'Pawn takes' was well worth considering as well. I think of any decision one has to make in chess, this type is the one I find the most difficult. 11.dxe5 seems to win a pawn, but at the cost of leaving White with a doubled e-pawn and the bishops a little shut in.

I'll add crucial to archduke's assessment. His mention of the disturbed pawn structure and slight restriction of the movement of the B's is enough to turn the tide in favor of 11.Bxe5

I was thinking Black was pretty well placed to think about Q-side operations. His K-side is well enough protected. So 12...Rc8 or 12...b6.

If I'm white here I'm liking the empty space on g7!

I actually am embarrassed to say this but I actually meant to play f6, but in a total mind block I pushed the pawn an hit the clock. It is hard to describe how I felt at the time.

id=knight-nite - I have done exactly that myself, or near as dammit. But the occasion (Queen's Birthday tournament of 1976 or 77) was an opponent who kept dropping off to sleep, and not being quiet about it. In a near on winning position I was in the act of moving my rook when he suddenly snorted and jerked just as I was setting the piece down. "Wrong square!" I murmured and quickly adjusted it to the right place. Boy, did my opponent wake up fast! But he couldn't know whether or not I had released the piece before spotting my mistake, and I wasn't saying I had. I went on and crushed him in short order.

I might have been inclined to look for perhaps a more opportune moment for this exchange. I rather like the look of g4, in principle, at any rate.

17.c4!? possibly?

I think Black was correct here to close down the h-file, and it had to be done at once. Sure, that leaves Black with a weak g6-square, but White has only limited avenues of attack to come at it.

Black underestimates what White can do against g6. Best was 18...Rf6 or ...Rg8 or ...Kh7.

A move too late, but that move was not to be got back. Marginally preferable, possibly, was 19...Bf7, But White has just about a winning edge, regardless.