Solving endgames with ChessBase 12 and Houdini

by ChessBase
12/23/2012 – How does one analyse fortress like positions with a computer? In the London Chess Classic round eight game Anand vs Nakamura the World Champion had set up a fortress which Nakamura was unable to penetrate. Using Let's Check and Houdini our ChessBase Magazine columnist GM Karsten Müller was able to do just that and find a clear win – specifically a missed mate in 93 moves.

ChessBase 18 - Mega package ChessBase 18 - Mega package

Winning starts with what you know
The new version 18 offers completely new possibilities for chess training and analysis: playing style analysis, search for strategic themes, access to 6 billion Lichess games, player preparation by matching Lichess games, download Chess.com games with built-in API, built-in cloud engine and much more.

More...

Anand's Airy Army

When pieces fight against a queen coordination is often of crucial importance.


A critical moment in the game between Hikaru Nakamura and Vishy Anand

[Event "4th London Chess Classic"] [Site "London"] [Date "2012.12.09"] [Round "8.3"] [White "Anand, Viswanathan"] [Black "Nakamura, Hikaru"] [Result "1/2-1/2"] [ECO "B90"] [Annotator "Müller,Karsten"] [PlyCount "130"] [EventDate "2012.??.??"] [EventCountry "ENG"] [SourceDate "2009.06.14"] 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. h3 e5 7. Nde2 h5 8. Bg5 Be6 9. Bxf6 Qxf6 10. Nd5 Qd8 11. Nec3 Nd7 12. Bc4 g6 13. a4 Bh6 14. a5 Rc8 15. Ba2 O-O 16. O-O Kg7 17. b4 Rc6 18. Qd3 Qg5 19. Rfd1 Rfc8 20. Bb3 f5 21. exf5 gxf5 22. Ne2 f4 23. Ndxf4 Bxb3 24. h4 Qg4 25. f3 Bxc2 26. fxg4 Bxd3 27. Nxh5+ Kg8 28. Rxd3 Bf8 29. Kh2 Rc4 30. Kh3 Rxb4 31. Rf3 Rc2 32. Nc3 e4 33. Rg3 Rd2 34. Re1 d5 35. Nf4 Rbd4 36. Ncxd5 Rxd5 37. Nxd5 Rxd5 38. Rxe4 Rxa5 39. g5 Ra3 40. Re8 Rxg3+ 41. Kxg3 b5 42. Rd8 Nc5 43. Rb8 Kf7 44. g6+ Kg7 45. Kg4 b4 46. h5 b3 47. Kf5 Bd6 48. Rb4 a5 49. Rb6 a4 50. Rxd6 b2 51. Rb6 a3 52. Kg5 Ne4+ 53. Kf4 a2 {Anand's Airy Army When pieces fight against a queen coordination is often of crucial importance:} 54. Rb7+ (54. Rxb2 a1=Q {and only now check to reach f7 safely also does not help due to} 55. Rb7+ Kf6 (55... Kh6 $4 56. Rh7#) 56. Rf7+ (56. Rb6+ Ke7 57. Kxe4 Qe1+ 58. Kf3 Qc3+ 59. Ke4 Qc4+ 60. Kf3 Qd3+ 61. Kf2 Qd4+ $19) 56... Ke6 57. Kxe4 {Now the rook is safe for the moment, but it does not help in the long run:} Qa8+ 58. Kf4 Qxg2 {This removes the backbone of White's fortress.} 59. Rh7 Qe2 60. Kg5 Qe3+ 61. Kg4 Qd4+ 62. Kg5 Qe5+ 63. Kh6 Kf5 64. Rf7+ Kg4 65. Kh7 Qxh5+ 66. Kg8 Qd5 $19) 54... Kf8 55. Rxb2 a1=Q 56. Rb8+ Ke7 $1 57. Kxe4 Qe1+ $6 {This drives White's king in the wrong direction.} (57... Qa4+ $1 58. Ke5 (58. Kd3 Qd7+ 59. Ke3 Qe6+ 60. Kf3 Qf5+ 61. Kg3 Qe5+ $19 ) 58... Qc6 $1 {does not allow White to coordinate his scattered forces and he loses quickly, e.g.} 59. g7 (59. Rb1 Qc3+ 60. Kf4 Qd2+ 61. Kf5 Qc2+ $19) (59. Rb2 Qf6+ $19) 59... Qd6+ $19) 58. Kf3 Qc3+ $2 {This allows White's king to reach the safety of the kingside pawns. But how to analyse this fortress like position with a computer? Here it is good that the new ChessBase 12 is connected to the Let's Check database and there I found that Houdini 3 x64 had evaluated the position after} (58... Qd1+ $1 {with Black mates in 93 moves. While I do not trust the number I believed that Black must be winning and after a short check I could confirm it. Houdini 3 indeed seems to be very strong at cracking such fortress like set ups. Some sample lines run} 59. Ke3 ( 59. Kg3 $6 Qd6+ $19) 59... Qc1+ 60. Kd3 (60. Kf3 Qc6+ 61. Ke3 Qe6+ 62. Kf3 Qf5+ 63. Ke3 Qe5+ $19) 60... Qf1+ 61. Ke3 Qxg2 {This removes the backbone of White's fortress. As his king is now caught in the middle of nowhere, White is lost (this is already a six-man tablebase position of course), e.g.} 62. Rb5 Qc6 63. g7 Kf7 64. Rg5 Qc1+ $19) 59. Kg4 $1 Qd4+ 60. Kh3 $1 Qd3+ (60... Qe3+ { does not help due to} 61. g3 Qe5 62. Rb7+ Kf6 63. g7 Qxh5+ 64. Kg2 Qd5+ 65. Kf2 Kg6 {and White can even play} 66. g8=Q+ Qxg8 67. Rb4 Qf7+ 68. Rf4 $11 {as with a knight's pawn this constellation with two safe anchor squares for the rook is also drawn despite the fact that the pawn has advanced to g3. In the position shifted one file to the left (with the pawn on f3) White would be lost.}) 61. Kh4 Qe4+ 62. g4 Qe1+ 63. Kh3 Qe3+ 64. Kh4 Qe1+ 65. Kh3 Qe3+ 1/2-1/2

Aronian's Attacking Amazons

It is very seldom that two queens fight to storm a fortress.


By far the longest game of round four, it had the audience at the edge until the end

[Event "4th London Chess Classic"] [Site "London"] [Date "2012.12.04"] [Round "4"] [White "McShane, Luke"] [Black "Aronian, Levon"] [Result "0-1"] [ECO "C55"] [WhiteElo "2713"] [BlackElo "2850"] [Annotator "Müller,Karsten"] [PlyCount "152"] [EventDate "2012.??.??"] [EventCountry "ENG"] 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. d3 b5 7. Bb3 d6 8. a4 Bd7 9. c3 O-O 10. Nbd2 Na5 11. Bc2 c5 12. Re1 Re8 13. Nf1 Nc6 14. Ne3 b4 15. h3 Rb8 16. Nc4 Be6 17. Bb3 h6 18. Be3 bxc3 19. bxc3 d5 20. exd5 Nxd5 21. Qc2 e4 22. dxe4 Nxe3 23. Rxe3 Rxb3 24. Qxb3 Na5 25. Nxa5 Bxb3 26. Nxb3 Qb6 27. Nbd2 Qb2 28. Rae1 Rd8 29. Nf1 c4 30. Ne5 Bc5 31. Rf3 Qb3 32. a5 Rd6 33. Nxf7 Rf6 34. Rxf6 gxf6 35. Nxh6+ Kf8 36. Ng4 Qxc3 37. Rd1 Qb4 38. Nxf6 Kf7 39. Nd5 Qb2 40. Nde3 c3 41. Rd5 Be7 42. Rf5+ Ke8 43. g4 c2 44. Nxc2 Qxc2 45. Ng3 Qc7 46. Nh5 Bd8 47. Kg2 Qc6 48. Re5+ Kf7 49. g5 Qa4 50. h4 Bxa5 51. Rf5+ Ke6 52. Nf4+ Kd7 53. f3 Qc2+ 54. Kh3 Qf2 55. Kg4 Be1 56. Ng6 a5 57. Rf7+ Kd8 58. e5 a4 59. e6 Bb4 60. e7+ Bxe7 61. Nxe7 Qc5 62. Kh5 a3 63. Kh6 a2 64. g6 {Aronian's Attacking Amazons It is very seldom that two queens fight to storm a fortress: } Qc4 {Aronian stops White's play first.} ({The direct promotion} 64... a1=Q { wins as well:} 65. g7 Qa6+ 66. Ng6 Qe3+ 67. Kh7 (67. f4 Qe8 $19) 67... Qe8 68. g8=Q Qxg8+ 69. Kxg8 Qxg6+ $19) 65. Nf5 a1=Q 66. Rf8+ Kc7 67. g7 {The pawn promotion comes too late, but White's pawns are also too slow after} (67. h5 $5 Qe5 $1 {A strong fortress breaker found quickly by Houdini 3. White has no escape against Black's centralised amazons.} 68. Kh7 Qcc5 {and Black wins, e.g. } 69. Rf7+ Kd8 70. Ng7 Qcd6 71. Nf5 Qde6 72. Rf8+ Kd7 73. g7 Q5xf5+ 74. Rxf5 Qxf5+ 75. Kh6 Qf7 76. Kh7 Qxh5+ 77. Kg8 Ke6 78. f4 Ke7 79. f5 Kf6 80. Kf8 Qf7#) 67... Qc6+ 68. Kh7 Qxf3 69. g8=N {The underpromotion does not help, but there was no salvation anywhere to be found anyway e.g.} (69. g8=Q Qh5+ 70. Nh6 Qb1+ 71. Rf5 Qbxf5+ 72. Kg7 Qhg6+ 73. Kh8 Qe5+ 74. Qg7+ Qgxg7#) (69. Rf7+ Kb6 $1 70. g8=Q Qh5+ 71. Nh6 Qb1+ 72. Kg7 Qbg6+ 73. Kf8 Qc5+ 74. Re7 Qc8+ 75. Re8 Qcxe8#) 69... Qh5+ 70. Ngh6 Qe5 $1 {Centralisation is often called for as the queen conrols so many important squares from the middle of the board.} 71. Ng7 (71. Rf7+ Qxf7+ 72. Nxf7 Qxf5+ $19) 71... Qxh4 72. Rf7+ Kb6 73. Ngf5 Qee4 74. Kg6 Qe6+ 75. Kg7 Qg5+ 76. Kf8 (76. Kh8 Qgg6 77. Rg7 Qe5 78. Ng8 Qgxf5 79. Ne7 Qf8+ 80. Kh7 Qh5#) 76... Qc8# 0-1

Tania's Turnaround

In pawn endings extreme care is often called for.


Wolfgang Uhlmann playing against Tania Sachdev at the Snowdrops-Oldhands event

[Event "Snowdrops vs Oldhands"] [Site "Podebrady CZE"] [Date "2012.12.13"] [Round "5.3"] [White "Uhlmann, Wolfgang"] [Black "Tania, Sachdev"] [Result "0-1"] [ECO "A29"] [WhiteElo "2319"] [BlackElo "2400"] [PlyCount "98"] [EventDate "2012.12.08"] [EventType "schev"] [EventRounds "8"] [EventCountry "CZE"] 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. g3 d5 5. cxd5 Nxd5 6. Bg2 Nb6 7. d3 Be7 8. a3 Be6 9. O-O O-O 10. Be3 f6 11. Rc1 Nd4 12. Bxd4 exd4 13. Na4 c6 14. Nc5 Bxc5 15. Rxc5 Bd5 16. Nd2 Bxg2 17. Kxg2 Qe7 18. Rc1 Rad8 19. Re1 Rd5 20. Nf3 Qd7 21. Qb3 Kh8 22. Qb4 Rd8 23. Rc5 Rxc5 24. Qxc5 Qd5 25. Qe7 Rd7 26. Qf8+ Qg8 27. Qb4 Qd5 28. Rc1 Kg8 29. a4 Qd6 30. Qa5 Qb8 31. Qb4 Qd6 32. Qxd6 Rxd6 33. a5 Nd7 34. b4 b6 35. Nd2 c5 36. Ne4 Rd5 37. bxc5 Nxc5 38. Nxc5 Rxc5 39. Rxc5 bxc5 40. Kf3 Kf7 41. Ke4 Ke6 42. g4 g6 43. h4 h6 44. h5 gxh5 45. gxh5 f5+ 46. Kf4 a6 $2 { Tania's Turnaround In pawn endings extreme care is often called for:} 47. e3 $2 {This runs into a mighty counterblow.} (47. e4 $1 {is the correct way to realise Uhlmann's plan:} dxe3 (47... c4 48. dxc4 d3 49. exf5+ Kf6 50. Ke3 $18) 48. fxe3 Kf6 49. e4 fxe4 50. dxe4 c4 51. e5+ Ke6 52. Ke4 c3 53. Kd3 Kxe5 54. Kxc3 Kd5 (54... Kf4 55. Kd4 Kg5 56. Kc5 Kxh5 57. Kb6 Kg4 58. Kxa6 h5 59. Kb6 h4 60. a6 h3 61. a7 h2 62. a8=Q $18) 55. Kd3 {This gains the opposition and White will win the resulting races:} Kc5 56. Ke4 Kb5 57. Kf5 Kxa5 58. Kg6 Kb4 59. Kxh6 a5 60. Kg6 a4 61. h6 a3 62. h7 a2 63. h8=Q $18 {as given in Chess Today 4422.}) 47... c4 $3 48. dxc4 (48. exd4 {is met by} c3 49. Ke3 f4+ 50. Ke2 Kd5 51. Kd1 Kxd4 52. Kc2 f3 $19) 48... d3 49. Kf3 (49. e4 d2 $19) 49... Ke5 (49... Ke5 50. e4 f4 $1 {This puts White's king in a typical zugzwang constellation.} ({However, not} 50... fxe4+ $2 51. Ke3 Kd6 52. f3 exf3 53. Kxd3 Kc5 54. Ke3 Kxc4 55. Kxf3 Kb5 56. Ke4 Kxa5 57. Kf5 Kb4 58. Kg6 a5 59. Kxh6 a4 60. Kg7 a3 61. h6 a2 62. h7 a1=Q+ 63. Kg8 $11 {and Black's king is outside the winning zone.}) 51. c5 Ke6 52. c6 Kd6 53. Kxf4 d2 $19) 0-1

Karsten Müller in ChessBase Magazine

Do you like these lessons? There are plenty more by internationally renowned endgame expert Dr Karsten Müller in ChessBase Magazine, where you will also find openings articles and surveys, tactics, and of course annotations by the world's top grandmasters.

Click to go to the ChessBase Magazine page

Apart from his regular columns and video lectures in ChessBase Magazine there is a whole series of training DVDs by Karsten Müller, which are bestsellers in the ChessBase Shop.


Reports about chess: tournaments, championships, portraits, interviews, World Championships, product launches and more.

Discuss

Rules for reader comments

 
 

Not registered yet? Register