Playing Against the Nimzo-Indian Defense: Nf3 (4...b6)

The pawn structure after 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3 b6 hasn't locked into anything rigid yet. Black is threatening ...Bb7, which would put pressure on your c4-pawn, and the bishop on b4 still pins the knight on c3 against your king. Your most important task is to decide how to handle the pin while maintaining your space advantage in the centre. The engine's preferred move, Qb3, does several things at once: it protects the c4-pawn, it attacks the b4-bishop indirectly (preparing a later a3 or Bd2 to kick it), and it gets the queen to a useful square away from the centre. The idea is to follow up with c5 and Bg5, building a harmonious setup before Black's fianchetto becomes too annoying. Black's plan will be to complete development with ...Bb7, ...d6, ...0-0, and eventually challenge your centre with ...c5 or ...d5. You want to stay flexible — don't commit to a king-side or queenside plan too early.

If you want the computer's seal of approval, Qb3 is the move. It leads to the line 5.Qb3 c5 6.Bg5 Bb7, which is exactly the sort of positional fight the Nimzo-Indian is famous for. The queen on b3 keeps an eye on the b7-bishop and discourages Black from playing ...a5 or ...Na6 too hastily. Meanwhile, Bg5 pins Black's f6-knight to the queen, preventing ...Ne4 and making it harder for Black to enforce ...d5. From a practical standpoint, Qb3 scores respectably at 49.2% for White across 17,410 games — not the highest win rate among the options, but it's the most principled. The idea is to keep the tension, develop naturally, and wait for Black to overreach. If Black plays ...0-0, you might consider a3 to ask the b4-bishop what it's doing, followed by e3 and Be2, with a comfortable position where you have slightly more space but no immediate winning plan.

The most popular move in the position is 5.Bg5 (56,036 games), and it scores 49.8% for White — essentially a coin flip. That's not surprising; Bg5 is a natural developing move that pins the knight, but it doesn't do anything about the immediate pressure on c4. The second most popular is 5.e3 (31,344 games, 47.2%), a solid choice that prepares Bd3 and kingside castling. 5.Bd2 (30,189 games, 46.3%) is a bit passive — it unpins the knight but doesn't challenge Black's setup enough. The best-scoring move by percentage is 5.g3 (8,307 games, 50.4%), which signals a Catalan-style fianchetto. That's a perfectly decent approach if you're comfortable with it, though the sample size is smaller. One clear warning in the data: 5.a3 is flagged as a mistake. It's been played 31,065 times but scores just 44.5% for White, and the engine says it loses about 0.6 pawns compared to the better move 5.e3. The problem is that after 5.a3 Bxc3 6.bxc3, you've given up the bishop pair and damaged your queenside pawns for very little gain.

The single most important thing to remember in this position is: don't play 5.a3. It's understandable — you see the annoying bishop on b4 and want to ask it a question. But the statistics are brutal: White scores only 44.5% after a3, and the engine considers it a clear inaccuracy. After 5.a3 Bxc3 6.bxc3, Black has the bishop pair, your c4-pawn is now backward on a semi-open file, and your pawn structure is compromised. Instead of a3, the engine recommends 5.e3 as the better alternative. That's a more modest move, but it's solid: it defends the c4-pawn indirectly (you can recapture with the bishop after ...Bb7 Bxc4), it keeps your pawn structure intact, and it leaves Black's bishop on b4 where it might end up misplaced once you develop properly. In the drill, try all the reasonable options and see for yourself how quickly Black can equalise if you give up the bishop pair.