The Discipline of the Empty Cell: The Courage to Admit What Cricket Data Doesn't Say

Last winter, sitting down to build a bowling-workload model for a county side, I opened a spreadsheet with four columns entirely blank—average rest per over, minutes between innings, travel distance, and a measure of consistency within a spell. A colleague said, “Just run the model without them, the result will come anyway.” I had no verifiable source for those four columns. So I did not run the model. Instead, beside the empty cells I wrote a single word: “unknown.” That night I understood that the hardest task in cricket analysis is not running a model; it is admitting the void when there is no information. A spreadsheet does not lie—it waits until I can catch the truth. Cricket's data revolution arrived fast, almost silently. Every delivery is now captured by tracking cameras, every run-rate shifts by the second, every field setting appears on a map. Yet inside this abundance an old crack remains: what cannot be measured easily is still assumed to have been measured. When a team loses we say its “mentality was weak”; when an innings collapses we say it “could not handle the pressure.” These sentences sound weighty, but behind them there is no sample, no verification. I work in the transfer market, where the moment a name is heard the price starts rising—yet no contract is signed. The same thing happens in cricket analysis. After one innings a story forms quickly, and that story is gradually accepted as truth. So I learned to write three things beside any claim—sample size, source, and cut-off date. Without these three, an analysis is incomplete to me. The habit began at eighteen. In August 2026 I logged every shot of 46 matches by hand in a notebook—distance, angle, body part, defensive pressure. Nobody paid me. Because that season's success was being explained entirely by “momentum.” My sheet said something else. But I did not throw the sheet away; I kept it, so the evidence could be checked at any time. When a number reaches me, I first ask—who wrote it, when, and what did they leave out. The most neglected question in cricket analysis, to my mind, is this: in what context did this data arise? A batter's average of 45 looks excellent, but if that average comes from only three innings, it is not information, only noise. A bowler's economy of 6.2 seems expensive, but if it comes in the powerplay, the number is not his fault—it is his job. So I write the format beside every statistic. Test, ODI, T20—these are really three different games. Success in one format cannot predict the future in another. Yet that is exactly what we do constantly: after a flashy T20 innings we say, “He is ready for Test cricket.” Behind that decision there is no sample, no verification. In the spring of 2026 I hand-coded 81 matches for a research project—crowd presence, referee decisions, stoppage time. The result was plain, and the sample was small: with empty stadiums, the home team's win rate fell. The effect was modest, and for that very reason I considered it worth trusting—because it was verifiable and free of romance. That habit has seeped into my writing: before blaming an individual or an emotion, I look at what the conditions are saying. Another habit of mine is deliberate delay. Sitting down to analyse, I do not rush to a conclusion; instead I hunt for evidence against each claim. Suppose a team's pressing has dropped—I check whether it is a tactical change, or the pressure of the schedule, or the fatigue of the bowlers. Seeing a relationship, I do not assume causation. A spreadsheet spots patterns easily, and cricket tactics easily misread those patterns. My biggest lesson came from a minute-counting ledger. In a major tournament one team played four consecutive matches into extra time, while another played normal lengths. I hand-counted every minute and wrote it down—450 versus 360. Fatigue is not a feeling; it is a calculation. That calculation taught me how a long schedule quietly decides outcomes. In cricket the lesson is sharper: back-to-back Tests, travel, absence of rest—these sit in some cell of a spreadsheet and silently change the result. This reckoning of time separates two kinds of teams in cricket—one sees rest as luxury, the other sees rest as an asset. For smaller cricket economies the problem is keener, because they have fewer analysts, less tracking, and less opportunity to rest. In the same schedule, when a big team enjoys the advantage of rotation, a small team plays back-to-back matches and tires. That inequality does not show on the scoreboard, but it is clear in the spreadsheet. The same inequality appears in youth development. Big clubs now see prodigies from small leagues as “satellite assets”—rather than growing them in their own home environment, they buy them quickly and send them out on loan. So a young player grows outside his own setting, and his foundation stays weak. I do not hate this system; I simply keep the account—who gets how many minutes on the field, and who gets how many on the bench. Still I stay cautious, because between correlation and causation there is a thin wall. When a team wins in a row, everyone calls it a “fairytale”; I check whether it is the luck of the schedule, the weakness of the opponent, or genuinely tactics. For me as an analyst, the most valuable moment comes when I prove my own favourite assumption wrong. If a model's output does not match my belief, I do not blame the model; I test my belief again. That is why the most dangerous moment for me comes when the information is partial. Given incomplete data, we often fill the remaining cells with our own imagination, and then pass that imagination off as analysis. Someone insists, “He is in the form of his life”—I want to know, by whose reckoning, at what time, on what sample. Popularity has no obligatory relationship with truth. A deal sold at a high price does not make a player great; it is just a column, beside which “confirmed” has not yet been written. My working style is slow, and I do not hide it. The most valuable part of analysis never reaches a headline—it is the methodology note, the list of sources, and the limits of the sample. This is why I say: data never decides anything by itself; the analyst decides, and if he becomes confident on incomplete information, the damage is enormous. Next season the county and Test calendars will grow denser, and teams will invest more in rest management. Then the question will change: who can decide quickly, and who understands—that sometimes not knowing is the most honest answer.