fjd wrote on 03/26/20 at 15:53:27:

Considering that 11... g5 has fallen out of fashion and that 12. Nxe5 gxf4 13. Nxd7 is perhaps the main reason why I find it very hard to believe that the exact same position with the moves ...f6 and 0-0 inserted (which arises after the mainline of 11... f6 12. 0-0 g5 13. Nxe5 gxf4 14. Nxd7) could be an attractive option for black. If you click through the lines it seems like pretty much every comparison of the two lines is in white's favour: after 14. Nxd7 black cannot reply with 14... 0-0-0 because white's king there are no checks on the d-file, after 14. Nxd7 Qxd7 white gets the extra options of 15. Qc1 and especially 15. a5 that would work well for black without 11... f6 12. 0-0 inserted, and even if white just treats the positions identically with 15. Qxd7+ it is not so obvious that black has really benefited (amusingly enough it seems that both sides would be better off without their free moves, since the white king is better placed on the queenside and the pawn on f6 restricts the scope of the dark-squared bishop that is likely to find itself e7). In all these lines it seems like black is just hoping to hold some depressing endgame with the bishop pair as compensation for the pawn, and ICCF practice both in this line and in similar endgames arising after 11... g5 suggest that even with engine assistance the draw won't be so easy to get.

In regards to the evaluation of the 5... Bf5 6. Ne5 Classical Slav lines other than this Caruana 9... e5 I think people might be interested to check out two recent editions of Jan's Opening Clinic (23 Season Finale, 10:05; 25 Part 5, 25:02). In both of those editions Gustafsson agrees that the line with 6... Nbd7 7. Nxc4 Nb6 is objectively black's best option, but that from a practical perspective he thinks white gets a lot of options and there isn't much upside for black. On the Morozevich line (6.. Nbd7 7. Nxc4 Qc7 8. g3 e5 9. dxe5 Nxe5 10. Bf4 Nfd7 11. Bg2 g5) he suggests that accepting the sacrifice with 12. Nxe5 is the problem, which is certainly consistent with the statistics from recent high level OTB and correspondence practice, although black shouldn't be too thrilled with the traditional 12. Ne3 either. As regards the famous piece sacrifice line that is the topic of this thread Gustafsson agrees that 8. Nxc4 0-0 9. Kf2 is the critical line, but of course does not mention Caruana's 9... e5. While I agree that this 9. Kf2 looks a critical test I am quite curious to hear what people's thoughts are on the traditional 8. e4, where everyone seems to be assuming that black is fine but white is doing quite well in my databases and there is no obvious concrete solution for black. (Certainly not obvious to me anyway). The fact that Giri went for 3. Nf3 Nf6 4. e3 against Caruana suggests that he wasn't too optimistic about white's chances against Caruana's line, which is definitely worth taking note of considering his abilities as an opening analyst.
Thanks in particular to Paul Cumbers for sharing his analysis. I had line 2 as leading to an advantage for white in my files but my engines weren't running at anywhere near that depth, so finding out that Stockfish eventually comes around to 0.00 is quite interesting. I would note that at least at low depths on my hardware LC0 is claiming a white edge after both 28. Qe1 and 28. Rd1, so I'm not so sure about the claim that Stockfish is the only engine claiming a white advantage.

stockhausen wrote on 03/24/20 at 16:43:16:

But Monokroussos doesn't consider Aagaard's earlier improvement for Black, i.e. 17...Rc8! 18.Ne3 and only then 18...Ng6 (Caruana played 17...Ng6 straight away). Aagaard suggests that Caruana simply mis-remembered his preparation, i.e. he forgot to move the rook move before continuing with ...Ng6. This sounds entirely plausible to me. Aagaard's main line then continues to a position where the engines - apart from Stockfish - see no advantage for White (see my earlier post). I found that Stockfish too says it's equal if you let it think for long enough.

I'm not really up to date on Slav theory, but what's supposed to be the problem with the ...Nbd7/...Qc7/...e5 lines?

Hello.

I imagine black has x-number of realistic lines after and possibly also before 9.Kf2, one being Caruana's. They might all be objectively difficult for black, they might not.

Do you want the discussion to be about why the lines should be deemed problematic or do you just want to say: I looked at (presumably) a lot of lines. Overall it didn't look good at all so watch out Slav players. It's sort of easier for others to reply in a constructive manner if things like this are clear.

Have a nice evening

trw wrote on 03/22/20 at 15:33:19:

No, I don't or you would have provided some lines.

According to FM Monokroussos GM Caruana lost hiw way when he started deep diving.

http://www.thechessmind.net/blog/2020/3/21/candidates-round-3-ding-beats-caruana...

"No preparation" looks too harsh, but perfect it apparently was not. At the other hand FM Monokroussos repaired the line.

"9.Kf2 was given a ! by Vigorito (I can't find the source, maybe ChessPublishing?)."

here on chesspublishing in the game Lashkin-Miron and also in New in Chess Yearbook 132