Empty Input, Zero Analysis: Where There Is No Data, There Is No Analysis

November 2026. After a piece ran on a niche analytics site, two club analysts and a scout emailed me. In my hands was a log of 96 matches — four Project Restart rounds, the four rounds before lockdown, and every soft-tissue injury, logged by minute, measured per thousand minutes played. For six weeks I had kept it hidden, convinced something so obvious would not be read. But one line I never crossed — inventing a story when there was no data.

Today that line is being tested. In front of me sits an analytical framework — nine dimensions, more than twenty cells, and in almost every cell the same sentence: insufficient information. No team. No player. No fee, no form, no points table. The very structure asked to analyse is telling me the raw material is zero.

It is worth being clear about what this document is. It is the second stage of a two-stage pipeline. Stage one breaks an article apart: title, source, information points, entities, viewpoints. Stage two lays deep analysis over those fragments — tactics, finance, results, league landscape, rules, dressing room, risk, media narrative, industry transmission.

But here stage one came back empty. No title, no source, not one information point, not one entity. Every dimension of stage two therefore stands on empty ground. What the document did is not an ordinary act — it refused to analyse. In every table, every row, it placed a single answer: I do not know.

That refusal is the real story here. A structure that can look at an empty input and say "I do not know" is a structure worth trusting. The others fill the empty cells with story — insert a name, insert a score, insert a "reportedly", and the reader never learns which part is fact and which is guess. Readers do not want to read an empty cell, so the filled cell sells better — whether or not it holds information.

I decode injuries by following the load, the tissue and the lie. The first requirement of this work is primary material. A club's injury bulletin is not data to me; it is a claim. The club will say "minor knock". But a minor knock that stretches six weeks stops being minor. Then I have to go to scan reports, surgeon interviews, frame-by-frame footage. In autumn 2026 I opened the Cazorla file expecting a foot — out came eight operations, twenty months, 8cm of Achilles tendon lost to post-surgical infection, and a skin graft harvested from his forearm. Without the mechanism, that story would have stayed nothing but a crying story.

Football media's oldest habit is filling the empty cell. In a transfer window that habit becomes an epidemic. One "reportedly" births ten headlines. No source, no date, just a verb and a club's name. When an analytical framework declines that habit, it has not failed — it has succeeded. The question is whether we can recognise success.

Look at my injury ledger. By December 2026, four hundred rows. Every row carries a mechanism, a minute, a return date. The value of that ledger is something else — it is my private baseline. When a club says "three weeks", I check my own rows for how long the same mechanism has taken before. Where it does not match, it does not match — that is a red flag. Without a source I do not print it; I only hold the question. In that St Petersburg mixed zone, my question about Mohamed Salah's shoulder was about the grade of the acromioclavicular joint, because the scan shape after a challenge five weeks earlier is what decides how fast anyone returns.

By the same logic, on 9 June 2026, Nabil Fekir's move to Liverpool collapsed on a medical over his knee — and he still played the final. Fee and fitness are not the same thing; that is the transfer market's largest invisible accounting. And on deadline day in 2026, Marc Guéhi's £35m move collapsed on a medical in London. Nobody asked which joint failed. Nobody asked which tissue. That is the journalistic gap, and a gap filled long enough becomes an invented truth.

Now imagine someone forcing this empty framework to be filled. First a team. Then a transfer. Then a fee. Then an injury. Every inserted fact becomes the next stage's input. Three stages later that invented player's name may surface in a headline — "according to a source". That is how one zero input becomes several invented truths. Misinformation does not spread in one step; it spreads step by step, becoming a little more credible at each one.

This is why null handling is not a courtesy but a safety ring. An empty input entering the pipeline must be stopped hard — not gently papered over with a template. The gentle path is more comfortable, and precisely for that reason it is the most dangerous.

And here I admit a weakness of my own. My faith in data is strong enough that the human can disappear inside it. Pain, fear, the boredom of six months of rehab, family pressure — none of these show up in a spreadsheet. In November 2026 in Qatar, when Sadio Mané was ruled out with a fibula injury — hurt nine days earlier in Bayern's win over Werder Bremen — Senegal still reached the last 16, but the man in that dressing room is not a number. Framework and human, both are needed. Without one, the other is incomplete.

The calendar is not neutral to me either. On 5 July 2026, in a Club World Cup quarter-final, Jamal Musiala fractured his left fibula and dislocated his ankle. To trace that injury's roots you have to travel through a season in which there was no such thing as an off-season. Injury here is not an accident; it is an invoice — one that simply arrives late.

This is where I part ways with most editors. In November 2026 I argued the real story was structural — five substitutions made permanent, a calendar with no reset, and hamstring data that would spike by February. Two editors told me to "stick to the football". Their case is realistic: readers want the club's name, not the calendar's arithmetic.

I did not comply. Because I had already made one decision — I will not write a story without data, and I will not abandon data without a story. That is the hard place. Filling an empty cell is easy, and the reader is pleased. A framework that says "insufficient information" looks like a failure. But it is not failure; it is honesty.

Confusing a filled template with a complete analysis is the biggest trap of all. Nine dimensions and nine cells do not make an analysis complete; every cell must carry tissue-level evidence. Otherwise it is a form in the costume of analysis, and a form never makes a decision.

I decode injuries by following the load, the tissue and the lie. Where not one of the three exists, there is no path for me to walk. To hunt the lie you need at least a claim; to measure load you need at least a minutes column; to read tissue you need at least a mechanism. An empty page holds none of them.

So what is needed now is not a new story. What is needed is a gate — a check at the pipeline's start that raises a red flag the moment it sees zero information points and zero entities. What the document did is right, but it is not the last line of defence. The last line is that this input returns to stage one — the original article is parsed again, the fetch and parse logs are read, and we learn where the information was lost.

My own rule is simple. A scan is a sentence; rehab is a revision of that sentence. Where there is no sentence, there is nothing to revise — only a blank page. And the most honest answer a blank page can give is to be left blank. So next time the question is not "what do I write on this page". The question is "why is this page blank". Whoever finds that answer will have written the real first sentence.