Why is Houdini 4 so clueless?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
mwyoung
Posts: 1636
Joined: Wed May 12, 2010 8:00 pm

Re: Why is Houdini 4 so clueless?

Post by mwyoung » Tue Mar 04, 2014 10:00 am

shrapnel wrote:Everyone havin' fun bashing H4 ? :roll:
H4 wouldn't make these mistakes given fast, modern hardware and the use of Syzygy EGTB and more time.

Hey Dude, the only one bashing H4 here is you. You are the one claiming H4 makes mistakes unless run on fast modern hardware. And needs Szygy EGTB to play well in the endgame, and just needs more time to play well.

When playing against the free program Stockfish.
Professing themselves to be wise, they became fools,
Take on me. foes 0

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Why is Houdini 4 so clueless?

Post by Sedat Canbaz » Tue Mar 04, 2014 10:33 am

mwyoung wrote: And needs Szygy EGTB to play well in the endgame, and just needs more time to play well.

When playing against the free program Stockfish.
Actually if we are testing the engines on same PC,
It is not necessary to set/use the large endgames
I think 4-men is quite enough

This issue (using large EGTB) is similar as using deep book in moves

I mean if we want to to see the real strength of the engines:
-It is a good idea the engines to be tested with a short neutral book
-And engines to be tested with small size of EGTB (4-men suits perfectly well)

Of course, if we are testing the engines on online severs or on official tournaments,
Then definitely: as far as possible large EGTB should be used
And there is no doubt that as far as possible the deepest and strongest books should be used too

Best,
Sedat

mwyoung
Posts: 1636
Joined: Wed May 12, 2010 8:00 pm

Re: Why is Houdini 4 so clueless?

Post by mwyoung » Tue Mar 04, 2014 10:41 am

Sedat Canbaz wrote:
mwyoung wrote: And needs Szygy EGTB to play well in the endgame, and just needs more time to play well.

When playing against the free program Stockfish.
Actually if we are testing the engines on same PC,
It is not necessary to set/use the large endgames
I think 4-men is quite enough

This issue (using large EGTB) is similar as using deep book in moves

I mean if we want to to see the real strength of the engines:
-It is a good idea the engines to be tested with a short neutral book
-And engines to be tested with small size of EGTB (4-men suits perfectly well)

Of course, if we are testing the engines on online severs or on official tournaments,
Then definitely: as far as possible large EGTB should be used
And there is no doubt that as far as possible the deepest and strongest books should be used too

Best,
Sedat
You need to address that comment to the dude.
Professing themselves to be wise, they became fools,
Take on me. foes 0

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Why is Houdini 4 so clueless?

Post by Sedat Canbaz » Tue Mar 04, 2014 10:49 am

mwyoung wrote:
Sedat Canbaz wrote:
mwyoung wrote: And needs Szygy EGTB to play well in the endgame, and just needs more time to play well.

When playing against the free program Stockfish.
Actually if we are testing the engines on same PC,
It is not necessary to set/use the large endgames
I think 4-men is quite enough

This issue (using large EGTB) is similar as using deep book in moves

I mean if we want to to see the real strength of the engines:
-It is a good idea the engines to be tested with a short neutral book
-And engines to be tested with small size of EGTB (4-men suits perfectly well)

Of course, if we are testing the engines on online severs or on official tournaments,
Then definitely: as far as possible large EGTB should be used
And there is no doubt that as far as possible the deepest and strongest books should be used too

Best,
Sedat
You need to address that comment to the dude.
Agreed.... :)

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Why is Houdini 4 so clueless?

Post by Sedat Canbaz » Tue Mar 04, 2014 10:53 am

mwyoung wrote:
Sedat Canbaz wrote:
mwyoung wrote: And needs Szygy EGTB to play well in the endgame, and just needs more time to play well.

When playing against the free program Stockfish.
Actually if we are testing the engines on same PC,
It is not necessary to set/use the large endgames
I think 4-men is quite enough

This issue (using large EGTB) is similar as using deep book in moves

I mean if we want to to see the real strength of the engines:
-It is a good idea the engines to be tested with a short neutral book
-And engines to be tested with small size of EGTB (4-men suits perfectly well)

Of course, if we are testing the engines on online severs or on official tournaments,
Then definitely: as far as possible large EGTB should be used
And there is no doubt that as far as possible the deepest and strongest books should be used too

Best,
Sedat
You need to address that comment to the dude.

But however,
I hardly doubt that Anıl will understand :)

jdart
Posts: 3816
Joined: Fri Mar 10, 2006 4:23 am
Location: http://www.arasanchess.org

Re: Why is Houdini 4 so clueless?

Post by jdart » Tue Mar 04, 2014 2:23 pm

Most engines are getting impressive depth at blitz time controls these days. And a blitz game in 2014 on current hardware is reaching depths that a standard time control game would have reached a few years ago.

I think some of the tactical success of Stockfish is possibly due to a different eval. Houdini eval is very good compared to most engines but there are cases where Stockfish is more effective at finding tactics.

--Jon

Dhanish
Posts: 42
Joined: Thu Jan 09, 2014 1:55 pm

Re: Why is Houdini 4 so clueless?

Post by Dhanish » Tue Mar 04, 2014 2:40 pm

Is the final position really won for White?
[D]8/8/3k4/1B3p1p/3K1P1P/6P1/8/3b4 w - - 45 77
Can the White king force an entry?

yanquis1972
Posts: 1762
Joined: Tue Jun 02, 2009 10:14 pm

Re: Why is Houdini 4 so clueless?

Post by yanquis1972 » Tue Mar 04, 2014 4:20 pm

good question, after about 10 minutes on my machine houdini at least evaluates it as winning for white but cant seem to find a path --

New game
[d]8/8/3k4/1B3p1p/3K1P1P/6P1/8/3b4 w - - 0 1

Analysis by Houdini 4 Pro x64:

77.Bc4 Ba4 78.Bf7 Bd1 79.Bg8 Bg4 80.Bd5 Be2 81.Bb3 Bg4 82.Bc4 Bd1 83.Ke3 Ba4 84.Be2
+/- (0.71) Depth: 7/12 00:00:07 6kN
+/- (0.84) Depth: 10/16 00:00:07 22kN
77.Ke3 Kc5 78.Bd3 Bg4 79.Bf1 Kd5 80.Ba6 Kc5 81.Bb7 Kc4 82.Bg2 Bd1 83.Bf1+ Kd5 84.Bh3 Ke6 85.Bg2 Ba4
+/- (0.90 ++) Depth: 10/16 00:00:07 24kN
+/- (0.82) Depth: 12/16 00:00:07 62kN
77.Bc4 Bf3 78.Ba6 Bd1 79.Bb7 Be2 80.Bd5 Bd1 81.Ba2 Bf3 82.Bc4 Bd1 83.Bf1 Bf3 84.Bd3 Bg4 85.Bb1 Bh3 86.Bd3 Bg4 87.Bb1
+/- (0.86) Depth: 12/19 00:00:07 84kN
+/- (0.86) Depth: 15/28 00:00:07 231kN
77.Bd3 Bg4 78.Bb1 Bh3 79.Bc2 Bg4 80.Bd3 Ke6 81.Kc5 Bf3 82.Bc2 Kf6 83.Kd6 Be2 84.Bb3 Bf3 85.Kd7 Bg4 86.Bc2 Be2 87.Ke8 Bg4 88.Bb3 Bf3 89.Bf7 Bg4 90.Kf8 Bf3 91.Be8 Bd1 92.Kg8 Bg4 93.Kh8 Ke7 94.Bg6 Kf6 95.Kh7 Bf3 96.Kh6 Bd5 97.Bxh5 Be6 98.Kh7 Bb3 99.Bf3 Be6 100.Bd1 Bf7 101.h5 Be6 102.h6 Kf7 103.Bh5+ Kf6 104.Kh8 Bb3 105.Be2 Bd5 106.h7 Be6 107.Bd3 Kf7 108.Be2 Bb3 109.Bf3 Kf6 110.Bc6 Be6 111.Bg2 Bf7 112.Bf1 Bd5 113.Bb5 Kf7 114.Be2 Kf6 115.Bd1 Bc4 116.Ba4 Be6 117.Bc6 Kg6 118.Bf3 Kf6 119.Be2 Bd5 120.Bd1
+/- (0.87) Depth: 15/28 00:00:07 280kN
+- (3.63) Depth: 37/80 00:09:15 2493MN, tb=3018508

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Why is Houdini 4 so clueless?

Post by Sedat Canbaz » Tue Mar 04, 2014 4:57 pm

Right now I noticed a game lost on time by Houdini 4 (actually Houdini was wining this game, but...)

Strange indeed...not sure why ?!

Has anybody seen a game forfeit on time by Houdini 4 ?

[Event "SCCT"]
[Site "?"]
[Date "2014.02.28"]
[Round "1.4"]
[White "Stockfish 260214 x64 6c"]
[Black "Houdini 4 x64 6c"]
[Result "1-0"]
[ECO "D47"]
[Annotator "0.37;0.12"]
[PlyCount "131"]
[EventDate "2014.02.27"]
[EventType "tourn"]
[Source "Sedat Canbaz"]
[TimeControl "180+2"]

{Intel(R) Core(TM) i7 CPU 970 @ 3.20GHz 3197 MHz W=33.2 plies; 9.
059kN/s; Perfect2014t.ctg B=20.6 plies; 8.614kN/s; Perfect2014t.ctg} 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]} c6 {[%eval 0,0] [%emt 0:00:00]} 3. Nf3 {
[%eval 0,0] [%emt 0:00:00]} d5 {[%eval 0,0] [%emt 0:00:00]} 4. Nc3 {
[%eval 0,0] [%emt 0:00:00]} e6 {[%eval 0,0] [%emt 0:00:00]} 5. e3 {
[%eval 0,0] [%emt 0:00:00]} Nbd7 {[%eval 0,0] [%emt 0:00:00]} 6. Bd3 {
[%eval 0,0] [%emt 0:00:00]} dxc4 {[%eval 0,0] [%emt 0:00:00]} 7. Bxc4 {
[%eval 0,0] [%emt 0:00:00]} b5 {[%eval 0,0] [%emt 0:00:00]} 8. Bd3 {
[%eval 0,0] [%emt 0:00:00]} Bb7 {[%eval 0,0] [%emt 0:00:00]} 9. e4 {
[%eval 37,23] [%emt 0:00:08]} b4 {[%eval 0,0] [%emt 0:00:00]} 10. Na4 {
[%eval 0,0] [%emt 0:00:00]} c5 {[%eval 12,22] [%emt 0:00:17]} 11. e5 {
[%eval 0,0] [%emt 0:00:00]} Nd5 {[%eval 14,21] [%emt 0:00:06]} 12. Nxc5 {
[%eval 42,24] [%emt 0:00:08]} Nxc5 {[%eval 4,21] [%emt 0:00:06]} 13. dxc5 {
[%eval 0,0] [%emt 0:00:00]} Bxc5 {[%eval 0,0] [%emt 0:00:00]} 14. Bb5+ {
[%eval 21,22] [%emt 0:00:12]} Ke7 {(Kf8) [%eval -27,20] [%emt 0:00:09]} 15. O-O
{(Bg5+) [%eval 5,23] [%emt 0:00:06]} h6 {[%eval -26,19] [%emt 0:00:04]} 16. Be2
{(Bc4) [%eval 13,23] [%emt 0:00:12]} Rc8 {(Qb6) [%eval -36,21] [%emt 0:00:09]}
17. Qd3 {(Ne1) [%eval -19,25] [%emt 0:00:05]} g5 {
(a5) [%eval -37,20] [%emt 0:00:07]} 18. a3 {[%eval -43,23] [%emt 0:00:06]} Qc7
{(bxa3) [%eval -53,21] [%emt 0:00:18]} 19. axb4 {[%eval -60,23] [%emt 0:00:05]}
Nxb4 {[%eval -54,19] [%emt 0:00:00]} 20. Qc3 {[%eval -55,25] [%emt 0:00:11]} a5
{[%eval -52,21] [%emt 0:00:11]} 21. Bd2 {(h4) [%eval -59,25] [%emt 0:00:09]}
Rhd8 {(Qb6) [%eval -52,19] [%emt 0:00:05]} 22. Rad1 {
[%eval -72,24] [%emt 0:00:07]} Rd5 {(Qb6) [%eval -53,20] [%emt 0:00:17]} 23. h3
{(h4) [%eval -105,22] [%emt 0:00:07]} Qb6 {[%eval -67,20] [%emt 0:00:25]} 24.
Qb3 {(Bc4) [%eval -120,26] [%emt 0:00:13]} Rcd8 {[%eval -89,21] [%emt 0:00:06]}
25. Bxb4 {(Kh2) [%eval -130,27] [%emt 0:00:14]} Qxb4 {
[%eval -93,23] [%emt 0:00:02]} 26. Qxb4 {[%eval -129,31] [%emt 0:00:05]} Bxb4 {
[%eval -89,25] [%emt 0:00:03]} 27. Rxd5 {[%eval -129,34] [%emt 0:00:03]} Rxd5 {
[%eval -80,26] [%emt 0:00:04]} 28. Rc1 {[%eval -129,35] [%emt 0:00:03]} Rc5 {
[%eval -81,25] [%emt 0:00:04]} 29. Rxc5 {[%eval -129,36] [%emt 0:00:03]} Bxc5 {
[%eval -79,24] [%emt 0:00:00]} 30. Kf1 {[%eval -129,36] [%emt 0:00:03]} Bd5 {
(a4) [%eval -78,24] [%emt 0:00:04]} 31. Bd1 {[%eval -129,31] [%emt 0:00:05]}
Bb6 {[%eval -77,24] [%emt 0:00:03]} 32. Ke2 {[%eval -123,32] [%emt 0:00:03]}
Bc4+ {[%eval -76,23] [%emt 0:00:03]} 33. Ke1 {[%eval -137,34] [%emt 0:00:05]}
Bc5 {(Bd3) [%eval -70,25] [%emt 0:00:05]} 34. Be2 {
(Ba4) [%eval -134,33] [%emt 0:00:09]} Bd5 {[%eval -73,24] [%emt 0:00:02]} 35.
Bd1 {(Kf1) [%eval -134,33] [%emt 0:00:03]} Bb7 {
(Kd8) [%eval -70,23] [%emt 0:00:04]} 36. Ke2 {
(Kf1) [%eval -137,29] [%emt 0:00:10]} Ba6+ {[%eval -77,24] [%emt 0:00:02]} 37.
Ke1 {[%eval -137,34] [%emt 0:00:03]} Bb5 {[%eval -79,25] [%emt 0:00:03]} 38.
Be2 {[%eval -137,34] [%emt 0:00:03]} Bc6 {[%eval -78,26] [%emt 0:00:06]} 39. b3
{(Bc4) [%eval -173,33] [%emt 0:00:32]} Bb4+ {[%eval -80,23] [%emt 0:00:02]} 40.
Kf1 {[%eval -155,34] [%emt 0:00:05]} Bc3 {(Bd5) [%eval -78,24] [%emt 0:00:05]}
41. Bc4 {[%eval -137,35] [%emt 0:00:03]} Kd7 {
(Bxf3) [%eval -89,22] [%emt 0:00:03]} 42. Ke2 {[%eval -177,31] [%emt 0:00:03]}
Kc7 {[%eval -104,24] [%emt 0:00:03]} 43. Kd3 {
(g3) [%eval -193,32] [%emt 0:00:09]} Ba1 {[%eval -147,23] [%emt 0:00:03]} 44.
Kc2 {[%eval -189,33] [%emt 0:00:02]} Kb6 {[%eval -147,21] [%emt 0:00:00]} 45.
Kb1 {[%eval -197,34] [%emt 0:00:04]} Bc3 {[%eval -143,23] [%emt 0:00:03]} 46.
Kc2 {[%eval -197,34] [%emt 0:00:02]} Bb4 {[%eval -143,21] [%emt 0:00:00]} 47.
Kd3 {[%eval -196,34] [%emt 0:00:02]} Ba8 {(Bb7) [%eval -139,24] [%emt 0:00:02]}
48. h4 {[%eval -242,35] [%emt 0:00:18]} Bxf3 {[%eval -139,22] [%emt 0:00:00]}
49. gxf3 {[%eval -351,32] [%emt 0:00:03]} gxh4 {[%eval -136,23] [%emt 0:00:03]}
50. Ke2 {[%eval -357,38] [%emt 0:00:04]} h3 {[%eval -143,15] [%emt 0:00:06]}
51. Kf1 {[%eval -372,40] [%emt 0:00:02]} Bd2 {
(Kc5) [%eval -136,18] [%emt 0:00:02]} 52. Bd3 {[%eval -418,40] [%emt 0:00:03]}
Kc6 {(Kc5) [%eval -136,17] [%emt 0:00:02]} 53. Bh7 {
[%eval -269,36] [%emt 0:00:04]} Kd7 {[%eval -135,12] [%emt 0:00:03]} 54. Bg8 {
(Bd3) [%eval -269,45] [%emt 0:00:02]} Ke7 {[%eval -147,16] [%emt 0:00:02]} 55.
Kg1 {[%eval -269,46] [%emt 0:00:01]} Bf4 {[%eval -152,17] [%emt 0:00:07]} 56.
Kf1 {(Bh7) [%eval -269,41] [%emt 0:00:01]} Bxe5 {
(h5) [%eval -166,15] [%emt 0:00:17]} 57. Bh7 {
(Kg1) [%eval -269,44] [%emt 0:00:02]} Bh2 {(Bf4) [%eval -157,16] [%emt 0:00:03]
} 58. Bb1 {(Bd3) [%eval -269,45] [%emt 0:00:01]} Kd6 {
(Bf4) [%eval -159,16] [%emt 0:00:02]} 59. Bh7 {
(Bd3) [%eval -269,46] [%emt 0:00:02]} Ke5 {(Kd7) [%eval -159,15] [%emt 0:00:10]
} 60. Bg8 {(Be4) [%eval -269,44] [%emt 0:00:02]} Kf6 {
[%eval -159,18] [%emt 0:00:03]} 61. Bh7 {[%eval -269,48] [%emt 0:00:01]} Kg5 {
(Bf4) [%eval -158,17] [%emt 0:00:02]} 62. Bd3 {
(Be4) [%eval -269,42] [%emt 0:00:03]} Kf4 {(h5) [%eval -156,15] [%emt 0:00:02]}
63. Bh7 {(Be2) [%eval -255,39] [%emt 0:00:02]} f5 {
(Kxf3) [%eval -155,18] [%emt 0:00:02]} 64. Bg8 {[%eval -390,30] [%emt 0:00:02]}
Ke5 {(e5) [%eval -152,16] [%emt 0:00:08]} 65. Bf7 {
[%eval -390,33] [%emt 0:00:02]} Kf6 {[%eval -151,16] [%emt 0:00:22]} 66. Be8 {
Time [%eval -464,36] [%emt 0:00:08]} 1-0

User avatar
sicilianquake87
Posts: 232
Joined: Wed Nov 14, 2012 2:24 pm
Location: Italy

Re: Why is Houdini 4 so clueless?

Post by sicilianquake87 » Tue Mar 04, 2014 6:40 pm

Sedat Canbaz wrote:Right now I noticed a game lost on time by Houdini 4 (actually Houdini was wining this game, but...)

Strange indeed...not sure why ?!

Has anybody seen a game forfeit on time by Houdini 4 ?
This is rare indeed. I thought Houdini TM was one of the best out there. :shock:
Someone spitting venom is annoying but harmless. He won't achieve anything. The real harm is done by nicely worded venom. (Ronald de Man)

Locked