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

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

At 3:12 a.m. in Tokyo, I reopened a tracking table for an athletics meet and found exactly one line: “N/A.” No metrics, no athlete names, no distance, no time, not even wind conditions. Only empty space. To someone who earns a living reading numbers, that emptiness is more frightening than any wrong number, because a wrong number can still be traced, while an empty space always invites people to fill it with guesswork. That night I sat a long time in front of the screen and asked myself: what happens to a sport when the data disappears and no one will admit it has disappeared?

That question returns to me whenever an analytics table is posted online with a full set of metrics but no sources. I have seen reports that look highly professional, with xG, PPDA, distance covered and top speed presented as though they were measured straight from the pitch. When I check them back, most of those figures have no clear origin. They were created to fill a gap, not to reflect reality. In a sport where one hundredth of a second decides a medal, filling a gap with fabrication is a quiet form of cheating. And once the gap is filled, no one remembers it was ever there.

The summer of Russia 2026 taught me the opposite. I was twenty, a second-year student in Tokyo, writing a World Cup blog built on data. Before the group-stage match between Germany and South Korea, I pointed out that Germany’s xG was 2.1 against South Korea’s 0.6, but South Korea recorded 121 sprint efforts and a PPDA of 7.8 in the second half. I predicted Germany could be eliminated. A male commentator online snapped at me: “What does a girl know about football to talk about pressing?” South Korea won 2-0 through goals from Kim Young-gwon and Son Heung-min, Germany went home, and my blog was shared thousands of times overnight. The lesson was not that data is always right, but that data is only right when people read it all the way to the end. When the data speaks, laughter becomes nothing but noise.

But data does not always speak. In sports analytics, people like to talk about data as an endless resource; few talk about its other side: silence. An athletics meet can feature hundreds of athletes and thousands of timing measurements, but if the recording system fails on the decisive lane, the whole picture behind it becomes distorted. The analyst faces two choices: state plainly that there is not enough data, or fill the gap with a story that sounds plausible. The second choice is always more attractive, and always more dangerous.

To understand why a gap is dangerous, look at how sports data is created. At a modern athletics meet, each lane has high-speed cameras, sensors at the start line, a photo-finish system, and a wind gauge beside the track. Data flows through several layers: the recording devices, the aggregation software, then the display for spectators. If one layer fails, the final layer still appears as normal, but a link has been lost. Viewers do not see the missing link; they only see a number. That is why I always cross-check at least two sources before citing any metric.

Take a concrete example from athletics. A mark is only ratified as a record when several conditions are met: the time is measured by a fully automatic timing system accurate to 0.01 seconds; wind speed does not exceed +2.0 metres per second; and the athlete passes a doping test. Miss one of the three, and the number still appears on the board, but it is no longer a record. That means the same string of digits depends entirely on the frame of measurement around it. A poor analyst reads only the digits. A decent analyst reads the frame as well.

In football the story is the same. PPDA, the number of opponent passes allowed before each defensive action, never appears on the scoreboard. It does not shoot, does not score, is not chanted by the crowd. But at Euro 2026, Italy reached the final with an average PPDA of 8.9, the most aggressive pressing in the tournament, while England’s was 11.4. When I presented that in a Tokyo meeting room, a male colleague laughed: “Japanese women only look at numbers, they don’t understand the psychology of Wembley.” I projected a chart of the last thirty matches and said: “The numbers do not lie. You will lose if you keep dropping deep.” The final on 11 July 2026 ended 1-1 after extra time, Italy won 3-2 on penalties, with Leonardo Bonucci scoring the equaliser. PPDA does not shoot, but it carried the Italians to the night they lifted the trophy.

There was a period when I learned the value of re-checking data sources the hard way. In 2026, when the pandemic halted global football, the Bundesliga returned in May with empty stadiums. I collected data from the first twenty-six matches and found home advantage fell from an average of 0.44 goals per match to 0.15. I built a “no-crowd” model, bet on under-priced away teams, and won 17 of 20 wagers that month. Home advantage is a hypothesis; COVID was an involuntary experiment. But I also learned that the 0.44 figure from the past cannot be applied directly to the present, because the context had changed. Every article of mine since then devotes a section to assumptions and the limits of the model.

The transfer market is where gaps are exploited most. An unsourced rumour can push a player’s valuation up by several million euros within hours. The transfer market has no rumours, only prices trying to find themselves again. Following deals in the J.League and Europe, I learned that the final figure rarely matches the number leaked at the start. The distance between those two numbers is the part that was fabricated, and someone always pays for it.

I once followed a major meet where the photo-finish result was disputed. Two athletes finished within the same hundredth of a second, and organisers needed almost ten minutes to determine the placings from high-speed images. During those ten minutes, social media already had every version of the result. None of them had the data. They had conclusions. The distance between a conclusion and the data is where information gets distorted, and it often lasts only a few minutes, yet it is enough for thousands of people to believe something false.

Back to that empty table. My point is not that data is useless, but that emptiness is itself a kind of data. When a metric does not exist, the right question is not “what can we guess?” but “why does it not exist?” The recording system may have failed. The athlete may have withdrawn mid-race. The event may not have taken place yet. Each possibility leads to a different conclusion, and telling them apart matters more than any interpolation. In data science this is called null handling: instead of inventing an average to fill the gap, you write plainly that there is insufficient information to assess. That is a humble rule, and also an ethical one.

The counter-intuitive angle sits here: most fans believe an analysis with more numbers is more trustworthy. The reality is usually the opposite. An analysis stuffed with ten unsourced metrics is more suspect than one with three metrics that clearly state their origin, timing and conditions. Quantity does not create credibility; traceability does. I have seen player rankings spread widely, presented as neatly as a financial report, yet when asked for their source, no one could answer. That is correlation drawn, not correlation measured. And correlation drawn is never causation.

There is another temptation I must guard against: dismissing emotion because I assume it is only noise. When an athlete fails in the final minute, people usually attribute it to psychology. But psychology is also a layer of data, just harder to measure. The silence of the stands, the gasp after the starting gun, the eyes lifted to the scoreboard, all are signals. A good analyst does not discard them but places them beside the hard numbers to see whether they align. In the meeting room, emotion asks, data answers. But if the data has no answer, the most honest thing is to say we do not yet know.

What I carry from that night staring at an empty table is a professional principle: better to leave a gap visible than to fill it with a number that sounds plausible. In a major-tournament season, when fans’ emotions are compressed and expectations run high, the pressure for a quick answer grows. Yet that is exactly when a data person must slow down. Every jeer is an unlabelled data column. And every gap, if acknowledged properly, can become the starting point for a better question.

I am not against using data to tell a story. I am against using a story to replace data. A good analysis must withstand the question: where did this number come from, when was it measured, under what conditions, and who verified it. If it cannot answer, that number is just a sentence written in digits. Fans deserve real data, because they invest money, time and trust in it. That is an ethical line, not a technical one.

I do not predict football; I measure the distance between expectation and goals. But I am also learning to measure the distance between what I know and what I think I know. The next round will again be full of numbers, and a few cells will be empty. When the data table is empty, I will wait for data, rather than fill it with a story that sounds good.