The Empty Array: When Cricket's Data Desk Reads N/A as Clean

There was a green tick burning on the screen. Below it, a button: "Analysis Complete". Below that, in a white cell, two square brackets, as empty as the Mohali press box on a rain day: [].

The record had eleven fields. Title — N/A. Source — N/A. Article type — unclassified. And the field called 'Information Points', the one the whole report stands on, was zero. The arithmetic is simple: zero of eleven fields were usable. The tick was green anyway, because the system does not check whether a file has content in it. It checks whether the file arrived.

This is not a report about a stadium. It is a report about the machinery that now writes about cricket — and about one evening when that machinery returned empty-handed and still filled in the entire form.

Context

Cricket news no longer comes out of a scorecard. It comes out of a pipeline. Franchise analytics desks, fantasy-platform models, broadcaster graphics, the agency night shift — all four run on the same raw material: match data and its text. When the text does not arrive, the pipeline does not stop. It fills its empty shell and picks up the next file.

In 2026, at thirty-six, I wrote and narrated a six-part digital series for a Bengaluru streaming startup, marking ten years since India's 2026 T20 World Cup win. I tracked down eleven of the fifteen squad members and built the third episode entirely around the bowl-out in Durban: 14 September 2026, Sehwag, Harbhajan and Uthappa all hit, Pakistan missed all three. My kinesiology training let me break down seam position on screen, but the viewer comments were all about the silence before Uthappa's run-up. After that day I stopped opening scripts with the score. I started opening with an object — a scuffed ball, a wet outfield, a sponsor board. And I learned that you keep the camera low so the boys can look tall.

That lesson inverts here. A text parser cannot read silence. The thing that is a writer's richest material — the pause — is, to a machine, only absence, and absence gets written as zero. An empty stadium and an empty file look identical to it.

Core

Now the part that never makes it into a published report.

The first condition of analysis is format. Test, ODI and T20 numbers are not interchangeable; a strike rate of 140 is ordinary in T20 and a different organism in a Test. Without a settled format, every benchmark dissolves. This file had no format field at all — only a category label, cricket_world, which tells you the subject is cricket and nothing else. No gender, no competition level, no bilateral series, no league.

The second condition is the player. The payload contained not one name — no cricketer, no coach, no support staff. Without a role you cannot even say whether you are looking at an opener or a finisher, a seamer or a spinner. And without a role, average, strike rate and economy all go inert, because none of the three means anything outside a role.

The third condition is the team. No squad, no selection list, no bench strength. The ICC keeps three separate ranking tables and they can never be merged; here it is not even established which format the question belongs to. The World Test Championship table cannot be examined when it is unknown whether the subject is Test cricket.

The fourth condition is money. No rights deal, no franchise valuation, no auction figure, no right-to-match reference. The most important discipline in this business is that a high IPL price is not evidence of international strength — but that discipline only bites when there is a transaction in front of you, a price on the record. Here there was no transaction at all.

The fifth condition is the dangerous one, and this is where the real story sits. When a file contains no governance event, no rule change, no DRS controversy, no sanction, then compliance risk is not "low" — it is "unassessable". The distance between those two is the distance between a finding and a confession. "Low risk" is a claim: I looked, I checked, I found nothing. "Unassessable" is an admission: I did not look.

That error propagates through the whole document. The overall risk rating came out not "low" but "not assessable" — because risk is a product, likelihood times impact. Zero risks were identified, so writing "low" would have been false comfort. Narrative analysis stalled for the same reason: measuring the gap between market expectation and fundamental reality requires at least one stated expectation on the record. And the author's stance? Also N/A. Which means it is not even known whether the missing piece would have been praise or scepticism, hype or doubt. Source quality could not be graded either, so no conclusion could carry a confidence weight.

Transmission — the path of a shock from talent supply through national teams and broadcast to derivative markets — came out as an empty map, because there was no shock. With no shock, there is no path to trace; only an arrow pointing nowhere.

And the tournament cycle belongs here, because the whole ecosystem is currently running under exactly that pressure. During a major tournament the demand for copy doubles and the time available to check it halves. In that weather, empty records are at their most dangerous, because the urgency to fill them is at its highest.

Now consider the human who receives this record. Three in the morning in a newsroom, forty slots to fill. A fantasy-platform model that treats every file as a signal. A team analyst carrying one page into the morning meeting. All three face the same problem: a record being present is not the same as content being present — and a database does not distinguish between the two. A field that looks filled is assumed to be filled.

Writing this, I think of Kazan in 2026. Thirty-two fans from Malappuram and Kochi had sold gold and taken loans to watch Argentina. France won 4-3, Pavard's volley in the 57th minute. I had nineteen hours of their songs, arguments and crying, and cut a twenty-two-minute film with no tactical graphic in it. The hardest decision was not to narrate over the crying. That silence was material, not emptiness.

My forty voice notes for The Empty Stand were built on precisely that distinction — what the voice did not say was also being said. But there the absence had a shape and a meaning. Here the absence means nothing. A selective gap tells a story; a total gap tells you only that the pipe broke. Which is why not one sentence about cricket can be extracted from this file — not that a team is strong, not that a player is in form.

Contrarian

The industry is worried about the wrong thing. We are all buying the fear that a machine will hand us bad cricket writing. Bad copy gets caught — a reader quotes it, a fan screenshots it, word reaches a player's family. What does not get caught is the complete copy: every heading in place, every subheading tidy, and nothing underneath. The danger is not that the machine writes badly; the danger is that the machine writes formally correct prose, and formal correctness is the best camouflage emptiness has.

The second inversion is that this was not an accident but an architecture. Cricket pays for volume. A Ranji round throws up a dozen matches at once, and who is going to staff a dozen separate analysts for them? So the pipeline is built to emit twelve records whether it read twelve matches or two. The incentive is completion, not provenance. And when the incentive is completion, an empty file will also present as "complete" — its green tick is not a witness to a match, only the tail end of a workflow.

The third truth is the least heard. The players who most need accurate coverage are the ones whose numbers get checked least. A superstar's average is cross-examined by a million people. A twenty-two-year-old's Ranji debut figure of 41.8 is checked by nobody, because there is perhaps one person reading it. Cricket built anti-corruption units in the 2000s to protect its image, and convened committees over spot-fixing. It has never once run a correction for an unsourced average.

Takeaway

What the system lacked was not technical. It needed one condition: if information points are fewer than one, quarantine the record and do not pass it downstream. That is a single line of code. The hard part is the editorial version of that line, which does not yet exist in any newsroom workflow — the person, or the process, able to say this report does not look whole, it does not go out.

The next time a table tells you a debutant averages 41.8, ask who saw it. And when the answer is that a machine saw it, ask what the machine saw.

I did not close that screen. I kept looking at it, and the green tick kept burning. On a rain day in a stadium the scoreboard shows exactly what it showed at the last over — what did not happen is recorded nowhere. The difference is that there, everybody knows it rained.