TN wrote on 03/19/19 at 13:25:07:



I don't trust the resulting endgame from Anand-Gelfand, 2013. I think that Black's two bishops do not fully compensate for his pawn structure. Indeed, the computer seems to give White a small, stable advantage after 13. d3.

I agree that White has several challenging responses to 3...d6. While the main line leaves Black with some problems to solve, even 3. Bxc6+ cannot be ruled out, e.g.

3...bxc6 4. e5 dxe5 5. Nxe5 Qd5 6. Nf3 Qe4+ 7. Kf1 g6 8. d3 Qf5 9. Bd2! Bg7 10. Bc3! with a small advantage.

I remember thinking that after 3...e6 4. 0-0 Nge7 5. d4 cxd4 6. Nxd4 Qb6 7. Nxc6 bxc6, both:

8. Bd3 Bc5! and

8. Be2 Ng6 9. c4 Be7 10. Nc3 O-O 11. Rb1 c5! 

were fine for Black. In the latter case, the computer tends to favor White, but underestimates the Hedgehog-like resilience of Black's position. 

After analyzing 9...a5 10. a4 Nf6 11. e5 dxe5 12. Nb3 Qd6 13. Nxe5 Nd5 14. Nc4 Qc7 15. Bd2 O-O 16. h3!? Bf5 some more, I came to the conclusion that Black has enough compensation, i.e.:

17. Nbxa5 Rfb8!, or
17. Bxa5 Qf4 18. Qe2 Qh4 19. Nbd2 Bxb2! 20. Nxb2 Rxa5.

Other alternatives:
12. Nxe5 Bf5! 13. c3 O-O 14. Nb3 Qb6 15. Be3 Qc7 16. Bf4 Qb7 with equality,

16. Nbxa5 c5! 17. Qf3 e6! with full compensation, or

16. Qf3 Bf5 17. Rac1 e5! 18. Qg3 Rfe8! with full compensation.

@ TN: I liked the positions Black gets in the mainline after 3...e6, so I analyzed it extensively. However, I wasn't able to "solve" 4. Bxc6! bxc6 5. b3!, as recommended by Kornev and Jones. The best I could find for Black was 5...f6! 6. O-O Ne7 7. Nh4 g6 8. f4! Bg7 9. Nf3 O-O 10. Nc3, when White has a small, long-term advantage due to his superior pawn structure.

In fact, it is this whole "superior pawn structure" thing that puts me off from defending the Rossolimo as Black. For example, in the mainline with 3...g6 and 5. Re1 Nf6 6. Bxc6, I'm struggling to find any plan, and it seems like Black is just sitting and hoping that White cannot break through.

I agree with the latest update that 5...Bg7 "is probably the critical try." However, I don't believe in the specific line that Black played in So-Maze, 2019, when 8. Rb1! is probably even better than So's 8. a3 (as I mentioned before).

The line in Kovalev-Praggnandhaa, 2019, however, was something I was looking at recently. The idea of setting up a solid pawn structure with a very quick ...cxd4 and ...c5 is certainly attractive, but Black seems rather inflexible in the positions that arise. For instance:

5...cxd4 6. Qxd4 f6 7. O-O d6!? 8. c4 c5 9. Qd3 Bg7 10. Nc3 Nh6 11. Nh4 Rb8 12. b3 O-O is certainly critical.

Now, instead of Kovalev's 13. Re1, I suggest bringing the other rook to the e-file with 13. Bd2! Qd7 14. Rae1. After 14...Bb7, White has the paradoxical 15. Bxh6! Bxh6 16. f4, which makes perfect sense when you realize that the fight revolves around White's pawn break on f5. 16...f5 17. exf5 gxf5 18. Qe2 and I think that White has at least a small advantage.

Pragg's line is probably one of Black's best retorts, but White has more space and a more flexible pawn structure, so I am not sure that the game is fully equal.

Hello.

This 5.d4 is interesting for sure.

Also somewhat non-forcing though, making it harder to analyse.

Probably black needs to be quite precise to equalise (if this is a possibility) and that's nice. It does also look like white needs to be very precise though. Playing against the two bishops in a situation where black may also be slightly ahead in development is maybe not always so easy.

But ok. Facing the bishop pair as white in the Rossolimo is not an uncommon ocurrence

Have a nice day.

5.d4 is clearly a good move because Black can't reply as he'd normally do in the Sicilian. The normal response against d4 is of course cxd4 but Qxd4 then attacks the rook. This can be compared to a line from the hyperaccelerated dragon 1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Qxd4 and after 4...Nf6 5.e5 Black can harass the queen with 5...Nc6 (so White loses a tempo in the mainline 6.Qa4 Nd5 7.Qe4, although not easy for Black to equalise here either), but in this Rossolimo line that knight is gone of course. And if you don't play 5...cxd4, White takes on c5 and you have to spend time with your queen getting the pawn back.
So for the moment I don't see how to equalize...

edit: maybe 5.d4 cxd4 6.Qxd4 Nf6 7.e5 Nd5 8.O-O Bg7 9.Qh4 h6 planning d6 is still the best way to go
edit 2: but then I can't find something against 10.Rd1 d6 11.c4

I was thinking about switching back to playing ...Nc6 sicilians recently and am searching for a reliable response to the Rossolimo. I'm leaning toward 3...g6 for now, with the idea of meeting Bxc6 with ...bxc6 lines and 4. O-O Bg7 5. Re1 with 5...Nf6. However, while analyzing I came upon an interesting sideline:

3...g6 4. Bxc6 bxc6 5. d4!? with the following variations:

5...d6!? 6. O-O cxd4 7.Qxd4 e5 8.Qd3 +/=

5...d5!? 6. exd5 Qxd5 and now both 7. dxc5 and 7. Nc3 Qc4 8. dxc5 are challenging for Black.

5...cxd4 6. Qxd4 and now:
6...Nf6 7. e5 Nd5 8. O-O Bg7 9. Qh4 is dangerous for Black.
6...f6 7. O-O Nh6 8. c4 Nf7 9. Nc3 Bg7 10. Rd1 O-O 11. c5!? with a bind on the center.

5...Ba6 6. dxc5 Qa5+ 7. Nbd2 and now:
7...Bg7 8. Rb1! and Black has an abysmal score from here.
7...Nf6 8. a3 Qxc5 9. c4 and I can't seem to find full equality after either 9...d5 10. exd5! or 9...Bg7 10. Qa4!? Qb6 11. O-O d6 12. e5 dxe5 13. Nxe5.

5...Bg7 6. dxc5 Qa5+ 7. Nbd2 Qxc5 8. O-O d6 9. Re1 and now:
9...Nf6 10. e5 dxe5 11. Nxe5 O-O 12. Nb3 +/=
9...a5 10. a4 Nf6 11. e5 dxe5 12. Nb3! Qd6 13. Nxe5 Nd5 14. Nc4 Qc7 15. Bd2 O-O 16. h3!? Bf5 17. Nbxa5 when I'm unsure whether Black has full compensation for the pawn.

Thoughts on Black's best route to equality?

(P.S. I'm willing to discuss my reservations with any of Black's other third moves as well.)