The Empty Board and the Games No One Recorded

THE EMPTY BOARD AND THE GAMES NO ONE RECORDED In the reading room of the library on Sadovaya Street, I opened a folder of paper gone the colour of tea. The crosstable of the Saint Petersburg tournament of 2026 lay between two stiff covers, black ink, slanted old German hand. On the row of an unremarkable competitor, the secretary lifted his pen and left the result blank. Not because he forgot. He had filled seventeen boxes already; by the eighteenth his wrist ached, or a bell rang, or, more simply, he never learned how that game ended. One hundred and twelve years later, I sat before a screen on the fourth floor of the newsroom and opened a chess database. There were no blanks. All fourteen games of the 2026 World Championship match in Singapore were stored in the machine, move by move, with clock records and engine evaluations down to hundredths of a pawn. Gukesh Dommaraju, eighteen years old, defeated Ding Liren 7.5–6.5 on 12 December 2026 according to the International Chess Federation, becoming the youngest world champion in the history of the game. I wrote about that night. The piece was clean, complete, verifiable line by line. I did not use a single blank. But the blank did not disappear. It only moved. WHEN CHESS LEARNED TO COUNT In 2026 the International Chess Federation formally adopted the rating system of Professor Arpad Elo, a Hungarian-born physicist teaching in Wisconsin. Before that threshold, a player's standing was a product of collective memory: a list compiled by a magazine's editorial board, a line of commentary in a specialist journal, a rumour carried from one coffee house to another newsroom. People ranked Capablanca above Alekhine, Rubinstein above Tarrasch, and nobody demanded proof. Elo arrived and gave chess a unit of measurement. He did not invent the truth; he supplied a convention tight enough to argue over. But attached to that convention was a subtler consequence: from then on, what was not recorded did not exist. A game without a scoresheet was a game without moves. A player without a rating was a player without strength. Based on my experience following matches over nearly five decades, chess is the only sport whose memory can be compressed into a file of a few hundred megabytes. Football has footage, but footage is not a goal. Chess is different. A game of chess, by its nature, is a sequence of symbols. When that sequence is typed into a machine, the game is reborn intact, not one move lost, not one clock beat wrong. So people believe chess is a sport of perfect data. I believed it for years. Then I went to Saint Petersburg and read handwritten sheets. The 2026 crosstable opened before me not as a page of data but as a page of diary. Emanuel Lasker arrived at forty-five, reigning world champion, and was led by José Raúl Capablanca — twenty-five, undefeated, at the height of his powers — after the first stage. Lasker came back, beat Capablanca in the final round, and finished on 13.5 out of 18, half a point ahead. Alexander Alekhine took third with 10. Siegbert Tarrasch and Frank Marshall shared the bottom two places of the final group. I read that table three times. On the third pass I realised I was reading it the way a twenty-first-century person reads: dividing points by games, computing differentials, comparing with other tournaments of the same year. I was treating Lasker as though he were a rated player, when the only thing I actually held was eighteen lines of handwriting and one blank box. WHAT DATA CANNOT MEASURE There is an asymmetry that modern chess commentary rarely names. We assess contemporary players by the whole body of their results, and we assess their predecessors by anecdote. For Gukesh we have month-by-month ratings, win rates by colour, average moves per game, even centipawn loss on every decision. For Mikhail Chigorin — founder of the Russian chess school, who played two world championship matches against Wilhelm Steinitz in 1889 and 1892 — we have a handful of famous games, a few contemporary remarks, and the rest is fiction handed down through three generations of textbooks. That asymmetry is not neutral. It leans toward the living. When a modern player contests three hundred games a year and we hold data on all three hundred, every error is exposed. When a nineteenth-century player contests forty games a year and only twenty are reprinted, his errors evaporate from history. Collective memory remembers what it has, forgets what it lacks, and then tells itself that what it forgot never mattered. Akiba Rubinstein is the most painful example I know. He dominated from 2026 to 2026, won major tournaments in succession, was regarded as the leading candidate for the world title, and never received a title match. The reason lay in money, in organisers, in the times — none of which appear in any crosstable. Harry Pillsbury won Hastings 1895, the strongest tournament of the nineteenth century, then fell ill and died at thirty-three. His final years have no scoresheets. We know he weakened; we do not know how, in which move, under what pressure. Here I must state plainly something my profession usually avoids: where there is no data, people fill the gap with story. And story always has a bias. Journalists tell the tale of a genius crushed by circumstance, because that tale sells. Historians tell the tale of a golden generation of a school, because that tale serves national identity. Nobody tells the tale of a middling player who competed for thirty years and won exactly the games he should have won. Yet the truth of chess, for the most part, lies precisely there. An empty stadium with no ball — and memory still keeping score. THE BLANKS OF THE TWENTY-FIRST CENTURY When I told this story to a young colleague in the newsroom, he laughed and said: that era is over. Every game is recorded now. I disagreed, and I had specific reasons. Online chess platforms today hold billions of games, but most of that number is played under anonymous accounts, in unrated mode, downloaded by no one, cross-checked by no one. A player who grows from twelve to eighteen years old may have played fifteen thousand blitz games online, and in his official record that number is zero. The games that shaped his style, the games that taught him how to lose, sit outside history. Closed training matches are the same. Exhibition games without arbiters. Internal club tournaments. Blitz games in hotel rooms at two in the morning, after the event has ended, between two players competing for the next invitation. Those games happened, someone won, someone lost, and they exist in no database. I understood this while reporting on the affair of 2026. That September, at the Sinquefield Cup in Saint Louis, Magnus Carlsen lost to Hans Niemann and withdrew from the tournament. A month later, in an online event, Carlsen resigned after a single move. The International Chess Federation issued a ruling on the matter, and a United States federal court dismissed the related lawsuit in June 2026. The truth of that entire story was not on the board. The game has a complete scoresheet, move by move, precise beyond dispute. What has no scoresheet is what happened inside two heads, in a hotel corridor, in deleted messages. Modern chess protects the data of the game and fails completely to record the human part of it. That is the new blank. It is not in the row of an unremarkable player in the 2026 crosstable. It sits in the middle of a story the whole world is watching. WHERE MEMORY KEEPS ITS OWN SCORE There is another field where the blank is growing, and I follow it with particular attention: the sporting citizenship of players. In chess, the so-called transfer window does not resemble football. There are no transfer fees, no January and July windows. There is something else: federation transfer. A player may change the country he represents, after paying a fee and waiting the required period. When that happens, his entire prior biography is reread along a new axis. Old wins are credited to the old country; new wins to the new one; and the years in between — the years in which he actually matured — fall into grey space. The transfer market is where people sell the future to buy hope. In chess, what is sold is not a player but a line in a results table. People buy a name, and the name drags along a history already written. I have sat in many press rooms and heard coaches speak about their students in the language of asset managers. They compare ratings the way others compare yields. They talk about youth as a commodity with an expiry date. They are not wrong — they are simply doing exactly what a measurement system encourages. When the stadium no longer holds a ball, memory begins to score. In chess, when the crosstable no longer holds a line, memory begins to rank. THE BLIND SPOT OF THE NUMBER I want to give the final part of this piece to something I consider more important than tracking down lost games. The Elo system is not a law of nature. It is an accounting convention resting on an assumption that the result of a game reflects the difference in strength between two players. That assumption holds over the long run and fails in the short run — something any player who has lost a game to sleeplessness knows. But what is said less often is this: Elo measures results against the pool you play in. Play in a weak pool, and you can reach a high score without getting stronger. Play in a dense pool, and you can improve while your number stands still. My generation grew up believing the number told the truth. I spent most of my career building that belief in my readers. Now, at sixty-five, on the far side of a career, I see the need to state the other half of the story. A ranking without blanks is not a complete ranking. It is only a ranking carefully bounded to look complete. Every chess database in the world — even one holding tens of millions of games — remains a subset of every game ever played. The missing part is never random. It is missing for economic reasons, for geopolitical reasons, for technological reasons, because a secretary in Saint Petersburg in 2026 had a tired hand. On the day a heart stops beating, I learned to beat slower but deeper. I learned that on 12 June 2026, when Christian Eriksen collapsed on a football pitch in Copenhagen. In chess, the equivalent moment happens more quietly: when a great player announces he will no longer defend his title, when a generation sets down the pieces and rises from the board. Nobody counts. No data records the moment a person decides to stop being the best. WHAT REMAINS AT THE END I did not write this piece to deny data. I wrote it to propose a different way of reading it. Every time I open a chess database and see a full row, not one move missing, I remind myself that the row is a decision, not an event. Someone chose to record. Someone chose not to. Someone chose to translate from Russian to English, and in that translation lost a footnote, a question mark, a doubt. As a man who has spent forty-nine years standing between lines of text and human beings, I hold that the task of a chess writer is not to make the crosstable look fuller, but to point precisely to where the blanks are — and to keep those blanks visible. A chessboard has sixty-four squares. None of them is empty. But the history of the board always has empty squares, and the games no one recorded are still being played, somewhere, between two people no one is photographing. If a later generation wants to understand us, they will not read our ratings. They will go looking for the blanks, and guess at what we left behind.