To give readers a taster of what is to come, I have published an article on the Allgaier Gambit, 1.e4 e5 2.f4 exf4 3.Nf3 g5 4.h4 g4 5.Ng5, a dubious but tricky knight sacrifice, offering an analysis of the key lines and five illustrative games, which are also available for PGN download.
As with many such unsound lines, it may pick up a fair number of points in casual and rapid games, since White always appears to get some sort of attacking chances, even in the most critical variations. I have occasionally faced the Allgaier and related lines with Black, and know that Black's defence is not quite as easy in practice as it should be in theory.
In summary, White leaves the knight without a safe retreat square, and thus after 5...h6, White must sacrifice the knight on f7 and bring the black king out into the open. The Hamppe-Allgaier Gambit (with 4.Nc3 Nc6 and then 5.h4 g4 6.Ng5) gives White rather better prospects.
For a second opinion on the Allgaier Gambit, Tim Harding's articles for ChessCafe.com, Kibitzer #78 and #79, are worth checking out, though he mostly reached similar conclusions.
5...f3!? (recommended by "micawber" at the Chesspublishing.com forum) is critical, since after 6.gxf3, cutting off the white queen's route to g4 and h5, Black can now get away with 6...f6. However the main line is 5...h6 6.Nxf7 Kxf7.
Theoretically White is struggling to prove enough compensation for the knight after 7.Qxg4?! Nf6, 7.Bc4+ d5 8.Bxd5+ Kg7 9.d4 f3!, 7.d4 f3, and 7.Nc3 d5. I tend to think that 7.d4 f3 8.Be3!?, preparing Qd2 and queenside castling, is White's best practical try, and 7.d4 also has the advantage that in practice Black often replies with 7...d5 or 7...Nf6, in which case White replies with 8.Bxf4 and may well get enough compensation for the sacrificed knight.
Meanwhile, here is an example of 7.Nc3:
Unfortunately for Allgaier Gambit fans, while the opening will always give practical chances, I don't think that it can be revived theoretically. However, it's probably sound enough for some players to have a fair amount of fun with it in rapid and casual games.
In the next few weeks I intend to cover a range of King's Gambit lines, focusing mainly on approaches with an early Nc3 and d4 followed by g2-g3 against an early ...g7-g5 from Black, for I consider these to be more promising than the Kieseritzky (5.Ne5 instead of 5.Ng5, above) and the Hanstein (4.Bc4, which allows Black to set up a f4-g5-h6 pawn chain). I am a bit busy with work right now, but will hopefully get a fair amount up in the next few weeks.