closed Sicilian game. Any advice greatly welcomed.
22.Rxb5 cxb5 23.Rc1 Qxc1 24.Nxc1 would have been interesting and probably winning for you. 25.Rxb5 or 25.Rxc6 Bxc6 26.Qxc6+ would also have given you good chances. Look out for those tactics especially when the enemy king is open like here and you have moves with check! Lots of stuff in this game, Bxc6 was probably a poor choice I would go d4 I think - at least wait until black plays a6 before Bxc6. No need to do it unprovoked. There was also 20.fxe6 en passent capture. As you say allowing your knight to be lost was a mistake but after that you still managed to create a good amount (and indeed winning IMO if you had been accurate) of counterplay so some good play there.
Didn't get very far, but wanted to add that 9. a3 is a fairly serious thought process error. Black's only real chance to get play is to play c4, so your idea of a3 & b4 is just besides the point. 9. a3 c4 10 d4 and you still have edge, but you are not getting any value from a3 either. 9. a4 doesn't end up yielding value either after 9. a4 c4 10 d4 e6 11. be3 and white is doing fine but a4/a5 looks besides the point. So where are the plays? One idea is b3...its a little odd looking, but it shuts down black's idea to get counterplay, giving white time to rip the queenside open subsequently. Stockfish confirms a big advantage. Another idea is f5, which makes it hard for black to finish developing in a coherent fashion. Stockfish again shows a big advantage. Anyway, my main point here is that a3 was an illusory move, that ignores black's plan, and actually has no benefit if you understand what black will do.
yes, looking back at it only God knows why I didn't see c4 as an obvious way for black to put pressure on d3! a3 was cripplingly bad! I sensed that my opponent wanted to harass the d3 square with his bishop by putting it on a6 earlier, so his c4 resource should of been on my mind. I guess this game is one for the comedy books!