Where There Is No Data, Confidence Is Football Analysis's Biggest Risk

Nine-thirty on a Monday morning. In an editorial room in London I open a spreadsheet with nine tabs: tactical, financial, results, league landscape, governance, dressing room, risk, narrative, industry transmission. Every tab carries the same entry — “insufficient information.” I watched the match last night, yet I do not know the club’s formation, its xG, its wage bill, or how long its coach’s contract runs. On the phone, the editor says: “Then write it from your opinion.”

This is precisely where football journalism and football analysis part company. My fear is not the empty cell. My fear is that sentence — “write it from your opinion.” Because when you pour opinion into an empty cell, what comes out is not analysis. It is a posture of confidence.

Football absorbed data over the past decade the way banking absorbed risk models: slowly, then suddenly, then everywhere. When I left my staff job in 2026 to run my own newsletter, a Wyscout subscription was my single largest investment. That August, after Liverpool beat Arsenal 4-0 at Anfield, I cut fourteen clips from Liverpool’s twenty-three high turnovers and wrote a piece arguing Arsenal’s problem was not the back three but Wenger’s fear. It drew 50,000 reads in forty-eight hours. The rule I took from it: clips first, opinion second.

That rule has since been inverted. A nine-dimension analytical framework now circulates in newsrooms, boardrooms and television studios — I use one myself. Formation to wage structure, governance risk to dressing-room chemistry. The framework is not the problem. The problem is the social pressure to fill every cell, which is so strong that we refuse to label an empty cell empty and dress it up as “sparse information.”

Two beliefs dominate. One: more data means better decisions.

Two, far more damaging: an analyst who says “I don’t know” is useless.

Any conclusion drawn from an empty set is not an inference; it is a fabrication. That is arithmetic, not ethics. Football breaks the rule daily because the punishment is never immediate.

What does xG actually measure? The quality of chances created in a match. It does not measure why a coach made his change on sixty-seven minutes. It does not measure why a right-back retreated two yards on a recovery run. It does not measure where a referee sets his card threshold. Yet boardrooms now settle a coach’s future on xG. The number is correct. But it answers a different question, and we have quietly reassigned it to this one. That is data debt — borrowed, never repaid.

PPDA tells the same story. It is the standard proxy for pressing intensity. If a side’s PPDA rises by two points across the first three fixtures, a narrative appears: they have abandoned the press. Nobody asks what that side’s baseline was last season, what the opponent’s build-up pattern looked like, what the game state was. Three matches against a zero baseline is a chasm. Narratives find chasms uncomfortable, so they are papered over.

The January window is where empty data becomes most expensive: the panic premium. Say a club pays £40m on a five-year deal. That is £8m of amortisation a year, plus wages, agent fees and a sell-on clause. If the buying club had no comparable valuation benchmark, it has lost £10-15m before the player kicks a ball. A panic premium is not an emotion. It is the money someone pays because they never wrote down their own list of unknowns.

In August 2026 Bayern beat Barcelona 8-2. Most wrote that this was Bayern’s peak. My ledger read differently: Bayern’s 5.2 xG against Barcelona’s 0.9 — the scoreline was not Bayern’s rise but a one-off settlement of Barcelona’s decade of accumulated data debt. In that same window Chelsea spent £200m: Kai Havertz £71m, Timo Werner £47.5m, Hakim Ziyech £33m. Many called it panic. I called it pandemic arbitrage. Both pieces stood on the same rule: measure the gap between price and value, then argue.

And yet there is one place where I distrust my own method. At the 2026 Qatar final, Argentina beat France on penalties. Everyone wrote about Kylian Mbappe’s hat-trick. My ledger pointed elsewhere — seven matches in twenty-eight days, Argentina’s average sprint distance down 11% in extra time. But I only knew that number mid-tournament; I never had the squad’s internal fitness data. I was drawing an interior picture from an exterior shadow. That column drew 1.2 million reads. Nobody wants to admit incomplete inputs under that much attention. I am admitting it now.

The real skill in football is not reading data. It is marking the place where data is absent. If a scouting report arrived without a player’s name, nobody would accept it as “sparse” — they would send it back. Why do we hold match analysis to a lower standard?

Now against myself. Saying “I don’t know” is easy, and the easiest sentences hide the biggest trap: cowardice. A club still has to buy players on partial information. An editor still has to decide on partial information. If I write “zero data, no comment” every week, nobody needs me — I become a cold, flawless, entirely useless machine.

The distinction matters: admitting uncertainty is not the same as avoiding accountability. The difference is disclosure. “This judgement rests on a fragile base, confidence is moderate, and it is falsified if the opponent overloads the right flank” is analysis. “It’s all complicated, nothing can be said” is evasion.

The market metaphor has a limit too. In equity markets a mispriced asset eventually punishes somebody; in football a shot hits the post and rolls out, the table changes, and no model captures it. The metaphor stops there.

So here is a testable claim. Across the next two January windows, the clubs that publish in writing what they do not know — which positions lack depth, which fees have no comparable — will pay a lower average panic premium. If the opposite happens, hold me to it; it goes in the ledger in red ink.

The mailbox is open for counter-evidence. One condition: bring data. Not empty cells.