When Sports Data Is Empty: The Line Between Analysis and Fabrication

On the night of November 22, 2026, the final whistle blew at Lusail, the scoreboard read 2-1 in favor of Saudi Arabia, and the world called it the biggest shock in World Cup history. I stayed behind alone, reopening 2,100 running actions from Saudi Arabia's three pre-tournament friendlies. No model on earth predicted this result — error is the ordinary tax of the trade. What kept me awake lay elsewhere: Saudi Arabia had rendered their own data meaningless.

In their friendlies, Saudi Arabia sat very deep, barely pressing, with running density more than 25% below their own average. Reading the table, anyone would conclude: a small, slow team with no path past Argentina. At the World Cup, they pushed their line unusually high, catching Argentina offside 10 times in the first half alone. The old table was not technically wrong. It was simply useless, because the opponent had deliberately distorted it. That night I told my four-person analysis team something I still repeat: old data is useless if the opponent deliberately distorts it.

Since then, I separate two kinds of risk. The first is bad data — the kind every analyst learns to cross-check. The second is empty data — the kind almost no one teaches you to face. In this trade, the second is the frightening one.

Every major tournament, empty data shows up more often than we think. A match not yet played has no metrics. A newcomer with no minutes in the domestic league has no sample. A national team that changed coach three weeks before a tournament has a tactical system with no reference value. These gaps are ordinary in football. The problem lies here: when the data table is empty, the pressure to fill it is greater than ever.

During a major tournament, millions await one verdict per match. Newsrooms need copy before kickoff. Bookmakers need numbers to price markets. Fans need a name to believe in. In that churn, an empty table is not treated as an answer. It is treated as the analyst's failure.

I understand that pressure better than most. In 2026, when the pandemic postponed every league until June, I had 90 football-free days and a betting company that still needed a working model. I did not invent a single match. I built a dataset on the rate of age-related decline, drawn from 3,200 players between 2026 and 2026. The result showed that wingers lose an average of 12% of their running distance after age 29. When football returned, that model let me read one deal correctly: Willian, aged 32, could not sustain Premier League intensity. I won that bet, but what I kept was not the money — it was a principle: when there is no data about the present, go find data about the pattern.

After Qatar, I rebuilt the noise-filtering process for my team. We removed from the sample any friendly with running density more than 25% below average, because that signals a team hiding its hand. We label the origin of every metric clearly: data we collected ourselves from video, data from external providers, and data that is only an estimate. Those three labels are never mixed inside one conclusion. A metric of unknown origin, to me, is worse than an acknowledged gap.

That is how I handled empty data for years. Only at Qatar 2026 did I realize empty data has a more dangerous form: empty data filled in unconsciously.

When an analyst receives an empty table and still has to file a report, the natural thing happens — he begins to infer. He remembers this team used to be strong. He remembers that player used to score. He stitches scattered memories into a plausible picture, then presents it in a confident tone. The report reads fluently. It lacks exactly one thing: a foundation.

The danger of empty data lies in the confidence manufactured to fill the gap.

I have seen this inside my own team. Once, before a knockout match, data on the opponent's lineup had not arrived in time. A team member built a full analysis from a match three months earlier, with passing-rate and pressing metrics that sounded thoroughly convincing. The analysis was so internally consistent it was hard to detect. But the lineup from three months earlier was not the lineup that day. We nearly made a wrong call because of a beautiful report.

I set a rule for the team: when data is empty, the correct answer is "not enough data to conclude." It sounds weak. But in an environment where everyone wants a decisive answer, daring to say "I don't know" is a professional capability, not a weakness.

In 2026, when I was an intern at a small tactical-analysis site in Shenzhen, I hand-calculated the xG for France's 12 shots against Argentina in the round of 16. Mbappe generated 1.8 xG from just four runs behind the defensive line. I wrote a piece with my own figures and my boss called it dull. A week later, a betting analyst shared it. The lesson was not that numbers are always right, but that data you verify yourself carries more weight than data you copy. When you cannot verify it, the best move is not to use it.

At Euro 2026, I applied that principle to the opposite situation. Italy faced Austria in the round of 16, and the crowd piled onto Italy. But Austria's PPDA was just 7.8 — meaning very intense pressing — while Italy completed only 21% of their passes into the final third. I recommended Austria +1 and Under 2.5. The match finished 2-1 to Italy after extra time, with Austria holding 48% of the ball against a major side. I won the handicap. What stays with me is not the money but the fact that data let me go against the crowd because it rested on a thick enough sample, not on a feeling.

I must also be honest about the flip side. Precisely because I built a brand on going against the crowd, I understand the temptation to defend a contrarian take just to protect an image. Once you are known for always saying the opposite, you begin to fear saying the same — even when the data says the same. I force myself to keep a public error log, recording every call I got wrong and why. Mistakes made because the data was too thin are forgivable. Mistakes made because I filled a gap with inference are not.

This principle is not limited to football. I also work in esports — reporting for the Chinese market — and there, empty data is even more common. A new balance patch lands, no official match has been played, and yet a verdict on the meta is still required. I learned that when there is no competitive data, the thing worth saying is not which team gets stronger, but a clear statement of the gap: how many matches are needed for the sample to thicken, how many weeks for the meta to settle. Naming the limits of data is part of the analysis, not an evasion.

That is why I never use a single match to conclude anything about a team. Every match is a confession of probability, and a lone confession is not enough to convict anyone. I need a series, a trend, a curve. The ball stops rolling, but the stream of numbers keeps flowing forward — and my job is to read that stream, not to paint it.

Here I want to speak plainly about a point the trade rarely admits. We often treat fan emotion as "noise" — something that pollutes clean data. The crowd falls asleep inside emotion; I stay awake with the table — I once thought that, and I was proud of it. But crowd emotion is data too. It shows where market expectation drifts from reality, and that drift is the opportunity. Ignoring it is blinding yourself to a real variable.

The paradox is this: the more I trust data, the more I must doubt data. A good model is not one that always gives an answer, but one that knows when to stay silent. A good analyst is not one who always has an opinion, but one who can tell the difference between a conclusion based on evidence and a conclusion based on wanting evidence.

Looking to the next round, I am betting on something that does not happen on the pitch. As major tournaments grow denser, as data grows larger while verification time grows shorter, the difference will not be who has more numbers, but who knows which numbers to trust. The most valuable skill for an analyst in the coming years will be the skill of refusal: refusing a sample that is too small, refusing a source of unclear origin, refusing a conclusion that sounds too plausible to re-check.

I do not believe in the hand of fate; I believe in the data curve. But I also know that curve is only trustworthy when it is drawn from real, thick-enough data and checked by a mind ready to say "I don't know." The biggest mistake is not placing a bet, but placing a bet with the crowd. The second mistake, rarely mentioned, is fabricating a table just to look like someone who knows everything.

The assumption that could be wrong in this piece: if tomorrow a model strong enough to predict even shocks like Saudi Arabia beating Argentina appears, then my principle of "when it is empty, say it is empty" will be challenged. Until then, I still choose honesty toward empty data over a beautiful but hollow report.