Silent Data Failure: When an Empty Table Still Looks Like a Clean Result

In March 2026, I received a data table for a Ligue 1 match from an aggregator. The table was fully formatted: 12 pitch zones, pressing frequency by half, average distance between lines. Not a single cell flagged an error. Every column carried a value. I opened highlight clips I had cut myself from two different video sources to cross-check, and at minute 63 the distance between the two defensive lines came out as 0 metres.

No team plays with a line distance of zero. A back line standing that tightly would be torn apart in three passes. That zero did not describe anything on the pitch; it described a gap in the data pipeline. What made me stop longer was its shape: an empty table presented exactly as if it were a finished result.

My tactical map was drawn on a France–Argentina night, where two shirt colours dissolved into a single intent. That night I learned that the thing worth reading on a pitch is the space, the thing that gets left behind. Years later I realised the same principle applies to data: a gap in a table is also a signal, as long as the reader does not fill it in with guesswork.

Football analysis in 2026 does not lack data. A single Ligue 1 match generates thousands of positional data points and hundreds of derived metrics, distributed faster than the match itself takes to finish. Most of those metrics come out of a processing chain the end reader never sees.

I began writing a weekly tactical journal in 2026, after France–Argentina in Kazan on 30 June. France held 39% possession and won 4-3; Mbappé scored twice from the space behind the South American back line. I was 17, sitting with a notebook, logging the position of every France player whenever his team did not have the ball. I did not record goals. I recorded standing positions.

In March 2026, European football stopped for four months. At 19, I rewatched 57 PSG matches from the 2026-20 season, built my own dataset splitting the pitch into 12 zones, and counted Verratti's pressing frequency match by match. For four frozen months I sat with PSG 57 times to hear them speak through the gaps. Before publishing my analysis of that team's pressing trap, I cross-checked every metric against two video sources and revised it three times after finding errors when measuring line distances.

That discipline came out of a very specific fear, and it leads directly to the table I received in March 2026.

Verifying football data has three layers, and each of them can fail silently.

The first layer is collection. Tracking cameras lose signal, a frame is dropped, an API returns an empty result. Failures at this layer make no sound. They return a valid structure with empty fields, and in many systems a valid structure with empty fields looks exactly like a match in which nothing happened.

The second layer is aggregation, where many metrics get distorted and nobody checks. A pressing rate computed over four minutes of live ball and over ninety minutes of live ball produces two completely different answers, yet both are printed under the same name. The denominator disappears from the headline. The reader gets a percentage, not a sample size.

The third layer is interpretation, and that is where I work. When Morocco reached the 2026 World Cup semi-finals with a single goal conceded — an own goal against Canada — the media called it negative defending. I measured the average distance between their lines, under the command of centre-back Saïss, and got 28 metres. Morocco built a wall, and I was the one writing a diary for each brick. That moving wall only means something because it is anchored to a published method: which lines were split, measured from which player's feet, across which minutes.

Remove the method and 28 metres becomes a decorative figure. The value of a football metric lies not in the number itself but in the definition and the denominator behind it. A metric with no definition can be neither wrong nor right — it simply sits there, waiting for someone to assign it a meaning.

Across those 57 PSG matches, the only thing they never rewatched was their own fear. I think the same holds for data tables: the most overlooked part is always the part that never appears.

The football analysis industry spends a great deal of attention on the risk of machines inventing numbers. That risk is real. But it is loud, and because it is loud it is easy to catch. A model that fabricates a metric produces absurd values, spikes that do not match the video. Anyone with the trade will notice.

The real blind spot sits on the opposite side: data that disappears without anyone announcing it. An empty field makes no noise, creates no spike, breaks no chart. It passes through in the exact shape of a clean result, and the entire burden of detection is pushed onto the end reader.

I once saw a match report in which a team's duel success count was zero in the second half. Nobody checked. A professional side cannot fail to win a single duel across 45 minutes. The metric was empty, but it sat in a row with a heading, and headings look trustworthy.

There is a structural problem in the industry here, and it bothers me more than the technical fault. Sources are ranked by name recognition, not by process. A large platform is presumed reliable, a small account presumed unreliable. Yet what determines whether a table can be trusted is whether it states its definition, its denominator and its extraction time.

Transfers are where people buy players, while coaching staffs buy time. In data analysis, the thing most heavily traded is the credibility of a name, and it is far cheaper than building a proper verification process.

I also have to state my own limits clearly. If a table carries a full definition, a denominator and an extraction time, and I still reject it, then the problem is me, not the data. That is the threshold I set so that caution does not harden into conservatism.

In March 2026, I sent that table back to its source and asked for the denominator. They did not have one. The match was still played, someone still scored, and my article on it was still written — from video, from a notebook, and from figures I counted myself.

Readers can carry one very cheap habit. Whenever a football metric appears, ask three things. How is it defined. How many minutes of live ball does it cover. And who checked it a second time.

The third is the hardest, because in modern football the person checking a second time is usually not paid to do it.