Moderator: Ras
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU
AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash
1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50
2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU
AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash
1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60
2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
Interesting.... what Contempt value is Houdini 3 using ?beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
Houdini's contempt assumes that everyone is far weaker. So Houdini is at a disadvantage playing against the strongest programs. But it's a trade-off. Komodo too has a 7 contempt by default which puts it at a disadvantage against an opponent that is nearly as strong or stronger.shrapnel wrote:Interesting.... what Contempt value is Houdini 3 using ?beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
If komodo at drawscore=0 gives good results, woudn't it give even better results at drawscore = -100 ?
Sorry, that should read drawscore= + 100.shrapnel wrote:Interesting.... what Contempt value is Houdini 3 using ?beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
If komodo at drawscore=0 gives good results, woudn't it give even better results at drawscore = -100 ?
Surely it is - 7 ?!Don wrote:Houdini's contempt assumes that everyone is far weaker. So Houdini is at a disadvantage playing against the strongest programs. But it's a trade-off. Komodo too has a 7 contempt by default which puts it at a disadvantage against an opponent that is nearly as strong or stronger.shrapnel wrote:Interesting.... what Contempt value is Houdini 3 using ?beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
If komodo at drawscore=0 gives good results, woudn't it give even better results at drawscore = -100 ?
There is nothing to be done about this because there is no way for a program to know what it's opponents strength is in comparison to it's own - at least not with the UCI protocol. So you have to take a guess.
And the match has ended with 52% for Komodo 5.1MP (with drawscore=0)beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 20m+10s 4 CPU
AMD Phenom II X6 1090T, Bram testsuite 2.0, 1024 Mb Hash
1 Komodo 5.1 64-bit +14/=24/-12 52.00% 26.0/50
2 Houdini 3 x64 +12/=24/-14 48.00% 24.0/50Hiberam wrote:This fashionable modern line of the Caro Kann, Komodo won both with white and black pieces
The black game was really nice to watch. After active opening play 5 ..e5! (Houdini played e6) Houdini choose to play rather risky 6 exd5 !? (6 Bd3 was played in a GM game this year by Christian Bauer – Markus Ragger) Komodo sacked two pawns for attacking chances on the exposed white king. And after the nice 17 ..Bg3 ! the advantage was clearly there and Komodo converted this advantage into a queen-pawns endgame with a decisive a-pawn passer.
[pgn]
[Event "Houdini3 - Komodo5.1, Blitz 20m+10s"]
[Site "AMD X6 1090T 3200 Mhz, (Fritz"]
[Date "2013.06.23"]
[Round "33"]
[White "Houdini 3 x64"]
[Black "Komodo 5.1 64-bit"]
[Result "0-1"]
[ECO "B12"]
[TimeControl "1200+10"]
{AMD Phenom(tm) II X6 1090T Processor 3207 MHz W=22.9 plies; 8.069kN/s B=21.
5 plies; 4.772kN/s} 1. e4 c6 2. d4 d5 3. f3 Qb6 4. a4 {Both last book move} e5
{[%eval 3,19] [%emt 0:00:27]} 5. dxe5 {[%eval 15,21] [%emt 0:00:55] (c3)} a5 {
[%eval -20,19] [%emt 0:00:19]} 6. exd5 {[%eval 2,21] [%emt 0:00:27] (Bd3)} Bc5
{[%eval -30,21] [%emt 0:00:29]} 7. Nh3 {[%eval 2,20] [%emt 0:00:00]} Bxh3 {
[%eval -17,21] [%emt 0:01:06]} 8. gxh3 {[%eval -2,21] [%emt 0:00:33]} Bf2+ {
[%eval -18,23] [%emt 0:01:20]} 9. Ke2 {[%eval -2,20] [%emt 0:00:00]} cxd5 {
[%eval -7,23] [%emt 0:00:58]} 10. Qxd5 {[%eval 0,22] [%emt 0:00:31]} Nc6 {
[%eval -11,23] [%emt 0:00:35]} 11. Qb5 {[%eval -5,24] [%emt 0:01:32] (Nc3)} Qa7
{[%eval -12,22] [%emt 0:00:27]} 12. Nc3 {[%eval 10,22] [%emt 0:00:28]} O-O-O {
[%eval -20,21] [%emt 0:01:06] (Nge7)} 13. Qc4 {[%eval -13,22] [%emt 0:01:41]}
Nge7 {[%eval -21,19] [%emt 0:00:10]} 14. Nb5 {[%eval -31,21] [%emt 0:00:49]}
Qb6 {[%eval 0,20] [%emt 0:00:10]} 15. Ra3 {[%eval -26,19] [%emt 0:00:58]} Nf5 {
[%eval 0,22] [%emt 0:00:38]} 16. Qe4 {[%eval -31,21] [%emt 0:02:26]} g6 {
[%eval 0,22] [%emt 0:00:32]} 17. Bg5 {[%eval -31,22] [%emt 0:00:33]} Bg3 {
[%eval -38,22] [%emt 0:01:09] (f6)} 18. Rd3 {[%eval -17,21] [%emt 0:00:27]}
Rde8 {[%eval -38,22] [%emt 0:00:49] (Rxd3)} 19. Be3 {[%eval -31,20] [%emt 0:00:
39] (Kd1)} Nxe3 {[%eval -54,22] [%emt 0:00:27]} 20. Nd6+ {[%eval -29,22] [%emt
0:00:19] (Qxe3)} Kb8 {[%eval -58,23] [%emt 0:00:28]} 21. Qxe3 {[%eval -29,21]
[%emt 0:00:00]} Rxe5 {[%eval -58,23] [%emt 0:00:24]} 22. Ne4 {[%eval -16,21]
[%emt 0:00:04]} Qxb2 {[%eval -60,23] [%emt 0:00:21]} 23. hxg3 {[%eval -16,20]
[%emt 0:00:00]} f5 {[%eval -64,24] [%emt 0:00:28]} 24. Rd2 {[%eval -30,23]
[%emt 0:00:21]} fxe4 {[%eval -67,21] [%emt 0:00:10]} 25. f4 {[%eval -30,24]
[%emt 0:00:28]} Re7 {[%eval -61,24] [%emt 0:00:37]} 26. Bg2 {[%eval -45,22]
[%emt 0:00:20] (c3)} Rc8 {[%eval -68,22] [%emt 0:00:35]} 27. c4 {[%eval -38,23]
[%emt 0:00:22] (c3)} Qb4 {[%eval -77,23] [%emt 0:00:29]} 28. Rc1 {[%eval -51,
25] [%emt 0:00:32]} Qxa4 {[%eval -71,21] [%emt 0:00:22]} 29. Kf2 {[%eval -40,
23] [%emt 0:00:24]} Qb4 {[%eval -63,20] [%emt 0:00:31] (Rf8)} 30. Rd5 {[%eval
-58,22] [%emt 0:00:23]} Rd8 {[%eval -60,20] [%emt 0:00:21]} 31. Rcd1 {[%eval
-84,23] [%emt 0:00:35]} Ree8 {[%eval -88,21] [%emt 0:00:33]} 32. Rxd8+ {[%eval
-61,23] [%emt 0:00:20] (Bf1)} Rxd8 {[%eval -98,20] [%emt 0:00:11]} 33. Rxd8+ {
[%eval -58,22] [%emt 0:00:18]} Nxd8 {[%eval -101,20] [%emt 0:00:06]} 34. Qxe4 {
[%eval -75,23] [%emt 0:01:06]} Qc5+ {[%eval -79,21] [%emt 0:00:43] (a4)} 35.
Ke1 {[%eval -73,23] [%emt 0:00:34]} Nc6 {[%eval -83,21] [%emt 0:00:53] (Ka7)}
36. Qe2 {[%eval -44,22] [%emt 0:00:19] (Qe8+)} Qd4 {[%eval -80,22] [%emt 0:01:
31] (Nd4)} 37. Bxc6 {[%eval -73,21] [%emt 0:00:18] (Kf1)} bxc6 {[%eval -87,21]
[%emt 0:01:08]} 38. Qe7 {[%eval -69,23] [%emt 0:00:16] (Qd2)} Qc3+ {[%eval
-126,19] [%emt 0:00:11] (Qg1+)} 39. Kf2 {[%eval -98,24] [%emt 0:00:40]} a4 {
[%eval -133,17] [%emt 0:00:06]} 40. Qd6+ {[%eval -98,23] [%emt 0:00:16]} Kb7 {
[%eval -144,18] [%emt 0:00:12]} 41. c5 {[%eval -98,24] [%emt 0:00:16]} Qc2+ {
[%eval -176,17] [%emt 0:00:06]} 42. Ke3 {[%eval -170,24] [%emt 0:01:14]} Qb3+ {
[%eval -224,19] [%emt 0:00:13]} 43. Qd3 {[%eval -225,24] [%emt 0:00:20] (Kf2)}
Qxd3+ {[%eval -568,21] [%emt 0:00:05] (Qe6+)} 44. Kxd3 {[%eval -309,23] [%emt
0:00:00]} a3 {[%eval -668,24] [%emt 0:00:09]} 45. Kc2 {[%eval -752,34] [%emt 0:
00:25] (Kc3)} Ka6 {[%eval -779,28] [%emt 0:00:16]} 46. Kb3 {[%eval -957,33]
[%emt 0:00:16]} Kb5 {[%eval -795,28] [%emt 0:00:11]} 47. h4 {[%eval -964,33]
[%emt 0:00:17] (Kxa3)} h5 {[%eval -933,26] [%emt 0:00:20]} 0-1
[/pgn]
Another 50 games match, but now Komodo 5.1MP with default setting, same testsuite, same time control 20m10s.beram wrote:And the match has ended with 52% for Komodo 5.1MP (with drawscore=0)beram wrote:Well it finally seems to be going to happen, a first win for Komodo 5.1MP against Houdini 3beram wrote:After another test with drawscore=0 settng for Komodo 5.1 I got better results for Komodo 45% against Houdini 3
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 512 Mb Hash 1 Houdini 3 x64 +17/=21/-12 55.00% 27.5/50 2 Komodo 5.1 64-bit +12/=21/-17 45.00% 22.5/50 Houdini3 - Komodo5.1 (Drawscore=0), Blitz 4m+2s 4 CPU AMD Phenom II X6 1090T, Noomen testsuite 2012 512 Mb Hash 1 Houdini 3 x64 +19/=28/-13 55.00% 33.0/60 2 Komodo 5.1 64-bit +13/=28/-19 45.00% 27.0/60
In my still running 50 games testmatch 4CPU at longer TC 20min 10sec, Komodo now leads with +11 =19 -8 after 38 games (53,9 % )
I don’t think Houdini can manage to equalize this in the upcoming twelve games, but I will keep you informed.
I have Komodo 5.1MP with the setting, drawscore=0 , because it seems to play better against Houdini with that modification.
So finally some good news, from my side, for Don and Larry
Thx for the free offered upgrade, your MP-engine has potention.
Code: Select all
Houdini3 - Komodo5.1 (Drawscore=0), Blitz 20m+10s 4 CPU AMD Phenom II X6 1090T, Bram testsuite 2.0, 1024 Mb Hash 1 Komodo 5.1 64-bit +14/=24/-12 52.00% 26.0/50 2 Houdini 3 x64 +12/=24/-14 48.00% 24.0/50
Code: Select all
Houdini3 - Komodo5.1 (Komodo default), Blitz 20m+10s 4CPU
AMD Phenom II X6 1090T, Bram testsuite 2.0, 1024 Mb Hash
1 Houdini 3 x64 +13/=28/-9 54.00% 27.0/50
2 Komodo 5.1 64-bit +9/=28/-13 46.00% 23.0/50