It is nice to see that your opponent also annotated the game...

I didn't study much material on the Halloween Gambit, but from all that I have seen it has a bad reputation, is not used by grandmasters and is definitely not recommended in correspondence chess. It is however considered a playable surprise weapon in blitz games and should not be underestimated. If you take a look at the statistics you will see, that white scores pretty well with the Schultze-Mueller Gambit as this opening is also called. Nonetheless it takes guts to play this in correspondence chess.

I agree with you that theoretically, this is bad but, the resulting positions suit my style perfect. You have to understand the interplay of dynamic & static weaknessess & strengths to have success in this gambit. A good understanding of initiative is vital here.

Someone - it might have been S. Tarrasch, or maybe Em. Lasker, someone active in the late 19th Century at any rate - once defined a Gambit as the 'sacrifice of material in the opening for the sake of getting a lost game.' That's something of an exaggeration. But the gambiter has to play very actively and imaginatively to make good the material investment. Paul Keres used therefore to recommend developing players to adopt such lines to improve the imagination. In his young days he played them in correspondence games...

White will achieve massive domination in the center and black gets a cramped and uncomfortable position. Is this worth giving up the piece? No doubt it might work in OTB chess, especially if there isn't too much thinking time (Rapid, Blitz). But in my opinion it is not to be recommended in correspondence chess with the use of databases and books.

I dont know... some of the more dubious openings (Fried Liver Attack, Latvian Gambit, Wilkes-Barre, Double Muzio) have their greatest moments in Corresspondence Chess.

I think that 6...Ng8 is the much stronger alternative here. Black should hang on to his material advantage. 6...Bb4 should produce equality, but why settle for equality, when you can have the advantage? If you look into databases there are two games with this Bb4 line that deserve attention: Sigfusson - Schubert, 2002 Reykjavik Open and Sigfusson - Nohr, 2005 Politiken Cup. Both these games were won by white, although it has to be noted, that according to the elo-ratings he was by far the stronger player both times. Sigfusson-Schubert: 6...Bb4 7.exf6 Qxf6 8.Be3 d5 9.Qd3 c6 10.O-O-O O-O 11.Kb1 Bg4 12.f3 Bf5 13.Qd2 Rfe8 14.Bg5 Qd6 15.g4 Bd7 16.h4 Bxc3 17.Qxc3 a5 18.h5 Nf8 19.h6 Ne6 20.Be3 g5 21.Qd2 f6 22.Bd3 Re7 23.Bf5 Nf8 24.f4 Rae8 25.Rh3 Bxf5 26.gxf5 Qd7 27.fxg5 Qxf5 28.Rg3 Kh8 29.gxf6 Qxf6 30.Bg5 Qd6 31.Qg2 1-0 Sigfusson-Nohr: 6...Bb4 7.exf6 Qxf6 8.Qd3 O-O 9.Be3 d5 10.O-O-O c6 11.Kb1 a5 12.Qd2 Re8 13.h4 a4 14.Nxd5 cxd5 15.Qxb4 a3 16.h5 Nf8 17.b3 1-0 Of course, 6...Ng8 doesn't look like a move anybody would want to play; it's just so counter-intuitive to move a piece back to it's starting square instead of getting on with developement. But if you play carefully, white doesn't have that much of any attack. A continuation I recommend for black is the following: 6...Ng8 7.Bc4 (white's most likely choice here) d5! (giving back a pawn) 8.Bxd5 c6 9.Bb3 and 9...Bb4 or even 9...Be6.

The continuation you give with 7 ...d5! is what I used to have the most difficult time with. It is seriously hard to stir up a serious attack in that line.

I think I'd prefer to take the pawn, but the text move isn't that bad.

Agreed, taking the pawn looks like a good option. There are no discovered attacks against the queen threatened. White would probably just castle and black can then castle too. The only moves black has to avoid after 8...Qxd4 9.O-O are d6, d5 and f6 which allow for Nb5+ or Nxg6+ with a discovered attack against the unprotected queen. And these are very unnatural and obviously bad moves, so I don't think anybody would feel tempted to play them.

8.Bd3 is a subtle positional trap that I'm certain is... incorrect. Originally, I was going to attack the weak c7 square with a tempo check (8.Qe2-c4) but, was worried about 8.Qe2+ Nge7 9.Qc4 Qe6+. I do not think a queen exchange is the best move for Black here but, I did not want to risk losing my queen. My idea with 8.Bd3 tempting Black to capture on d5 is: 8 ...Qxd5 9.O-O O-O 10.Bd2! & Black is already facing some serious strategical decisions to make. This is the kind of position that computers do not understand (I am sure everybody's Fritz would say White is slightly worse here) but, the truth is... positionally White is just about winning. Black is forced to play 10 ...c6 (or 10 ...a6) but, after 10 ...c6 11.a3 Be7 (If 11 ...Bxc3 than what was the point of 10 ...c6?) 12.Re1 Qf6 13.Ne4 with 14.Bc3 to follow. And, IF 10 ...Bxc3 THAN 11.Bxc3! and it should be easy to see that White has a very strong attacking game with control of the e file, the 2 Bishops & an undeveloped Black army. Now, the d4 pawn (after 8.Bd3) might be bitter tasting but, the poisoned one was the b2 pawn that I was hoping Black would grab; i.e. 8 ...Qxd4 9.O-O Bxc3 10.bxc3 Qxc3? 11.Re1+! & the Black King is stuck in the center of the board while his army is still undeveloped... not really good.

No need to consider, you did the right thing

The white pair of bishops is a force alright and the black knight is somewhat awkwardly placed on g6. Still, black does have a good pawn structure and is not in any immediate danger. Game looks fairly equal, with slightly more comfortable play for white.

The nice pawn structure is an illusion as it is easily attacked & weakend. My friend Gavin here is correct, the half open b file causes some difficulties for Black.

Agreed, Nf4 looks better. I also considered Bf5 right away, but the problem with that is, that white will not allow the exchange of bishops but will rather play Qe2+ after which Be6 appears to be the best move for black, making Bf5 look stupid. It might even lead to lines such as: 10...Bf5 11.Qe2+ Be6 12.Bxg6 Qxg6 13.Qb5+ c6 and 14.Qxb7 ...

10 ...Be6 is best here I think. In your analysis; 10 ...Bf5 11.Qe2 Be6 12.Bxg6 Qxg6 White has some exposed weaknesses & does not have time for queen adventures like 13.Qb5. For example; 13.Qb5+ c6 14.Qxb7 0-0! & now it is White who is stuck in the center since IF 15.0-0 THEN 15 ...Bh3! Also, what to do about the c2 pawn & the g2 pawn?

It also gets the rook activated. If white can gain the e-file for his rook that will give him the advantage.

I can totally understand the temptation to play 11...Bf5 and might have done the same. But I think 11...O-O is a better alternative here.

12.Qh5! very strong move. I also like 12.Re1+ which might have been my first choice for white. Either way black is under considerable pressure. After 12.Qh5 black cannot castle and he cannot play g6. Both are met with Bg5, of course. Probably h6 is the best defence for black. But then follows Re1+ and black has to either move his king, losing the right to castle, or if he plays Ke6 loses the d5-pawn which is not protected by the pinned bishop.

I think I was expecting 12. Re1+, and was preparing to play a dangerous 12... Kd7 and 'force' a castle on queen-side to allow me to connect my rooks and challenge for the e-file.

I like the defensive idea of ...h6! I also like that you are thinking about challenging for the e file. That is good thinking unfortunately, the e file cannot be contested as I have demonstrated in my annotation of this game.

Er... nobody seems to talk about the threat Bg5. Maybe I am missing something.