Losing won positions
Finally I was able to win a piece up position! Of course not without some mental strain.
At least I am seeing a common thread in my issues with winning won positions, Opponent rook activity. This motivates me to be more agressive with my rooks when I am losing and to put make sure my opponent does not get to active with his.
A good player once said that Kasparov was able to beat very strong players by knowing how to get his rooks involved quicker. I notice that if my rooks are first active and I achieve a winning position I can often win it. If my rooks are dormant and I get a winning position I often do not convert
I think your opponent went downhill after 14...Qd7. NxN and 17...Nc5 is a decisive move. 35.Rd6 looks a lot better to bring curtains. You should analyze the end game with an engine to understand what were the alternative moves. I believe you could've checkmated him many moves ago.
Its very frusterating to be able to find the best (or close to the best) moves to reach a winning position and then realizing that you are not able to find good moves after that. this was a 30 minute time control game without increment and in this time control I tend to shuffle while preserving the win instead of caluclating once I get low on the clock.
I completely agree on move 15 moving the knight was not the correct approach, I had misguided idea's about rolling the f and e pawns. In fact the second after he played Bg5 I knew my move was not the way I should have played it. I do think h6 is a good alternative.
21. . . c6 was bad as well, I think during the game I thought it was winning a pawn. I meant to play 23. . . Nxd5 to carry this out. Something about the pin's on the g8-a2 diagonal put me off from doing that.
I think you did hit the main points of why that game went bad for me. Thank you for your thoughts