There is not a single true chessplayer in the world whose heart does not beat faster at the mere sound of the words “gambit games.” — Bronstein
1. d4 d5 2. c4
White threatens to knock Black out of the center with 3. cxd5, and if 3…Qxd5, then 4. Nc3 plus 5. e4 is an excellent start.
2…dxc4
Black accepts the gambit, which isn’t truly a “gambit” at all because if White wishes, he can recover the pawn right off with 3. Qa4+, forking.
3. e4
Here we are. If White achieves Bf1xc4, he gets the better of the pawn trade for two reasons: the central deflection, and the gain of time for recapturing with a developing move.
The Queen’s Gambit is excellent chess. It has a sound positional foundation, and makes a real threat at move 2. On the other wing: 1. e4 e5 2. f4 White is not threatening 3. fxe5 because 3…Qh4+ is bad news. That makes 2. f4 an unreal threat, and Black is free to play however he wishes. 2…d5 is logical, but the average player is materialist by nature, inclined to capture anything that isn’t nailed down. They feel empowered by world champion Steinitz’s advice — true, in theory — “The best way to refute a gambit is to accept it”, which U.S. champion Evans repeated many times in Chess Life magazine.
2…exf4
As in the Queen’s Gambit, White’s ideal is to play d2-d4 to seize the center, then Bc1xf4 to gain time by recapturing with a developing move. The bug in White’s program is the check looming at h4. It can be a killer.
In sum: The Queen’s Gambit is one of the two or three most venerable openings in chess. The King’s Gambit is plainly “bad”.
2. f4 makes an unreal threat, does nothing to aid White’s development, and exposes to white king to a fearsome check. Dr. Tarrasch said in the classic handbook The Game of Chess: “1. e4 e5 2. f4 is a decisive mistake … it is almost madness to play the King’s Gambit”. My suggestion is to embrace the madness, and get used to making decisive mistakes on purpose.
Maybe you ask: Why bother? Poker players have an expression: “betting on the come”, meaning you don’t have the cards you want or need at the moment, but you’re betting that they might come in the future.
The King’s Gambit is bad — a decisive mistake, Tarrasch said — and we agree that’s true … for the moment. Players who gamble on 1. e4 e5 2. f4 are betting on the come, wagering that they’ll overcome the immediate tactical and positional obstacles, and reap the rewards — sometimes brilliant and immortal — in the future.
The list of great masters — world champions and challengers — who have risked the King’s Gambit at the highest levels of competition includes Fischer (1963 US championship) Bronstein (1948 interzonal; 1945, 1947, and 1952 USSR championships), Spassky (1985 candidates tournament, 1960 USSR championship), and Tchigorin (1892 world championship).
Some modern cynics can’t accept anything as legitimate unless the current king of the hill does: Carlsen (six games in the online databases, including two at the 2012 world blitz championship).