Mars

Sat Jul 12, 2014 10:21 am by · Sat Jul 12, 2014 10:21 am by

@mcdiniz wrote:tks.. majortom... some improvement strength in Elo? ...

My 120"+2" test is not completed yet:

Code:: # PLAYER : RATING ERROR POINTS PLAYED (%) 1 Houdini 4 Pro x64 : 3153.9 11.9 1361.5 1935 70.4% 2 Stockfish 190514 64 SSE4.2 : 3137.4 15.1 1044.0 1529 68.3% 3 Mars 2.0 AVX x64 : 3018.6 31.6 98.0 300 32.7% 4 Mars 1.5 AVX x64 : 3003.5 20.0 327.0 1064 30.7% 5 Mars 1 AVX x64 : 3000.0 ---- 633.5 2100 30.2%

Mon Jul 14, 2014 2:32 pm by · Mon Jul 14, 2014 2:32 pm by

@majortom wrote:Mars 2.0 x64

Fast 24 hours Test of Mars 2 against Fire 3. 1000 games (500 opening-positions, repeated, 70''+700ms, 1 core, no ponder, i7-2630QM-notebook (2.0 GHz).

Mars 2 -Fire 3: 503-497 (+174 -168 =658)

Nice result. Older Mars-versions scored below 45% against Fire 3 in my tests...

Stefan

Mon Jul 14, 2014 8:39 pm by · Mon Jul 14, 2014 8:39 pm by

Few elo added:
Mars 2.1 x64

Mon Jul 14, 2014 8:42 pm by · Mon Jul 14, 2014 8:42 pm by

@pohl4711 wrote:

@majortom wrote:Mars 2.0 x64

Fast 24 hours Test of Mars 2 against Fire 3. 1000 games (500 opening-positions, repeated, 70''+700ms, 1 core, no ponder, i7-2630QM-notebook (2.0 GHz).

Mars 2 -Fire 3: 503-497 (+174 -168 =658)

Nice result. Older Mars-versions scored below 45% against Fire 3 in my tests...

Stefan

Thanx Stefan for the test!

Mon Jul 14, 2014 9:15 pm by · Mon Jul 14, 2014 9:15 pm by

First results of new version:

Code:: # PLAYER : RATING ERROR POINTS PLAYED (%) 1 Houdini 4 Pro x64 : 3154.2 12.9 1640.0 2338 70.1% 2 Stockfish 190514 64 SSE4.2 : 3137.1 14.7 1287.5 1895 67.9% 3 Mars 2.1 AVX x64 : 3036.5 33.3 108.0 309 35.0% 4 Mars 2.0 AVX x64 : 3007.3 23.0 237.0 760 31.2% 5 Mars 1.5 AVX x64 : 3003.4 19.0 327.0 1064 30.7% 6 Mars 1 AVX x64 : 3000.0 ---- 633.5 2100 30.2%

Mon Jul 14, 2014 9:57 pm by · Mon Jul 14, 2014 9:57 pm by

TimeControl "120+2":

[pgn][Event "?"]
[Site "?"]
[Date "2014.07.14"]
[Round "7"]
[White "Stockfish 190514 64 SSE4.2"]
[Black "Mars 2.1"]
[Result "0-1"]
[ECO "E81"]
[Opening "King's Indian"]
[Variation "Sämisch, 5...O-O 6.Be3 Nbd7 7.Qd2 c5 8.Nge2 a6"]
[TimeControl "120+2"]
[Termination "normal"]
[PlyCount "128"]
[WhiteType "human"]
[BlackType "human"]

1. d4 {book} d6 {book} 2. e4 {book} Nf6 {book} 3. f3 {book} e5 {book} 4. d5
{book} Be7 {book} 5. Be3 {book} Nh5 {book} 6. c4 {book} Bg5 {book} 7. Bf2
{book} g6 {book} 8. Nc3 {book} O-O {book} 9. g4 {+0.40/22 8.4s} Nf4
{-0.25/20 5.9s} 10. h4 {+0.39/24 3.1s} Bh6 {-0.39/20 25s} 11. Nge2
{+0.34/22 6.6s} c5 {-0.39/19 6.8s} 12. g5 {+0.58/18 5.3s} Bg7 {-0.21/17
0.098s} 13. Qd2 {+0.67/18 6.2s} Nd7 {-0.30/17 1.6s} 14. O-O-O {+0.81/20
25s} Nh5 {-0.31/17 1.8s} 15. Ng3 {+0.63/21 17s} Nf4 {-0.31/18 1.9s} 16. Kb1
{+0.59/21 18s} a6 {-0.31/18 6.4s} 17. Nge2 {+0.59/20 2.3s} Nh5 {-0.16/16
5.7s} 18. b3 {+0.60/19 2.3s} Rb8 {-0.15/16 7.2s} 19. Ng3 {+0.45/18 5.3s}
Nf4 {0.00/14 0.047s} 20. Be3 {+0.58/20 3.8s} Re8 {-0.08/16 4.4s} 21. Qh2
{+1.21/18 4.0s} b6 {-0.18/16 7.4s} 22. Rc1 {+1.19/19 7.7s} Qe7 {-0.26/16
4.3s} 23. Rc2 {+1.26/19 8.0s} b5 {-0.41/16 4.5s} 24. cxb5 {+1.85/19 4.4s}
axb5 {-0.40/16 1.3s} 25. Nxb5 {+1.86/19 7.3s} Nb6 {-0.59/16 8.1s} 26. h5
{+1.99/16 3.1s} c4 {-0.79/16 3.1s} 27. bxc4 {+2.02/16 6.0s} Bd7 {-0.82/17
16s} 28. Rb2 {+1.61/17 6.3s} Na4 {-0.83/17 5.3s} 29. Rb3 {+1.67/18 3.4s}
Rec8 {-1.05/16 14s} 30. hxg6 {+1.77/18 5.7s} hxg6 {-1.06/16 3.6s} 31. Ne2
{+1.18/20 7.3s} Bxb5 {-1.06/16 1.9s} 32. cxb5 {+1.28/19 4.2s} Qxg5
{-0.69/18 11s} 33. b6 {+1.50/16 2.5s} Qf6 {-0.13/16 7.7s} 34. Ka1 {+0.98/17
2.0s} Nc5 {0.00/15 5.3s} 35. Rb5 {+1.13/16 2.0s} Ncd3 {0.00/16 2.6s} 36.
Kb1 {+1.17/17 2.0s} Ne1 {+0.01/16 3.8s} 37. Ng1 {+0.41/17 2.0s} Nc2
{+0.19/15 1.4s} 38. Bc1 {+0.12/18 2.0s} Nd4 {+1.36/15 2.3s} 39. Rb2
{0.00/17 1.9s} Nxd5 {+1.36/13 0.005s} 40. Ba6 {-0.06/18 2.1s} Rc6 {+2.49/15
4.1s} 41. Bd3 {-1.79/17 2.0s} Nf4 {+2.49/13 0.007s} 42. Bxf4 {-2.42/19
2.0s} exf4 {+4.36/10 0.47s} 43. b7 {-8.72/19 2.0s} Nc2 {+8.26/14 0.69s} 44.
Qh7+ {-9.72/20 1.7s} Kf8 {+8.26/12 0.001s} 45. Qxg7+ {-10.22/22 2.3s} Kxg7
{+9.64/13 0.69s} 46. Bxc2 {-10.47/23 2.0s} Qc3 {+9.64/12 0s} 47. Ne2
{-10.76/22 2.0s} Qxf3 {+11.16/14 1.6s} 48. Rh2 {-10.96/21 1.7s} Qf1+
{+11.16/12 0s} 49. Nc1 {-11.10/22 2.1s} f3 {+11.44/14 0.92s} 50. Rd2
{-12.57/20 2.2s} f2 {+12.22/14 0.62s} 51. e5 {-13.20/22 2.0s} Qg1
{+11.85/11 0.004s} 52. Rxf2 {-14.42/21 1.8s} Qxf2 {+12.74/13 0.58s} 53. Ba4
{-15.51/21 2.2s} Rb6 {+12.95/11 0.003s} 54. Bb3 {-16.42/22 2.1s} Qf5+
{+16.94/13 1.2s} 55. Rc2 {-18.58/23 2.0s} Rc6 {+16.94/12 0s} 56. Ne2
{-105.82/26 2.0s} Rxc2 {+M33/13 1.9s} 57. Nd4 {-110.51/18 2.0s} Rc1+
{+M21/17 1.4s} 58. Kxc1 {-M18/17 2.0s} Qf4+ {+M15/15 0.84s} 59. Kb1
{-M16/18 1.1s} Qxd4 {+M13/16 3.1s} 60. a4 {-M10/22 1.9s} Qd2 {+M9/17 3.3s}
61. Bd5 {-M8/100 3.0s} Rh8 {+M7/16 0.64s} 62. a5 {-M6/120 0.027s} Rh2
{+M5/17 3.1s} 63. Bg2 {-M4/120 0.007s} Rxg2 {+M3/17 0.59s} 64. a6 {-M2/120
0.032s} Qd1# {+M1/128 0.047s, Black mates} 0-1
[/pgn]

Last edited by majortom on Mon Jul 14, 2014 10:17 pm; edited 1 time in total

Mon Jul 14, 2014 10:12 pm by · Mon Jul 14, 2014 10:12 pm by

hum.. good game... marjortom when we can have a hand .. to be able to test? ... will do soon my particular tournament ... your engine would be very welcome ... best

Mon Jul 14, 2014 10:20 pm by · Mon Jul 14, 2014 10:20 pm by

@mcdiniz wrote:hum.. good game... marjortom when we can have a hand .. to be able to test? ... will do soon my particular tournament ... your engine would be very welcome ... best

That game was played by Mars 2.1, follow the link:
http://treu.ru/mars/Mars_2.1_x64.7z

Wed Jul 16, 2014 12:34 pm by · Wed Jul 16, 2014 12:34 pm by

Thanks majortom. thumb up

Sun Jul 20, 2014 10:23 pm by · Sun Jul 20, 2014 10:23 pm by

Few elo added:

Mars 2.2 x64

Tue Jul 22, 2014 1:27 am by · Tue Jul 22, 2014 1:27 am by

Mars 2.2 looks very strong in test still early but this engine is strong! thumb up

Fri Jul 25, 2014 8:46 am by · Fri Jul 25, 2014 8:46 am by

Few elo added:

Mars 2.5 x64

Fri Jul 25, 2014 1:55 pm by · Fri Jul 25, 2014 1:55 pm by

Majortom, Mars x64 gauntlet will start 2/3 days after BooksWar 11 is finished. Be Ready with the final version. smile

Fri Jul 25, 2014 9:14 pm by · Fri Jul 25, 2014 9:14 pm by

Vladimir wrote:Majortom, Mars x64 gauntlet will start 2/3 days after BooksWar 11 is finished. Be Ready with the final version.

Vladimir, we are ready and we are waiting for:

Mars 2.6 x64

Mon Jul 28, 2014 9:37 pm by · Mon Jul 28, 2014 9:37 pm by

Mars 2.7 x64

TC 30"+1"

Code:: Program Elo + - Games Score Av.Op. Draws 1 Mars 2.7 AVX x64 : 3004 17 17 600 51.1 % 2996 63.8 % 2 Fire 3.0 x64 : 2999 25 25 300 49.3 % 3004 60.7 % 3 PanChess 00.611 x64 : 2993 23 23 300 48.5 % 3004 67.0 %

Code:: Individual statistics: 1 Mars 2.7 AVX x64 : 3004 600 (+115,=383,-102), 51.1 % PanChess 00.611 x64 : 300 (+ 54,=201,- 45), 51.5 % Fire 3.0 x64 : 300 (+ 61,=182,- 57), 50.7 % 2 Fire 3.0 x64 : 2999 300 (+ 57,=182,- 61), 49.3 % Mars 2.7 AVX x64 : 300 (+ 57,=182,- 61), 49.3 % 3 PanChess 00.611 x64 : 2993 300 (+ 45,=201,- 54), 48.5 % Mars 2.7 AVX x64 : 300 (+ 45,=201,- 54), 48.5 %

Games

Wed Jul 30, 2014 2:04 pm by · Wed Jul 30, 2014 2:04 pm by

@majortom wrote:

Vladimir wrote:Majortom, Mars x64 gauntlet will start 2/3 days after BooksWar 11 is finished. Be Ready with the final version.

Vladimir, we are ready and we are waiting for:

Mars 2.6 x64

Which one is the final version? wink

Wed Jul 30, 2014 2:44 pm by · Wed Jul 30, 2014 2:44 pm by

Vladimir wrote:

@majortom wrote:

Vladimir wrote:Majortom, Mars x64 gauntlet will start 2/3 days after BooksWar 11 is finished. Be Ready with the final version.

Vladimir, we are ready and we are waiting for:

Mars 2.6 x64

Which one is the final version?

I think the last one - Mars 2.7 wink

Sat Aug 02, 2014 5:24 pm by · Sat Aug 02, 2014 5:24 pm by

@pohl4711 wrote:

@majortom wrote:Mars 2.0 x64

Fast 24 hours Test of Mars 2 against Fire 3. 1000 games (500 opening-positions, repeated, 70''+700ms, 1 core, no ponder, i7-2630QM-notebook (2.0 GHz).

Mars 2 -Fire 3: 503-497 (+174 -168 =658)

Nice result. Older Mars-versions scored below 45% against Fire 3 in my tests...

Stefan

Hello Stefan, are you planning to add Mars 2.7 in your tournament? please give me a link to your list.

Sun Aug 03, 2014 11:31 am by · Sun Aug 03, 2014 11:31 am by

waelsae wrote:

@pohl4711 wrote:

@majortom wrote:Mars 2.0 x64

Fast 24 hours Test of Mars 2 against Fire 3. 1000 games (500 opening-positions, repeated, 70''+700ms, 1 core, no ponder, i7-2630QM-notebook (2.0 GHz).

Mars 2 -Fire 3: 503-497 (+174 -168 =658)

Nice result. Older Mars-versions scored below 45% against Fire 3 in my tests...

Stefan

Hello Stefan, are you planning to add Mars 2.7 in your tournament? please give me a link to your list.

I am doing a small tournament with Mars 2.7 and some other Ippolit-derivatives at the moment. Will take some days.

For a rankinglist-result of Mars (2.2) and PanChess 00.611 take a look at Andreas Strangmüllers excellent website/bullet-ratinglist: http://fastgm.de/60+0.60%20-%20Universal.html
There you will see, that the new PanChess is no progress and Mars 2.2 is a regression (related to Mars 1)

Regards - Stefan

Mon Aug 04, 2014 6:23 am by · Mon Aug 04, 2014 6:23 am by

Mars 2.7 x64 Gauntlet (finished)
https://www.chess2u.com/t8642p15-mars-tests#58877

Code:: EnginesWar (4 Aug 2014) Program Elo + - Games Score Av.Op. Draws Mars 1 x64 : 3168 9 9 2400 47.7 % 3184 58.1 % Mars 2.7 x64 : 3167 12 12 1500 39.6 % 3240 53.4 % Mars 1.1 x64 : 3160 8 8 3300 42.7 % 3211 51.9 %

https://www.chess2u.com/t6886p45-engineswar-rating-list#58879