A File Labelled Tennis With No Tennis In It: What One Wrong Label Reveals About a Broken Archive

It was nearly eleven at night in the Manchester flat. The scorebook lay open on the desk, and beside it the pencil — another of the same ones that once counted the 2026 Wimbledon final. A file landed in the inbox, tagged ‘Tennis’. Years of habit made me pick up the pencil first, then open the file.

Inside were 43 information points. Not one serve, not one return, not one ranking, not one court. There was the State Bank of Pakistan, Governor Jameel Ahmed, the IMF and World Bank Annual Meetings, foreign exchange reserves of $21.4 billion, a current account deficit of $874 million, a $3 billion Eurobond at 7.5 and 7.9 percent. Not a letter of tennis — no player, no coach, no tournament, not a single scoreline.

The score sheet stayed blank. And that blankness is the story.

What a label actually promises

Our pipeline runs in two stages. Stage one extracts information points and core viewpoints, then assigns a domain label — tennis, football, cricket, economics. Stage two reads that label and decides which analytical framework applies. A label is therefore not a filing convenience; it is a promise about the kind of attention a file deserves.

Now imagine the promise is broken. Nine tennis dimensions — technical and tactical, data and form, tournament system and schedule, tour landscape, rules and governance, team and player management, risk, media narrative, industry transmission. All nine returned one verdict: insufficient information. None was partially satisfied. The gap is categorical, not partial.

The entity list makes it plain. JPMorgan, Citibank, Deutsche Bank, Barclays, Standard Chartered, MUFG appear as banks hosting an investor event. No stated role in tour ownership, sponsorship or event investment attaches to any of them. Yet in tennis economics a bank’s name usually means a tournament sponsor or a prize-money ecosystem. That bridge is absent here because there is no tennis on the other bank.

How did the label arrive? Not verifiably. The likeliest explanation is a lexical trap — words like ‘buffers’, ‘resilience’, ‘defence’ live in both dictionaries, and an automated classifier counts words instead of reading content. That is inference, not proof. What is certain is that nobody asked the file a single question: does this actually contain tennis?

Our own archive has the same disease

This is where my professional discomfort begins, because I know the disease. Digging through Bangladesh’s tennis record, I have learned repeatedly that the problem is not only lost information — it is that nobody checks what name the information is filed under.

Bangladesh’s Davis Cup debut came in 2026. Since then, 41 ties, 178 rubbers. Sree-Amol Roy’s singles results were scattered across newspaper cuttings, club registers and a few memories. When the tour stopped in March 2026 and Wimbledon was cancelled on 1 April — the first time since 2026 — I had no matches to chart. In those four months I rebuilt all 41 ties and 178 rubbers into an open spreadsheet. That spreadsheet, not memory, is now my source of record.

That work was necessary because no public, dated, verifiable version of the record existed anywhere. The file was not mislabelled; the file had no label at all. When a sport runs for three decades and its primary record lives in a handful of heads, every generation has to rebuild it from scratch. That is not a journalism problem. It is an infrastructure problem.

By contrast, consider something small. On 16 July 2026, Roger Federer beat Marin Cilic 6-4, 6-4, 6-3. From a Rusholme bedroom I charted all 29 games on paper — Cilic winning 18 of 44 second-serve points, Federer converting 3 of 7 break points. I had taught myself to chart a year earlier as a ball boy at the Northern Lawn Tennis Club in Didsbury, so the hand was trained.

That sheet of paper remains my most trustworthy source, because live broadcast forgets and wire copy corrects, but a hand-kept margin knows where the break point is coming before it arrives. A number written by hand never forgets its own label.

Six days, eleven women’s matches, and a 2026 gallery

In July 2026 a family trip put me inside the National Tennis Championship at the Ramna complex for six days. Russia 2026 played on every screen outside the gate. Inside, I charted 41 matches, 11 of them in the women’s draw. In the clubhouse I got five minutes with Khaled Salahuddin, the 2026 inaugural champion, who told me the first final filled a gallery at Ramna — a detail no online archive carried then, and does not carry now.

The resulting 2,600-word feature ran on a Dhaka English daily’s sports desk and paid 3,000 taka, my first cheque. More useful than the money was a habit: from that week I kept a ‘federation watch’ file, dated entries on every board announcement, court closure and calendar change, and I made it a rule to name the decade a problem began in, not merely the problem.

In September 2026 Carlos Alcaraz beat Casper Ruud 6-4, 2-6, 7-6, 6-3 to take the US Open and the world No. 1 ranking at 19; on 23 September, Federer’s last match at the Laver Cup. My piece for a Dhaka outlet was not about their greatness. It asked what Alcaraz’s drop shot and defensive speed mean for a Bangladeshi junior with one hard court, and what Federer’s exit means for fans raised on Indian pay-TV. It drew 18,000 reads.

That experience produced a standing rule: no global story leaves my desk without a ‘what this means for us’ paragraph. Note that this is precisely the check the pipeline lacked. There, the label was trusted to match the content. Here, I do not trust — I demonstrate.

Three analogies I refused to write

Now the hardest part of working with this file. The temptation is strong. ‘Reserves versus three months of imports’ could be dressed as a ‘points-defence buffer’. The ‘25th IMF programme’ could become a ‘coaching-carousel churn rate’. Circular energy debt could be painted as ‘a long-tail off-court liability’.

I did none of it, and that refusal is the most important professional decision in this piece. An analogy that sounds best is the most dangerous kind of error. ‘Three months of imports’ is a specific calculation — which goods, at what price, in which window. A ‘points-defence buffer’ is an entirely different calculation — a 52-week rolling ledger of where points must be defended. Joining them in one sentence makes the sentence elegant and the information false.

And one more thing I did not do: I did not write the name of that fourteen-year-old girl. In the spring of 2026, the teenager I had charted at Ramna in 2026, from a BKSP feeder group, lost her courts. I quietly paid her board registration and never wrote her name. The rule of the trade is that the reporter’s own presence must not leak into the record.

That rule is the whole point. The same discipline that keeps my name out of the record would have caught seven ‘opinion’ points sitting at fact-layer parity in that document. When a writer’s opinion and a verified fact occupy the same tier, analysis stops being analysis and becomes advocacy.

The outside reading is wrong; the inside reading is worse

The easy reading is that the classifier erred, the tag should be fixed, and the matter closed. That admits the problem while missing the disease. The tag was a symptom.

First, there is no provenance. No outlet, no byline, no publication date, no URL. Four fields, all missing. Without them it is not journalism, it is assertion. A tennis report without a date is a rumour; an economics report without a date is worse, because stale numbers produce wrong decisions and the reader never finds out.

Second, seven of the 43 points are explicitly marked as opinion and yet carried at fact-layer parity. When more than twenty percent of a premise set is opinion, the conclusion rests on position, not analysis.

Third, the dates point forward: ‘July–September 2026’, an ‘IMF document dated 7 October 2026’. A record dated in the future cannot be verified, only believed. And a record beyond verification is not an archive — it is an essay.

Fourth, and this is where my interest sits: the easy cricket-shadow explanation does not hold here. Nobody took tennis’s space. The mechanism is that on a Dhaka sports desk, tennis copy is read by whoever is free. There is no tennis specialist in the room, so nobody is there to suspect a file tagged tennis of not being tennis. The classifier did exactly what the desk does — it trusted the label.

And the most counter-intuitive finding of all: the original document is not badly built. Its architecture is honest — an official’s positive claim rebutted with contrary data, closing on policy critique. A well-built file caused the worst failure, because good form conceals the need for verification.

What must change before the next batch

A mandatory keyword gate should run before Stage 2 — player, tournament, ranking, serve, return, break point. If fewer than ten percent of entities match the label, halt the process, correct the label, and route the file to economics. A wrong tag costs seconds of work. A fabricated tennis report costs the credibility of the whole pipeline.

And my own notebook keeps its rule unchanged: every archival detail must carry a date, a score, or a name; if it only evokes, cut it. That rule is why the 2026 Davis Cup semi-final and Zarif Abrar’s 2026 junior title can sit in the same paragraph without either being asked to shout.

The next ingestion batch will face the same test. The Davis Cup squad is now being rebuilt around three under-21 players — their names, their courts, their calendar. The question is no longer whether the label is right. The question is how many files in our archive carry the correct label and are empty inside, and how much longer we will read them as reliable.