Five Variables That Make Athletics Results Misread

An eleven-second clip of a 100-metre run landed in my inbox four times in a single week, from four people in four different provinces, all asking the same question: is this result real?

I watched the clip seventeen times. The frame showed the lane, the stands, and a digital clock in the right corner. Three things were missing. No wind reading. No information about the altitude of the venue. No data on the athlete's shoes.

To a casual viewer, those are minor details. To anyone working with data, they are mandatory variables. A sprint result is not a single number; it is a coordinate.

The day football stopped, I started counting strides again. The more I counted, the more I realised that most arguments about athletics results have nothing to do with an athlete's ability. They have everything to do with missing data.

I entered the profession through a football team's spreadsheet. In 2026, while working as a data consultant for Hai Phong Football Club, I reviewed the youth team's metrics and found a midfielder with an average PPDA of 6.8 — the highest in the academy. He pressed better than anyone in the squad but nobody noticed him, because his physique was modest. I brought the table into the meeting room. Against Ha Noi FC on matchday 17 of the V.League, he won the ball 14 times, provided one assist, and the team won 2-1. Since that day, I no longer read players through their reputations.

In 2026, when competitions stopped, I spent four months reviewing 2,300 matches from the V.League and three major European leagues, building a pressure index that combined PPDA, defensive distance and pressing speed. Teams with a PPDA below 8.5 averaged 1.8 points per match in my sample, well above the rest. That model left me with a habit I carried intact into athletics: before trusting a number, know how it was measured.

Athletics has an advantage over football in that almost everything is standardised. The track is 400 metres long, the starting line has sensors, timing is electronic, and World Athletics stores the results of most sanctioned meets in its own system. Because of that standardisation, athletics creates an illusion: viewers believe a number on a screen is a complete truth.

It is not complete. A running performance is a coordinate made of at least five variables: wind speed, altitude above sea level, shoe type, competition density within the week, and split data. Remove one variable and I misread the entire chain of comparison that follows.

The +2.0 m/s wind threshold is a hard condition for ratifying any record. World Athletics only recognises a record when the measured wind speed does not exceed +2.0 metres per second. This is not a technical recommendation; it is a legal condition.

The clearest marker is Usain Bolt's 9.58 seconds in Berlin on 16 August 2026, with a wind reading of +0.9 m/s. That number has survived more than fifteen years because it came with a valid wind measurement. Imagine the same athlete running 9.80 seconds in a +3.4 m/s wind. A wind difference of roughly 2.5 m/s can be worth 0.10 to 0.12 seconds over 100 metres. Most of the gap between the two runs sits in the weather, not in the legs.

I do not see an athlete losing. I see a number that does not lie — and a number that cannot introduce itself by explaining the conditions it was measured in.

Altitude is the second variable, and it is routinely ignored at domestic meets. Thin air reduces drag. Bogota sits at roughly 2,640 metres above sea level; Mexico City at about 2,240 metres. In those places, the same athlete can sprint faster and jump farther than at sea level. I always check altitude before I check speed.

This does not make altitude performances worthless. It places them in a different data class, and a different data class requires a different reference table.

Shoes are the third variable, and the youngest in the sport's history. In January 2026, World Athletics issued regulations on competition shoes. For track events, the maximum sole thickness is 25 mm. For road events, the figure is 40 mm. Another clause requires the model to have been available on the market for at least four months before an athlete uses it in official competition.

Those rules followed a genuine dispute across 2026 and 2026, when a wave of road records fell and analysts split into two camps: one arguing it was the product of training, the other arguing the carbon plate did most of the work. When reading a new performance, I always ask which shoe model the athlete wore. Skip that question and I am adding two different quantities into the same column.

The fourth variable is sample size. One good run does not create a class. When evaluating an athlete, I need at least five to ten races over twenty-four months, under comparable conditions, and then I calculate the standard deviation. An athlete whose personal best is 10.20 seconds but who ranges between 10.20 and 10.90 is a completely different case from one who runs a steady 10.45 for two years.

Peak performance attracts attention. Consistency is what predicts the next round.

The fifth variable is split data, and it is the biggest gap in Vietnamese reporting. For 100 metres, I need reaction time, top speed and the position where top speed occurs. For 400 metres, I need four split points. For the 400 metres hurdles, ten hurdle gaps and the stride rhythm between them show exactly where an athlete loses time. A runner like Nguyen Thi Huyen, a multiple SEA Games champion in the 400 metres hurdles, competes through a rhythm structure; without split data I only see the final result and never the cause.

At SEA Games 31 in Hanoi in 2026, Nguyen Thi Oanh won gold in both the 1,500 metres and the 3,000 metres steeplechase in the same competition session. Only a few dozen minutes separated the two races, meaning the second was run on legs that had not recovered. Competition density is a tax on the body, and any comparison table missing that variable is comparing two things that are not the same.

There is one more category I keep separate: marks that were never ratified. On 12 October 2026 in Vienna, Eliud Kipchoge ran a marathon in 1 hour 59 minutes 40 seconds, the first time a human being covered 42.195 km under two hours. That performance was not ratified as a world record. The run was purpose-built: not an official race, with a pacing car and a rotating group of pacers shielding him from the wind. Kipchoge's official world record is 2:01:39, set in Berlin in 2026, later improved by him to 2:01:09, also in Berlin, in 2026.

All three numbers are real. They belong to three different categories, and mixing them into one sentence is a serious data error. The same thing happens when training clips circulate as if they were competition results.

Now I have to argue against myself. Checking data too strictly is also a form of error, and it costs more than I once assumed. In 2026, had I evaluated only crude indicators such as height and weight, I would have overlooked the best pressing midfielder in the academy. A data filter must serve discovery, not only rejection.

A more common mistake is reading correlation as causation. Teams with a PPDA below 8.5 averaged 1.8 points per match in my sample — that does not prove pressing produces points. Strong teams tend to press well, and strong teams also tend to win. To separate the two, I have to control for squad quality, fixture list and home advantage. In athletics the same logic appears: an athlete who runs faster after changing shoes may not be faster because of the shoes. It could be a higher training volume, or simply more races that year.

Data is a mirror. Most of the market looks into it and sees only itself.

Three signals I will track in the next cycle: the wind reading for every published run; a minimum sample of five races across twenty-four months before any evaluation; and split data for all events from 400 metres upward. If a clip lacks all three, I do not argue. I wait.

People call me a data monk. A monk needs no cathedral – only the truth.

The missing data in this article is substantial: I do not yet have routine wind readings for domestic athletics meets, a competition shoe inventory for each athlete, or standardised split data for most events. When those gaps are filled, my conclusions may change. That is a condition of the work, not an apology.