FCP: Reorganization ...

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

Frank Quisinsky
Posts: 7228
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

FCP: Reorganization ...

Post by Frank Quisinsky »

Hi there,

very soon I will start a small reorganization of my test systems! A good time to do that because I have very soon in my Rating List 42 current engines. Rybka and Houdini will not participant for different reasons.

At the moment:
In FCT1 participant engines not older as three years. This one is to complicated to follow and different engines participant at the moment in FCT0 and FCT1.

In the future:
League 1 (Place 1-21)
League 2 (Place 22-42)
Special 2900+ (Engines stronger as around 2.900 Elo have to play vs. my TOP-32 = 1.600 games, no changes here).

With the result that I have a better structur and different engines have to play in League 1 and League 2. For that reason different not tested engines I am testing so far.

After Sjeng, Thinker, Critter, Nemo and Murka ... MinkoChess will be follow.

I hope of new releases of Booot and Hakkapeliitta (can include the updates with my reorganization) ... hope so ... more interesting to follow with two new engines not tested engines from my point of view).

Before I do that I will test the new Fizbo and Protector versions. At the moment the last 500 games from the still running Murka 3 x64 test-run are to play.

Have fun with it or all the others interesting things around computer chess.

:-)

Best
Frank
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: FCP: Reorganization ...

Post by Ferdy »

I have a suggestion regarding your rating list.

You have the existing FCP Rating List.

Code: Select all

"FCP Rating List"
 -----------------

 Date             : June 06th, 2015 (23:45)
 Games            : 53200, 32150 games sorted out!
 Version          : 2.08.s10
 Hardware         : 2x Intel Core i7-4770K *4.3Ghz, , 40/10 "repeatedlyâ€￾
 Engines          : Maximum allowed = two versions of same engine!

 Calculation      : Ordo 1.0, "Hiarcs 14 WCSC w32" with 2.825 Elo / News 057 & News 062
 Parameter        : Ordo_10.exe -a 2825 -A "Hiarcs 14 WCSC w32" -p 1.pgn -o rating.txt -W -s1000 -D -E -V
 Conditions       : http://www.amateurschach.de/main/_conditions.htm


 
      Program                          Elo    +    -   Games   Score  Av.Op.  Draw
  01. Stockfish 26.04.15 BMI2 x64      3150   16   16  1800    82.9%   2846   30.7%, New +16,  1492,5 Points
  02. Komodo 9 x64                     3146   15   15  1800    82.6%   2846   29.8%, New +41,  1487.0 Points
  ---------------------------------------------------------------------------------
  Fight for Place 1, Stockfish 26.04.15 & Komodo 9 have allways the same opponents!
  ---------------------------------------------------------------------------------
  --. Stockfish 6 BMI2 x64             3134   17   17  1700    83.8%   2823   28.9%
  --. Komodo 8 x64                     3105   15   15  1650    80.7%   2834   31.8%
  03. Fire 4 x64                       3044   13   13  2050    73.7%   2843   36.0%
  04. GullChess 3.0 BMI2 x64           3037   13   13  1900    71.8%   2856   41.2%
  05. Equinox 3.30 x64                 2989   12   12  2050    67.9%   2844   42.6%
  06. Critter 1.6a x64                 2980   13   13  1700    67.6%   2840   44.5%, older engine added!
  07. Sting SF 4.8.4 x64               2933   12   12  2000    60.9%   2849   45.8%
  08. Protector 1.7.0 x64              2910   12   12  1800    55.8%   2868   46.0%
  --. Critter 0.90 SSE4 x64            2900   13   13  1750    56.1%   2856   44.9%
  09. Nirvanachess 2.1c POP x64        2899   13   13  1600    57.2%   2846   48.1%, New +61
  10. Hannibal 1.5 x64                 2897   12   12  2000    56.9%   2846   45.8%
  11. Chiron 2.0 x64                   2896   10   10  3050    61.2%   2811   40.7%
  12. Texel 1.05 x64                   2893   11   11  2050    56.5%   2845   43.0%
  13. Naum 4.6 x64                     2883   11   11  2300    57.5%   2827   43.3%
  --. Hannibal 1.4b x64                2862   11   11  2450    58.9%   2796   43.3%
  --. Texel 1.04 x64                   2846   10   10  2250    57.0%   2795   42.7%
  --. Nirvanachess 2.0a x64            2838   11   11  2300    52.4%   2824   44.7%
  14. Senpai 1.0 SSE42 x64             2829    9    9  3050    52.5%   2814   42.7%
  15. Hiarcs 14 WCSC w32               2825    9    9  3050    52.0%   2814   43.2%
  16. Andscacs 0.72 POP x64            2813   11   11  2300    48.7%   2828   42.5%
  17. Sjeng c't 2010 w32               2796   11   11  2200    46.1%   2833   42.0%, older engine added!
  18. Shredder 12 x64                  2794   11   11  1950    42.5%   2858   42.5%
  19. Vajolet2 2.0 POP x64             2785   12   12  1700    44.4%   2834   41.2%, New +56
  20. Junior 13.3.00 x64               2784    9    9  2950    46.1%   2819   40.4%
  21. Spike 1.4 Leiden w32             2771   12   12  1850    38.6%   2866   41.9%
  ---------------------------------------------------------------------------------
  Current TOP-21, FCT0
  ---------------------------------------------------------------------------------
  22. Quazar 0.4 x64                   2758   12   12  1700    37.4%   2865   39.6%
  22. Deuterium 14.3.34.130 POP x64    2758   11   11  2350    42.0%   2828   42.6%
  24. DiscoCheck 5.2.1 x64             2755   12   12  1850    37.0%   2864   37.4%
  25. Cheng4 0.38 x64                  2754   13   13  1800    43.6%   2810   41.7%
  26. iCE 2.0 v2240 POP x64            2751    9    9  3250    43.5%   2807   43.8%
  27. SmarThink 1.70 SSE3 x64          2747    9    9  3050    41.8%   2816   38.1%
  27. Spark 1.0 x64                    2747    9    9  3050    42.7%   2809   41.8%
  29. Zappa Mexico II x64              2744   15   15  1450    34.6%   2876   39.6%
  30. Fizbo 1.3.1 x64                  2742   12   12  2000    41.5%   2816   41.3%
  31. Thinker 5.4d Inert x64           2733   11   11  2150    38.0%   2835   41.0%, older engine added!
  --. Vajolet2 1.45 POP x64            2729    9    9  3350    41.3%   2803   41.9%
  --. Deuterium 14.2.33.276 x64        2721   13   13  1450    41.9%   2787   42.0%
  32. Atlas 3.80 x64                   2719   13   13  1700    39.6%   2808   40.6%
  33. Gaviota 1.0 AVX x64              2715    9    9  3450    39.3%   2804   38.0%
  ---------------------------------------------------------------------------------
  Qualificated for Special 2900+ test-runs
  ---------------------------------------------------------------------------------
  34. Arasan 17.5 POP x64              2714   13   13  1400    42.5%   2774   43.6%
  35. Rodent 1.7 Build 1 POP x64       2688   14   14  1200    40.9%   2756   42.8%
  36. Nemo 1.01 Beta POP x64           2685   12   12  1600    38.8%   2770   42.8%, older engine added!
  --. Arasan 17.4 POP x64              2684   12   12  1750    38.5%   2774   40.9%
  --. Cheng4 0.36c x64                 2681   13   13  1550    38.9%   2768   42.1%
  --. Fizbo 1.2 x64                    2676   15   15  1200    36.0%   2789   37.7%
  37. DisasterArea 1.54 x64            2674   13   13  1550    38.6%   2759   43.4%
  38. EXchess 7.51b x64                2672   12   12  1700    36.5%   2780   41.3%
  39. Pedone 1.1 BMI2 x64              2663   14   14  1250    38.0%   2753   44.2%
  --. Rodent 1.6 Build 6 POP x64       2658   14   14  1350    34.8%   2778   37.4%
  40. Crafty 24.1 SSE42 x64            2647   12   12  2000    35.5%   2760   36.8%
  
  
  
 
 Move average                : 176.44 / 88.22
 White advantage             : 39.65  +/- 1.17
 Draw rate (equal opponents) : 48.90% +/- 0.25
 
 White Wins   :  18204 (34.2%)
 Black Wins   :  13130 (24.7%)
 Draws        :  21866 (41.1%)
 Unfinished   :      0

 White Perf.  :  54.8%
 Black Perf.  :  45.2%

 ECO A        =   8976 Games (16.9%)
 ECO B        =  12455 Games (23.4%)
 ECO C        =  11381 Games (21.4%)
 ECO D        =  10442 Games (19.6%)
 ECO E        =   9946 Games (18.7%)
My suggestion is to create a lighter version of it like the following. Something that is easy to read. It is also better if you can remove the +/-, and Av.Op. columns. So what remain are the rank, Program, Elo, Games, Score and Draw columns.

Your existing FCP rating list should still be there, that would be of help when someone wants to see the details of the rating list and engines tested.

Code: Select all

"FCP Rating List"
 -----------------

 Date             : June 06th, 2015 (23:45)
 Version          : 2.08.s10.best
 Hardware         : 2x Intel Core i7-4770K *4.3Ghz, , 40/10 "repeatedlyâ€￾

 Calculation      : Ordo 1.0, "Hiarcs 14 WCSC w32" with 2.825 Elo / News 057 & News 062
 Parameter        : Ordo_10.exe -a 2825 -A "Hiarcs 14 WCSC w32" -p 1.pgn -o rating.txt -W -s1000 -D -E -V
 Conditions       : http://www.amateurschach.de/main/_conditions.htm

 
      Program                          Elo    +    -   Games   Score  Av.Op.  Draw
  01. Stockfish 26Apr2015 BMI2 x64     3150   16   16  1800    82.9%   2846   30.7%
  02. Komodo 9 x64                     3146   15   15  1800    82.6%   2846   29.8%
  03. Fire 4 x64                       3044   13   13  2050    73.7%   2843   36.0%
  04. GullChess 3.0 BMI2 x64           3037   13   13  1900    71.8%   2856   41.2%
  05. Equinox 3.30 x64                 2989   12   12  2050    67.9%   2844   42.6%
  06. Critter 1.6a x64                 2980   13   13  1700    67.6%   2840   44.5%
  07. Sting SF 4.8.4 x64               2933   12   12  2000    60.9%   2849   45.8%
  08. Protector 1.7.0 x64              2910   12   12  1800    55.8%   2868   46.0%
  09. Nirvanachess 2.1c POP x64        2899   13   13  1600    57.2%   2846   48.1%
  10. Hannibal 1.5 x64                 2897   12   12  2000    56.9%   2846   45.8%
  11. Chiron 2.0 x64                   2896   10   10  3050    61.2%   2811   40.7%
  12. Texel 1.05 x64                   2893   11   11  2050    56.5%   2845   43.0%
  13. Naum 4.6 x64                     2883   11   11  2300    57.5%   2827   43.3%
  14. Senpai 1.0 SSE42 x64             2829    9    9  3050    52.5%   2814   42.7%
  15. Hiarcs 14 WCSC w32               2825    9    9  3050    52.0%   2814   43.2%
  16. Andscacs 0.72 POP x64            2813   11   11  2300    48.7%   2828   42.5%
  17. Sjeng c't 2010 w32               2796   11   11  2200    46.1%   2833   42.0%
  18. Shredder 12 x64                  2794   11   11  1950    42.5%   2858   42.5%
  19. Vajolet2 2.0 POP x64             2785   12   12  1700    44.4%   2834   41.2%
  20. Junior 13.3.00 x64               2784    9    9  2950    46.1%   2819   40.4%
  21. Spike 1.4 Leiden w32             2771   12   12  1850    38.6%   2866   41.9%
  22. Quazar 0.4 x64                   2758   12   12  1700    37.4%   2865   39.6%
  22. Deuterium 14.3.34.130 POP x64    2758   11   11  2350    42.0%   2828   42.6%
  24. DiscoCheck 5.2.1 x64             2755   12   12  1850    37.0%   2864   37.4%
  25. Cheng4 0.38 x64                  2754   13   13  1800    43.6%   2810   41.7%
  26. iCE 2.0 v2240 POP x64            2751    9    9  3250    43.5%   2807   43.8%
  27. SmarThink 1.70 SSE3 x64          2747    9    9  3050    41.8%   2816   38.1%
  27. Spark 1.0 x64                    2747    9    9  3050    42.7%   2809   41.8%
  29. Zappa Mexico II x64              2744   15   15  1450    34.6%   2876   39.6%
  30. Fizbo 1.3.1 x64                  2742   12   12  2000    41.5%   2816   41.3%
  31. Thinker 5.4d Inert x64           2733   11   11  2150    38.0%   2835   41.0%
  32. Atlas 3.80 x64                   2719   13   13  1700    39.6%   2808   40.6%
  33. Gaviota 1.0 AVX x64              2715    9    9  3450    39.3%   2804   38.0%
  34. Arasan 17.5 POP x64              2714   13   13  1400    42.5%   2774   43.6%
  35. Rodent 1.7 Build 1 POP x64       2688   14   14  1200    40.9%   2756   42.8%
  36. Nemo 1.01 Beta POP x64           2685   12   12  1600    38.8%   2770   42.8%
  37. DisasterArea 1.54 x64            2674   13   13  1550    38.6%   2759   43.4%
  38. EXchess 7.51b x64                2672   12   12  1700    36.5%   2780   41.3%
  39. Pedone 1.1 BMI2 x64              2663   14   14  1250    38.0%   2753   44.2%
  40. Crafty 24.1 SSE42 x64            2647   12   12  2000    35.5%   2760   36.8%
If you can rename the version of Stockfish from

Code: Select all

Stockfish 26.04.15 BMI2 x64
to

Code: Select all

Stockfish 26Apr2015 BMI2 x64
that would be very good. Again it is easy to distinguish when you test other version of it.
Seeing

Code: Select all

Stockfish 12.11.15 BMI2 x64
would be unclear, when you are not following sf development.
Frank Quisinsky
Posts: 7228
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP: Ordo ...

Post by Frank Quisinsky »

Hi Ferdinand,

1. It's possible to display with Ordo information you like only? Maybe with special parameters?

Rank, Program, Elo, Games, Score and Draw ...

You are right, because error, points, av. opp. and other information from Ordo and EloStat aren't really interesting, much (good example "error") are really nonsens in my opinion.

2. To rename SF on all my pages is a lot of work. I have around 100 stats HTML pages. Sitting a complete day on it.

Will thinking about it, yes to rename Stockfish by myself was a bad idea.

---

Maybe I should try to create all much more easy. Delete stats and playing style pages, link pages (only two or three programmers set a link to my site, I should do the same). The time I update my page, I can better use for my opening book.

On the other hand, all was so many work and with time ... a Rating list will be bigger and bigger: It's quit clear that with time all is more complicated if I have interest to Display more as Elo only.

One of the main problems in displaying a rating list are the games I should use for it. Only games from max. two versions of same engine allowed or more? I think more stable and more exactly will be a Rating list with lesser games and not so many versions of same engine ... with a lot of opponents.

Interesting in my opinion is this one ...

1. Deuterium x 2.800 Elo / 2.000 games / 50% / 40% draw quote / 36 opponents

If I display my TOP-40 only - without older versions you wish -, which games I should used for an Elo calculation? I think what I do here (maximal two versions of same engine allowed) is perfect.

And important for myself is to display the ECO code stat.

I can create a complete new webpage about it but I am not sure what should I display and what not. Much users like the Information with stats, others with playing style ... ohters are looking on live tables only. Others like to read a bit in a News-Ticker. Others will know ... which engine will be test and started to which time (ToDo Calendar).

I try to give all information but all is again and again a lot of work for myself.

What IPON, CEGT and others do with displaying results is to boring and not interesting for myself because Elo isn't the most interesting information for myself. So stats and playing style other things ... I like it.

Best
Frank

Best
Frank
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: FCP: Ordo ...

Post by Ferdy »

Frank Quisinsky wrote:Hi Ferdinand,

1. It's possible to display with Ordo information you like only? Maybe with special parameters?

Rank, Program, Elo, Games, Score and Draw ...
Miguel can do this.
2. To rename SF on all my pages is a lot of work. I have around 100 stats HTML pages. Sitting a complete day on it.
I mean for future engine versions.
One of the main problems in displaying a rating list are the games I should use for it. Only games from max. two versions of same engine allowed or more? I think more stable and more exactly will be a Rating list with lesser games and not so many versions of same engine ... with a lot of opponents.
You can include many version of engines in the calculation but only display the best version of the engine in the rating list I mentioned. Ordo has switch -g to report for isolated group or players that have not enough games with other engines. Include as many players as you want, then enable the -g switch.
Interesting in my opinion is this one ...

1. Deuterium x 2.800 Elo / 2.000 games / 50% / 40% draw quote / 36 opponents
That is fine too.
If I display my TOP-40 only - without older versions you wish -, which games I should used for an Elo calculation? I think what I do here (maximal two versions of same engine allowed) is perfect.
Try all, then use the -g switch. If older versions are isolated according to report try to remove the games of that older version. Or perhaps add new games if that would not require too much games, Asks Miguel how many games are required so that isolated players are no longer reported by Ordo. Then run ordo again.
Frank Quisinsky
Posts: 7228
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP: Maybe Miguel read that ...

Post by Frank Quisinsky »

Hi Ferdinand,

at first ... you have to know that your helps are allways welcome (important for myself).

OK ...

1. I renamed Stockfish versions in my current rating list. Also in my game databases (with the next game database update to see). I will not changed the older stat html files on my webpage.

---

Yes, Miguel can do that but I am sure on this changes he has no interest. If yes ...

It would be great to see ...

1. Rank
2. Name of Engine
3. Elo
4. Quantity of opponents
5. Games
6. Score
5. Draw

Example:
Gavitoa played vs. 20 engines, but two x different versions of Vajolet2.
I would like such an Information

Opponents
19 + 1

Or Gaviota played vs. 40 opponents
3x vs. Nirvanachess, 2x vs. Fizbo
37 + 3

This one:

Code: Select all

   # PLAYER                           : RATING   dif.Opp.  doubleOpp.   Games   Score(%)  Draws(%)
  29 Deuterium 14.3.34.130 POP x64    : 2758.0     37         10         2350    42.0      42.6
  41 Gaviota 1.0 AVX x64              : 2715.2     40         11         3450    39.3      38.0
With more opponents the calculation is faster - with lesser quantity of games - stable. Much more interesting as "Error".

Ferdinand, for the moment I will no other things changed. It's too many work for myself. Maybe I do that if my League 1 and League 2 System is ready for takeoff.

Best
Frank
Frank Quisinsky
Posts: 7228
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP: League 1 and 2 ... this one I have in my brain!

Post by Frank Quisinsky »

Hi Ferdinand,

this one I have in my brain ...
Main Rating List for the future ...

And the for the calculation not more as two versions from the same engine without to display older versions (good idea by Ferdiand).

Only Error / Points ... not important.
And draw Quote / move average I missed, Quantity of opponents I missed here.

Such a list is perfect for me, means with the important information.
If Miguel will not changed Ordo, no Problem I can edit the main information and detele the other one.

Best
Frank

Code: Select all

   Still running test-run with Murka 3 x64
   After round 35/50
   June, 11th, 2015 (19:45)

   ---------------------------------------------------------------------------
   FCT League 1: Different place 1-21 = 378,7 Elo
   ---------------------------------------------------------------------------
   # PLAYER                           : RATING  ERROR   POINTS  PLAYED    (%)
   1 Stockfish 26Apr2015 BMI2 x64     : 3149.5   15.4   1492.5    1800   82.9%
   2 Komodo 9 x64                     : 3145.9   15.7   1487.0    1800   82.6%
   3 Fire 4 x64                       : 3044.5   13.0   1511.5    2050   73.7%
   4 GullChess 3.0 BMI2 x64           : 3037.2   13.3   1363.5    1900   71.8%
   5 Equinox 3.30 x64                 : 2989.2   12.1   1391.0    2050   67.9%
   6 Critter 1.6a x64                 : 2979.7   13.1   1149.5    1700   67.6%
   7 Sting SF 4.8.4 x64               : 2932.5   12.0   1217.0    2000   60.9%
   8 Protector 1.7.0 x64              : 2910.8   12.0   1033.5    1835   56.3%
   9 Nirvanachess 2.1c POP x64        : 2899.4   12.3    943.5    1635   57.7%
  10 Hannibal 1.5 x64                 : 2896.3   11.4   1163.0    2035   57.1%
  11 Chiron 2.0 x64                   : 2896.0    9.4   1892.0    3085   61.3%
  12 Texel 1.05 x64                   : 2892.3   11.5   1184.0    2085   56.8%
  13 Naum 4.6 x64                     : 2882.3   10.8   1348.0    2335   57.7%
  14 Senpai 1.0 SSE42 x64             : 2828.5    9.4   1625.0    3085   52.7%
  15 Hiarcs 14 WCSC w32               : 2825.0    9.4   1611.0    3085   52.2%
  16 Andscacs 0.72 POP x64            : 2812.6   10.5   1143.0    2335   49.0%
  17 Sjeng c't 2010 w32               : 2796.4   11.0   1038.0    2235   46.4%
  18 Shredder 12 x64                  : 2793.7   11.2    850.0    1985   42.8%
  19 Vajolet2 2.0 POP x64             : 2786.5   12.6    779.5    1735   44.9%
  20 Junior 13.3.00 x64               : 2784.0    9.3   1384.0    2985   46.4%
  21 Spike 1.4 Leiden w32             : 2770.8   12.1    734.5    1885   39.0%
  ----------------------------------------------------------------------------
  FCT League 2: Different place 1-21 = 111,2 Elo 
  ----------------------------------------------------------------------------
   # PLAYER                           : RATING  ERROR   POINTS  PLAYED    (%)
   1 Quazar 0.4 x64                   : 2759.0   12.5    657.5    1735   37.9%
   2 Deuterium 14.3.34.130 POP x64    : 2757.5   11.1   1005.5    2385   42.2%
   3 DiscoCheck 5.2.1 x64             : 2755.2   12.2    705.5    1885   37.4%
   4 Cheng4 0.38 x64                  : 2752.7   12.0    802.5    1835   43.7%
   5 iCE 2.0 v2240 POP x64            : 2751.6    8.9   1437.5    3285   43.8%
   6 Spark 1.0 x64                    : 2748.1    9.1   1327.5    3085   43.0%
   7 SmarThink 1.70 SSE3 x64          : 2747.7    9.3   1298.0    3085   42.1%
   8 Zappa Mexico II x64              : 2742.5   13.8    518.0    1485   34.9%
   9 Fizbo 1.3.1 x64                  : 2742.2   11.6    849.0    2035   41.7%
  10 Thinker 5.4d Inert x64           : 2732.9   11.2    837.5    2185   38.3%
  11 Atlas 3.80 x64                   : 2717.1   12.7    689.0    1735   39.7%
  12 Gaviota 1.0 AVX x64              : 2715.7    8.5   1377.0    3485   39.5%
  13 Arasan 17.5 POP x64              : 2712.1   13.8    610.0    1435   42.5%
  14 Rodent 1.7 Build 1 POP x64       : 2688.8   14.3    509.0    1235   41.2%
  15 Murka 3 x64                      : 2686.6   14.6    448.5    1155   38.8%, still running after round 35/50
  16 Nemo 1.01 Beta POP x64           : 2685.2   13.3    638.0    1635   39.0%
  17 DisasterArea 1.54 x64            : 2672.1   13.0    612.5    1585   38.6%
  18 EXchess 7.51b x64                : 2671.8   12.7    635.5    1735   36.6%
  19 Pedone 1.1 BMI2 x64              : 2664.8   13.8    494.0    1285   38.4%
  20 Crafty 24.1 SSE42 x64            : 2647.8   11.4    726.5    2035   35.7%
  21 ...
  ----------------------------------------------------------------------------
  not qualificated
  ----------------------------------------------------------------------------

White advantage = 39.54 +/- 1.17
Draw rate (equal opponents) = 49.01 % +/- 0.24

Frank Quisinsky
Posts: 7228
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: FCP: Example: I can more do as visitor with this one ...

Post by Frank Quisinsky »

Such an information as human I like!
I am not a Computer and like to see "Error" or "Points".

Because I have vs. all this engines only one chance ... to win after the opening book moves. OK, vs. engines bad in endgames (good example is Spark ... to win in endgames).

And only this one is interesting in my opinion ...
And not "Error", Points and all this ... can nothing do with it.

With all the years we are looking on a rating list we await error and other nonsense as output ... but again ... nothing I can do with such an information.

Spark ... 400 Elo weaker as SF and such a high Attacker level. Often Spark can see the combination ... Stockfish, Komodo, Fire see nothing.

Engine Spark is really great ... I love Spark and Cheng too, Junior and on different times Hiarcs, EXchess, ICE ... great ideas. Nirvana too ... such things I will see in a rating list and not boring Elo as main information.

Stockfish have really nice stats ... but all the very fast draw games are terrible.

In the beginning of the game ... Shredder, Senpai, Hannibal and Equinox do nothing in interesting attack chess. More or less random or blunder from the other engine if a game will be won.

All this stats are allways the same ... with 1.000 game stats or 5.000 game stats per engine.

Best
Frank

Code: Select all

   --------------------------------------------------------------------------------------------------------------
   FCT League 1: Different place 1-21 = 378,7 Elo
   --------------------------------------------------------------------------------------------------------------
   # Player                           : Elo      Games   Score   Draws  MoveØ  won<50 lost<50   won(%), Attacker?
   1 Stockfish 26Apr2015 BMI2 x64     : 3149.5    1800   82.9%   30.7%   80!!    48       0      2.666%   Place 1
   2 Komodo 9 x64                     : 3145.9    1800   82.6%   29.8%   85      36       2      2.000%   Place 2
   3 Fire 4 x64                       : 3044.5    2050   73.7%   36.0%   86      39       1      1.902%   Place 3
   4 GullChess 3.0 BMI2 x64           : 3037.2    1900   71.8%   41.2%   88      31       3      1.631%   Place 4
   5 Equinox 3.30 x64                 : 2989.2    2050   67.9%   42.6%   89       9       3      0.439%
   6 Critter 1.6a x64                 : 2979.7    1700   67.6%   44.5%   89      22       3      1.294%   Place 6
   7 Sting SF 4.8.4 x64               : 2932.5    2000   60.9%   45.8%   90      14       1      0.700%
   8 Protector 1.7.0 x64              : 2910.8    1835   56.3%   45.5%   88      17       4      0.926%
   9 Nirvanachess 2.1c POP x64        : 2899.4    1635   57.7%   47.9%   87      14       4      0.856%
  10 Hannibal 1.5 x64                 : 2896.3    2035   57.1%   45.8%   92?!     8      15?     0.393%
  11 Chiron 2.0 x64                   : 2896.0    3085   61.3%   40.7%   87      36       8      1.166%   Place 8
  12 Texel 1.05 x64                   : 2892.3    2085   56.8%   43.1%   88      20       2      0.959%   and good king safty
  13 Naum 4.6 x64                     : 2882.3    2335   57.7%   43.5%   89      18       6      0.770%
  14 Senpai 1.0 SSE42 x64             : 2828.5    3085   52.7%   42.7%   89       9      48???   0.291%
  15 Hiarcs 14 WCSC w32               : 2825.0    3085   52.2%   43.2%   88      34      23?     1.102%   Place 9
  16 Andscacs 0.72 POP x64            : 2812.6    2335   49.0%   42.7%   89      14      13      0.599%
  17 Sjeng c't 2010 w32               : 2796.4    2235   46.4%   42.1%   87       8      18      0.357%
  18 Shredder 12 x64                  : 2793.7    1985   42.8%   42.6%   87       7      39???   0.352%
  19 Vajolet2 2.0 POP x64             : 2786.5    1735   44.9%   41.3%   89       6      12      0.345%
  20 Junior 13.3.00 x64               : 2784.0    2985   46.4%   40.3%   85      36      17      1.206%   Place 7
  21 Spike 1.4 Leiden w32             : 2770.8    1885   39.0%   42.1%   87      10      20      0.530%
  ---------------------------------------------------------------------------------------------------------------
  FCT League 2: Different place 1-21 = 111,2 Elo 
  ---------------------------------------------------------------------------------------------------------------
   # Player                           : Elo      Games   Score   Draws  MoveØ  won<50 lost<50   won(%), Attacker?
   1 Quazar 0.4 x64                   : 2759.0    1735   37.9%   39.6%   97?!     4     15       0.230%
   2 Deuterium 14.3.34.130 POP x64    : 2757.5    2385   42.2%   43.0%   83!     10     41??     0.419%
   3 DiscoCheck 5.2.1 x64             : 2755.2    1885   37.4%   37.7%   93?!     5     35??     0.265%
   4 Cheng4 0.38 x64                  : 2752.7    1835   43.7%   41.8%   87      18      2       0.980%   and good king safty
   5 iCE 2.0 v2240 POP x64            : 2751.6    3285   43.8%   43.9%   88      23     19       0.700%
   6 Spark 1.0 x64                    : 2748.1    3085   43.0%   41.9%   89      42     11       1.361%   Place 5
   7 SmarThink 1.70 SSE3 x64          : 2747.7    3085   42.1%   38.1%   90      12     24       0.388%
   8 Zappa Mexico II x64              : 2742.5    1485   34.9%   39.7%   89       8     21       0.538%
   9 Fizbo 1.3.1 x64                  : 2742.2    2035   41.7%   41.6%   83!     11     12       0.540%
  10 Thinker 5.4d Inert x64           : 2732.9    2185   38.3%   41.4%   87      19     38??     0.869%
  11 Atlas 3.80 x64                   : 2717.1    1735   39.7%   40.8%   88       4      8       0.230%
  12 Gaviota 1.0 AVX x64              : 2715.7    3485   39.5%   38.3%   87      13     68???    0.373%
  13 Arasan 17.5 POP x64              : 2712.1    1435   42.5%   43.6%   90      10      5       0.696%
  14 Rodent 1.7 Build 1 POP x64       : 2688.8    1235   41.2%   42.8%   90       8      6       0.647%
  15 Murka 3 x64                      : 2686.6    1155   38.8%   48.2%   88       0      5       0.000%
  16 Nemo 1.01 Beta POP x64           : 2685.2    1635   39.0%   43.1%   88       4     16       0.244%
  17 DisasterArea 1.54 x64            : 2672.1    1585   38.6%   43.6%   92?!     2     11       0.126%
  18 EXchess 7.51b x64                : 2671.8    1735   36.6%   41.3%   88      10      9       0.576%
  19 Pedone 1.1 BMI2 x64              : 2664.8    1285   38.4%   44.7%   91       1     19       0.077%
  20 Crafty 24.1 SSE42 x64            : 2647.8    2035   35.7%   37.0%   87      11     49??     0.540%
  21 ...
  ---------------------------------------------------------------------------------------------------------------
  not qualificated
  ---------------------------------------------------------------------------------------------------------------
Ferdy
Posts: 4851
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: FCP: Maybe Miguel read that ...

Post by Ferdy »

Frank Quisinsky wrote: Yes, Miguel can do that but I am sure on this changes he has no interest. If yes ...

It would be great to see ...

1. Rank
2. Name of Engine
3. Elo
4. Quantity of opponents
5. Games
6. Score
5. Draw
Just remove item 4 perhaps. It would just distract the eyes when looking at the rating list.