Classic games
Reading a great game in a book means never testing your own idea at move nineteen. Here the coach speaks where a strong annotator had something to say, the evaluation bar keeps score, and at any position you can step off the page, try your idea against Stockfish, and see what history did.
10 games, every note traced to a published annotator. 5 more are still being sourced and stay off the shelf until they check out.
The Opera Game
Start here · Paris Opera, 1858
Morphy1–0Duke & Count
Edward Lasker presents Morphy's famous consultation game as a study of mobility and the compounding cost of lost time. His explanation follows one continuous causal thread: Black's early choices obstruct development, each threat then demands another passive move, and Morphy invests material because the undeveloped pieces cannot participate in the defence.
The rest of the shelf
9 games- Rotlewi1907
Rubinstein's Immortal
Łódź, 1907
Rotlewi0–1Rubinstein
Queen's Gambit, Symmetrical Tarrasch
Two tempi is a winning advantage if you spend them on the same squares your opponent needed. Nothing in the finish was calculated on move fifteen; the pieces were simply put where they belonged.
Petrosian's Invitation
World Championship, Moscow, 1966
Petrosian1–0Spassky
King's Indian Defence, Fianchetto Variation
An attack is only as strong as the pieces that can join it. Track the defender or attacker stranded on the wrong wing.
Dufresne1852The Evergreen Game
Berlin, 1852
Anderssen1–0Dufresne
Evans Gambit
Castling is a move you have to spend, not a move you get to postpone. Black is never lost in the opening here and is offered the king's escape at least four times.
Photos: Frans Peeters · CC BY-SA 2.0 · Stefan64 · CC BY-SA 3.0
The Hour on Move Eleven
World Championship, Singapore, game 11, 2024
Gukesh1–0Ding
Reti Opening, reversed Blumenfeld
A worse position is a problem to be solved, not a verdict. Gukesh spends an hour admitting he is worse, then plays for the squares his opponent has stopped watching.
Photos: Frans Peeters · CC BY-SA 2.0 · ChessKelly · CC BY 4.0
Playing the Other Man's Clock
Candidates Tournament, Paphos, round 6, 2026
Wei Yi0–1Sindarov
Vienna Game
A weakness you chose is not the same as a weakness you allowed. Sindarov takes doubled pawns and an uncastled king on purpose, because both of them come with open lines attached.
Photos: Harry Pot for Anefo · CC BY-SA 3.0 nl
The Move Tal Did Not Mean to Play
Bled International Tournament, round 2, 1961
Fischer1–0Tal
Sicilian Defence, Taimanov Variation
One omitted pawn move in a Sicilian is not a small thing. Tal skips it by accident, and from that moment every natural-looking move he has costs him more than it should.
Photos: Miroslav.vajdic · CC BY 4.0 · Frans Peeters · CC BY-SA 2.0
The Win No Human Was Going to Find
World Championship, London, game 6, 2018
Carlsen½–½Caruana
Petroff Defence
Nothing dramatic loses this position. A bishop sent to the wrong square, a rook on the wrong file, one careless stretch of five moves, and a symmetrical endgame has become a defensive job that lasts six hours.
Four Reasons for One Move
27th USSR Championship, Leningrad, 1960
Spassky1–0Bronstein
King's Gambit Accepted, Modern Defence
A sacrifice is a question put to a particular person at a particular moment. Spassky played one he knew was second best, because the answer that refuted it asked Bronstein to walk his king into the open with twenty minutes left for twenty-six moves.
Model game for King's Gambit
Nezhmetdinov1958The Sacrifice He Could Not Calculate
18th RSFSR Championship, Sochi, 1958
Polugaevsky0–1Nezhmetdinov
Old Indian Defence
Some combinations are too long to check and have to be judged instead. Nezhmetdinov gave his queen up on a position he could see rather than a line he had verified, and he said afterwards that the trusting was the hard part, not the counting.