Nimzowitsch Defense: Kennedy Variation with 3.c3 – You Play Black

After 1.e4 Nc6 2.d4 e5 3.c3 d5, you've created a pawn tension in the centre that White has to resolve. Your idea is simple: you want to open lines for your pieces. If White captures on d5 (exd5), you recapture with the queen — or if White captures on e5 (dxe5), you recapture with the knight or push forward. The engine's top choice is dxe5, which leads to the line 4.dxe5 dxe4 5.Qxd8+ Kxd8. That queens early trade settles the position into a roughly equal endgame. Your main job when playing Black is to stay alert to which centre pawn White captures — each choice leads to a different type of middlegame, and some of White's alternatives are outright bad for them.

White's strongest move here is dxe5 (12,636 games in the database, White scores 47.9%). The engine gives the continuation: 4.dxe5 dxe4 5.Qxd8+ Kxd8. After the queens come off, the position simplifies into an endgame where your king is stuck in the centre — but Black's control of e5 with the knight and the open d-file gives you active play. In practice, Black scores better than White even in this line, so don't fear the queen trade. Your king can find safety, and your pieces develop naturally. If White plays dxe5, you should reply 4…dxe4 and be ready to recapture the queen with your king; the pawn on e5 and your knight on c6 give you a solid, equal game.

By far the most popular move is exd5 (23,826 games, White scores only 45.1%). After 4.exd5, you can recapture with 4…Qxd5 — the queen comes out early but is safe, since White cannot easily attack it. You'll develop your kingside and castle quickly. The other frequent move is Bb5 (6,175 games, White scores 48.5%), pinning your knight on c6. Here you can respond with 5…dxe4, breaking the centre and threatening a discovered attack on the bishop. All of White's main options keep the position roughly level — Black has nothing to fear. But pay special attention to the next section, because some White moves cost them dearly.

The engine has identified three moves that hand Black an advantage. If White plays Nf3 (4,160 games, White scores just 42.2%), the engine calls it an inaccuracy worth about 0.7 pawns in your favour — White should have played dxe5 instead. You can meet 4.Nf3 with 4…dxe4 5.Nxe5? (a natural but losing move — 5.Qxd5 is necessary) and after 5…Nxe5 6.Qxd8+ Kxd8, Black has won a pawn. The real blunders are f3 (3,608 games, White scores 40.7%) and Bd3 (2,422 games, White scores 43.3%). Against 4.f3, the engine says White loses about 1.4 pawns — you can simply play 4…dxe4 5.fxe4 Bc5, developing with a huge lead. Against 4.Bd3 (a mistake worth about 1.1 pawns), play 4…dxe4 5.Bxe4 Nf6, winning a tempo and leaving White's centre shaky. When you see one of these subpar moves, strike immediately by capturing on e4 or attacking the weakened centre.

The numbers tell a clear story: Black is not just surviving in the Kennedy c3 — Black is winning. In the full database of 58,717 games, Black wins 50.9% of the time while White wins only 45.0%. That's a remarkable statistic for an opening that the engine rates as dead equal. The reason is practical: White has many tempting ways to go wrong, while Black's plan is straightforward. The worst-scoring White moves in practice are f3 (White wins only 40.7%) and Nf3 (White wins 42.2%). Even the most popular move exd5 gives White a below-par 45.1% score. When you know the proper responses, you'll be adding to that Black win percentage. This variation suits players who enjoy early central tension, tactical chances, and positions where one mistake by the opponent can cost the whole game.