QueenEndings

Queen and pawn endings are notoriously difficult for human players - even top grandmasters - to understand and play correctly. The second diagram, where the ending actually arises, should be a forced win for Black. The correct transition to the queen endgame can also pose problems, as both the examples in this article demonstrate.

In the first diagram, from a recent top tournament, GM Karjakin (White) has been under pressure for a long time in a difficult bishop endgame. He has just sacrificed his bishop on f4 at move 69 to force promotion of his e-pawn but world number two Fabiano Caruana has a resource to reach a theoretically won queen ending.

If Caruana had just played 61...g1Q the tablebase shows that the resulting position would be drawn despite Black's extra piece. By giving back his bishop, Black drove the white king away from the defence of the d-pawn so that his new queen could take that pawn with check.

The second diagram shows the position that then arose. Would you be able to win this position in practice, or even put your opponent under pressure? The commentators thought it would be very hard for Karjakin to save this ending and in fact the tablebase shows that Black can force mate in 61 moves. Caruana, however, had thought too long about capturing the bishop (even repeating the position rather than taking it at the first opportunity) and he had no time to play properly.











(1) Karjakin,Sergey (2753) - Caruana,Fabiano (2803) [D41]
FIDE Grand Prix Khanty-Mansiysk (9.2), 24.05.2015
[Harding,T.]

1.Nf3 Nf6 2.c4 c5 3.Nc3 Nc6 4.g3 d5 5.d4 e6 6.cxd5 Nxd5 7.Bg2 cxd4 8.Nxd4 Nxc3 9.bxc3 Nxd4 10.Qxd4 Qxd4 11.cxd4 Bd6 12.0-0 Rb8 13.e4 b6 14.Bb2 Bb7 15.Rac1 f6 16.Rfd1 Ke7 17.f4 Rhc8 18.e5 fxe5 19.fxe5 Bb4 20.Rxc8 Bxc8 21.Be4 g6 22.Rc1 Kd8 23.Kf2 Bb7 24.Kf3 Bd5 25.Bxd5 exd5 26.Ke3 Rc8 27.a4 Rxc1 28.Bxc1 Kd7 29.Bd2 Be7 30.Kd3 h5 31.a5? bxa5 32.Bxa5 g5 33.h3 g4 34.hxg4 hxg4 35.Ke3 Bg5+ 36.Kd3 Kc6 37.Bd2 Be7 38.Ke3 Bg5+ 39.Kd3 Be7 40.Ke3 Bf8 41.Ba5 Bh6+ 42.Kd3 Kd7 43.Bd2 Bf8 44.Ke3 Ke6 45.Ba5 Kf5 46.Kd3 Be7 47.Ke3 Bg5+ 48.Kd3 Ke6 49.Bd2 Bd8 50.Ke3 Kf5 51.Bb4 a5 52.Ba3 a4-/+ 53.Bb4 Bg5+ 54.Kd3 Bc1 55.Bd2 Ba3 56.Kc3 Be7 57.Bh6 a3 58.Kb3 Bb4 59.Bf4 Ke6 60.Bc1 Be1 61.Be3 Bxg3 62.Kxa3 Kf5 63.Kb4 Be1+ 64.Kc5 Ke6 65.Bf4 g3 66.Bg5 Kf5 67.Bf4 [67.Bh4? g2 ] 67...Ke6 Caruana repeated. 67...Kxf4 could have been played immediately. 68.Bg5 Kf5 69.Bf4 Kxf4 The 7-man tablebase shows that Black has an objectively winning position - but the follow-up is not simple! 70.e6 g2 71.e7 Bb4+ 72.Kxb4 g1Q 73.e8Q Qxd4+ The ending of queen and pawn against queen should normally be much easier for the player with the extra pawn. 74.Ka5 Qe5? Giving White a chance to save the game, with Karjakin having about 12 minutes on his clock. [74...Kf3 best says tablebase. Other winning moves; 74...Qc3+ ; 74...Qc5+ ; 74...Qd2+ ; 74...Qe4 offering Q exchange in centre] 75.Qc8! Karjakin finds one of the defences (so far). Also sufficient, if correctly followed up, would be [75.Qf8+ ; 75.Qd7 ; 75.Qc6 or; 75.Qg6 ] 75...d4+ 76.Kb4 d3 Probably not the best practical chance. Four moves hold here: 77 Qd8, 77 Qc1, 77 Qh3 and 77 Kb3. Karjakin found one of them. 77.Qh3! Ke4 Many moves hold here. 78.Qh1+ Now it is a simple draw as White can swap queens and get his K back. 78...Kd4 79.Qa1+ Kd5 80.Qxe5+ Kxe5 81.Kc3 Ke4 82.Kd2 Kd4 83.Kd1 Kc3 84.Kc1 d2+ 85.Kd1 Kd3 stalemate 1/2-1/2



 

In this example, from the recent French Team Championship at Montpellier, White has overpressed in a bishop versus knight endgame, again with both players running short of time. He could have kept his bishop where it protceted the a-pawn. Now his problem is that Black has chances to create passed pawns on both wings.

At move 56 Black has to decide whether it is safe to win the bishop. Can he cope with the white passed d-pawn when his king is on holiday at h7 and his knight jaunts to the rim to eat the a-pawn? It seems that he can but best play is only a draw.

The second diagram shows the queen endgame that arose after White's 66th move. It should be drawn but the tablebase shows that three serious errors were committed (two by White, one by Black) before the game was decided.

The first two mistakes are far from obvious but White's final blunder, allowing the exchange of queens, was surprising. Tiredness and clock pressure presumnably explain.

 











(2) Donchenko,Alexander (2570) - Mista,Aleksander (2556) [D90]
Top 12 French Team Championship (10.1), 08.06.2015
[Harding,T.]

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.Bd2 Bg7 6.e4 Nxc3 7.Bxc3 0-0 8.Nf3 Bg4 9.Be2 c5 10.d5 Bxc3+ 11.bxc3 e6 12.0-0 exd5 13.exd5 Bxf3 14.Bxf3 Qd6 15.Rb1 b6 16.Be2 Nd7 17.Ba6 Nf6 18.c4 Rae8 19.Bb5 Re4 20.Rb3 Rc8 21.Re1 Rxe1+ 22.Qxe1 Rc7 23.Re3 Kf8 24.h3 Re7 25.f4 Nh5 26.g3 Rxe3 27.Qxe3 Qe7 28.Qd3 Qe1+ 29.Kg2 Nf6 30.Qf1 Qe4+ 31.Qf3 Qd4 32.Qf2 Qe4+ 33.Qf3 Qd4 34.Qf2 Ne4 35.Qxd4 cxd4 36.Kf3 f5 37.Ke2 Nxg3+ 38.Kd3 Ke7 39.Kxd4 Kd6 40.Be8 Nh5 41.Ke3 Kc5 42.Bf7 Nf6 43.Kd3 a6 44.a4 a5 45.h4 Ne4 46.Bg8 Nf6 47.Be6 Kd6 48.Kd4 Nh5 49.Ke3 Ke7 50.Bc8 Nf6 51.Kd4 Ne4 52.Be6 Nc5 53.Bg8?! h6 54.Bh7 Kf6 55.d6 Black had a think, time running low for both? 55...Kg7 The best practical chance. [55...Ke6 56.d7 (56.Bxg6? Nxa4 57.Bxf5+ Kxf5 58.Kd5 (58.d7 Nc5 59.d8N Ne6+ 60.Nxe6 Kxe6 wins for Black, as discussed by the French online commentators.) 58...Nc5 59.Kc6 Kxf4 seems to win for Black, e.g. 60.Kxb6 Ne6 61.c5 (61.Kxa5 and the 7-man tablebase shows that Black's h-pawn will ultimately win the game, starting with 61...Ke5 ) 61...Ke5 coming back to stop the pawns) 56...Kxd7 57.Bxg6 Ke6 58.Be8 looks drawish, as does 57...Nxa4 58 Bxf5+.] 56.Kd5 Kxh7 57.Kc6 g5 58.hxg5 hxg5 59.Kxb6 Nd7+ 60.Kc6 Nf8 61.c5 gxf4 62.d7 Nxd7 63.Kxd7 f3 64.c6 f2 65.c7 f1Q 66.c8Q The Probe tablebase says this should be a draw. 66...Qe1 67.Kd6[] Qg3+ 68.Ke7? [68.Kd5[] ] 68...Qe5+? [68...f4! mate in 44 says Probe] 69.Kf7[] Qg7+ 70.Ke6[] f4 71.Qa8?? Terrible blunder; now Black exchanges queens. [71.Kf5 or; 71.Kd5 can hold.] 71...Qg8+ And the f-pawn will queen. 0-1



All games on this page as PGN here

Generated with ChessBase 10
Download CBLight for free here