FACED a German yesterday afternoon in round three of the Bodensee Open at Bregenz, Austria.
 |
| Approaching Bregenz Oberstadt (upper town) |
 |
| The other side of the gateway |
Spanton (1860) - Marius Kaiser (2046)
French Exchange
1.e4 e6 2.d4 d5 3.exd5
My 24th attempt to play against the French Defence in the style of Paul Morphy.
3...exd5 4.Nf3 Nf6 5.Bd3 c5!?
This for quite some time is Stockfish14.1's top choice, along with the strange-looking 5...Bb4+!? 6.c3 Bd6. Komodo12.1.1 prefers the mainline 5...Bd6, and, given enough time, Stockfish14.1 agrees.
6.0-0 Be7
The commonest continuation in ChessBase's 2022 Mega database is 6...c4!? 7.Re1+ Be7 8.Bf1 0-0 9.b3 cxb3 10.axb3 Nc6 with what the engines reckon is an equal position.
7.Re1 0-0 8.dxc5!? Bxc5 9.Be3 Qb6
This may be a novelty. The engines prefer 9...Bxe3. But note 9...d4? fails to 10.Bxd4 Bxd4 11.Nxd4 as 11...Qxd4?? loses to 12.Bxh7+ etc.
10.Bxc5 Qxc5 11.h3!?
Denying the g4 square to both the black queen's bishop and king's knight.
11...Nc6 12.c3 Be6 13.Nbd2 Rad8 14.Nb3 Qb6 15.Qc2
I missed that 15.Nbd4 is possible as 15...Qxb2? loses to 16.Nb5, trapping the black queen.
15...Rdc8
Threatening 16...Nb4.
16.a3
Probably better is the engines' 16.Ng5!?, one point being 16...Nb4? gets hit by 17.Bxh7+. Best, according to the engines, is 16...h6 17.Nxe6 Nb4 18.Qd2 Nxd3 19.Qxd3 fxe6, when Black's hanging central pawns mean White has the advantage.
16...Bd7!?
An interesting way of meeting the threat of 17.Ng5.
17.Rad1
Not 17.Ne5?? Nxe5 18.Rxe5 Ba4 etc.
17...Rfe8 18.Nbd4 g6 19.Rxe8+ Rxe8 20.Nxc6!?
Transforming the pawn-structure, something that often happens in an IQP game.
20...bxc6 21.c4 Be6?!
The engines prefer 21...dxc4 22.Bxc4 Be6, but give White a slight edge.
 |
| How should White proceed? |
*****
*****
*****
*****
22.c5!
The engines agree this is best, even though Black gets a protected passed pawn. One of the main ideas is to keep Black's light-square bishop bad.
22...Qc7 23.Nd4 Bd7 24.Qd2 Qe5 25.Nf3 Qh5 26.Qg5?!
White is on top after the consolidating 26.b4!?, eg 26...Bxh3 27.Qf4 Bf5 (indirectly defending the black knight) 28.Bxf5 Qxf5 29.Qxf5 gxf5 30.Nd4.
26...Qxg5 27.Nxg5 a5 28.Kf1 Rb8 29.Rd2!?
Probably slightly better is 29.Rb1.
29...Nh5 30.Nf3
Not 30.g3? h6 31.Nf3 Bxh3+.
30...Nf4 31.Ne5 Be8 32.Be2
Keeping guard over the b5 square.
32...Ne6 33.Nd3 Bd7 34.b4 axb4 35.axb4 Ra8 36.Ne5 Be8 37.h4 f6 38.Nd3 h6 39.g3 Bd7 40.h5 g5
 |
| What should White play? |
*****
*****
*****
*****
41.Bg4?
This falls to a tactical trick. The engines reckon 41.Ne1!? is equal.
41...Nxc5! 42.Bxd7 Nxd7 43.Rc2 Nb8?
Unnecessary passivity. Correct is 43...Ra6. Black need not fear 44.b5?! as 44...cxb5 45.Rc8+ Kf7 46.Rc7 Ke6 (not 46...Ke7? 47.Rxd7+ Kxd7 48.Nc5+ etc) 47.Rxd7?? loses to 47...Ra1+ etc.
44.Ke2 Kf7 45.Ke3 Ke7 46.f4?!
Aiming to free the e5 square for a white piece, but it is almost certainly too slow. Instead 46.Kd4 leaves Black only slightly better, according to the engines.
46...Ra3!?
The engines much prefer 46...Kd6.
47.fxg5 fxg5 48.Kd4 Nd7 49.g4?
Not 49.Rxc6? Rxd3+! 50.Kxd3 Ne5+ etc, but much better, as Ray Kearsley pointed out when we went over the game, is 49.Re2+, when 49...Kd6 50.Re8 gives White strong counterplay against the black kingside.
49...Kd6 50.Rf2 Rb3?
Better is the engines' 50...Ra4, threatening 51...c5+. White probably needs to play 51.Kc3, but then 51...Ra1 is strong.
51.Rf1 Ra3 52.Rf2?
52.Nc5 gives excellent counterplay, eg 52...Rg3 53.Nxd7 Rxg4+ 54.Kc3 Kxd7 55.Rf6.
52...Ra8?
Still very strong is ...Ra4.
53.Rf7 Rf8 54.Ne5!
This completely equalises, and gives Black a chance to go wrong.
54...Rd8?
Drawing is 54...c5+ 55.bxc5+ Nxc5! 56.Rxf8 Ne6+; also 54...Rxf7 55.Nxf7+ Ke6 56.Nxh6 Ne5; finally the engines reckon 54...Ke6!? 55.Rxf8 Nxf8 56.Nxc6 also draws despite the white outside passed pawn.
55.Nxd7
Only drawing is 55.Rh7? Nxe5 56.Rxh6+ Kc7 57.Kxe5 d4.
55...Rxd7 56.Rf6+ Kc7 57.Rxh6 Rf7 58.Kc5 Rd7 59.Rxc6+ Kb7 60.Rd6?
Throwing away the win, which is to be had by, for example, 60.h6 d4 61.Rg6 d3 62.Rg7 Rxg7 63.hxg7 d2 64.g8=Q d1=Q 65.Qd5+. The text leads to a similar line, except for the key fact that Black promotes with check.
60...Rxd6 61.Kxd6 d4 62.h6 d3 63.h7 d2 64.h8=Q d1=Q+ 65.Kc5 Qxg4 66.Qg7+ Kc8 67.Qf8+ Kd7 68.Qd6+ Ke8 69.b5 Qg1+ 70.Kc6
 |
| So far, so correct, according to the Syzygy endgame tablebase, but here Black has only one move that draws |
*****
*****
*****
*****
70...Qc1+?
The tablebase shows 70...Qh1+ draws.
71.Kb7 Qf4 72.Qc6+
72.Qxf4? only draws.
72...Ke7
 |
| White has one winning move |
*****
*****
*****
*****
73.b6?
73.Qc5+ wins.
*****
*****
*****
*****
*****
73...Qf5?
Black has three drawing moves: 73...Qb4, 73...Qd4 and 73...Qf2.
74.Qc7+ Kf6 75.Ka7 Qa5+ 76.Kb8 g4 77.b7 Qa4 78.Kc8 Qa6 79.Qf4+ Ke7 80.Qg5+ Kf7 81.Qh5+ Ke7 82.Qh7+ Kf8 83.Qf5+ Ke7 84.Qd7+ Kf8 85.Kb8 g3 86.Qf5+ Ke7 87.Qg5+ Qf6 88.Qxg3 Qh8+ 89.Kc7
74.Qc7+ Kf6 75.Ka7 Qa5+ 76.Kb8 g4 77.b7 Qa4 78.Kc8 Qa6 79.Qf4+ Ke7 80.Qg5+ Kf7 81.Qh5+ Ke7 82.Qh7+ Kf8 83.Qf5+ Ke7 84.Qd7+ Kf8 85.Kb8 g3 86.Qf5+ Ke7 87.Qg5+ Qf6 88.Qxg3 Qh8+ 89.Kc7
89.Ka7? only draws.
89...Qd8+ 90.Kc6 Qd7+ 91.Kc5 Qf5+
Not 91...Qxb7?? 92.Qg7+ etc.
92.Kd4 Qd7+ 93.Ke3 Qe6+ 94.Kf2 Qa2+ 95.Kg1 Qb1+ 96.Kh2 Qh7+
 |
| White to play and win |
*****
*****
*****
*****
97.Qh3?
Winning are 97.Kg2 and 97.Kg1, although the lines are very long (but within the 50-move drawing limit).
It is long gone midnight here in Austria and I need to get some sleep, so I fear I have not gone into any depth in analysing this queen-and-pawn ending.
The game finished:
97...Qc2+ 98.Qg2 Qh7+ 99.Kg3 Qg6+ 100.Kf2 Qc2+ 101.Kg1 Qb1+ 102.Kf2 Qc2+ 103.Kg3 Qg6+ ½–½
ADDENDUM: the last capture occurred on move 88, so the 50-move rule would count from there. The Syzygy tablebase includes what are called "cursed" wins, ie positions in which one side would win eventually, but which in a game would be drawn because of the 50-move rule.
No comments:
Post a Comment