Chessy
The Dutch Defense: Janzen-Korchnoi Gambit – A Shock Weapon for White
The Dutch Defense: Janzen-Korchnoi Gambit (1.d4 f5 2.h3 Nf6 3.g4) is a sharp, offbeat line where White immediately challenges Black's control of the centre and kingside. The engine rates the position at -0.21, a tiny edge for Black, meaning you are essentially dead level — neither side is better out of the opening. But don't let that fool you: across over 12,000 games on Lichess, White scores an eye-popping 63.6% at this exact position. That gap between evaluation and practical results is the whole story — your opponents will not know what hit them. The drill below puts you in White's shoes against an adapting engine, so you can learn exactly how to turn this early aggression into a full point.
Play the Dutch Defense: Janzen-Korchnoi Gambit against the engine
Free, no signup — you play white, the engine adapts to your level.
Jump into the interactive drill below to face the Janzen-Korchnoi Gambit as White against an engine that adapts to your level. Put those 63.6% winning odds to a
Create a free account →The Big Idea: Why the Gambit Works
By playing 2.h3 and 3.g4, White asks a direct question: can Black's king's pawn really hold the f5-square? Most Dutch players expect quiet positional play, but the Janzen-Korchnoi Gambit rips up that script. White sacrifices a pawn (or at least offers one) to gain rapid central control and open lines against the Black king, which often gets stuck in the centre or on the kingside. The statistics prove the gambit is psychologically devastating: Black's most popular move, fxg4, has been played in 7,482 games, yet White scores a crushing 72.9% against it. That is a huge number for a line where theory says Black is fine. The practical reality is that Black's position is harder to play, and the gambit gives you excellent winning chances without needing deep memorisation.
Engine's Top Reply: The Positional Approach
If you are curious what the engine wants, Stockfish (at depth 16) points to d5 as Black's best continuation — and that only appears in 758 of the 12,249 games on record. The follow-up runs d5 c4 Nc6 Nf3, building a classical pawn centre for Black. Notice that this move avoids taking the bait on g4. When Black plays d5, White's practical results dip to 41.6%, which shows that disciplined opponents can neutralise the gambit. But here is the good news: most club players will not find d5 over the board. The vast majority grab the pawn or play a flexible developing move, and that is where your winning chances skyrocket.
Punishing Black's Three Big Mistakes
The database flags three poor continuations that you, as White, should be ready to punish. Each one has a concrete cost in evaluation and a clear best response from you: fxg4 — This is the most common move and an inaccuracy that loses roughly 0.9 pawns. Black takes the pawn but lets you open lines at will. Follow up with active development and central play; the engine says something better was available for Black, but your practical winning chances are excellent. g6 — Also an inaccuracy with a similar ~0.9-pawn cost. Black fianchettoes but neglects the centre. Seize space with your pawns and pieces before Black can consolidate. f4 — This is a full mistake, costing roughly 2.3 pawns. Black advances the f-pawn again, leaving huge weaknesses. You can dominate the centre and target the overextended Black pawns immediately. In all three cases, the engine reports that Black should have played d5 instead — but they did not, and that is your chance.
Why Practical Performance Beats the Evaluation
The Stockfish evaluation of -0.21 says Black is microscopically better in a perfect world. Yet White's actual win rate of 63.6% (against just 34.0% for Black, with only 2.4% draws) tells a completely different story. This is the hallmark of a good surprise weapon: the position is harder for Black to navigate than the computer thinks. Your opponents will face unfamiliar structures, potential king-safety issues, and the psychological pressure of being down time and space after they take on g4. You do not need to outplay Stockfish — you need to outplay the human on the other side of the board. The Janzen-Korchnoi Gambit gives you excellent odds of doing exactly that.
Results across 12,249 Lichess games
| Most-played continuation | Games | White wins |
|---|---|---|
| fxg4 | 7,482 | 72.9% |
| e6 | 2,149 | 47.8% |
| d6 | 934 | 50.2% |
| d5 | 758 | 41.6% |
| g6 | 511 | 56.6% |
| f4 | 204 | 61.3% |
Frequently asked questions
Is the Janzen-Korchnoi Gambit sound for White?
The engine gives -0.21, a tiny edge for Black, meaning the position is dead level — neither side is better out of the opening. So it is perfectly playable, not refuted. In practice, White scores a massive 63.6% across over 12,000 Lichess games, making it a dangerous practical weapon.
What is the best move for Black against this gambit?
Stockfish recommends d5 as Black's strongest reply, aiming for a classical pawn centre with c4 and Nc6 to follow. However, this move appears in only 758 out of 12,249 games — most opponents will grab the pawn on g4 or play a different developing move instead.
Should I be afraid if Black takes the g4 pawn?
Not at all. fxg4 is the most popular reply (7,482 games), but White scores a phenomenal 72.9% against it. The engine marks fxg4 as an inaccuracy, and you get open lines and a lead in development for the pawn. Keep pushing forward and your opponent will regret taking the bait.
How do I handle the position after 3...d5?
After 3...d5, Black has found the engine's top choice, and you face a tougher fight. White's winning percentage drops to 41.6% in those games. Focus on solid development and central play — the gambit has been neutralised, but the position remains complex and you can still outplay your opponent in the middlegame.