55-  :    64.5  :   5172.00  :  37.50%      1        1        82        91       88
32. Mantissa 3.7.2 NN            :   172  :   12+  :   96=  :   64-  :    60.0  :   4584.25  :  34.88%      0        2        97        80       82
33. Midnight 8 NN                :   172  :   15+  :   89=  :   68-  :    59.5  :   4573.25  :  34.59%      0        9        88        94       91
34. Booot 6.4                    :   172  :   13+  :   91=  :   68-  :    58.5  :   4564.75  :  34.01%      2        2        74        80       83
35. Xiphos 0.6                   :   172  :   14+  :   87=  :   71-  :    57.5  :   4466.25  :  33.43%      1        0        93        98       92
36. DanaSah 9.1 NN               :   172  :   12+  :   91=  :   69-  :    57.5  :   4302.00  :  33.43%      2        4        82        85       85
37. Winter 2.0 NN                :   172  :    7+  :   97=  :   68-  :    55.5  :   4413.00  :  32.27%      1        1        87        89       90
38. Nalwald 18 NN                :   172  :    7+  :   96=  :   69-  :    55.0  :   4256.50  :  31.98%      0        5       102        95       91
39. Andscacs 0.95                :   172  :   10+  :   86=  :   76-  :    53.0  :   4079.00  :  30.81%      1        1        96       104       98
40. Shredder 13                  :   172  :   13+  :   79=  :   80-  :    52.5  :   3992.50  :  30.52%      0        6        95        95       91
-----------------------------------------------------------------------------------------------------------------------------------------------------
41. Laser 1.7                    :   172  :   11+  :   83=  :   78-  :    52.5  :   3964.25  :  30.52%      0        0        77        95       91
42. Chiron 5.01                  :   172  :   12+  :   80=  :   80-  :    52.0  :   3872.50  :  30.23%      0        4        90        91       88
43. Hiarcs 15.2 (aggr.)          :   172  :   11+  :   81=  :   80-  :    51.5  :   3879.50  :  29.94%      0        2        86        92       93
44. Fizbo 2.0 NN                 :   172  :    7+  :   82=  :   83-  :    48.0  :   3783.00  :  27.91%      0        2        99        87       87


White Wins =    989 ( 26.14% )
Draws      =  2.286 ( 60.41% )
Black Wins =    509 ( 13.45% )
Average    = 177.22 ( 88,61 moves )

[Frank Quisinsky]

Hi there,

If I look in my blitz results of newer releases, Starzix and the last Akimbo are interesting for this tournament. That's right, two people asked for possible updates. Unfortunately, any changes here need too much time. I will not change anything. After round 10 I have three free places in the tournament table. I will use one of them for the next Stockfish release. Might be interesting to see how much more Elo SF can give with neural-network and longer time controls.

This tournament = state at the end of 2023 with engines I think the move average is low. In addition some of the older and in my opinion interesting engines with classic eval.

Interesting can be engines with different playing styles and a low move average to test other / new releases.
Example: 7 with around 3400 or higher, 7 with 3300 or higher, 7 with 3200 or higher for testing new releases with a test set of positions.

Example: Texel, Velvet and Wasp don't play the same way. Only with a quick view we can think it. Velvet is stronger than Texel and Wasp for attacking moves in the middle of the board, Texel has better statistics with white pieces than Wasp and Wasp has better statistics with black pieces than Texel. Furthermore is Wasp most aggressive with pawn moves, open the position.

Revenge, SlowChess and Uralochka are also attackers with different styles but stronger than Texel, Wasp and Velvet. Important are the programs where I can find no weaknesses ... the all-rounders ... like RubiChess.

The problem is ...
The attackers like open positions and often the pawn structures after attacking chess are not good. So many games are lost in the late middlegames. This group of engines must have weaknesses and they are often to be found in the earlier endgames.

To see ... wow, Wasp has so many quick wins is good for first time players or for people like me who work on opening analysis or like quick wins. But such engines often have weaknesses in endgames. Another good example is the older Spark or even Booot 6.4 / 6.5.

Important to have Caissa or Seer in the group, no attackers, but very strong in the late midgame and early endgame. What I like to write is, possible to find out a strong group of 21 Engines, can build the reference for testing all the other new releases with longer time controls.

With engines that like to produced "chewing gum draws" I lost too much time in testing with time controls like x moves in x minutes. So I also have to produce blitz games, because the information I am looking for I cannot read in rating list systems like CEGT or CCRL. Stefan Pohl tests with unbalanced positions and only the strongest.

Again, I will nothing change in that 66+6 tournament!
And again ... move-average is really important because much more games for interesting statistics can be produced with engine comes with a low move-average.

So if I have a test-group of engines (21 reference engines) all of the engines like to produce "No chewing-gums draws" it's not important what the new releases do if they have to play against the group. End of the day I lost not to many time for testing and can produced some interesting stats later. So the 21 reference engines must have a low move-average and must have different playing styles. Not easy to find out it.

Best
Frank

[Frank Quisinsky]

It is a completely new way of testing engines that I will be taking.
Not with my current tournaments, but what I want to do in the future.

To have a strong test set with 50 balanced positions (working on it, need more stats here).
To build a team of 21 engines that can test the others, new releases.

When new engine releases, produce games with very long move averages ... must not be bad. It is not my intention to give such information. Every engine is interesting to me. But I am very sure I can see more in the stats if I build a test team of engines (different styles, perfect move-average and so one) vs. all the new releases.

More and more powerful engines are available.
I have lost the overview for a while.

So I need a new plan to find out more about it, and that very quickly.
I've been doing this for too long and I've got tester's disease. I don't want to miss anything interesting.

So, I am working on it and have no problems to give the information why I do this and that.

[Frank Quisinsky]

To the test-team ...
I am thinking on ... all with a good move-average and interesting and quiet different playing-styles. All are very danger not easy to beat ... OK DanaSah lose more games as the others fast. But the style is just fantastic.

Newer releases must have a difficult life.



01. RubiChess 20240112 NN
02. Seer 2.8.0 NN
03. Caissa 1.16 NN
04. Revenge 3.0 NN
05. Uralochka 3.40 NN
06. Igel 3.5.0 NN
07. Arasan 24.1 NN
08. SlowChess 2.9 NN
09. Fritz 19 NN (Gingko) NN ... first Fritz version with a really interesting style of play
10. Fire 9.2 NN
11. Starzix 4.0 NN
12. Wasp 6.50 NN
13. Velvet 6.0.0 NN
14. Texel 1.11 NN
15. Pawn 3.0 NN
16. Nemorino 6.11 NN
17. chess.cpp 4.0 NN ... like the engine a lot, very balanced style, often with interesting ideas in late midgames
18. Marvin 6.20 NN
19. DanaSah 9.1 NN
20. Hiarcs 15.2
21. Booot 6.4

---

Yes, not in the field are Rebel or CSTal or Dragon ...
For different reasons! Rebel or CSTal produced from time to time GUI hang-ups.
To add Dragon made no sense, must not have the strongest engines in the field.

---

Now a new Stormphrax 5.0.0 NN is out ...
And have to play vs. this 21 engines the same set of balanced opening positions = 2100 games
The results goes in a rating list without Elo. I am working only with points. Ratings are so boring!

Now a new Marvin 7.00 NN is out ...
So Marvin have to play vs. the same 21 engines, in one match vs. Marvin 6.20 NN.
The results goes in a rating list.

The 21 "Team-worker-engines" are not in the rating-list!
Used only as reference test-engines!

Later in the rating list I will add some interesting stats, not Elo.
I have here different ideas what I can do and working on it.

Thats the idea I had since sommer 2022 after I closed for some reasons my KI-Rating List.
Time control will be 40 in 8 + 2 seconds (time control I like most for testing engines).

I can start with it middle of the year.
I am not sure to 100% with the test-group ... but I am thinking exactly on this group of engines.
Furthermore, I need to different points more results and time for thinking about it.

Best
Frank