A Null Is Also a Result: The Hole in Cricket's Record That Nobody Logged

On a Monday morning in Pune, my desk had two things open: a spreadsheet and a blank screen. The spreadsheet had forty-seven columns. Forty-six of them carried the same phrase — insufficient information. The forty-seventh carried nothing but a tag, cricket_world, which does not match the label our desk is supposed to use, Cricket. No title. No source. No publication date. Article type: Unclassified. And at the bottom, the list of information points, empty — two bare brackets, as though someone had started counting and walked away.

That is not a match report. It is the autopsy of a report that could never be written. I opened the file three times that morning, re-fetched from the server twice, and once turned the screen towards a colleague to confirm the fault was not in my eyes. The fault was not in my eyes. That payload did not even contain a zero. A zero is a number. This was absence. So I did the thing my trade does least: I left the blank blank, and I published it.

What follows is the record of that decision, and more usefully, the record of the habit that made it necessary. The real problem with Indian cricket argument is not that people do not watch enough. It is that what is watched is not logged. And what is not logged gets replaced by memory — memory that is always generous, always builds a story, and always leans towards the player the commentary box has already decided to like.

I started at The Indian Express in Madras in June 2026 as a trainee sub-editor. A rule was taught there that appears in no style guide: any sentence carrying a number must have paper behind it. I thought it was about proofreading. It was about journalism.

What I learned between the ages of eighteen and thirty-three is that no claim about a match is complete without a ball-by-ball log. That sounds obvious. In practice it means slowing down. It means giving up the thrill of running alongside the scorecard. It means surrendering the moment when the commentary box has already delivered the verdict and wants only your confirmation.

Which is why I feel nothing like pride about that blank screen — only the plainness of a process that held. Read correctly, that null row is the cleanest X-ray we have of an illness nobody in the cricket industry is treating: the habit of not keeping records.

To understand where the screen came from: for three years I have worked on a pipeline that lifts hundreds of cricket articles a day, breaks them down, and pulls out five things — headline, source, article type, information points, time sensitivity. An information point in this pipeline is a strict unit: an atomic, checkable fact containing a number, a quote, or a named event. These points are the foundation of every metric downstream. Without them, whatever gets added on top is inference wearing technical clothing.

That day the return was zero. I had two ways to fill it. One: write something. Plausible cricket content is among the easiest products in the world to manufacture, because no reader can verify every claim. Two: explain what the zero is a zero of. I took the second, and this article is what came out of it.

Now the least entertaining and most necessary part — method. I recorded the screen. I noted the time. I pulled the null rate for that outlet: how many of its articles returned an empty information-point array. A single empty return is not a crisis; every system fails a little. The real question I asked that morning was: is this a failure of the newsroom or of the machine?

Three possibilities. The source page used a template the parser could not read. Or the page opened, text was extracted, and the output was truncated — half an answer arriving. Or there was no text to begin with: a video, a paywall, a live-score widget. None is good news, but one distinction matters. In the first and third cases the fault is elsewhere. In the second, the fault is at home.

The evidence I had pointed at one odd thing: there was not a single named entity in the payload. A serious cricket article — match report, governance story, auction arithmetic — almost always leaves at least one name behind: a board, a league, a broadcaster, a player. Zero names, combined with an Unclassified type, suggests not that half the text was read but that no text arrived. In my internal taxonomy that is not an accident. It is instrument failure.

Here I have to name the addiction every data journalist falls into, and which this article exists to resist. When data is missing, the hand reaches for the nearest number: a spectator's memory, a Twitter thread, an old season's figure, another format's average. That is proxy inflation. I have watched it happen in my own work. Editors ask how much pressure there was, how badly someone wanted it, who stood up in the clutch. A convenient estimate slides quietly into the seat of the quality it cannot measure. Name every proxy, state what it does not measure, and never let it stand in a sentence as if it were the thing itself.

The root of that illness is the same place the blank screen came from. The Ranji Trophy is still argued about in sentences, not in deliveries. Domestic fielding data, dropped catches, wicketkeeping speed, the revolutions on a spinner's delivery — much of it is either uncaptured or unavailable. Selection debates therefore default to reputation. When you do not have three years of a batsman's record against domestic spin, you sit down with an average of 45 that never recorded what kind of bowler those runs came against.

The hole is wider in the women's game. After the 2026 home World Cup and the WPL, the biggest question is not attendance but whether broadcast promises convert into a pipeline — and an honest answer needs domestic women's data that barely exists. I will not extrapolate from internationals, because that shows me what I want to see rather than what is there. When data is missing, the professional sentence is: unknown, and the absence is itself a finding.

The root of the habit is 25 February 2026. First Test of the Border-Gavaskar Trophy at the MCA Stadium, Pune. I was in the third row of the press box, twenty-four years old, a junior data writer. India were bowled out for 105 and 107. Australia won by 333 runs. Steve O'Keefe took 12/70 — 6/35 in each innings. The commentary box closed the match with two words: rank turner. I did not.

I logged every ball. Spin deviation, bat-pad catches, shot type, release point. When the sheet filled, one number stood out: 14 of India's 20 wickets fell to spin or bat-pad pressure. That did not overturn the rank-turner story; it marked its limit, because the remaining wickets belonged to the game the story was hiding. I published the table with a methodology note and no conclusion sentence. I logged every ball before I allowed myself an opinion. Editors found it slow. From that day my copy was treated as audit-proof, which became my signature.

The following year, 2 July 2026, Rostov. Belgium beat Japan 3-2 after Japan led 2-0 at fifty-two minutes, Nacer Chadli finishing a counter in the 94th. I was on the football data desk and built a live PPDA, xG and distance dashboard. My model recorded Japan's PPDA at 8.9 through sixty minutes and 15.2 after, with sprint count down 18%. When a broadcaster graphic mislabelled an xG figure, I published the correction and the source table. The log remembers what the highlight reel edits out. The correction cost me speed and made me the desk's reliable methodologist.

Then came the year that made me doubt the habit: 2026, the empty-stadium IPL in the UAE. Mumbai Indians beat Delhi Capitals by five wickets in the final. I coded all sixty matches, powerplay and death, and tested home advantage and toss effect. Designated home sides won 27 of 60 — exactly the noise band I had set before analysis. I refused to call empty stadiums a revolution and published a null result with confidence intervals. Editors disliked the headline. Coaches read the table.

That is how I look at the blank screen. I went hunting for a pattern and came back with its absence. Cricket fears that sentence because we have mistaken it for defeat. In a pipeline it is the most valuable thing there is: proof the instrument works, the threshold held, and nobody is building a career story on a thin sample.

My trigger for empty returns is fixed: a null rate above 5% for any single outlet. Below that is noise; above it is a signal. I set the threshold before writing this, not while writing it. A second trigger fires with it — zero named entities across several articles from one source, which confirms a fetch failure rather than a classification one. A third is label drift. That article returned cricket_world when our taxonomy says Cricket. It looks trivial. It is the exact mechanism that routes a piece to the wrong template and lets it disappear unread.

In the office the line thrown at me is always the same: when did your piece land? My answer never changes: the press broke a full ten minutes before the scoreline did. That ten-minute edge is worth nothing if the number printed is wrong. The same applies to a pipeline. The speed with which an empty payload becomes an estimated one is so easy that it starts to feel natural.

Here is what undercuts this entire article. I know nothing about my instrument's fault beyond the fact that the payload was empty. My strongest hypothesis is that text was never extracted, not that it was misread. But a hypothesis is not a fact. When I label something an estimate, I mean that if a colleague proves the text arrived and the classification failed, I will change the conclusion without argument. The best thing I did that morning was publish my own reasoning with a confidence level attached.

The first objection is fair: if the article is empty, why write about it? Because an empty return is also a report, and it carries a cost nobody prices. What if that piece was a retirement, an auction record, a sanction? It would have vanished from the monitoring set silently and nobody would have known. Five per cent is the tolerance for outlets we lift five articles a day from. On the day that line is crossed, no scoreline prints wrong — some stories simply never print. That is the real loss, and it never shows up in the ratings.

So the fix is a human one. Log the source link and headline by hand. Set a 72-hour window, after which news-type pieces lose their relevance. Keep three fallback routes for re-extraction: RSS body, AMP page, archive snapshot. And set the threshold in advance: if fewer than 70% of empty returns are recovered through those routes in a week, the problem is source access, not our fairness. The response then is licensing, not engineering.

We have long assumed in cricket that what is watched most is recorded best. Ranji is played most and logged least. The women's domestic circuit is now visible but still uncounted. The highlight reel is the fastest thing made and the first thing forgotten. Before the roar, there is a spreadsheet nobody thought to film.

That evening I closed the screen and entered the null row in the register with a note: null result, cause unknown, filling forbidden. Next month, in the batch null-rate review, that row will be a small red light in a table. I do not know what was in that article. I know that filling it would have added one more fake number to a cricket argument that nobody ever retracts. The question is not what that article contained. It is how many more will drop out of our pipeline in the next six months — and who has been assigned to count them.