On further examination, here's what frightens me the most:

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Nxf6+ gxf6 6.Nf3 which shouldn't cause much grief 6...Bf5 7.Bd3!? Bg6 8.0-0 Qc7 9.c4 Nd7 10.d5! There's nothing outwardly problematic here beyond the fact that Black is about to get wiped off the board in very short order. 0-0-0 11.Be3 c5 12.Be2 Kb8 13.Rc1 f5 (13...c5 is better) 14.c5 Nxc5 15.Bxc5 e4 16.Bd4 exf3 17.Bxh8 fxe2 18.Qxe2 Bd6 19.Bf6 Bxh2+ 20.Kh1 Bf4 21.Rc4! 1-0 Matulovic-Larsen(!!) (Hastings 1972). I'm coming to terms with the fact that Larsen didn't actually play B16 particularly well. His efforts are far less theoretically significant or consistent than Bronstein's. Or Turov's, for that matter.

In short, I think there are some obvious improvements that can be made to Black's play in the game above, but rather than the 6.c3 lines with a fianchetto or with 7.Ne2 as previously discussed, I find this very natural 6.Nf3 and the rapid c2-c4 try more potentially dangerous. The absence of a bishop on c4 makes Black's thematic Nd7-b6-d5 seem slow and awkward and irrelevant. The natural try is 6...Bg4, but I think it's interesting that Turov has shifted to ...Bf5 away from ...Bg4. Nevertheless:

6...Bg4 7.Be2 Qc7 8.Be3 Nd7 9.c4 e6 10.0-0 0-0-0 11.Qa4 Kb8 looks okay for Black.

I just looked at Turov's games. 6...Nd7 also looks completely viable and interesting. What I like about this whole opening is it seems like there is plenty of room for creativity and expression.

I also found it interesting that instead of 6.c3 when White plays 6.Nf3 instead of the universally recommended 6...Bg4 Turov simply replies 6...Bf5 and seems to get on just fine.

HgMan wrote on 09/22/17 at 22:53:19:


Interesting. I have attended lectures/training sessions with GM Turov twice, and both times he showed a Caro-Kann game ... but alas with Capablanca's 4...Bf5 main line, not 4...Nf6.

Does he play both lines? And the Bronstein-Larsen more as a surprise weapon or against weaker opponents, perhaps?

I spent a good portion of the afternoon with Maxim Turov's games, and especially his tries with 6.c3 Nd7!? He makes the Bronstein-Larsen look very very easy. More so than 6...h5--though it frequently transposes, I think his games offer a valuable model.

Stefan Buecker wrote on 09/20/17 at 16:51:55:


I appreciate this thought. But I wonder if the other rook to g8 at move 21 makes a tangible difference? 21...Rdg8 22.Qf3 f5 (I think my favourite part about playing the Bronstein-Larsen is being able to lash out with ...f5 twice in one game!). But maybe Black still has a tricky endgame after 23.Rfd1 f4 24.bxc6

HgMan wrote on 09/19/17 at 02:58:05:


7...e5!? Jeremy Silman doesn't like this idea for Black, which is understandable. Optically it looks all wrong. But with a bit of patience, Black might be able to uncover some counterplay. 8.Ng3 Be6 The point, I guess. Black tries to open the position while providing some cover for the king along the opening e-file. 9.Be3 Qc7 10.Bd3 Capitalizing on the departure of Black's bishop from this diagonal. As Silman notes, "the e4 and f5 squares are now in White's hands. Black's game is already very bad." 10...Nd7 11.O-O O-O-O 12.Qh5 ( 12.Qf3 h5 13.Bf5 was eventually drawn in Altshuler-De Carbonnel (ICCF WC05/final 1965). 12...Kb8 may not be as bad as Silman claims. 13.Rfd1 Bd6 14.Nf5 exd4 15.Bxd4 Be5 16.Bc2 Rdg8 This is playable for Black. In fact it looks rather good.

Not a critical line under discussion, but I stumbled across this lovely miniature from David Bronstein last night, which I think demonstrates the potential for activity in Black's position. I thought the 10...f5 & 11...Qh4 was a particularly pretty idea worth knowing about.

Aseev-Bronstein, Moscow 1982
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Nxf6+ gxf6 6.Bc4 Bf5 7.Ne2 Nd7 8.Ng3 Bg6 9.O-O e6 10.h4 f5 11.h5 Qh4 12.Qf3 Bd6 13.hxg6 hxg6 14.Re1 Bxg3 0-1

Keano wrote on 09/21/17 at 19:09:33:

Or perhaps your idea is too fast? 

Personally I like e7-e6 too, but now I think 11...0-0-0? is just premature. It seems better to delay castling: 11...e6! 12.b4 Qc7 13.a5!? (13.b5?! is hardly possible here; 13.Nxh5 a5!; 13.Bd3 Ne5!) 13...Bd6 14.Nxh5 a6!. Freezing the structure is more urgent than castling. For example: 15.g3 0-0-0 16.Nf4 Be4 17.Bf3 f5, and there is nothing wrong with Black's position. 

There are circa a dozen games with 11...e6 in the database, but none of them problematic for Black.