9. Nbd7 is correct here.

Theory continues with Nd7, b5 and Bb7 I believe. For white the theory was 10.g4 if I am not mistaken. It has been a long time since I studied this opening though, sio I can be wrong.

I think the main line does not have the castling before Nd7. More common is 9...Nbd7 10. Qg3 h6 11. Bh4 g5 12. fxg5 and 12...Nh5 or 12...Rg8. Such moves would be too risky after the castling. But it seems that these variations have not been played recently.

10...Nbd7 is a better choice than what was played.

Sensible, but it seems kinda slow in an opening where two side should race to attack the others king. Since the white Queen is now blocking her own g pawn, I am not sure how white should start the attack here. For black an expansion on the Queen side with b5 seems the logical way to go.

11. Kb1 isn't sensible, it's paranoid! 11. Bd3! puts indirect pressure on h7 and gives White the advantage.

Think this is thematic in this line jkarp and is much more aggressive - don't think there's anything wrong with Nc6 though

This is too slow. You are advancing your pawns and you have four pieces sitting on the wall doing absolutely nothing. What's the point? Besides, 11...b5 allows the unexpected 12. Nf5! exf5, 13. Bxf6 Bxf6, 14. Nd5 Qa5 15. Nxf6+ Kh8 and everything that Black owns is on the back rank except his Queen.

White should consider 13. e5 which I believe is better than 13. Bf3 in this position. Black's position is an uncoordinated mess so White might as well get this thing going.

Normal development with Nd7 seems the rigth way to go, to me. Ra7 doesnt achieve much, and I always prefer a quick development. Plus it is looking a bit vulnerable on a7.

I certainly thought about Nd7, because the basic setup in this position is Nd7, Qc7, Bb7. But then I began to worry about 13...Nd7 14. e5 dxe5 15. fxe5 Nd5 16. Nxd5 or 15...Nxe5 16. Bxb7 losing a piece. I could not solve this, so I avoided putting the knight between the queen and the bishop. With the rook on e8, this would not have been a problem. More thoughts with some computer assistance: I can avoid losing a piece by playing 15...Bxf3, but after 16. Nxf3 Nh5 17. Qh3 Bxg5 18. Nxg5 my knight's bad position will probably cost me something, for example 18...h6 19. Nxf7! Kxf7 20. Qh5+ Kg8 21. Rd6 Nxe5 22. Rxe6 and White is a pawn up. I could choose a different 13th move like Nc6 or b4. But for example 13...Nc6 14. Nf5! Ne8 (14...exf5 15. Bxf6 Bxf6 16. Nd5 winning back the piece with advantage) 15. Nxe7 Nxe7 and White has a better position. 13...b4 14. e5 dxe5 15. fxe5 Ne8 16. Bxe7 Qxe7 17. Ne4 and here, ironically, the computer suggests 17...Ra7.

I'm convinced at this point that Black has been playing without a plan this whole time. Did you think of 13...Ne8 trying to trade your bad Bishop for his good Bishop? All White needs to do is move his Queen to f2 aiming at your rook on a7 and you have to move it again, admitting your previous move was a lemon. Even 14. a3 is strong here.

Through the center, instead of going for the kingside with g4. It definitely looks like a very sensible plan.

I'm looking at 14...Rc8, increasing the pressure down the c-file.

Good question. Perhaps I thought something like 15. e5 dxe5 16. fxe5 bxc3 17. exf6 Qxg3 18. hxg3 gxf6 with a slight advantage. However, White has better moves as shown in the game.

15. Nc3-d5! e6xd5, 16. e4xd5 Nf6xd5, 17. Bf3xd5 Be7xg5, 18. Bd5xb7

The pin on g7, plus allowing the bishop to f3, were the reasons behind the move 10.Qg3 it seems.

Wow, that is a fantasic move.

When you're hoping for the opponent to mess up in chess you're in a bad way - that being said this is brilliant!

black_cat_hamlet, that is not what I meant. The options were either resign now or try to make the game more complicated for the opponent. The position is lost anyway. All options are bad, so it is better to go for the one that gives the opponent most headache.