When the Athletics Data Sheet Comes Back Empty

The clock in the corner of the newsroom read 2:14 a.m. I opened the results file that had just been pushed through from a weekend athletics meeting. Every field was there: event name, discipline, weather, wind speed and direction, track surface, athlete list, marks, placings, technical notes. Complete structure. Correct format. Every content field empty.

No athlete names. No marks. No competition date. No source. Not a single line to hold on to. I scrolled down and back up three times, as if the fourth pass might be different.

What made my skin cold was the status line at the end of the file: the system had flagged it as valid. It had cleared every automated checkpoint in the building. Those checkpoints were written to catch data in the wrong format. They were not written to catch data that does not exist.

I sat still for about two minutes. In those two minutes I was not thinking about any athlete. I was thinking about whoever would fill the gap. If that file landed in a night-shift queue, a tired editor would open it, see nothing, and do what humans always do in front of a blank space: fill it in. Not with data. With memory, with inference, with something that looks very much like knowledge.

An empty file does not produce errors. It produces temptation.

From the grandstand to the pipeline

I came into this trade through a spreadsheet. In 2026, aged 33, I left a statistics firm for a young sports media platform in Shanghai. I built a pressing model for Barcelona, calculated success rates by pitch zone, cross-checked against midfield turnovers. When Neymar moved to Paris for 222 million euros, I published a set of proprietary figures: Barcelona's pressing success rate fell from 34.2 percent to 28.7 percent across the first half of the 2026-18 season. Coaches read it. I understood that data can tell a story.

I used to build World Cup spreadsheets before I learned that the grandstand never appears in the formula.

The summer of 2026 in Russia taught me the rest. I covered 17 matches, and my most-read piece was not about the prettiest team at the tournament. It was about a team that played dead football. France won the World Cup with 9 of their 14 goals coming from set pieces. I set those figures beside Didier Deschamps' pragmatism, and the piece reached 2 million reads. For the first time I was not purely a man with a spreadsheet.

In 2026 the stadiums froze. I was in Shanghai, writing about absence, interviewing an athlete from Wuhan who had lost family to the pandemic, and then losing three weeks in which I could not write at all. The essay 'When the Lights Went Out' came out of that silence. An empty stadium does not lack matches. It lacks the soul it borrows from the roar.

Now I report on athletics for the Chinese market from an apartment overlooking the Huangpu. And I have learned one thing about the sport I follow: athletics is the most thoroughly quantified discipline in the entire sports system. Swimming has splits. Football has expected goals. Only athletics turns every extension of a human body into a quantity measured to two decimal places, and then attaches a string of conditions to it: wind, altitude, track surface, temperature, shoe type, start technique.

Because of that, my industry has shifted. A sports reporter used to say 'I was there.' A sports reporter now says 'the system sent it.' Between those two sentences sits a pipeline: collection, extraction, analysis, publication. And on that night, the pipeline ran clean from end to end without anyone noticing it was transporting air.

Nine layers of a post-event analysis

To understand how serious an empty file is, you need to know what a decent post-event analysis must contain. It is not a match report with a few numbers bolted on. It is a nine-layer structure, and each layer has its own failure mode when data goes missing.

Layer one: the true value of a footprint. An athletics mark never stands alone. A legal wind reading does not exceed +2.0 m/s; beyond that, the mark is excluded from any record comparison. Altitude above 1,000 metres thins the air and gifts something to the sprints. Modern synthetic tracks are worth a few hundredths of a second over 100 metres compared with older surfaces. And since 2026, World Athletics has capped road shoe sole thickness at 40 mm and required that any model used in competition be available to the general public. Each of those factors is an unsettled debit or credit.

Noah Lyles won the men's 100 metres at Paris 2026 in 9.79 seconds with a wind reading of +1.0 m/s. Legal. But legal is not the same as neutral. In the marathon, Eliud Kipchoge ran 2:01:09 in Berlin in 2026 and Kelvin Kiptum ran 2:00:35 in Chicago in 2026, both on flat courses, both near sea level, both in the carbon-plate era. Strip those conditions away and you do not have a performance. You have an unverified fragment.

Based on my experience following races and athletics meetings, I keep one rule: any mark that arrives without its measurement conditions goes into the pending drawer. Not the false drawer. The pending drawer.

Layer two: the curve and the age window. An athlete is not a data point but a curve. You need a year-by-year personal best series, a season's best, the gap between the two, an injury history, and a periodisation signal. The peak window in athletics is narrow and fairly stable: sprints around 24 to 29, middle and long distance around 26 to 31, throws around 28 to 33. Place an athlete outside that window without explanation and you are not analysing, you are guessing.

This layer contains the check I trust most and see skipped most often: the personal-best explosion screen. If an athlete's single-year improvement exceeds roughly three times their own historical annual average, that is a signal requiring cross-validation against the competition calendar, coaching changes, training group changes, and the anti-doping layer. Not to accuse. To ask.

Layer three: the qualification architecture. Since 2026, World Athletics has run a dual pathway: hit the entry standard or accumulate enough world ranking points, with a three-per-country quota per event at the Olympics. Those three places create a beautiful, cruel paradox: the world's third-best nation may have nobody in a final, while the world's best nation may have to eliminate the world number four at its own domestic trials. The American system, where one race decides everything, maximises drama and maximises occupational hazard. Comprehensive selection systems flatten the risk and trade it for the danger of picking the wrong athlete in the most important month.

Competition density is another variable. An athlete racing five times in six weeks before a World Championships is trading cash for opportunity. Without the calendar in hand, nobody has the right to judge that trade.

Layer four: the competitive map. Any event splits into four tiers: dominant, title contention, finals, and qualification fringe. Drawing those tiers requires personal bests and season bests for eight to twelve names. To talk about the future you need the talent pipeline: the American collegiate system, the Jamaican school system, the East African altitude camp chain, or China's professional team system. Those four models produce four different kinds of athlete, with different ages of maturity, different tolerance for volume, and different ways of celebrating and of accepting defeat.

Layer five: rules and anti-doping. This is the layer where emptiness is most dangerous. The athlete biological passport, whereabouts obligations, a sample storage window of up to ten years that allows retrospective retesting and medal reallocation, differences of sex development regulations, the three-year waiting period on nationality changes, authorised neutral athlete status. Each is a gate an athlete can hit, and each requires a named subject to mean anything.

There is an asymmetry I want to state plainly: a screen that could not run must never be reported as 'no issue found'. It must be reported as 'not screened'. In practice, those two sentences are read identically. That is the most serious error an analysis can make.

Layer six: the training system. The altitude camps at Kunming Haigeng, the Duoba base in Qinghai, the East African winter camps: those places determine training calendars and competition calendars alike. An athlete who spends six weeks at 2,300 metres has a structurally different season from one training at sea level. Then there is technology adoption: lactate testing, GPS load monitoring, wind tunnels, 3D motion capture. Above all sits the biggest single risk variable before a championship: a coaching change.

Layer seven: the risk matrix. Six categories, covering competition, doping, finance and career, eligibility, public opinion, and system-level exposure, plus a seventh I only learned on that night: process risk. An empty analysis, used without a gate, will generate its own conclusions. High probability, high impact, and the only mitigation is to stop.

Layer eight: the public narrative. Every event travels a cycle: germination, acceleration, peak, backlash. In athletics that cycle is governed by a very specific exposure law: cold in odd years, red hot in Olympic years. April of an odd year and April of an Olympic year are different worlds, even though the track is still 400 metres long. The most important check here is the hype kill: when someone is labelled a successor, that label must be placed beside a value-adjusted mark. If the label cannot stand next to the mark, the label belongs to the newsroom, not to the track.

Layer nine: industry transmission. Super shoes are born on the elite track and flow into the mass running market within about eighteen months, and a city marathon can remake a local tourism economy. The Diamond League is redefining itself against a new circuit founded by a former athlete. The naturalisation value chain operates on rules nobody has written down. Every one of those channels needs an anchor: an athlete, a meeting, a contract, a sourced figure.

The paradox of emptiness

We have built an entire industry to catch wrong numbers. There are cross-checks, copy editors, confirmation requests to federations. Almost nobody has built anything to catch missing numbers.

That is a structural blind spot. A validation system is designed to answer the question 'is this value correct'. It is not designed to answer 'does this value exist'. When every field is empty, the system returns 'no errors'. That sounds like a clean bill of health. It is actually a blank sheet of paper.

In sports journalism the most dangerous abbreviation is not a swear word. It is the three letters used to mark an empty cell. Those letters get read as 'no problem' when their real meaning is 'nobody has looked'. An athlete with no injury history in the database might be the healthiest person in the field, or might be someone nobody ever recorded. On screen, those two people look identical.

There is a commercial reason this blind spot survives. A 24/7 publishing cadence rewards filling. A piece with content, even content built on sand, generates reads today. A piece saying 'we do not yet have the data to conclude' generates nothing and is read as weakness. The incentive structure runs against quality, and it runs very quietly.

I would argue the greatest danger is not wrong data. Wrong data leaves traces: a record that does not reconcile, a competition date that never happened, a name nobody can trace. Empty data leaves nothing at all. It goes undetected because there is nothing to detect. And so it spreads silently through bulletins, rankings and commentary, until a system downstream cites it back and turns blank space into fact.

One more thing needs saying about the writer's side, not just the machine's. I have stood in a grandstand and watched an athlete limp off the track, then opened the results sheet and found no line recording it. The results sheet records marks. It does not record the price paid. A reporter who only reads the sheet will write a completely different piece from the one who was there. Both are correct in their own way, but only one of them is journalism.

That is why I still arrive two hours early, stand by the warm-up area, and watch how people tie their spikes. Not out of ritual. Because that is the only data the pipeline cannot take from me.

At 42, I have learned that every athlete is a poet who never gets published.

What to build next

There is a small technical fix that could eliminate an entire class of failure: a two-field gate that blocks any file with a null title or fewer than one information point. It is cheap, it is fast, and it would make nights like that one impossible to repeat in silence.

The technical fix is the easy part. The hard part is building a culture in which recording what you do not know counts as part of the product rather than an apology. An honest athletics report should have room for the sentence 'this is undetermined', placed on equal footing with marks, placings and wind readings. When blank space is publicly marked, it becomes part of knowledge. When it is filled in, it becomes part of a false history.

Sport is the common language of people who do not share a language, and for exactly that reason it carries a higher-than-normal duty of accuracy. A 100-metre track is 100 metres in every country. If we let numbers that do not exist slip between the athlete and the audience, we damage the only language the whole world can read.

That night I published nothing. I wrote one line in my notebook: today the pipeline delivered air, and nobody noticed for six hours. Tomorrow, someone may write a complete article about that air. My job is to make sure they cannot.