Now Try This...

Here's a game with some ideas. It takes too many moves and the final position isn't quite right, but maybe it will help:

Another attempt:

oh that's nice - by having the bishops taken by black pieces, you can castle, and also get the rooks out!

Close but no banana...

Originally posted by thaughbaer
Close but no banana...

[pgn]
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]

1.Nf3 Nf6 2.Ng5 Ne4 3.Nxh7 Ng3 4.Nc3 Nxf1 5.Nb5 Ne3 6.a4 Nf5 7.O-O d5 8.Nxa7 Qd7 9.Re1 Qxa4 10.Kf1 Qxa1 11.Nc6 Qxc1 12.Nd4 Qa1 13.Qb1 Nh6 14.Rd1 Ra6 15.Ke1 Ng8 16.Qa2 Rh6 17.Rb1 e6 18.Rd1 Bc5 19.Nf8 Qxa2 20.Nd7 Bxd4 21.Rc1 Bxd7 22.Ra1 Qxa1# *
[/pgn]

The only thing I can think of is maybe the black queen gets to a1 via a8, i.e. Qd8-a8xa1, saving a move on the Qd8-d7-a5xa1 maneuver. However this would involve the b8 knight moving and it seems that it will take that night much longer to capture the bishop on f1 and return to b8 than it would the g8 knight.

How about a rogue promoted a-pawn in the form of a knight? It wouldn't have to waste time getting back to g8 if it was captured somewhere.

If the rogue knight captures the bishop on c1, all of White's pieces can come streaming out of the a-file:

If the promoted pawn knight is captured on f1, then we save a move or two. I bet this isn't a million miles from the correct solution:

Is this it?

There's something about these type of problems that draws one in spite of themselves. I know this one took me a few days (off and on, of course).

Here's a near miss with a rook promotion. Can anyone shave a move off?

I give up - can you post the solution SwissGambit?

Originally posted by Data Fly
I give up - can you post the solution SwissGambit?

Quite possibly the hardest Proof Game I've solved.