The Empty Notebook: When a Cricket Feed Comes Back Blank

Colombo, 18 February 2026. Behind the nets sits the scorer's wooden table, one leg propped up on a bundle of old newspapers. On it: a laptop, a small bowl of linseed oil, and a glass of tea with the sugar settled into silt at the bottom. In the flat half-past-seven light, a left-arm spinner is having his wrist strapped — same strap, same spot, third day running. Twenty feet away a trainer is laying out cones. The kettle sits at the corner of the table and nobody has switched it on.

I open the laptop. The tab is already open, already refreshing.

Title: N/A. Source: N/A. One-sentence summary: blank. Information points: empty list. Entities: not extracted. Time sensitivity: not assessed. Source quality: not assessed. Domain label: cricket_world.

I have seen blank scoreboards before. Rain-washed ones, chalked over with fresh overs the next morning. The old board whose flaps somebody had pulled off, which the groundsman tied to a tree instead. Emptiness has its own seasons. But this morning's emptiness was not weather. It was the kind that arrived carrying a full, neatly formatted, eight-dimension report — every box filled, and the letters N/A inside every one of them.

The beat was there long before anyone thought to listen for it. This morning, the machine that listens was simply silent.

---

First, how a modern cricket desk actually works, because this is a story not about a match but about the pipe the match travels down.

No major cricket outlet begins with a person any more. A scraper pulls the raw page. A parser strips out headline, date, teams, players, numbers. An entity-extraction layer decides who is in the piece — which board, which side, which cricketer, which ground. Only then does a writer sit down, and these days a model sits down before the writer does.

That chain ran this morning and returned a structure that was complete: every field present, every field type-safe, every field hollow. Zero information points. No core viewpoint. No entities. No date. And at the top, the label that gives everything away — cricket_world, where the schema demanded Cricket.

It is a small thing, and I liked it, because it is the scorebook page where the bowler's name is spelled two different ways in two different places. Nobody reads the whole book; the man who fills it every week catches it at a glance. When even the label does not match the contract, everything beneath it deserves suspicion.

Which brings me to the centre of this piece, so I will write it plainly: formatting is not evidence. The shape of a report says nothing about the truth inside it. Eight dimensions, risk matrices, star ratings, red rows marked high risk — readers see all of it and conclude the work was done. This morning the work was not done. And the report was so tidy that forgetting this was easy.

You will find no claim about any team, player, board or match in this piece. That is its first truth and I will not hide it. The payload behind it held not one cricket fact. Had I dropped in a score, an average, a ranking, that would not have been journalism. It would have been typing that looks like the truth.

---

Club cricket's scorebook is the thing people take lightly and without which a match has no past. The league table comes from it, the averages come from it, next season's promotion comes from it. For twenty years I have watched the same man fill it, writing left-handed, turning his pen after every over as if turning a key.

Once I asked him what he does on a day when he cannot remember something. He said he never leaves a space empty; he writes 'not seen'. That difference is enormous. 'Not seen' is information. A blank is a lie, because somebody later will read it as a zero and build an average on top of it.

A record that holds no claim inside it is not a record, it is a form. And that is exactly what reached me this morning; it simply hid inside its own neatness.

Where does the pipe break? In cricket, four ways, and I meet three of them constantly. Paywalls. Pages built by JavaScript — every modern scorecard is assembled in the browser, not served as HTML, so a scraper without a browser sees a clean, empty shell. Video-only or image-only pages with no text at all. And the most deceitful: a truncated payload that starts well and dies in transit.

All four share one quality: the system did not lie. It stopped. That is correct behaviour.

What frightens me is the alternative, where the system starts helping — filling absent information with likely information because likely information reads well. Cricket has an old name for this: the umpire's guess. When a review system has no clear evidence, a decision must still be delivered, and that is precisely where subjectivity enters. I have never been able to swallow football's 'clear and obvious error' phrase because it claims clarity while being definitionally foggy; cricket's umpire's call bothers me the same way. In a room where 'clear' is never defined, the claim to clarity belongs to the loudest voice.

This morning there was no voice in the room at all. Zero information points meant no decision could be reached. The correct decision was no decision — and that is what came, provided anyone bothered to read it.

---

Now the thing most likely to do damage and least likely to be taken seriously: format.

The report says Test, ODI, T20 and The Hundred all remain possible. That is not a technical gap. That is the doorway every wrong number walks through.

A Test average of forty and a T20 average of forty are the same number and not the same thing. A strike rate of fifty in Tests tells one story; in the first six overs of a T20 it tells another; in the death overs it is a joke. A bowler's economy can be identical in two formats with opposite meanings — in Tests economy is often beside the point because wickets are bought with runs, in T20 economy is the whole point. Until format is fixed, no metric can be quoted, and any metric quoted anyway is not evidence but decoration.

I have a clean example from my own notebook. In the summer of 2026 there were five Tests in England against India. England won 4–1 and Virat Kohli made 593 runs at 59.30. Those are Test numbers. Lift them into a T20 argument and they turn dishonest instantly, because ball, ground, law and intent are all different. One player, one year, two different games.

An untagged format is not a small omission — it is the doorway every wrong number walks through. My rule is blunt: before any player metric is printed, the format is printed. If it cannot be printed, the metric does not go in.

The second thing that spoke loudest was the zero in the entity column.

Almost any cricket text names somebody — a board, a team, a league, a player, a ground, a coach. Cricket's language is built from names; writing with none is hard work. So when zero entities come out of a cricket article, the likeliest explanation is not that the article contained no names. It is that what reached us was not the body of an article at all: a cookie notice, a paywall skeleton, an empty template, a video page with three surviving lines.

Zero entities is not a cricket story — zero entities is the story of a story that never arrived. And that check is free. It needs no model. It needs a count.

---

Here I have to go back to my notebook, because the most useful lesson in all this is not technical but habitual.

That summer of 2026 the newsroom was empty — the whole sports desk was in Russia and I was handed five Tests alone. Five Tests, an empty newsroom, and a kettle nobody switched on. I abandoned the press-conference circuit and embedded at training grounds: two hours at Trent Bridge watching Jasprit Bumrah's seam position, two minutes with a bowling coach in a Southampton corridor, a 250-word nets notebook filed before lunch.

What travelled furthest that series was not a quote. It was a strapped wrist and twenty extra minutes of throwdowns — and when I put those two details inside the match copy three days later, readers recognised them.

Then July 2026. Empty stadiums, three hundred people inside the bubble. I asked readers to send thirty-second voice notes describing the sound of a full ground. Eleven hundred and eighty arrived. I listened to every one. One thousand one hundred and eighty voice notes is just a person refusing to lose the thread, over and over.

That week taught me something that runs against the entire data business: a thousand small verifiable inputs outweigh one confident-looking summary. This morning I had the inverse — a confident-looking summary with zero inputs behind it.

Which is why, when people ask how reporting has changed in the age of data, my answer is slightly disappointing: the pipelines have multiplied, the proximity has not. What worked in 2026 was not a feed. It was standing at Trent Bridge for two hours.

---

And then the reader, for whom all of this is supposedly written.

At Lord's in July 2026, the Women's World Cup final: England 228 for 7, India 219, nine runs. I talked my editor into filing it as a live blog instead of the usual 900-word match report. Fourteen posts across six hours. 412,000 page views, more than the paper's entire county cricket output that month. And three readers emailed to ask why nobody had told them the tournament existed.

I kept the blog open the way you keep a door open. After that day I stopped treating the match report as the unit of work and started treating the reader's day as the unit — pre-match explainer, in-play notebook, post-match mailbag.

That habit produced the question this piece is built on: what does a fan ask at six in the evening? He asks what happens next, and why. He does not ask what your pipeline's domain label was.

Which is why a hollow payload is dangerous. The reader will never know that the answer to his six o'clock question was assembled out of a blank. He will read, nod, and repeat it in an argument next week. The error travels downstream and is visible to nobody on the way.

---

Now let me walk briefly into the room I usually keep shut — the money room.

One line in the report stung most: source quality not assessed. Which means that if an auction price had come down the same pipe tomorrow, it would be equally useless. Price is the easiest thing to print without grading and the most wrong.

A franchise auction is cricket's transfer window, and its register is football's: a figure sourced to an agent, a number hiding behind the word 'reported', noise that scales with traffic. I stay out of that register because the writer stops being a writer there and becomes a participant in the industry he covers from a slight distance.

A price is a traffic number; sporting strength is settled somewhere else. The most expensive buy does not become the best player because somebody wrote a large cheque. And every transfer is a conversation somebody was too tired to finish, which somebody else then declared complete.

So a pipe that came back with no figures at all was, in one sense, a relief. There is only one thing worse than a blank payload: a payload that looks full, where the number belongs to an agent and the analysis belongs to nobody.

---

There is one more field that is easiest to skip by calling it unassessable — integrity and governance screening.

The report states plainly that no anti-corruption signal was found because there was nothing to screen. I want that sentence kept in view at all times: not screened is not the same as passed. A blank page is not a clean page. It is an unread page.

In cricket that distinction is the whole game. Twenty years ago board politics ran on tea-stall rumour; it still does, except now there is a dashboard in the middle claiming to know everything. A claim and a check are different objects.

---

And finally, the top of the pipe.

Cricket's economic chain is simple: talent at the bottom, national teams and leagues in the middle, broadcast and derivative markets at the top. If the news at the top arrives empty, every floor beneath inherits the emptiness. A dry pipe upstream leaves every tap downstream dry.

One clarification, because this did not come from the article that came back empty. It is the market around me, offered strictly as context: global cricket revenue flows mainly from the Indian market, and the IPL-owner capital network now reaches leagues in South Africa, the United States and the Gulf. That matters because when a number arrives from above, it is worth asking whose pipe it came down.

---

Now the thing every outsider will get wrong.

He will say the machine failed.

The machine did not fail. The machine stopped, which is what a good machine does. The failure was at the gate, where no check stood to say that a payload with empty information points must not travel onward as a report. Nobody was at the gate, so a default object walked in wearing a report's clothes and everyone treated it as a guest.

The second misreading is that the cure is more data. In my experience the cure is something else. What I got in 2026 came from no feed; it came from being in the right place. In 2026 it came from a hotel corridor. After more than three decades in this trade I will say it with confidence: the most reliable data source in cricket journalism is still a person standing in the right place at the right time with a notebook in his pocket.

The third misreading is the costliest: assuming a hollow thing weighs nothing. A hollow thing that arrives looking like a full page weighs the most. The arithmetic is one-sided — saying 'no data' correctly costs one re-run; inventing something costs every decision built on top of it. I will take the first, every time.

---

So three things go into the notebook for tomorrow morning, on the same page that stayed open today.

When the article is pulled again — and it should be pulled again within the same working day — format gets fixed first. Test, ODI, T20 or The Hundred. Until format is fixed, no average, no strike rate, no economy rate gets quoted.

Entities get counted. Fewer than two names means the source was headline-grade and its reference value drops accordingly. No publication date means no claim gets called current. No figure gets printed before source quality is graded.

And any page that looks completely filled gets one question asked of it: what is actually written inside this box?

I will leave one question open, because I do not have the answer and perhaps nobody does. How much of the cricket copy published today rests on a blank payload that was never shown to the reader? Nobody keeps that count. The scorebook has a man who writes 'not seen'. There is no column for a feed that never arrived.

Nets are breaking up. The spinner has taken the strapping off and is turning his wrist slowly, left-handed, checking where today's pain lives. The kettle is still cold.

Most of what happens here is repetition, and that is the story. Today's repetition was that the machine sent nothing and I stopped it from sending. Tomorrow it may send. And on the day it does, my job will be exactly what it was today — counting how many names are inside, how many dates, and how much of it I have seen with my own eyes.