Empty Fields, Full Format: An Audit of the Cricket Coverage Failure That Passes for Success

Twelve fields. In all twelve, a single answer — N/A.

Title N/A. Source N/A. One-sentence summary blank. Information points an empty list. Entities: nothing recorded. Time-sensitivity: unassessed. One field was populated: the domain tag, cricket_asia.

That record did not stop at an error page. It travelled. The next stage accepted it as valid input, because the structure wasn't broken — the JSON parsed, the fields existed, they simply contained nothing. The system didn't distinguish 'empty' from 'correct', because nobody had asked it to.

I start with the ledger, not the legend. And in twenty-five years of notes I have collected a fair number of these records — press releases with every box filled and nothing inside; 'sources say' lines with no source; retention lists that are lists of names and not a cap sheet. Which is why this empty sheet didn't startle me. It did something else: it opened the question that sits under cricket's entire content supply chain. How does a null ID pass the whole system?

A business where the news is a unit

Cricket coverage is produced today by three separate economic demands, and each wants something different.

The first is broadcast rights. When the IPL media rights e-auction concluded on 14 June 2026, the 2026-27 cycle sold for ₹48,390 crore — the digital package to Viacom18 at ₹23,758 crore, the TV package to Disney Star at ₹23,575 crore, and the rest-of-world rights to Times Internet at ₹1,057 crore. A business of that size does not leave the news free; it treats the news as a content unit whose supply must not stop.

The second is the franchises and the teams. How many 'stories' a franchise needs in a year is not decided by its cricket; it is decided by its sponsor contracts. The social team, the video team, the content agency — all three are measured on one number: items published. Quality is not a column in that report.

The third is the aggregators and the wires, whose product is pageviews. And pageview arithmetic is brutal: an empty page earns a view, and so does a painstaking analysis. To the pipeline the two are identical, until somebody senior says otherwise.

The night of 20 February 2026 still reads to me like a template. In a Mumbai hotel ballroom, two of roughly two hundred journalists in the media room were women, and the room's job was to collect quotes. Mine was to keep the cap arithmetic — the $5 million franchise ceiling, Dhoni's $1.5 million to Chennai, Symonds' $1.35 million to Deccan, and the icon-player deductions nobody had explained on the record. I was told I would be 'better at player profiles'. I filed 1,400 words of cap math. It was the most-read sports piece we published that week.

Since that night I have had exactly one method: number first, story after. In a supply chain where the number comes first, an empty ID cannot pass — because the count does not balance when the slot is left blank.

Zero needs a schema too

There was a rule on the old newsroom desk: the copy editor's job is not what has been written, it is what has been left out. When I later ran a transfers desk the rule doubled — in transfers the biggest story is often the one that didn't happen. The deal that never reached the file explains the most.

In a schema where an empty field and a filled one are typed identically, audit stops meaning anything. Here the zero wore the costume of a filled field, and the whole system kept reading the costume.

Three questions attach to every tip I take, whether it arrives from an agent or a scraper: where is the document, what is the number, when is the deadline. This record failed all three. No document, because no snapshot of the source page was retained. No number, because the information-point count was zero. No deadline, because there was no publication schedule to attach it to.

The empty-field format and the 'sources say' line are two shapes of one disease. Both wear a wrapper that feels institutional rather than authorial — a format, an attribution tone — and both are hollow when you open them. In 2026, during the spot-fixing investigation, I did the opposite. I ran a piece with eleven documents attached — agent commission receipts, third-party payments, undisclosed side letters between parties. Not one anonymous quote. A story built on anonymous quotes leaves the reader no way of checking whether the error was the reporter's or somebody else's.

Zero needs a schema of its own. Until the system decides that 'zero information points' is a FAILURE and not a valid record, every empty object will keep moving downstream. One simple condition kills the problem — if the information-point count is zero, or the body text falls under a set length, mark the stage EXTRACTION_FAILED and stop. Retain the raw HTML snapshot so it can later be established whether the page was empty or the parser was blind.

Now look at the one field that was populated. cricket_asia was the only piece of information in the whole record, and it is suspect in itself. A domain tag sitting inside an empty record may well be a default value — evidence not that this is cricket news, but that the system needed a label and chose the safest one available. Transfer reporting has a close cousin: attaching a player to 'interest from a big club' and publishing it. The information is replaced by a tag. Big club, bigger audience. Label first, fact later.

In cricket the same label failure runs in another form. A team with a large market behind it does not need its news hunted down — the news arrives by itself, and is manufactured even in its absence. A team without that market needs a match preview written the hard way. The result is that coverage volume is set not by the quality of the cricket but by the supply of press releases. A system that does not select news on merit will not select out empty news either. It is one weakness, appearing once in editorial judgement and once in code.

And here the lesson from my own beat attaches directly. When a franchise announces its retention list, television presents it as a victory. The list only means something with the cap sheet beside it — who was released, who was held through a right-to-match, how many crore of space was left open, and what that space could buy at the next auction. Issuing a list of names and calling it the game is precisely what this extraction stage did. A list, filled out of zero.

The fear is not the thing that lies

Now to the objection that arrives first in every argument: the machine lies. The real danger, the claim goes, is hallucination — the model will invent a score, invent a citation, and the reader will never know.

My experience is the reverse. Hallucination makes noise. It gets seen, caught, screenshotted, and corrected by the next day. What does not get seen is the clean empty output. There is no wrong fact in it for you to catch. There is no false name that requires a player's lawyers to be called. There is an N/A sitting in a cabinet that has passed every validation.

A story that contains a lie gives its reader something to take revenge on. A story that contains nothing gives its reader no suspicion at all — he turns the page and blames himself for not reading it properly.

A second objection follows: if the fault was in the first stage, complaining at the third is a letter to the wrong address. That argument is small. The first stage's error is never as expensive as the third stage's silence. A parser that returns empty text could not have done better — it reported that it found nothing. The real fault belongs to the controller that gave an empty response the same standing as a filled one.

In 2026, with the World Cup running in Russia, I was on the Mumbai desk. I built a 32-team valuation model — contract expiry, release clauses, sell-on percentages, wage-to-revenue ratios — and ran it live through the tournament. The remote desk out-trafficked the on-site filing three to one. The lesson was not that remote is better. The lesson was that a model is better — and the first virtue of a good model is that it tells you when its own input is empty.

The whole of this has been running in cricket media for years, only without the technology. A franchise press release has every field filled — quote, date, photograph, caption. Inside there is not one item that can be tied to a document. On several occasions I have read an entire release and then had to decide what to publish, and the answer came back: nothing. That silence is the most expensive thing on the beat, because empty space gets filled quickly. When nobody covers a smaller board's match, the space fills with rumour and invented reporting.

The window closes for clubs; the audit never does. Before I work a tip there is one question left: who else benefits from this appearing? This time the answer was clean. The empty ID being published benefits the pipeline whose daily job is counting stories. The loss fell on a reader who looked at that page and concluded that nothing happened in cricket that day.

The next domino

The next time a story reaches your feed, pick one field out of it and ask: where is the document? If there is no answer, it was not a story. It was a format.

The rule on my desk stays as it was — the number gets published first and the denial second. The day cricket's content chain learns to treat its own empty IDs as failures, half its stories will disappear. For this industry that will be the best news of the year. The question is whether anyone will be counting by then.