Here's another HORRIBLE HORRIBLE line someone recently used on me, that just makes me nauseous.

1. e4 e5 2. f4 Nc6 3. Nf3 d6 4. Bc4 Be7 5. O-O Bg4 6. d3 Qd7 7. Nc3 Nf6 8. fxe5 dxe5 9. Qe1 Bxf3 10. Rxf3 Nd4 11. Rf2 Ng4 12. Rd2 Bg5 13. Rd1

At which point I'm in "oh my god. this is terrible" mode.

This starts to look like a refutation. The bishop pin ties down the knight AND the queen, and if the queen moves to unpin the knight, black can forcibly win the exchange? (unless white wants to expose his king)

Anyone care to draw a comparison between 3. Bc4 and 3. Nf3? Lately, I've been attracted to the King's Gambit and to 3. Bc4, as many tend to fall to 3... Qh4+ 4. Kf1 Bc5 5. d4 Bb6 6. Nf3 when I gain two important tempi.

6...Bg4 7.c3 is a move i have used before to prevent any ideas of Nc6-d4. it will also allow you to play b4 to prevent Na5 which will chase your bishop of its preferred diagonal. if however 7.c3 Na5 8.Be2 BxNf3 9.RxBf3 0-0 (Qd7) 10.b4 this move will come in handy as well if your opponent is thinking about castling queenside. these moves have worked well for me in fast games, but in a slower time controls you can most likely find better moves.

7.Nc3? Nd4 either winning the exchange or causing you to play g2xf3 which leaves your king in serious jeopardy.

after 7.Nc3? Na5?? you can play 8.Bxf7+! KxB 9.fxe dxe 10.Nxe5 winning the bishop plus two pawns.

i think 6.d3 locking your bishop out of the kingside and Qe1 were the problems in the line you gave. even 6.d4 exd 7.Qd3 allowing a later Qb3 would've done better for you i think.

9.Be3 seems to be called for here. i dont think 9.h3 would work as 9...Bh5 then if 10.g4 Nxg4 11.hxNg4 Bxg4 and your king is about as exposed as it gets.

Anyway, the entire point of the King's Gambit is to generate a quick attack, yes? Therefore, I believe it is fine, perhaps better, to play Bc4, as it already hits key point f7. White has defenses against the mate threats after Qh4+.

The spoiling of the king's position is a minor detail which will be difficult for black to exploit, as white will gain rapid development after Nf3, d4, Nc3, etc. Perhaps one problem with this variation, however, may be the difficulty to bring the h-rook in play, as the tempi gained may not be enough to develop it. On the other hand, the queenside and kingside will be cleared, thus the a-rook may take action on the e-file.

I play 3.Bc4 over 3.Nf3 many chess writers think this Kings Gambit is so analyzed so one can only get a draw but there is still lots of paths that are unexplored in the most important line.

When Gallagher reintroduced the King's Gambit with his book <Winning with the King's Gambit> he only analyzed the line 3.♘f3 and wrote in his foreword <...my apologies to those fans of the Bishop's, and various other third move gambits, but my advice to you is to start playing 3♘f3>.

The KG strategic goal is to open the f file to start an attack on the weak f7 square. 3.Bc4 is an aggressive move, aiming directly to f7.

<Isn't castling early a cardinal 'rule' of chess? > I think this is a bad advice: I've often made the mistake of castling too early!

If you really want a 'cardinal rule' :) to follow, I think this one is more precise:

The defender should always castle before the attacker (by 'attacker' I mean the player who has the initiative and/or an advantage in development) to avoid giving the attacker an 'objective' to aim his stategy to, before the defender has a developed position.

So, do not hesitate to keep your king in the middle sometimes: wait and see how the position evolves before choosing a side for castling!