Cricket's Empty Dataset and the Provenance Crisis: When 'No Verdict' Is the Most Trustworthy Report

A pipeline ran all night. In the morning the table opened, and every cell read N/A. Article title N/A, source N/A, one-sentence summary empty, author stance N/A, and the entire information-points section empty — that was the field that invalidated everything else. The only surviving signal was a label: cricket_world. Which told me the subject was cricket, and nothing else.

I looked at that empty table twice. The first time I was annoyed. The second time I decided it was the most honest document I have read this season. A table that says "I have no facts" cannot mislead anyone. A table with twelve numbers in it and no one who knows where they came from misleads everyone, and does it with total confidence.

The most dangerous number in cricket coverage is not the wrong one. It is the orphan — no provenance, no definition, no sample size, no era. This piece is an audit of that orphan. And the audit starts with an empty table, not a cricket ground.

Context: what a cricket desk actually receives

When I sat on a Kolkata digital desk, calling a cricket piece "data-driven" meant passing it through three layers that everyone conflates. The first is the event layer: ball-by-ball. Who bowled to whom, in which over, for how many runs, and how the wicket fell. This layer is raw; it contains no interpretation. The second is the aggregation layer: strike rate, average, economy, phase splits. Every number here is the product of a definition, and change the definition and the number moves. The third is the interpretation layer: value, impact, "the shape of the match." Readers want the third layer immediately, because that is what gets called insight. And that is precisely where orphan statistics are manufactured, because by the time anyone reaches the third layer they have forgotten the first.

My definition of provenance is deliberately narrow: definition plus sample size plus era. If any one of the three is missing, the number is not quotable. The strike rate needs its definition — are wides and no-balls in it, is it simply runs divided by balls. The sample needs its N — how many innings, how many balls, how many matches. The era needs its phase — powerplay, middle, death. What remains without all three is not a statistic. It is a tone of voice.

So when the pipeline returned empty fields, my first instinct was not to fill them. It was to note that a system which can declare its own failure is more trustworthy than a system that converts failure into product. A television desk does not have the option of returning an empty table. It has a 22-yard camera and four minutes of air. So it fills the gap with orphan statistics. And that is why its numbers cannot be honest.

The import audit: a metric crossing a border brings its luggage

Let me be clear about the most expensive lesson of my career. When football vocabulary enters cricket — expected goals, possession share, esports-style pace ratings — the word is not the heavy object. Words are light. What is heavy is the assumption hiding behind the word, because that is what actually crosses.

I did not port the words; I ported the assumptions hiding inside them.

xG was built to convert shot quality into goal probability. Its founding logic was this: football produces few events, goals are rare, therefore you need a bridge that measures the quality of chances in a low-event sport. Now look at cricket's structure. Every ball is a complete, closed, recorded event whose outcome is known instantly and which changes the state for the next ball. Cricket does not need a bridge to measure chances. The count is already there.

Which is why the xG logic breaks at the exact point where a wicket gets treated as a turnover. In football, losing the ball is recoverable. In cricket, a dismissed batter does not return in that innings. A wicket is simultaneously a cost and a resource. A model that cannot handle that double nature can forecast a score, but cannot price a situation.

Here is the thing I repeat until people get annoyed. Every clean dashboard begins with a dirty question nobody wanted to ask. When a metric arrives, the first question is not what it tells you. The first question is what it was designed to predict, and whether cricket's event structure can supply it. Until that question is asked, every metric is an entry without a visa.

Core: three cases, one disease

Ahmedabad, 19 November 2026: no century, and the truth elsewhere

Recall the scorecard. India bowled out for 240; Australia 241 for 4 in 43 overs. Not a single century in the match. Travis Head made 137. Virat Kohli made 54 off 63. Rohit Sharma made 47 off 31. KL Rahul made 66 off 107. Anyone reading only the big box on the card — the hundred — saw a low-scoring final with no hero.

That reading is convenient and completely wrong. The phase split shows the architecture of the match, and that architecture is built from blocks of overs, not from round numbers. Kohli's 54 off 63 is the kind of figure that builds trust while hiding cost: it was the longest stretch in which India's tempo held, and it was also the stretch that compressed them through the middle overs. Rahul's 66 off 107 is the clearest evidence of that compression — sixty-six runs in more than a hundred and fifty balls, in an innings with a ceiling of three hundred.

The problem is that the hundred is a counting convention inherited from print-era scorecards, invented when a reader needed one line to summarise an innings. It does not measure match value. It counts. In a match with no hundred, the whole scale goes blank and the commentator is forced into "the batting didn't fire" — a sentence that is not analysis but a caption pasted into an empty cell.

What actually happened is that India traded runs for wickets through the middle: they let the wicket count weigh more than the scoring rate. On a surface where 240 was roughly par-defensive, that trade went the wrong way. Australia's chase was built on Head's 137, but the more important number sits beside it — only four wickets down in 43 overs, six wickets in hand at the finish. That gap in wickets in hand is the real provenance that Head's hundred hides, because we all look at the 137 and nobody looks at the six.

Dubai, 9 March 2026: a second final, a second empty century box

In the Champions Trophy final, New Zealand made 251 for 7. India made 254 for 6 in 49 overs and won by four wickets with six balls remaining. No century in the chase either. Rohit Sharma made 76 and was player of the match.

Notice that the scorecard's big box is empty and so is commentary's big sentence. And yet this is the most controlled knockout chase of the cycle. The evidence is not in the hundred; it is in the arithmetic. On a slow Dubai surface, 251 was not a soft target — the required rate sat near six, which sounds trivial and plays expensive. When the required rate drops, runs stop being the instrument of victory; wickets become it. India kept six in hand to the end, and that is the chase's actual story.

Where the required rate falls, wickets in hand become currency; where the required rate climbs, the same wickets become collateral locked away. That one sentence carries a framework the century count can never supply, because a hundred is a batter's property and wickets in hand are a team's position. They sit on different layers.

One thing should be recorded plainly: Shreyas Iyer's 48 and KL Rahul's unbeaten 34 were not worth less than Rohit's 76. Perhaps nobody read the chase that way. Fine. I am not arguing Rohit's innings was small. I am arguing that 76 was the most visible number, and the most visible number is not automatically the most important one.

Kolkata, 2026: the sheet was right for the wrong reason

In 2026 I spent four months on a model called Expected Runs Added, built on 2026 IPL ball-by-ball data. Its conclusion was that KKR were wasting Sunil Narine in the middle order. I published that piece before the season. In April, KKR opened with him, and he finished IPL 2026 with 224 runs at a strike rate of 172.3. The piece got roughly four hundred thousand reads, and an unsolicited email arrived from the franchise's analyst asking for my methodology sheet. He asked for the method, not the conclusion. That distinction is the whole health of the market in one line.

This story gets retold constantly, and almost always wrongly. The convenient version is: the model was right, the franchise listened, Narine flew. That is not what happened. The sheet argued against wasting Narine inside a middle order where his ball-by-ball return was lowest. The franchise read the same sheet and drew a different inference — he opens, because his edge is in the powerplay matchup, not the middle. Two different claims that happen to point the same way.

The outcome validated the decision; the outcome did not validate the model's mechanism. This is cricket analytics' classic trap — the place where correlation gets promoted to causation. After that season I cannot stand on any platform and claim my model predicted the opening call. I can claim my sheet opened a door, and that the room the franchise walked into was not my model's room.

The auction is the selector

One more thing Indian coverage keeps ignoring: purse arithmetic. If anything now gives a public signal about what Indian cricket values, it is retention and auction decisions rather than zonal selection. That leads to an unfashionable conclusion — India's bench strength is a selection problem, not a talent problem. The hardest team in world sport is India's second XI.

But provenance is mandatory here too. An auction price is not evidence on its own; the price only means something alongside the purse size, retention rules and right-to-match mechanics. Quoting a fee without the purse is the orphan-statistic disease in expensive clothing.

Contrarian: state the strongest version first

Now the position I am attacking, in its strongest form, because beating a weak version of it would be dishonest.

The case is this: cricket coverage's floor has risen over the past decade. Phase awareness is common knowledge. Matchup language reaches ordinary viewers. Auction economics gets debated on television. All of that is analytics, and the claim is correct. Coverage is better than it was.

Grant all of it, and what remains is this: the same pipeline that raised the floor also opened an orphan-statistic factory. Once you have proved that "data talks," any number can talk. "Impact" arrives — defined how? "In form" — over how many innings? "Value" — against which era, when boundary dimensions, bat edges and the two-new-ball rule have all moved? I have seen comparisons printed on desks I worked at where a 2026 strike rate was set against a 2026 strike rate with no acknowledgment that the game between them had changed shape.

And the empty-table incident reveals the industry's default: when fact extraction fails, fill the gap with narrative. It fills with story rather than fact because story is cheap. The pipeline breaks and the desk writes "the batting had no life in it." That is not failure management. That is failure production.

So my position is the ENTJ position: I want a verdict. Where the data does not exist, the correct verdict is no verdict. That is not weakness; it is the highest output an audit can produce. An analyst with an answer to every question probably has not framed a single question properly.

Takeaway: the next-round signal is a convention, not a metric

The next signal will not be a new impact index. It will be a habit: writing N in the first line. A desk that writes "across 42 innings" can be checked. A desk that writes "he's in form" cannot — and what cannot be checked is not journalism, it is a weather forecast.

One last question, for myself. The next time the pipeline returns an empty table, will we publish it in that state? Or will we fill the cell with a graceful sentence and tell the reader the batting had no life in it? The day we can print an empty table is the day the rest of our numbers start meaning something.

---

Disclosure: datasets used here are public scorecards (official match records for Ahmedabad, 19 November 2026 and Dubai, 9 March 2026), the public ball-by-ball extract of IPL 2026-17, and the 2026 Expected Runs Added sheet built on a Kolkata desk. I have no access to any franchise's internal scouting or tracking database, and I cannot report what happened inside BCCI selection meetings or contract negotiations. Where a figure is rounded, it is marked as rounded.