When the Match Log Came Back Empty: The Discipline of Filing 'Insufficient Information' as a Verdict

Hook: Fourteen Fields, Fourteen Blanks

11:40 pm, Manchester. A deconstruction template sits open on screen: fourteen fields. Title: N/A. Source: N/A. Article type: 'Unclassified'. Information points: zero. Entities: none. Time sensitivity: not assessed. Source quality: not assessed. Fourteen out of fourteen are blank, and not one of them says 'unknown'.

Every reporter's first instinct in front of that screen is identical: go and find it. Which match, which club, which dispute, which decision. In real research that instinct is the whole job. In an audit it is the first instinct to be suppressed, because pulling a club name out of a log that cannot even record a headline isn't gathering information — it is manufacturing it, and the first casualty of manufactured information is always whoever gets named.

I counted the minutes nobody counted, and the stoppage-time gap appeared. This time the counting surface was not the pitch but the pipeline, and the result was that an empty log is still data. Read properly, it tells you almost as much as a full article would.

Context: What an Audit Contract Actually Says

Football's rulebook carries a principle that holds far beyond the touchline: what has not been measured is not assumed to exist. IFAB Law 7 — the duration of the match — requires allowance for time lost in both halves: substitutions, assessment of injury, removal of injured players, time-wasting, disciplinary sanctions, medical stoppages, VAR checks. That list is not decorative. It is a checklist. A referee who does not count those items separately ends up announcing a single number that he cannot defend afterwards.

The emphasis on calculating lost time in the 2026-24 Laws of the Game, effective 1 July 2026, is the direction of travel that followed Qatar 2026, where average added time rose visibly across the tournament. That change made an old truth public: whoever owns the clock is a party to the result. And the clock belongs to whoever records it.

I have coded sheets going back to 2026. As a statistics undergraduate at the University of Manchester I volunteered on a refereeing audit at the FIFA U-17 World Cup in India and coded all 52 matches for fouls, cards and added time. Referees in India averaged 4.1 minutes of stoppage time; the same year's UEFA U-17 Championship averaged 2.9. Eighteen of the 52 matches showed inconsistent restart management.

A fan-site editor asked me to drop the tables and use initials because readers would see bias. I refused both and published a 2,000-word report with a methodology appendix. It was shared 4,000 times. The lesson I took from that day is the one now doing the work in front of this empty screen: an audit's credibility comes from its procedure, not its conclusion. A report that can publish its own weakness is the only kind whose strength is worth trusting.

Core: The Fingerprint of an Empty Log

Nine dimensions — tactics, club finance, results cycle, league landscape, rules and governance, management, risk, media narrative, industry transmission. Nine out of nine blank. The surface reading is a single sentence: nothing was found. But surface reading is not an auditing instrument.

When I made the blankness itself the subject, three fields turned out to be empty together in a way that is not coincidence: a null title, a null source, and a type of 'Unclassified'. That almost never happens with a real article. Any football report — match report, transfer story, feature, federation statement — names at least one competition in passing. 'Unclassified' does not mean the genre was ambiguous. It means the text that reached the parser had no genre — boilerplate, an error page, or a consent wall.

The distinction matters. 'Genre unknown' is an analytical problem, fixable by reading again. 'Genre-less text' is a fetch failure, fixable only by retrieving again. In the first case the analyst is at fault; in the second, the pipeline.

Four Windows, One Outcome

I listed four candidate failure routes, because one outcome can have four causes and each needs a different repair. First, a paywall or login wall: content exists, the server returned 401 or 403, and the parser got an empty body. Second, a non-Latin-script or PDF source that broke the parser — plenty of Indian football documents are in Devanagari or Bengali, and many federation circulars are scanned PDFs. Third, a live blog or ticker page where 95 per cent of the text is '19:45 — teams emerge' and the actual analysis is buried far below. Fourth, a 302 redirect into a consent page that goes no further.

All four share one property: the loss occurred at the acquisition layer, not the parsing layer. And if the loss is at acquisition, the nine dimensions do not need nine separate repairs — one fix upstream can restore all nine in a single pass. That is the only good news in this file, and I am writing it as a possibility, not a finding.

The Second Split

One trap recurs in my work, and I named it for myself: cause bias. Home teams get more penalties, so the easy conclusion is referee bias. Easy because the split looks clean, and the cleanliness feels like the finding. So I always split a second time.

In the ISL 2026-21 Goa bio-bubble I had to. The whole season was played in one state under one set of conditions. I coded 40 of the 115 matches. Home teams received 0.09 penalties per game, down from 0.21 in 2026-20; away teams received 0.18. The bio-bubble looked neutral until I separated home penalties from away penalties — and I did not stop there. I split again by referee, by penalty-box entries, by score state, and where the effect shrank under the second split, I wrote that it had shrunk, in the same paragraph.

I ran the same procedure on this empty report. The first split says: the article contained no information. The second asks: at which layer? My answer is acquisition, on the strength of the three-field fingerprint. Had the failure been at extraction, title and source would have survived, because they sit in the page header. Their absence places the failure before the header was reached.

The Discipline of n = Zero

My hardest professional habit is writing n beside every percentage. A pattern built on ten penalties is not a pattern; it is ten events somebody bothered to count. One season of stoppage-time data is not a trend; it is a record of one season.

Here n is zero, and at n = zero the statistics are unambiguous: every ratio is undefined, every mean is undefined, and even random variation cannot be computed, because variation needs two observations. This is where the honest analyst and the fast analyst part company. The fast analyst says the data was thin but the signal is clear. The honest analyst says there was no data, so the question of a signal does not arise. The wording differs by a few words; the accountability differs by an entire career.

I applied to myself the rule I apply to referees: do not write the causal sentence until the sample can carry it. In this report the causal sentence would have been 'this club, or this institution, has this problem'. I cannot write it, because its subject is unidentified.

The Pipeline's Law 7

There is a detail here I am proud of, and I am writing it as a record rather than a boast. Every rule leaves a trail; my job is to timestamp it. In refereeing, my standing complaint is that the rule on paper and the rule in force are different objects. Law 12 states the penalty criterion clearly; the real question is how often that criterion was invoked in a season, how often it was waived, and who was sanctioned. Where no enforcement log exists, I write that the log does not exist — not that the rule is ignored, because the second claim needs evidence and the first needs only arithmetic.

In this pipeline, the instruction to write 'insufficient information' behaves exactly like Law 7 on the pitch. It is the clause that authorises an analyst to come back empty-handed and say so. This report exercised that authority nine times. That is not a weakness; it is compliance.

The measurable question that follows is how often that clause is invoked and how often the data arrives on a second attempt. If a re-run returns content, the rule worked and the source was slow. If a re-run also returns nothing, the fault sits in the source, not the rule. The two outcomes need different treatments, so my protocol starts with a re-run rather than with an assignment of blame.

The 54-Transfer Test, and Why It Cannot Apply Here

On 14 August 2026 FIFA suspended the All India Football Federation over third-party interference; the suspension was lifted on 26 August 2026. I wrote a 1,200-word FAQ to make the two-week timeline legible, then audited 54 transfers in the 2026-23 ISL window against the salary cap. Three clubs had registered players under force-majeure clauses without written AIFF approval, and two had misclassified agent fees. The AIFF issued a circular on 12 September 2026 requiring written approval for such clauses.

That audit ran because it had three things: a document, a date, and the clause applicable to that document. This report has none of the three. Which is why work that took twelve days in August 2026 cannot be attempted here even in theory. I make the comparison so the reader can see that the gap between the two is not effort. An audit is not the piece with the most work in it; an audit is the piece with a document under every claim.

The Clock Nobody Puts on Screen

Broadcast graphics do not show stoppage time. They show the score, the minute, and briefly '90+4'. What stays hidden is the question of who decided the fourth minute, which incident it was awarded for, and who benefited.

Move that question onto the pipeline. Everything the article contained is missing from the extraction report, and the graphic reads 'N/A'. Added time on the pitch is an allocation decision with beneficiaries; so is this blankness. Somebody decided the material would not be captured. Usually that somebody is not a person but a default setting, and the outcome is the same either way: what was never recorded does not travel downstream, and what does not travel downstream never becomes anybody's responsibility.

I count restarts rather than look at the scoreboard. That habit found inconsistency in 18 of 52 matches in 2026. The same habit logged 112 second-half cooling-break stoppages in 2026-21, several of which had no clean relationship to a Law. Both times the same thing happened: what nobody counts belongs to nobody's account. In this pipeline, the empty field is playing that role.

What an Empty Report Costs

It is tempting to say that no data means no damage. The damage is specific. Had the template been filled by inference — a club, a fee, a formation, an injury — the first paragraph would have been false and everything after it a continuation of the same falsehood. The larger danger sits in the governance box: a club name inserted there becomes the sentence 'club X faces a compliance question'. In my trade that is the most expensive sentence available. I have deliberately kept Manchester City's 115 charges, the Everton and Nottingham Forest points deductions and the Juventus case out of this report, because there is no allegation here, no accounting period, no charge sheet. Attaching precedents to nothing is not analysis; it is accusation.

The third cost is the least visible: time. A report that returns empty and is never re-ordered makes its subject invisible for a season. In rules reporting that bites hardest, because circulars and approvals go stale in weeks. The AIFF circular of 12 September 2026 still exists; its relevance is tied to the window that produced it.

Contrarian: An Empty Report Is Not the Failure

The industry's biggest mistake is not the empty report. It is the hurry to fill it. Nine blank dimensions are an unpleasant outcome; three of them filled from 'sources' is permanent damage. The first can be re-ordered. The second cannot, because it has been published.

So this report is not a retreat. It is the same work I did in 2026 when an editor asked me to remove the tables. Removing them was easy, and at the time it would even have been called reader-friendly. I did not remove them, because the tables were the claim a reader could check. Forced to choose between verifiability and comfort, I choose verifiability — politely, and without a scene. I do not tell an editor he is a bad man. I send a three-row table.

The second inversion is that a system which can return an empty report earns more trust, not less. A system that always returns something is permanently suspect, because we never learn where it was guessing. A system that can say 'insufficient information' wins belief for every positive claim it makes. The capacity to come back empty is the hardest qualification an audit can hold.

I see the same thing on the pitch. A referee who decides every doubtful collision must produce a decision gives one in the fifth minute that television later disproves. A referee who can say 'nothing there' and keep his hands down protects his credibility for the next seventy. Refereeing is an audit that never blinks — but a good referee also knows that whistling every challenge is not judgement. It is a habit.

Takeaway: What Gets Measured Next

I will close not with a result but with a measurable question. Three things next round: did the title return, did the source return, and did the type move past 'Unclassified'. Three yeses and this conversation ends and the work starts. Three nos and the verdict stands, however unsatisfying it reads.

I have kept one rule since 2026: publish only the number I counted myself. 52 matches, 4.1 minutes, 18 inconsistencies, 40 coded matches, 0.09 penalties, 54 transfers, three force-majeure registrations — every figure has a sheet behind it and every sheet has a date. Today the sheet is empty. An empty sheet yields exactly one honest sentence, and this report has now written it nine times. Until the sheet fills, I have nothing further to say — and anyone who says more should be asked for their source, and its date.