Three games are quoted more than any others in chess literature: the Immortal Game (Anderssen–Kieseritzky, London 1851), the Evergreen Game (Anderssen–Dufresne, Berlin 1852), and Morphy's Opera Game (Paris 1858). All three are short, all three end in checkmate against a defender who was ahead on material, and all three were played casually rather than in a tournament hall.

They survive because they are teachable. Each one shows a single idea so cleanly that you can absorb it in one sitting: activity beats material, a double check cannot be blocked, and a lead in development is only worth something if you cash it in. Below is each game with its story, the critical position on a real board you can open in the engine, and what it is actually trying to teach you.

Why the 1850s produced games like these

Opening theory in the mid-nineteenth century barely existed, and the prevailing etiquette pushed players towards attack. Declining a gambit was considered timid; grinding out a small endgame edge was considered dull. The result was a generation of players who developed pieces fast, opened lines at almost any cost, and hunted the enemy king.

That style has an obvious weakness — a well-prepared defender refutes most of it — but it also produced the clearest illustrations ever recorded of what piece activity can do. Modern engines have found improvements for the losing side in every game below. This does not make the games worthless; it makes them honest. You are watching two humans calculate, and the winner is the one who saw one move further into a forcing line.

The Immortal Game: Anderssen vs Kieseritzky, London 1851

Adolf Anderssen, a mathematics teacher from Breslau, played this casual game against Lionel Kieseritzky in London in June 1851, during the period of the first great international tournament — an event Anderssen went on to win. The game was never part of the competition. It was a friendly, played at speed, and Kieseritzky thought enough of it to send the moves to a Paris chess magazine. The nickname arrived a few years later, in print, and stuck.

Anderssen opened with the King's Gambit, offering the f-pawn to rip open lines towards the black king before anyone had finished developing:

1. e4 e5 2. f4 exf4 3. Bc4 Qh4+ 4. Kf1 b5 5. Bxb5 Nf6 6. Nf3 Qh6
7. d3 Nh5 8. Nh4 Qg5 9. Nf5 c6 10. g4 Nf6 11. Rg1

Two things are worth noticing here rather than memorising. First, Anderssen accepted a permanently displaced king on f1 in exchange for the initiative — in this era that was a normal trade. Second, Kieseritzky's early queen excursion won a tempo with the check but then cost several more as the queen was chased around by pawns and knights. Bringing the queen out early is one of the oldest ways to hand your opponent free development.

The move where both rooks were on offer

By move 18 Anderssen had already left a bishop on b5 to be taken and a rook on g1 to be taken. Then he played the move the game is remembered for.

11... cxb5 12. h4 Qg6 13. h5 Qg5 14. Qf3 Ng8 15. Bxf4 Qf6
16. Nc3 Bc5 17. Nd5 Qxb2 18. Bd6
After 18.Bd6 — both white rooks hang, but every white piece is aimed at the black king while Black's queen sits on b2 doing nothing. Play this position →

Count the pieces doing work. White has a knight on d5, a knight on f5, a bishop on d6, a queen on f3 and a pawn on e4, all pointing at the black king still sitting on e8. Black has a queen deep in White's territory, a bishop on c5, and four pieces on their starting squares. The material score says Black is fine. The position says Black is lost.

Kieseritzky took on g1, then took the other rook with check:

18... Bxg1 19. e5 Qxa1+ 20. Ke2 Na6 21. Nxg7+ Kd8 22. Qf6+ Nxf6 23. Be7#

The point of 19.e5 is easy to miss: it does not attack anything, it blocks. The pawn cuts the a1–h8 diagonal so that the black queen, once it lands on a1, can never come back to defend. After that, 22.Qf6+ is a deflection — the knight is dragged to f6 where it no longer guards e7, and the bishop mates on the square the knight just abandoned.

Checkmate. White has given up a bishop, both rooks and the queen, and mates with two knights and a bishop. Play this position →

Add up the final position and Black is ahead by a queen, two rooks and a bishop — the largest material deficit in any famous mating position. The black king on d8 has no legal move because the bishop on e7 gives check, the knight on d5 covers c7 and e7 is defended, and the knight on g7 covers e8.

What the Immortal Game teaches

  • Material only counts if it can reach the fight. A queen on a1 and two rooks on their original squares are worth nothing to a king being mated on d8. This is the core of piece activity versus material.
  • Blocking moves can be as strong as attacking moves. 19.e5 wins the game by taking a defensive route away, not by threatening anything.
  • Chasing an enemy queen is free development. Every pawn move that gains a tempo on the queen also improves White's position permanently.
  • A king in the centre is the target. Kieseritzky never castled. Every combination in the game exists because of that.

The Evergreen Game: Anderssen vs Dufresne, Berlin 1852

A year later, in Berlin, Anderssen played another informal game — this time against Jean Dufresne, a Berlin player and chess author. Wilhelm Steinitz is credited with the description that gave the game its name: an evergreen leaf in Anderssen's laurel wreath. Where the Immortal Game is a wild brawl, the Evergreen Game contains one long forced combination that can be calculated to the end, which is exactly why it has been printed in tactics books for over a century.

The opening was the Evans Gambit, a sharp branch of the Italian Game in which White throws in b4 to deflect the black bishop and then builds a big centre with c3 and d4:

1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Ba5 6. d4 exd4
7. O-O d3 8. Qb3 Qf6 9. e5 Qg6 10. Re1 Nge7 11. Ba3 b5 12. Qxb5 Rb8
13. Qa4 Bb6 14. Nbd2 Bb7 15. Ne4 Qf5 16. Bxd3 Qh5 17. Nf6+ gxf6
18. exf6 Rg8

At this point the game looks genuinely double-edged. Anderssen has given a knight to shatter the pawns in front of the black king, but Black has an open g-file pointing straight at g2 and a queen already on h5.

The quiet move

Instead of defending his knight on f3, Anderssen played a rook move that ignores the threat entirely.

After 19.Rad1 — White adds a rook to the d-file and leaves the knight on f3 to its fate. Play this position →

This is the hardest kind of strong move to find, because it does not check, capture or threaten anything obvious. It simply puts the last inactive piece on the file where the black king will end up. Dufresne took the knight with 19...Qxf3, which was not greed — it threatens 20...Qxg2 mate, since the rook on g8 defends the queen when it lands on g2. Anderssen had to be faster, and he was, by exactly one move.

The combination

19. Rad1 Qxf3 20. Rxe7+ Nxe7 21. Qxd7+ Kxd7 22. Bf5+

Each move has one job. 20.Rxe7+ forces the knight to recapture on e7, where it blocks its own king's escape square and stops guarding d7. 21.Qxd7+ then hands over the queen to drag the king onto d7 — a decoy, not a sacrifice for its own sake. And 22.Bf5+ is a double check: the bishop checks from f5 and, by moving off d3, uncovers the rook on d1 along the file.

Double check on d7. Two pieces give check at once, so blocking and capturing are both impossible — the king has only Ke8 and Kc6. Play this position →

A double check is the most forcing move in chess. You cannot capture both checking pieces in one move and you cannot block two lines at once, so the king must step. That reduces Black's entire position to two legal moves, and both lose:

  • 22... Kc6 23. Bd7#
  • 22... Ke8 23. Bd7+ and now both Kf8 and Kd8 run into 24.Bxe7#

Dufresne chose 22...Ke8, and the game ended 23.Bd7+ Kf8 24.Bxe7#.

Mate by two bishops. The f6 pawn covers g7, the bishop on d7 covers e8 and defends its partner on e7, and Black's own rook blocks g8. Play this position →

What the Evergreen Game teaches

  • Double check ends arguments. When you are looking for a forcing finish, ask which of your pieces can move and reveal a second attacker at the same time. See discovered attacks for the mechanics.
  • Sacrifices are usually about squares, not material. The rook on e7 buys the e7 square as a blockade; the queen on d7 buys the king's position on d7. Neither is speculative.
  • Count tempi in mutual attacks. Anderssen allowed a mate threat against himself because he had calculated that his own forcing sequence arrived first. That calculation is the whole game.
  • Bring the last piece. 19.Rad1 is the move most club players would never consider, and it is the move that makes everything after it work.

The Opera Game: Morphy vs Duke of Brunswick and Count Isouard, Paris 1858

Paul Morphy, a 21-year-old lawyer from New Orleans who had won the first American Chess Congress the year before, was in Paris on the European tour that would make his reputation. He was invited to a performance of Bellini's Norma at the Italian opera house, and shared a box with the Duke of Brunswick and Count Isouard, who insisted on playing him as a consulting pair while the music went on. Morphy, reportedly trying to watch the stage, produced the most quoted teaching game in history.

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

The whole game is a lesson in counting developed pieces. By move 12 Morphy has castled, connected his rooks, and has four pieces in play. His opponents have a queen shuffled to a passive square on e7, a knight pinned on f6, a knight pinned on d7, and a king still on e8 with the f8 bishop and h8 rook completely undeveloped.

After 12.O-O-O — White has given a knight for a pawn and has every piece working. Both black knights are pinned and the king cannot castle. Play this position →

The knight sacrifice on move 10 is worth understanding, because it is not a flourish. Black's 9...b5 tried to gain time by attacking the bishop on c4; Morphy answered by taking on b5 with the knight so that after the recapture, Bxb5+ comes with check and opens the d-file for the rook that is about to arrive there. He paid a knight to keep the enemy king in the centre for four more moves, and four moves is all he needed.

The finish is a chain of removals. Morphy takes off every piece defending d7 and d8 until the rook can land there:

12... Rd8 13. Rxd7 Rxd7 14. Rd1 Qe6 15. Bxd7+ Nxd7 16. Qb8+ Nxb8 17. Rd8#
Mate in 17 moves. The rook checks on d8, the bishop on g5 covers the escape to e7, and Black's own knight and bishop seal the back rank. Play this position →

Note the final queen sacrifice on b8. Its only purpose is to force the knight onto b8, where it blocks the black king's last flight square — the same deflection idea that finished the Immortal Game, run in reverse.

What the Opera Game teaches

  • Every move should add a piece or add a threat. Morphy made seventeen moves and effectively never repositioned a piece twice for no reason.
  • A lead in development is a wasting asset. It has to be converted into an attack while it exists, which is why Morphy was willing to spend material to keep the position open.
  • Open the file your rook is going to use. 11.Bxb5+ and 12.O-O-O are one idea, not two.
  • Pins immobilise defenders. The bishop on g5 does nothing dramatic all game, but it is the reason the knight on f6 never defends d7 or e8.

If you want the same game annotated move by move rather than in summary, we have a longer treatment in Morphy's Opera Game, and a wider look at the player himself in Paul Morphy, chess genius. It is worth remembering that Morphy defeated Anderssen in a match in Paris later that same year — the romantic era's two great names met, and the more disciplined developer won.

Beyond the romantic era: Fischer vs Spassky, 1972

The most famous modern game is Game 6 of the 1972 World Championship in Reykjavík, where Bobby Fischer — a lifelong 1.e4 player — opened with 1.c4 and steered the game into a Queen's Gambit structure that Boris Spassky had every reason to expect he would never face. There are no sacrifices to admire. Fischer simply accumulated tiny structural advantages and converted them without ever allowing counterplay.

It is remembered as much for what happened afterwards: Spassky stood and applauded his opponent along with the audience. The full story is in Fischer vs Spassky 1972. Put next to the three games above, it marks the change in what the chess world admires — from the brilliance of a forced mate to the harder brilliance of never giving your opponent a chance.

How to study a famous game properly

Replaying moves from a list teaches almost nothing. What works:

  1. Set up the position before the key move and try to find it yourself. Give yourself five minutes on 18.Bd6, 19.Rad1 and 10.Nxb5 before you look at the answer.
  2. Ask what the losing side should have played. Every one of these games contains defensive resources that were missed. Finding them is more instructive than admiring the winner.
  3. Count developed pieces at the critical moment. In all three games the winner had roughly twice as many pieces in play. That single number explains more than any variation.
  4. Play the position out against the engine from both sides. Winning a won position is a separate skill from reaching one, and it is where most club games are actually decided.

If the notation itself is the obstacle, start with how to read chess notation, then work through how to analyze your own games using the same questions.

Play these positions yourself

Every diagram above has a "Play this position" link that loads the exact position into the free board on LocalChess. Take the Immortal Game after 18.Bd6 and try to hold it as Black. Take the Evergreen position after 19.Rad1 and see whether you can find the four-move finish before the engine does. Take the Opera Game after 12.O-O-O and try to survive as the Duke and the Count.

You will learn more from ten minutes of struggling with one of these positions than from an hour of watching the moves scroll past. Nothing to install, no account required — open the board and start.