Play the Nimzo-Indian Defense: Qb3 as Black

Your bishop on b4 pins the knight on c3, which already pressures the delicate Qb3 setup. White's queen came out early — that's the whole point of 4.Qb3 — but it also gives you a target. Your move 4...c5 strikes at the centre immediately, challenging d4 before White can consolidate. The resulting position is razor-sharp: whoever understands the central tension better will come out ahead. For you as Black, the main idea is to complete development quickly (usually ...0-0 and ...Nc6), keep the pressure on d4, and be ready to recapture on c5 if White takes there. Your bishop on b4 remains a nuisance; White often spends a tempo chasing it with a3, which is exactly what you want.

The computer's top choice for White is 5.Nf3, continuing with ...0-0 6.dxc5 Nc6. In that line White gives up the centre to relieve the pin and develop naturally. After 6.dxc5, you recapture with the knight (6...Nc6, not the bishop — you want to keep that diagonal active) and reach a comfortable IQP (Isolated Queen's Pawn) position where your piece activity compensates for the pawn weakness. You can target d5 later and enjoy easy development. If White avoids Nf3 and tries something else, your task stays the same: get your pieces out, castle quickly, and keep an eye on that d4-square.

The numbers from 27,971 games reveal a clear pattern: White struggles no matter what they try. Here is how the most common replies work out for White: - a3 (8,037 games): White scores just 43.3%. Chasing your bishop with a3 costs a tempo and doesn't solve the central tension. - dxc5 (7,754 games): White scores 45.7%. This is the most principled try, leading to the IQP structure mentioned above — which you handle well. - e3 (4,029 games): White scores 46.4%. A solid but passive move; you get easy equality. - Nf3 (3,429 games): White scores 45.4%. The engine's choice, yet White still underperforms in practice. - d5 (1,907 games): White scores 44.4%. This is actually a mistake, losing about 0.6 pawns of evaluation. White's best here was dxc5 instead. No matter which path White picks, you as Black are outperforming the engine's dead-equal assessment by a meaningful margin.

The move 5.d5 is the most instructive error White can make in this position. It's an inaccuracy that loses roughly 0.6 pawns of advantage (the engine says dxc5 was better). Why is d5 bad? You simply move your knight — ...Ne7 or ...Nd4 — and suddenly White's queen on b3 looks awkward while your bishop on b4 remains active. White has pushed a central pawn without adequate support, and you get to plant a piece in the centre or start a quick ...b5 or ...e5 break. If your opponent plays 5.d5 in a game, you can be confident you've already outplayed them in the opening. The drill below will let you practise finding the right responses.