The Arithmetic of Empty Cells: The Invisible Risk Inside Cricket Analytics

Around ten at night last month, at my desk in Rangpur, I opened a dashboard. I had three hours before filing a match preview, and I needed at least one trustworthy number before anything went to print. The spreadsheet loaded — expected runs, progressive-shot rate, powerplay economy. The columns were empty. The data feed had collapsed at eight that evening, and somewhere along the pipeline, nobody had noticed.

I sat staring at those blank cells. My problem was not that numbers were scarce. My problem was that I had never learned to read the language of an absent number. I had been taught that an empty cell means there is no information. Nobody taught me that an empty cell is itself a piece of information — proof that a process broke somewhere.

This piece is about those empty cells. The faster cricket's data economy grows, the faster its invisible risk grows with it. And that risk does not live on the pitch. It lives inside the pipeline.

Cricket is no longer just a game of bat and ball; it is a data-producing industry. Every delivery generates five to seven events — the line, the length, the batter's footwork, the fielder's position, the bowler's release point, the ball's speed, the bat's angle. An ODI contains roughly two thousand legal deliveries across two innings, meaning more than ten thousand events per match. T20 franchise leagues play dozens of matches a week, meaning hundreds of thousands of data points weekly.

That data now functions as collateral in three markets at once. Selection — who plays, who is dropped. Commerce — what a franchise pays for a player, what a sponsor or broadcaster commits. Entertainment product — fantasy leagues, broadcast graphics, social-media threads.

I have watched the inner arithmetic of cricket for ten years. During the 2026 World Cup in Russia, I was a first-year student. My classmates argued about passion and momentum; I opened Excel and counted a midfielder's progressive passes. That 900-word analysis drew four thousand readers — more than my entire department published that month. Since then, I begin a match preview with a single number, never with narrative colour.

But beginning with a number comes with a condition I did not understand then. The number must be true. And whether it is true depends on the pipeline behind it. In 2026 I interviewed Soumya Sarkar as a Daily Star reporter; the piece was later picked up by Prothom Alo. Back then I thought a reporter was a collector of stories. Today I understand a reporter is a keeper of sources.

Modern cricket data is built in four stages. Collection — scorers, tracking cameras, hawk-eye systems. Cleaning — removing duplicates, fixing missing values, reconciling team IDs. Modelling — expected runs, death-over projections, player valuation. Decision — selection, price, strategy.

Most discussion happens at the final stage. People talk about outcomes, not process. But the risk hides in the cleaning stage, because the work between collection and modelling receives the least scrutiny.

Consider a six-match T20 series. In the third match the stadium loses power and tracking dies for ten minutes; nobody fills in those overs. In the fourth, a new scorer spells an old team differently, so the database now holds two team IDs for one team. In the sixth, a bowler's delivery-speed sensor loses calibration, so every ball reads four kilometres per hour faster.

These small gaps accumulate and can change one player's valuation. A bowler's true economy is 8.2, but drop the missing over and it reads 7.4. The selection committee sees 7.4 and picks him; nobody knows where four overs of data went.

So my first rule: before any decision, measure the completeness rate of the data, not only its quality. Whether a column is full matters as much as whether it is accurate. If 12% of ball-by-ball data is missing, I do not call a model built on the remaining 88% a basis for decision. I call it a decision risk.

One thing needs clearing up. The spreadsheet did not vanish. It moved to the screen. A coach once kept a ball-by-ball ledger and, seeing an empty cell, trusted his own eyes. Now a coach reads a dashboard where every cell is full, coloured, confident. The empty cell is erased — the system inserts a zero, or last match's average, in its place. The system does not lie; it goes silent. And silence is more dangerous than a lie.

In late 2026, when global sport shut down, I was a second-year student. My stadium-atmosphere thesis collapsed. Instead I built a financial model for fourteen clubs — matchday revenue loss from empty stadiums, hospitality, merchandise. I calculated one club's wage-to-revenue ratio at 74%, a figure that later proved terrifyingly accurate. I sent it to five sports editors. Three never replied. One ran it as a guest column in a regional business daily.

That experience taught me the biggest enemy of data is not missing data — it is false confidence in present data.

Now look at this in the Bangladesh context. Domestic cricket here still collects data largely by hand. A Dhaka Premier League match may have two scorers, one per innings. Tracking-camera density is nowhere near the IPL's. Decisions here are often made on visual memory and scorecards — average and strike rate. Meanwhile international franchise leagues decide using pressure index, impact index, bowling matchup matrices.

So what happens? A player performs brilliantly at home, but his data is produced on an incomplete pipeline. At an IPL auction, that sells him short. A gap in information does not only ruin analysis, it ruins the market — a player's price is set not by his skill but by how much data was collected about him.

That one line is the foundation of my entire career. When I read a scouting report for a franchise, I do not only check how many runs the kid scored. I check who built his data, from how many matches, how much of it is missing. Because I know the missing portion is his real risk.

I still open a spreadsheet before every match. My rule is simple. To write a number, I must know where it came from. Who tracked it, on which system, how much data was lost. If I cannot name the source, I do not write the number.

In January 2026, I was a junior finance analyst at a Bangladesh Premier League club. The board wanted a 31-year-old foreign striker on an annual $180,000. I ran the numbers. His goals-per-90 had fallen 40% across two seasons, and the deal would breach the league's salary cap by 8%. I proposed a domestic alternative — 24 years old, 0.67 per 90, at 60% of the cost. The board approved my recommendation within twenty minutes.

Those twenty minutes were my career's biggest lesson. The club did not agree with my analysis; it agreed with the sourcing behind my numbers. I had not merely said the target is 0.42, the alternative 0.67 — I had said by which source, at which time, on how much data. Decisions do not change on beauty; they change on transparency.

This is why I keep a source-redundancy protocol. Any major decision needs three independent data streams. Editors call it paranoid. I call it prepared.

Why three? Because at the 2026 Qatar World Cup I nearly got burned. Forty-eight hours before publication, my primary source — a stadium construction worker — withdrew in fear. I had no backup. Instead of scrapping the piece, I cross-referenced FIFA's own sustainability reports against three NGO datasets and built a timeline of contractual violations. The 2,200-word story was my first nationally syndicated report.

Since that night my rule has held: a source who vanishes leaves a trail of questions you should have asked. I now arrange every number so it cannot stand alone — I verify it from at least two places. In cricket that means checking a player's average against his strike rate, and his strike rate against his pressure situations.

One lesson is the most valuable of all. I learned more from the missing columns than from the final report. The report tells me what happened; the missing column tells me where my blind spot is.

As the fantasy-sports and broadcast-graphics market grows, dependence on data grows with it. A fantasy app shows hundreds of thousands of users a projection every day. If that projection is wrong, nobody dies, but a market's trust erodes. When a broadcaster puts a bowler's death-over economy on screen as the league's lowest, that number is a pipeline's liability. If it is wrong, it is not a lie — it is negligence.

And here is my most uncomfortable conclusion. We assume empty data is bad and full data is good. In the cricket market the reverse holds. A full, clean, colourful dashboard is more dangerous, because it manufactures confidence — and confidence wants belief, not questions.

When a model is 100% full, the user stops asking where that over went. An empty cell would have prompted exactly that question. An empty cell makes us cautious; a full one puts us to sleep.

The largest signal from an empty cell concerns system health. If 15% of a tournament's delivery-speed data disappears, the problem is not that 15% — the problem is a broken instrument, camera, or training process in the pipeline. If we merely erase the empty cell, we erase the broken instrument with it. This is why half my writing is about empty cells.

One more thing I am obliged to say. An analyst who hides an empty cell and inserts a guess is not analysing data; he is inventing a story. In the cricket industry this is the most unlawful act of all, because his story is priced in the market — a player pays for it in being dropped or bought. Based on my years of watching matches, I can say this: where the eye stops, data begins; and where data falls silent, the liability is heaviest.

So my position on empty columns is simple. If I write a number as the basis of a decision, I write more than the number — I write its source, its gaps, its limits.

The question is not really mine. It belongs to league owners, broadcasters, selection committees. When a name is finalised at the end of your data pipeline, do you know how much data was lost behind that name? If you do not, you are not judging a player — you are betting on an empty cell.