The Philidor Defense is what Black plays after 1.e4 e5 2.Nf3 d6. Black defends the e5 pawn with a pawn instead of the knight, keeps a solid wall in the centre, and accepts that the position will be a little cramped. The pages we opened describe it well: Wikipedia calls it a solid but passive choice for Black, and chess.com calls it passive but playable. They do not tell you what White actually plays against it at your rating, and that is what decides which Philidor you will face. We ran the maia3-5m checkpoint, see how Maia chess actually works, on two positions at seven rating bands from 1000 to 2200 and graded every legal move with Stockfish 18 at depth 20. Percentages on this page are the model's predicted share of human moves at that rating. They are not game database frequencies and they are not win rates.
Every evaluation is from White's side, so +0.65 means White is better by about two thirds of a pawn. The two dataset positions are the one after 2...d6 with White to move, and the one after 3.Bc4 h6 with White to move. The line checks are separate local runs of ten seconds per line, which we call a local Stockfish check, so a number can differ by several hundredths from other methods.
What the Philidor is, and where the name comes from
The position after 2...d6 is ECO C41. The opening is named after the 18th century player François-André Danican Philidor. Chess.com says he analysed it in his 1749 treatise, and Wikipedia cites that book as well. Wikipedia adds that Philidor promoted 2...d6 as an alternative to the common 2...Nc6, and that his original idea was to challenge White's centre with the pawn thrust ...f5. That idea survives as the Countergambit, which modern sources regard as inferior, and we check it below.
The reputation is modest. Chess.com lists the drawbacks as a rather passive central presence and the blocked dark squared bishop, and 365chess says the biggest disadvantage is that the bishop gets trapped behind the d6 pawn. On the other side, chess.com notes that 2...d6 avoids the Ruy Lopez that follows 2...Nc6 3.Bb5. A ChessBase article on Alekhine is a counterweight: it found 15 games where he played the Philidor and argued his results did not support the passive reputation. At the top the opening is rare: Wikipedia's text, from around 2017, says no top players employ it with regularity, and we did not confirm any current elite usage.
The first decision: what White plays after 2...d6
Stockfish ranks 3.d4 first at +0.65. Then comes 3.Nc3 at +0.42, 3.Bc4 at +0.37 and 3.h3 at +0.26. The gap between the best move and the one people actually choose is the interesting part. Here is what the model predicts people play.
| Rating | 1000 | 1200 | 1400 | 1600 | 1800 | 2000 | 2200 |
|---|---|---|---|---|---|---|---|
| 3.Bc4 (+0.37) | 40.9% | 46.4% | 48.1% | 47.7% | 46.2% | 42.0% | 31.9% |
| 3.d4 (+0.65) | 21.8% | 22.3% | 24.7% | 28.6% | 33.7% | 41.8% | 56.0% |
| 3.Nc3 (+0.42) | 21.1% | 16.1% | 11.9% | 9.0% | 7.0% | 6.0% | 5.6% |
From 1000 to 2000 the model's likeliest move is 3.Bc4, not the engine's 3.d4. At 1000 it is predicted 40.9 percent of the time against 21.8 for 3.d4. 3.Bc4 peaks at 48.1 percent at 1400, and at 2000 the two are almost level, 42.0 against 41.8. Only at 2200 does 3.d4 become the likeliest move, at 56.0 percent, with 3.Bc4 down to 31.9. The share for the engine's choice rises at every step, 21.8, 22.3, 24.7, 28.6, 33.7, 41.8 and finally 56.0. Averaged over both dataset positions, the probability of the engine's best move, which we call findability, climbs from 19.8 percent at 1000 to 43.7 percent at 2200.
As of October 2026, Chessiverse publishes rating banded Philidor data too, with 3.Bc4 as its top reply at 1200 and 3.d4 at 2500, so the broad picture agrees. Our contribution is different: the likeliest human move set against the engine's grade for every legal move, so you can see how much a typical choice at your rating costs.
The cost here is small. 3.Bc4 is +0.37 against +0.65, a difference of 0.28 of a pawn, and the sources agree the move is respectable: Wikipedia names it as White's main alternative to 3.d4, and the local check gives 3.Bc4 Nf6 as +0.37 and 3.Bc4 Be7 as +0.33. 3.Nc3 is the one that fades, from 21.1 percent at 1000 to 5.6 at 2200. For a player choosing a repertoire, the lesson is practical. Below 2000 you will mostly face 3.Bc4, so that is the line to prepare, and 3.d4 matters more as the opposition improves.
Why people play 3.Bc4 so often
It is the Italian Game move, and it develops a piece, eyes f7 and keeps the centre pawn structure flexible. Black's usual 2...d6 is a set up that does not stop it. White also gets traps. Wikipedia says 3.Bc4 is White's route to a possible Légal trap, and chess.com says Black must avoid Légal's mate after it. If you want the same bishop move in its more usual setting, the Italian Game page has the data for 2...Nc6.
After 3.Bc4: the second dataset position
The second position follows 3.Bc4 h6. We picked it because at 1400 the model's likeliest Black reply to 3.Bc4 is 3...h6, at roughly a quarter of the time, and the model gave 3.Bc4 itself 48.1 percent. That probability exists for the 1400 band only, so we do not quote it for other ratings. White now has a choice of calm moves, and the engine wants the centre.
Stockfish ranks 4.d4 first at +0.85. 4.O-O is +0.49, 4.c3 is +0.39, 4.Nc3 is +0.33 and 4.d3 is +0.20. Here is how the bands choose.
| Rating | 1000 | 1200 | 1400 | 1600 | 1800 | 2000 | 2200 |
|---|---|---|---|---|---|---|---|
| 4.d4 (+0.85) | 17.8% | 19.6% | 22.7% | 25.3% | 26.5% | 27.4% | 31.4% |
| 4.O-O (+0.49) | 23.6% | 21.7% | 19.4% | 18.4% | 19.0% | 20.7% | 21.6% |
| 4.d3 (+0.20) | 22.2% | 20.9% | 17.9% | 14.8% | 12.6% | 11.6% | 11.7% |
| 4.Nc3 (+0.33) | 20.6% | 20.4% | 19.5% | 17.7% | 15.0% | 11.9% | 9.8% |
| 4.c3 (+0.39) | 8.6% | 10.1% | 11.9% | 13.9% | 16.0% | 17.5% | 17.5% |
This table is flatter. At 1000 and 1200 White's choice is split almost evenly between castling, 4.d3 and 4.Nc3, with 4.O-O leading at 23.6 and 21.7 percent. From 1400 the engine's 4.d4 becomes the likeliest move at 22.7 percent and keeps growing, reaching 31.4 at 2200. The lower bands play more quietly than the engine wants. The slow 4.d3 loses most of its support, from 22.2 percent at 1000 to 11.7 at 2200.
Findability of the best move in this position runs from 17.8 percent at 1000 to 31.4 at 2200, lower than after 2...d6, which fits a position with several reasonable plans and no forcing one. The practical reading is that after 3.Bc4 h6 you should expect a quiet game below 1400, and a centre fight from 1400 up.
The main line: 3.d4 exd4 and the Nf6 setups
When White does play 3.d4, Black has several systems. As of October 2026, chess.com says the Exchange Variation, 3...exd4, is reached in more than half of Philidor games. The usual route is 4.Nxd4 Nf6 5.Nc3 Be7, which Wikipedia names the Antoshin Variation, and 365chess gives the main line as 6.Be2 O-O 7.O-O Re8 8.f4 Bf8 9.Bf3 c5 10.Nb3 Nc6. After 3...exd4 our local Stockfish check gives 4.Nxd4 as the best move at +0.61.
The other main reply is the Nimzowitsch Variation, 3...Nf6, named in chess.com's list. Here Black can fall into a small error: 4.dxe5 Nxe4 5.Qd5 is +0.79 in our local Stockfish check, the queen attacking e4 and f7 together. It is not a refutation, but it shows why the knight move needs care. A calmer setup is 4.Nc3 Nbd7, where our check gives +0.56 for White. That is the sort of position the Philidor is meant to reach: a solid structure, a modest White edge, and no immediate tactics. As of October 2026, chess.com adds that after 3...Nf6 White scores about 58 percent with 4.dxe5 in its database, a figure that is not split by rating, so we treat it as general background.
The Hanham Variation: 3.d4 Nd7
The Hanham Variation, 3...Nd7, is named after the American master James Moore Hanham. It keeps the e5 pawn defended without committing the g8 knight, and it avoids the exchange on d4. It also has a well known pitfall. After 4.Bc4, the natural looking 4...Be7? runs into 5.dxe5, and our local Stockfish check puts White at +2.23. Black cannot recapture on e5 without loosening f7, which the c4 bishop is already aiming at, and that is the kind of detail that decides games at club level.
The Légal trap, checked
A well known Philidor trap comes after 3.Bc4. The setup is 3...Bg4 4.Nc3 g6, and both of Black's moves are plausible-looking developing moves. Then White plays 5.Nxe5. Our local Stockfish check gives the position after 4...g6 as +1.86 for White, with 5.Nxe5 the best move.
- 5...Bxd1. Black takes the queen. Then 6.Bxf7+ Ke7 7.Nd5# is checkmate. We verified the final position as mate with local Stockfish, so the trap is sound as written.
- 5...dxe5. Black declines. Then 6.Qxg4 takes the bishop back, White is a clean pawn up, and the local check puts White at +1.97.
Wikipedia's Légal trap article gives the game Légal against Saint Brie as the sequence 1.e4 e5 2.Nf3 d6 3.Bc4 Bg4 4.Nc3 g6 5.Nxe5 Bxd1 6.Bxf7+ Ke7 7.Nd5#, played at rook odds. The article itself flags its 1750 date as probably wrong, so we leave the year out and say only that the trap pattern bears Légal's name. Its main illustration is a related move order with 2...Nc6, so the Philidor version is the same idea reached by a different route.
The lesson for Black is plain. When White has a bishop on c4 aimed at f7, taking the queen on d1 is not free. If you play the Philidor, the cure is simple: do not put a bishop on g4 while the c4 bishop is aimed at f7, and meet 3.Bc4 with 3...Be7 or 3...Nf6, which our check gives as +0.33 and +0.37.
The Opera Game: 3...Bg4 against 3.d4
The Philidor's best known game is Paul Morphy's Opera Game, which Wikipedia calls one of the most famous games ever played. Wikipedia describes it as a consultation game in a box at the Salle Ventadour in Paris in October or November 1858, played by Morphy as White against the Duke of Brunswick and Count Isouard. The moves went 1.e4 e5 2.Nf3 d6 3.d4 Bg4 4.dxe5 Bxf3 5.Qxf3 dxe5 6.Bc4 Nf6 7.Qb3 Qe7 8.Nc3 c6 9.Bg5 b5 10.Nxb5 cxb5 11.Bxb5+ Nbd7 12.O-O-O Rd8 13.Rxd7 Rxd7 14.Rd1 Qe6 15.Bxd7+ Nxd7 16.Qb8+ Nxb8 17.Rd8#. Sources disagree on which opera was being performed, so we leave it out.
Chess.com says Black's problems started with the weak third move 3...Bg4, and Chessiverse and 365chess also flag it as dubious. Our local check agrees without drama: after 3.d4 Bg4 White is +1.17 with 4.dxe5, and after 6...Nf6 the position is +2.24 with the quiet 7.Qb3. So Black was already in serious trouble before the famous sacrifices began. Black spent moves on a pin and fell behind in development. That is the useful moral for a Philidor player: the structure is solid, but only if you spend your moves on development and not on pins.
The Philidor Countergambit: 3...f5
Philidor's own idea was to hit the centre with ...f5, and Wikipedia calls the Countergambit 3.d4 f5 the most aggressive approach. Chess.com and 365chess say it is now regarded as inferior or out of favour. Our local Stockfish check agrees strongly: after 3...f5, White is +1.96 with 4.Bc4. There is nothing wrong with playing it for fun against a weaker opponent, but a player who sets out to use the Philidor should not choose it as a main line.
What the numbers say about how to play it
Put the two tables together. As Black at 1000 to 1800, you will meet 3.Bc4 about 41 to 48 percent of the time, so learn the Légal trap first and then the safe replies 3...Be7 and 3...Nf6. Both leave White near +0.35, which is a comfortable result for the defender. After 3.Bc4 h6 4.d4 is the engine's move, but lower rated players will often castle or play d3, and a solid reply will keep you level.
As White, the data suggest a split. If you want the engine's best score, 3.d4 at +0.65 is the move, and 4.Nxd4 after 3...exd4 keeps it at +0.61. If you want a quick developing game with traps, 3.Bc4 gives up only 0.28 of a pawn and the model says it is what most players choose below 2000. It also sets up the Légal trap, where Black's natural 3...Bg4 followed by ...g6 walks into mate. 3.Nc3 is playable at +0.42 but loses popularity as ratings climb, from 21.1 to 5.6 percent.
Plans matter more than move lists, the argument in why you should stop memorizing openings and learn plans. In the Philidor, Black's plan is to complete development with ...Nf6, ...Be7 and castling, keep ...d6 and ...e5 as a pair, and break with ...exd4 or ...c6 and ...d5 only when ready. White's plan is to take space with d4, trade into a lead in development, and keep the bishops pointed at f7 and e5. For other ways to meet the same 1.e4 e5 start, see the Petrov Defense page, chess openings by rating and the Ruy Lopez page.
Is the Philidor Defense a good opening?
It is a reasonable, slightly passive choice, and the evidence is consistent. Wikipedia calls it solid but passive, chess.com calls it passive but playable, and our own checks put the sound main lines between +0.33 and +0.65 for White. One chess.com forum poster put it this way: there is nothing wrong with the Philidor, it is not great, but it is fully playable. We cannot check that opinion, but it matches the numbers.
Choose it if you want a structure you understand and are happy to defend for a while. Avoid it if you want to seize the initiative as Black, because the d6 pawn blocks the bishop and White keeps the space.
Where other resources are better
This page is narrow on purpose. The dataset holds two positions, and Maia3 5m is a small model of human choice, so its percentages are predictions and can be wrong at the margins. The line checks are quick local runs, not a substitute for an opening database or a repertoire course. We have no game results by rating, so we cannot say how Black scores at your level. Chessiverse publishes rating banded results and is the better source for that, though its own game counts do not add up cleanly. Wikipedia is the better reference for the full list of named variations, and the Lichess explorer, which we did not open, is the place for real game counts. None of the pages we opened compares the engine's best move with the likeliest human move at each rating, which is the gap this one fills.
What to take from it
The engine prefers 3.d4 at +0.65, but Maia3 predicts 3.Bc4 as the likeliest move from 1000 to 2000, at 40.9 to 48.1 percent, and only at 2200 does 3.d4 lead, at 56.0 percent. After 3.Bc4 h6, castling and d3 are likelier below 1400 and 4.d4 is likeliest from 1400 up. The traps are real: the Légal mate on d5 is verified, 3...Bg4 costs Black +1.17 and the Hanham 4...Be7? costs +2.23. Maia3 is a model of human choice trained on online games, not a count of real games, and every percentage above should be read that way.
Frequently asked
- Is the Philidor Defense good?
- It is playable and a little passive. Wikipedia calls it a solid but passive choice for Black, and chess.com calls it passive but playable. Our local Stockfish check backs the second half: after 3.d4 exd4 White is +0.61, and after 3.Bc4 Nf6 it is +0.37, both modest. The risk is in the sidelines, where careless Black moves such as 3...Bg4 or 3...f5 are punished by +1.17 and +1.96.
- What does White play against the Philidor at 1200?
- On our model, 3.Bc4 is the likeliest move at 46.4%, with 3.d4 at 22.3% and 3.Nc3 at 16.1%. These are Maia3 predictions of human play from a model trained on online games, not game counts. The engine's top choice is 3.d4 at +0.65, so at 1200 the likeliest human move is the third best of the three main ones.
- What is the Légal trap in the Philidor?
- After 1.e4 e5 2.Nf3 d6 3.Bc4 Bg4 4.Nc3 g6, White plays 5.Nxe5. If Black takes the queen with 5...Bxd1, then 6.Bxf7+ Ke7 7.Nd5# is checkmate, which we verified as mate with local Stockfish. If Black declines with 5...dxe5, White plays 6.Qxg4 and is +1.97. The pattern bears Légal's name; we do not tie it to a specific date.
- What is the Hanham Variation?
- Hanham is the Philidor setup that starts 3.d4 Nd7, named after the American master James Moore Hanham. It keeps the e5 pawn defended and avoids exchanging on d4. It has a pitfall of its own: after 4.Bc4 Be7?, 5.dxe5 is +2.23 for White in our local Stockfish check, because Black cannot recapture on e5 without loosening f7 to the c4 bishop.
- Is the Philidor Defense good for beginners?
- It is simple to learn, and Wikipedia calls it suitable for amateurs seeking a simpler defensive strategy. The cost is a cramped position and a dark squared bishop that the d6 pawn blocks in. If you play it, learn the Légal trap and the 3...Bg4 problems first, because those are the lines where our checks show the biggest penalties.
Want to face the Philidor the way your rating does?
Chessdrive runs the real maia3-5m checkpoint on-device as your opponent at any rating you set from 600 to 2600, so you can play the Philidor from either side against a model that answers 2...d6 the way your band does, then see where the game turned with a separate Stockfish review. Available now on iOS.