A nine-part analysis of a swim lane just landed on my desk in Brisbane. Nine sections. Three data tables. Not a single real number. Every cell repeated the same sentence: "Insufficient information, cannot assess." I have read thousands of reports on xG, on PPDA, on injury metrics throughout my career. But I had never received an analysis honest enough to be empty. And what made me stop was this: it was right. Completely right. In an industry where everyone wants to fill every gap with some number, a report brave enough to say "I don't know" is rarer than a world record.

In Australia, swimming goes far beyond a sport. It is national identity. Every Olympic cycle, the whole country holds its breath with each lane of the Dolphins. Behind every lane lies a vast data ecosystem: split times every 50 metres, underwater-phase data, turn speed, stroke rate, distance per stroke cycle. World Aquatics publishes official results, while timing systems such as Omega supply data down to the hundredth of a second. At the Tokyo 2026 Olympics, Australia's swim team won 9 gold medals — its best result in Olympic history.

The swimming betting market in Australia is not as large as football or cricket, but it exists and operates on precisely the numbers I just listed. Every Olympics or Commonwealth Games, bookmakers price each event, and those prices are built from historical data, recent form, and sometimes complex prediction models. When the input data is complete, the model works well. When it is empty, the model cannot generate a price — and some bookmakers, rather than admit that, fill the gap with intuition and call it a model.

But here is the problem few are willing to state: most swimming data is valuable only for an instant. A football match leaves behind 90 minutes of footage, hundreds of passes, thousands of positional data points. A 100-metre freestyle lane ends in under 50 seconds. No passes. No xG. No positional heat map. All that remains is a series of times, a few slow-motion angles, and a great many gaps the camera lens cannot touch.

That is why, when a swimming analysis returns zero, it carries a different meaning: an honest confession that the raw material simply does not exist. What I want to dissect here is whether our analytical industry has the courage to accept that.

I have spent most of my career hunting anomalies in data. On the day Germany collapsed in Kazan in 2026, I pointed out that their team managed only 11 passes into the box and an xG of 0.7, lower than South Korea's. Kazan was the day I learned that a 99% probability can still die on the betting table. But the larger lesson from Kazan is not about football. It is this: a number only has power when it exists, and when it does not, the real power lies in daring to admit it.

The structure of an in-depth swimming analysis has nine layers: technical analysis, performance and data analysis, competition system and participation mechanism, the world swimming landscape map, rules and anti-doping governance, athlete career and team system, risk profile, public narrative and expectations, and finally the industry ripple effect. Each layer needs its own type of input data: split times, performance indices, seasonal rankings, injury history, sporting-nationality transfer records.

When the input source is empty — no athlete name, no distance, no time, no meet — all nine layers collapse at once. Not because the analyst is lazy. But because the foundational principle of any serious analysis is no unfounded speculation. The moment you start inventing an athlete, a result, an event to fill the gap, you are no longer doing analysis. You are writing fiction and stamping it with a scientific label.

In swimming, this is more dangerous than in any other sport, because the margin of error here is measured in hundredths of a second — and a hundredth of a second can be the distance between gold and fourth place.

I remember watching a youth swim meet in Queensland. A coach asked me to predict who would win the 200-metre individual medley. I opened the data table: no split times for the breaststroke leg, no underwater-phase data, no competition history for three of the four swimmers. I answered: "I don't know." He was disappointed. But that was the only honest answer the data permitted. Three weeks later, with complete data from another meet, I predicted accurately to the tenth of a second.

That is the crux. The heat map has become the new fortune-telling of the sports industry. People paint red and blue squares onto a pitch, then declare they have decoded the match. In swimming, the equivalent is a chart comparing average speeds between swimmers — a flat average that conceals the entire tactic of pacing, turn technique, and the tactical decision in the final 15 metres. An average does not tell you that a swimmer went out faster than the world record over the first 50 metres and then collapsed in the final 25.

I do not trust emotion. I trust a data series longer than your emotion. But I also believe a long data series does not mean a complete one. And when it is not complete, the only right thing is to state clearly what is missing, by how much, and how that affects the conclusion.

At a deeper layer, swimming data is bound to something far more human. Scouting networks in developing countries both discover geniuses and produce sports lottery tickets and broken families. A 12-year-old is measured for arm span, VO2 max, potential — and from that, an entire future is priced by a table. When that table is empty or wrong, the price paid is not in hundredths of a second, but in a human life.

This is the counter-intuitive angle I want to defend: an analysis that returns zero can still be a success, in the humblest way.

The sports-analytics industry is sick with one disease: a fear of gaps. When data is missing, the first reflex is to fill it — with a model, an assumption, a half-remembered figure. I have seen 3,000-word reports built on a single number nobody verified at source. I have seen analysts assign an injury index to athletes who have never been injured, simply because the model needed a variable to run.

Data has no gender. But the numbers have no gender, while those who read them do. Readers carry deadline pressure, bosses demanding reports, audiences demanding answers. And that pressure turns a data gap into a lie presented beautifully.

There is a probability paradox I learned after Kazan: having witnessed the 99% fail, people easily swing to the opposite extreme — believing nothing is predictable. Both extremes are wrong. The balance lies in between: predict when you have enough data, and stay silent when you do not. A 99% success still happens hundreds of times a day. But to say "99% success", you need a sufficiently large sample. You cannot derive probability from thin air.

In swimming this is especially clear. A swimmer can break a world record at one meet and swim 3 seconds slower at the next, for reasons the table does not record: pool condition, water temperature, psychological pressure, an unhealed shoulder injury. If you have data from only one meet, you are building a house on a single brick. And if you have no brick at all, build nothing.

Limits of the data: I must admit there are things in swimming the table will never measure. A swimmer's feeling underwater in the final 15 metres. The fear of an opponent who has beaten you before. The silence of an empty pool. These are a different kind of data — not measured in milliseconds, but data nonetheless. An honest analysis must map the boundary between the zone data can assert, the zone that is ambiguous, and the zone that must rest on intuition. That nine-part analysis did exactly this: it pointed precisely to where the data ends and the silence begins.

What I leave behind sits in a different question: how to build an analytical culture that dares to say "insufficient information" without being seen as weak. A swim lane ends in 47 seconds. A wrong decision built on fabricated data can last for years. Between those two spans of time, I choose honesty.