Chessy
English Opening: King's English Variation, Kramnik-Shirov Counterattack – a3
After 1.c4 e5 2.Nc3 Bb4, White plays 3.a3 — inviting you to give up your dark-squared bishop for the knight on c3. This is the starting point of the Kramnik-Shirov Counterattack, a sharp and principled way for Black to fight for the initiative. You voluntarily trade bishop for knight, doubling White's c-pawns and creating long-term structural pressure. Stockfish evaluates the position at -0.19, a tiny edge for Black — essentially dead level. Across over 90,000 games, Black actually scores 48.6% to White's 47.2%, so you are in excellent practical shape right from the move. The drill below will help you nail the correct follow-up and punish White's most common mistakes.
Play the English Opening: King's English Variation, Kramnik-Shirov Counterattack: a3 against the engine
Free, no signup — you play black, the engine adapts to your level.
Now it is your turn. Play the position as Black against the engine drill below, practise punishing White's mistakes, and build your confidence in the Kramnik-Sh
Create a free account →The Big Idea: Why Black Wants to Play Bxc3
The whole point of 3.a3 is to force you to decide immediately what to do with your bishop on b4. You take on c3, which might look like a concession — you're trading a bishop for a knight, after all. But look at what you get in return: White's pawn structure becomes damaged. After you play ...Bxc3, White recaptures with either the b-pawn (bxc3) or the d-pawn (dxc3). Either way, White ends up with doubled c-pawns that can become a long-term target. Meanwhile, your own pawn structure stays clean, and you keep your king's bishop and both knights for a lively middlegame. This is a classic 'exchange a minor piece to damage the opponent's structure' trade — a fundamental idea that rewards positional understanding.
The Critical Moment: Your Recapture Choice
After 3...Bxc3, White has to decide how to take back. The engine's best move is bxc3, which appears in over 67,000 games. White scores 47.3% from there — statistically your best result too. The alternative recapture dxc3 (about 23,000 games) scores 47.0% for White, almost identical. Either way, you're fine. The engine's suggested continuation after bxc3 is Nf6 f3 Nh5 — Black develops the king's knight, White prepares e4 with f3, and you reposition your knight to h5, often eyeing the f4 square or preparing ...Ng3 possibilities. This is a flexible, aggressive setup that keeps the pressure on White's centre and doubled c-pawns.
Punish These Common Blunders
Statistics show that many White players go wrong at this exact position. Three moves are outright blunders according to the engine, and they all involve avoiding bxc3 or dxc3. If White plays b3, the engine says White loses about 6.6 pawns — it's a terrible move. b4 is similarly awful, losing about 6.3 pawns. Even Nf3, which looks natural, is a blunder losing roughly 3.0 pawns. The reason is simple: by not recapturing on c3, White leaves that knight hanging and also fails to fix the doubled pawns. If you see any of these moves from White, you have a serious advantage. The drill will train you to spot them and respond correctly.
What the Statistics Tell You
Out of 90,282 games at this position, Black wins 48.6% of the time, White wins 47.2%, and draws are rare at just 4.3%. That is a fantastic practical score for Black — you win more often than White, and you lose less. The position is objectively near-equal (the engine says -0.19, a microscopic edge for Black), but the asymmetry in the pawn structure means the two sides play very different chess. You have a clear plan: develop naturally, target the doubled c-pawns, and enjoy the freer piece play. White has to be careful not to let the position open up too quickly. For a club player, these are exactly the kinds of positions where understanding beats memorisation.
Results across 90,282 Lichess games
| Most-played continuation | Games | White wins |
|---|---|---|
| bxc3 | 67,170 | 47.3% |
| dxc3 | 22,964 | 47.0% |
| b3 | 77 | 16.9% |
| b4 | 24 | 25.0% |
| Nf3 | 11 | 18.2% |
| g3 | 10 | 30.0% |
Frequently asked questions
Is the Kramnik-Shirov Counterattack good for Black?
Yes. The engine evaluates the position at -0.19, a tiny edge for Black, and across over 90,000 games Black wins 48.6% of the time — slightly more than White. It is a sound and principled opening choice.
Should I capture with bxc3 or dxc3 as White?
The engine prefers bxc3, and it is by far the most common move (over 67,000 games). White scores nearly identically with dxc3 (47.0% vs 47.3%), so both are fine. The key is that White must recapture on c3 — moves like b3, b4, or Nf3 are all blunders.
What is Black's plan after 3...Bxc3 bxc3?
The engine's top continuation is Nf6 f3 Nh5. You develop the knight, White prepares e4, and you reposition your knight to h5. From there you can target f4, threaten Ng3, or keep pressure on White's centre and doubled c-pawns.
Why is b3 a blunder for White in this position?
b3 avoids recapturing on c3, leaving the knight undefended and failing to fix the doubled pawn structure. The engine says it loses roughly 6.6 pawns — White is simply giving away material or a decisive positional advantage.
How many games feature the English Opening: King's English Variation, Kramnik-Shirov Counterattack: a3?
Over 90K Lichess games have reached the English Opening: King's English Variation, Kramnik-Shirov Counterattack: a3 position. White wins 47.2%, Black wins 48.6%, with 4.3% draws — based on real rated games.