Wednesday, 23 October 2019

Guernsey Round Four

Pal Magnussen (1983) - Spanton (1927)
Accelerated Dragon
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 5.Nxc6!?
Other moves - 5.Nc3, 5.c4, 5.Be3 - are more popular, but there are almost 2,100 examples of the text in ChessBase's 2019 Mega database.
5...bxc6 6.Qd4 Nf6 7.e5 Nd5 8.e6 f6 9.c4 Nb4 10.Qc3 Na6
10...a5 was played in James Hart - Jeremy Silman, Flagstaff (Arizona), 1991, and is Stockfish10's choice. 10...c5 is at first preferred by Komodo10, but it later switches to 10...a5.
11.exd7+ Qxd7 12.Be3 e5 13.a3 Bc5 14.Be2 Bd4 15.Qc2 Rb8 16.Nc3 Bxe3?
The wrong idea. Yes, White gets an isolated e pawn, but it is not worth giving up such a strong bishop, especially as the isolani is not easy to get at and White might be able to use the half-open f file. The engines give 16...Nc5 17.0-0 Kf7!? with a slight edge for Black.
17.fxe3 Nc5?!
White now gets a strong initiative and ends up taking over the d6 square. Better is 17...Nc7, and if, as in the game, 18.b4, Black can counterattack on the queenside with 18...a5!?, or simply castle and be ready to meet any White attempt to take over d6. However, White would still be better.
18.b4 Ne6 19.Rd1 Qe7 20.Bf3 Bd7 21.Ne4?!
Going after the d6 square, but even stronger seems to be 21.Qa4, hitting both black isolated pawns.
21...0-0 22.c5 a5 23.Qd2
Black to make his 23rd move
*****
*****
*****
*****
*****
23...Rfd8
I strongly considered 23...axb4!?, which does eventually become the engines' top choice. The idea is to meet 24.Qxd7 with 24...Qxd7 25.Rxd7 bxa3. The engines reckon White is better, but the position is not clear-cut. I just was not sure, and there was also the practical point that White would have the choice of taking the bishop, if he thought the sac did not work, or of declining it with 24.axb4, although then the position is equal, according to the engines.
24.0-0 f5 25.Nd6 axb4 26.axb4 Ng5 27.Be2
27.Qc3!? Nxf3+ 28.gxf3 leaves White with an unchallengeable knight, but it is not clear if White can effectively support it.
27...Be6 28.Qc3 Bd5?!
The engines prefer 28...Nf7 29.Nxf7 Bxf7 30.Rxd8+ Rxd8 31.b5 cxb5 32.Bxb5, when the white passed pawn is dangerous but probably manageable.
29.Bc4?!
Now the pawn push b5 is even more dangerous as Black cannot capture because the d5 bishop would hang.
29...Bxc4 30.Qxc4+ Qe6
White has a very strong continuation
*****
*****
*****
*****
*****
31.Qc3
Stronger is the engines' 31.Qa6!! The main point is that 31...Rxb4 runs into 32.h4! (the immediate 32.Nxf5 is also good) Rxh4 33.Nxf5!, when Black is losing material, or even getting mated, eg 33...Rxd1 34.Qa8+ Kf7 35.Nd6+ Kg7 36.Qf8#. Black can go for counterplay with 31...Qb3, when one line runs 32.Rb1!? Qxe3+ 33.Kh1 Nf7 34.Nxf7 Kxf7 35.Qxc6, but White's connected passed pawns are clearly more threatening than Black's 4-2 kingside majority. Stockfish10 eventually gives 31...Nf7!? 32.Qxc6 Nxd6 33.Rxd6 Rxd6 34.cxd6, but White is a pawn up and has two dangerous passed pawns, so it is hard to see the game continuing much longer.
31...Nf7 32.Nc4 e4 33.h3 Rxd1?!
Almost certainly better was 33...h5 as the text gives White an initiative.
34.Rxd1 Rd8 35.Rxd8+ Nxd8 36.Qd4
Stockfish10's 36.Nd6! looks strong, eg 36...Nf7 37.b5! cxb5 (37...Nxd6 38.cxd6 Qxd6 39.b6! wins for White - in queen-and-pawn endings a far advanced passed pawn is often more important than an extra pawn) 38.Nxb5, with a difficult defence ahead for Black.
36...Nf7 37.Nd6
Too late - the extra tempi Qd4 and …Nf7 favour Black.
37...Ne5 38.b5 cxb5 39.Nxb5 Nc6 40.Qa4 Kg7 41.Kf1 Qd5 42.Nd4 Nxd4 43.Qxd4+
Or 43.exd4 e3 44.c6 Qe4 45.Qb4 Qf4+ when eventually one player or the other has to take a draw by perpetual check.
43...Qxd4 44.exd4
PM had thought his pair of passed pawns would win, but he forgot Black too has a protected passed pawn.
44...Kf6 45.c6 ½–½

No comments:

Post a Comment