Strelka 5.5 x32 is released....

[gleperlier]

Here is a patched version with its internal database disabled:
http://www.mediafire.com/download.php?f1jb8hd989hv8ja

Playing strength shouldn't be affected (I patched only 2 bytes in DB probing code).

Hi Richard,

Sorry if I missed a train but is Strelka Open source so that you can modify the code ? Or are you also part of Strelka team and not only father of Critter ?

Sorry I feel a little stupid to ask

[rvida]

Hi Richard,

Sorry if I missed a train but is Strelka Open source so that you can modify the code ? Or are you also part of Strelka team and not only father of Critter ?

Sorry I feel a little stupid to ask

I don't have Strelka5.5 sources. I hex-edited the binary.

[Laskos]

Thank you for doing the test. The results confirm what I above: it's not a good idea to use a ultra-fast pure increment TC like 0.05"/move to compare engines with different Time Management strategies.

Generally speaking I don't recommend using pure increment matches for Houdini, its TM is not well adapted to this situation. The shortest TC I use for testing Houdini is 1"+0.05", but even for this TC it doesn't make sense to play engine matches between different engines. The timings are too fast, and random choices like the timing safety buffer begin to influence the result.

Looking at all this, it appears that Strelka 5.5 is just a late April Fools joke... well done, Yuri!

Robert

I assummed that Strelka 5.5 has the same time management (and timing buffer) as Houdini 1.5 has . If I remember well, I had no such stringent problems with Strelka 5.1, that's why I pitted S5.5 exactly against H1.5. And those test suites incredible results were not helping too.

Kai

[ATOMICC]

...
Here is a patched version with its internal database disabled:
http://www.mediafire.com/download.php?f1jb8hd989hv8ja

Playing strength shouldn't be affected (I patched only 2 bytes in DB probing code).

Awesome, Richard!! If you could do the same when the 64-bit version comes out, that'd be great. Then I, and others who use opening suites, could test it safely.

BIG +1 here.

[Houdini]

I assummed that Strelka 5.5 has the same time management (and timing buffer) as Houdini 1.5 has . If I remember well, I had no such stringent problems with Strelka 5.1, that's why I pitted S5.5 exactly against H1.5. And those test suites incredible results were not helping too.

Kai

The eval and search of Strelka 5 are 99% identical to Houdini 1.5a, but all the surrounding code (Time Management, UCI input/output handling etc) appears to be Ivanhoe-inspired or original.

Certainly original is forgetting to divide by 2 to produce centipawns in the output (Houdini 1.5 uses 1/200 pawn internally), this means that all Strelka evals are a factor of 2 too high.

Robert

[pichy]

Hello,

Here is a game I just played with Vitruvius (2 CPU) versus Strelka on Playchess :


[Event "Rated game, 60 m + 15s"]
[Site "Engine Room"]
[Date "2012.04.24"]
[Round "?"]
[White "Strelka 5.5"]
[Black "Vitruvius_1.11_HEM"]
[Result "0-1"]
[ECO "C13"]
[WhiteElo "2498"]
[Annotator "0.09;0.00"]
[PlyCount "51"]
[EventDate "2012.04.24"]
[TimeControl "3600+15"]

{Strelka 5.5: 20.9 plies; 1 957kN/s Intel(R) Core(TM) i7 CPU X 980 @ 3.
33GHz, (6 CPUs),} 1. e4 {B 0} e6 {B 0} 2. d4 {B 0
} d5 {B 0} 3. Nc3 {B 0} Nf6 {B 0} 4. Bg5 {B 0} dxe4 {B 0} 5. Nxe4 {B 0} Be7 {
B 0} 6. Bxf6 {B 0} Bxf6 {B 0} 7. Nf3 {B 0} O-O {B 0} 8. Bc4 {B 0} Nc6 {B 0} 9.
c3 {B 0} e5 {B 0} 10. d5 {B 0} Ne7 {B 0} 11. O-O {B 0} Ng6 {0.00/23 115} 12. a4
{B 0 (Ne4xf6+)} Be7 {-0.10/24 295} 13. a5 {B 0} Kh8 {-0.08/23 49} 14. b4 {0.09/
21 130} Rb8 {-0.12/24 0} 15. Nc5 {-0.15/20 185 (Rf1-e1)} b6 {-0.46/23 182
(Lxc5)} 16. axb6 {-0.90/20 84 (Nc5-e4)} axb6 {-0.50/23 84} 17. Nd3 {-0.90/21
58 (Nc5-a6)} e4 {-0.60/23 95 (Ld6)} 18. Nd4 {-1.28/21 85 (Nf3-e5)} Bb7 {-0.89/
24 270} 19. Nb2 {-1.87/22 123} Bxd5 {-0.89/23 0} 20. Qc2 {-1.93/21 78 (Qd1-b3)}
Bb7 {-0.93/22 141} 21. Rfd1 {-1.96/21 31 (Bc4-a6)} Qc8 {-1.15/23 118} 22. b5 {
-2.01/22 0} Ra8 {-1.18/23 53} 23. Rxa8 {-2.09/22 99} Qxa8 {-1.18/24 0} 24. Bb3
{-1.94/22 82} Nf4 {-1.22/24 0 (Se5)} 25. Nc4 {-1.99/22 95} Nd3 {-1.13/24 0} 26.
Ne3 {-2.12/21 60,Strelka 5.5 gibt auf} 0-1

Not that impressive ....

One game or 100 games does not tell you much, but you should reverse the engines playing the same opening to make it fair and that will remove any possibility of one side being played better in the opening than the other side or if you want to make it more fair let both engines play without Opening

[gleperlier]

Hello,

Here is a game I just played with Vitruvius (2 CPU) versus Strelka on Playchess :


[Event "Rated game, 60 m + 15s"]
[Site "Engine Room"]
[Date "2012.04.24"]
[Round "?"]
[White "Strelka 5.5"]
[Black "Vitruvius_1.11_HEM"]
[Result "0-1"]
[ECO "C13"]
[WhiteElo "2498"]
[Annotator "0.09;0.00"]
[PlyCount "51"]
[EventDate "2012.04.24"]
[TimeControl "3600+15"]

{Strelka 5.5: 20.9 plies; 1 957kN/s Intel(R) Core(TM) i7 CPU X 980 @ 3.
33GHz, (6 CPUs),} 1. e4 {B 0} e6 {B 0} 2. d4 {B 0
} d5 {B 0} 3. Nc3 {B 0} Nf6 {B 0} 4. Bg5 {B 0} dxe4 {B 0} 5. Nxe4 {B 0} Be7 {
B 0} 6. Bxf6 {B 0} Bxf6 {B 0} 7. Nf3 {B 0} O-O {B 0} 8. Bc4 {B 0} Nc6 {B 0} 9.
c3 {B 0} e5 {B 0} 10. d5 {B 0} Ne7 {B 0} 11. O-O {B 0} Ng6 {0.00/23 115} 12. a4
{B 0 (Ne4xf6+)} Be7 {-0.10/24 295} 13. a5 {B 0} Kh8 {-0.08/23 49} 14. b4 {0.09/
21 130} Rb8 {-0.12/24 0} 15. Nc5 {-0.15/20 185 (Rf1-e1)} b6 {-0.46/23 182
(Lxc5)} 16. axb6 {-0.90/20 84 (Nc5-e4)} axb6 {-0.50/23 84} 17. Nd3 {-0.90/21
58 (Nc5-a6)} e4 {-0.60/23 95 (Ld6)} 18. Nd4 {-1.28/21 85 (Nf3-e5)} Bb7 {-0.89/
24 270} 19. Nb2 {-1.87/22 123} Bxd5 {-0.89/23 0} 20. Qc2 {-1.93/21 78 (Qd1-b3)}
Bb7 {-0.93/22 141} 21. Rfd1 {-1.96/21 31 (Bc4-a6)} Qc8 {-1.15/23 118} 22. b5 {
-2.01/22 0} Ra8 {-1.18/23 53} 23. Rxa8 {-2.09/22 99} Qxa8 {-1.18/24 0} 24. Bb3
{-1.94/22 82} Nf4 {-1.22/24 0 (Se5)} 25. Nc4 {-1.99/22 95} Nd3 {-1.13/24 0} 26.
Ne3 {-2.12/21 60,Strelka 5.5 gibt auf} 0-1

Not that impressive ....

One game or 100 games does not tell you much, but you should reverse the engines playing the same opening to make it fair and that will remove any possibility of one side being played better in the opening than the other side or if you want to make it more fair let both engines play without Opening

Well this is not a game played on my computer vs my computer but one played on Playchess, I just gave it for example.

Gab

[pichy]

I assummed that Strelka 5.5 has the same time management (and timing buffer) as Houdini 1.5 has . If I remember well, I had no such stringent problems with Strelka 5.1, that's why I pitted S5.5 exactly against H1.5. And those test suites incredible results were not helping too.

Kai

The eval and search of Strelka 5 are 99% identical to Houdini 1.5a, but all the surrounding code (Time Management, UCI input/output handling etc) appears to be Ivanhoe-inspired or original.

I found this to be very surprising that the evaluation of Strelka is so close to Houdini
[d]8/8/5N2/8/8/8/4Kpb1/5Qbk w - - 0 1
FEN: 8/8/5N2/8/8/8/4Kpb1/5Qbk w - - 0 1

Strelka55:
2 00:00 95 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
3 00:00 160 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
4 00:00 281 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
5 00:00 475 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
6 00:00 636 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
7 00:00 1.000 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
8 00:00 1.294 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
9 00:00 1.807 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
10 00:00 3.109 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
11 00:00 4.544 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
12 00:00 6.809 0 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
13 00:00 9.981 665.400 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
14 00:00 14.614 974.266 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
15 00:00 20.957 1.397.133 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
16 00:00 27.912 900.387 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
17 00:00 37.249 1.034.694 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
18 00:00 51.876 1.179.000 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
19 00:00 71.465 1.299.363 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
20 00:00 98.799 1.372.208 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
21 00:00 132.663 1.474.033 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
22 00:00 178.850 1.490.416 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
23 00:00 239.351 1.595.673 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
24 00:00 317.578 1.689.244 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
25 00:00 403.062 1.737.336 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
26 00:00 518.591 1.800.663 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
27 00:00 688.663 1.861.251 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
28 00:00 888.629 1.898.779 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
29 00:00 1.179.605 1.943.336 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
30 00:00 1.512.902 1.969.924 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
31 00:01 1.916.930 1.970.123 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
32 00:01 2.344.046 1.948.500 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
33 00:01 2.893.032 1.966.711 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
34 00:01 3.586.839 1.992.688 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
35 00:02 4.707.630 2.002.394 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
36 00:02 5.641.239 1.996.898 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
37 00:03 7.029.369 2.002.669 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
38 00:04 8.524.237 1.995.373 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
39 00:05 10.582.457 1.987.315 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
40 00:06 13.177.133 1.972.034 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
41 00:08 15.716.587 1.949.948 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
42 00:10 19.649.998 1.926.281 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
43 00:25 41.975.776 1.676.817 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
44 00:32 52.720.922 1.650.882 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
45 00:43 70.086.600 1.617.096 +0.03 Qf1xf2 Bg1xf2 Ke2xf2
46 00:55 88.967.694 1.600.224 +0.03 Qf1xf2 Bg1xf2 Ke2xf2

==================================================================================================================================================

And here is Houdini Analysis very Similar to Strelka 5.5
FEN: 8/8/5N2/8/8/8/4Kpb1/5Qbk w - - 0 1

Houdini_15_w32:
4/8 00:00 363 0 0.00 Qf1xf2 Bg1xf2 Ke2xf2
5/8 00:00 562 0 0.00 Qf1xf2 Bg1xf2 Ke2xf2
6/8 00:00 727 0 0.00 Qf1xf2 Bg1xf2 Ke2xf2
7/12 00:00 1.104 0 0.00 Qf1xf2 Bg1xf2 Ke2xf2
8/12 00:00 1.422 0 0.00 Qf1xf2 Bg1xf2 Ke2xf2
9/12 00:00 1.955 122.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
10/15 00:00 3.555 222.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
11/17 00:00 5.300 331.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
12/17 00:00 7.855 490.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
13/18 00:00 11.780 736.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
14/20 00:00 18.041 1.127.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
15/20 00:00 26.474 1.654.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
16/21 00:00 34.632 402.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
17/24 00:00 45.656 475.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
18/24 00:00 63.785 596.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
19/24 00:00 90.013 750.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
20/25 00:00 125.981 926.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
21/25 00:00 177.499 1.130.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
22/27 00:00 234.842 1.326.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
23/30 00:00 313.370 1.536.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
24/30 00:00 417.257 1.703.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
25/30 00:00 548.463 1.938.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
26/32 00:00 703.919 2.159.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
27/33 00:00 929.274 2.407.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
28/33 00:00 1.264.410 2.678.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
29/36 00:00 1.569.807 2.849.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
30/36 00:00 2.006.951 3.022.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
31/37 00:00 2.651.064 3.240.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
32/38 00:01 3.184.524 3.366.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
33/39 00:01 4.024.214 3.505.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
34/39 00:01 4.876.713 3.609.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
35/39 00:01 6.162.355 3.633.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
36/43 00:02 7.948.507 3.363.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
37/43 00:02 9.308.228 3.464.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
38/43 00:03 11.676.647 3.568.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
39/43 00:03 13.805.699 3.635.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
40/45 00:04 17.133.892 3.732.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
41/45 00:06 22.217.232 3.759.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
42/47 00:07 26.959.751 3.787.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
43/48 00:08 32.234.626 3.813.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
44/48 00:18 64.510.973 3.523.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
45/49 00:27 94.020.365 3.434.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2
46/53 00:43 143.131.007 3.322.000 0.00 Qf1xf2 Bg1xf2 Ke2xf2

[Dicaste]

Wow, it's quite awkward.

8/8/5N2/8/8/8/4Kpb1/5Qbk w - - 0 1

Analysis by Houdini 1.5a x64:

1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 7/12 00:00:00 1kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 8/12 00:00:00 1kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 9/12 00:00:00 2kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 10/14 00:00:00 3kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 11/16 00:00:00 5kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 12/16 00:00:00 8kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 13/19 00:00:00 11kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 14/20 00:00:00 17kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 15/20 00:00:00 25kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 16/22 00:00:00 34kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 17/22 00:00:00 44kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 18/23 00:00:00 62kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 19/24 00:00:00 86kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 20/24 00:00:00 123kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 21/26 00:00:00 168kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 22/28 00:00:00 229kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 23/28 00:00:00 299kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 24/30 00:00:00 401kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 25/32 00:00:00 546kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 26/32 00:00:00 714kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 27/32 00:00:00 881kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 28/32 00:00:00 1203kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 29/33 00:00:00 1423kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 30/38 00:00:00 1935kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 31/38 00:00:00 2579kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 32/38 00:00:00 3195kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 33/38 00:00:00 4004kN
1.Qxf2 Bxf2 2.Kxf2
= (0.00) Depth: 34/40 00:00:01 5051kN

(Eren Yavuz, 26.04.2012)

------------------------------------------------------------------------

New game
8/8/5N2/8/8/8/4Kpb1/5Qbk w - - 0 1

Analysis by Houdini 1.5a x64 w/ Gaviota Tablebases:

1.Ng4
+/= (0.48 ++) Depth: 7/15 00:00:00 3kN
1.Ng4
+/= (0.62 ++) Depth: 7/15 00:00:00 3kN
1.Ng4
+/- (0.94 ++) Depth: 7/15 00:00:00 3kN
1.Ng4
+- (1.96 ++) Depth: 7/15 00:00:00 3kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kf1 6.Qd1+ Kg2 7.Qd5+ Kf1
+- (4.48) Depth: 7/15 00:00:00 3kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kf1 6.Qd1+ Kg2 7.Qd5+ Kf1
+- (4.48) Depth: 8/15 00:00:00 3kN
1.Ng4
+- (4.67 ++) Depth: 9/19 00:00:00 4kN
1.Ng4
+- (5.04 ++) Depth: 9/21 00:00:00 4kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Ke3 Kg3 6.Qf3+ Kh2
+- (5.13) Depth: 9/21 00:00:00 5kN
1.Ng4 Bf3+
+- (4.92 --) Depth: 10/21 00:00:00 7kN
1.Ng4
+- (5.35 ++) Depth: 10/21 00:00:00 7kN
1.Ng4
+- (5.78 ++) Depth: 10/21 00:00:00 8kN
1.Ng4
+- (6.94 ++) Depth: 10/21 00:00:00 8kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kh2 5.Qf3 f1R 6.Qxf1 Bc5 7.Qe2+ Kg1
+- (7.47) Depth: 10/21 00:00:00 9kN
1.Ng4 Bf3+
+- (7.16 --) Depth: 11/23 00:00:00 11kN
1.Ng4 Bf3+
+- (6.54 --) Depth: 11/24 00:00:00 16kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Ke3 Kg3 6.Qf3+ Kh4 7.Ke2 Kg5 8.Kf1 Kh4 9.Qe4+ Kg3 10.Qd3+ Kf4
+- (6.74) Depth: 11/24 00:00:00 16kN
1.Ng4
+- (7.01 ++) Depth: 12/24 00:00:00 19kN
1.Ng4
+- (7.57 ++) Depth: 12/24 00:00:00 21kN
1.Ng4
+- (9.12 ++) Depth: 12/27 00:00:00 27kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Qf4+ Kg2 9.Qe4+ Kf2 10.Qc2+ Ke3 11.Qb3+ Ke4 12.Qxb6
+- (10.68) Depth: 12/27 00:00:00 32kN
1.Ng4 Bf3+
+- (10.23 --) Depth: 13/29 00:00:00 48kN
1.Ng4
+- (11.12 ++) Depth: 13/29 00:00:00 48kN
1.Ng4
+- (12.00 ++) Depth: 13/35 00:00:00 61kN
1.Ng4
+- (14.54 ++) Depth: 13/35 00:00:00 66kN
1.Ng4
+- (22.03 ++) Depth: 13/35 00:00:00 83kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 13/35 00:00:00 87kN
1.Ng4 Bf3+
+- (142.81 --) Depth: 14/35 00:00:00 91kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 14/35 00:00:00 92kN
1.Ng4 Bf3+
+- (142.81 --) Depth: 15/35 00:00:00 99kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 15/35 00:00:00 102kN
1.Ng4 Bf3+
+- (142.81 --) Depth: 16/35 00:00:00 110kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 16/35 00:00:00 114kN
1.Ng4 Bf3+
+- (142.81 --) Depth: 17/35 00:00:00 128kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 17/35 00:00:00 138kN
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 18/35 00:00:00 157kN, tb=1
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Qg3+ Kh1 7.Qh3+ Bh2+ 8.Kf3 f1R+ 9.Qxf1+ Bg1 10.Qg2#
+- (#10) Depth: 19/35 00:00:00 192kN, tb=1
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 20/35 00:00:00 256kN, tb=1
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 21/35 00:00:00 322kN, tb=1
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 22/35 00:00:00 421kN, tb=2
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 23/35 00:00:00 581kN, tb=3
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 24/35 00:00:00 781kN, tb=11
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 25/35 00:00:00 1052kN, tb=21
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 26/35 00:00:00 1368kN, tb=62
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 27/35 00:00:00 1762kN, tb=85
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 28/35 00:00:00 2370kN, tb=127
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 29/37 00:00:00 3472kN, tb=170
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 30/37 00:00:00 4547kN, tb=250
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 31/37 00:00:00 5864kN, tb=363
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 32/42 00:00:01 8193kN, tb=532
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 33/42 00:00:01 12307kN, tb=653
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 34/46 00:00:02 17591kN, tb=1054
1.Ng4 Bf3+ 2.Ke3 Bxg4 3.Kf4 Be2 4.Qxe2 Kg2 5.Qf3+ Kh2 6.Kg4 f1B 7.Qxf1 Bb6 8.Kf3 Bg1 9.Qg2#
+- (#9) Depth: 35/46 00:00:02 20709kN, tb=1387

(Eren Yavuz, 26.04.2012)

Strange... Houdini 1.5a x64 with Gaviota working fine but without "NO!"

[Karol Majewski]

Strange... Houdini 1.5a x64 with Gaviota working fine but without "NO!"

That's because Houdini 1.5a evaluates minor vs minor endgames as a draw. And so does Strelka.