King's Gambit
1.e4 e5 2.f4 Bc5 3.Nf3 d5!?
A pet line of mine. There are 110 examples of the move in ChessBase's 2026 Mega database - 10 of the games are mine.
 |
| How should White proceed? |
*****
*****
*****
*****
4.exd5
This is the commonest continuation in Mega26, and is the second choice of Stockfish18 and Dragon1. Stockfish18 prefers the relatively rare 4.Nc3!?; Dragon1 the more popular 4.Nxe5.
4...e4 5.d4!?
5.Qe2 and especially 5.Ne5 are more common, but the engines agree the text is best.
 |
| How should Black respond? |
*****
*****
*****
*****
5...exd3
Dragon1 reckons the text and 5...exf3 have an almost identical value, but Stockfish18 suggests an apparent-novelty in 5...Bb4+!?, eg 6.Bd2!? Bxd2+ 7.Nfxd2 Nf6 8.Nc3 0-0, when the engines agree Black has enough for a pawn.
6.Qxd3!?
The engines are unsure whether to prefer this move, which is not in Mega26, or the known 6.Bd3, but given enough time, they come down in favour of the latter, reckoning Black should play 6...Ne7, when they find it hard to decide whether Black has enough for a pawn.
6...Nf6!?
Although 6.Qxd3!? is not in Mega26, the move transposes into a known position, where 6...Ne7 has been played. However, in this case the engines prefer the text.
 |
| What should White play? |
*****
*****
*****
*****
7.Nc3!?
The engines' main line runs 7.Qe2+ Qe7 8.Bd2!? (they also like 8.Nc3) Qxe2+ 9.Bxe2 Nxd5 10.Nc3 Nxc3 11.Bxc3 Nc6, when Stockfish18 calls the position equal, but Dragon1 fluctuates between equal and slightly favouring White. If 8.Qxe7+, then 8...Kxe7, with ...Re8 to come.
7...0-0 8.Bd2 Ng4?
I was disappointed to discover after the game I made exactly the same mistake at Hastings 1991. Better is 8...Re8+, with equality, according to the engines, although the game remains tense, eg 9.Be2 c6!? 10.0-0-0 (10.dxc6 Qxd3 11.cxd3 Nxc6 12.0-0-0 Bf5, with good play for Black) cxd5 11.Ne5 Nc6, with what they reckon are equal chances.
 |
| White to play and gain what the engines agree is a winning position |
*****
*****
*****
*****
9.Ne4??
This turns a winning position into a losing one. Instead, 9.Ng5 seems to flummox Black in all lines, eg 9...Re8+ 10.g6 Nce4, or 9...g6 10.Nce4.
 |
| After the immediate 9.Ne4??, White has defended f2 and attacks c5, but there is a fatal flaw |
*****
*****
*****
*****
9...Bf5!?
Seeing this was one reason why I mistakenly liked 9.Ne4??
10.0-0-0
Perhaps slightly better is 10.Ng5, but 10...Nf2 11.Qc4 Re8 12.Qxc5 Nxe4 13.Nxe4 Rxe4+ gives Black strong play for a pawn, eg 14.Kf2 (14.Be2? Bg4) Nd7 15.Qc3 Nf6 16.h3 Nxd5, when Black has got the pawn back, and continues to have play against White's king and uncoordinated pieces.
10...Nf2 11.Qb5 Nxe4 12.Qxb7 Nd7
 |
| White has two pawns for a knight, but that is rarely sufficient in an early-middlegame |
*****
*****
*****
*****
13.Bd3 Rb8 14.Qa6 Rb6 15.Qc4 Nxd2 16.Nxd2 Qf6 17.c3 Rfb8
 |
| As well as being ahead on material, Black is fully developed and has the safer king - advantages adding up to the equivalent of being roughly a rook ahead, according to the engines |
*****
*****
*****
*****
18.Nb3 Be3+ 19.Kb1 Bxd3 20.Qxd3 Bc5!?
The engines reckon capturing on f4 is playable, but I saw no need to give Black counterplay for the sake of a pawn.
21.Rhf1 Bd6 22.g3?!
Missing Black's next move. The engines suggest 22.Ka1 or 22.Rd2, but still have Black winning.
22...Nc5 23.Qd4 Qxd4 0-1
No comments:
Post a Comment