FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hello Alayan,

I correted my mistakes.
Of course Ethereal (not ETChess).

:-)

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

to Elo ...

- Nirvana plays the first 500 games clearly over performance and found at the moment the right place.
- Rybka plays the first 800 games clearly over performance and more and more the different to Wasp, Pedone, Winter, Vajolet and Chiron goes down.
- Wasp plays the first 1.000 games around 10-15 Elo over performance. I think Pedone is a bit stronger. Wasp lost the first 5 games vs. Senpai for an example. Maybe Wasp can hold place 21 because 20 minutes games and 5 seconds per move "Fischer" is good for Wasp. With 40 moves in 20 minutes "repeatedly" the opponents have more time in endgames and Wasp more problems (bad for Wasp, weaknesses in endgames).

Chiron plays the first 1.000 games clearly under performance. In the last round 25 a good place for Chiron. I don't know whats the reason for the "bad" Chiron results. Normally Chiron should figthing for place 20 with Pedone and Wasp.

John and myself await place 23. With the current place 21 the _World of Wasp_ is fully OK but end of the tournament I am quite sure that Pedone is on place 20 and Wasp have to fight for place 21 with Vajolet2 and Winter.

Other ratings after 1.000 games looks OK for me.
A hard fight between Komodo and Houdini for place 2, or from Rodent and Monolight for place 40.

More information to Elo-trends can be found in the Excel files by Klaus.

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Code: Select all

   # Player                          :      Elo  Games  Score%  won  draw  lost  Points  Draw%  Error   OppAvg   OppE   OppD
   1 Stockfish 11 BMI2 x64           :  3268.26    212    80.4  129    83     0   170.5   39.2  35.32  3010.60  24.23   34.8
   2 Houdini 6.03 Pro x64            :  3197.81    202    71.3   86   116     0   144.0   57.4  28.22  3028.86  23.99   36.3
   3 Komodo 14.0 BMI2 x64            :  3167.82    238    67.2   82   156     0   160.0   65.5  24.04  3029.58  24.52   37.4
   4 Ethereal 12.25 PEXT x64         :  3156.78    162    68.2   59   103     0   110.5   63.6  28.59  3012.09  24.08   34.8
   5 SlowChess BC 2.2 x64            :  3127.82    220    62.5   57   161     2   137.5   73.2  21.96  3034.54  24.48   36.9
   6 Xiphos 0.6 BMI2 x64             :  3115.98    191    61.8   45   146     0   118.0   76.4  22.96  3027.03  23.77   37.1
   7 Defenchess 2.2 POP x64          :  3111.34    174    60.6   39   133     2   105.5   76.4  24.05  3030.85  24.34   36.9
   8 Andscacs 0.95 BMI2 x64          :  3104.62    205    60.0   48   150     7   123.0   73.2  22.79  3026.58  24.45   37.1
   9 Booot 6.4 POP x64               :  3104.13    246    61.0   54   192     0   150.0   78.0  20.58  3026.74  23.64   36.3
  10 Fire 7.1 POP x64                :  3097.20    184    57.6   30   152     2   106.0   82.6  21.39  3036.78  23.86   35.8
  11 Fritz 17 (Ginkgo) x64           :  3095.61    181    59.4   40   135     6   107.5   74.6  22.69  3024.00  24.00   34.9
  12 Fizbo 2.0 BMI2 x64              :  3095.35    179    59.2   47   118    14   106.0   65.9  23.33  3025.60  24.37   36.9
  13 Schooner 2.2 SSE x64            :  3092.96    185    59.5   36   148     1   110.0   80.0  21.60  3021.62  23.88   36.4
  14 Laser 1.7 BMI2 x64              :  3092.20    203    58.6   35   168     0   119.0   82.8  20.65  3025.80  24.00   37.8
  15 rofChade 2.3 BMI x64            :  3091.05    138    57.2   26   106     6    79.0   76.8  24.87  3039.35  23.98   33.6
  16 Wasp 4.00 Modern x64            :  3061.23    164    54.3   20   138     6    89.0   84.1  21.07  3031.57  23.61   35.6
  17 Shredder 13 x64                 :  3050.90    181    51.7   17   153    11    93.5   84.5  21.13  3036.09  24.01   35.8
  18 RubiChess 1.7.3 x64             :  3044.34    180    51.4   16   153    11    92.5   85.0  19.69  3033.52  24.38   36.5
  19 Arasan 22.0 BMI2 x64            :  3043.85    177    52.3   25   135    17    92.5   76.3  21.42  3032.14  23.87   34.6
  20 Pedone 2.0 BMI2 x64             :  3035.84    182    51.4   23   141    18    93.5   77.5  20.79  3026.67  23.67   35.7
  21 Chiron 4 x64                    :  3032.18    159    50.9   23   116    20    81.0   73.0  22.78  3028.43  24.25   34.3
  22 GullChess 3.0 BMI2 x64          :  3031.11    162    46.6   12   127    23    75.5   78.4  22.32  3052.35  23.74   32.5
  23 Vajolet2 2.8 BMI2 x64           :  3027.93    154    49.4   12   128    14    76.0   83.1  22.31  3034.59  23.96   35.2
  24 Igel 2.5.0 BMI2 x64             :  2998.13    193    44.0    7   156    30    85.0   80.8  20.58  3043.99  23.79   36.5
  25 Nirvanachess 2.4 POP x64        :  2989.05    203    42.9    9   156    38    87.0   76.8  20.94  3047.77  24.24   36.1
  26 Protector 1.9.0 x64             :  2988.42    154    43.2   13   107    34    66.5   69.5  24.22  3039.97  24.27   33.9
  27 Nemorino 5.00 BMI2 x64          :  2987.50    151    40.7   11   101    39    61.5   66.9  25.43  3057.91  24.59   33.1
  28 iCE 4.0 v853 Modern x64         :  2983.44    157    40.8    7   114    36    64.0   72.6  24.35  3054.51  23.90   35.0
  29 Critter 1.6a x64                :  2980.46    160    41.6   10   113    37    66.5   70.6  22.94  3046.41  23.85   33.5
  30 Equinox 3.30 x64                :  2975.52    171    38.6    6   120    45    66.0   70.2  23.99  3065.10  24.02   34.4
  31 Texel 1.07 BMI2 x64             :  2975.04    159    41.8   11   111    37    66.5   69.8  24.01  3040.40  24.10   36.0
  32 Fritz 16 (Rybka) x64            :  2970.99    201    38.3   11   132    58    77.0   65.7  22.76  3058.37  23.67   35.3
  33 Demolito 2020-05-14 PEXT x64    :  2965.55     90    37.2    9    49    32    33.5   54.4  33.91  3062.33  24.07   30.5
  34 Hannibal 1.7 x64                :  2951.74    130    35.0    4    83    43    45.5   63.8  29.80  3065.00  24.08   31.7
  35 Winter 0.8 x64                  :  2949.01    192    35.9   11   116    65    69.0   60.4  24.30  3054.57  24.05   35.8
  36 Rodent IV 0.22 POP x64          :  2948.44    167    36.8    4   115    48    61.5   68.9  24.38  3053.44  24.04   34.3
  37 Minic 2.33 x64                  :  2924.49    190    33.7    6   116    68    64.0   61.1  25.13  3049.74  23.72   35.7
  38 Senpai 2.0 BMI2 x64             :  2916.01    181    32.6    6   106    69    59.0   58.6  26.86  3055.42  23.86   35.0
  39 Monolith 2 PEXT x64             :  2909.77    198    32.8    3   124    71    65.0   62.6  24.38  3048.26  23.76   36.4
  40 Combusken 1.2.0 x64             :  2889.90    205    28.0    4   107    94    57.5   52.2  28.33  3060.62  23.81   36.0
  41 SmarThink 1.98 AVX2 x64         :  2885.18    189    26.5    7    86    96    50.0   45.5  30.23  3067.11  23.87   33.9

White advantage = 41.96 +/- 2.71
Draw rate (equal opponents) = 100.00 % +/- 1.12
The different from Stockfish to Wasp = 367 Elo.
The different for the first 59 moves from Stockfish to Wasp = 207 Elo.
The strength of Wasp is clearly to see!

This and other stats can be found in the *.zip file.
At the moment around 100Mb.

But after all ...
Wasp can win clearly more short attacking games with lesser time vs. the same group of opponents!

After FCP Tourney-2020 I will start the same tourney with 5x lesser time.
Need around 1 months with 4 minutes game + 1 second move.
So we can compare the results and stats to all of the 41 engines!

Best
Frank

Enough chess and QuisiQuasi blabla?!?!
Next updates will be available start of November 2020.

Have a good time and many fun in testing chess programs!
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

I am for 5 weeks 1 time only at home.
The reason for the break.

Unfortunately, I lost 10 days because the USB-Stick, connected with the monitor, cann't save the games.
Not thinking on it to hold the monitor power=on or to connected the USB-Stick with the computer.

The mistakes for such a big power tournament comes from a human.
Maybe the biggest chess touranment in chess history every played ... think so!

But I added the results from the last 4 weeks on my website.
Still running is now round 37/50.

1.440 of 2.000 games for each of the 41 engines are ready = 29.520 of 41.000 games with fischer time control 20 minutes per game / 5 seconds per move on the i9-10900 Intel 5.3Ghz system.

Frank's Chess Page:
www.amateurschach.de

Have a nice day!

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

Excel files updated by Klaus Wlotzka (last 11 rounds).

In the next days the tournament organization for FCP-Tourney-2021 will be online.
I wrote here messages in an other thread about it.

Go Wasp go ...

:-)

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

the results from Round 37/50 are online.

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

the results from round 38/50 are online.

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

the results from round 39/ 50 are online.

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

now the results from round 40/50 are available.

New material in Excel files by Klaus Wlotzka are added.

How many games for +- Elo, calculated with our tournament rules and 41 participate engines (each one against ech others). I think nothing is new but interesting to study.

Best
Frank
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k

Post by Frank Quisinsky »

Hi there,

a lot of Excel statistics are available to the still running FCP Tourney-2020.

On the detail page you can jump to all information around the tourney!

Two important downloads:
1. My files, games / stats (after 1.600 of 2.000 games = 80%, around 150Mb).
2. Excel files by Klaus Wlotzka (round 7Mb).

Detail page:
http://www.amateurschach.de/main/_fcp-tourney-2020.htm

Here what I wrote before ...
At the moment Klaus is working on statistics to the question:
How many games needed!
I think 800-1000 games for a good Elo calculation for 41 participate engines, each one vs. each other are enough.

Best
Frank

Image