Here's the way the board looks as of PN's last message:
Here's a board generated with Portable Game Notation which has our entire game laid out in full.
0-1
I'd like to thank PN for an excellent game. As I told him in chat, even though I scored a victory here, he made me sweat and grind my teeth for it. It was a hell of a run. Even more amazingly to me is how well he plays after years away from the game!
Both his and my play has given me a great deal to think about. Now the mistakes in my play have led me to conclude that I should pick a style of opening and stick with it, regardless of whether or not it's been codified as part of traditional opening theory. If I had wanted to keep the game open—the opening called "Giuoco Piano" that we ended up in is considered a closed game—I should have played 4. ... d5?! in response to his 4. d3, even though it's not accepted by traditional opening theory.
Secondly, I should have the courage to sustain an attack when I make it, rather than frittering away valuable tempo by turning my attentions elsewhere when I've got a piece of his on the run. I should have followed up 8. Bh4 with 8. ... g5 and even 9. ... g4.
Thirdly, beware of a premature attack. PN's 26. Nf5? was premature, and could have been refuted, if he had persisted in Nxh6+, with Qxh6, then Bxh6? would have resulted in Re1#. I know he'd agree, because we're talking about it in chat. He cursed when he figured out what a position he was in, but he got impatient, as he explained to me. Impatience can be as problematic as my overcautiousness was.
Fourth, castle early. PN should have castled when I did, on his 7th move, or earlier. The pin I had on his knight because he didn't castle caused him no end of problems.
Fifth, the sacrificing pieces or using check for positional or strategic advantage is no bad thing. I did that multiple times in this game. 12 Bb4+ was simply to get him to move his knight into a corner and pin it there (note that this check wouldn't have been possible with an earlier castle). My Bxd2 move was just to get the rook off the back rank, and then taking that rook with my knight was designed to either move the bishop or the queen out of the way. My c5 move was designed to break up the hold white's pawns had on the center, either by white doing dxc5 or me doing cxd4. And note that I didn't try to stop the queen taking my pawn at c5, because I knew that would keep her from causing problems elsewhere to my e8 or a8 rooks (Qa4! would have caused me serious problems), and indeed she didn't move from that c5 square the rest of the game.
Again, thank you PN for a wonderful, thrilling, and challenging game!