Nunn, John DM (2605) - Shirov, Alexei (2690)
Bundesliga 9596 (Germany), 1996
[Shirov]

1.e4 g6 2.d4 Bg7 3.Nc3 c6 4.Bc4 d6 5.Qf3 e6 6.Nge2 b5 7.Bb3 a5 8.a3 Ba6 Despite my terrible loss against Judit Polgar in the Donner memorial (Amsterdam 1995) I still employ this setup. It's hard to believe that with such a bishop on b3 White should get the advantage.
9.O-O Judit played

[9.d5 and after 9...cxd5 10.exd5 e5 11.Ne4 Qc7 12.c4 bxc4 13.Ba4+ Nd7 14.N2c3 Ke7? 15.Nxd6! I got crashed with the incredible speed. Of course, this time I hoped to improve somewhere but Nunn decided not to test my preparation.]

9...Nd7 10.Bf4 Qe7 11.Rad1

[11.e5!? d5 12.Qe3 was quite interesting as it would create some difficulties for Black to develop.]

11...e5 12.Bg5

[12.dxe5 would be answered by
12...Nxe5! with the unclear game.]

12...Ngf6 13.d5?! A very dubios plan but I don't know what to recommend White.
13...c5 14.a4 b4 15.Nb5 Nb6!? As usual I am being quite ambitios. I saw that simple

[15...Bxb5 16.axb5 Nb6 would be very good but I wasn't sure that I would establish a clear advantage after
17.Qd3! with the idea to answer
17...a4 with 18.Bc4 . The text might objectively be even better but still it's not as strong as it seems. After White finds a few very good moves (most of which I of course hadn't seen) the game becomes rather sharp and complicated.]

16.Qd3! c4! 17.Bxc4 Nxa4 18.Ra1! This I overlooked.

[18.Rb1 Nc5 19.Bxf6 Bxf6 20.Qf3 O-O 21.Ra1 Rfc8! 22.c3 Qd8! would be more according to my plan.]

18...Nxb2 19.Qb3 Nxc4 20.Qxc4 Rc8 21.Qd3 Qd8 22.c4! O-O!? I was not at all sure about the position arising after

[22...bxc3 23.Nexc3 Rxc3

(23...O-O 24.Rfb1)

24.Qxc3 Bxb5 25.Bxf6 Bxf6 26.Rfb1 Bd7 27.Rxa5 O-O 28.Ra7! #. Black has a clear material advantage but his pieces don't feel extremely comfortable. Meantime White's activity yields him very good drawing chances.]

23.Ra2! Another surprise. Here I realised that if I choose now a 'normal' continuation, White would soon get full compensation for the pawn. But then I saw a rook sacrifise. It was difficult to convince myself to do it - just couple of moves ago I thought I had had a clear advantage and now I have to take such risks! Some minutes of doubts and I decided let it be. It's simply the best way.
23...Rxc4!! Variations like

[23...Bxb5 24.cxb5 Ra8 25.Rfa1 would just demonstrate White's positional power. But now Black's pawn are going to be really atrong.]

24.Bxf6

[24.Qxc4 Qb6 would finally lead to the same thing.]

24...Bxf6 25.Qxc4 Qb6 26.Qc7 Qxb5 27.Qxd6 Kg7

[27...Bd8!? would probably also keep Black's advantage and be less concrete. The text leads to positions where such a small thing as White's pawn on h2 and not on h3 will sometimes become a decisive factor.]

28.Rc1! Unprotected first rank leads to White's loss in the line

[28.Rfa1?! b3! 29.Rxa5? b2 30.Qxa6 b1=Q+]

28...Rd8! 29.Qc6

[29.Qc5 b3 30.Rb2 a4 looks clearly in Black's favour since the queen's exchange -
31.Qxb5 Bxb5 32.Ra1 Rc8 is catastrophic for White.]

29...b3! Not

[29...a4? since in the endgame arising after
30.Rxa4 Qxc6 31.dxc6 Bxe2 32.c7 Rc8 33.Rxb4 only White has chances to win.]

30.Rb2 # Here I got a feeling of horror as if something was completely misplayed. But it was definitely my day since I quickly found a move that justified all my previous play.
30...Qd3!! But not

[30...a4? that would fail to
31.Nc3 Qxc6 32.dxc6 .Now White's pieces are out of coordination, his back rank is still a shame on him and 31...Rc8 is threatening.]

31.Rcb1? Being short of time Nunn loses immediately. He had to try

[31.Rxb3! and it's not sure that in the game I would respond correctly. My home analysis convinced me that 31...Qxe2 would give White good practical drawing chances while 31...Qxb3! should maintain Black's edge intact. Here are the variations:
31...Qxb3!

(31...Qxe2 32.Rf3 Be7 (32...Bg5? 33.Qc7 Rf8 34.Qxe5+ f6 35.Rc7+ Kh8 36.h4 Qd1+ 37.Kh2 Qxf3 38.Rxh7+ Kxh7 39.Qc7+ Kg8 40.gxf3) 33.Qc7 Qb2 34.Re1! (34.h4 Rf8 (34...Rd7? 35.Qxd7 Qxc1+ 36.Kh2 Qc5 37.Qe8 (37.h5!?)) 35.Rcc3 (35.Rc6 Bd8 36.Qd6 Bb7 37.Rc4 Bxh4 38.Rc7 Bc8) 35...Bxh4 36.Qxe5+ Kg8) (34.h3 Rd7 35.Qxd7 Qxc1+ 36.Kh2 Qc5 37.Qe8 Bf6) (34.Rd1? Re8) 34...Re8 35.Qxa5 Bb5 36.Qa1 Qd4!)

32.Qxa6 a4

(32...Bg5 33.Qc4!)

33.Ng3! Be7!

(33...a3 34.Rc7 (34.Nf1? Be7 35.Rc7 Bd6 36.Rb7 Qc2 37.Qa7 Rf8) 34...a2 35.Ra7 Qb1+ 36.Nf1 Qxe4 37.Qxa2 Qxd5)

34.Rc7

(34.Qa7 Bd6 35.Qa5 Rb8 36.Qa6 Rb6)

34...Bd6 35.Rb7 Qd1+!

(35...Qc2 36.Qb6!! Bc5 37.Rxf7+! Kxf7 38.Qe6+ Kf8 (38...Kg7 39.Qxe5+) 39.Qf6+ Ke8 40.Qe6+ Be7 41.Qg8+ Kd7 (41...Bf8 42.Qe6+) 42.Qe6+ Kc7 43.Qxe7+)

36.Nf1 Qd4! In this position Black should definitely win due to his a-pawn but it still requires a good technik.

(36...a3 37.Qb6! Rb8 (37...a2? 38.Qa7!) 38.Rxb8 Bxb8 39.Qxb8 a2 40.Qxe5+ Kh6 41.Qf4+)]

31...Rc8 The rest is easy to understand.
32.Qa4

[32.Qd6 a4]

32...Rc2 33.Rxb3 Qxe2 34.Rf3 Bd3 35.Qd7 Bxe4 36.Rxf6 Kxf6 37.Qd6+ Kf5 38.Rf1 Bd3 39.Qd7+ Kf6 White resigns. If my analysis is correct then it's one of my best games so far. The complications starting with 15...Nb6!? went in right direction.

0-1

Created with PGNtoJS

1.e4 g6 2.d4 Bg7 3.Nc3 c6 4.Bc4 d6 5.Qf3 e6 6.Nge2 b5 7.Bb3 a5 8.a3 Ba6 Despite my terrible loss against Judit Polgar in the Donner memorial (Amsterdam 1995) I still employ this setup. It's hard to believe that with such a bishop on b3 White should get the advantage. 9.O-O Judit played
	[9.d5 and after 9...cxd5 10.exd5 e5 11.Ne4 Qc7 12.c4 bxc4 13.Ba4+ Nd7 14.N2c3 Ke7? 15.Nxd6! I got crashed with the incredible speed. Of course, this time I hoped to improve somewhere but Nunn decided not to test my preparation.]
9...Nd7 10.Bf4 Qe7 11.Rad1
	[11.e5!? d5 12.Qe3 was quite interesting as it would create some difficulties for Black to develop.]
11...e5 12.Bg5
	[12.dxe5 would be answered by 12...Nxe5! with the unclear game.]
12...Ngf6 13.d5?! A very dubios plan but I don't know what to recommend White. 13...c5 14.a4 b4 15.Nb5 Nb6!? As usual I am being quite ambitios. I saw that simple
	[15...Bxb5 16.axb5 Nb6 would be very good but I wasn't sure that I would establish a clear advantage after 17.Qd3! with the idea to answer 17...a4 with 18.Bc4 . The text might objectively be even better but still it's not as strong as it seems. After White finds a few very good moves (most of which I of course hadn't seen) the game becomes rather sharp and complicated.]
16.Qd3! c4! 17.Bxc4 Nxa4 18.Ra1! This I overlooked.
	[18.Rb1 Nc5 19.Bxf6 Bxf6 20.Qf3 O-O 21.Ra1 Rfc8! 22.c3 Qd8! would be more according to my plan.]
18...Nxb2 19.Qb3 Nxc4 20.Qxc4 Rc8 21.Qd3 Qd8 22.c4! O-O!? I was not at all sure about the position arising after
	[22...bxc3 23.Nexc3 Rxc3
		(23...O-O 24.Rfb1)
	24.Qxc3 Bxb5 25.Bxf6 Bxf6 26.Rfb1 Bd7 27.Rxa5 O-O 28.Ra7! #. Black has a clear material advantage but his pieces don't feel extremely comfortable. Meantime White's activity yields him very good drawing chances.]
23.Ra2! Another surprise. Here I realised that if I choose now a 'normal' continuation, White would soon get full compensation for the pawn. But then I saw a rook sacrifise. It was difficult to convince myself to do it - just couple of moves ago I thought I had had a clear advantage and now I have to take such risks! Some minutes of doubts and I decided let it be. It's simply the best way. 23...Rxc4!! Variations like
	[23...Bxb5 24.cxb5 Ra8 25.Rfa1 would just demonstrate White's positional power. But now Black's pawn are going to be really atrong.]
24.Bxf6
	[24.Qxc4 Qb6 would finally lead to the same thing.]
24...Bxf6 25.Qxc4 Qb6 26.Qc7 Qxb5 27.Qxd6 Kg7
	[27...Bd8!? would probably also keep Black's advantage and be less concrete. The text leads to positions where such a small thing as White's pawn on h2 and not on h3 will sometimes become a decisive factor.]
28.Rc1! Unprotected first rank leads to White's loss in the line
	[28.Rfa1?! b3! 29.Rxa5? b2 30.Qxa6 b1=Q+]
28...Rd8! 29.Qc6
	[29.Qc5 b3 30.Rb2 a4 looks clearly in Black's favour since the queen's exchange - 31.Qxb5 Bxb5 32.Ra1 Rc8 is catastrophic for White.]
29...b3! Not
	[29...a4? since in the endgame arising after 30.Rxa4 Qxc6 31.dxc6 Bxe2 32.c7 Rc8 33.Rxb4 only White has chances to win.]
30.Rb2 # Here I got a feeling of horror as if something was completely misplayed. But it was definitely my day since I quickly found a move that justified all my previous play. 30...Qd3!! But not
	[30...a4? that would fail to 31.Nc3 Qxc6 32.dxc6 .Now White's pieces are out of coordination, his back rank is still a shame on him and 31...Rc8 is threatening.]
31.Rcb1? Being short of time Nunn loses immediately. He had to try
	[31.Rxb3! and it's not sure that in the game I would respond correctly. My home analysis convinced me that 31...Qxe2 would give White good practical drawing chances while 31...Qxb3! should maintain Black's edge intact. Here are the variations: 31...Qxb3!
		(31...Qxe2 32.Rf3 Be7 (32...Bg5? 33.Qc7 Rf8 34.Qxe5+ f6 35.Rc7+ Kh8 36.h4 Qd1+ 37.Kh2 Qxf3 38.Rxh7+ Kxh7 39.Qc7+ Kg8 40.gxf3) 33.Qc7 Qb2 34.Re1! (34.h4 Rf8 (34...Rd7? 35.Qxd7 Qxc1+ 36.Kh2 Qc5 37.Qe8 (37.h5!?)) 35.Rcc3 (35.Rc6 Bd8 36.Qd6 Bb7 37.Rc4 Bxh4 38.Rc7 Bc8) 35...Bxh4 36.Qxe5+ Kg8) (34.h3 Rd7 35.Qxd7 Qxc1+ 36.Kh2 Qc5 37.Qe8 Bf6) (34.Rd1? Re8) 34...Re8 35.Qxa5 Bb5 36.Qa1 Qd4!)
	32.Qxa6 a4
		(32...Bg5 33.Qc4!)
	33.Ng3! Be7!
		(33...a3 34.Rc7 (34.Nf1? Be7 35.Rc7 Bd6 36.Rb7 Qc2 37.Qa7 Rf8) 34...a2 35.Ra7 Qb1+ 36.Nf1 Qxe4 37.Qxa2 Qxd5)
	34.Rc7
		(34.Qa7 Bd6 35.Qa5 Rb8 36.Qa6 Rb6)
	34...Bd6 35.Rb7 Qd1+!
		(35...Qc2 36.Qb6!! Bc5 37.Rxf7+! Kxf7 38.Qe6+ Kf8 (38...Kg7 39.Qxe5+) 39.Qf6+ Ke8 40.Qe6+ Be7 41.Qg8+ Kd7 (41...Bf8 42.Qe6+) 42.Qe6+ Kc7 43.Qxe7+)
	36.Nf1 Qd4! In this position Black should definitely win due to his a-pawn but it still requires a good technik.
		(36...a3 37.Qb6! Rb8 (37...a2? 38.Qa7!) 38.Rxb8 Bxb8 39.Qxb8 a2 40.Qxe5+ Kh6 41.Qf4+)]
31...Rc8 The rest is easy to understand. 32.Qa4
	[32.Qd6 a4]
32...Rc2 33.Rxb3 Qxe2 34.Rf3 Bd3 35.Qd7 Bxe4 36.Rxf6 Kxf6 37.Qd6+ Kf5 38.Rf1 Bd3 39.Qd7+ Kf6 White resigns. If my analysis is correct then it's one of my best games so far. The complications starting with 15...Nb6!? went in right direction. 0-1