Strange 16-move semi-slav loss (black)

Question

[FEN ""]
[Date "6/9/2018"]
[White "Opponent"]
[Black "Me"]
[Result "1-0"]
[WhiteElo "2217"]
[BlackElo "1874"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. Bg5 h6 6. Bxf6 Qxf6 7. e4 dxe4 8. Nxe4 Bb4+ 9. Ned2 O-O 10. a3 Ba5 11. Qc2 Nd7 12. O-O-O c5 13. Ne4 Qf4+ 14. Kb1 cxd4 15. Rxd4 Nf6 16. Nd6 1-0

My rating: 1874

Opponent's rating: 2217

This was an odd game with a quick blunder but I found myself not understanding much of the opening. Feel free to go over the following points or any other important pieces of analysis that I left out:

• I understand from looking over some of the theory that it's best to play c5 sooner than later. 9...c5 would have been straight to the point or maybe even I could have played 10...Bxd2+ 11.Qxd2 c5.
• Perhaps 10...Ba5 was a mistake because if white played b4 then he would have superior control over the c5 square
• Once again 11...c5? Or is playing Nd7 first plausible?
• White's queenside castling almost comes unexpected to me because of all the open space there. Should my plan still be to play c5 like I did in the game or should it shift to something else with perhaps a b5 push? I feel like white may be the one who is now better prepared to handle c5.
• 13...Qf4+ seems a little odd, maybe Qe7 instead?
• Maybe I can try holding with 14...b6 instead of taking so soon in the center? With the rook on d4 I find that my only actual safe square is b8.
• Obviously 15...Nf6 seems to be a blunder and I should have pulled my queen out of harm's way immediately.

Also, no computer analysis please unless there's something that can hardly be explained or justified without it. Thanks!

Answer 1

Considering c5: Yes at some point Black needs to break in the center with c5 or e5 but you need to prepare such breaks. You are behind in development at move 12 and should consider that before moving a pawn twice.

While 14...cd4 spends another move and brings White's rook to a threatening square. In general you made too many pawn moves.

Answer 2

9...c5 seems too early to me. Castling into safety is better.

10...Ba5 is reasonable. It's often played in the Marshall Gambit of the QGD.
I prefer Bd6 to maximize it's potential and to protect the center dark squares.

11...Nd7 is fine. I think c5 first is wrong as after 12. dxc5 Nd7 13. Rd1 (or 0-0-0) the threat of b4 forces you to give up your two bishop advantage.

12...b5 looks very interesting; opening the lines even at the cost of a pawn is worth it. c5 is also fine, I would consider e5, just to get my LSB to e6.

13...Qf4+ seems best, as it puts more pressure on white--especially the f3-knight. 13...Qe7 may have problems after 14. d5. This requires much calculation with many nuances.

I definitely prefer 14...b6. The LSB needs to get active and out of the way to the rooks. However, you must still watch for dxc5, with a discovered attack on the knight (after the bishop moves), and d5 with problems getting your pieces to an effective square.

On move 7, opening the center while behind in development is not recommended. Qd8 is better on general principles, but hard to make OTB. 7...Bb4 is an option.

Answer 3

12...b5! is the correct pawn break here. You see this theme in the Caro-Kann as well.