[BrianWall-ChessList] DC Gentle on Dana MacKensie's Queen sac
Brian Wall
brianwallchess3 at taom.com
Mon Nov 30 18:39:09 MST 2009
I mentioned Dana Mackensie's Queen sac at Chess_Improvement at Yahoogroups.com and
DC Gentle, a frequent contributor/refuter, did some heavyeight analysis with
computers and a friend - You can pop the whole thing into an engine which is
what Chris Peterson keeps asking me to do.
BW
-------------------------------------------------------------------------------
----- Forwarded message from Dc Gentle <dcgentle at freenet.de> -----
Date: Tue, 1 Dec 2009 00:07:07 +0100
From: Dc Gentle <dcgentle at freenet.de>
Reply-To: Chess_Improvement at yahoogroups.com
Subject: [Chess Improvement] Re: Dana MacKensie's Queen sac
To: Cimp <Chess_Improvement at yahoogroups.com>
On Dana's website, that is
http://www.danamackenzie.com/game1.htm
you can find the commented version of the
following game:
[Event "Western States Open"]
[Site "*"]
[Date "2006.08.10"]
[Round "?"]
[White "Mackenzie, Dana"]
[Black "Pruess, David"]
[Result "1-0"]
[ECO "B21"]
1. e4 c5 2. f4 d5 3. Nf3 dxe4 4. Ng5 Nf6 5. Bc4 Bg4 6. Qxg4
Nxg4 7. Bxf7+ Kd7 8. Be6+ Kc6 9. Bxg4 e6 10. Nc3 Na6 11. a3
Bd6 12. O-O Nc7 13. Ncxe4 Qe7 14. Nxd6 Qxd6 15. d3 Raf8 16.
Bf3+ Kd7 17. c3 Nd5 18. g3 h6 19. Ne4 Qc7 20. b4 cxb4 21. axb4
b6 22. Bd2 Rf7 23. c4 Nf6 24. Bc3 Ke7 25. Be5 Qd7 26. Nd6 Rd8
27. Nxf7 Kxf7 28. d4 Kg6 29. g4 Rc8 30. c5 Qb5 31. Rxa7 Qd3
32. h4 h5 33. g5 Ne8 34. Kg2 b5 35. Re1 Kf5 36. Be4+ Qxe4+ 37.
Rxe4 Kxe4 38. Ra5 Nc7 39. Bxg7 Kxf4 40. Be5+ Kg4 41. g6 Kxh4
42. g7 Rg8 43. Ra7 Nd5 44. Rf7 Ne3+ 45. Kf3 Ng4 46. c6 1-0
Dana comments on the queen sac at move 6 (quotation):
"I first started playing this queen sacrifice in 2004, as a
crazy experiment against Fritz, the computer chess program.
After playing perhaps a hundred games with it over two years
against the computer, I believe that the queen sacrifice is
completely sound. It is far too complex, however, for an
exhaustive analysis, which makes it an ideal weapon to play
against the computer. This game proves that it works pretty
well against humans, too!"
Well, I can only say that a sacrifice of a queen at such an
early stage of the game is most likely not sound in any way
because I learned the hard way that even sacrificing a
simple pawn in any gambit opening has only a chance of
success if this sac is followed by a sustainable initiative.
Dana however tries a positional game (11. a3) against the
versatile queen, which may only deliver some chances
if the opponent doesn't really know what to do.
And if Dana thinks a hundred games against a computer
will prove anything, she is wrong, because computers
don't understand this opening, as you'll notice later.
I showed the above game to Peter, my friend, who already
helped me analysing the King's Gambit earlier this year.
I claimed that Black will win but he said, ok, but only with
a fight, and we started to send variants to each other.
Both of us used computers and so we noticed where these
engines went wrong. The reasons for the chess engines'
failures are manifold: Because material balance is already
gone at move 6, they have to find compensation by try
and error during their search, and often they can't get deep
enough to find anything, and then they don't really know
how important either the initiative or the center is during
the opening phase of the game.
Many members of this group believe that the play of
computers is almost perfect, but the following analysis
will prove otherwise. All subvariants will start with inferior
computer moves which are just blunders sometimes.
As always with my analysis, you can copy and paste it
into a PGN-Reader like Chessbase (light) or Fritz.
__________________________
[Event "Analysis of the Budapest Gambit with reversed colors"]
[Site "Chess Improvement"]
[Date "2009.11.30"]
[Round "?"]
[White "Peter"]
[Black "Dc Gentle"]
[Result "0-1"]
[ECO "B21"]
[Annotator "Gentle,DC"]
1. e4 c5 2. f4 d5 3. Nf3 dxe4 4. Ng5 Nf6 5. Bc4 Bg4 6. Qxg4
{This queen sacrifice is questionable if Black knows how to
deal with it.}
6... Nxg4 7. Bxf7+ Kd7 8. Be6+ Kc6 9. Bxg4
{Black's next move is already decisive, because White has
achieved a superior development, but only for the time being.}
9... e5
{! Usually Black plays 9... e6 here, but I felt this move is
inferior. The text move attacks f4 and strengthens Black's
influence over the center. Now White has to fight for a draw
already.}
10. Nf7 Qe8
{Engines want to play 10... Nxh8.}
11. Nxe5+
{Because White is striving for the better development, it's
not good to take rook h8, awkwardly placing the knight in the
corner. Much better is it trying to keep the initiave by
delivering check and maintaining the knight in the center.}
(11. Nxh8
{is the move that most engines would play, not
realizing that Black will get an easy game after this
mistake.}
11... exf4 12. Nc3 Nd7 13. O-O Nf6 14. Be2
{White has to retreat already.}
14... a6 15. d3 f3 16. gxf3 exd3 17. Bxd3 Qh5
{The strength of the black queen and the sorry state of
the white kingsite will decide the outcome here.}
18. Be2 Bd6 19. f4 Qe8 20. Bd2 Qd7 21. Bf3+ Kc7
22. Rad1 Rxh8 23. Be1 Qf5 24. Bxb7 Kxb7 25. Rxd6 Qxc2
26. Rd2 Qf5 27. Rg2 g6 28. Rg5 Qh3 29. Rg3
{White cannot annoy the agile queen.}
29... Qd7 30. Bf2 Qf5 31. Rg5 Qxf4 32. Rg3 Nh5 33. Be3
Qc4 34. Rgf3 Re8 35. Kf2 Re6 36. Kg1 Qg4+ 37. Kh1 Re5
38. h3 Qe6 39. Bc1 Rf5 40. Kh2 Qc6 41. Rxf5 gxf5
42. Ne2 Qe4 43. Rf2 Qc2 44. Be3 Qxb2 45. Nd4 Qa3
46. Nxf5 c4 47. Kg2 Ng7 48. Bh6 Nxf5 49. Rxf5 Qxa2+
50. Kf3 c3 51. Be3 c2 52. Rc5 Qb1 53. h4 a5 54. Rxa5
Qh1+ 55. Ke2 Qd1+ 56. Kf2 c1=Q 57. Bxc1 Qc2+ 58. Kg3
Qc7+ 59. Kg4 Qxa5
{and Black will deliver checkmate in 12 moves. Did you
see the queen hastening to and fro collecting white material?})
11... Kc7 12. Nc3 Bd6 13. O-O h5
{is the only move that prevents White from maintaining the
initiative.}
(13... Bxe5
{will be played by most engines. They underestimate the
white chances in the center.}
14. fxe5 Nd7 15. d4 cxd4
{The black center pawns are weak and will soon fall.}
16. Nb5+ Kb6 17. Nd6 Qd8 18. Bxd7 Qxd7 19. Rd1 Kc7
20. Rxd4 Raf8 21. Rxe4
{White now rules the center and has got no problems to
improve his position.}
21... Kb8 22. Be3 Qe6 23. c4 Rd8 24. Re1 Ka8 25. Bf2
Rhf8 26. Rd4 Rd7 27. Rd5 Re7 28. b3 Rd7 29. a4 h5
30. Red1 Rfd8 31. a5 Kb8
{Black can only wait for the things to come.}
32. Rf1 Rg8 33. h3 a6 34. Bc5 Ka8 35. b4 Rdd8 36. b5
axb5 37. Nxb5 Rc8 38. Nd6 Rc6 39. Bd4 h4 40. Rf3 Rb8
41. Be3 Rd8 42. Rf1 Rh8 43. Rfd1 Rf8 44. Nb5 Qe7
45. Bb6 Kb8 46. e6 Rcc8 47. Na7 Rce8 48. Rd6 Qf6
49. Bd4 Qxd4+ 50. R1xd4 Kxa7 51. Rxh4 Re7 52. Rd7
Rfe8 53. Rxe7 Rxe7 54. Kf2 Re8 55. Rf4
{and Black can resign.})
14. Bf5 Rf8 15. Bg6 Qe6 16. Nb5+ Kd8
(16... Kc8
{would be preferred by most engines, but is also a
blunder because of}
17. Nf7 Qxg6 18. Nbxd6+ Kc7 19. f5 Qf6 20. d3 Kc6
21. Bg5 Qd4+ 22. Kh1 exd3 23. Rad1 Qxb2 24. Rxd3 Nd7
25. Re1 Qxa2 26. c4 b5 27. Rd2 Qb3 28. h3 bxc4 29. Re6
c3 30. Nc4+ Kb5 31. Rxd7 Qb1+ 32. Kh2 Kxc4 33. Nd6+
Kb3 34. Rb7+ Ka2 35. Rxb1 Kxb1 36. Re3
{and draw.})
17. Nxd6 Qxd6 18. Nf7+ Rxf7 19. Bxf7 Nc6 20. c3 h4 21. d3
exd3
{Pawn d3 is White's worry already.}
22. f5 Qf6 23. Be6 Kc7 24. Bf4+ Kb6
{The black king has found a safe haven here and rook h8 is
the last piece to be developed. White has lost any initiave by
now.}
25. Rad1 Rd8 26. h3 g6 27. Rf2
{in order to double rooks on the d- or f-file.}
27... Ne5 28. Be3 gxf5 29. Bxf5 Nc4 30. Bf4 Qxf5
{! This move is not easy to find, but it's the only way to
prevent a white fortress, as you can see from the variant
starting with 30... d2 below. Of course White has got a
discovered check now.}
(30... d2
{preferred by most engines, this move tries to keep the
biggest material advantage and doesn't realize the
consequences.}
31. Bg4 Qe7 32. b3 Ne3 33. Re2 Rd3 34. c4 Ka5 35. Bf3
Rc3 36. Bxe3 Rxe3 37. Rdxd2 Kb4 38. Kh2 Qc7+ 39. Kh1
Rxe2 40. Rxe2
{By all these exchanges Black has lost any chance to
win already.}
40... b6 41. Kg1 Qg7 42. Kh2 Ka3 43. Bd5 a5 44. Bf3 Qf6
45. Kh1 Kb4 46. Rc2 a4 47. bxa4 Qd4 48. Kh2 Ka3
49. Bd5 Kxa4 50. Re2 b5 51. Re4 Qd3 52. cxb5+ Ka3
53. Bc6 c4 54. Rxh4 c3 55. Be4 Qd6+ 56. g3 c2 57. Bxc2
Qd2+ 58. Kg1 Qc1+ 59. Bd1 Qxd1+ 60. Kf2 Qc2+ 61. Kf3
Qd3+ 62. Kf2 Qd2+ 63. Kf3 Qd5+ 64. Kf2 Qxa2+ 65. Kf3
Qb3+ 66. Kf2 Qb2+ 67. Kf3 Qc3+ 68. Kg2 Qd2+ 69. Kf3
Qd3+ 70. Kf2 Qf5+
{A lone queen, only miserably supported by the own
king, cannot do much in this position.}
71. Kg2 Qd5+ 72. Kf2 Qxb5
{and draw according to the endgame tablebases.})
31. Bc7+
{winning Black's queen, but not without conceding own
material.}
31... Kxc7 32. Rxf5 Ne3
{A nice fork winning a rook.}
33. Rxc5+
{White will sacrifice a rook in order to get pawn d3 and to
get further chances to build a fortress.}
(33. Rf7+
{is the move that would be played by most engines
ending the game rather quickly:}
33... Kc8 34. Rd2 Nc4 35. Rd1 Nxb2 36. Rd2 Nc4 37. Rd1
d2 38. Rf2 b5 39. Kf1 Ne3+ 40. Ke2 Nxd1 41. Kxd1 Rd3
42. Rf4 Rxc3 43. Rxh4 Rd3 44. Rf4 c4 45. h4 c3 46. Rf8+
Kd7 47. Rf1 b4 48. Kc2 Rd5 49. Rh1 Re5 50. Rd1 Re1
51. h5 Rxd1 52. Kxd1 Ke6 53. h6 Kf7 54. Kc2 Kg6 55. h7
Kxh7 56. g4 a5
{and the strong pawn phalanx will win the game for
Black.})
33... Kb6 34. Rc8 Rxc8 35. Rxd3 Nf5 36. Kf2 Rf8 37. Rf3 Kc5
38. Rf4 b6
{Advancing the queensite pawns is Black's key idea to win
this ending.}
(38... Kd5
{The following engine variant displays the inability of
even top engines to realize the conditions that Black
has to avoid if he wants to win:}
39. Kf3 b6 40. Ra4 Rf7 41. Re4 Nd4+ 42. Ke3 Nc2+
43. Kd3 Rf2 44. c4+ Kd6 45. Kc3 Rxg2 46. b4 Na3
47. Rxh4 Rg3+ 48. Kd2 a5 49. Rh6+ Kc7 50. Rh7+ Kc8
51. Rh8+ Kd7 52. Rh7+ Kd8 53. Rh8+ Ke7 54. Rh7+ Ke8
55. c5
{Black must prevent the white queensite pawns from
marching but engines don't see this.}
55... Nc4+ 56. Kc2 axb4 57. cxb6
{Now the black pawn b4 is weak.}
57... Rc3+ 58. Kb1 Nxb6 59. Rb7 Rc6 60. h4 Kf8 61. Rb8+
Ke7 62. h5 Kd7 63. Rg8 Rc4 64. Rg3 Rh4 65. a3 bxa3
66. Rxa3 Rxh5
{and draw according to the EGTBs.})
39. Kf3 Rf7 40. Ke4 Nd6+ 41. Kf3 Rh7 42. Rd4 a5
{This move is also missed by most engines.}
43. Rg4 Kd5 44. Rg5+ Ke6 45. Rg6+ Ke5 46. Rg5+ Kf6
47. Rg4 Rf7 48. Ke2 Nf5
{Black only waits fo the chance to invade the white camp.}
49. Kf3 Rf8 50. Rc4 b5 51. Rg4 Ke5
(51... Rd8 {would be wrong:}
52. a4
{It's Black who should advance to a4, not White.}
52... Rd2 53. b4 Rd3+ 54. Ke2 Re3+ 55. Kf2 axb4 56. Rxb4
bxa4 57. Rxa4
{Again all black queenside pawns have vanished and
White will survive:}
57... Rxc3 58. Rf4 Rb3 59. Re4 Ra3 60. Rf4 Ke5 61. Rc4
Nd4 62. Rc7 Kf4 63. Rc4 Ra2+ 64. Kg1 Ke4 65. Rc8 Rb2
66. Re8+ Kf4 67. Rc8 Kf5 68. Rf8+ Ke5 69. Kh2 Ra2
70. Re8+
{Although Black has got a knight against White's
surplus pawn, it's a draw because the white fortress will
hold. If one side has less material, often the only way to
defend a draw is to set up a fortress, although current
engines don't "know" what a fortress is after all.})
52. Re4+ Kd5 53. Rg4 a4 54. Rf4 Rf7 55. Rb4 Nd6+ 56. Ke2
Rg7 57. Kf2 Rh7 58. Rg4 Ne4+
{Finally the time is ripe to invade.}
59. Kf3 Rf7+ 60. Ke3 Re7 61. Kf3 a3
{The white queensite pawns are doomed.}
62. bxa3 Nxc3 63. Rxh4 Ra7 64. g4 Rxa3 65. Rh8 Nxa2+
66. Kf4 b4
{The b-pawn will decide the day.}
67. g5 b3 68. g6 Ra7 69. Rb8 Kc4 70. Kf5 Ra8 71. Rb7 Nb4
72. Rc7+ Kd4 73. Rc1 b2 74. Rb1 Kc3 75. g7 Nd5 76. h4
Ne7+ 77. Ke6 Ng8 78. Kf7 Kc2 79. Re1 b1=Q 80. Rxb1 Kxb1
{and Black will win in 15 moves according to the EGTBs:}
81. Kg6 Kc2 82. Kg5 Kc3 83. Kf5 Kd4 84. Kg4 Ra7 85. Kg5
Rxg7+ 86. Kf5 Rg1 87. Kf4 Rf1+ 88. Kg3 Nf6 89. Kg2 Rf4
90. Kg1 Ke4 91. Kg2 Nh5 92. Kg1 Ke3 93. Kh2 Kf2 94. Kh3
Kf1 95. Kh2 Rxh4# 0-1
____________________
Have fun,
DC
More information about the BrianWall-ChessList
mailing list