Delphil 2.9e beta 7 - Pro Deo 1.8 : 1 - 0

[gleperlier]

Hello,

Just wanted to share with you a nice long game (40 moves in 120 minutes + 60 minutes rest) between Delphil 2.9e beta 7 and Pro Deo 1.8.

Delphil 2.9e beta 7 was promoted earlier today to official Delphil 2.9e.

So the game was really nice and I guess it can be usefull for Ed.

So the game (Sicilian Scheveningen : Keres attack) started like this avec 7...Be7, the two engines are out of book in this position :

[D]rnbqk2r/pp2bpp1/3ppn1p/8/3NP1PP/2N5/PPP2P2/R1BQKB1R w KQkq - 0 8

In this positon Delphil plays 8.Qf3 which is the second move in Playchess livebook.

[D]rnbqk2r/pp2bpp1/3ppn1p/8/3NP1PP/2N2Q2/PPP2P2/R1B1KB1R b KQkq - 0 8

Now Pro Deo goes out the known lines and go for 8...g6, only played 1 time by Kasjanov in URS-chT in 1988. Kasjanov lost the game versus Mihail Rodin.
Delphil played exactly like Mihail Rodin : 9.g5! hxg5 10.hxg5*

[D]rnbqk2r/pp2bp2/3ppnp1/6P1/3NP3/2N2Q2/PPP2P2/R1B1KB1R b KQkq - 0 10

Now where Kasjanov choosed 10...Rxh1, ProDeo choosed 10...Nh5

[D]rnbqk2r/pp2bp2/3pp1p1/6Pn/3NP3/2N2Q2/PPP2P2/R1B1KB1R w KQkq - 0 11

After 11.Be3 a6 12.Qg2 b5* we arrive to a key position

[D]rnbqk2r/4bp2/p2pp1p1/1p4Pn/3NP3/2N1B3/PPP2PQ1/R3KB1R w KQkq - 0 13

In this position Delphil played the very good looking move for a human (Philippe Fabiani likes it very much) : 13.Nf5!?

[D]rnbqk2r/4bp2/p2pp1p1/1p3NPn/4P3/2N1B3/PPP2PQ1/R3KB1R b KQkq - 0 13

Then 13...Nc6 14.Ng7+ Kd7 15.Nxh5 gxh5*

Another nice position :

[D]r1bq3r/3kbp2/p1npp3/1p4Pp/4P3/2N1B3/PPP2PQ1/R3KB1R w KQ - 0 16

Again Delphil is in this position playing a very nice move that his author Philippe likes very much again : 16.e5!?

[D]r1bq3r/3kbp2/p1npp3/1p2P1Pp/8/2N1B3/PPP2PQ1/R3KB1R b KQ - 0 16

As Philippe said, thoses risky moves are perhaps lost versus 3000+ engines but they are so nice versus humans and middle class engines.

Many moves later, avec 27.Kb1 we are now in this position :

[D]1kr5/1b2bp2/p3p2R/1p2P1P1/3n1P2/P1N1Q3/1Pr5/1K3B2 b - - 0 27

ProDeo choose the nice 27...Bxa3.

[D]1kr5/1b3p2/p3p2R/1p2P1P1/3n1P2/b1N1Q3/1Pr5/1K3B2 w - - 0 28

Now position is very complexe for a human (like me) but chances are fairly equal. Vitruvius gives it a 0.00 after 1 minute pondering.

We were then going to a draw until this position :

[D]8/kb6/p3B1p1/1p2P3/8/3Q4/N1r2r2/K7 b - - 0 38

ProDeo made a little mistake and played 38...Rcd2?

[D]8/kb6/p3B1p1/1p2P3/8/3Q4/N2r1r2/K7 w - - 0 39

Then Delphil jumped and played 39.Qe3+! and took a first little advantage in this endgame.

Now it's only a matter of technics for Delphil.

[D]k1r5/1b5r/p3P3/1p2N1p1/8/KB4Q1/8/8 w - - 0 51

51.Nd7! and game is already won !

[D]k6r/1bQN4/p3P3/1p1B2p1/r7/8/1K6/8 b - - 0 54

ProDeo resigned.

What a nice game isn't it ? And I am sure Ed will be able to find so improvements for Pro Deo.

You can find the PGN bellow.

Cheers,

Gab

PGN

[Event "40 coups en 120 min; reste en 60 min"]
[Site "Engine Match"]
[Date "2012.07.02"]
[Round "1"]
[White "Delphil 2.9e Beta 7"]
[Black "ProDeo 1.8"]
[Result "1-0"]

1. e4 {Biblio. 0s} c5 {Biblio. 0s} 2. Nf3 {Biblio. 0s} e6
{Biblio. 0s} 3. d4 {Biblio. 0s} cxd4 {Biblio. 0s} 4. Nxd4
{Biblio. 0s} Nf6 {Biblio. 0s} 5. Nc3 {Biblio. 0s} d6
{Biblio. 0s} 6. g4 {Biblio. 0s} h6 {Biblio. 0s} 7. h4
{Biblio. 0s} Be7 {Biblio. 0s} 8. Qf3 {+0.49/15 10:55m} g6
{0.00/1 0s (e5)} 9. g5 {+0.93/13 3:24m} hxg5 {0.00/1 0s}
10. hxg5 {+0.79/14 3:24m} Nh5 {+0.16/17 9:15m (Txh1)}
11. Be3 {+0.89/12 3:24m (Dg2)} a6 {+0.12/17 5:33m (Fd7)}
12. Qg2 {+0.95/14 3:24m} b5 {+0.41/17 14:59m (e5)} 13. Nf5
{+1.10/13 10:13m} Nc6 {+0.56/17 25s (exf5)} 14. Ng7+
{+1.42/13 3:09m} Kd7 {+0.83/17 0s} 15. Nxh5 {+1.31/13
3:09m} gxh5 {+0.41/15 29s} 16. e5 {+1.31/13 2:39m (O-O-O)}
d5 {+0.49/17 5:12m (Fb7)} 17. O-O-O {+1.35/13 6:21m} Kc7
{+1.05/16 0s} 18. f4 {+1.16/12 3:02m} Bb7 {+1.03/16 0s}
19. Rh3 {+1.22/11 3:02m} Rc8 {+1.04/17 2:34m (d4)} 20. Qh2
{+1.19/13 3:02m} Kb8 {+0.64/17 10:04m} 21. Rxh5 {+0.76/15
0s (Df2)} d4 {+0.46/17 4:11m} 22. Rh6 {+0.46/14 2:12m} dxe3
{+0.08/17 4:16m (Txh6)} 23. Rxd8 {+0.69/16 9:45m} Rhxd8
{+0.44/19 0s} 24. a3 {+0.64/15 2:52m} Rd2 {0.00/18 2:11m}
25. Qh3 {+0.88/16 6:24m (Dg3)} Nd4 {+1.00/18 7:23m (b4)}
26. Qxe3 {+0.79/15 7:54m} Rxc2+ {+0.70/20 0s} 27. Kb1
{+0.79/14 2:15m} Bxa3 {+0.78/19 2:55m (Fc5)} 28. Qxd4
{+1.20/16 6:46m} Rxb2+ {+0.20/20 0s} 29. Ka1 {+1.20/15
1:52m} Rc2 {+0.17/19 4:51m (b4)} 30. Na2 {+1.30/15 5:38m}
Bc5 {+0.24/19 0s (Fd5)} 31. Qd1 {+1.82/15 4:30m} Bb6
{+0.19/18 0s (Tf2)} 32. g6 {+1.48/14 3:30m} fxg6 {0.00/19
0s} 33. Bh3 {+0.48/14 2:37m (Dd6+)} R8c4 {-0.44/19 5:43m}
34. Rh8+ {+0.71/16 0s (Rb1)} Ka7 {-0.36/19 4:59m} 35. Rd8
{+0.62/17 0s} Bxd8 {-0.31/17 1:31m} 36. Qxd8 {+0.62/16 0s}
Rxf4 {0.00/19 5:13m (Tf2)} 37. Qd3 {+0.91/14 1:05m} Rff2
{0.00/19 4:43m (Tcf2)} 38. Bxe6 {+1.07/14 1:05m} Rcd2
{0.00/19 7:42m (Tfe2)} 39. Qe3+ {+1.22/16 1:05m} Ka8
{+0.69/19 3:12m} 40. Nc1 {+1.40/18 0s} Rc2 {+0.93/20 7:42m}
41. Qe1 {+1.63/18 0s (Fb3)} Rh2 {-0.45/12 1s} 42. Bb3
{+1.67/17 1:30m} Rc8 {+0.27/17 0s} 43. Ne2 {+1.63/16 1:28m}
Rh1 {+0.47/16 0s} 44. Ng1 {+1.99/15 1:12m} Rh5 {+0.71/17 0s
(g5)} 45. e6 {+2.12/13 10s} Rh7 {+0.54/14 0s} 46. Kb2
{+2.28/14 1:24m} Rh2+ {+0.71/10 1s (g5)} 47. Ka3 {+2.21/10
2s (Rb1)} Re8 {+0.58/10 1s} 48. Nf3 {+2.85/13 1:21m (De5)}
Rh7 {+0.88/10 1s} 49. Ne5 {+3.19/13 1:19m} g5 {+1.94/16 0s}
50. Qg3 {+3.48/13 1:17m (De3)} Rc8 {+1.43/10 1s (Teh8)}
51. Nd7 {+5.66/12 11s (Dxg5)} Rch8 {+1.77/10 1s (Te8)}
52. Qc7 {+6.79/10 3s (Dxg5)} Rh4 {+6.03/10 1s (Ra7)}
53. Bd5 {+8.73/9 0s (e7)} Ra4+ {+9.53/10 1s} 54. Kb2
{+17.20/10 2s} 1-0

[carldaman]

Very nice game, Gabriel, especially the Fischeresque 13.Nf5!? , just beautiful.
Nice exposition of all the highlights

CL

[lucasart]

Very nice game, Gabriel, especially the Fischeresque 13.Nf5!? , just beautiful.
Nice exposition of all the highlights

CL

Very nice indeed!
Nf5 may look like a Tal sacrifice, but actually, it's a very precise move (not just beatufil because it is bold). IvanHoe 999946h gives black a score of -0.99 after Nf5! Although IvanHoe would have played the more solid and patient a3 instead of Nf5 (my engine DiscoCheck would also play a3).

[gleperlier]

Thanks Lucas & Carl

[Graham Banks]

Very nice game, Gabriel, especially the Fischeresque 13.Nf5!? , just beautiful.
Nice exposition of all the highlights

CL

Yes - good to see this posted.
Many people don't seem to realise just how much enjoyment you can get from the lesser engines.

[carldaman]

I ran Houdini 2.0c on my i3 3.3ghz quad with 1 GB hash enabled and it liked 13.a3, then 13.O-O-O after a few minutes. However, when I doubled its hash to 2 GB, this was the result:

299: Delphil 2.9e Beta 7 - ProDeo 1.8, 40 coups en 120 min; reste en 60 min 2012
rnbqk2r/4bp2/p2pp1p1/1p4Pn/3NP3/2N1B3/PPP2PQ1/R3KB1R w KQkq b6 0 1

Analysis by Houdini 2.0c x64:

13.a3 e5 14.Nf3 Nc6 15.Nd5 Kf8
+/= (0.42) Depth: 6/18 00:00:00 11kN
13.a3
+/= (0.48 ++) Depth: 7/18 00:00:00 13kN
13.a3 e5 14.Nf3 Nc6 15.Nd5 Kf8 16.0-0-0
+/= (0.57) Depth: 7/18 00:00:00 15kN
13.a3
+/= (0.64 ++) Depth: 8/20 00:00:00 25kN
13.a3 e5 14.Nf3 Be6 15.0-0-0 Nc6 16.Nd5 Bxd5 17.Rxd5
+/= (0.59) Depth: 8/20 00:00:00 28kN
13.a3
+/= (0.67 ++) Depth: 9/23 00:00:00 34kN
13.a3 e5 14.Nf3 0-0 15.Nd5 Nd7 16.Be2 Bb7 17.0-0-0 Bxd5 18.Rxd5
+/= (0.67) Depth: 9/23 00:00:00 37kN
13.a3 e5 14.Nf3 Be6 15.0-0-0 Nd7 16.Kb1 Rc8 17.Nd5 Nc5 18.Nd2 Bxd5 19.exd5
+/= (0.68) Depth: 10/23 00:00:00 62kN
13.a3 e5 14.Nf3 Nc6 15.Nd5 Kf8 16.0-0-0 Be6 17.Kb1 Rc8 18.Bb6 Qd7 19.Be2 Bxd5 20.Rxd5
+/= (0.62) Depth: 11/23 00:00:00 111kN
13.a3 e5 14.Nf3 Be6 15.0-0-0 Nd7 16.Kb1 Rc8 17.Nd5 Kf8 18.Be2 Bxd5 19.Rxd5 Qc7 20.c3 Nb6
+/= (0.62) Depth: 12/34 00:00:01 280kN
13.a3 Bb7 14.0-0-0 Qc7 15.Kb1 Nd7 16.Nb3 Rd8 17.Bd4 e5 18.Be3 Kf8 19.Be2 Rc8 20.Qh3 Kg8
+/= (0.62) Depth: 13/35 00:00:01 759kN
13.a3
+/- (0.72 ++) Depth: 14/35 00:00:01 1554kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Qc7 16.Nb3 Rd8 17.Bd4 e5 18.Be3 Kf8 19.Be2 Rc8 20.Qh3 Kg8
+/= (0.62) Depth: 14/35 00:00:01 1780kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Qc7 16.Nb3 Rd8 17.Bd4 e5 18.Be3 Kf8 19.Be2 Rc8 20.Qh3 Kg8
+/= (0.62) Depth: 15/35 00:00:01 3073kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.f4 Nc4 17.Bxc4 bxc4 18.Nf3 Qd7 19.Qe2 Rc8 20.Bd4 0-0 21.Be5 a5
+/= (0.70) Depth: 16/42 00:00:02 7426kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.f4 Nc4 17.Bxc4 bxc4 18.Nf3 Qd7 19.Qe2 Rc8 20.Ne5 Qc7 21.Ng4 Kf8 22.Bd4 Rh7
+/- (0.72) Depth: 17/42 00:00:03 11421kN
13.a3
+/- (0.82 ++) Depth: 18/42 00:00:07 23428kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.f4 Nc4 17.Bxc4 bxc4 18.Nde2 Qc7 19.Ng3 Rb8 20.Qe2 Nxg3 21.Rxh8+ Kd7 22.Qh2 Nxe4 23.Ne2 Bd5 24.Qh3 Rxh8 25.Qxh8
+/- (0.75) Depth: 18/54 00:00:09 32477kN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.f4 Nc4 17.Bxc4 bxc4 18.Nde2 Qc7 19.Ng3 Rb8 20.Qe2 Nxg3 21.Rxh8+ Kd7 22.Qh2 Nxe4 23.Ne2 Bd5 24.Qh3 Rxh8 25.Qxh8
+/- (0.75) Depth: 19/54 00:00:15 56119kN
13.Nf5
+/- (0.84 ++) Depth: 19/54 00:00:33 138mN
13.Nf5 gxf5
+/= (0.65 --) Depth: 19/54 00:00:41 175mN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.f4 Nc4 17.Bxc4 bxc4 18.Nde2 Qc7 19.Ng3 Rb8 20.Qe2 Nxg3 21.Rxh8+ Kd7 22.Qh2 Nxe4 23.Ne2 Bd5 24.Qh3 Rxh8 25.Qxh8
+/- (0.75) Depth: 19/54 00:00:41 175mN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.Be2 Bf8 17.Bxh5 Rxh5 18.Rxh5 gxh5 19.Qh3 Qd7 20.Qxh5 Nc4 21.Qf3 Rc8 22.Nde2 Qe7 23.Qf4 Nxe3 24.Qxe3 Bg7 25.f4
+/- (0.75) Depth: 20/54 00:00:46 197mN
13.a3 Bb7 14.0-0-0 Nd7 15.Kb1 Ne5 16.Be2 Bf8 17.Bxh5 Rxh5 18.Rxh5 gxh5 19.Qh3 Qd7 20.Qxh5 Nc4 21.Qf3 Rc8 22.Nde2 Qe7 23.Qf4 Nxe3 24.Qxe3 Bg7 25.f4
+/- (0.75) Depth: 21/54 00:01:08 296mN
13.Nf5
+/- (0.84 ++) Depth: 21/54 00:01:27 379mN
13.Nf5 b4 14.Ng7+ Kf8 15.Nxh5 bxc3 16.Nf6 Bxf6 17.gxf6 Rxh1 18.Qxh1 Kg8 19.bxc3 Bb7 20.0-0-0 Qxf6 21.Rxd6 Nc6 22.e5 Qxe5 23.Rxc6 a5 24.Bd3 Bxc6 25.Qxc6 Rd8 26.a3 Rb8 27.Be2 a4 28.Bd3 Kf8 29.Kd2 Rb2 30.Qa8+ Kg7 31.Qxa4
+/- (0.87) Depth: 21/54 00:01:52 493mN
13.Nf5
+/- (0.96 ++) Depth: 22/55 00:02:23 632mN
13.Nf5 Bb7 14.Ng7+ Kf8 15.Nxh5 gxh5 16.0-0-0 Nd7 17.a3 Rc8 18.f4 Qa5 19.Kb1 Rxc3 20.Bd4 Rxa3 21.bxa3 Rg8 22.Rxh5 Qxa3 23.Rh3 Qa4 24.Qh2 Ke8 25.Qe2 Nc5 26.Bxc5 dxc5 27.Bg2 Qb4+ 28.Rb3
+/- (0.96) Depth: 22/56 00:03:37 961mN
13.Nf5
+/- (1.06 ++) Depth: 23/59 00:06:06 1638mN

(Doe, 14.07.2012)