Tournament Game- Caro-Kann

I thought since I mentioned the Caro-Kann yesterday, I would share a game that I played in the Hamilton Summer Open 2009 in the U-1900 section. Unfortunately, this is the Panov-Botvinnik attack, not the Cappablanca variation. And I still am unsure about the procedure of inserting diagrams.
Gillis, Doug (1656)-O'Sullivan, Patrick (unr.)(at the time)
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 -The Panov-Botvinnik attack
4.... Nf6 5.Nc3 Nc6 6.Nf3 Bg4 -I particularlly like this move. It allows me to keep an indirect eye on d5, with a potential Bxf3
7.cxd5 Nxd5 8.Qb3 -This is in the Rybka opening book, but I think my response justifes the exclusion of this move
8.... Bxf3! -Still book, but I found it over the board and was very pleased with the resulting position
9.gxf3 Nb6 10.Bb5?! -Out of book now, Rybka gives Be3 with the idea of O-O-O and d5
10.... e6 -stopping the threat of d5, and preparing to develop my dark bishop. I did not want to venture out with 10... Qxd4, because after Be3 followed by Rd1, I am getting pressured. A quick check with the engine confirms my suspicions
11.Be3 a6 -I decide that if I can get off the queenside pressure from the exchange, and the isolated c-pawn that results will bolster my control of d5, allowing my b6 Knight to reside there permanently
12.Bxc6+ bxc6 13.O-O-O -Nothing much better as far as I can see
13... Nd5 -In my opinion, the best move. The computer likes Rb8, eying the white Queen
14.Nxd5? -I assign a "?" because this move, although it removes my best piece, locks in his dark bishop, and fixes my pawn structure. Qa4+ does not worry me, as the Queen exchange is favourable for me. I want a good bishop vs bad bishop ending. Rybka likes Nxd5 and still thinks white has a slight edge, but I can't see it
14... cxd5 15.Rd3 Bd6 -Puts pressure on h2, making the rook deployment undesirable
16.Qa4+ Qd7 17.Qxd7+ Kxd7 18.f4? -Anti-strategy, locking in the dark bishop. I now have an edge, and my next idea is to make sure the f-pawn stays put
18.... Rc8+ 19.Kb1 g6 -activating a rook before locking in the pawn. I chose g6 instead of f5 because I was concerned about rook pressure on g7. When you have a strategic edge, do not harm your structure, and force the opponent to give ground in order to create ideas.
20.h4 -another pawn on a dark square, his bishop is bad. He is struggling to create an attack
20.... h5! -I do not miss the opportunity to fix this pawn
21.Rb3 Rb8 22.Rxb8 Rxb8 23.Rc1 Rc8 24.Rxc8 -in my eyes an admission of defeat, but to my surprise this move is coupled with a draw offer. I decline and coast to victory...
24.... Kxc8 25.Bd2 Be7 -wins h4, my passed pawn will be unstoppable
26.f5 exf5 -adding to my pawn mass
27.Bf4 Bxh4 28.Bg3 Bxg3 29.fxg3 f4 -a nice shot, the King can't stop the pawns
30.gxf4 h4 0-1 -A nice game in which I use the Caro-Kann as it was meant to be used- a slow eventual squeeze in the endgame. The key idea used in the game was limiting the dark squared bishop. I made moves designed to worsen my opponent's mobility and structure, and earned a win in my first official tournament game.
If you have comments or questions, please don't hesitate to write them. Sorry f I was too biased or if my analysis lacked enough concrete variations. I hope you enjoyed my analysis. My next step will be to figure out how to insert diagrams of positions.
"The tactician must know what to do when something needs doing; the strategist must know what to do when nothing needs doing" (Savielly Tartakower)