How to Play Black in the Italian Game: Paris Defense – 4...exd4

The Paris Defense (3...d6) tells White you are not rushing into the knife-fight of a Two Knights or a Greco Gambit. You keep a compact, solid pawn chain and wait for White to overreach. After 4.d4 exd4, you have given back the centre, but your structure is clean: you still have a pawn on d6 controlling e5, your light-squared bishop can develop to e6 or g4, and kingside castling is uncomplicated. Your main drawback is a slight space disadvantage — White's knight on f3 and bishop on c4 eye your kingside, and the engine says you are slightly worse. Your job is to complete development accurately and exploit any aggressive move White makes. Most of White's tries here are harmless if you respond correctly. The statistics back this up: White scores only 49.6% after the quiet O-O, so when White plays cautiously you have nearly equal chances.

The computer's preferred continuation is 5.Nxd4 Be7 6.Nc3 Nf6. That knight on d4 is a big piece — it pressures f5 and b5, and can retreat or jump again later. You should aim to reach a position where your pieces flow naturally: bishop to e7 (safe, flexible), knight to f6, castle short. Notice that White hasn't attacked anything directly; you have time to finish development. This line is also the most played in practice by a huge margin — over 1.27 million games. White scores 52.1% from here, so you are still slightly worse but the game is very much playable. Trust the standard developing moves and do not panic if White's knight looks threatening on d4.

Your opponents will not always find Nxd4. Three common tries are outright inaccuracies or worse, and each one gives you a chance to improve your position. Here is what to watch for: - c3 (293,937 games, White scores 54.1%): This is an inaccuracy that loses roughly 0.6 pawns. White hopes to recapture with a pawn and open lines, but you can simply maintain your extra pawn or develop quickly. - Ng5 (235,445 games, White scores 48.9%): Also an inaccuracy losing about 0.6 pawns. This move threatens nothing serious after ...d6 covers f7 a second time. You can ignore the threat and continue developing — the engine says Nxd4 was better for White. - Bxf7+ (35,863 games, White scores only 45.1%): This is a blunder that costs roughly 3.9 pawns. After you recapture with ...Kxf7, White has traded a bishop for a pawn and your king is safe. Black's winning chances skyrocket — White's score here (45.1%) is the worst of any move in the database. If you see this sacrifice on the board, accept it and enjoy your nearly winning position.

The two safer options for White are O-O (341,689 games, 49.6% for White) and h3 (28,531 games, 48.3% for White). These are not mistakes — they are just slower. After O-O, you can develop normally with ...Nf6 and ...Be7, and White's score barely edges above 50%. After h3, White's score is actually below 50%, meaning Black scores slightly more than half the points. Both moves reveal that White is happy to play a quiet positional game. That is fine for you too: finish development, contest the centre with ...Bg4 or ...Be6 when appropriate, and do not feel forced to grab back the pawn on d4 immediately. Your position is solid, your king will be safe, and the middlegame will be a fair fight.