[Event "CG.com Masters vs Machines Invitational (2011)"]

[Site "Yahoo! chess"]
[Date "2011.04.16"]
[Round "1"]
[White "Houdini 1.5a w32/<Golden Executive>"] [Black "Stockfish 2.0/<Kutztown46>"]

[Result "1/2-1/2"]
[ECO "B51"]
[PlyCount "120"]
[TimeControl "105 minutes/no increments"]

I started the game using the fixed time approach: 105 less 10 minutes for operator lag time, for a 70 moves game.

For the first 8 moves I used the Opening Explorer in combination with the suggested move by Houdini (in this approach I just can not select by myself directly a move from the OE, that is only possible in the Infinite Analysis approach).

After the 8 move, the engine was on its own and I just forced to move (pressing my spacebar) if I felt the engine reached a reasonable depth or the move was so obvious.

The opening was Sicilian, Canal-Sokolsky (Rossolimo) Attack (B51)

White Move/Eval/Depth/Time(seconds)/ Black Move/Time(seconds)

1. e4 0.30/17 7 c5 97
2. Nf3 0.32/1928 d6 35
3. Bb5+ 0.31/17 14 Nc6 51
4. O-O 0.25/19 7 Bd7 35
5. c3 0.21/20 56 Nf6 40
6. Re1 0.21/18 26 a6 47
7. Bf1 0.36/19 28 Bg4 41
8. d3 0.31/18 19 e6 40
9. h3 0.17/21 93 Bh5 54
10. Nbd2 0.19/21 41Be7 83
11. g4 0.18/22 4 Bg6 27
12. Nh4 0.18/22 25 Nd7 93
13. Nxg6 0.17/21 74 hxg6 27
14. Nf3 0.16/22 140

click for larger view

That was the last move you can find in the OE

14...O-O 161
15. d4 0.27/21 99 Re8 166
<16. Be3 0.32/21 94> cxd4 87

I think that was when white lose the slight advantage. Better is <16.d5> it´s more agressive and gain more space for white, also puts a pawn at d5:

click for larger view

Analysis by Houdini 1.5a w32 <depth 25>

16..17.exd5 Nce5 18.Nxe5 dxe5 19.b3 Bd6 20.Qc2 Qh4 21.Bg2 Nf6 22.c4 Rad8 23.Rb1 e4 24.g5 Nh5 25.Rxe4 Nf4 26.Bxf4 Bxf4 27.b4 b6 28.bxc5 bxc5 29.Rbe1 Rxe4 30.Qxe4 Qxg5 31.Qe7 Qxe7 32.Rxe7 Rb8 33.Ra7 Rb6 34.Bf1 < (0.34) Depth: 25>

17. cxd4 0.22/21 127 d5 70
18. exd5 0.22/20 44 exd5 42
19. Rc1 0.24/18 34 Bd6
20. Bg2 0.11/21 96 Nf6
21. a3 Qd7
22. b4 -0.06/18 6 Bc7 183
23. Nd2 0.17/19 80 Qd6 150
24. Kf1 0.18/21 69 Re7 77
25. Nb3 0.23/21 89 Rae8 49
26. Nc5 0.11/21 124 Nd8 78
27. g5 0.11/21 81 Nh5 67
28. Qg4 0.15/21 31 Ne6 59 offering a pawn

1546907: Houdini 1.5a w32,G - Stockfish 2.0,K, 44m/80 Mexico

click for larger view

If White takes the pawn with the knight,the knight has to go far away from the action and the next moves are kind of forced leading to a lot of exchanges and a drawish position (like ocurred at the end of the game).

After 29.Nxb7 Qh2 30.Na5 Nef4 31.Bxf4 Nxf4 32.Rxe7 Rxe7 33.Nc6 Re6:

1546907: Houdini 1.5a w32,G - Stockfish 2.0,K, 44m/80 Mexico

click for larger view

Analysis by Houdini 1.5a w32 <depth 26>

1. = (0.00): 34.Qg3 Qxg3 35.fxg3 Nd3 36.Rd1 Bxg3 37.Bxd5 Re3 38.Ne5 Nxe5 39.dxe5 Bxe5 40.a4 Kf8 41.b5 axb5 42.axb5 Rxh3 43.b6 Rh2 44.Bg2 Ke7 45.b7 Rh8 46.Ke2 f6 47.Bc6 Rd8 48.Bd5 Rh8 49.Ke3 Kd7 50.Bf7+ Kc7 51.gxf6 gxf6 52.Bxg6 Kb8 53.Ke4

2. = (0.00): 34.Ne5 Bxe5 35.Rc8+ Kh7 36.dxe5 Rxe5 37.Qh4+ Nh5 38.Bf3 Rxg5 39.Qxg5 Qxh3+ 40.Kg1 Qxc8 41.Bxh5 gxh5 42.Qxh5+ Kg8 43.Qxd5 Qg4+ 44.Qg2 Qd1+ 45.Qf1 Qg4+ 46.Qg2 Qd1+

29. Bd2 0.12/21 77 Qh2 114

30. Nd3 0.18/21 53 Kh7 78

Here the position reachs its critical point, so complicated:

click for larger view

31. Re2 0.21/20 91 Rd8 89
32. Rce1 0.26/18 44 Bd6 70
33. Be3 0.27/18 50 Nc7 76
34. a4 0.30/20 63 b6 107
35. Bd2 0.21/20 137 Ne6 44
36. Rc1 0.23/18 45 Rc7 85

After some maneuvering, Stockfish offers to begin with the exchanges with Rc7:

click for larger view

37. Rxc7 0.22/20 34 Bxc7 28
38. Re1 0.19/20 35 Nef4 108
39. Bxf4 0.15/25 33 Nxf4 33
40. Nxf4 0.14/25 13 Qxf4 35
41. Qxf4 0.13/25 5 Bxf4 24
42. Re7 0.14/25 22 Bxg5 52
43. Rxf7 0.11/25 9

After <Rxf7> the position and the opposite colour bishops starts to look drawish.

click for larger view

43..Bd2 116
44. b5 0.11/28 7 axb5 50
45. axb5 0.11/28 8 Bc3 111
46. Rf4 0.10/28 4 g5 110
47. Rg4 0.11/27 6 Kg6 42
48. h4 0.11/27 53 Kf6 117
49. Rxg5 0.13/24 64 Bxd4 81
50. Bxd5 0.14/26 41

With 50.Bxd5 Houdini is a pawn up, but seems like is not enough to win the game:

click for larger view

50..Be594
51. Be4 0.14/26 19 Rd4 69
52. f3 0.14/25 50 Ke6 55
53. h5 0.14/25 50 Rb4 102
54. Bc6 0.14/25 25 Bf6 82
55. Rg4 0.14/27 96 Rb2 45
56. Re4+ 0.14/26 22 Kd6 52
57. Rc4 0.14/25 23 Ke5 185
58. Be8 0.14/2563 Kd6 71
59. Re4 0.14/25 0 Bg5 84
60. Rg4 0.14/26 115 Bf6 55
1/2-1/2

After 60..Bf6 <kutztown46> offered a draw and I accepted.The last 10 evals for Houdini were 0.14.At the end of the game clocks were: 37:38 for Stockfish and 42:01 for Houdini.

This is the final position:

click for larger view

Some thoughts and analysis on the Houdini - Stockfish game from April 16:

Here is the game score:

1.e4 c5 2.Nf3 d6 3.Bb5+ Nc6 4.0–0 Bd7 5.c3 Nf6 6.Re1 a6 7.Bf1 Bg4 8.d3 e6 9.h3 Bh5 10.Nbd2 Be7 11.g4 Bg6 12.Nh4 Nd7 13.Nxg6 hxg6 14.Nf3 0–0 15.d4 Re8 16.Be3 cxd4 17.cxd4 d5 18.exd5 exd5 19.Rc1 Bd6 20.Bg2 Nf6 21.a3 Qd7 22.b4 Bc7 23.Nd2 Qd6 24.Kf1 Re7 25.Nb3 Rae8 26.Nc5 Nd8 27.g5 Nh5 28.Qg4 Ne6 29.Bd2 Qh2 30.Nd3 Kh7 31.Re2 Rd8 32.Rce1 Bd6 33.Be3 Nc7 34.a4 b6 35.Bd2 Ne6 36.Rc1 Rc7 37.Rxc7 Bxc7 38.Re1 Nef4 39.Bxf4 Nxf4 40.Nxf4 Qxf4 41.Qxf4 Bxf4 42.Re7 Bxg5 43.Rxf7 Bd2 44.b5 axb5 45.axb5 Bc3 46.Rf4 g5 47.Rg4 Kg6 48.h4 Kf6 49.Rxg5 Bxd4 50.Bxd5 Be5 51.Be4 Rd4 52.f3 Ke6 53.h5 Rb4 54.Bc6 Bf6 55.Rg4 Rb2 56.Re4+ Kd6 57.Rc4 Ke5 58.Be8 Kd6 59.Re4 Bg5 60.Rg4 Bf6 1/2-1/2

I was playing black, operating Stockfish. I took advantage of my option to choose the first eight moves of the game. I wanted Stockfish to have winning chances as black against Houdini, so I chose a Sicilian against 1. e4. I also wanted a complicated game. I was not disappointed in this regard. A patzer such as myself would stand no chance in a game this complicated.

<Golden Executive> has stated that he believes 16. Be3 was not the best for white and that 16. d5 would have been stronger. Stockfish agrees that 16. d5 would have been stronger. However, Stockfish contends that 16. Bf4 would have been strongest of all.

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (30-ply):

1. = (0.24): 16.Bf4 cxd4 17.cxd4 Qb6 18.Rc1 d5 19.e5 Na5 20.Qa4 Qd8 21.Qc2 Nc6 22.Qd1 Nb6 23.Bd2 Nd7 24.Qb3 Qb8 25.Bf4 Qa7 26.Red1 Rac8 27.a3 Bd8 28.Qe3 Bc7 29.Bd3

2. = (0.20): 16.d5 exd5 17.Qxd5 Qc7 18.Bc4 Nde5 19.Nxe5 Nxe5 20.Rd1 g5 21.Be3 Bf6 22.Be2 Re6 23.b4 cxb4 24.cxb4 Ng6 25.Rac1 Qe7 26.Qd2 Rxe4 27.Bd3 Re6 28.Bxg6 fxg6 29.Rc2 Re8 30.a3 Qf7 31.Rdc1

3. = (0.12): 16.Be3 cxd4 17.cxd4 d5 18.exd5 exd5 19.Qb3 Nf6 20.Ne5 Na5 21.Qd3 Bd6 22.Bg2 Bxe5 23.dxe5 Rxe5 24.Rad1 Nc6 25.g5 Ne4 26.Qb3 Na5 27.Qb4 Re6 28.Bf4 Nc6 29.Qxb7 Nc5 30.Qc7 Qxc7 31.Bxc7 Rc8

Sliding forward: <16. Bf4 cxd4 17. cxd4 Qb6 18. Rb1 Rac8 19. Be3> brings us here:

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (28-ply):

1. ² (0.52): 19...Qc7 20.Rc1 Qb8 21.Qb3 Qa8 22.a4 Bd8 23.d5 exd5 24.exd5 Nce5 25.Nxe5 Nxe5 26.Be2 Rxc1 27.Rxc1 Nd7 28.Rc2 Ba5 29.Qa3 Qb8 30.Bf4 Ne5 31.b4 Bb6 32.a5 Bd4 33.b5 Bc5 34.Qa4 axb5 35.Qxb5

2. ² (0.56): 19...Qd8 20.Rc1 Nf6 21.d5 exd5 22.exd5 Ne5 23.Rxc8 Qxc8 24.Bf4 Nxf3+ 25.Qxf3 Qc2 26.Qa3 Ne4 27.Re2 Qc4 28.g5 Bf8 29.Qb3 Qc8 30.Qf3 Nc5 31.Rxe8 Qxe8 32.Qe3 Qa4 33.a3 Nb3 34.Kg2 b5 35.Bg3

3. ² (0.60): 19...d5 20.exd5 exd5 21.b4 Qd8 22.b5 axb5 23.Rxb5 Nf6 24.Rxb7 Ne4 25.Ne5 Nxe5 26.dxe5 Bc5 27.Bxc5 Nxc5 28.Rb5 d4 29.Bc4 Re7 30.Qd2 Rd7 31.Rb4 Ne6 32.Bxe6 fxe6 33.a4 d3 34.Rb2 Rd5 35.Re3

4. ± (0.80): 19...Qa5 20.b4 Qc7 21.b5 axb5 22.Rxb5 Qb8 23.Qa4 Rc7 24.Bd3 Rec8 25.Reb1 b6 26.d5 Nce5 27.Nxe5 dxe5 28.dxe6 fxe6 29.Qb3 Kf7 30.Qd1 Rc3 31.a4 Qc7 32.R5b2 Rd8 33.Qf3+ Kg8 34.Qe2 Nc5 35.Bc2

I reviewed the entire game with Stockfish, and white's 16th move seems to represent the single best place where either side could have greatly improved upon what was played.

Other possible improvements are:

14...Rc8 instead of O-O

19. Qb3 instead of Rc1

27. Qc2 instead of g5

30...Bb6 instead of Kh7

34. Bc1 instead of a4

As I mentioned, this was a complicated game. For quite a while, there were very few forced or semi-forced moves, and both sides frequently had a large number of plausible moves to choose from. For example, look at black's 21st move:

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (25-ply):

1. = (-0.04): 21...Qd7 22.Qd3 Re7 23.Bg5 Rxe1+ 24.Nxe1 Ne4 25.Bxe4 dxe4 26.Qxe4 Nxd4 27.Nf3 Ne6 28.Be3 Re8 29.Qc4 Bc7 30.Kf1 Bd8 31.Ne5 Qd6 32.Nf3 Bb6 33.Bxb6 Qxb6 34.b4 Rd8 35.Qe4 Qb5+ 36.Kg1 Qa4 37.Qxb7

2. = (0.00): 21...Re7 22.Qb3 Qd7 23.Ne5 Bxe5 24.dxe5 Rxe5 25.Red1 Rd8 26.Bf4 Re7 27.Be3 Re5

3. = (0.00): 21...Qb8 22.Qd3 Qc7 23.Qb3 Qd7 24.Nd2 Rad8 25.Qd3 Bc7 26.Bg5 Qd6 27.Rxe8+ Rxe8 28.Be3 Bb6 29.Kf1 Bc7 30.Qb3 Na5 31.Qc2 Nc6 32.Qb3

4. = (0.00): 21...Bc7 22.Qb3 Qd7 23.Ne5 Bxe5 24.dxe5 Rxe5 25.Red1 Rd8 26.Bf4 Re7 27.Be3 Re5

5. = (0.00): 21...Rc8 22.Qd3 Qc7 23.Qb3 Qd7 24.Nd2 Bc7 25.Qd3 Qd6 26.Kf1 Rb8 27.Bg5 Bb6 28.Rxe8+ Rxe8 29.Be3 Bc7 30.Qb3 Na5 31.Qc2 Nc6 32.Qb3

6. = (0.00): 21...Qc7 22.Qb3 Qd7 23.Nd2 Rad8 24.Qd3 Bc7 25.Bg5 Qd6 26.Rxe8+ Rxe8 27.Be3 Bb6 28.Kf1 Bc7 29.Qb3 Na5 30.Qc2 Nc6 31.Qb3

7. = (0.00): 21...Rb8 22.Qd3 Bc7 23.Bg5 Qd6 24.Rxe8+ Rxe8 25.Qb3 Na5 26.Qc3 Nc6 27.Qb3

8. = (0.00): 21...Qe7 22.Bf4 Qc7 23.Be5 Rad8 24.Qb3 Qb8 25.Bxd6 Rxe1+ 26.Rxe1 Rxd6 27.Qd1 Qa7 28.Qc1 Rd8 29.Qf4 a5 30.Qc7 Rf8 31.Ng5 Qa6 32.Bf1 Qa7 33.Bg2

9. = (0.00): 21...Qc8 22.Ne5 Bxe5 23.dxe5 Rxe5 24.Qb3 Qd7 25.Red1 Rd8 26.Bf4 Re7 27.Be3 Re5

One very interesting fork in the road occurred at black's 29th move. Stockfish was torn between 29...Qh2 (which was played) and 29...Nhf4. I was happy to make the move when Qh2 was on top because it looked like an aggressive move. Events proved, however, that this move's bark was worse than its bite. As part of the post-mortem, I decided to slide forward on 29...Nhf4. This leads to a completely different line with completely different but fascinating tactics, but with the same result.

<29...Nhf4 30. Nxe6 Nxe6 31. Re5 Qd7 32. Rxd5 Qa4 33. Qf3 Bh2 (the bishop goes to h2 in this line!) 34. Qd3 Nf4 35. Bxf4 Bxf4 36. Rb1 Re2 37. Bf3 Ra2 38. Qb3 Qxa3 39. Ra5 Qxb3 40. Rxb3 Rd2 41. Bxb7 Rb8 42. Bc6 Rd1+ 43. Kg2 Rxd4 44. Rxa6 Rbxb4 45. Rxb4 Rxb4>

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (32-ply):

1. = (0.00): 46.Bd5 Rb5 47.Ra8+ Bb8 48.Kf3 Rxd5 49.Rxb8+ Kh7 50.Rb7 Rf5+ 51.Ke3 Re5+ 52.Kd4 Rf5 53.Ke3

2. = (0.00): 46.Ra8+ Kh7 47.Ra7 Bxg5 48.Rxf7 Bh4 49.Kf3 Rb2 50.Kg2 Kg8 51.Bd5 Kh7 52.Rf3 Kh6 53.Rf4 g5 54.Rf3 Rd2 55.Be4 Rd4 56.Bf5 Rd2 57.Be4

3. = (0.00): 46.Ra5 Kf8 47.Kf3 Bd6 48.Ra8+ Bb8 49.Bd5 Ke7 50.h4 f6 51.Be4 Be5 52.Ra7+ Ke6 53.Ra6+ Kf7 54.Ra7+ Ke6

4. = (0.00): 46.h4 Bb8 47.Kh3 Rb2 48.Kg2 Kf8 49.Ra4 Ke7 50.Bd5 f6 51.gxf6+ gxf6 52.Rg4 g5 53.hxg5 fxg5 54.Rxg5 Kd6 55.Bc4 Ba7 56.Rg6+ Ke5 57.Rg7 Bxf2 58.Kf3 Kd6 59.Ke4 Bc5 60.Kf3 Bb6 61.Ke4 Bc5

<Stockfish agrees that 16. d5 would have been stronger. However, Stockfish contends that 16. Bf4 would have been strongest of all.

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (30-ply):

1. = (0.24): 16.Bf4 cxd4 17.cxd4 Qb6 18.Rc1 d5 19.e5 Na5 20.Qa4 Qd8 21.Qc2 Nc6 22.Qd1 Nb6 23.Bd2 Nd7 24.Qb3 Qb8 25.Bf4 Qa7 26.Red1 Rac8 27.a3 Bd8 28.Qe3 Bc7 29.Bd3

2. = (0.20): 16.d5 exd5 17.Qxd5 Qc7 18.Bc4 Nde5 19.Nxe5 Nxe5 20.Rd1 g5 21.Be3 Bf6 22.Be2 Re6 23.b4 cxb4 24.cxb4 Ng6 25.Rac1 Qe7 26.Qd2 Rxe4 27.Bd3 Re6 28.Bxg6 fxg6 29.Rc2 Re8 30.a3 Qf7 31.Rdc1>

Just for curiousity, I ran "my" Stockfish and this is the result from <depth 6 to 34>:

Analysis by Stockfish 2.0.1 JA <32 bit> with <512 MB Hash Tables>

<16.Bd3> Qb6 17.Qb3 Qxb3 18.axb3 cxd4 19.Nxd4 Nxd4 20.cxd4 <= (0.08) Depth: 6/9 00:00:00 8kN>

<16.Bf4> cxd4 17.cxd4 Qb6 18.d5 exd5 19.exd5 Nce5 20.Rb1 Nxf3+ 21.Qxf3 Bh4 22.Be3 Qb4 <= (0.16) Depth: 8/14 00:00:00 29kN>

16.Bf4 Qb6 17.dxc5 Qxc5 18.Be3 Qa5 19.Qb3 Qc7 20.Rad1 Nde5 21.Bg2 Nxf3+ 22.Bxf3 Ne5 23.Be2 b5 24.a4 Nc4 25.axb5 axb5 = (0.16) Depth: 14/20 00:00:03 907kN

<16.d5> exd5 17.Qxd5 Qb6 18.Qb3 Qc7 19.Bf4 Nde5 20.Be2 Nxf3+ 21.Bxf3 g5 22.Be3 Ne5 23.Be2 Qc6 24.Qd5 Qxd5 25.exd5 <= (0.12) Depth: 15/19 00:00:06 1997kN>

<16.Be3> cxd4 17.cxd4 d5 18.exd5 exd5 19.a3 Nf6 20.Ne5 Qb6 21.b4 Bd6 22.Rc1 Bxe5 23.dxe5 d4 24.Rxc6 Qxc6 25.Bg2 Nd5 26.Qxd4 Rad8 27.Rd1 <= (-0.08) Depth: 18/23 00:00:15 8234kN>

<16.d5> exd5 17.Qxd5 Qb6 18.Rd1 Nf6 19.Qc4 Qc7 20.Bg2 Nd7 21.Qb3 Nde5 22.Qd5 Qb6 23.Nxe5 dxe5 24.Be3 Rad8 25.Qb3 Qc7 26.h4 b6 <= (0.16) Depth: 19/22 00:00:20 11604kN>

<16.d5> exd5 17.Qxd5 Qc7 18.a4 Nde5 19.Be2 Nxf3+ 20.Bxf3 Ne5 21.Bg2 g5 22.Be3 Bf6 23.Red1 Rad8 24.Ra2 Ng6 25.Ra3 Ne5 26.a5 Nc6 27.Rda1 Rc8 28.Ra4 Re5 29.Qd2 Qd8 30.b4 Re6 31.bxc5 dxc5 < = (0.24) Depth: 30/32 00:32:32 1351mN>

16.d5 exd5 17.exd5 Nce5 18.Nxe5 Nxe5 19.f4 Nd7 20.g5 Bf8 21.c4 Rb8 22.a4 Nb6 23.Bd2 Nc8 24.a5 Rxe1 25.Qxe1 Qd7 26.Qf2 Ne7 27.b4 cxb4 28.Bxb4 Nf5 29.Bd3 Re8 30.Re1 Rxe1+ 31.Qxe1 Qa4 32.Kg2 Nd4 33.Qc3 Nf5 (0.64) Depth: 32/36 01:29:19 3792mN
16.d5 exd5 17.exd5 Nce5 18.Nxe5 Nxe5 19.f4 Nd7 20.g5 Bf8 21.c4 Rb8 22.a4 Nb6 23.Be3 Nc8 24.Qd2 Ne7 25.Kg2 Nf5 26.Bf2 Rxe1 27.Rxe1 Qd7 28.b3 Re8 29.Rxe8 Qxe8 30.Bd3 Be7 31.Qe1 Qd7 32.a5 Bd8 33.b4 cxb4 34.Qxb4 Qe7 35.Kh2 Qd7 (0.64) Depth: 33/40 01:49:27 4687mN

<16.d5> exd5 17.exd5 Nce5 18.Nxe5 Nxe5 19.f4 Nd7 20.g5 Bf8 21.c4 Rb8 22.a4 Nb6 23.Be3 Nc8 24.Qd2 Ne7 25.Kg2 Nf5 26.Bf2 Rxe1 27.Rxe1 Qd7 28.b3 Re8 29.Rxe8 Qxe8 30.Bd3 Be7 31.Qe1 Qd7 < (0.64) Depth: 34/32 02:01:42 5192mN>

"My" Stockfish found <16.Bf4> at depths 8 to 14 and then never appeared again. So at depth <30/32> gives <= (0.24)> for <16.d5>, the same eval of <kutztown46>´s Stockfish for <16.Bf4>.

Weird engines !

I remember <AylerKupp> gave some explanations about different results between "his" Rybka and the Rybka of another member of the World Team vs NPogonina.

<kutztown46> just for fun, could you run "your" Stockfish up to depth 34 just to see how compares with "mine"?

<I reviewed the entire game with Stockfish, and white's 16th move seems to represent the single best place where either side could have greatly improved upon what was played> <In conclusion, it was a fun game to play. Both sides played very well, and the biggest missed opportunity seemed to be white's 16th move.>

After 15. d4:

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (26-ply):

1. = (0.20): 15...Re8 16.d5 exd5 17.Qxd5 Qb6 18.Qb3 Qc7 19.Rd1 b5 20.Be3 c4 21.Qc2 Nce5 22.Nd4 Qb7 23.b3 Nc5 24.Bg2 Rad8 25.Rab1 Qc7 26.b4 Ncd3 27.a4 Rb8

2. = (0.24): 15...cxd4 16.cxd4 Re8 17.a3 Rc8 18.Rb1 Na5 19.Bf4 Nc4 20.Bd3 Nf6 21.Qb3 b5 22.a4 d5 23.e5 Ne4 24.axb5 axb5 25.Qxb5 g5 26.Be3 Rb8 27.Qa6 Rb6

3. ² (0.36): 15...Qc7 16.d5 exd5 17.exd5 Nce5 18.Nxe5 dxe5 19.a4 Bd6 20.g5 f5 21.gxf6 Nxf6 22.Bg2 Qd7 23.c4 e4 24.Re2 Rae8 25.Ra3 Bc7 26.Kh1 Bd6 27.Kg1

4. ² (0.68): 15...d5 16.exd5 exd5 17.Bg2 Nf6 18.Bg5 cxd4 19.Nxd4 Qc7 20.Qd3 Rfd8 21.Rad1 Rd7 22.Re2 Bd6 23.Rde1 Bh2+ 24.Kh1 Ne5 25.Qc2 Nc6 26.Nb3 Rdd8 27.Qd2 Rac8 28.Rd1 Bd6 29.f4 Ne7 30.Qd4 Re8

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (27-ply):

1. = (0.16): 16...Re8 17.a3 d5 18.exd5 exd5 19.Qb3 Nf6 20.Bf4 Na5 21.Qa2 Qb6 22.Rad1 Rac8 23.b4 Nc6 24.Ne5 Bd6 25.Qb2 Re6 26.Nxc6 Rxe1 27.Rxe1 Rxc6 28.Bxd6 Rxd6 29.Bd3 Re6 30.Rxe6 Qxe6 31.Qc1 b5 32.Qc7

2. = (0.20): 16...Rc8 17.a3 Re8 18.Be3 Nf6 19.d5 exd5 20.exd5 Ne5 21.Bd4 Nxf3+ 22.Qxf3 Nh7 23.Qb3 Qd7 24.Bg2 Bg5 25.Rxe8+ Rxe8 26.Qb4 Re2 27.Bf3 Re8 28.Bc3 Bf6 29.Re1 Bxc3 30.Rxe8+ Qxe8 31.Qxc3 Qd8 32.Kg2

3. ² (0.36): 16...Qb6 17.Rb1 d5 18.exd5 exd5 19.b4 Qc7 20.b5 axb5 21.Rxb5 Bb4 22.Re2 Ra5 23.Qd3 Nf6 24.Rc2 Bd6 25.Qb3 Rxb5 26.Bxb5 Qd7 27.Ne5 Qe7 28.Nxc6 bxc6 29.Rxc6 Qe1+ 30.Bf1

4. ² (0.40): 16...d5 17.exd5 exd5 18.Qb3 Bb4 19.Re2 Nf6 20.Ne5 Bd6 21.Nxc6 bxc6 22.Qb7 c5 23.dxc5 Bxc5 24.Bf4 Bd6 25.Bxd6 Qxd6 26.Rc1 Rab8 27.Qc7 Qb4 28.b3 a5 29.Re7 Rbe8 30.Rxe8 Rxe8 31.Rc5 Qd4 32.Rxa5

Then after forcing 16...d5 17. exd5 exd5:

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (28-ply):

1. ² (0.40): 18.Qb3 Bb4 19.Re2 Nf6 20.Ne5 Bd6 21.Nxc6 bxc6 22.Qb7 c5 23.dxc5 Bxc5 24.Bf4 Bd6 25.Bxd6 Qxd6 26.Rc1 Rab8 27.Qc6 Qf4 28.Qc7 Qd4 29.Qc5 Qxc5 30.Rxc5 Rfc8 31.Rxc8+ Rxc8 32.Rd2 g5 33.Bxa6

I could spare only 4 hours, which was enough time for Stockfish to complete its 33rd ply. Here is what we have - interesting:

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (33-ply):

1. ² (0.28): 16.d5 exd5 17.Qxd5 Qb6 18.Kg2 Rad8 19.Qb3 Qc7 20.Bc4 Rf8 21.Bd5 Nb6 22.c4 Bf6 23.Be3 Rfe8 24.Rad1 Ra8 25.g5 Be7 26.Bd2 a5 27.Qc3 Nb4 28.a3

2. ² (0.28): 16.Be3 cxd4 17.cxd4 d5 18.exd5 exd5 19.Qb3 Nf6 20.Ne5 Bd6 21.f4 Na5 22.Qd3 Rc8 23.Bg2 Nc4 24.Bf2 Qa5 25.Re2 Re7 26.a3 Rec7 27.Rae1 Qb5 28.Rc2 Qa5 29.Ree2 Qb5 30.Rc3 Qb6 31.Bxd5 Nxd5

3. = (0.24): 16.Bf4 cxd4 17.cxd4 Rc8 18.Rc1 e5 19.dxe5 Ndxe5 20.Nxe5 dxe5 21.Qxd8 Rexd8 22.Be3 Nd4 23.Red1 b5 24.Kg2 Kf8 25.a4 bxa4 26.Bxa6 Rb8 27.Rc4 Rxb2 28.Rxa4 Ra8 29.Rc1 Bd8 30.Rc5 f6 31.Rc8 Rxc8

1. ² (0.28): <<16.d5 exd5 17.Qxd5>> Qb6 18.Kg2 Rad8 19.Qb3 Qc7 20.Bc4 Rf8 21.Bd5 Nb6 22.c4 Bf6 23.Be3 Rfe8 24.Rad1 Ra8 25.g5 Be7 26.Bd2 a5 27.Qc3 Nb4 28.a3>

"My" Stockfish considered 17.Qxd5 as the best reply for 16..exd5 up to depth 30/32, the same for "your" Stockfish up to depth 33.

The eval jumped from <(0.24)> to < (0.64)> at depth: 32/36 when <17.exd5> was find.

The game changes significantly. The game opens and whites´s bishop pair and pawns have space to maneuver:

click for larger view

After <16.d5 exd5 17.exd5> Nce5 18.Nxe5:

click for larger view

Analysis by Houdini 1.5a w32:

<1. (0.28)>: 18...dxe5 19.Bg2 Bd6 20.Qc2 c4 21.b3 Rc8 22.Be3 Qh4 23.Qd2 cxb3 24.axb3 e4 25.c4 f6 26.Bd4 Nc5 27.Rab1 g5 28.b4

2. (0.41): 18...Nxe5 19.f4 Nd7 20.g5 Bf8 21.Be3 Rb8 22.a4 Nb6 23.Qb3 Qc7 24.Bf2 Rxe1 25.Rxe1 Nc8 26.Bd3 Ne7 27.c4 Qa5 28.Re2 Nf5 29.Kg2 Nd4 30.Bxd4

Houdini gives only a slight advantage for white at depth 24.

<kutztown46> Thank you very much for your feedback, I really appreciated.

My last question about this game is:

Who is evaluating properly the position? Stockfish or Houdini?