The Opera Game: Development With a Body Count

The short version: Paul Morphy develops his pieces. Black develops excuses. Morphy sacrifices one rook, offers the queen, and uses the other rook to deliver mate on move seventeen. This is not merely a famous old attack. It is the cleanest available demonstration that development is stored tactical energy.

Some chess games are famous because champions played them. Some are famous because the final combination is beautiful. Some survive because teachers can use them to explain ten ideas without changing the board.

The Opera Game is all three.

Paul Morphy played White in Paris in 1858 against two opponents consulting over Black’s moves: Karl II, Duke of Brunswick, and Count Isouard de Vauvenargues. The game is traditionally associated with an opera performance, hence the title that has followed it for more than a century and a half.

The historical scenery is excellent. The chess is better.

Morphy wins in seventeen moves without relying on an obscure trap. He occupies the center, develops with threats, castles, connects the rooks, opens the only file that matters, and refuses to spend time rescuing material when the king can be mated instead.

This is the extremely famous game for Game #4. We are not easing into the classics. We are walking directly into the opera box and asking why Black’s queenside is still wearing pajamas on move twelve.

Replay the Opera Game

This Lichess Study preserves the historical score and gives us an interactive board.

Open the Opera Game study on Lichess for the original study page.

The complete score

[Event "Opera Game"]
[Site "Paris"]
[Date "1858.??.??"]
[Round "-"]
[White "Paul Morphy"]
[Black "Duke of Brunswick and Count Isouard"]
[Result "1-0"]
[ECO "C41"]
[Opening "Philidor Defense"]

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# 1-0

Seventeen moves. White ends with a queen available for capture, one rook already sacrificed, and every remaining piece doing exactly the job the position requires.

What Stockfish adds—and what it does not

Morphy did not have Stockfish. The game does not need an engine to become good.

The engine is useful here for a different reason. Famous games accumulate slogans. “Develop your pieces.” “Do not move too many pawns.” “Morphy sacrificed everything.” Those slogans are directionally correct and tactically incomplete.

I ran the same local Stockfish 18 harness introduced in The Queen Walked Into the Fork: five million nodes for every search, three candidate lines before each played move, and a separate search after each move. The famous final combination produced a disagreement between those searches, so I did not select the prettier number. I ran two targeted 50-million-node searches on the critical position.

That deeper audit matters at move fifteen. More on that when the orchestra reaches maximum volume.

All engine scores below are from White’s perspective.

1.e4 e5: claim the center

Morphy opens 1.e4. Black mirrors with …e5.

Both sides occupy a central square and open lines for a bishop and queen. There is nothing mystical here. White has used the first move to improve two pieces that have not moved yet.

That is the first theme of the game: development begins before a piece leaves its original square. A pawn move can create a future route. A useful move does more than one job.

2.Nf3 d6: defense without development

White plays 2.Nf3, developing a knight and attacking e5.

Black protects the pawn with 2…d6, entering the Philidor Defense. Stockfish prefers 2…Nc6, which develops a piece while defending e5. The engine moves from roughly +0.3 under its preferred defense to +0.6 after …d6.

That is not a losing mistake. Philidor structures are playable. The comparison simply introduces an efficiency question:

Can I answer the threat while developing a piece?

Black answered the threat. Black did not develop.

3.d4: Morphy asks for the whole center

Morphy immediately plays 3.d4.

White now has pawns on e4 and d4, two pieces with open diagonals, and a knight attacking the center. Stockfish agrees with the move.

Black can clarify the center with 3…exd4 and remain only modestly worse. Instead, Black plays 3…Bg4?!.

The bishop pins White’s f3-knight to the queen. It looks like development with tempo. Stockfish’s evaluation grows to about +1.0, roughly half a pawn worse than the immediate exchange.

Why does a developing move deserve suspicion?

Because it does not solve the center. White’s d-pawn is still attacking e5. Black has developed a bishop outside the pawn chain while leaving the central tension and queenside pieces unresolved. The pin is real, but it is trying to make a tactical detail compensate for a strategic delay.

4.dxe5: call the bluff

Morphy takes the pawn with 4.dxe5.

If Black responds automatically with 4…dxe5?, White can play 5.Qxd8+ Kxd8 6.Nxe5, winning material because the pinned knight was never actually unable to move after the queens came off.

Black first exchanges the knight with 4…Bxf3. White recaptures 5.Qxf3, and only then comes 5…dxe5.

The move order prevents the immediate queen trade tactic, but look at what is left on the board:

• White’s queen is active on f3.
• White’s light-squared bishop has an open path.
• Black’s kingside knight and queenside pieces remain home.
• Black has exchanged the bishop that moved twice for a knight that moved once.

The material is equal. White’s position is not.

This is Dan Heisman’s question about what remains after a trade appearing in an 1858 evening dress.

6.Bc4: another piece, another threat

Morphy develops 6.Bc4, aiming directly at f7.

Stockfish slightly prefers the immediate 6.Qb3, but Bc4 is the move that makes the coming queen attack powerful. The bishop and queen will cooperate against f7. Development is not a ceremonial checklist; Morphy places the bishop on a square with tactical meaning.

Black plays 6…Nf6?, finally developing a knight but blocking the queen’s defense of f7 and allowing the double attack. Stockfish’s evaluation rises from a best-defense neighborhood around +1.4 to approximately +2.3.

The engine’s defensive choice is the strange-looking 6…Qf6, meeting queen pressure with queen defense. Black is already solving problems rather than building a position.

7.Qb3: two targets, one queen

White plays 7.Qb3.

The queen attacks f7 with the bishop and attacks b7 along the diagonal. Black cannot casually defend both pawns with one ordinary move.

Black chooses 7…Qe7, protecting f7. Morphy could capture b7 immediately. Stockfish prefers 8.Qxb7, but Morphy instead plays 8.Nc3.

This is one of the instructive details that disappears when the game is summarized as a tactical explosion. Morphy declines a pawn to bring out another piece.

The knight on b1 becomes active. Queenside castling becomes possible. The rook on a1 gains a future route to d1. White’s advantage is not the b7-pawn. White’s advantage is that the entire army can arrive before Black’s queenside wakes up.

8…c6: Black defends the pawn, White builds the attack

Black plays 8…c6, finally protecting b7.

White responds 9.Bg5. Stockfish considers Be3 similarly strong, but the game move pins the f6-knight to the queen on e7 and adds another developed piece to the attack.

Count the participation:

• White queen: active on b3.
• White bishops: c4 and g5.
• White knights: c3 and one exchanged.
• White king: one move from queenside castling.
• White rook on a1: waiting for the d-file.

Black’s bishop on f8, knight on b8, rook on a8, and king on e8 have not moved.

Material remains equal. Time does not.

9…b5?: the pawn that opens the stage

Black attacks the bishop with 9…b5?.

The move looks energetic. It gains space and forces the c4-bishop to respond. Stockfish instead wants 9…Kd8, an ugly admission that king safety matters more than castling rights. After …b5, the evaluation rises from about +2.2 to +3.4.

The problem is not merely “too many pawn moves.” The problem is that …b5 creates a forcing tactical target on b5 and opens lines toward Black’s uncastled king.

Morphy does not retreat the bishop.

10.Nxb5!: destroy the pawn shield before it exists

White plays 10.Nxb5!.

The knight captures the pawn and offers itself. Black can decline with the forcing 10…Qb4+, after which White remains clearly better but Black resists. The historical move is 10…cxb5?.

Black’s c-pawn captures the knight. Material arithmetic says White has lost a knight for two pawns after the next capture.

Position arithmetic says the c-file and diagonals are opening around a king that cannot castle comfortably.

Stockfish moves from approximately +3.4 under the best defense to +5.3 after accepting the knight.

11.Bxb5+: develop by capturing with check

Morphy’s bishop takes on b5 with check.

This is the perfect attacking move because every word in its description matters:

• the bishop remains active;
• White recovers a pawn;
• the king is checked;
• Black must respond;
• and another line toward the center opens.

Black blocks with 11…Nbd7. The b8-knight finally develops, but it develops into a pin and becomes the plug in the d-file.

White has not rushed to recover the sacrificed material completely. White has used the sacrifice to purchase forcing moves.

12.O-O-O!: castle into the attack

Morphy castles queenside.

The move makes the king safer and places the rook from a1 directly on d1, opposite Black’s queen and king infrastructure. It is king safety, development, and rook activation in one notation symbol.

This is why castling can be an attacking move. The king steps away. The rook steps onto the only file that matters.

Stockfish gives White roughly +5.4. Black answers with 12…Rd8, meeting rook with rook and reinforcing the pinned knight on d7.

Black has finally placed a major piece on the file. Morphy removes it by force.

13.Rxd7!: the first rook goes in

White plays 13.Rxd7!.

The rook from d1 captures the knight on d7. Black’s rook from d8 captures it with 13…Rxd7.

Morphy has traded a rook for a knight. If the calculation stops there, White has lost an exchange.

But the calculation cannot stop while Black’s king is uncastled, the d-file is open, and White has another rook.

Stockfish approves Rxd7. The move is not romantic decoration attached to a winning game. It is the engine’s first choice.

Black’s recapture is natural but inaccurate. Stockfish prefers 13…Nxd7, keeping the rook protected on d8 and limiting the invasion. After 13…Rxd7, White’s advantage grows to about +7.1.

14.Rd1: the understudy walks onstage

White’s remaining rook began on h1. The squares g1, f1, and e1 have been cleared by development and castling. It slides all the way to d1.

This is the move that explains the whole game.

The first rook could sacrifice itself because development gave the second rook a route. If White’s bishop still sat on f1 or king remained on e1, Rd1 would not exist. The combination on move thirteen was funded by moves six, eight, nine, and twelve.

Development is stored tactical energy.

Black moves the queen to 14…Qe6?, stepping away from the d-file and defending. Stockfish prefers …Qb4, seeking counterplay with check. The game move leaves White around +8.7 in the initial report.

Now comes the combination printed in chess books, videos, lessons, studies, and probably at least one tasteful opera-themed necktie.

15.Bxd7+!: the deep audit

The bishop from b5 captures the rook on d7 with check.

Our ordinary five-million-node report behaved strangely here. Its MultiPV search valued Bxd7+ near +8.8, while the independent search after the move briefly reported only +3.7. On Black’s turn, one MultiPV line even ranked the losing …Nxd7 above …Qxd7, while the separate post-move search immediately found mate.

That is not a contradiction to hide. Fixed-node searches can reach different horizons, and MultiPV divides attention among several lines. So I reran the position at 50 million nodes before the move and 50 million nodes after it.

The deeper result:

• 15.Bxd7+ is Stockfish’s first choice at approximately +8.93.
• 15.Bxf6 follows at approximately +6.84.
• 15.Qc3 follows at approximately +5.46.

The famous move is not merely pretty. It is best under the deeper search.

Black has one non-mating reply

After 15.Bxd7+, Black has three legal responses in the audited lines.

15…Qxd7

This is the only move that avoids forced mate. Stockfish still gives White approximately +9.27 after a 50-million-node search. Its principal variation begins:

15…Qxd7 16.Qb8+ Ke7 17.Qxe5+ Kd8 18.Bxf6+ gxf6 19.Qxf6+ Kc8 20.Rxd7

Black survives the immediate mating net but loses enormous material under continued checks.

15…Ke7

Stockfish finds mate in four:

15…Ke7 16.Qb4+ Qd6 17.Qxd6+ Kd8 18.Qb8+ Ke7 19.Qe8#

15…Nxd7??

This is the historical move. It permits mate in two.

The knight from f6 captures the bishop on d7. Black now appears to have removed the checking piece and defended the back rank.

Morphy offers the queen.

16.Qb8+!: please take the queen

White’s queen moves to b8 with check.

The queen is directly available to Black’s knight on d7. Black plays 16…Nxb8.

Calling this a queen sacrifice is correct and incomplete. The queen is not exchanged for vague attacking chances. It is offered to force the knight away from d7 and onto b8, where it can no longer control the d8-square.

The queen’s value is irrelevant because the game has one move left.

This is Heisman’s trade lesson at maximum volume: do not ask only what leaves the board. Ask what remains.

After the queen and knight leave their previous functions:

• Black’s king remains on e8.
• White’s rook remains on d1.
• The d-file remains open.
• No Black piece can interpose on d8.
• The king has no legal escape.

The queen was worth nine points. Checkmate is worth the game.

17.Rd8#: the quietest possible ending

The rook moves from d1 to d8.

Checkmate.

No capture. No promotion. No second queen. The remaining rook simply occupies the square that sixteen earlier moves prepared for it.

The final position is visually simple because the preceding game removed every complication that could stop the rook.

That simplicity is the masterpiece.

The development ledger

It is tempting to say Black lost because Black moved too many pawns. Let us be more exact.

By the time Morphy castles on move twelve, White has:

• moved both central pawns;
• developed both knights, even though one has been sacrificed;
• developed both bishops to attacking squares;
• activated the queen;
• castled;
• and placed a rook on the open d-file.

Black has:

• moved five different pawns;
• developed the light-squared bishop and exchanged it;
• developed the g8-knight;
• moved the queen to e7;
• left the f8-bishop home;
• left the a8-rook home;
• and left the king in the center.

The problem is not a moral prohibition against pawn moves. Black’s pawn moves failed to create enough safety, development, or counterplay for the time they consumed.

Morphy’s moves accumulated. Black’s moves happened.

What beginners are usually told

The standard lessons from the Opera Game are correct:

1. Develop pieces quickly.
2. Fight for the center.
3. Castle.
4. Do not move the same piece repeatedly without a reason.
5. Open lines against an uncastled king.
6. Bring every piece into the attack.

But the game becomes more useful when those principles stop sounding independent.

Morphy does not develop because a checklist says “knights before bishops.” He develops because each new piece creates another forcing possibility. He does not castle merely to hide the king. He castles because the rook lands on d1. He does not sacrifice because romantic chess demands a blood offering. He sacrifices because the second rook can replace the first.

The principles are one system:

Use time to place pieces where opening the position will make them dangerous.

What stronger players can still study

The Opera Game is often assigned to beginners, which can make experienced players treat it like a children’s book they have outgrown.

That is a mistake.

Stronger players can study the precision of the move ordering:

• Why 8.Nc3 instead of immediately taking b7?
• Why does 9…b5 make the knight sacrifice possible?
• Why is 12.O-O-O stronger than collecting material?
• Why can the first rook disappear?
• What changes if Black chooses 13…Nxd7?
• Why does 15…Qxd7 avoid mate but not defeat?
• Why is the queen offer on b8 forcing rather than speculative?

The moves are easy to remember. The dependencies between them are the real study.

The honest engine verdict

Stockfish does not declare every Morphy move uniquely best.

It prefers 6.Qb3 to 6.Bc4 by a few tenths. It prefers taking b7 on move eight to Morphy’s development. It offers Black defensive resources that the consultation team did not find: …Qf6, …Kd8, …Qb4+, …Nxd7 on move thirteen, and …Qxd7 on move fifteen.

That does not diminish the game. It makes the achievement specific.

Morphy consistently chose moves that developed, forced, or connected. Black repeatedly chose moves that solved one local problem while leaving the king and queenside behind. Once the position opened, White’s pieces arrived with tempo. Black’s pieces required invitations they never received.

The final combination is objectively winning. The deeper engine confirms 15.Bxd7+ as the strongest move. The historical 15…Nxd7?? is not forced, but every alternative is already grim; only …Qxd7 prevents immediate mate, and Stockfish evaluates that position around nine pawns in White’s favor.

So the sober verdict is somehow still theatrical:

Morphy did not force Black to choose the prettiest death.

He did build a position in which Black was choosing among deaths.

Why this game remains famous

The Opera Game survives because nothing in it is wasted.

The queen attack explains development. The knight offer explains open lines. Castling explains rook activation. The first rook sacrifice explains coordination. The bishop capture explains deflection. The queen offer explains that material values stop mattering when mate is forced. The final rook move explains every earlier clearance.

You can show the game to someone who learned the rules yesterday. You can place it under Stockfish 18 at fifty million nodes. Both viewers find something.

That is not merely fame.

That is repertoire.

Historical and game references

• The Opera Game on Lichess
• Opera Game historical overview
• Paul Morphy overview
• Stockfish 18 release