Most notable about this opening is that the reverse of the opening (e4 e5 nf3 nc6 nc3 nf6 g3 Nxe4!?) is actually more sound.

In a recent game, at move 5 my knight went to c6. White got his piece back fairly quickly, but the freewheeling tactical melee that followed led to a double-rook ending that seemed favorable to Black. My opponent timed out, though, before it could be tested.

Playing the knight to c6 is likely worse because usually White can play d5 with tempo and hence can play d6, crippling Black's position.

Seems like a reasonable decision in the circumstances. However, it might have been as well to have let Black exchange in his own time, and play something else, such as 10.0-0.

Being a piece ahead it doesn't make sense to weaken the pawn structure, much bettr would have been to play Bb4.

Mining operations. Black seeks to undermine White's e-pawn.

More subtle - though Black gets a pawn out of it - was to complete development: 14.0-0-0 Nxe5 15.Qh5+ Ng6 16.Bxd4. It's not easy for Black to find a move here. If, say, 16...Qc6 17.Rhe1, White's lead in development and aggressive stance gives him a very reasonable game.

That is right ! o-o-o is the better move : developping a piece, putting pressure on d4 and on the Black Queen, hindering the rokade ( because of Rxd4).

'...til Stneman's cavalry came, and tore up the tracks again...'

Instead, White would have better to have exchanged Queens first, and then advanced the h-pawn. That would have weakened Black's Q-side a little, and the White knight seems better on c3 for the moment, as will soon become apparent.

...which would not be the case had the W knight stood at c3.

Why? His rating is 442 higher yours. He sees more than you do. Play for simplicity. GL

No: I agree with Joanne here. Keeping it simple against a much stronger opponent is just about a sure recipe to get ground down. Have you noticed how highly rated players against opponents much less numerically endowed will avoid complications? Complications is where 'accidents' can happen. True, Freddy is very likely to come out ahead in the ensuing melee, whatever Joanne manages to conjure up. But it is a dead-set, 100% guarantee that if Joanne 'keeps it simple', she will lose.

I agree with adowman. When I am playing a lesser rated opponent I don't like to sacrifice much and complicate the position because I feel with normal play I can easily win a pawn or two and the game. On the other hand, when playing a stronger opponent I often like to play more abnormal(but still in my style) openings because I feel like I have nothing to lose.

21.Ne4

Better was to keep the K active, though Black already has just about a winning advantage. However bleak the situation, it is almost always a good idea to carry the fight to your opponent. To that, you need active pieces.

Wrong rook...

After 29. Rg1 I see Freddy playing Ne3 threatening Rd1# if you take Rxb7. So 29...b3 30. Nf5 Outpost & you take nothing. What is does prove is the unsoundness of playing 2 classes higher looking for a win instead of looking for instruction.

The way to receive the instruction is to look for the win. As it happens I think Joanne played a bit below her usual standard, especially in the second half of this game. However, if Joanne wanted to test out the Halloween Gambit, I would have recommended she try it first against opponents with about her own rating, and maybe some rather weaker (unrated if necessary). The idea there is to get a feel for what the line seeks to achieve, and what avenues of attack are likely to open up.