When the Page Stays Blank: The Data Discipline Tennis Is Losing

New York, the last days of August 2026, before the US Open began. In the media room at the Billie Jean King National Tennis Center, a spreadsheet was projected onto a screen. Thirty-two rows. Sample dates. Sample codes. Laboratory names. The independent tribunal's findings. The date of appeal. Every row had been checked against the original document, with reference numbers, timestamps and signatures. The final column — the one meant to give the reading public a verdict — was blank.

A reporter raised a hand: "So is he guilty or not guilty?" The answer was that the question did not live inside the spreadsheet, because the spreadsheet only held what could be verified. I was sitting in the fourth row. I opened my forty-page notebook and wrote a single line: "Insufficient data to conclude." The forty-page notebook never lies. It only knows how to say little, or to say nothing.

That was the moment I remember most from that season. Not a serve in a tiebreak. Not a save of three championship points. A blank cell on a spreadsheet, and a room full of people uncomfortable with it. Eighteen months later, I understand that blank cell was the biggest lesson professional tennis handed to the people who cover it.

When noise has more sources than data

December is tennis's transfer window. There is no crowded market like football's, but there is an underground one operating on exactly the same logic: coaching contracts, wild cards, the build-up to the Australian Open, exhibition swings in the Gulf, and hundreds of aggregator accounts that live off screenshots. December is when the signal-to-noise ratio hits its annual low.

The 2026 and 2026 seasons pushed that market to a new level. The Jannik Sinner case, the Iga Świątek case, the end of Rafael Nadal's era in Malaga in November 2026, the termination of the Novak Djokovic–Andy Murray coaching partnership in May 2026 — each event generated at least three competing versions. The version shared most widely was almost always the shortest.

In more than four decades in this trade, from a London newsroom in 2026 to quiet mornings at Indian Wells and Miami, I have kept one rule: I only stand on a practice court when there is at least one verifiable fact to record there. A practice court has no spectators, but every answer is sitting on it. The problem with the past season was not a shortage of practice courts. The problem was too many people writing without opening a notebook.

Break-point conversion and the small-sample trap

Start with the most quoted and most misunderstood metric in the sport.

In a three-set match at the second round of a Masters event, a player might generate four break points. Convert three and the coverage calls it nerve. Convert one and the coverage calls it a mental weakness. Four chances. Four. Statistically, four observations cannot separate a great clutch player from an average one who got lucky.

Reverse the test. A player converting exactly 25 percent of break points across a full season, on roughly two hundred chances, gives you a sample thick enough to say something. But 25 percent over a season and 75 percent in one match are different animals, even though both are written with the same sentence: "break-point conversion: 75 percent."

A fact only means something when you publish the sample size alongside it. Strip out the sample size and you are no longer doing statistics. You are doing prose.

The 2026 French Open final is the finest example I have watched from inside the grounds. Carlos Alcaraz beat Sinner 4-6, 6-7(4), 6-4, 7-6(3), 7-6(10-2), after five hours and twenty-nine minutes, after saving three championship points. Read only the scoreboard and you see a match that was broken open and then inverted. Sit inside the stadium and you see something else entirely: two players splitting almost every service game across the first two sets, the difference living inside seven tiebreak points, and a final-set scoreline of 10-2 that fails to represent the four hours of balance before it.

People look at the winner. I look at the gap that opened before the ball left the strings. That gap never appears on a scoreboard. It sits in the return position of the eventual loser in the seventh game of the fourth set, when he retreated half a step to his left and exposed his entire forehand side.

The metrics that never reach the scoreboard

Three columns in my notebook never get printed by any aggregator.

The first is second-serve points won. It tells the truth far more honestly than first-serve percentage, because it measures the exact moment a player must choose: play safe and accept being attacked, or take risk and accept the double fault. At the top level, the swing between those two choices inside one set can reach fifteen points. Fifteen points is three games. Three games is a set.

The second is return depth. No official body publishes it in raw form, yet it governs almost the entire shape of a match. A return landing half a metre from the baseline turns the server from aggressor to defender by the second shot. A return landing mid-court opens the door to a finishing shot. Same rate of returns in play, two opposite outcomes.

The third is the number of points lasting beyond nine shots. This is a biological metric, not a technical one. It measures accumulated tolerance. In a three-set match running two hours and forty minutes, the total number of points past nine shots is usually under twenty. But where those twenty points fall — which set, which scoreline — decides who is still standing in the third.

The blank spreadsheet: two verification tests from the 2026 season

Back to the empty cell in New York.

On 20 August 2026, the International Tennis Integrity Agency announced that Sinner had produced two samples positive for clostebol in March 2026 — one collected in competition at Indian Wells, one out of competition. An independent tribunal accepted his explanation of unintentional transdermal contamination via a physiotherapist using a spray containing the substance, found no fault or negligence, imposed no suspension, but stripped 400 ranking points and 325,000 dollars in prize money from Indian Wells. On 26 September 2026, the World Anti-Doping Agency appealed to the Court of Arbitration for Sport. On 15 February 2026, the two sides reached a settlement: three months of ineligibility, from 9 February to 4 May 2026. Sinner missed Indian Wells, Miami, Monte Carlo and Madrid, and returned in Rome.

Read that spreadsheet and you have fourteen facts. Read the headlines and you have two sentences: "cleared" or "banned." Both are blanks filled with guesswork. The first ignores that the original ruling was under appeal. The second ignores that the outcome came from a settlement, not from a full hearing.

The second case ran in the opposite direction. On 28 November 2026, the integrity agency announced that Świątek had produced a positive sample for trimetazidine — a banned metabolic modulator — from an out-of-competition collection in August 2026. She accepted a one-month period of ineligibility covering a stretch the media had previously described as "personal reasons."

There is a professional paradox here that I have watched play out at least four times in my career. When a player withdraws for personal reasons, nobody checks. When the real reason surfaces, nobody apologises. Neither direction costs anyone credibility, because in the attention economy a wrong headline still draws more readers than a correct one filed two days late.

The counterintuitive part: more data does not mean more certainty

This is where I want to slow down.

Modern tennis readers are told that more numbers bring them closer to the truth. That holds for season-long cumulative metrics and fails badly for metrics drawn from a single match or set. The same 6-4 can be a set of total control or a set in which the winner won only the two points that mattered. A scoreboard cannot tell those apart.

Worse, adding data can raise confidence without raising accuracy. With four metrics on a player, you know you are missing something. With twenty, you forget you are still missing the twenty-first — the one that would have inverted the conclusion.

As a woman who has worked in this trade since newsrooms used typewriters, I have been challenged with a sentence I will not forget. In 2026, in a stadium corridor in Russia, a male colleague laughed and said women only know how to count tackles. I did not argue. I handed him a forty-page notebook from training sessions, containing pressing counts, off-ball movement that created space, and defensive actions inside the player's own half by Croatia's number nine. A European football body later cited my piece. The notebook does not argue with anyone. It simply stands there.

Our trade is not punished for inventing, it is punished for being slow

That is the real motive behind every filled blank. Nobody is disciplined for writing that a player is "rediscovering his feel for the ball." But a newsroom falls behind if it arrives after a sourced-less aggregator.

What results is a system of skewed incentives: the professional risk of being wrong is priced low, the risk of being slow is priced high. In that system, the blank cell is always the losing option. It is also the only correct one.

There is a standard I consider exemplary, and I have seen it applied by newsrooms built around sports data: publish your own working sheet, even when it is empty. A report with nine sections, all nine reading "insufficient information, cannot assess," is worth as much as a two-thousand-word analysis built on guesswork. Readers have a right to know where nobody knows anything.

One fact, two ways of using it

Try applying that discipline to a question currently hot in the winter coaching market: a former major champion opening an academy or taking an advisory role with a young player.

The common approach reads names. A name is a story. A name is a headline. The other approach reads three things: how many grassroots coaches are actually certified inside that system, how many supervised training hours the junior receives per week, and the contract structure — paid per match day or paid per result. A results-based contract tells you who genuinely carries the risk.

The first fact can be verified in a day. The third often takes six months to surface. That six-month lag is where the prettiest stories live longest.

Signals to track instead of rumours to read

For the rest of this transfer window, I will track three signals and ignore everything else.

The first is coaching contract structure. When a partnership ends, read the statement in the language of the people who drafted the contract: term, termination clause, who holds the right to announce. The shorter the statement, the higher the chance of a dispute.

The second is the appearance of metrics with sample sizes. A piece claiming a player "wins 78 percent of break points" without saying out of how many chances is not data. It is decoration.

The third is corrections. A newsroom that publishes a correction is a newsroom still holding the line. A newsroom that stays silent when proven wrong has already changed professions.

My notebook is thicker this season than last, not because I wrote more, but because I crossed out more. The crossed-out lines are the ones I nearly filed before I had enough. If there is one thing I want to leave with the young reporters blanketing tournaments with phones and speed, it is this: a blank cell is not a failure. A blank cell is proof that you still know the difference between what you know and what you want to believe.