Belgrade Round Two

Nemanja Milanović (2102) - Spanton (1831)

34th Belgrade Trophy International Round 2

Sicilian Hyper-Accelerated Dragon

1.e4 c5 2.Nf3 g6 3.c3

White has four main plans against the Hyper-Accelerated Dragon - play a normal Open Sicilian with 3.d4 cxd4 4.Nxd4, keep the game in pure Hyper lines with 3.d4 cxd4 4.Qxd4, set up a Maróczy Bind, or build a classical centre with c3 and d4. This game starts as if it will be the fourth option, but then diverts.

3...Bg7

This is by far the most-popular move, but Capablanca as far back as 1928 played 3...d5, which has been recommended by Andrew Greet, and turns the game into a line of the Alapin Variation.

4.Bc4

Building a classical centre with 4.d4 is normal, but the text has also been played by grandmasters.

4...e6

The moves ...g6 and ...e6 often do not go well together as Black can be sensitive on the dark squares, but here it is perfectly reasonable as Black will gain a tempo with ...d5.

5.0-0 Nc6 6.Re1 Nge7 7.d3!?

White can still build a classical centre, and indeed 6.d4 is the commonest move in ChessBase's 2020 Mega database.

7...0-0 8.Nbd2

The game is looking like a King's Indian Attack, except the white light-square bishop has been developed at c4 instead of g2.

8...d5 9.Bb3 d4!?

Liviu-Dieter Nisipeanu (2611) - Claude Adrian (2360), French Team Championship 2000, went 9...b5 10.e5 a5 11.Bc2 Qc7 12.Qe2, which Stockfish12 and Komodo11.01 reckon is slightly better for Black (0-1, 58 moves).

10.cxd4 Nxd4 11.Nxd4

How should Black capture on d4?

*****

*****

*****

*****

*****

11...cxd4!?

Possibly better is 11...Qxd4 12.Nf3 Qd8, or 11...Bxd4 12.Nf3 Bg7, when White temporarily, at least, has a backward d pawn. It can be eliminated with 13.Be3 b6 14.d4, but Black is fine.

12.f4 e5?!

Now White gets enhanced pressure against f7. The engines reckon Black should play on the queenside, eg 12...Nc6 13.Nf3 a5, although they prefer White.

13.Nf3 Nf6 14.Bd5 Bg4?!

The engines prefer 14...exf4 15.Bxf4 Be6, which prevents the white light-square bishop from becoming dominant.

15.h3 Bxf3 16.Qxf3 Nb4 17.Bb3 Rc8 18.Qd1!?

Probably too passive. The engines prefer 18.Re2 or 18.Qe2.

18...Qh4

NM said afterwards that 18...Qc7 is better, but this is far from certain.

19.Rf1 Bh6! 20.a3

On 20.fxe5 I planned 20...Rxc1 21.Rxc1 Be3+ 22.Kh1 Bxc1 23.Qxc1 Nxd3, which the engines reckon is dead-equal.

20...Nc6 21.f5 Na5?

I rejected 21...Be3+ because of 22.Bxe3?, missing that after 22...dxe3 the threat of ...Nd4 is huge. Indeed, Black is winning, according to the engines, and therefore White needs to play 22.Kh1, after which the engines have difficulty evaluating the position, but reckon it is roughly level.

22.Bd5 Rxc1?

Better is 22...Be3+ 23.Bxe3 dxe3, but without the knight threatening to occupy d4 it is good for White.

23.Rxc1 Be3+ 24.Kh1 Bxc1

Or 24...Bf4 25.Qe1.

25.Qxc1

Black has no threats on the kingside and is lost on the other flank.

The game finished:

25...Kg7 26.f6+ Kg8 27.Qc7 Nc6 28.Bxc6 bxc6 29.Qxe5 1-0