Athletics Analysis Without a Data Foundation: The Fragile Line Between Inference and Fabrication

At a press centre in Nairobi, I had fifteen minutes to file a story on the women's 3000m steeplechase. On the big screen, the home runner crossed the line amid cheers. In the folder handed out by the organisers, there was a single final time. No 400m split table. No wind reading. No note on how she handled the water jump on each lap. Just one number, and a vague feeling that the decisive moment had happened on the penultimate lap, though I could not point to exactly where.

We are taught to write with a coach's eye. But that eye, without data to anchor it, drifts very quickly toward fiction. I have read pieces about inner endurance and final-stretch character as if they were measurable quantities. I used to write that way too. Until an older editor placed a draft on my desk and called it by two words: no foundation.

At forty-two, after twenty-six years following tracks from Vietnam to Kenya, I understand that the value of an athletics analysis does not lie in its length or elegance. It lies in its testability: a measurable figure, a verifiable condition, and a traceable date. When those three disappear, all that remains is feeling. And feeling is the fastest thing to spoil in a crowded press room.

More than two decades ago, when I began writing for a running magazine, athletics analysis was almost purely visual. You sat in the stands, counted strides, noted positions each lap, then wrote it up at home. That method had its own honesty, but also a fatal flaw: the human eye cannot distinguish three-tenths of a second per lap, and certainly cannot detect a tailwind that has just crossed the legal threshold.

Today everything is different. Tracks carry sensors. Wind gauges sit beside the straight in sprint and jump events. Real-time positioning systems produce splits down to the metre. Data is no longer a luxury of major meets; it has become the foundation of every serious conclusion. And precisely because of that, a paradox appears: as data becomes abundant, analysis without data becomes more dangerous, not less.

I have lived through this in Kenya, where athletics is an informal industry that nonetheless dominates sporting life. In Iten, in Eldoret, training camps stand at more than two thousand four hundred metres above sea level. Young runners log long mornings before sunrise, and their results travel by word of mouth faster than they are recorded. I once sat in a tea shop in Iten listening to people describe a seventeen-year-old who had just run an unbelievably fast session, and realised that if I retold that story without a stopwatch, I was doing the work of a storyteller, not an analyst.

In Vietnam the story differs. The training system is more organised, but the data infrastructure is thinner. At domestic meets, obtaining an official split table is still a luxury. Reporters often time runners themselves from the stands, then cross-check hours later against published results. The difference between Kenya and Vietnam is not talent. It is the availability and reliability of data. And that difference determines which kind of analysis can be written honestly.

This is why I want to spend this piece on something rarely discussed: the cost of analysing without a data foundation. In recent years I have watched more and more athletics writing assembled from scattered fragments, then draped in the cloak of certainty. That misleads readers, and it erodes the very thing this trade needs most: trust in numbers.

If I had to compress serious athletics analysis into one image, I would choose a bridge with five pillars. Remove any one, and the bridge still looks fine in a photograph, but collapses the moment real weight is placed on it.

The first pillar is the mark. It is the most basic unit, and the most misunderstood. A mark says nothing on its own. It only means something against a reference frame: personal best, season's best, national record, world record. The gap between the season's best and the personal best indicates current form relative to one's own peak, and it often tells a story the mark itself cannot. A wide gap means the athlete is rebuilding rhythm. A narrow gap means near-peak form.

When I write about athletes such as Eliud Kipchoge or Faith Kipyegon, names that put Kenyan athletics on the world map, I always begin by placing their marks in a temporal context. Kipchoge has run marathons for years, and his value lies not in one fine race but in consistency across dozens. Kipyegon has won multiple titles in middle-distance events, and each mark only means something compared with her own seasons. That is how numbers become stories.

The second pillar is the wind reading. It is the most neglected in popular writing, and the one that can overturn a conclusion in a single line. In sprint and jump events, a mark is recognised only when the tailwind does not exceed two metres per second. Beyond that threshold, the mark is flagged as wind-assisted, cannot count as a record, and cannot be compared directly with a mark set in still air. A sprinter who runs 100m with a +2.1 m/s tailwind may be raised by media into a phenomenon, when in truth most of the difference came from the wind.

I remember an evening at a youth meet, when a young man ran a mark that brought the crowd to its feet. People around me cheered. But the wind official bent over the gauge, then shook his head. The reading was over the limit. That beautiful mark became, in the system's eyes, a number that could not be used for comparison. The boy still ran fast. But the story about the boy changed entirely, because of one data line almost nobody in the stands noticed.

The third pillar is altitude. At stadiums above one thousand metres above sea level, the air is thinner, drag is lower, and sprint and jump marks benefit naturally. Endurance events, conversely, suffer, because oxygen supply to the muscles falls. This means a middle-distance mark set in Bogotá or Addis Ababa cannot be placed alongside an equivalent mark at sea level without a correction step.

Understanding this is exactly why Kenyan training camps sit at altitude. They want athletes' bodies to adapt to oxygen scarcity so that, when they descend to race at sea level, they have extra headroom. But that effect does not automatically convert into marks. It merely creates a baseline condition, and baseline conditions must be measured, not assumed.

The fourth pillar is venue and competition conditions. The same athlete over the same distance can run faster on a firm, springy track, in cool temperatures, in a race with a suitable leading pace. This factor is usually lumped under form, but it is in fact a set of observable variables: temperature, humidity, wind, track quality, and the tactics of those running alongside. A mark set in an evenly paced race is fundamentally different from one set in a slow race that ends in a sprint.

The fifth pillar is the date. In athletics, time is not merely a calendar entry. It is the qualification window. Each Olympics or World Championships recognises marks achieved within a specific period. A mark outside that window, however beautiful, has no selection value. A mark set at a meeting not on the recognised list is the same. This is one of the most common sources of error in media: confusing achieving the standard with being eligible.

Once the five pillars stand, analysis can begin. And only then may higher tools be used.

The first tool is the age curve. Every athlete has a trajectory: some peak at twenty-three, some at thirty. But there is a signal analysts call an abnormal performance explosion, when an athlete in a single year improves many times faster than their own historical rate. That is not automatically a negative sign, but it is always a point requiring re-checking, because the human body does not improve in a straight line.

The second tool is the qualification mechanism. At major meets today, entry can come via two routes: achieving the standard, or accumulating world ranking points. The two demand entirely different strategies. A standard-chasing athlete will pick a meet with ideal conditions to run one qualifying mark. A ranking-chasing athlete will spread marks across many meets to accumulate points steadily. Misreading an athlete's route leads to misreading their entire schedule.

I saw this with Vietnamese athletes such as Nguyen Thi Oanh, known for racing across multiple events. Her multiple distances at a single Games were not an impulsive choice. They were the result of calculations about schedule, recovery capacity, and medal targets. Reading those calculations correctly requires more than looking at a results table.

The third tool, and the most sensitive, concerns competition rules and anti-doping. Some rules look small but carry enormous weight. Since 2026, a single false start in a sprint is enough for disqualification. In relays, an error in the exchange zone can wipe out an entire race. And deeper still, an athlete's biological passport tracks blood and steroid markers over time, detecting anomalies a single test would miss. In the past decade, more than a few medals have been reallocated to runners-up after the original medallists were disqualified for doping.

All these tools only operate when data exists. No mark, no age curve. No wind reading, no valid assessment. No qualification window, no conclusion about selection. That is why I say an analysis without a data foundation cannot be called a weak analysis. It is a fabricated analysis, dressed in the cloak of professionalism.

This is where I want to tell a story from my own trade. A few years ago, I received a draft analysing a race in Africa. It was fluent, full of fine phrases: explosive speed, iron will, a champion's character. But when I opened the attached data sheet, I found a void. No meet name. No date. No specific mark. No athlete name. The data page was blank, while the interpretation was overflowing. I returned the draft with a single question. What exactly are you analysing. The writer could not answer.

I tell this story not to belittle anyone. I tell it because I have been in that position. In content work, the pressure to file is real. And when there is no data, the fastest way to fill the gap is to write about emotion. Emotion is always available. Data is not.

If the piece stopped here, it would fall into another trap: the worship of data. That is what I want to argue against.

In recent years a new analytical school has emerged, built on advanced indices. In football, expected goals is used to assess chances. In athletics, predictive models use splits and fatigue indices. These tools have real value. But they share a fatal limit: they measure what happened, not what almost happened.

An expected-goals figure can say how often a shot from that position ends in the net. It cannot say that the opposing defender got a fingertip to the ball in the final second. A running-performance model can produce a fine number, but it does not know the athlete slept three hours the night before from anxiety. Data measures the tip. The submerged mass of the iceberg, decisions, hesitation, the moment of changing one's mind, lies beyond its reach.

This is the central paradox of athletics analysis. Data is a necessary condition, but never a sufficient one. An analysis with only data is like a map with coordinates but no terrain. Technically accurate, yet useless to anyone who wants to cross that land.

I once witnessed this at a meet I was following. One athlete was rated by the models as the top contender, based on steady season form. But in the final she ran unusually cautiously, and lost in the closing metres. The models could not explain it. Only a human could. I later learned she had just been through a family crisis, and choosing to run safe rather than gamble was a rational decision of her own, even if it contradicted every forecast.

That story taught me something. When data and the human story collide, do not rush to trust the data. Check both. The gap on the track is a living thing, and it changes when someone dares to believe. The gap in the data is the same. It is not proof of emptiness. It is an invitation to look further.

And this is where I want to go a step further. In modern sport there is a troubling tendency: treating the absence of data as confirmation. When an analysis file is empty, people easily read it as no problem. When no red flags are raised, they easily assume no risk. But the absence of information does not equal the absence of risk. It merely means we have not yet seen anything.

In anti-doping, this is a life-or-death principle. A negative sample does not mean an athlete is absolutely clean. It only means that sample detected nothing. This is exactly why the biological passport was created, to track markers over time rather than rely on a single test. And by the same logic, an analysis report full of insufficient information should not be read as everything is fine. It should be read as not yet determinable.

I think this is the greatest lesson a sports content maker can learn from the world of data. Honesty does not lie in having many numbers. Honesty lies in admitting when you have none.

Back to that evening in Nairobi, with fifteen minutes and a single number on the screen. That day, I did not write an analysis. I wrote a short news item, stating the mark, stating the conditions I lacked, and leaving the interpretation open. I knew readers would find it drier than other pieces. But I also knew it was the only piece I could defend if someone asked again.

If you are a reader who loves athletics, I want to leave you a self-check. Next time you read an analysis of a race, ask yourself: does this piece give me a specific mark, does it mention conditions, does it mention the date. If all three are missing, you are reading a piece built on feeling. And feeling, however beautiful, cannot run one hundred metres.

Athletics is the sport of gaps. The gap between two athletes on the track. The gap between a mark and a record. And the gap between what we know and what we think we know. A good analyst is not the one who erases that last gap. He is the one who keeps it open, so people can step in and check again. The gap on the track is a living thing, and it changes when someone dares to believe. The same is true of the gap in data. Both only become useful when we are brave enough to look straight at them, instead of filling them with elegant but hollow words.