The Header Survived, the Scorecard Didn't: A Silent Failure in Cricket's Data Chain

On a February morning at Gymkhana Ground in Hyderabad, I turned to an old page in my notebook. The date was written at the top, then "net spell — Siraj", and after that twelve blank lines. Twelve overs had been allotted a space; twelve overs had never been written. Rain had cut the session short and I left the page as it was — header alive, content absent.

Seven years later the same page returned, not on a field but on a screen. An analysis file arrived with "Domain — Cricket" written at the top. Below it: title blank. Source blank. Format undefined. The list of information points: empty. In other words, an envelope with a clear address on the outside and nothing inside.

The subject of this report is not the match whose analysis was requested. The subject is the process that took an empty file, stuck a "cricket" label on it, and passed it downstream — which is the most cricket-like part of the entire episode.

From match to scorecard, cricket's data chain is less complicated than it is sold as being. A scorer at Gymkhana writes with a pen on paper. An operator feeds it. A server carries it to an aggregator, which carries it to the dashboard you check at night. There are four points in that chain where "nothing" can be recorded, and none of the four looks the same. The scorer's blank space is forgivable because it looks blank. An error code looks like an error. But at the fourth point — where analysis begins — the failure is invisible. There a structure is built that looks complete and contains nothing.

I learned to test that structure at eighteen, on an Under-19 World Cup data desk, using a three-column template: event, team consequence, fan consequence. No row was complete until all three were filled. It is why my piece on Shubman Gill's 372 runs in 2026 read differently: the scoreboard showed the runs, but the team consequence was that his strike rotation protected the middle order, and the fan consequence was India's eight-wicket final win over Australia.

Cricket already carries the distinction that matters most here. "Did not bat" and "out for 0" are different things. "Absent hurt" and "retired hurt" are different entries. The scorecard has preserved these distinctions for sixty years because it knows a blank is information, not a zero.

The file that started this argument lost that distinction. Everywhere it said "insufficient information". On the surface that looks like honesty. It is honesty, but incomplete — because one field did not say that. It said: Domain — Cricket.

That is the fault line. An empty file without a cricket label is waste paper nobody opens. An empty file with a label is a hazard, because it looks exactly like a full one. To the system, the label is a validity stamp. The moment it is affixed, the file becomes a processed object.

I learned this the day Hyderabad's assistant coach corrected a fact in my school-blog piece on Siraj's 41-wicket season. I re-reported rather than defended. The lesson was not that I had written a number wrong. It was that I had no second source to verify it — only my notebook and my memory. The label "analysis" was attached; the basis was not.

And if the format is undefined, every statistic after it is a gamble. Test, ODI and T20 have separate statistical grammars. An opener is judged on average and balls faced; a finisher on a 180-plus strike rate. Comparing across formats is the commonest error in cricket analytics. Without a format you cannot locate the powerplay, cannot assess home advantage, cannot even know whether rain revised a target under DLS.

The 2026 World Cup final proves the point. Match tied, super over tied, and the title decided on boundary count — England 22, New Zealand 16. The rule produced a result the scorecard could not express. A sport in which outcomes can be settled by something outside the scorecard should not be read as if the scorecard were the whole truth.

In 2026, when the country was shut and the Ranji Trophy was stopped, I spent six weeks at the Rajiv Gandhi International Stadium, interviewing fourteen groundstaff and nine club cricketers. The hardest part of that 4,000-word oral history was not the writing. It was deciding what would not be written. That gave me a habit I now apply to every file: identify who can be harmed, then state what was withheld and why. A report that does not say what it withheld is incomplete, however many numbers it carries.

This file did not say why the information was missing — and that distinction matters. Either the source article was genuinely hollow, in which case declining to analyse is correct; or it was full and never reached the parser, in which case what was lost is not an article but a player's record. I lean to the second for a technical reason: title, source and type were all blank, yet the domain label survived. A thin article keeps its title and source. When the body never reaches the parser, the title and source go first, and the externally applied label lasts longest.

What is solid is that the file contained not a single name — no team, no player, no venue, no coach. No cricket match report or preview can be written without two team names. That silence is itself information.

Whenever a data feed fails, the reflex is: this is an engineering matter, hand it to the engineers, it has nothing to do with cricket. That is the central misreading. Data is no longer external to cricket. A bowler who takes 41 wickets in a Ranji season wins on flat decks in August heat, bowling seventy or eighty overs across four days, and his entire season compresses into one line: 41 wickets, average whatever. The moment that line returns empty from a feed, his season does not exist on anyone's screen. The work happened; the record did not. And the absence of the record shows up in next year's auction, next year's contract, next year's selection.

My notebook has a page for every ball he bowled that season. But my notebook does not reach a selection committee. The database does. When the database returns empty, that is not merely a technical event; it is a man's work quietly erased.

The second misreading is that more data has made cricket better understood. More data has made cricket better measured, and measuring is not understanding. When four metrics are generated per ball, some fill themselves in whether or not they mean anything. A fast bowler's economy can look good over thirty wicketless overs — if those overs came after the match was decided, that good economy is a blank page with a fine headline on it.

The third misreading, the one that troubles me most, is that the system must show something. Returning an empty file feels like failure to a system, so the system's instinct is to fill in something. Humans share it. A journalist short on facts reaches for adjectives; an analyst short on numbers reaches for inference. Both are fabricating, one in words and one in digits.

I am not arguing that returning an empty file is good work. I am arguing that it is honest work, and honesty has to be designed in. The present architecture does not do that. A zero-information object with a label on it simply passes through.

The biggest risk here is informational, not sporting — which is why it is also sporting. A wholly empty file makes everyone alert. A half-empty file makes no one alert. The most dangerous state is a file whose skeleton looks complete — eight sections, subheadings everywhere — with not one defensible sentence inside. It takes time to read, is easy to quote, and contains nothing that would stand up in court.

Where I first learned the trade, there was a rule I thought pointless: every statistic had to be published with its base — how many matches, which format, what conditions. That rule was not protecting the number. It was protecting the person behind the number. The moment you start writing the base, you discover whether you have one.

Applied here, "insufficient information" is a good start but needs a reason attached: source not found, source found but unreadable, or source genuinely contained nothing. That three-way distinction is what could save the chain next time.

My notebook still has that page — date, name, twelve blank lines. I have not torn it out. It stays so that every time I think I have information, I look at it and remember that information and the appearance of information are two different things.

The next big fight in cricket's data chain will not be about faster feeds or better graphics. It will be about what the system does when there is no information. Does it quietly pass an empty envelope along with a label on it, or does it stop and say nothing can be written on this page? The first route is cheap and looks tidy. The second is expensive and requires someone to admit they have nothing.

The day a report says "this information is unavailable, and that is the most accurate information available right now", cricket journalism will have taken a real step. When the next blank page arrives, I do not know. That it will arrive, and that it will have a headline on it, I do know.