SOS - Secrets of Opening Surprises - April 2013

No time to study opening theory? Shock your opponent with an SOS!
With an SOS you deviate early (usually before move 6!) from regular lines in mainstream openings. So you will reach positions you have actually studied without having memorized tons of stuffy theory, while gaining time on the clock! And you will have fun watching the horror on your opponent's face... 

SEE FOR YOURSELF HOW AN SOS CAN SHOCK AND CONFUSE!
Every month, the editor of the SOS Secrets of Opening Surprises book series, IM Jeroen Bosch, annotates a game which was recently played with an SOS-variation.

SOS Game of the Month: April 2013
See also SOS–3, Chapter 3, page 28
The Candidates Tournament in London has just finished, and what a great success it has been. Lots of high-level chess, excitement and some real drama in the final rounds to top it off. It seems strange perhaps to concentrate on a relatively short draw, but Alexander Grischuk deserves all the praise for taking on Magnus Carlsen with 5.h4!? in a Grünfeld. Our interest was clearly awoken by this belated SOS-thrust.

Here's the complete text of this game analysis:

Grünfeld D90 
Alexander Grischuk
Magnus Carlsen
London Candidates 2013

1.d4 Nf6 2.c4 g6 3.Nc3 d5

4.Nf3

4.h4 was the line I wrote on originally for SOS-3. Since then a lot of games (hundreds) have been played with this Grünfeld flank attack. I quite liked the fact that the natural 4...Bg7 is met by 5.h5 Nxh5 6.cxd5 assuming control in the centre via the pushing of the h-pawn. What put me off the line is 4...c5 5.dxc5 d4, as I reported in the SOS Files of SOS-12 (see also the SOS Files in SOS-10).

4...Bg7

5.h4!?

The main idea of Grischuk's move order is that ...c5 is by no means as strong as it is after 4. h4. After 5... c5 6. dxc5 White controls square d4 so ...d4 is not on, as it is after 4. h4 c5 5. dxc5 d4.
Note that 5.Bf4 c5 6.dxc5 Qa5 7.cxd5 Nxd5 8.Qxd5 Bxc3+ 9.Bd2 is a well-known theoretical line. Below we could end up in a similar position but with the h2-pawn on h4.

This favours White. Take a note of 9...Bxd2+ (9...Be6) 10.Qxd2 Qxc5 11.Rc1 Qf5, and now Petrosian actually played 12.h4!? (not the strongest move in this position perhaps, but interesting to compare with our line) 12...Nc6 13.h5, and White won in a training game T.Petrosian-Shamkovich, Moscow 1966.

5...c6

Carlsen admitted at the press conference that he could not remember anything on the sharper lines, and therefore chose a solid response. 5...c5 6.dxc5 and now Black has to make 6...Qa5 work. However, it seems that this favours White in all lines following 7.cxd5!:

- 7...Ne4 8.Bd2 Nxc3 9.Bxc3 Bxc3+ 10.bxc3 Qxc3+ 11.Nd2 Na6 12.Rc1 Qa5 13.Qc2 0-0 14.h5 Bd7 15.hxg6 fxg6 16.Qc3 (16.e4! Rac8 17.c6 bxc6 18.d6

and 19.Qc4+ is a nasty threat) 16...Qxc3 17.Rxc3 Rac8 18.Ne4 Nb4 (18...Bf5) 19.d6 exd6 20.Nxd6, White better, Simkin-Gerasimchuk, Kiev 1998. 
- 7...0-0 8.Bd2 Qxc5 9.Qb3 e6 10.h5 exd5 11.hxg6 hxg6 12.Rc1 Nc6 13.e3 Re8 14.Nb5 Qe7 15.Bc3 Bf5 16.Be2 Rad8 17.Nbd4 Bc8 18.Qa4 Ne4 19.Nxc6 Bxc3+ 20.Rxc3 bxc6 21.Qd4 f6 22.Rxc6, with a clear plus in Manfredi-Andreoni, Milan 2012.
- 7...Nxd5 8.Qxd5 Bxc3+ 9.Bd2 Bxd2+ 10.Qxd2 favours White, as I argued above.

5...h5 looks wrong, and must give White an opening edge. For example 6.Bg5 c6 7.e3 (7.cxd5 Nxd5 8.Qb3 Qa5 9.Bd2 0-0 10.e4 Nxc3 11.bxc3 (11.Bxc3) 11...Nd7 12.e5, and White is slightly better, Meyer-Säring, Germany 2006/07) 7...0-0 8.Rc1 Be6 9.cxd5 Nxd5 10.Qd2, and White is slightly better, Kadas-Bakos, Hungary 2005/06.

5...0-0 has to be investigated of course:
- 6.h5 Nxh5 7.cxd5 is actually less strong for White than 4. h4 Bg7 5. h5 Nxh5 6. cxd5, because here 6... c6 is answered by 7. e4! cxd5 8. e5, when because of the threat 9. g4 Black saw nothing better in Shliperman-Ady, New York 1999, but to play the sad 8...Bf8. Here after 7...c6

White played 8.Qb3 (8.e4 cxd5 9.e5 is no good because g4 is not a threat; 8.dxc6 Nxc6 9.e3 is probably about equal) 8...cxd5 9.Bg5 (9.Qxd5!?) 9...Nc6 10.e3 Be6 11.Qxb7 in Aderito-Makoto, Windhoek 2007.
- Instead, 6.cxd5 Nxd5 7.h5!? c5 8.hxg6 hxg6 9.e4 looks like a very interesting try to me.

5...dxc4 6.e4 Bg4 (6...c5!? 7.d5 b5 is unclear but worth studying) 7.Bxc4 0-0 8.e5 Nfd7 9.h5 gxh5 10.Qd3 h6 11.Qe4

was a lot better for White in Morozevich-Giri, Beijing blitz 2012 (while 11.Bxh6! was actually winning on the spot).

6.Bg5

This is not new, but Grischuk's try certainly looks more entertaining than the exchange Slav way of playing it, although White may hope for a very minor edge after 6.cxd5.
6.cxd5 turning the game into a kind of exchange Slav (where the dark fianchetto bishop is somewhat misplaced) is played more often: 6...cxd5 7.Bf4 Nc6 8.e3 0-0, and now:

- 9.Be2 Bg4 (9...Qb6 10.Qd2 Bf5 11.Ne5 Nxe5 12.Bxe5 Ne4 13.Nxe4 Bxe5 is given as 0-1, Kadas-Szeberenyi, Nyiregyhaza 2004, in the database. At present White seems to be better though after 14.Nc3 Bg7 15.Nxd5) 10.0-0 e6 11.Rc1 Rc8 12.Qb3 Qe7 was equal in Melkumyan-Erdos, Berlin 2013. It stayed that way until move 58 when the game ended in a repetition.
- In a blindfold game Morozevich played 9.Ne5 against Giri. The ending after 9...Qb6 10.Qb3 Qxb3 11.axb3 Nb4 12.Kd2 does not appear to give White anything. However, after

12...h5?! 13.Nd3 Nc6 14.f3 Bf5 15.Ne5 Rfc8 16.Bb5 Nd8 17.Bd3 Bxd3 18.Nxd3 White did have a little something, and eventually won: 18...e6 19.b4 Nd7 20.Ra2 Nb6 21.b3 a6 22.Rha1 Bf8 23.Na4 Nxa4 24.Rxa4 Rc6 25.b5 Rb6 26.Ra5 Bd6 27.Bxd6 Rxd6 28.Nc5 Rc8 29.bxa6 b6 30.Nb7 (30.Rb5! bxc5? 31.a7, winning) 30...Rd7 31.Nxd8 bxa5 32.Nb7 Rdc7 (32...Rxb7! 33.axb7 Rb8 34.b4!? a4 35.Rxa4 Rxb7, White better) 33.Nc5 Kf8 34.Rxa5, winning, Morozevich-Giri, Beijing blindfold rapid 2012.

6...0-0

6...Ne4, and now:

- 7.cxd5 cxd5 8.Qb3 Nxc3 was agreed drawn in Wocke-Chandler, Wiesbaden 1988. A bit early, but play certainly is even.
- 7.e3 could be a try for more: 7...Qa5 (7...Nxc3 8.bxc3 0-0 9.Qb3!?; 7...Nxg5?! 8.hxg5, White slightly better) 8.Rc1, intending 8...Nxc3 9.Rxc3!? Qxa2 10.Qc1 (threatening to trap the queen) 10...Qa5 11.cxd5 Qxd5 12.Bc4, with superior development for the pawn.

7.Bxf6

7.e3?! Nh5 would be a little awkward.

7...Bxf6 8.h5

8...Bf5!?
8...dxc4 9.hxg6 hxg6 10.e4 is attractive to the human eye, but the engines are still OK with the Black position.
8...Bg4 9.hxg6 hxg6 10.Qd2!?, and castling queenside would be very original.

9.e3

9.Qb3 is met by 9...dxc4 10.Qxb7 Qb6!.

9...e6

10.hxg6

10.Ne5 was investigated by Grischuk during the game. After 10...Bxe5 11.dxe5, 11...gxh5!? was mentioned by Carlsen after the game, but Grischuk is right in calling it risky. (11...Qb6! 12.Qd2 Nd7 looks quite OK for Black, and was the reason why Grischuk went for 10.hxg6) 12.Qxh5 Bg6 (12...Nd7? was a suggestion by Carlsen after the game. It loses to the spectacular

13.Bd3! Bxd3 14.e4! Re8 15.Qxh7+ Kf8 16.Qh6+ Ke7 17.Qg5+) 13.Qh3, with a position that is pleasant for White.

10...Bxg6

White's pawn structure is a little better. However, Black is quite solid.

11.Bd3 Nd7 12.Qe2

12...e5?!

This gives White something to bite on. Both players called it a mistake after the game. Useful moves were 12...a6 and 12...Qe7.

13.Bxg6 hxg6 14.dxe5 Nxe5 15.cxd5 cxd5

So Black has an isolated pawn. However, it still does not seem to promise White all that much.

16.Rd1 Qa5! 17.Kf1

17.Rxd5!? Nxf3+ 18.Kf1!? (18.Qxf3 Bxc3+ 19.bxc3 Qxc3+ 20.Kf1 Rfd8 21.Rxd8+ Rxd8 22.g3=) 18...Nd2+ (18...Nh2+!?, Grischuk; after 18...Qc7 19.Qxf3 Bxc3 20.bxc3 Qxc3 21.g3 Rfd8 22.Rh4!? White is slightly better) 19.Qxd2 Rad8!, and Black has sufficient compensation for the pawn (it's even).

17...Rad8 18.Nd4 Qc5 19.Nb3

Both players now saw nothing better than to repeat the position, and it seems that they were absolutely right. If White wants to play, he can move his king, but say that after 19.Kg1 Black does the same with 19...Kg7 How to continue now?

19...Qc6 20.Na5 Qc7 21.Nb3 Qc6 22.Na5 Qc7 23.Nb3 Qc6

Draw.