FCP: Sting SF 4.8.4 x64 is still running ...

[lech]

Frank, thank you for your effort.
The end result 2931 is very good for Sting SF 4.8.4.
It means that my idea to make complete engines and the way to get it are correct.

[Frank Quisinsky]

Code: Select all

 Calculated with Ordo 0.9.9
 Stats after 100, 200 ... games!       Elo   Games  Score  Draws   White Black Points  w/ d/ l
 06. Sting SF 0.8.4 x64                2931  1.600  63.0%  43.9%   55,0  49,5  104,5   72/65/23
 06. Sting SF 0.8.4 x64                2929  1.440  62.7%  44.3%   54,0  46,5  100,5   62/77/21
 06. Sting SF 0.8.4 x64                2929  1.280  62.7%  43.8%   47,0  44,5   91,5   53/77/30
 06. Sting SF 0.8.4 x64                2935  1.120  63.5%  43.2%   54,0  43,0   97,0   64/66/30
 06. Sting SF 0.8.4 x64                2940    960  64.0%  43.5%   53,0  47,5  100,5   62/77/21
 06. Sting SF 0.8.4 x64                2941    800  64.2%  42.6%   54,5  46,0  100,5   66/69/25
 06. Sting SF 0.8.4 x64                2944    640  64.5%  42.5%   58,0  44,0  102,0   70/64/26
 06. Sting SF 0.8.4 x64                2947    480  64.8%  43.3%   57,5  46,0  103,5   72/63/25
 06. Sting SF 0.8.4 x64                2947    320  64.8%  45.3%   55,0  45,5  100,5   62/77/21
 06. Sting SF 0.8.4 x64                2964    160  66.9%  42.5%   58,0  49,0  107,0   73/68/19

     Program                           Elo    +    -   Games   Score  Av.Op.  Draw
 01. Komodo 8 x64                      3108   17   17  1600   81.3%   2825   31.0%
 02. Stockfish 03.08.14 BMI2 x64       3103   16   16  1700   82.1%   2811   31.1%
 --. Komodo 7a x64                     3078   19   19  1250   79.6%   2816   31.9%
 --. Stockfish 5 SSE42 x64             3077   20   20  1000   79.6%   2818   35.4%
 03. Fire 4 x64                        3054   16   16  1700   79.8%   2791   29.9%
 04. GullChess 3.0 BMI2 x64            3040   14   14  1900   74.5%   2830   37.9%
 05. Equinox 3.30 x64                  2993   15   15  1650   73.5%   2797   40.3%
 --. Fire 3.0 AVX x64                  2951   14   14  1700   66.9%   2816   39.9%
 06. Sting SF 4.8.4 x64                2931   13   13  1600   63.0%   2831   43.9%
 07. Protector 1.7.0 x64               2908   16   16  1250   56.6%   2859   45.5%
 08. Texel 1.05 x64                    2900   13   13  1650   59.7%   2827   43.8%
 09. Chiron 2.0 x64                    2895   10   10  3100   62.0%   2803   40.1%
 10. Critter 0.90 SSE4 x64             2891   14   14  1450   56.8%   2839   44.3%
 11. Naum 4.6 x64                      2887   16   16  1250   65.6%   2767   42.0%
 --. Protector 1.6.0 x64               2882   14   14  1600   58.7%   2816   44.4%
 12. Hannibal 1.4b x64                 2863    9    9  3100   58.0%   2804   43.0%

[Frank Quisinsky]

Hi Marek,

no problem ... nice to have Sting now in my rating list. I saw very nice games and ideas your engine have in openings (most important for me are the moves after my book). Games are now available in my selection "Download games by Players" ... if you like you can download.

All is fine, I think the endgame can be improved?!

Have a look in my News 142. After the fight around place 1 between Stockfish and Komodo I will start FCT0. So I can create for the TOP-21 engines also stats to middlegame, transposition into endgame and endgame. I forget in my Wirrwarr testing.

So, more interesting to follow my 2900+ round robins if I do that.

Furthermore, I find really a lot of bad lines in my book. In detail 39 games I replayed. Each bad line will be deleated with "F" in my book and the games replayed. That is a lot of work, means to check each game in my Rating List. Sting will help me a lot here (also in the future) to create a strong opening book for eng-eng or a rating list. The work you do here are really very good. No lost game up to move number 50 is a super result.

I hope 1-2 in the year I can test a newer Sting version and you continue with the development of Sting.

Thanks!

Best
Frank