An Empty Athletics Dossier and Nine Verification Layers: The Line Between Data Analysis and Fabrication

The screen lit up at 6:40 a.m. on August 13, 2026. The dossier sent back by the data collection team had nine sections, and all nine were blank. Performance section: no measurements. Athlete status section: no names. Competition structure section: no competition. Risk section: no risks listed, only a single line of notes at the foot of the page — "no input data available."

I read that dossier three times over roughly forty minutes before typing the first line of the draft. That line was not a judgment. It was a refusal: you cannot analyse what does not exist.

In twenty-five years in this trade I have seen dossiers fail in every conceivable way. One missing a wind-gauge column. One with times recorded by three different people with hand stopwatches, disagreeing by four tenths of a second. One that copied results from a newspaper article, which had itself copied from an unattributed scoreboard. But never before had I received a document complete in all nine sections and blank in all nine.

Forty minutes before the first line

People assume a data consultant's job is about numbers. It is not. It is about gaps. Most of my actual work is determining where the data stops, so that the rest of the article does not step past that line.

If I had received this blank dossier on August 13 and still insisted on writing two thousand words, I would have had to do exactly one thing: fabricate. I would have had to pick an event, pick a name, assign that name a statistic, and build a plausible-sounding story about that person's performance curve. The draft would have flowed. Readers would have nodded. And every sentence in it would have been wrong in the worst possible way — wrong in a manner nobody could check, because there was no source to check against.

That is why I stopped. Not for lack of ideas, but for sufficient awareness.

Before I believe in a reputation, I need to see the data behind it. A big name is just a big name until someone traces the training logs and junior results of the person carrying it. Without that chain, every compliment is guesswork with decoration.

Context: what Vietnamese athletics is actually recorded with

To understand how a dossier can be blank in all nine sections, you have to understand the data infrastructure Vietnamese athletics currently operates on.

At major international meets, every start generates a complete data package: reaction time off the blocks, hundred-metre splits, airborne time over each hurdle, peak speed, deceleration over the final twenty metres, wind direction and speed measured on site, temperature, humidity, track condition. Those packages are captured automatically and published alongside results.

At most domestic meets, that chain is far shorter. Finish results are captured electronically on some tracks and by hand on others. Wind gauges exist, but not always, and not always read at the correct moment. Splits essentially do not exist at domestic level — nobody measures them, nobody stores them, nobody publishes them.

This does not mean Vietnamese athletics lacks results. It means Vietnamese athletics lacks the kind of results that analysis requires. A SEA Games gold medal, a ticket to a continental championship — those facts are enough to write a news item. They are not enough to write an analysis. The gap between those two kinds of writing is the entire distance between a sports reporter and a data person.

Based on my experience following domestic athletics meets over many years, one recurring feature stands out: data is abundant at large time scales — year, season, Games — and barren at small ones. Everyone remembers a final's result. Almost nobody has retained the average speed of that final's last twenty metres. Yet the small scale is exactly where a performance curve is written.

Nine verification layers, and what collapses first when data is blank

The framework I use for an athletics analysis has nine layers. With complete data, the nine interlock like nine checkpoints on a long-distance run: only after clearing one can you compute the next. With blank data, all nine collapse at once — and notably, they collapse in a very specific order.

Layer one: performance and adjustment

Every number on a track is a conditional number. A 100-metre mark only means something with wind speed attached: a tailwind above 2.0 metres per second means that mark is no longer counted as an official performance. The same athlete on the same track, separated by wind, can differ by several tenths of a second — more than the gap between gold and fourth at a regional meet.

Altitude is another variable, since thinner air reduces drag. Equipment is the third, and this is the point under-discussed in Vietnam. When the generation of super-shoes with highly resilient foam became the international standard, record thresholds shifted, but domestic data does not automatically record whether an athlete was wearing them. Comparing an old result with a new one, without an equipment field, is a comparison with a hole in it.

When layer one is blank, nothing can be done. No measurement, no adjustment. No adjustment, no conclusion about class.

Layer two: the athlete's curve

Layer two asks a simple question: where is this person in their career?

The answer is not in age, but in the year-by-year progression of personal bests. A middle-distance athlete like Nguyen Thi Oanh cannot be read through a single result, because her value lies in her capacity to absorb a dense competition load — something that only appears when an entire schedule is laid out alongside the performance decline between races held days apart. Likewise, a hurdler like Quach Thi Lan needs airborne time over each hurdle to show where time is being lost, not a single finishing mark.

Layer two also asks about injury and peaking. Was a season designed to peak for one meet, or stretched across four consecutive ones? Without a schedule and without workload data, neither question has an answer.

For a distance runner like Nguyen Van Lai, the variables are stricter still: pace distribution across laps, late-race deceleration, the physiological cost of a qualifying round. Without split data, a victory and a defeat can look nearly identical on paper.

Layer three: competition structure and qualification

Athletics has no single door. It has at least three: hitting a qualifying standard, accumulating ranking points, and national federation selection. These three run on three different clocks, and missing one window costs a whole season.

When layer three is blank, entry strategy cannot be assessed. Should an athlete concentrate on one meet to break the standard, or spread across several to accumulate points? That question is answerable only if you know the applicable standard, when the points window closes, and how dense the calendar is in that period.

One further variable is rarely discussed: the choice between an individual event and a relay. At national-team level, an athlete strong on the opening leg of a 4x100-metre relay may be redirected to the team event, and the cost of that decision lies not in individual performance but in recovery time. Without schedule data, that decision cannot be evaluated.

Layer four: the event landscape

Every athletics event has its own landscape: one dominant figure, a two-horse race, a five-deep contest, or a generation in transition. Reading that landscape relies not on names but on group depth.

Bui Thi Thu Thao in the long jump illustrates the kind of data required: not just the best mark, but the distribution of valid jumps within a competition. A long jumper with a very high peak but a high foul rate is an athlete with large variance — and variance, not peak value, determines win probability.

This is where I want to state a principle clearly: a season should be read as a sequence of probabilities, not a sequence of events. A sequence of events tells you who won. A sequence of probabilities tells you who wins next, and with what confidence.

When layer four is blank, there is no landscape map. No sense of who leads the group, where the gaps lie behind, or what threat is rising from the junior pipeline.

Layer five: rules and anti-doping

This is the layer where missing data does the most damage, because it does not merely ruin an analysis — it can ruin a career.

The regulatory framework of international athletics has multiple tiers: technical competition rules, eligibility rules, equipment rules, and an anti-doping system with specific provisions governing eligibility for women's events. A valid check requires knowing which category an athlete falls under, how many out-of-competition tests they have undergone, and which therapeutic use exemptions are in force.

When those fields are blank, the only conclusion available is: no conclusion can be drawn. In this profession, the sentence "no conclusion can be drawn" must be written out, not left blank for the reader to fill in. A gap in a text is always filled with the worst thing the reader already has in mind.

Layer six: team and training system

Layer six looks behind the athlete: coach, training group, periodisation, technology adoption, medical and recovery capacity.

This is a layer where Vietnam has narrative advantage but weak evidential data. Major training centres have gyms, doctors, periodised programmes. But the data proving those programmes work — weekly volume, intensity by heart-rate zone, active rest days — is often not stored as a series. Without a series, nobody can prove anything, including the things that are true.

Layer seven: the risk matrix, where "not assessable" is a verdict

Here I want to pause a little longer, because this is the most misunderstood part.

In a standard risk matrix, each cell has three parameters: level, probability, impact. When input data is blank, every cell reads "not assessable." Many readers see that phrase and assume it is an omission, something unfinished.

It is not. In data analysis, "not assessable" is a complete verdict. It states that the available dataset cannot separate signal from noise, and that any conclusion drawn from it has a higher probability of being wrong than right.

With the dossier of August 13, the biggest risk lay not in any of the nine sections. It lay in the fact that all nine were blank. That is a process-level risk, not a subject-level one — and process risks are always more dangerous, because they do not incriminate themselves. A wrong result about an athlete will be caught. A process that produces wrong results will keep producing them until somebody stops it.

Layer eight: public narrative and expectation

Layer eight asks: what story is being told about this athlete, and does it stand on underlying data?

Familiar narrative labels — emerging prodigy, record assault, comeback, farewell — each have a lifespan. A record-assault label survives only if the underlying series is long enough. A prodigy label survives only if the sample is large enough.

This is where I apply a simple test: strip every adjective out of the story. Does what remains stand? For a young athlete with one good competition behind them, it usually does not. And a narrative that cannot stand will eventually fall — usually onto the athlete.

Layer nine: transmission into the wider industry

The final layer links a competition result to larger things: event commercialisation, equipment technology, individual sponsorship, the junior talent chain, the national-team ecosystem.

This is a layer where I see a genuine gap in Vietnam — not a data gap but a structural one. A good result at regional level does not automatically convert into a funding stream for junior development. Without data on that flow, nobody can measure the effect of a medal on the number of children joining provincial tracks.

When layer nine is blank, an article loses its industry significance. It can still be a news item. It cannot be an analysis.

The counter-intuitive angle: blank data is not neutral

There is a widespread newsroom belief: when data is thin, write lighter, use fewer numbers, lean on emotion. That belief sounds reasonable and is wrong in a very specific way.

A data gap is not neutral. It is a suction. Writers, however careful, tend to fill it with pre-existing archetypes — the prodigy, the comeback, the turning point. Those archetypes do not come from data; they come from collective memory. And collective memory is always available, always appealing, always easy to write.

I worship data, but I pray through real-world verification. That means every time I am forced to fill a gap with interpretation, I set a test question: if a critic held the original dataset, would this argument stand?

There is an example I use when training young reporters, and it sits exactly on the line between measurement and instinct. Athletics starting rules state that a reaction time below 0.100 seconds counts as a start before the signal. The threshold exists to protect fairness, and it does that job well. But it also produces a side effect: athletes are forced to learn to wait, to restrain their natural reflex slightly, because a reflex that is too fast and entirely honest can still be penalised. Measurement, once precise enough, does not merely record behaviour — it changes it.

The same is true of writers. When you know your figures will be cross-checked, you write differently. When you know nobody will cross-check them — because there is no source to check against — you write more loosely, and usually do not notice you are doing it.

This is also why I object to using millimetre-level measurement lines as a moral standard rather than a technical one. A high-resolution measuring system is only good when paired with a matching standard of interpretation. Otherwise it turns writers into editors of reality: adjusting a result to match a number, rather than reading a number to understand a result. In athletics this phenomenon is less famous than in sports with offside lines, but the mechanism is identical.

And here is the most counter-intuitive part: the greatest risk in Vietnamese athletics analysis is not a shortage of results. It is an abundance of results with no underlying source. A dense results table that cannot be traced to its original recording point is worse than a blank table. A blank table is at least honest about its own ignorance.

Luck is the residual the model cannot explain — and I never set it to zero. But luck is not something to be stuffed into gaps at will. When an athlete produces a result better than every forecast, the first task is to re-examine the model's inputs, not to write a tribute to fighting spirit.

Three fields required before writing a single word

After many years, I have reduced my process to three mandatory fields. Without all three, I do not write analysis.

Field one: measurements with conditions. A mark must come with date, time, venue, wind speed, track condition and equipment type. This is the field that was entirely blank in the dossier of August 13.

Field two: a series of at least three consecutive competitions. A single result is a point. Three results are a trend. Only with a trend may the word "currently" be used.

Field three: the original recording source. Who recorded it, when, with what equipment. Without a source, every citation is a citation of a citation, and I have a rule never to let my writing stand on a citation chain with no anchor.

These three fields are not administrative procedure. They are a fence against myself — against the instinct to write a good piece even when there is nothing yet to write about.

Signals to track in the next cycle

From that blank dossier, three signals are worth tracking, and all three sit on the process side rather than the subject side.

First, the arrival of an input dataset with a split field. If domestic athletics meets begin publishing hundred-metre splits for running events, the analytical door opens at an entirely new layer. The trigger condition is clear: at least one meet publishing split data alongside official results.

Second, the arrival of an equipment field in results tables. This is a small technical change with large analytical consequences, because it allows year-on-year comparison without distortion from changing equipment standards.

Third, the arrival of a unified publication standard across organisers. When every meet records in its own way, stitching data into a series remains feasible but costly and error-prone. With a common standard, stitching costs fall and the number of people capable of doing analysis rises.

None of these three signals depends on any athlete's performance. They depend on whether the people organising competitions decide to keep what they are already seeing.

What remains after a blank dossier

The dossier of August 13 did not give me an analysis. It gave me something else, and in a sense something more useful: a reminder that in this trade, the hardest part is not finding the answer, but knowing precisely that you do not yet have enough to answer.

If I had to choose one sentence to place at the top of every athletics report I write, I would choose what I still tell young reporters in training sessions: what cannot be measured should not be written. Not because it is meaningless, but because writing it takes the space of what ought to have been measured.

Tomorrow, the collection team will send another version. I will open it at 6:40 a.m., and the first thing I will do is not read the judgments. I will count the blank fields.