Insane move

[michiguel]

This is Exchess - Gaviota played today in the ACCA - WRCCC

[d]rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 8

In this position, Gaviota played h5?!?!?
Generally, black tries to make room for his bishop to retreat to h7 if white tries Nh4. So, a natural move will be h6... but h5?

My question is, for your favorite top engine, how far down is h5 in the list of choices?

It was hard to believe, but Gaviota got away with this craziness and won the game. In fact and ironically, the h-pawn ended up helping in the attack. The game is below if you want to see it.

Miguel

[pgn]
[Event "ICS unrated standard match"]
[Site "207.99.83.228"]
[Date "2013.07.20"]
[Round "-"]
[White "EXchess"]
[Black "Gaviota"]
[Result "0-1"]
[WhiteElo "2365"]
[BlackElo "2392"]
[TimeControl "1500+4"]
[Annotator "2... -0.09"]

1. d4 d5 2. Nf3 Nf6 {-0.09/12 0.4} 3. g3 c6 {-0.15/13} 4. c4 Bf5
{-0.26/11 0.4} 5. Nbd2 e6 {-0.17/10 0.2} 6. Bg2 Be7 {+0.08/10} 7. O-O O-O
{-0.17/12 0.1} 8. b3 h5 {+0.19/20 1:58} 9. Ng5 Nbd7 {+0.16/19 19} 10. e4
dxe4 {+0.06/20 38} 11. Ndxe4 Nxe4 {+0.18/21 46} 12. Nxe4 Nf6 {+0.20/20 44}
13. Nxf6+ Bxf6 {+0.21/20 46} 14. Be3 h4 {+0.19/20 43} 15. g4 Bh7
{+0.11/20 2.4} 16. Qd2 Qd7 {+0.09/19 43} 17. g5 Be7 {+0.00/19 43} 18. Qe2
Qd8 {-0.09/20 1:44} 19. f4 Bf5 {+0.35/19 43} 20. d5 exd5 {+0.49/20 49} 21.
cxd5 h3 {+0.44/20 9} 22. Bf3 c5 {+0.40/20 26} 23. Qb5 Bd6 {+0.72/20 2:09}
24. Rae1 Qc7 {+0.92/22 1:29} 25. Qc4 Qa5 {+0.92/20 28} 26. a4 a6 {+0.95/23}
27. Rc1 Rac8 {+1.19/20 31} 28. Qc3 Qc7 {+1.12/19 27} 29. Qd2 Rfe8
{+1.43/19 27} 30. Be2 Qe7 {+1.51/18 26} 31. Kf2 Bc7 {+1.94/19 25} 32. Bc4
Ba5 {+1.78/18 44} 33. Qe2 Qd7 {+1.87/19 22} 34. Qf3 Bb4 {+2.17/17 23} 35.
d6 b5 {+3.01/18 22} 36. Bd5 c4 {+3.48/17 22} 37. Kg3 cxb3 {+3.48/18 20} 38.
Rxc8 Rxc8 {+2.97/17 21} 39. axb5 Rc2 {+3.40/17 20} 40. Rg1 axb5
{+3.75/18 19} 41. Bxb3 Be1+ {+2.74/20} 42. Rxe1 Rg2+ {+3.19/20} 43. Qxg2
hxg2 {+3.53/18 21} 44. Bf2 Qxd6 {+4.08/19 19} 45. Kxg2 Qxf4 {+4.93/17 18}
46. Re8+ Kh7 {+4.93/18 2.3} 47. g6+ Kxg6 {+5.31/17 18} 48. Re3 b4
{+5.48/17} 49. Kg1 Be6 {+6.24/17 11} 50. Bc2+ Kh6 {+6.54/18 17} 51. Kg2
Qg4+ {+8.60/17 23} 52. Bg3 Bd5+ {+8.89/18} 53. Kf2 f5 {+9.10/18 16} 54. Ke1
b3 {+10.27/16 1.2} 55. Bxb3 Qb4+ {+11.48/16 16} 56. Kf2 Bxb3 {+12.26/16 14}
57. Be5 f4 {+12.26/14 15} 58. Rh3+ Kg6 {+17.98/16} 59. Rxb3 Qxb3
{+327.22/14 23} 60. Bxf4 Kf5 {+327.54/15 6} 61. Bc1 Ke4 {+327.56/13 1.3}
62. Bg5 Qf3+ {+327.58/11 1.0} 63. Kg1 Qe2 {+327.60/9} 64. Be3 Kxe3
{+327.62/7} 65. h3 Kf3 {+327.64/2} 66. h4 Qg2# {+327.66/2 0.1}
{EXchess checkmated} 0-1
[/pgn]

[kiroje]

Well not really insane, as long as d4 also is attacked white cannot take the pawn at h5. The move h5 have merits in going to h4 and disrupt/annoy whites king position. And its not trivial/easy to punish black for h5 primarily because there isnt anything wrong with it

This is of course with my myopic view of things.

[zullil]

This is Exchess - Gaviota played today in the ACCA - WRCCC

[d]rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 8

In this position, Gaviota played h5?!?!?
Generally, black tries to make room for his bishop to retreat to h7 if white tries Nh4. So, a natural move will be h6... but h5?

My question is, for your favorite top engine, how far down is h5 in the list of choices?

Stockfish (8 threads, 4GB hash, MultiPV=10) has it second-best at depth 25:

Code: Select all

info depth 25 seldepth 38 score cp 30 nodes 922654672 nps 6615576 time 139467 multipv 1 pv h7h6 f3e5 b8d7 c1b2 d7e5 d4e5 f6e4 d2e4 d5e4 d1d8 f8d8 b2c3 d8d7 f1d1 a8d8 d1d7 d8d7 b3b4 h6h5 a2a4 h5h4 g3h4 e7h4 a4a5 d7d8 a5a6 b7a6 a1a6

info depth 25 seldepth 38 score cp 28 nodes 922654672 nps 6615576 time 139467 multipv 2 pv h7h5 f3g5 b8d7 e2e4 d5e4 d2e4 f6e4 g5e4 d7f6 e4f6 e7f6 c1e3 h5h4 g3g4 f5g6 d1d2 d8d7 a1d1 a8d8 f1e1 f8e8 f2f3 e8f8 f3f4 f6e7

info depth 25 seldepth 38 score cp 20 nodes 922654672 nps 6615576 time 139467 multipv 3 pv f6e4 d2e4 f5e4 c1f4 b8d7 d1d2 f8e8 f1d1 a8c8 g2h3 a7a5 f3g5 e4g6 g5f3 h7h6 h3g2 b7b6 c4d5 c6d5 a1c1 e7b4 d2b2 d8e7 c1c8 e8c8 d1c1 c8c1 b2c1 d7f6 c1c8 e7f8

info depth 25 seldepth 38 score cp 16 nodes 922654672 nps 6615576 time 139467 multipv 4 pv b8d7 f3h4 f5g4 h2h3 g4h5 g3g4 f6e8 g4h5 e7h4 e2e4 e8c7 h5h6 g7g6 f2f4 f7f5 e4e5 d5c4 d2c4 h4e7 a2a4 c7d5 c1d2 d7b6 a4a5 b6c4 b3c4

info depth 25 seldepth 38 score cp 14 nodes 922654672 nps 6615576 time 139467 multipv 5 pv d8b6 c4c5 b6a5 a2a3 h7h6 b3b4 a5d8 c1b2 b8d7 f3e5 a7a5 e5d7 f6d7 f2f3 a5b4 a3b4 a8a1 b2a1 f5g6 e2e4 f7f5 a1c3 d8e8 c3b2 f5e4 f3e4 f8f1 d1f1

info depth 25 seldepth 38 score cp 6 nodes 922654672 nps 6615576 time 139467 multipv 6 pv e7b4 c1b2 h7h6 c4c5 b8d7 a2a3 b4a5 f3e5 a5c7 e5d3 f8e8 d2f3 f6e4 a1c1 e4g5 f3g5 h6g5 f2f3 d7f6 b2c3 g5g4 d3e5 g4f3 e2f3 f6d7 e5d7 d8d7 f3f4

info depth 25 seldepth 38 score cp 0 nodes 922654672 nps 6615576 time 139467 multipv 7 pv f5g6 f3e5 g6f5 e5f3

info depth 25 seldepth 38 score cp 0 nodes 922654672 nps 6615576 time 139467 multipv 8 pv f5g4 f3e5 g4f5 e5f3

info depth 25 seldepth 38 score cp 0 nodes 922654672 nps 6615576 time 139467 multipv 9 pv e7d6 f3h4 f5g4 c1b2 b8d7 h2h3 g4h5 f1e1 a8c8 a2a3 d6e7 a1c1 f6e8 h4f3 e8f6 f3h4

info depth 25 seldepth 38 score cp 0 nodes 922654672 nps 6615576 time 139467 multipv 10 pv f6e8 f3e5 e8f6 e5f3

info nodes 922654672 time 139467
bestmove h7h6 ponder f3e5

[carldaman]

This is Exchess - Gaviota played today in the ACCA - WRCCC

[d]rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 8

In this position, Gaviota played h5?!?!?
Generally, black tries to make room for his bishop to retreat to h7 if white tries Nh4. So, a natural move will be h6... but h5?

My question is, for your favorite top engine, how far down is h5 in the list of choices?

It was hard to believe, but Gaviota got away with this craziness and won the game. In fact and ironically, the h-pawn ended up helping in the attack. The game is below if you want to see it.

Miguel

The move isn't half-bad. It actually fits in with the character of the position. This may be a case of the engine teaching us something.

Regards,
CL

[gerold]

h7-h6. Houdini,Fruit,Critter,Rybka, Strelka. Seems most of the Fruit line sees same best move.

Nb8-d7. Dragon,Ktulu,Murka3. Other top engines from the pass seen different moves.

[zullil]

Stockfish has it third after a long MultiPV=3 search:

Code: Select all

info depth 38 seldepth 64 score cp 22 nodes 263573622895 nps 7488886 time 35195304 multipv 1 pv h7h6 f3e5 b8d7 c1b2 d8b6 e5d7 f6d7 a1c1 d7f6 d2f3 b6a5 a2a4 e7b4 f3e5 b4d2 c1a1 f8d8 e2e3 f6d7 e5d7 d8d7 d1e2 d2b4 a1c1 a8e8

info depth 38 seldepth 64 score cp 16 nodes 263573622895 nps 7488886 time 35195304 multipv 2 pv d8b6 c4c5 b6d8 f3h4 f5g4 c1b2 b8d7 a2a3 h7h6 b3b4 a7a5 f1e1 g4h5 d1c2 d8c7 g2f3 h5f3 h4f3 a5b4 a3b4 b7b6

info depth 38 seldepth 64 score cp 12 nodes 263573622895 nps 7488886 time 35195304 multipv 3 pv h7h5 c1b2 b8d7 e2e3 f8e8 f1e1 a7a5 a2a3 f6e4 h2h4 d7f6 a1c1 e4d2 f3d2 f5g6 d1e2 d8b6 g2h3 g6f5 h3f5 e6f5 e2d1 g7g6 d1c2 b6a6 a3a4 e7b4 b2c3 b4a3

[Sean Evans]

New game
rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 1

Analysis by Komodo 5.1r2 64-bit:

1. = (-0.17): 8...h6 9.Ne5 Re8 10.Bb2 Nbd7 11.Nxd7 Nxd7 12.e4 dxe4 13.Bc3 Nf6 14.Re1 e3 15.Rxe3 Qc7 16.Ne4 Red8 17.Qc2 a5 18.a4 Ng4 19.Re2 b6 20.f3 Nf6 21.Rf1 Nxe4
2. = (-0.08): 8...a5 9.Nh4 Bg4 10.a3 Bh5 11.Bb2 Nbd7 12.Re1 b5 13.c5 Qc7 14.h3 a4 15.b4 Rac8 16.Qc2 Rfe8 17.Qd3 Qb7 18.Rac1 Rcd8 19.Ndf3 Ne4 20.Bc3
3. = (-0.07): 8...h5 9.Bb2 Nbd7 10.Nh4 Bh7 11.Nhf3 Re8 12.Qc1 Qc7 13.Ng5 Bf5 14.Ndf3 Ne4 15.Nxe4 Bxe4 16.Rd1 Nf6 17.c5 Rad8 18.Qg5 Bg6 19.Bc1 Ne4 20.Bf4 Qa5 21.Qe5

[Don]

New game
rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 1

Analysis by Komodo 5.1r2 64-bit:

1. = (-0.17): 8...h6 9.Ne5 Re8 10.Bb2 Nbd7 11.Nxd7 Nxd7 12.e4 dxe4 13.Bc3 Nf6 14.Re1 e3 15.Rxe3 Qc7 16.Ne4 Red8 17.Qc2 a5 18.a4 Ng4 19.Re2 b6 20.f3 Nf6 21.Rf1 Nxe4
2. = (-0.08): 8...a5 9.Nh4 Bg4 10.a3 Bh5 11.Bb2 Nbd7 12.Re1 b5 13.c5 Qc7 14.h3 a4 15.b4 Rac8 16.Qc2 Rfe8 17.Qd3 Qb7 18.Rac1 Rcd8 19.Ndf3 Ne4 20.Bc3
3. = (-0.07): 8...h5 9.Bb2 Nbd7 10.Nh4 Bh7 11.Nhf3 Re8 12.Qc1 Qc7 13.Ng5 Bf5 14.Ndf3 Ne4 15.Nxe4 Bxe4 16.Rd1 Nf6 17.c5 Rad8 18.Qg5 Bg6 19.Bc1 Ne4 20.Bf4 Qa5 21.Qe5

On a deep enough search Komodo shows h5 as the second choice.

info multipv 1 depth 24 time 4364657 nodes 5770381590 score cp 17 nps 1322069 pv h7h6 f3e5 f8e8 d2f3 b8d7 c1b2 d8c7 e2e3 a8d8 d1e2 c7a5 f1c1 e7d6 b2c3 a5c7 e5d7 c7d7 c4c5 d6c7 f3e5 c7e5 d4e5 f6e4 c3d4 d7c7 f2f3 e4g5
info multipv 2 depth 24 time 4364657 nodes 5770381590 score cp 10 nps 1322069 pv h7h5 c1b2 a7a5 d1c1 b8a6 b2c3 a5a4 f1d1 a4b3 a2b3 a6b4 c4c5 b4c2 a1a8 d8a8 c1b2 a8a3 b2a3 c2a3 d1a1 a3b5 c3b2 f6e4 d2e4 f5e4 f3e5 e4c2 b3b4 g7g6 e5d7 f8e8 d7b6 g8g7 e2e3
info multipv 3 depth 24 time 4364657 nodes 5770381590 score cp 8 nps 1322069 pv f8e8 f3h4 f5g4 h2h3 g4h5 g3g4 h5g6 h4g6 h7g6 g4g5 f6h5 d2f3 b8d7 a1b1 a8c8 d1d3 c6c5 c4d5 e6d5 f1d1 c5c4 b3c4 d5c4 d3c2 d8c7 c2c3 e7d6 e2e3 c7c6

[Maharadja]

This is Exchess - Gaviota played today in the ACCA - WRCCC

[d]rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 8

In this position, Gaviota played h5?!?!?
Generally, black tries to make room for his bishop to retreat to h7 if white tries Nh4. So, a natural move will be h6... but h5?

Hi,

thx for the game!! I play this opening a lot with black. I will for certain try h5 now.

[Terry McCracken]

This is Exchess - Gaviota played today in the ACCA - WRCCC

[d]rn1q1rk1/pp2bppp/2p1pn2/3p1b2/2PP4/1P3NP1/P2NPPBP/R1BQ1RK1 b - - 0 8

In this position, Gaviota played h5?!?!?
Generally, black tries to make room for his bishop to retreat to h7 if white tries Nh4. So, a natural move will be h6... but h5?

My question is, for your favorite top engine, how far down is h5 in the list of choices?

It was hard to believe, but Gaviota got away with this craziness and won the game. In fact and ironically, the h-pawn ended up helping in the attack. The game is below if you want to see it.

Miguel

[pgn]
[Event "ICS unrated standard match"]
[Site "207.99.83.228"]
[Date "2013.07.20"]
[Round "-"]
[White "EXchess"]
[Black "Gaviota"]
[Result "0-1"]
[WhiteElo "2365"]
[BlackElo "2392"]
[TimeControl "1500+4"]
[Annotator "2... -0.09"]

1. d4 d5 2. Nf3 Nf6 {-0.09/12 0.4} 3. g3 c6 {-0.15/13} 4. c4 Bf5
{-0.26/11 0.4} 5. Nbd2 e6 {-0.17/10 0.2} 6. Bg2 Be7 {+0.08/10} 7. O-O O-O
{-0.17/12 0.1} 8. b3 h5 {+0.19/20 1:58} 9. Ng5 Nbd7 {+0.16/19 19} 10. e4
dxe4 {+0.06/20 38} 11. Ndxe4 Nxe4 {+0.18/21 46} 12. Nxe4 Nf6 {+0.20/20 44}
13. Nxf6+ Bxf6 {+0.21/20 46} 14. Be3 h4 {+0.19/20 43} 15. g4 Bh7
{+0.11/20 2.4} 16. Qd2 Qd7 {+0.09/19 43} 17. g5 Be7 {+0.00/19 43} 18. Qe2
Qd8 {-0.09/20 1:44} 19. f4 Bf5 {+0.35/19 43} 20. d5 exd5 {+0.49/20 49} 21.
cxd5 h3 {+0.44/20 9} 22. Bf3 c5 {+0.40/20 26} 23. Qb5 Bd6 {+0.72/20 2:09}
24. Rae1 Qc7 {+0.92/22 1:29} 25. Qc4 Qa5 {+0.92/20 28} 26. a4 a6 {+0.95/23}
27. Rc1 Rac8 {+1.19/20 31} 28. Qc3 Qc7 {+1.12/19 27} 29. Qd2 Rfe8
{+1.43/19 27} 30. Be2 Qe7 {+1.51/18 26} 31. Kf2 Bc7 {+1.94/19 25} 32. Bc4
Ba5 {+1.78/18 44} 33. Qe2 Qd7 {+1.87/19 22} 34. Qf3 Bb4 {+2.17/17 23} 35.
d6 b5 {+3.01/18 22} 36. Bd5 c4 {+3.48/17 22} 37. Kg3 cxb3 {+3.48/18 20} 38.
Rxc8 Rxc8 {+2.97/17 21} 39. axb5 Rc2 {+3.40/17 20} 40. Rg1 axb5
{+3.75/18 19} 41. Bxb3 Be1+ {+2.74/20} 42. Rxe1 Rg2+ {+3.19/20} 43. Qxg2
hxg2 {+3.53/18 21} 44. Bf2 Qxd6 {+4.08/19 19} 45. Kxg2 Qxf4 {+4.93/17 18}
46. Re8+ Kh7 {+4.93/18 2.3} 47. g6+ Kxg6 {+5.31/17 18} 48. Re3 b4
{+5.48/17} 49. Kg1 Be6 {+6.24/17 11} 50. Bc2+ Kh6 {+6.54/18 17} 51. Kg2
Qg4+ {+8.60/17 23} 52. Bg3 Bd5+ {+8.89/18} 53. Kf2 f5 {+9.10/18 16} 54. Ke1
b3 {+10.27/16 1.2} 55. Bxb3 Qb4+ {+11.48/16 16} 56. Kf2 Bxb3 {+12.26/16 14}
57. Be5 f4 {+12.26/14 15} 58. Rh3+ Kg6 {+17.98/16} 59. Rxb3 Qxb3
{+327.22/14 23} 60. Bxf4 Kf5 {+327.54/15 6} 61. Bc1 Ke4 {+327.56/13 1.3}
62. Bg5 Qf3+ {+327.58/11 1.0} 63. Kg1 Qe2 {+327.60/9} 64. Be3 Kxe3
{+327.62/7} 65. h3 Kf3 {+327.64/2} 66. h4 Qg2# {+327.66/2 0.1}
{EXchess checkmated} 0-1
[/pgn]

Imho it looks like a good move and I doubt it really compromises Black.