The Blank Scoresheet: The Chess Games That Vanished From Collective Memory

On a late-year Wednesday afternoon, in the second-floor room of a chess club tucked deep inside an old alley in Jing'an, Shanghai, I sat for two hours without writing a single line. Four men past sixty shared two blitz boards. They used a worn mechanical clock with no screen, no connection port, no software. Nobody brought paper. Not one move was ever notated. By the time the last of them put on his coat and walked down the stairs, hundreds of games played in that room had vanished from the world, clean as if they had never existed.

That evening, a data file arrived in the newsroom inbox. It had a full name, a title, a source, a section, and a domain label that read, in two characters, chess. Every field in the table had a heading. None of them had content.

The room in Jing'an and that data file sat fifteen kilometres apart, and they told the same story.

The backdrop of a trimmed memory

Chess holds the largest game archive of any sport on earth. FIDE adopted the Elo rating system in 2026 and published its first rating list in 2026. Since then, tens of millions of games have been recorded, numbered, and classified by opening code. ChessBase, Mega Database, Lichess, Chess.com — each platform keeps a share of this discipline's memory. A game at the 2026 World Championship can be re-evaluated move by move within seconds of a player lifting a hand off a piece.

In April 2026, in Toronto, Gukesh Dommaraju — seventeen years old — won the Candidates Tournament and became the youngest challenger in history. Seven months later, on December 12, 2026, in Singapore, he beat Ding Liren in the fourteenth game and became the youngest world champion at eighteen. Every moment of that event was preserved almost past the point of completeness: thinking time per move, the rate at which moves matched engine recommendations, average centipawn loss, even the heart rates of online viewers.

Vietnam keeps its own record in its own way. In 2026, Nguyen Ngoc Truong Son won the world under-10 title in Spain. In 2026, Le Quang Liem won the World Blitz Championship in Khanty-Mansiysk, one of the biggest titles Vietnamese chess has ever held, alongside his repeated dominance at the HDBank Cup. Those games still sit in international databases, searchable, replayable, available for commentary twenty years later.

The problem lies elsewhere.

What the engine cannot read

A simple calculation nobody wants to perform. An open international chess event can draw several hundred players and produce several thousand games in ten days. Over the same stretch, in Shanghai alone, I counted dozens of small clubs running every evening, a few dozen boards each, and not one of them keeping a record. Multiplied across the world's cities, the games played in a single year outnumber the games archived in the entire history of modern chess.

Put another way, chess's archive is a photograph taken through a crack in the door.

Based on my experience covering events in Shanghai and Ho Chi Minh City, whether a game is properly recorded depends on almost everything except how good the game was. A game at an event with a bank sponsor, a notation desk, an arbiter keeping scoresheets, and a screen feeding the outside world will last forever. A better game, a sharper game, played on the third floor of a primary school at seven in the evening, exists for nobody beyond the two people sitting across the board.

A chess game truly exists twice: once on the board, once in the memory of the person sitting opposite. Scoresheets, databases, indices are copies, and a copy has the right to be left blank.

Modern metrics do not fill that gap. They only thicken the part that is already recorded. Average centipawn loss, engine match rate, time spent per move — these are excellent measures for answering how precisely a player performed. They cannot answer why someone chose an imprecise move. A bad move may come from miscalculation, from time pressure, from a phone call at eleven the night before, from a small child with a fever at home. No engine evaluates the breathing of the player holding the white pieces.

There is one moment I keep returning to when I think about Gukesh in Toronto. When a seventeen-year-old prodigy still did not know he was about to become a legend, he sat at the board wearing the face of a schoolboy working through a hard exam. Every game he played there has been archived, analysed, annotated. The one thing never archived is that he did not yet know what was waiting for him. Legends get recorded only after the legend has already formed, and by then the fragility has disappeared from every source.

In 2026, while Chinese football talked about Oscar's sixty-million-euro transfer to Shanghai SIPG, I skipped a newsroom meeting to sit at a training pitch and watch the youth squad shoot. What I brought back was not a metric. It was the breathing of a seventeen-year-old standing over a ball, and the fact that he hit the inside corner seven times out of ten. Seven out of ten is data. The two seconds he stood still after each miss, looking down at the grass before walking back to his spot, is in no statistics table anywhere.

Some space should also go to the boring part, because the boring part occupies most of this discipline's time. Most elite games, seen from outside, are balanced, repetitive, colourless sequences running forty moves before anything worth mentioning happens. Draws at the Candidates are treated as defective products. Yet if chess history were written from those draws, it would be far more honest than a history written from a handful of celebrated reversals.

Meanwhile, another gap persists at the very centre of the sport. Magnus Carlsen reached a peak Elo of 2882 in May 2026, the highest ever recorded. After 2026 he declined to defend the world title. The crown went to Ding Liren in 2026 and then to Gukesh in 2026, while the highest-rated player remained Carlsen. The rating list and the honours list, the two official records of the same discipline, now tell two different stories. A gap like that had not appeared in nearly a century of modern chess, and no amount of additional data can patch it.

The blind spot of collective memory

Our professional instinct treats a data gap as a defect to be fixed. Without a scoresheet, an article is not credible enough. Without engine evaluation, commentary is not weighty enough. A game absent from the database was never played.

The reasoning sounds tight, but it reverses cause and effect. Absence of evidence gets read as evidence of absence. In chess, that rule operates in favour of a very small group of players: those whose federations can afford rated events, afford notation staff, afford sending players abroad for extra rated games. The rest of the chess world plays unrated games, and unrated games go unremembered.

Another aspect rarely discussed: organisations publish only what suits them. A federation will post news about a rising junior's rating climb and stay silent about why a veteran withdrew. I once received a press release for a Southeast Asian international event with a complete entry list and not a single line about the budget being cut and the number of rounds reduced. In the database, that tournament looks entirely normal. A reader twenty years from now will never know what happened.

This is where perfect data becomes more dangerous than missing data. An empty field declares itself empty. A field stuffed with figures that have been trimmed declares nothing at all.

There is one more paradox, tied to the nature of recording. When every move is evaluated, players change how they play. They choose safer lines, less criticisable lines, lines easier to explain at the post-game press conference. Recording is part of the room, and it changes the room.

It should also be said plainly that some games disappeared for reasons so mundane nobody bothers to tell the story. A federation's hard drive failed after a leaky roof. An archivist resigned and took the external drive with him. A provincial tournament held in 2026 kept handwritten scoresheets in a wooden cabinet, and the cabinet was moved out during an office renovation. There is no tragedy here. Only busy people and cabinets.

What remains after the game

There is one detail I keep from the 2026 World Championship, and it appears in no technical report. After the fourteenth game ended in Singapore, when every camera had turned toward the eighteen-year-old champion, the loser stayed at the board a few minutes longer, his hand resting on a king that had been checkmated. Then he stood, walked toward the back of the hall, and for that instant the entire room behind his back went quiet.

I have written about chess for more than twenty years, and most of that work is resistance against disappearance. The resistance does not come from recording more. It comes from recording carefully the things machines cannot keep: a Wednesday afternoon at a second-floor club in Jing'an, a worn mechanical clock, the face of an unknown player in the moment before anyone hands him a title.

An empty stadium still holds the applause of the dead. An empty archive does too, except nobody bothers to go looking for the applause inside it.

After three decades at the desk, I have realised I never wrote about sport. I wrote about what remains after the game ends, and about how much of it will remain nothing at all.