@jkarp, rookie879 is White in this game. This is a fine move, though.

Because after 32. Qc7 Black has to play Qxc7 and after Rxc7, I will win my pawn back at the very least. My knight is suffocating Black on d6 and I have a rook on the 7th rank. White has to be better there

Absolutely. White is of course still looking strong.

Black is solid, though, and has the extra pawn. White has, I think, to play very actively even to stay on level terms.

It would seem that 32...a5 is wrong - or maybe Black's Q-side play was a mistaken plan. What alternative was there then? Maybe ...f6 or ...f5, seeking to crack White's centre and undermine the Knight.

Oops. A very unusual miscalculation by one of Black's calibre (rating).

I don't think it's overconfidence really but I simply missed the move in my calculations. I envisioned this position in mind because I knew I can get the pawn back with a least a draw. However, with such a good position, I should've taken a little more time to look for a knockout

A case of 'believing your opponent', maybe? Not so much suspecting a trap as not investigating the possible consequences.

Probably Black was hoping for that scenario, but he would pretty much have counted upon the move you did play.

@jkarp: this sort of thing is called a 'desperado.' The rook is lost so you sell him for as much as you can get him. This capture is 'check', which ensures that White after all does not lose a piece.

This was not a blitz time control. We both had an hour and 10 minutes to play this game... However, in such a tense game, people tend to run low on time

6 minutes is more than enough people play whole great games in 6 minutes

This position should not be a draw: Black wins! He has a choice, at that. Killer blow 'A' is 47...b3!! followed by 48...a3 (This would be the line I'd favour). Killer blow 'B' you will see next turn. The mistake White made here is assuming that bishops of 'opposite colour' pretty much automatically confer a draw. +++++++++++++++ Six minutes left? Plenty of time!

Instead of 48...bxa3, how about 48...b3!! The game might have concluded 49.f4 b2!! 50.Bxb2 Kxb2 51.Ke3 Kxa3 and now 52.Ke4 Bc6ch or 52.g4 Bb5 53.gxh5 gxh5 54.h4 Bf1! 55.Ke4 Bh3 and Black's remaining pawns are safe. Even after the text move, Black should still be winning.

The one move that renders Black's pawns unassailable. But although 52.Kf2 is the only good move White has, it is not clear that it would have saved the game in the long run (if there were a long run): 52.Kf2 Bb5! 53.Kg3 Be2 54.Kh4 a2 55.Kg5 Bf3!! 56.g3 Kb1 57.kf6 Bd5 58.Kg5 a1=Q 59.Bxa1 Kxa1 60.Kxg4 Kb2 61.Kg5 ... -(or 61.f5 e5 62.f6 Kc3 63.Kf5 Kd4) 61...Kc3 62.Kf6 Kd4 63.g4 Ke4 64.Kg5 Bc4 'zugzwang!' 65.f5 e5 66.f6 Kd3 67.Kf5 e4 68.g5 e3 69.g6 e2 70.f7 e1=Q 71.Kf6 ... -(Not 71.f8=Q Qb1ch 72...Qxf8) 71...Qa1ch 72.Ke7 Qg7 73.Ke8 Qxg6 and wins. This 20-move sequence is just a sample, and Black could very easily slip into a drawing line, but it does go, I think, to suggest that with best play Black could still have won after 52.Kf2!