When Tennis Data Goes Silent: The Trap of Emptiness in Sports Analysis

On a summer evening at a Masters 1000 hard-court event, I sat in the newsroom in front of three screens. The first screen carried the live match. The second ran the point-by-point scoreboard. The third displayed a model of win probability on every point — what I jokingly call my own "pressing scanner", a machine that reads the rhythm of a match faster than the human eye.

Then the third screen went blank. No red warning, no error line, no connection-lost notice. Just a quiet, clean emptiness, like a page with nothing written on it.

What struck me was our reaction. Nobody panicked. Nobody stopped broadcasting. We kept commentating, kept analysing, kept offering judgements — as though the emptiness were a confirmation that everything was unfolding exactly as expected. Only when the match ended, when I reopened the recording to check, did I realise I had spent nearly forty minutes talking about a dataset that did not exist.

Emptiness is not safety. I learned that through a specific mistake, and it changed the way I see the entire business of tennis analysis.

To understand why that mistake is dangerous, it has to be placed in the context of an industry that has turned numbers into a religion. Over two decades, tennis has moved from a sport people analysed with their eyes to a sport in which every shot is logged, every point is encoded, every player is built into a data profile.

Hawk-Eye tracks the ball to within a millimetre of error. Platforms such as StatsBomb and Tennis Abstract record serve location, spin, speed, bounce point, and even the movement cadence of each player. The analytical departments of the ITF, ATP and WTA run ever more sophisticated forecasting models, to the point where a Grand Slam semi-final can now be dissected point by point like a chess game.

In that setting, data becomes the foundation of every conclusion. A coach no longer says "I feel he's out of form". He says "his second-serve points won have fallen from 54% to 47% over three months". A journalist no longer writes "she has nerves of steel". He writes "her tie-break win rate is 68%, among the highest on tour".

But there is a hole almost nobody in the industry will admit to: the data system can fail, and when it fails, it fails silently. No alarm sounds. No red text appears. There is only emptiness, and that emptiness looks exactly like calm.

That is what I call the trap of emptiness. It runs on a logic that is simple and lethal: when a metric does not appear, the human brain defaults to thinking there is nothing to worry about. When a "risk" field is left blank, we read it as "no risk". When an injury record lacks data, we assume the player is healthy.

In an analysis class I once attended, a lecturer called this "the blank-page bias". A blank page, he said, is not a clean page. The difference between failing to detect a problem and there being no problem is the difference between two worlds.

In tennis, data gaps appear at three different layers, and each demands its own way of being read.

The first layer is the source. A Hawk-Eye sensor can fault during a game. A camera can be blocked. A scoreboard can miss a point. When the data source breaks, everything behind it — models, forecasts, judgements — becomes a castle built on sand. But because there is no warning, the analyst keeps working, keeps drawing conclusions, except those conclusions are no longer anchored to reality.

The second layer is extraction. Even when the source data is intact, the process of turning raw data into meaningful information can fail. A shot-recognition algorithm may fail to classify a drop shot. A filter may swallow an important point. The result is a dataset that looks complete but is in fact missing vital pieces.

The third layer is labelling. This is the most dangerous layer, because it involves people. When an analyst looks at a gap and automatically labels it "nothing here", he has turned ignorance into a conclusion. This is the moment data stops being a tool and becomes a curtain.

These three layers explain why the worst analytical mistakes in tennis rarely come from wrong data. They come from missing data that is read as though it were full.

Take a concrete example of how emptiness deceives us: the 52-week ranking-points cycle.

Every tennis ranking runs on a rolling mechanism. The points a player earns at a tournament expire in the corresponding week of the following year. If the player fails to reproduce an equivalent result, those points evaporate. This mechanism creates "points-defence cliffs" — windows in which a player must defend an enormous haul of points within just a few weeks.

The problem is that the cliff is usually invisible on the ranking list. A player can sit at No. 5, looking utterly secure, while in truth standing before a points cliff that will hit within weeks. The ranking shows the position. It does not show the structure of that position. And what is not shown is where the truth lives.

I have been fooled by this kind of gap. In one season I looked at a ranking list and saw a young player climbing steadily. I wrote that he was on course for the top 10. A few weeks later, after a run of results that were not bad at all, he dropped. The reason was not form. The reason was that points from a big tournament a year earlier had expired en masse, while the ranking list I was reading showed none of that pressure. I read a gap as a positive signal. And as I still tell myself whenever I get it wrong, an analyst's arrogance is like an own goal — nobody can save you, you can only learn from it.

The second example, and the most dangerous one: injury.

There is no field in which data gaps are more damaging than injury. Official injury reports in tennis are notoriously sparse and vague. A player can withdraw citing a "physical issue" without specifying the detail. A cancelled practice can go unreported. A minor surgery can be revealed only months later.

When there is no injury news, fans and media alike tend to assume the player is healthy. But the absence of injury news is not evidence of health. It is only evidence that the information has not been published. In many cases that gap hides a recovery process unfolding quietly, or a chronic injury the player and the team have decided not to disclose.

Schedule density is the biggest culprit. When a player has to play two matches a week for months on end, the body has no time to regenerate. No medical team, however good, saves a player from that schedule if he still has to take the court enough times to hold his points. And when the injury finally arrives, it usually arrives with no warning sign in the public data — because the public data never tracked that accumulation in the first place.

This is where I want to stress one thing: the silence of injury data is not the absence of injury, it is the absence of an observer.

The third example: form and sample size.

In tennis, every judgement about form rests on sample size. A player who wins five matches in a row looks to be in form. But five matches in a sport where a single match can last three hours, with every point a variable, is far too small a sample to conclude anything with confidence. When data is thin, every model becomes hypersensitive to noise, and confidence intervals widen until they lose meaning.

The problem is that the media rarely admits this. We like tidy stories. "She's back." "He's found himself again." Those stories fill the gap with emotion instead of evidence, and the gap disappears — until reality speaks up.

There is a deeper layer few discuss: the era of the Federer–Nadal–Djokovic trio is closing, and the generation of Carlos Alcaraz and Jannik Sinner is taking over. Every time a major name withdraws from a tournament, every time a generation hands over, we face the same data gap: a tournament missing the players our forecasting models are used to handling, and new players for whom we lack a sufficient sample. In those periods, it is precisely the absence of familiar names that is the most important information.

Now the counter-intuitive part.

I do not sell predictions; I sell hypotheses. There is an ocean between the two. And in that ocean, emptiness is not only the enemy. Sometimes emptiness is the story.

Every tactical schematic is an orderly lie — I go looking for the truth behind it. But there is a truth that only appears when you accept that you do not know: that sometimes the most notable thing about a dataset is not what it contains, but what it lacks.

When a player has an ordinary second-serve points-won rate but an unusually high tie-break win rate, the gap between those two numbers is where the story lies. When a player has no injury data at all for months, that absence may signal an extremely discreet fitness-management regime, or a problem being concealed. When a tournament lacks marquee names, that absence says something about the calendar, about prize money, about the fatigue of an entire generation.

The paradox is this: modern tennis analysis teaches us to trust data, but does not teach us how to read missing data. We are trained to extract information from what exists, not to interrogate what does not. As a result, when the system falls silent, we keep talking — and saying things with no foundation.

There is a solution, and it does not require complex technology. It requires a change in how we frame questions.

Instead of asking "what does the data say", ask "what does the data not say". Instead of reading a blank field as "no risk", read it as "risk not yet assessed". Instead of treating silence as confirmation, treat it as an unanswered question.

In my work as a documentary screenwriter, I learned that gaps are precious material. A scene in which no character speaks can say more than a stretch of dialogue. A photograph missing its subject can evoke more than a complete one. It is the same in tennis. The absence of data is part of the data. And a good analyst is one who listens to what is not said.

Back to the evening with the blank screen. I no longer cheat myself by filling gaps with baseless judgements. Now, when a dataset goes empty, I stop. I flag it. I tell the audience that the data source has a problem, that the following judgements rest only on eye observation, and that I will re-check when the system is back online.

That is not a weak admission. It is methodological honesty. And in an industry where everyone wants to appear certain, methodological honesty becomes a competitive advantage.

Because fans do not need an expert who is always right. They need an expert who is honest about what he knows and what he does not. The distance between those two things is where trust is built.

Tennis is a sport of numbers, but also a sport of gaps. Every time a player withdraws from a tournament without explanation, every time a statistics table is missing an important metric, every time a forecasting model returns a confidence interval too wide to conclude anything — that is when we must choose between filling the gap with guesswork and admitting that some things cannot yet be known.

I choose the second. Not because I enjoy ambiguity, but because I believe the truth deserves more respect than tidy stories. In an industry growing ever more dependent on data, the most valuable skill may not be the skill of mining data, but the skill of recognising when data is insufficient to say anything at all.

There is an ocean between knowing and thinking you know. A good analyst is one who can swim in that ocean without pretending to have reached the far shore.