English
1.Nf3 Nc6!? 2.c4 e5 3.Nc3 g6 4.g3 Bg7 5.Bg2 d6 6.0-0 Nge7 7.a3 a5
 |
| This might seem like a fairly standard position from the English, but it occurs only five times in ChessBase's 2023 Mega database |
*****
*****
*****
*****
8.d4!?
IK afterwards said he was following a game of Tony Miles', but he almost certainly got his positions mixed. Stockfish16 and Komodo14.1 suggest 8.Rb1, 8.d3 and 8.e3, all of which have been played.
8...Nxd4 9.Nxd4 exd4 10.Nb5 Nc6 11.e3!?
Offering to make the pawn-sacrifice permanent.
The engines suggest 11.Bg5!?, when 11...f6 weakens the black kingside, while 11...Qd7 leads to complicated play, eg 12.Bh3!? Qxh3!? (12...f5 is possible, but weakening) 12.Nxc7+ Kf8 13.Nxa8 h5!? with a very unclear position.
 |
| How should Black respond to White's offer? |
*****
*****
*****
*****
11...d3!?
Returning the pawn but making it difficult for White to develop his queen's bishop.
11...dxe3 12.Bxe3 0-0 (almost certainly not 12...Bxb2?! 13.Bh6) 13.c5 gives White promising compensation, but the engines reckon is playable.
12.Qxd3 0-0 13.Rd1?!
This looks natural but fails to take into account the weakness of the c4 pawn. The engines suggest 13.f4 or 13.Qc2, although Stockfish16 gives Black a slight edge after either move.
13...Ne5 14.Qe2
The engines prefer 14.Qe4 or 14.Qc2.
14...Be6
Even stronger, according to the engines, is 14...Bg4 15.f3 Be6.
15.Rd4 Nc6 16.Rd2 Ne5 17.Bd5?!
If 17.Rd4 I was considering playing on with a pawn offer of my own, namely 17...Qe7!? The engines reckon Black has more than enough compensation after, for example, 18.Bxb7 Rxb8 19.Bg2 c6 20.Nc3 Rfc8. Another try is 17...c6!?, when both 18.Nxd6 and 18.Rxd6 can be met by 18...Qe7.
The engines reckon best is the passive 17.Rc2, but they prefer Black.
17...c6 18.Bxe6 fxe6 19.f4
White can save the knight and cover the f3 square with 19.Nd4, but the engines reckon both 19...Qf6 and 19...Qe7 leave Black well on top.
19...cxb5 20.fxe5 Bxe5 21.cxb5 Qg5
 |
| What should White play? |
*****
*****
*****
*****
22.e4?
Missing the main threat contained in my last move. The engines suggest covering the back rank with 22.Rd1.
22...Bxg3! 23.Kh1?!
Not 23.hxg3? Qxg3+ 24.Qg2 Qe1+ 25.Kh2 Rf4 etc, nor 23.Qg2? Qe3+ 24.Kh1 Qe1+ etc, but probably best is 23.Rd1 Bf4+ 24.Kd1 Qh4, although Black emerges a pawn up and with much the safer king.
23...Rf2 24.Qxf2 Bxf2 25.Rxf2 Qxb5
 |
| How would you assess this late middlegame? |
*****
*****
*****
*****
White's extra rook and bishop are no match for Black's queen and two pawns. The engines reckon Black is the equivalent of more than a rook ahead.
26.Bh6 Qh5 27.Bd2 Qe5
Somewhat greedily going after another pawn. Much simpler is 27...Rf8.
28.Re1 Qxb2 29.Ref1 Qxa3 30.Bh6
 |
| White threatens mate in two - there is only one defence, but it is enough |
*****
*****
*****
*****
30...d5 31.exd5 exd5 32.Rf7 Re8 33.Rg7+ Kg8 34.Rgf7 Qb4 35.Rf8+!?
The bishop-v-pawns ending arising after this is lost for White, but perhaps it is the best try.
35...Rxf8 36.Rxf8+ Qxf8 37.Bxf8 Kg8 38.Ba3 Kf7 39.Kg2 Ke6 40.Kf3 Kf5 41.Bc5 b5 42.h3 g5 43.Ke3 h5 44.Kf3 b4 45.Ke3 g4 46.hxg4+ hxg4 47.Bd6
 |
| The bishop cannot cope with four pawns - indeed, of all Black's possible moves, only 47...g3?? fails to win |
*****
*****
*****
*****
47...b3
Bizarrely, the top move given by the Syzygy endgame tablebase is 47...a4!?, and if 48.Bxb4 then an only-move, 48...g3 wins, but I doubt if any human would play that way.
The game finished:
48.Ba3 g3 49.Kf3 g2 50.Kxg2 Ke4 51.Kf2 Kd3 52.Kf3 Kc3 53.Ke3 d4+ 54.Ke2 Kc2 55.Ke1 d3 0-1
No comments:
Post a Comment