I happened to play a very nice attacking game the other day, in a Petrov Defence with black.
I saw an interesting sacrifice to get the king out of his castle, but was not sure that it was correct.
In the game, my opponent allowed me to checkmate him very quickly (it seems he didn't calculate much).
In the analysis, I was proud and happy to see that my Nxg2 was the move chosen by Stockfish and that the game itself was flawless!
However, I frankly can't say it's my best game, because 1) my opponent played the opening very shily and 2) he didn't fight a lot once the attack was on.
With almost 8 minutes on his clock, in a tense situation, he thought just 1 minute before committing his final mistake. Anyway, it seemed that White was already lost even after Nf1 or Ng1 or Qe2, although I would have had to find the right moves.
On move 16, the idea Nxg2 already crossed my mind, but it was leading to nothing after Nxg2 Kxg2 Fh3 Kg3 and there is no way to continue the attack. So I played a "normal" developing move (Re8). On the next move, I thought that with the rook joining the attack on e4, which added the threat of Qf4+, I could go for it. And so I did. I was mostly afraid of his move 20. Nd5 (which covers f4 and threatens my queen) because of lines like 20... Qg6+ 21. Ng5 Rg4+ 22. Kf3 and I would have had no time to mess around because of the threat of Ne7+. But I thought there should be ways to fight on. But then, just when the Knight was actually about to land on d5, I saw that it was mate in 2. And that is what was played.
One of those games that give you a huge satisfaction!
Hope you enjoy it.