FCT1: Komodo 8 x64 is still running ...

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 12:05 pm

What for a fight!
This Round-Robin will go in my history for a Long time.
Very rarely: 2 good series behind one another!

Only 0.3% more and Komodo 8 is on first-Position!
3 Elo missed only !!

The same I produced in my Long time tourney for a while.
After 996 of 1.000 games ... Komodo 7a got the place for Stockfish 5. During the complete tourney are Stockfish 5 on one.

Have a look here:
http://www.amateurschach.de/ftptrigger/ ... 1-t02.html

And read my older News: 032
Perhaps with Komodo 8 now the same situation!

Komodo is to be good value for Money!
For every Cent!

Opinion switched (for the Moment)

Frank are thinking ... in LIVE mode!
Compare to each good Football game ...

Now we have to wait of he final result ...
Again, a very hard fight in FCT1 !!
Again ... 81,9% Komodo need 0.3% missed or 3 Elo only!
SF will fall to 81,9 with the finish!

And the prognosis are wrong too.
In 1 of 12 cases so far the prognosis is wrong!
Great fight ...

Code: Select all

     Programs                      Score       1-0  =  0-1          EloS  Baye  Ordo  +   -  Draws  MoveØ won<50 lost<50
 01. Stockfish 03.08.14 BMI2 x64   819.0/1000  660 318  22   81.9%  3085  3063  3117  19 19  31.8%   81     30       0
 02. Komodo 7a x64                 794.5/1000  635 319  46   79.5%  3059  3045  3089  19 19  31.9%   84     25       1
  

 Stats after   100,   200 ... games!   Elo   Games  Score  Draws   White Black Points  w/ d/ l
 --. Komodo 8 x64                      ----  1.000  --.-%  --.-%   --,-  --,-  --,-   --/--/--
 --. Komodo 8 x64                      ----    900  --.-%  --.-%   --,-  --,-  --,-   --/--/--
 02. Komodo 8 x64                      3082    800  81.6%  30.1%   42,5  43,0  85,5   74/23/03  +3.9% Elo = go up
 02. Komodo 8 x64                      3076    700  81.0%  31.1%   44,5  40,5  85,0   73/24/03  +4,0% Elo = go up "WRONG"
 02. Komodo 8 x64                      3069    600  80.3%  32.3%   41,5  39,0  80,5   64/33/03
 02. Komodo 8 x64                      3069    500  80.3%  32.2%   42,5  37,0  79,5   60/39/01
 02. Komodo 8 x64                      3071    400  80.5%  30.5%   42,0  36,0  78,0   60/36/04
 02. Komodo 8 x64                      3080    300  81.3%  28.7%   44,0  38,0  82,0   67/30/03
 02. Komodo 8 x64                      3077    200  81.0%  28.0%   43,5  38,0  81,5   66/31/03
 02. Komodo 8 x64                      3072    100  80.5%  25.0%   42,0  38,5  80,5   68/25/07
  

                                       Points  Score  EloS  Ordo  +   -  Draws
 01. Stockfish 03.08.14 BMI2 x64    813.0/990  82.1%  3085  3117  19 19  31.5%     
 02. Komodo 8 x64                   652.5/800  81.6%  3082  3114  22 22  30.1%
 03. GullChess 3.0 BMI2 x64         748.5/990  75.6%  3020  3047  18 18  35.9%
 04. Fire 3.0 AVX x64               657.5/990  66.4%  2946  2964  17 17  42.5%
 05. Chiron 2.0 x64                 578.5/990  58.4%  2890  2901  17 17  40.1%
 06. Protector 1.6.0 x64            575.5/990  58.1%  2888  2898  16 16  45.6% 
 07. Hannibal 1.4b x64              545.0/990  55.1%  2867  2875  16 16  43.4%
 08. Texel 1.04 x64                 527.0/990  53.2%  2855  2861  17 17  40.8% 
 09. Senpai 1.0 SSE42 x64           492.0/990  49.7%  2831  2835  16 16  42.6%
 10. Hiarcs 14 WCSC w32             484.5/990  48.9%  2826  2829  16 16  44.5%
 11. Shredder 12 x64                445.0/990  44.9%  2800  2800  17 17  41.4%
 12. Junior 13.8.04 Yokohama x64    431.5/990  43.6%  2791  2789  16 16  43.3%
 13. DiscoCheck 5.2.1 x64           409.5/990  41.4%  2776  2773  17 17  38.5% 
 14. SmarThink 1.70 SSE3 x64        397.5/990  40.2%  2767  2763  18 18  35.5% 
 15. Quazar 0.4 x64                 394.5/990  39.8%  2765  2761  17 17  42.1% 
 16. Zappa Mexico II x64            390.5/990  39.4%  2762  2758  16 16  42.9%
 17. Vajolet2 1.45 POP x64          379.5/990  38.3%  2755  2750  17 17  39.9%
 18. Spark 1.0 x64                  373.0/990  37.7%  2750  2744  17 17  38.2%
 19. Gaviota 1.0 AVX x64            344.5/990  34.8%  2730  2722  17 17  37.7%
 20. Deuterium 14.2.33.276 x64      343.0/990  34.6%  2728  2721  17 17  40.2%
 21. Nirvanachess 1.7 x64           317.5/990  32.1%  2709  2700  17 18  37.7%

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 12:40 pm

If Carlos can see that ...
Much better commantator as my humble self in Frankish!

Carlos wrote the best comments in Winboard times, as we organice and play the first Eng-Eng tourneys under Winboard in good and old times.

Komodo will get the most important Swedish medal if Komodo will reach place 1.
Thinking on the 4:4 Football match ... Germany - Sweden!
Better not thinking on this match ...

Best
Frank

Ozymandias · Post by **Ozymandias** » Sat Sep 06, 2014 3:29 pm

For a while there, it looked like this was going to be just another Komodo release, with a meager Elo increase. Now there's hope it will reach (and hold) the +25 Elo bar, on 1 CPU. That means it could reach the +40 Elo (*) I'm counting on, as an excuse for buying ($1 for each Elo point).

* SMP is supposed to be improved, I'll have to wait for CCRL 40/40 on that one.

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 3:32 pm

And the results vs. all the IPPs are very strong if I am looking on the results by others. This Group of programs arn't Play in FCT1. Could be the reason that I can see around 25 Elo, others can see 40 Elo more. Today we can explain a lot ... too many persons are testing.

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 6:03 pm

[pgn][Event "FCT1, 40/10, p=off, i7-4770k, 4.3GHz"]
[Site "Trier"]
[Date "2014.09.06"]
[Round "45.2"]
[White "GullChess 3.0 BMI2 x64"]
[Black "Komodo 8 x64"]
[Result "0-1"]
[ECO "E60"]
[PlyCount "80"]
[EventDate "2014.??.??"]
[EventType "tourn (rapid)"]
[EventRounds "50"]
[EventCountry "GER"]
[Source "Frank Quisinsky"]
[SourceDate "2014.09.06"]

1. d4 {[%eval 0,0] [%emt 0:00:00]} Nf6 {[%eval 0,0] [%emt 0:00:00]} 2. c4 {
[%eval 0,0] [%emt 0:00:00]} g6 {[%eval 0,0] [%emt 0:00:00]} 3. f3 {[%eval 0,0]
[%emt 0:00:00]} Bg7 {[%eval 0,0] [%emt 0:00:00]} 4. e4 {[%eval 0,0] [%emt 0:00:
00]} O-O {[%eval 0,0] [%emt 0:00:00]} 5. Ne2 {[%eval 0,0] [%emt 0:00:00]} c6 {
[%eval 0,0] [%emt 0:00:00]} 6. Nbc3 {[%eval 0,0] [%emt 0:00:00]} d5 {[%eval 0,
0] [%emt 0:00:00]} 7. cxd5 {[%eval 38,19] [%emt 0:00:27]} cxd5 {[%eval 14,21]
[%emt 0:00:15]} 8. e5 {[%eval 38,18] [%emt 0:00:00]} Nfd7 {[%eval 14,23] [%emt
0:00:24]} 9. h4 {[%eval 46,19] [%emt 0:00:14]} Nc6 {[%eval 14,22] [%emt 0:00:
16] (f6)} 10. h5 {[%eval 61,17] [%emt 0:00:17]} Nb6 {[%eval 14,21] [%emt 0:00:
16] (e6)} 11. b3 {[%eval 68,19] [%emt 0:00:26]} f6 {[%eval 27,22] [%emt 0:01:
10] (Ld7)} 12. hxg6 {[%eval 51,17] [%emt 0:00:16]} hxg6 {[%eval 14,18] [%emt 0:
00:11]} 13. exf6 {[%eval 75,18] [%emt 0:00:30]} Bxf6 {[%eval 14,18] [%emt 0:00:
07] (exf6)} 14. Qd2 {[%eval 63,16] [%emt 0:00:11]} Rf7 {[%eval 14,21] [%emt 0:
00:18]} 15. Ba3 {[%eval 68,17] [%emt 0:00:07] (Lb2)} e5 {[%eval -19,18] [%emt
0:00:05] (Th7)} 16. Qh6 {[%eval 57,18] [%emt 0:00:20] (dxe5)} exd4 {[%eval 14,
22] [%emt 0:00:21]} 17. Nb5 {[%eval 65,18] [%emt 0:00:08] (Dxg6+)} d3 {[%eval
14,20] [%emt 0:00:18]} 18. Nf4 {[%eval 65,18] [%emt 0:00:03] (Dxg6+)} Bf5 {
[%eval -6,18] [%emt 0:00:16] (Lxa1)} 19. Rd1 {[%eval 113,17] [%emt 0:00:26]
(Lxd3)} Ne5 {[%eval -81,18] [%emt 0:00:12] (Sd4)} 20. Kf2 {[%eval -1,18] [%emt
0:00:36] (Sxd3)} Nc8 {[%eval -79,18] [%emt 0:00:09] (Tc8)} 21. Bc5 {[%eval 0,
16] [%emt 0:00:22]} a6 {[%eval -75,17] [%emt 0:00:05] (g5)} 22. Nc3 {[%eval 0,
15] [%emt 0:00:13]} Nb6 {[%eval -50,18] [%emt 0:00:08] (Se7)} 23. Bb4 {[%eval
-24,16] [%emt 0:00:17] (b4)} a5 {[%eval -294,19] [%emt 0:00:09] (Sc6)} 24. Ba3
{[%eval -274,17] [%emt 0:01:16] (Lc5)} Rc8 {[%eval -337,20] [%emt 0:00:14]
(Dc7)} 25. Nb5 {[%eval -197,17] [%emt 0:00:16]} Qe8 {[%eval -424,20] [%emt 0:
00:11] (Tc2+)} 26. Nh3 {[%eval -539,16] [%emt 0:00:53] (Sd6)} Bxh3 {[%eval
-641,18] [%emt 0:00:51] (Th7)} 27. Qxh3 {[%eval -670,17] [%emt 0:00:25]} Ng4+ {
[%eval -2319,19] [%emt 0:00:06] (Tc2+)} 28. Qxg4 {[%eval -765,15] [%emt 0:00:
03]} Rc2+ {[%eval -2563,20] [%emt 0:00:46]} 29. Kg3 {[%eval -820,17] [%emt 0:
00:41]} Be5+ {[%eval -32733,16] [%emt 0:00:36]} 30. f4 {[%eval -820,17] [%emt
0:00:27]} Rxf4 {[%eval -32743,15] [%emt 0:00:45] (Lxf4+)} 31. Rh8+ {[%eval
-3247,17] [%emt 0:00:02]} Kxh8 {[%eval -32745,15] [%emt 0:00:19]} 32. Qxf4 {
[%emt 0:00:04] -M-1370/19} Bxf4+ {[%eval -32747,15] [%emt 0:00:01]} 33. Kxf4 {
[%emt 0:00:07] -M-1371/21} Rf2+ {[%eval -32749,15] [%emt 0:00:00]} 34. Kg3 {
[%emt 0:00:08] -M-1372/23 (Kg4)} Qe3+ {[%eval -32753,17] [%emt 0:00:01]} 35.
Kh2 {[%emt 0:00:02] -M-1373/23} Rf7 {[%eval -32755,17] [%emt 0:00:01]} 36. Bb2+
{[%emt 0:00:01] -M-1374/23} Kg8 {[%eval -32757,18] [%emt 0:00:00]} 37. Bg7 {
[%emt 0:00:01] -M-1374/23} Rxg7 {[%eval -32759,17] [%emt 0:00:01]} 38. g4 {
[%emt 0:00:09] -M-1376/27 (g3)} Qf2+ {[%eval -32761,18] [%emt 0:00:01]} 39.
Kh1 {[%emt 0:00:09] -M-1377/29} Rh7+ {[%eval -32763,19] [%emt 0:00:02]} 40. Bh3
{[%emt 0:00:00] -M-1377/6} Rxh3# {[%eval -32765,19] [%emt 0:00:02]} 0-1[/pgn]

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 9:01 pm

Only 0.22% more and Komodo 8 x64 is on first ...
All is possible!

Important now that Komodo can win the last 4 games vs. SF maybe with 2.5 : 1.5 ... means during game number 47-50!

Now the last 100 games!

Best
Frank

Code: Select all

     Programs                      Score       1-0  =  0-1          EloS  Baye  Ordo  +   -  Draws  MoveØ won<50 lost<50
 01. Stockfish 03.08.14 BMI2 x64   819.0/1000  660 318  22   81.9%  3085  3063  3117  19 19  31.8%   81     30       0
 02. Komodo 7a x64                 794.5/1000  635 319  46   79.5%  3059  3045  3089  19 19  31.9%   84     25       1
  

 Stats after   100,   200 ... games!   Elo   Games  Score  Draws   White Black Points  w/ d/ l
 --. Komodo 8 x64                      ----  1.000  --.-%  --.-%   --,-  --,-  --,-   --/--/--
 02. Komodo 8 x64                      3083    900  81.7%  30.1%   41,5  41,0  82,5   67/31/02
 02. Komodo 8 x64                      3082    800  81.6%  30.1%   42,5  43,0  85,5   74/23/03  +3,9% Elo = go up, WRONG
 02. Komodo 8 x64                      3076    700  81.0%  31.1%   44,5  40,5  85,0   73/24/03  +4,0% Elo = go up, WRONG
 02. Komodo 8 x64                      3069    600  80.3%  32.3%   41,5  39,0  80,5   64/33/03
 02. Komodo 8 x64                      3069    500  80.3%  32.2%   42,5  37,0  79,5   60/39/01
 02. Komodo 8 x64                      3071    400  80.5%  30.5%   42,0  36,0  78,0   60/36/04
 02. Komodo 8 x64                      3080    300  81.3%  28.7%   44,0  38,0  82,0   67/30/03
 02. Komodo 8 x64                      3077    200  81.0%  28.0%   43,5  38,0  81,5   66/31/03
 02. Komodo 8 x64                      3072    100  80.5%  25.0%   42,0  38,5  80,5   68/25/07
  

                                       Points  Score  EloS  Ordo  +   -  Draws
 01. Stockfish 03.08.14 BMI2 x64    816.0/995  82.0%  3085  3118  19 19  31.8%     
 02. Komodo 8 x64                   735.5/900  81.7%  3083  3116  21 21  30.1%
 03. GullChess 3.0 BMI2 x64         750.0/995  75.4%  3019  3046  18 18  36.0%
 04. Fire 3.0 AVX x64               659.0/995  66.2%  2946  2964  17 16  42.6%
 05. Chiron 2.0 x64                 580.0/995  58.3%  2890  2901  17 17  40.0%
 06. Protector 1.6.0 x64            576.0/995  57.9%  2887  2898  16 16  45.4% 
 07. Hannibal 1.4b x64              546.5/995  54.9%  2867  2875  16 16  43.5%
 08. Texel 1.04 x64                 527.5/995  53.0%  2854  2861  17 17  40.7% 
 09. Senpai 1.0 SSE42 x64           492.5/995  49.5%  2831  2834  16 16  42.5%
 10. Hiarcs 14 WCSC w32             485.0/995  48.7%  2826  2829  16 16  44.4%
 11. Shredder 12 x64                446.0/995  44.8%  2800  2800  17 17  41.4%
 12. Junior 13.8.04 Yokohama x64    431.5/995  43.4%  2790  2789  16 16  43.1%
 13. DiscoCheck 5.2.1 x64           410.0/995  41.2%  2776  2772  17 17  38.4% 
 14. SmarThink 1.70 SSE3 x64        398.5/995  40.1%  2768  2764  17 18  35.5% 
 15. Quazar 0.4 x64                 395.0/995  39.7%  2765  2761  17 17  42.0% 
 16. Zappa Mexico II x64            391.5/995  39.3%  2763  2758  16 16  42.9%
 17. Vajolet2 1.45 POP x64          380.0/995  38.2%  2755  2749  17 17  39.8%
 18. Spark 1.0 x64                  374.5/995  37.6%  2751  2745  17 17  38.3%
 19. Gaviota 1.0 AVX x64            344.0/995  34.6%  2729  2721  17 17  37.4%
 20. Deuterium 14.2.33.276 x64      343.0/995  34.5%  2728  2720  17 17  40.0%
 21. Nirvanachess 1.7 x64           318.0/995  32.0%  2709  2700  17 18  37.6%

Jesse Gersenson · Post by **Jesse Gersenson** » Sat Sep 06, 2014 9:33 pm

It was +41 on IPON. That's one core.

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 9:42 pm

Hi Jesse,

yes, with all the IPPs included!!
Double Rykba ... (Naum)

If good result vs. one of this engines, engine produce good results vs. oll of these engines. Futhermore, I think my i7 4.3Ghz have the double power as the older AMD hardware Ingo used. Clearly more time with my conditions. I am sure you can not compare the results directly. I have more different opponents too ... next reason! In my tourney engines can Play with BMI2, AVX and so one. Not possible with the older AMD Hardware. Ingo Plays with ponder = on, my actual FCT1 is with ponder = off. Other book idea, other time controls ... and so one!

Not to compare the results.
But nice to see that under the conditions from IPON Komodo made +41!!

I am sure Mark made a better optimazion for older AMD processors! Can be 3-5 Elo, not sure here!

In my FCT1 K8 is really very strong!
K7a was strong too ... used BMI2 or AVX all the time (if K7a / K8 set it automatically, think so after all I read about it).

In my list a newer SF is playing, stronger as the official Version 5.0 Ingo used. If Komodo have the same strength as the actual dev. version of SF ... that is great!

With my conditions K7a have the same strength as SF5 release version!

All in all ...
CEGT is testing and all the others ...
On the end of the day we all produced the same ... maybe not easy to read and often to understand if we are looking on Elo only, but the results are allways the same.

Best
Frank

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 10:29 pm

Hi,

means after all what I can read here is Rybka very similar to Naum. If both playing ... not good ... because the results will be very similar too.

Jesse, I have here an other problem in testing. Not Komodo 8! The dev. Version from SF gave to fast draw vs. clearly weaker engines. Information can be found in my News 056.

I hope SF can improved a bit to this topic.

Basically it seems very clear that the actual SF and Komodo version have around the same playing strength.

Interesting only, both have an other style ...

After the still running round robin:
Fizbo will replaced Zappa.
Zappa have strong results vs. SF.

If I delete in my Continuous Tourney the older Zappa version and if Komodo is after the still running round robin on 1 ... Komodo will lost directly place 1 if Zappa is out of tourney.

Other opponents, other results ...
Not important how many games you have!

To the topic ...
On IPON +41

Best
Frank

Frank Quisinsky · Post by **Frank Quisinsky** » Sat Sep 06, 2014 10:58 pm

Code: Select all

   1 Komodo 8 x64                   : 3120.7   20.5    766.0     934   82.0%
   2 Stockfish 03.08.14 BMI2 x64    : 3117.9   20.5    816.5     997   81.9%
   3 GullChess 3.0 BMI2 x64         : 3047.4   18.8    751.0     997   75.3%
   4 Fire 3.0 AVX x64               : 2964.5   16.7    659.0     997   66.1%
   5 Chiron 2.0 x64                 : 2901.2   16.5    580.0     997   58.2%
   6 Protector 1.6.0 x64            : 2898.5   16.0    576.5     997   57.8%
   7 Hannibal 1.4b x64              : 2875.5   16.1    546.5     997   54.8%
   8 Texel 1.04 x64                 : 2861.1   16.4    527.5     997   52.9%
   9 Senpai 1.0 SSE42 x64           : 2834.8   15.6    492.5     997   49.4%
  10 Hiarcs 14 WCSC w32             : 2829.7   16.1    485.5     997   48.7%
  11 Shredder 12 x64                : 2800.0   16.3    446.0     997   44.7%
  12 Junior 13.8.04 Yokohama x64    : 2789.5   15.9    432.0     997   43.3%
  13 DiscoCheck 5.2.1 x64           : 2772.8   16.1    410.0     997   41.1%
  14 SmarThink 1.70 SSE3 x64        : 2764.0   16.5    398.5     997   40.0%
  15 Quazar 0.4 x64                 : 2761.4   15.9    395.0     996   39.7%
  16 Zappa Mexico II x64            : 2758.7   16.4    391.5     996   39.3%
  17 Vajolet2 1.45 POP x64          : 2750.2   16.3    380.5     997   38.2%
  18 Spark 1.0 x64                  : 2745.5   16.4    374.5     996   37.6%
  19 Gaviota 1.0 AVX x64            : 2721.4   16.9    344.0     996   34.5%
  20 Deuterium 14.2.33.276 x64      : 2720.6   16.8    343.0     996   34.4%
  21 Nirvanachess 1.7 x64           : 2700.3   16.9    318.0     996   31.9%

FCT1: Komodo 8 x64 is still running ...

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: Where is Carlos E. A. Drake

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: FCT1: Zack, Zack, Zack and mate ... after 40 vs. Gull3

Re: FCT1: K8 after 900 of 1.000, +24 Elo ... all is possible

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: FCT1: K8 after 800 of 1.000, +23 Elo ... incredible!

Re: FCT1: K8 after 934 games, first time, Komodo on 1