Empty Cells, Confident Conclusions: The Biggest Hole in Indian Cricket's Data Age

The thread began at 1 a.m., and the replies told the real story. A number was doing the rounds on the timeline — a bowler's death-overs economy — and three hundred people were arguing about it. Nobody asked where the number came from, which format it belonged to, how many overs the sample held. I opened the file I keep titled "narratives that will be wrong by Friday" and added a new column: source. Nine of twelve claims left it blank.

This is not a new disease in cricket, but the scale has changed. Once, estimates were made and called opinions. Now, estimates are made and called data. The difference matters: an opinion can be argued with. Data makes people bow — even when the data is sitting on an empty cell.

I opened a spreadsheet to test how empty the empty space really was. The answer: very. And the least expected finding was that nobody had invented those numbers. They had simply outsourced the filling of a blank to somebody else, who had outsourced it again.

Context: three pillars and a fourth nobody audits

Indian cricket's power structure rests on three things — the calendar, the auction and the broadcast. The BCCI decides who plays when and against whom; the IPL decides what a player is worth; the broadcast decides which story gets air and which gets buried. A fourth pillar has grown up between them, politely called "analytics," and it is the least audited of the four.

The problem is structural. Cricket has four principal formats — Test, ODI, T20 and The Hundred — and their numbers do not translate. A Test strike rate of 50 and a T20 strike rate of 150 describe different competencies, yet our talk shows and our threads put both in the same sentence. The ICC rankings sit in three separate tables, which means "ranked third in the world" is half a sentence until the format is named. That is not a technicality; it is the foundation of the entire argument.

The second layer is selection. Selectors now carry tablets, slides, and an "impact metric" glowing on every slide. But the largest input into that machinery is the same one it was in the 1990s: franchise need. A player goes for a large sum at auction and spends the next month on the national bench — a contradiction we happily admit over tea and quietly omit from the model.

The third layer is the home-away differential, the single largest explanatory variable in team performance: spin dominance at home in India, the difficulty of adapting to pace away. A model that does not isolate that variable is not a model. It is an average. And teams are not picked from averages; the habit of picking from averages is what teams become.

Core: four empty cells carrying the weight of the argument

First, format contamination. This is the most common and most expensive error. A batter's T20 record says nothing about Test technique; a bowler's Test economy says nothing about death-over capacity. When I audit a claim, my first question is: which format, how many matches, in what situation. Two times out of three, there is no answer. And we treat the absence of an answer as "obviously."

Second, mistaking an auction price for proof of performance. The auction is a market, not a measurement. It contains demand, availability, the domestic quota and panic bidding. A player can go expensive because his skill is scarce, and cheap because four others offer the same skill. The same logic governs retention and Right to Match — those rules do not set a player's value, they set a franchise's bargaining power. An analyst who conflates the two is doing accountancy, not cricket.

Third, home numbers being repackaged as overseas claims. A spinner's figures in India look magnificent because conditions are on his side. When he travels, we discover how much of that success was his fingers and how much was the pitch. That audit is hard, so it is skipped. And what is skipped in analysis does not stay neutral — it goes missing in selection.

Fourth, the "next Tendulkar" forecast. The historical fulfilment rate of that framing in Indian media is very low, and the reason is technical: a T20 league explosion on a small sample translates poorly to international cricket, because both the quality of bowling and the pressure change. Virat Kohli and Sachin Tendulkar appear here not as comparisons but as markers of a language we load onto a teenager every three years and then spend a decade waiting to unload.

And fifth, the most dangerous: reading "no risk identified" as "no data supplied." These are entirely different sentences, and Indian selection debate merges them constantly. When there is no recent information on a player, he is treated as risk-free, when the truth is that we know nothing about him. That is not a failure of argument; it is a failure of data discipline. I have watched an empty column make a large decision, and I have watched nobody audit the decision afterwards.

So when a slide reads "risk: low," the question should not be how much risk there is. The question should be whether risk was ever assessed. Both print in the same colour.

Contrarian: where I could be wrong

Now the part where I attack my own argument, because a hot take that does not survive its own audit is just noise.

The first objection: dashboards have reduced guesswork, not increased it. That may be right. A clear metric, even an imperfect one, is often better than the instinct produced by a pre-formed impression. If selectors now at least write down the basis on which they picked someone, a door to transparency has opened — even if the room behind it is empty. I concede it, because an audit is not a claim that everything is broken.

The second objection applies to me rather than my subject. I was born in the US, I sit in Britain, and I write about Indian cricket. That distance lets me perform detachment, and the same distance can make me forget that selection decisions are not made in a room but in a set of relationships — coach, captain, franchise owner, state association. An analyst who does not hear those voices writes a clean piece and reaches a wrong conclusion.

The third objection is the strongest: the WPL and the new leagues have genuinely changed the talent supply in women's cricket over the past few years, and the constraint there was opportunity, not data. That weakens my thesis that the framework itself is the problem. Sometimes an empty cell simply means nobody has yet walked out to the middle to fill it. I accept that.

Takeaway: a prediction you can check

I have learned more from empty columns than from final scores, and this time was no different. What looked like a hot take was actually a paper trail — we have simply lost the habit of reading it.

So here is my testable prediction. Watch the next major selection press conference and count how many numbers are spoken with a source and a format attached. My estimate: no more than three out of ten. And if, within the next two IPL cycles, at least one big-money auction buy is out of the T20 squad inside eighteen months, you should be asking how that price bought a place it never earned.

The question is not whether we have data. The question is what we put in its place when we did not.