Don't Read 'N/A' as 'Low Risk': The Real Cost of a Silence Inside Cricket's Data Supply Chain

Eleven-forty on a Delhi morning, day four of the transfer window. A file sits open on the desk — eight columns, and every column carries the same sentence: "Insufficient information, assessment not possible." Twenty minutes later the same file reaches a dashboard and turns green. The dashboard does not read empty cells. It reads "no risk recorded."

This is not the first time in my career that a blank sheet has turned green. But every time it has happened, the cost was not paid by one man — it was paid by the whole chain that forwarded the green.

Let me be precise about what was in that file. Title: N/A. Source: N/A. Article type: unclassified. Information points: zero. Core viewpoint: blank. Entities: zero. Time sensitivity: never assessed. Source quality: unassigned. One field was populated — a regional label: cricket_asia.

That was not a cricket report. It was a packet that died before it became a report.

In a transfer window, a blank sheet is the most expensive object in the room

Every transfer window opens with a number nobody trusts yet. This one is no different — retention lists, release-clause structures, agent calls, and three competing versions of the wage bill circulating before lunch. The reader is not short of information. The reader is short of a filter.

Building that filter is the data desk's job: tell people which rumour arrives with a document and which arrives with only a tone. That job can be done honestly only if we say the word "empty" out loud when our own pockets are empty.

A cricket data desk runs in two stages. Stage one breaks an article down — information points, entities, time sensitivity, source quality. Stage two builds on the extracted material: format benchmarks, player technique, squad structure, auction valuation, governance, a risk matrix. Stage two is in debt to stage one. If stage one extracted nothing, stage two has nothing to write except an honest statement that it has nothing.

That is exactly what happened here. All eight dimensions were opened — format and match, player technique, team landscape, league and commercial ecosystem, rules and governance, risk, public narrative, industry transmission — all filled, all returning the same verdict: insufficient information.

That is not a failure. That is discipline. But discipline survives only if it travels. If it turns into "low risk" at the next desk, we have merely swept an empty room rather than lit it.

Three tests that apply to a blank sheet

One: zero and absent are not the same number

In 2026 I was feeding ball-by-ball data to Cricinfo from a cramped vendor office in Connaught Place. After Cronje's confession and the CBI report that December, every anomalous innings invited suspicion. I pulled eleven seasons of one-day scorecards — 1,047 matches — and built a collapse index. Eleven innings were flagged.

Two flags died under scrutiny. One was a rain-reduced chase where the over count itself had changed, so what looked like a batting collapse was arithmetic collapsing. The other was a No. 8 batter playing with a broken finger — an anomaly, not a fix. I retracted both flags in print, by name, in the same column.

The lesson lands directly on this file: a flag is only a flag if its sample survives scrutiny. Here the sample is zero. A risk score built on zero sample is not low risk. It is an absent risk, and the difference between those two is the whole argument.

Two: without a format, there is no benchmark

If you cannot establish whether you are looking at a Test, an ODI or a T20, you cannot evaluate a batter. Average and strike rate mean different things in each. In Tests, an average of 45 is excellent. In T20, an average of 45 at a strike rate of 118 is what teams politely call an anchor problem. Same number, two different courts.

I made exactly this mistake at the 2026 World Cup, and I have carried it since. On 27 June in Kazan, Germany out-shot South Korea 26–11 and lost 0–2; my sheet had Germany well ahead. On 1 July at Luzhniki, Spain completed 1,029 passes and lost to Russia on penalties; my xG sheet read Spain 2.4, Russia 0.6.

I filed on PPDA that tournament, arguing that Russia's 28.4 marked them passive rather than defensive. That held. But when I carried the same logic home into T20 powerplays, I learned the model dies the moment you change format without re-specifying every component. Intent in a powerplay is not intent in the last thirty minutes of a football match.

So when a file reports that the format could not be established, it is not reporting a technicality. It is reporting that any conclusion built above it will be a decision without a yardstick — and a decision without a yardstick in cricket is always confident, because it has no way of finding out it is wrong.

Three: a label is not a story

Only one field in that packet was populated: cricket_asia. That is a bucket, a filing cabinet. It tells you where an article was shelved far more than it tells you what the article said.

This is the most common and most expensive error in sports data. Quoting a batter's home average without saying which ground produced it is the same mistake. One bowler's economy at Kotla and at Dharamsala are two different stories. A label tells you which city you are in; it does not tell you which street.

Reading cricket_asia as evidence that the article concerned an Asian team, board or league is not inference. It is invention. And I invented once, in 2026, when two flags failed. After that I made a rule: put the conclusion where the evidence stands, and where there is no evidence, leave the space visibly empty.

Four: without a source tier, a narrative has no weight

Author stance and article purpose were both N/A. That means it could not even be established whether the piece was reporting, opinion, rumour amplification or promotion.

This distinction matters most in cricket, because three separate source-worlds run in parallel: Indian domestic media framing, mainstream English-language cricket media, and self-media. Their speeds differ, their verification processes differ, and the cost of their errors differs. Weigh all three on one scale without grading them and you are not measuring, you are guessing.

Five: and here the auction comes back

February 2026, Mumbai — the inaugural IPL auction. I built a replacement-level valuation model for one of the eight founding franchises, trained on 2026 T20 World Cup ball-by-ball data and two seasons of the Syed Mushtaq Ali Trophy. My sheet called MS Dhoni's $1.5 million price fair, Kolkata's $950,000 for Ishant Sharma a 40% overpay, and Rajasthan's $475,000 for Yusuf Pathan the best value in the room.

Yusuf was Player of the Match in the 2026 final. Two of three calls held.

Now notice why the two held. Not because the model was elegant — because the input sheet was complete. Format stated, role stated, sample limits written down, and beside every number a note on what it was made from. On a day when the input sheet is incomplete, it is not the model that bids. It is the guess, and the guess has the loudest voice in the room.

That is why what this season's reader needs is not another xG-style composite score. It is a column that says how many documents sit behind a claim. A price is only a rumour until the data agrees — and the data's agreement carries a date.

Now hear the other side

Here is where I have to challenge my own posture. Thirty years of being right about the counter-intuitive call teaches a man that the number nobody trusts must be the true one. That is not rigour. That is laziness with a track record.

So read it upside down: perhaps this was not a pipeline failure at all. Perhaps the source article genuinely was empty — a post with a photograph and two lines, the kind produced daily, which yields exactly nothing when deconstructed. If that is the case, the analysis did its job perfectly. It called a non-story a non-story.

What would falsify my extraction-failure read? A clean test: run the same article through stage one again. If the second pass also returns zero information points and no named entity, the source was empty and my read is wrong. If it returns three or more points and at least one named entity, the failure was mechanical, not editorial. I am publishing that test alongside the finding, because a contrarian claim without a falsification condition is just another opinion with better vocabulary.

Second, and this is what the industry needs to hear. Every conversation about cricket analytics ends with "more data." But this file did not lack data. It lacked provenance. Volume and lineage are different problems. We spent a decade solving the first and never touched the second. A data desk that cannot say where a number came from gains nothing from forty pages. I ran the numbers forty pages deep before I learned to listen — and listening did not mean the data was scarce; it meant I finally knew whose voice it was speaking in.

Third, and least comfortable. The person who converts "insufficient information" into "low risk" is not usually dishonest. He is on deadline, he has eighteen other files, and his tooling offers him a colour to drop into an empty cell. That is not a character problem, it is a design problem — and design problems are fixed by changes, not sermons. The number that should sit at the top of every dashboard is not a risk score. It is the input-completeness percentage. If that percentage is zero, every other colour on the page is a lie.

Fourth, a complaint about my own trade. The data desks have won the argument — selection, workload, auction pricing, the models are inside all of it. The commentary box has not filed its response. Even now the fastest analysis after a big match is often the analysis we got thirty years ago, with two new acronyms bolted on. This blank-sheet story is a photograph of that gap. A profession holding this much data, which still cannot separate absent risk from absent evidence, has a problem of habit rather than of numbers.

What to watch from here

Three things, all countable.

First, for every report that links a player to a franchise "according to sources": does it state a format? If not, it has no benchmark, and a claim without a benchmark is a tone.

Second: is there at least one verifiable named entity — board, franchise, agent, contract date? In Delhi, statistics was never math; it was patience in disguise. Patience means hunting a document behind a name instead of a feeling behind a tweet.

Third, and most important: did the outlet grade its source tier — official board statement, confirmed journalist, general media, or aggregator? Those four carry different prices. An outlet that does not distinguish them is asking the reader to bring his own scales.

The spreadsheet said one thing; the stadium said another. This time the spreadsheet said nothing at all — and that was the loudest statement made that day. The question is not how to fill a blank sheet. It is whether, the next time a blank sheet arrives on your desk already coloured green, you will ask what was written on it before you believe the colour.