When an Empty Data Sheet Gets Read as a Clean Report

When an Empty Data Sheet Gets Read as a Clean Report

It was 11:40 on a Tuesday morning in a 14th-floor meeting room in Boston with a view of the harbour. On the screen sat a 41-slide deck about a transfer worth eight million dollars. I was in the second chair from the end of the table, next to an analyst who had been awake all night stitching together wage data and medical data from three clubs.

Slide 12 held a three-column table. Column one: player name. Column two: proposed salary. Column three: contract risk. The third column was completely empty, and the cell borders were still green, because the spreadsheet's conditional formatting rule had been written with a single line of logic: empty cell equals pass.

Nobody in the room asked why the column was empty. The meeting closed after 74 minutes. The minutes recorded exactly one sentence: “No material risks identified.”

Three weeks later I learned the cause. The feed from the medical records provider had dropped at 2:47 a.m., but the export still ran normally: eleven header rows, zero data rows. The spreadsheet did not report an error. It reported green.

In the sports industry, the most expensive mistake is not misreading a figure. It is reading the absence of data as safety. Data does not lie, but it needs someone who knows how to listen.

The industry has moved past the argument about whether analytics helps. That argument ended long ago. What remains is a question of power: who controls the data pipeline, who gets to interpret it, and who is accountable when the pipeline goes quiet.

In Major League Soccer, every club now has at least one analyst, a payroll that must comply with a salary budget, and a player list priced by metrics rather than a scout's instinct. In esports, where I report for the US market, the structure is similar but the tempo is far faster: a single game update can wipe out the value of a roster within 48 hours, and organisations must decide on contract renewals using files collected remotely, through screens, with no in-person tryouts at all.

Money in both sectors does not stop when the stands empty. Fans leave the stadium, but the money never sleeps. That is exactly why the quality of a data pipeline becomes a financial variable rather than a technical detail.

I started this work with an Excel spreadsheet, and I still end it with questions.

In 2026, aged 16 and still in high school in Boston, I launched a blog called MLS Moneyball on Medium. My only source was the salary list published by the MLS Players Association. I downloaded the whole file, built a pivot table, and took apart the New England Revolution payroll. The result kept me up until two in the morning: the club was spending 71 per cent of its salary budget on five players, while the league average was 55 per cent. The piece, headlined “New England is betting on the wrong thing”, drew 12,000 reads in a week and was shared by a local reporter.

The first lesson was not the 71 per cent. It was that I had to reopen the source file four times to be certain I had read the right column, the right season, and the right currency unit. A percentage without a source, a publication date and a clear definition is just an opinion dressed up with a percentage sign. From then on, every piece I wrote began by stating the source, the date and the sample scope before the first sentence was drafted.

In July 2026, aged 17, I watched the World Cup quarter-final between France and Uruguay. Based on my experience tracking knockout matches that summer, I could see France pressing at a rhythm no other side had matched. I pulled tracking data from a public analytics account, hand-counted every sequence again, and came away with two figures: France made 27 pressing sequences, above the tournament average of 19, and their transition time from defence to attack was 0.8 seconds faster than Uruguay's. The match finished with goals from Raphaël Varane and Antoine Griezmann.

I wrote “How Deschamps digitised the press” within two hours of the final whistle. It was shared more than 3,000 times. But what I kept was not the number 27. What I kept was the process: I counted by hand first, cross-checked against automated data second, and published only the overlap between the two sources. The difference between an analyst and someone retelling a match is who is willing to spend 40 minutes re-checking what they already saw.

In 2026, when the pandemic halted MLS, I was interning at a Boston sports analytics firm and was assigned to build scenario models for a club. I calculated that twelve matches behind closed doors would cost 14.2 million dollars in ticket revenue and 2.8 million dollars in food and beverage sales. My report proposed cutting academy spending by 20 per cent and delaying the signing of a foreign striker. It was later sent up to the league office as a reference document.

The structure of that report is still the structure I use: quantify the damage, list the available actions with their costs, and state plainly the residual risk the model cannot price. That third section is the most important. It exists to remind a board that a full spreadsheet can still be missing the variables that decide the outcome, such as season-to-season supporter loyalty, or the chance that a sponsor walks away mid-season.

In 2026, I reported that Arsenal were prepared to pay 7.5 million dollars for New England Revolution goalkeeper Matt Turner, with a 15 per cent sell-on clause. The selling club denied it outright. I published anyway, under the headline “Sources close to the deal confirm Turner to Arsenal”. Three days later Arsenal made it official, and the fee was confirmed to the dollar.

That story brought 50,000 reads, but its real value was the three-step process I have applied ever since: check the source, cross-check both sides independently, and timestamp the verification inside the copy. In the article I wrote “verified at 21:10 Eastern”, with a note that the club had not confirmed. Confident language and absolute language are two different things, and only one of them survives a fact-check.

Every example above leads back to the principle the 14th-floor meeting room violated: an empty cell does not carry the same value as a zero. In a database those are usually two different states, and in analytical thinking they must be treated as opposites. Empty means not collected, not verified, not yet assessable. Zero means measured, with a result of nothing. When a formatting rule paints both the same colour, the system produces a false conclusion without generating a single warning.

The trap appears everywhere. An esports scouting database missing an injury-history field shows a blank, and the reader of the file defaults to assuming the player is healthy. A contract compliance tracker missing an age-verification document shows silence, and legal may conclude the file is complete. A player valuation model missing data on contract bonus clauses returns a figure below true value, and nobody catches it because the output looks reasonable.

The danger of missing data is not that it is wrong. It is that it looks exactly like data that is right.

The same failure of reasoning shows up in public, where audiences and reporters read numbers without checking definitions. Possession percentage is the clearest case. A side with 62 per cent of the ball may simply be circulating it sideways in its own half, generating no meaningful entries into the box. When I recounted the matches of one such team over three months, the number of sideways passes in their own half was typically double the number of line-breaking passes. The possession table does not lie about who had the ball. It stays silent about what the ball was worth.

VAR runs on the same mechanism. Technology does not remove controversy from football; it relocates controversy from the pitch to the review room, where the crowd cannot see who is drawing the line, who is choosing the frame, and who is deciding which instant counts as the moment of contact. The grey areas of the law are not narrowed. They are moved into a more closed room.

That is why I keep a template document open on my second screen whenever I write. It has four lines: what data is missing, who holds it, the verification deadline, and which conditions would force the current conclusion to be withdrawn. Modern football is not won on the pitch. It is won in the meeting room. And in the meeting room, the winner is usually the only person willing to say the sentence nobody wants to hear: we do not have enough data.

That sentence costs money. It delays a transfer, slows an announcement, forces the board into another session. In an industry where competitive advantage is measured in decision speed, saying “not yet assessable” reads as incompetence. That is precisely why dashboard vendors design in the opposite direction: every cell filled, every cell coloured, every gap patched with an inferred value. The product looks immaculate. The decision system is poisoned.

Here is the counter-intuitive part: the scarce resource in sports analytics is not data. Data has become cheap. The scarce resource is the ability to write a report with holes in it, and to defend those holes in front of an impatient board. A spreadsheet that is empty in the right places is worth more than one stuffed with unsourced numbers.

There is one more variable analysts routinely drop because it is hard to measure: fan emotion. I once treated it as soft data not worth a column. After 2026 I changed my mind. When twelve matches were played in empty stadiums, average spend on club-branded merchandise fell clearly among younger supporters while it rose among the most loyal. That is data. Emotion can be measured through behaviour and through cash flow, and a model that ignores it will misprice next season's revenue.

Which brings me back to the blank cells on slide 12. The problem that morning was not a lack of tools. That board had better data infrastructure than most clubs in the league. The problem was that the system was designed to always produce an answer, and a system built that way will automatically convert every silence into an assertion. The risk was not the feed dropping at 2:47 a.m. The risk was the colour rule.

Over the next three years I expect a new role to appear at clubs and esports organisations: a head of data integrity with the power to block a board report when the share of empty cells crosses a threshold. That role will not sit inside the analytics department, because the person checking cannot be the person who built the model. It will sit next to legal and internal audit, where people are paid to be suspicious.

The question I leave readers with is the one I ask myself every time I open a new data file: if the most important column in your report were entirely blank, who in the room would notice before the minutes were signed?