The Empty Board: Data Discipline in the Craft of Chess Writing

October night in Saint Petersburg. The newsroom is so empty that the keystrokes sound like knuckles on wood. I open the database I have used for years, type in a tournament name, press search. The screen returns a blank line. No games. No results. No opponent names. Just an empty cell sitting there, cold as a lake that has not yet frozen over.

Thirty years beside the board taught me to read what a position throws up: Elo ratings, time controls, draw rates, every move graded by an engine. I never had to learn how to read a blank space. At sixty-five, I finally understand that a blank space is also a form of data – and the form most likely to fool a writer.

The day the heart stopped beating, I learned to beat slower but deeper. I wrote that line in the summer of 2026, and since then I have applied it to everything I write, even the driest news item. An empty cell on a screen deserves to be read just as slowly.

Context: a sport where every fact can be looked up

Chess today runs on data. The International Chess Federation (FIDE) publishes its rating lists monthly, split into three categories: classical, rapid and blitz. Live ratings update game by game on specialist tracking sites. Game archives sit inside the ChessBase and TWIC databases. Online platforms such as Chess.com and Lichess publish statistics for every account. A fan needs one click to hold more information than an editor did twenty years ago.

That has changed my job completely. The first task of a chess writer now is source triage. An official FIDE release, a specialist outlet's report, a mainstream sports desk piece, an anonymous forum post – those four carry radically different evidentiary weight. That weight must be established before writing, not after.

In this field, a lack of data does not mean there is no story. But it also does not license the writer to build a story out of thin air. A fabricated rating, a fabricated head-to-head record, a fabricated prize fund – all can be checked within minutes. On the Luzhniki night, the old bird learned that landing is also a way of flying.

Reading a rating: a photograph, not an essence

A rating is not a player's essence; it is a photograph of a moment. The classical rating measures strength at long time controls, where a single game can run six hours. Rapid and blitz ratings measure reflex. A player can be fifth in the world in classical and twentieth in blitz. The list published at the start of the month is a still frame of a moving stream, and readers are entitled to know when the frame was taken.

Live ratings are something else entirely. They jump with every game, and across a two-week event a number can peak and collapse. When a player crosses 2800 while a tournament is in progress, that is a real fact, but the fact may expire within three days. The rule I set for myself: never quote a live rating without a date.

Performance rating is the subtler trap. It is calculated from results and opponents' ratings. The problem is the sample. Seven games are far too few to conclude anything about a human being. A player who wins four games against weaker opposition will show a prettier performance rating than one who draws six against the elite. A result in one event does not measure strength; it measures a run of luck plus a schedule.

My experience following thousands of games shows this repeats with metronomic regularity: after every major event a young name appears with a glittering performance rating, and three months later that name returns to its true position. The inexperienced writer turns them into an icon; the professional states the sample size.

The engine grades moves, not people

Analytical tools have changed how we see a game. Average Centipawn Loss (ACPL) measures move quality, lower being better. Engine match rate measures fidelity to the strongest choice. Both are useful, and both are misread.

ACPL only means something alongside a time control. In blitz, nobody maintains a low ACPL; in classical chess, a move that does not match the engine may be practically stronger. The engine does not know the opponent is short of time. It does not know that an objectively inferior move forces the opponent to think for another forty minutes. The engine grades moves, not people.

A chess game is also a war of clocks. At classical controls each side may have ninety minutes plus increment per move. In blitz, a few minutes each; in bullet, one or two. When a player falls into time trouble, move quality collapses fast. Any analysis written afterwards must state that context, or every technical conclusion hangs in the air.

Opening preparation is a submerged industry

Behind every elite player stands a team of seconds who build prepared lines for specific opponents. A novelty never seen in the databases can be worth half a point, sometimes a whole game. But its value depends on who is sitting opposite: the same novelty can be a weapon against one player and suicide against another.

This is the easiest place to err. Writers praise a prepared line while forgetting to ask whom it was aimed at. Opening preparation is opponent-specific, and an analysis that names no opponent has analysed nothing.

Tournaments, qualification and draw rates

The tournament system is layered. Qualification runs through the World Cup, the Grand Swiss, or a rating spot. The Candidates Tournament selects the challenger. The Olympiad is national team competition. Each tier has its own rules on time, number of games and tiebreaks.

Draw rate is the most underrated metric. A high-draw event is usually dismissed as dull, but it can also signal a field so strong that nobody dares take a risk. The Sofia Rules, designed to limit early draws, were an administrative attempt to rescue watchability – and like every administrative attempt, both effective and crude.

When two players draw a standard match, they move to rapid tiebreaks. If still level, Armageddon: one side gets more time but must win. It is the most beautiful and most cruel rule in chess.

The most sensitive zone: anti-cheating

No subject demands greater caution than cheating. The 2026 affair between Magnus Carlsen and Hans Niemann pushed the story off the board and into the court of public opinion. Online platforms have also run ban waves based on statistical models and security screening.

This is the zone where a wrong line cannot be repaired with an apology, because it attaches to a named individual.

Absence of evidence is not evidence of innocence, and it is not evidence of guilt either. Both directions are true, and both must be stated explicitly rather than left to the reader's inference.

At the governance level, FIDE makes the rules, publishes the ratings and handles federation transfers. A player changing the country they represent is not merely paperwork; it is a story about resources, scholarships and qualification places. The transfer market is where people sell the future to buy hope – and chess has its own market, just a quieter one than football.

One more line must be drawn clearly: over-the-board results and online results are not directly convertible. A player who dominates online may not hold that form through a two-week classical event, where nobody types for you and there is no off switch.

Contrarian view: when a blank is presented as a completed check

Here I want to speak plainly.

The most dangerous thing in chess writing today is not a wrong fact. A wrong fact gets caught, usually fast, because my readers carry a database in their pockets.

The most dangerous thing is an empty report presented as a completed check.

When a source returns zero, the inexperienced writer concludes: no issues found. But an empty cell only says the pipeline broke somewhere, or the source was blocked, or the original contained no text at all. It does not say the subject is clean. The absence of a red flag is not a certificate of health.

I nearly made that mistake once. Years ago, when a game vanished from the archive because of a technical fault, I almost wrote that the match had gone smoothly. My old editor knocked on my door and put a sheet of paper on the desk bearing a single line: the blank has not been answered. I have never forgotten it.

In a sport where ratings, head-to-head records and prize funds are all retrievable, an error is permanent. Readers can open the database while they read my piece. So my trade has an unwritten rule: when there is nothing to analyse, say there is nothing to analyse. It is the hardest sentence to write, and the one that keeps the profession honest.

What remains

An empty board is not a full stop. When the stadium no longer holds a ball, memory begins to score – and a professional writer is one who can tell memory apart from evidence.

I am still here in Saint Petersburg, waiting for the database to answer. Tomorrow I will re-run the query. If the screen is still blank, I will write about the blank itself: honestly, slowly, and without adding a single name I have not verified. In a sport where every fact can be looked up, honesty toward the blank is the highest form of accuracy the craft allows.