The Empty Column and the Filled Illusion: A Null-Result Report from Cricket's Data Desk

At the old Eden Gardens block, the fan turned slowly that morning and the spreadsheet on the monitor held 4,180 rows — one row per delivery. Seam position, release angle, pitch-landing zone on a nine-box grid. An entire Ranji Trophy season tagged by hand, frame by frame, past midnight.

On the third day I noticed a column that was empty. The header existed — ‘false shots induced’ — and every cell beneath it was white. Nobody had complained for three days, because the populated columns were all familiar ones: wickets, economy, dot balls, boundaries. Nobody missed the one column that might have said something new.

Not one of those 4,180 rows explained the noise, and the noise was the reason I had been given a chair.

I count deliveries the way other people count regrets — quietly, past midnight. The habit is twenty-four years old, and its real benefit is not that I find answers. It is that I recognise, quickly, where the answer is not. That empty column was the most honest report in the building, and we had no vocabulary for it.

How the system gets built

In 2026, after a third unpaid season of camera work, the Cricket Association of Bengal hired me as its first full-time video analyst. The post had not existed before. Bengal cricket carries an old, slightly dynastic grievance about data, and its root sits in 2026, when the BCA–Mafatlal Bowling Scheme trained Abey Kuruvilla, Paras Mhambrey, Sairaj Bahutule, Salil Ankola and Nilesh Kulkarni under Frank Tyson. The fast-bowling tradition the association claims as its grand inheritance was in fact a planned technical intervention built on outside expertise. That detail never appears in official brochures.

My job was to measure the current edition of that tradition. Thirty-odd bowlers, one season, 4,180 deliveries, each coded for seam position (upright, inclined, cross-seam, wobble), release angle, nine-zone landing area, and outcome.

A data desk is built not by its questions but by its columns. The moment you fix the column, you have fixed what the world is allowed to show.

Wickets came first in the sheet but not in my evenings. A wicket is a lagging indicator: it records what happened, not what was happening. The useful moment was the half-inch of misjudgement — the batter committing to a high-angle ball that beat him without the edge. No runs, no wicket, no scoreboard change. And yet the same bowler comes back to that ball again and again.

So the ledger’s most valuable column was the process variable, not the outcome. When it finally filled, one name surfaced that no bowling programme had prioritised: Mukesh Kumar. His release angle sat measurably wider than the squad’s other right-arm seamers, his wobble-seam frequency clearer than anyone’s. He was not the trial’s bright star. In the sheet he was a row.

I wrote a 2,000-word newsletter piece on him. Nobody at the association read it. Some years later that name appeared in the national Test squad.

This is not a story about talent being spotted. It is a story about correct data that failed to become a decision.

Why the wicket is the wrong column to bet on

Cricket’s entire public machinery runs on outcomes, because the spectator needs an event every three minutes and the wicket is the only unit that measurably changes a match. But selection decisions should not be taken on the most visible unit; they should be taken on the most predictive one. For a fast bowler, wicket counts over three months are noisy — a dropped catch, an umpiring call, a pitch, and the average swings. False-shot frequency is the lower-variance measure of action quality and minor skill, the thing that persists from season to season.

India’s largest invisible deficit sits here: process data is scarce at the bottom of the pyramid and outcomes are abundant at the top. In Ranji, hundreds of bowlers play each year for a match fee and a scholarship, and nobody records who induced the most false shots. The video stops when the match ends. The next season the same bowler is no longer a hero at trials, and nobody remembers which month was his best.

From inside a coaching staff, one learns: telling a player he made 40 is easy. Telling him that eleven of those balls should have ended his innings is hard. The first motivates; the second improves. Our system is built entirely around the first.

The auction ledger

January 2026, Bengaluru, the IPL mega auction. Kolkata Knight Riders bid ₹7.4 crore for Dinesh Karthik and built a spin-centred squad around him.

Every auction paddle goes up in seven seconds. Behind it sit thousands of rows, some of them blank — and the blank ones are the least-read part of a player’s season. Every paddle rises before the logic arrives; my job is to arrive earlier.

₹7.4 crore is not a certificate of Karthik’s batting ability. It is the market-clearing price of the role Kolkata needed to fill: keeper-batter, middle order, and a face the home crowd would trust. In market language, it priced the gap they were brave enough to admit, not the man they were buying.

This is where cricket journalism errs most consistently — treating price as proof of performance. Two-thirds of auction bids resolve pre-identified role gaps; the remaining third is spent when a franchise buys a second weakness to hide the first.

And there is a question the entire IPL economy has left unasked: when a ₹7.4 crore call is made, how often can the analysis underneath it say ‘we do not have enough data’? Empirically, never — because that sentence produces a null result, and in cricket the null is the worst-paid product in the market.

The broken link: right report, wrong room

25 May 2026, Eden Gardens, IPL Qualifier 2. Sunrisers Hyderabad 174/7, Kolkata Knight Riders 160/9 — a 13-run defeat. Rashid Khan took 3/19.

My brief carried one line I have not forgotten: in the middle overs Rashid’s googly frequency was no lower than his leg-break, and his real weapon was that batters could not read length. The recommendation was to score before he arrived and reduce risk against him from the seventh over.

In the team meeting that recommendation was an eight-line bullet sheet. The coach had ninety seconds, the bus was waiting, and the one sentence a player heard became his truth. The plan asked batters to attack him anyway, because courage was the story the room needed. Rashid stopped at 3/19.

The largest failure in analysis is not the model; it is transmission — turning a correct line into the dressing room’s language.

We analysts produce PDFs, send them to each other, and consider the job delivered. Delivery happens only when the information changes one drill, one field setting, one bowling change. If it changes nothing, the analysis has not entered production.

A Russian summer taught me that weather is a variable nobody puts in the model. July 2026, the World Cup, France’s pressing triggers: the moment a defender opens his body to play infield, three players step two yards forward together. I argued the same logic applied to a T20 ring field — start the circle two metres closer and the batter’s reading time drops by a third. I wrote it. It was published. No coach called.

That was not rejection. It was the absence of a language in which two metres between runs and wickets could be spoken.

The bigger problem is selection, not transmission

Indian cricket’s most repeated error is the claim that talent is scarce. The opposite is true. India’s second XI is the hardest team to make in world sport. Yet the surplus is managed by incentives that steadily erode it.

The Impact Player rule has broken the all-rounder economy: if a team never needs batting at seven, why develop a player with both skills? Three seasons later the national side needs a fourth seamer who can bat ten balls and cannot find one. Ranji’s competitive meaning is thinning even as its performance meaning holds. And auction visibility has become a filter that correlates poorly with red-ball durability — not because the auction is wrong, but because it answers a different question and we keep reading it as an answer to ours.

The reversal

Seven times I have caught the question early, months before the broadcast. Seven times builds a habit, and the habit is a trap. The biggest risk in this profession is starting to hurry toward the answer you have already found once.

One season I refused the temptation and wrote plainly: our pitch mapping showed no clear difference in wobble-seam frequency, and what we measure may not be what matters. That report got no meeting. A null result ends the conversation.

With rare exceptions, negative findings are never published, so the data is never edited, and a flattering archive accumulates in which everything we ever said turned out right.

The consequence is quietly dangerous: an unfilled column and a filled column showing no effect look identical on a screen. Both are empty cells. They are two different diseases, and our formats have no separate words for them. So every column in my team now carries three states — filled, ‘no effect detected’, and ‘not measured’. The third is the one that gets re-checked.

Broadcast returns the same disease in another face: the fifty-over milestone reflex. I remember a hundred built at a strike rate that let the side hit one boundary in the final ten overs, and the post-match question was whether it was his finest innings, not what the team needed from him in those ten overs. A milestone reflex gives you a trophy and a systemic blind spot.

And my own habit? I shared the successful reports and filed the failures in the corner of a disk nothing returns from. The truest line in my 4,180 rows was a blank cell, and it took me three days to see it.

What to check next

Three things, none of them visible on a scoreboard. First, how often batters misread length against spin in the middle overs — if that number rises, control improves before economy does. Second, the starting distance of the ring field: if a side brings the circle two metres closer and the powerplay run rate does not rise, the essay was right and what we called ‘it didn’t work’ was really ‘it was never measured’. Third, the second XI’s composition — a fourth seamer who can bat ten balls, or another specialist.

If selection always drifts toward the specialist, the surplus will always return as a weak fourth bowler.

Results are not lies. They are incomplete testimony, and the court is running on the witnesses who happen to be on camera. When the scoreboard says one thing and your sheet says another, which will you trust? The question is not how much data you have. It is whether your sheet still contains the one empty cell you have not yet spent two hours filling.