Forty-Seven Empty Cells: When a Swimming Analysis Has Not a Single Data Point

The nine-dimension analysis board came up with forty-seven empty cells. Not a single timestamp, not a single athlete's name, not a single lane. I stared at it for about four minutes, then reopened the source file — still blank. It was the first time in six years of doing this work that I received a brief whose input data did not exist.

My first reflex was to blame the tool. I rechecked the extractor, re-ran the processing line, cross-referenced three older pieces that used the same framework. All three behaved normally. By the fourth attempt I understood: the problem was not the model. The problem was that someone had sent me a shell.

The nine-dimension framework and the conditions that keep it alive

This framework was not built in an afternoon. It began in 2026, when I built an xG model from 180,000 shots across five European leagues and correctly predicted 14 of 16 World Cup knockout matches. It was hardened in 2026, when 312 Bundesliga matches played in empty stadiums became an enormous laboratory: home advantage fell from 54 percent to 47 percent, and home teams' PPDA rose by 0.9. Then in 2026 I carried it into Euro and Tokyo, writing a twelve-part series on Italy with a PPDA of 7.6 and a midfield that covered 4,200 kilometres after the group stage.

Every time it was applied, the framework survived because something fed it: information points. A swim time, a 50m split, a placing, a meet name, a competition date. That is the smallest unit of fact any conclusion must cite. Without it, every dimension is just a table waiting.

This time, the information-point list was empty. Not thin — empty. Even a two-sentence results bulletin yields at least one information point: the line "athlete A won the 200m freestyle" already gives you a name, an event, a distance and a result. Here there was nothing.

Worse, the entity-extraction instruction said to "identify from the information points above" — while above there were none. A self-referential loop. That signal does not come from a thin article; it comes from a cascading fault in the extraction layer.

What each dimension needs in order not to be blank

I tried to sketch the minimum condition for each dimension, and the empty table began to mean something.

The technical dimension needs splits, start reaction time, the first 15 metres, stroke count per length. The performance dimension needs a time plus a course length — 50m or 25m — because short course adds turns and is ratified as a separate record category, never compared directly. The competition-system dimension needs only a meet name and a date, which is enough to locate the result in the Olympic cycle. The world-landscape dimension needs one named entity. The rules and anti-doping dimension needs an identified rule system. The career dimension needs an athlete. The risk dimension needs a subject to hang risk on. The narrative dimension needs a headline — and the headline is the heaviest loss of all, because in swimming the headline carries almost the entire narrative frame.

The notable part: three dimensions — performance, competition system, landscape — can be recovered almost completely from external sources. Given a meet name and a date, the World Aquatics results database and electronic timing sheets will return most of it. A single field — the headline — reopens roughly half the framework. This failure is cheap. It only becomes expensive if we fill it with something else.

The temptation to fill the cells

The nine-dimension framework has one lethal weakness: its form is too persuasive. A table with headings, columns, rows and confidence notes — anyone glancing at it assumes analysis has taken place. If I filled it with a prodigy story, a doping controversy, a record chase, I would have a document that looks authoritative and rests on no factual basis whatsoever. The greatest risk in sports analysis is not a wrong number; it is an empty table presented beautifully enough that people stop asking questions.

That is the number-one risk, and it does not belong to me alone. It belongs to an entire sports press forced to produce faster than it can verify.

Based on my experience tracking matches, a metric is only allowed into print once it has passed three sources. In 2026, analysing all 26 rounds of the V-League for Becamex Binh Duong, I had an average PPDA of 8.4 — lowest in the league — and an xGA of 0.68 per match, with 14 clean sheets. I published with 17 charts, each stamped with its statistical window. The piece passed 250,000 reads, but what held it up was that every data cell could be traced back to a source.

In swimming, the equivalent discipline barely exists in Vietnam. Most domestic swimming reports stop at the medal, the result, one emotional quote from the athlete. No splits, no course length, no reaction time. Nobody is wrong to write that way — but nobody can analyse from it either.

And this is where I have to warn myself: forcing a European model straight onto Vietnamese data is the familiar trap. This framework was born from 180,000 shots in five European leagues, from 312 Bundesliga matches without crowds. It is not automatically correct when placed beside a domestic meet with three conflicting result sources and none of them recording course length. A model is a compass, never a map.

An empty cell is not a clean bill of health

This is what I want readers to remember most, and also what is easiest to misread.

When all seven risk screens return "insufficient information", that does not mean the athlete carries no risk. It means unknown. In anti-doping, the distance between four things — a confirmed violation, a contamination dispute, a procedural violation such as evading testing, and an allegation on the internet — is the distance between four different fates. Merging them is an error that cannot be undone. Silence is not a clean bill of health.

Numbers do not lie, but people always find a way to lie with them. And when there is no number at all, the lying gets easier.

I once treated models as scripture. Now a model is only a compass — but without it, we are lost. What I did not expect was a third state: a compass, but no map, no coordinates, no starting point. That is exactly the state of the forty-seven-cell analysis.

The signal for the next cycle

When the stadium empties, every model collapses. I rebuild from the half-burnt data. This time the half-burnt pile had nothing to rebuild from, so the work was to point precisely at the gap and state the shortest path to closing it.

That path is not long. It starts with a minimum requirement: every swimming report must carry the meet name, the competition date, the course length, and at least one split. Four fields. With those four, a results bulletin becomes data, and data becomes analysis.

The signal I will track next season is not a record. It is the first Vietnamese swimming report that records course length and a 50m split. On the day that happens, the nine-dimension framework will stop returning forty-seven empty cells.

Until then, every "Vietnamese swimming analysis" you read deserves exactly one question: where is your first information point?