Empty Cells on the Data Sheet: Why the Most Honest Athletics Report Is the One That Says "Insufficient Information"

On August 4, 2026, at the Stade de France, the men's 100m final at the Paris Olympics. The scoreboard flashed 9.79 seconds for two names at once: Noah Lyles and Kishane Thompson. Nobody in the stadium knew who had just won. For nearly thirty seconds while the photo-finish was examined, eight athletes, eight coaching teams and millions of viewers shared a state that sports analysis rarely dares to name: insufficient information to assess.

The result came moments later: Lyles 9.784, Thompson 9.789. Five thousandths of a second, roughly the gap between two chests over 100 metres. But what I kept from that evening sits elsewhere. During those thirty seconds, no commentator dared to declare a winner. And I believe that was the most honest moment of my profession.

I remember the opposite feeling very clearly. In early 2026, still a first-year student, I sat down after the World Athletics Championships in London to rewatch Usain Bolt's final 100m. Justin Gatlin finished in 9.92 seconds with a 0.138-second reaction time; Bolt took 0.183 seconds for the same movement. The gap at the start: 0.045 seconds. I built a frame-by-frame video and titled it "Bolt isn't old, he's just a blink slower". It drew about 50,000 views. From then on I believed the smallest technical detail decides the largest outcome.

That belief was right. But it led me to a different mistake: believing that enough frames always produce a conclusion.

Nine rows, four empty cells

Every week I receive data reports from analytics desks. They look polished: form tables, personal-best curves, comparisons with world records, season-ranking cells. The structure is so uniform you can predict which cell sits where.

The first thing I do is read upward, from the cells marked "insufficient information to assess".

A typical case: a file on a young athlete who ran 9.9x seconds at a domestic meet. The table has nine rows. The "this performance" cell reads 9.9x. The "gap to record" cell reads a distance. But the "split data" cell is empty, so we do not know whether he accelerated at 60m or faded at 80m. The "wind reading" cell is empty, so we do not know whether the mark is valid. The "shoe model" cell is empty, so we do not know how much of it came from equipment. The "opponents in the race" cell is empty, so we do not know how much pressure the lane carried.

Nine rows of data, four empty cells, and a conclusion at the bottom. That conclusion rests on 55 percent of the information.

My industry lives in that condition permanently, but only during a transfer window does the noise grow dense enough for people to notice. A player valued at 100 million euros after fewer than 50 top-flight matches. A contract announced with a release clause nobody can confirm. A midfielder rumoured injured before a semi-final. Every cell in the table has a figure, and most of them cannot be verified.

Four data types you cannot skip

In athletics, four categories of data exist without which an entire assessment table becomes meaningless.

The first is split data. An athlete who runs 200m in 19.9 seconds can do it in two opposite ways: blasting the first 100m and holding, or running evenly and breaking away in the final 50m. Those two speed distributions predict two different futures. Without splits, you do not know whom you are reading about.

The second is the wind reading. World Athletics sets a maximum assisting wind of 2.0 metres per second for the 100m, 200m and long jump. Above that threshold, a mark still appears on the results sheet but is not counted as a record, and is flagged as wind-aided. A 2.5 m/s gust down the home straight can hand a 100m sprinter roughly one tenth of a second for free. In an event where the men's world record stands at 9.58 seconds, set by Usain Bolt on August 16, 2026 in Berlin, one tenth of a second is an entire decade of progress.

The third is equipment data. In January 2026, World Athletics capped sole thickness at 40mm, permitted at most one carbon plate, and required a shoe model to have been available at retail for at least four months before an athlete competes in it. The rule arrived because of a fact nobody could deny. On October 12, 2026, Eliud Kipchoge ran a marathon in 1 hour 59 minutes 40 seconds in Vienna, but that was a staged event with a pace car and rotating pacers, so it was never ratified as a world record. Four years later, on October 8, 2026, Kelvin Kiptum ran 2 hours 00 minutes 35 seconds at the Chicago Marathon, and that mark was ratified as the men's world record.

Place the two milestones side by side: in 2026, Paul Tergat set the men's marathon record at 2:04:55 in Berlin. Twenty years later, the record was 2:00:35. Four minutes and twenty seconds erased in two decades, and most of that erasure falls after 2026, when carbon-plated shoes spread through the elite group. I do not conclude that shoes create records. I only say that without deducting the equipment dividend, we are weighing two different eras on the same scale.

The fourth category, and the one I treat most carefully, is training marks. A few times each season I get a message like "this athlete just ran 9.7 in training". No officials, no calibrated wind gauge, no crowd, no racing pressure. Such a figure can be useful to the athlete and coach as an internal sign of progress. Publishing it is something else entirely.

Beyond those four categories sits a variable the tables almost never display: sample size. One fast run does not establish a level, just as one win does not describe a season. In 2026, when European football returned to empty stadiums, I tracked the first 62 Bundesliga matches and compared them with pre-pandemic data. The home win rate fell from 43 percent to 35 percent. Goals from counter-attacks rose 12 percent. The cause is easy to understand: with no crowd noise, away teams no longer carried the psychological burden, and they pushed higher up the pitch. The data gave me a conclusion the naked eye could not, but only because I had 62 matches to compare, not two.

One more variable belongs to competition structure. Qualification for the men's marathon at the Paris 2026 Olympics ran through a window from November 1, 2026 to April 30, 2026, with an entry standard of 2 hours 08 minutes 10 seconds, plus places via the world ranking. When a selection system offers two parallel routes, races get designed specifically to hit the standard: flat courses, low wind, pacemakers. Marks born in those races are real, but they are produced under deliberately optimal conditions, unlike a knockout race at a championship. Placing them in the same ranking column is an operation that deserves a footnote.

Based on my experience following competitions, I have kept the frame-by-frame method since 2026, and it helped me in one very specific situation.

The contrarian angle

Today's sports media system rewards those who assert. Rankings, views and shares flow toward certain statements. An analysis that ends with "insufficient data to conclude" will almost certainly get no engagement.

I think the opposite.

The most valuable report does not lie in its conclusion. It lies in the list of what it declares it cannot conclude. That list tells you where to go back, what to measure, and how many more samples are needed before anyone is allowed to speak loudly.

A beat slow, and I see the match begin at the twelfth frame.

In 2026, I mispronounced Luka Modrić's name during the World Cup semi-final between Croatia and England, and I got it wrong three times in the first half. I spent the following month rewatching footage at 0.25 speed to relearn the names of all 32 national teams. But what I learned most was elsewhere: watching footage in slow motion, I began to read the gaps in defensive lines, something I had previously applied only to running tracks. In 2026 I got Modrić wrong. That was the most honest analysis of my life.

By the same principle, when the 2026 World Cup was held in Qatar, I tracked the Morocco national team and found their defensive line pushed an average of 52 metres from goal, the highest at the tournament. Before the semi-final against France, rumours of a Sofyan Amrabat injury spread across every outlet. I pulled GPS data from open training sessions, compared running speed and active time, and concluded he would start. Two days later, Amrabat started. My principle is that data outranks rumour, but with a condition few people quote: GPS data from open training is still a small sample. If I had been wrong that time, I would have had to admit it publicly.

There is one more point my industry rarely states. Athletics is a sport measured to the hundredth of a second, so the public assumes it is objective. Being measured and being complete are two different things. A table with thirty figures can still be missing exactly the data point that makes the other thirty mean anything.

Count the empty cells before reading the conclusion

When the stadium is empty, I finally hear the numbers rolling over every metre of grass.

An empty cell in a data report is not a slot for stuffing in a guess to fill the table. It is a coordinate. Next time you read an analysis of an athlete, a transfer deal, or an Olympic qualifying slot, try counting the empty cells before you read the concluding line. If the empty cells outnumber the sourced facts, you are reading a hypothesis presented as a result.

And if you want to know what I am waiting for, I am waiting for a report with a single row: not enough to say. Whoever writes that row deserves more trust than ten people writing ten conclusions.