Nine Layers of Athletics Data: The Discipline of Saying 'Not Enough' Before Concluding

5:40 a.m., 14 July 2026, Nha Trang. Outside the window the sea is still grey. I reopen the data sheet from a national athletics meet held three months earlier, looking for the men's 100m final. First place: 10.28 seconds. The stands erupted that afternoon, the roar bouncing from block B to block D, and I remember I did not shout. My eyes were fixed on an empty cell sitting right beside that number — the wind column.

That blank was never a typo. It is the entire difference between two different athletes. With a tailwind of 3.1 metres per second, 10.28 seconds is worth roughly 10.45 seconds in still air. With a headwind of 0.4 metres per second, it is worth 10.23. One athlete has never broken 10.4. The other is knocking on the door of 10.2. Same line of text, two men, two entirely different futures.

When the numbers speak, I simply listen. But that morning, the number said nothing. It stayed silent, and I had to stay silent with it.

Context: twenty-three years learning to keep quiet

I entered the trade in 2026 at Runner's World, when a piece about running had to carry at least one comparison table to get printed. I then spent nearly a decade covering athletics for SI, and in 2026 I received the SJA Young Sports Journalist of the Year award. People assume that award came from some grand investigation. It actually came from a very small piece about how one 800m runner distributed his pace across two laps, in which I gave almost half the space to explaining why I could not draw a conclusion.

In 2026, aged thirty, I became a data consultant for a new football site. Hanoi Club were being called boring. I broke down their first fifteen matches: an average xG of 1.2 but 2.1 goals per game, with the surplus of 0.9 concentrated almost entirely in shots from inside the box taken by Nguyen Van Quyet. I wrote that they would win the title through chance quality rather than luck. The piece got buried in criticism. Then they lifted the trophy. The lesson I took was not that I was right, but that data only tells the truth when you are willing to read all of it.

In 2026, aged thirty-one, I worked as an analyst for a television channel at the World Cup in Russia. Before the semi-final I tallied Luka Modric at 88.3 km covered across the tournament, with sprint speed falling 21 per cent after the seventieth minute. I wrote that Croatia would reach the final if they used all three substitutions to cover that distance. Colleagues laughed. Croatia beat England after extra time, and the editor handed me a new match-by-match data column.

Moving into athletics, I brought the old habits but had to replace the whole toolkit. Football has xG and PPDA — the number of passes an opponent completes before you win the ball back, lower meaning more aggressive pressing. Athletics has none of that. Athletics has only one raw truth: time, distance, and the conditions that produced them. So I built a nine-layer framework, and its logic runs against instinct. People use tables to fill gaps. I use tables to find which gaps are empty, and what that emptiness means.

Based on my experience tracking matches and athletics meets over more than two decades, I can state one thing plainly: most errors in sports analysis come not from reading a number wrongly, but from reading a number that does not exist.

Layer one — Event and performance

This is the foundation, and it demands more than people assume. To assess a mark, I need the discipline: track, field, throws, or combined events. The distance. The round — heat, semi-final, final. The mark to the hundredth. And most importantly, the conditions that produced it.

In sprints and jumps, wind speed is decisive. World Athletics rules state that a tailwind above 2.0 metres per second still allows a result to stand as a competition performance, but not as a record. Over 100m, one metre per second of tailwind is worth roughly 0.05 to 0.06 seconds. That sounds small, but in an event where a national record is often improved by hundredths, it is a wide horizon.

Altitude matters too. Competing above 1,000 metres clearly helps sprints and jumps, because thinner air means less drag. Over 400m and beyond the advantage grows. Then there is equipment. The generation of carbon-plated shoes with super-elastic foams, widespread from around 2026, reshaped every distance record between 2026 and 2026. A marathoner in 2026 and a marathoner in 2026 ran on two different technical surfaces, at the same distance.

Finally, the track itself. Every surface returns energy differently, and some are built to give back far more than average.

At this layer I have not said whether the athlete is good or bad. I have only established what the number means. Distance never lies; we simply have not been patient enough to listen.

Layer two — Athlete condition

If layer one answers what the mark is, layer two answers where it sits within an entire career.

The first tool is the personal-best curve year by year. Not one number, but a string. I need at least four or five consecutive seasons to see the shape. A rising talent has a steep curve. Someone at their peak has a flat one. Someone declining has a curve slipping to the right.

Here is a check I use often and always phrase with great care: if a personal best improves in one year by more than roughly three times the average annual gain across the career to that point, that is a signal to look closely. To look, not to accuse. Some jumps are fully explained by a coaching change, a new training group, a shift in distance, or simply eating and sleeping properly for the first time.

The second tool is the age curve. Sprints peak around 24 to 29. Middle and long distance around 26 to 31. Throws later, around 28 to 33. Placing a mark on that curve tells me whether it is ordinary, early, or late.

The third is competition and injury history. An athlete who withdrew from two consecutive seasons is a high-risk case, and every later mark needs a note attached. No exceptions.

The fourth is current-season form, measured as the gap between this season's best and the career best. If that gap narrows meet by meet, the peaking is on schedule. If the season best came in March and never moved again, something happened in the months between.

Layer three — Competition structure and qualification

A common mistake is judging a national team by who finishes fastest. What decides who competes on the biggest stage is the qualification system, and it is far more complex than fans imagine.

At the highest level each event has two parallel paths. The first is hitting a qualifying standard inside a set window. The second is accumulating enough world ranking points across a season. Open either door and the ticket is yours.

In some countries the internal selection model is harsher than the global standard. The American model is the clearest example: the ticket comes from a single race, which means a reigning world champion in fine form can miss the team after one bad afternoon. For forecasting, this is structural risk rather than random risk.

One more variable: each country may enter a maximum of three athletes per event at the world championships. In events dominated by one nation this creates brutal internal competition, and that nation's fourth-best — sometimes the planet's fourth-best — stays home.

Then there is competition density. An athlete racing five times in eight weeks and one racing twice in eight weeks cannot be compared directly. A heavy schedule can produce a run of good marks, but by late season the cost appears in the final laps.

Layer four — Event landscape and national strength

Every athletics event has its own geography, and that geography shifts slowly but it does shift. Jamaica and the United States hold the sprints. Kenya and Ethiopia dominate middle and long distance. The United States has deep field-event talent, while Europe stands out in the throws. China has a tradition in race walking and women's throws.

Within that landscape, a memorable marker is the 9.83 seconds run by Su Bingtian at the Tokyo Olympics on 1 August 2026, setting an Asian 100m record. Another is the durable shot put cycle of Gong Lijiao, spanning multiple Olympic quadrennials.

What I look for at this layer is the shape of the contest. There are four basic types. One absolute ruler, where the title is nearly predetermined. A two-horse race, where two names create a rivalry lasting years. A wide-open melee, where six or seven athletes can win. And a generational transition, where the leading group is ageing and the next one is arriving.

The fourth is the hardest to read and the most often misdiagnosed. The tell is not in the leader's mark but in the average age of the top ten. If that average rises steadily over three years while the leading mark stays flat, a transition is coming, and it will arrive sooner than expected.

Layer five — Rules and anti-doping

This is the layer where journalism taught me most. One point must be stated clearly: finding no doping-related content in an article does not mean the article certifies anyone as clean. When data is missing, this layer's status is "unassessed", not "cleared".

The central concept is the athlete biological passport. Rather than waiting for a banned substance to appear in a urine sample, anti-doping authorities build a biological baseline for each athlete over time, and any abnormal deviation from that baseline is a signal to investigate. Samples are stored for up to ten years, meaning a sample taken this year can be reanalysed in 2036 and medals reallocated long after the fact.

On technical rules, athletics has hard thresholds. A start before the gun means instant disqualification, with no second chance. Stepping on or over a lane line in a lane-restricted event means elimination. In relays, the baton exchange has a limited window and a handover outside it voids the result. In pole vault, both equipment specification and bar placement are regulated.

Three further areas belong here: eligibility related to biological sex, nationality transfer procedures, and equipment control. Each can completely change how a mark should be read, and each is routinely skipped in short news items.

Every number is a confession the competition cannot deny. But that confession is only worth something if the interrogator knows the room is empty.

Layer six — Team and training system

No track athlete produces marks without a system behind them. This layer asks four questions: who coaches, where they train, on what cycle, and who surrounds them.

Development models vary. The state-centralised model, where athletes live and train in one centre. The club model. The American collegiate model, where athletes study and compete inside a dense school calendar. The private group led by one coach. And the altitude-camp model, most common in East Africa, training year-round at roughly 2,000 to 2,500 metres.

Each has strengths and weaknesses. Centralisation allows control but breeds dependency. The collegiate model offers competitive depth but a punishing schedule can erode long-term physical foundations. Private groups are flexible but rest on one person.

Two indicators I always look for here are staff stability and the level of recovery technology. Constant staff turnover signals instability. Strong recovery support explains many apparently strange phenomena, such as a thirty-two-year-old maintaining the racing frequency of a twenty-five-year-old.

And then there is key-person risk. An athlete who works with one coach for eight years, then loses that coach, faces a bigger event than any minor injury.

Layer seven — Risk landscape

This layer gathers everything into a matrix. Each row is a risk type, each column an attribute: probability, impact, mitigation.

Competitive risks include opponent form, the sudden arrival of a new name, and schedule changes. Health risks include accumulated injury and overtraining syndrome. Legal and anti-doping risks include whereabouts and testing-schedule violations. Media risks include public pressure after a defeat. Organisational risks include funding shortfalls, lost sponsorship, and federation infighting.

The key here is not letting the matrix become a decorative list. Every row must attach to an observable event. If a risk cannot be observed, it belongs in layer eight, not here.

Layer eight — Hidden information

This layer reflects what every data sheet leaves unsaid. An athlete can have a beautiful progression curve, a sensible schedule, a stable team, and still be carrying an injury nobody outside the clinic knows about.

I approach this layer through indirect indicators rather than free speculation. The time gap between consecutive appearances. A withdrawal without official announcement. A sudden change in coaching staff. A season in which an athlete contests exactly one signature event and then stops.

Each indicator only counts when it appears at least twice. Once is coincidence. Twice is a pattern. I do not believe in luck; I believe in what has been repeated enough times.

At the seventieth minute, the crowd sees collapse; I see a structure being rebuilt. But to say that sentence, I need at least two independent indicators pointing the same way, and I need to be sure both come from real data.

Layer nine — Source quality and the chain of evidence

This final layer is the least discussed and it determines the value of the eight before it.

Primary sources include official organiser records, wind readings logged by automatic devices, electronic timing systems, and athlete entry files. Secondary sources include press reports, statements, and interviews. Non-recognised sources include hand-timed watches and any mark recorded in training.

This is the point I want to stress most. A mark recorded on a hand-held watch, with no wind gauge, no officials, no written record, does not exist. It may be a lovely memory from a lovely morning, but it is not data. Every season I receive at least a few messages along the lines of: I just ran 10.1 in training. I always reply with the same question. Which watch, how much wind, who pressed the button.

A chain of evidence strong enough must satisfy three conditions. Each link must be traceable to a source. The links must be independent, meaning they do not all descend from a single article. And the chain must survive a reverse test: if the original hypothesis is wrong, does the data still hold?

People ask me why I stay silent; I am reading the lines the field writes.

The counter-intuitive angle: a blank sheet is not a clean bill of health

Let me return to that empty cell.

A young writer's first reflex is to fill it. With memory, with the account of someone who was in the stands, with feeling. That is the moment the piece becomes worthless, because a question without an answer has been converted into an assertion.

The second reflex, more dangerous, is to fill it with a contrarian conclusion for effect. When everyone says an athlete is finished, we declare he is restructuring. It sounds very Data Monk. But if only one indicator supports it, we have traded truth for style.

The third reflex, the most dangerous, is turning correlation into causation. An athlete changes shoes and breaks a personal best. The easiest conclusion is that the shoes bought half a second. In reality, the first athletes issued a new shoe generation tend to be the best sponsored, training in the best conditions, with the best support staff. The shoes come with an entire ecosystem. Separating the shoes from that ecosystem is bad arithmetic.

My self-imposed rule is simple: before reversing a judgement, I must find at least two independent indicators pointing the same way, and both must pass layer nine.

What I carry into the next cycle

That blank cell from the morning of 14 July 2026 is still blank. I have sent three emails to the meet organisers asking for the wind readings from the men's 100m final. No reply yet.

But I will not publish a claim about the man who ran 10.28 until I have an answer. If the tailwind exceeded 2.0 metres per second, it is a story about reading numbers correctly. If the air was still, it is a completely different story, and possibly the biggest one of the season.

While I wait, I write three lines in my notebook. First, track his mark sequence across at least four more competitions. Second, check whether there has been any coaching change in the past eighteen months. Third, reread the regional entry lists to see whether he is being selected at all.

That evening, his coach sent me a message. He wrote: thank you for not asking how he feels. You asked how much wind. I read it twice before replying: it is the least I can do.

The field does not need a cheerleader. The field needs someone willing to sit with the empty cell, to admit not knowing, and to wait. The annual season is long. And the wind in Nha Trang, as I have learned over years of watching outdoor competition, has never blown in a fixed direction.

People ask me why I stay silent; I am reading the lines the field writes. This time the line is still an empty cell, and I choose to leave it that way.