Chess Opening Theory/1. e4/1...e5/2. Nf3/2...f5/3. Nxe5

= Latvian Gambit =

3. Nxe5
White is threatening a fearsome queen check on h5, with the e5-knight ready to pounce on the g-pawn should Black advance it to g6. There's also the incidental threat of exf5. So that's the Latvian Gambit refuted then? Most top players would say "yes". But there is a particularly optimistic breed of chess player that looks at the pawns on f5 and e4 and thinks that if the knight could just be driven away from e5, it's Black who will be capturing a pawn (fxe4), and possibly even defending it with ...d5.

And so the move /3...Qf6/ was invented. Yes, it's not usual to move the queen so early, but then again, following up a wild lunge like 2…f5 with solid-looking moves is a recipe for disaster.

There's also /3...Nc6/, the Fraser Variation. "You're threatening to win the exchange? Coward! I shall FORCE you to win the exchange!"

/3...Nf6/ has recently been suggested, preventing Qh5+ but allowing exf5. This may or may not be suicidal.