Empty Cells, False Calm: The Invisible Workers Sitting Inside Cricket's Data Pipeline

Five-forty in the morning. The gallery at Barishal Divisional Stadium is still asleep. Two groundstaff are dragging the pitch covers — one with a cup of tea, the other with a folded scoresheet in his pocket. The match is six hours away, yet the scoresheet has already gone to work: toss, names, order, roles, every cell filled in advance. Nobody knows the man's name. The broadcast camera never turns towards him. And yet every serious decision in the tournament eventually stands on those cells.

Last week a data report landed in my hands with almost every field empty. No headline, no information points, no named entities, no time-sensitivity assessed. At the bottom, one cold line: insufficient information, cannot assess. Some of the people who saw it read it as: nothing was found, so there is no risk. That misreading is what this piece is about.

Cricket now manufactures numbers by the hour. Bowling economy, strike rate, powerplay runs, death-over economy, drop-catch conversion, run-out conversion. Auction tables price players in the language of these numbers. Television graphics, fantasy points, selection committees — all run through the same pipeline.

The mainstream view is simple and, at first glance, correct: more data means better decisions. Every franchise unit hires analysts; no squad is announced without ball-by-ball coding. Over a long season, a number that repeats is more reliable than an eye that remembers — because you cannot read a pattern without repetition, and repetition can only be measured if someone wrote it down. I accept that argument.

My forty-eight years of watching from the ground tell me the gap is not in the numbers. It is in the room where the numbers are made. Analysis usually runs in two stages. Stage one breaks the raw material into information points — which date, which figure, which entity, which quote. Stage two builds tactics, form and risk out of those points. If stage one returns empty, what stage two writes is not analysis. It is a framework. And in that framework every cell receives the same sentence: cannot assess.

That is where the real danger sits, because an empty cell is never neutral in cricket.

An empty cell and a zero are not the same thing — yet almost every cricket database stores them in the same box. Picture two bowlers. One did not bowl a single ball. The other bowled and took no wicket. Both have zero wickets. About the first you know nothing: injured, out of the XI, or simply not a bowler in this format. About the second you know something specific: in these conditions, against these batters, in these overs, he could not take a wicket. One means absence. The other means failure. In the database, both are 0.

Batting has the same trap. A batter who faced no balls has a strike rate of zero; a batter who made 20 off 30 has a strike rate of 66.6. One is a blank, the other is a verdict — and on the selection spreadsheet they sit side by side.

In my press-box years nobody caught this, because the record lived in a hand-written book, and the person writing it knew exactly which cell he had left blank. Today the blank cell enters the pipeline and quietly becomes a number, and nobody asks who left it blank.

I climbed down from the press box to the balcony and found the crowd had sharper eyes. The people in the stands know who dropped the catch at third man, who stopped chasing towards the boundary, who dived at the rope and opened his knee. They have no spreadsheet, and they do not lie.

Nobody keeps a ledger for the invisible workers of the data pipeline, yet every cricket number begins with their hands. The scorers I know in Barishal sit up the night before with the squad list, because the online spelling is wrong. Some keep two books side by side — one for the office, one for themselves — because if the two disagree, the blame lands on them.

In 2026, when the grounds went quiet, I ran a twenty-four-hour live stream called Barishal Virtual Terrace. Twelve retired players and eight women journalists came on camera, and together we raised 220,000 taka for forty-five stranded groundstaff at Barishal Divisional Stadium. When the stadiums emptied, the sixth fielder turned out to be all of us.

But how do you measure that money? It never appeared on a graphic. Those forty-five names are not in any player index. The people who pull covers, paint lines, change the scoreboard — they live outside the data. And their work enters the data without their names attached.

From this a larger claim follows. A cricket system that does not record the names of its own scoring workers will also lie about its players — because both records are written by the same hands.

The next layer is subtler. Suppose an auction-season bowling-depth index is built. Bowlers from a big city academy have four balls per over on video, pace measured, seam movement measured. A tape-ball bowler from a mofussil town has a name and two second-hand stories. The index seats them side by side, and into the small-town boy's cell drops a zero. That is not the zero of his ability. It is the zero of the system's blindness. The index does not record the difference.

I have walked into this trap myself. In 2026, during the Russia World Cup, I hosted fourteen watch parties in Barishal and argued in a video that France's 4-2 win was decided not by Croatian fatigue but by nineteen-year-old Kylian Mbappe's 65th-minute goal and his four goals in the tournament. The video travelled. But I forgot to credit my editor. I had to publish a correction, and from that day I built a checklist: name, date, source, credit, cross-check. A checklist is not a humiliation. A checklist is a substitute for memory.

One line on that checklist applies directly to the report in front of me: when you write a zero, ask whether it is a zero or a blank — and if it is a blank, put a flag on it in the pipeline so nobody counts it as neutral trend data.

Now the metrics. Cricket runs on expected runs, win probability, impact scores. The problem is not that these metrics exist. The problem is what they claim. A strike rate of 145 looks superb — but 145 in the powerplay against the new ball and 145 in the nineteenth over against a part-timer are two different sports. A model that does not know ball type, field placement or injury news builds its error silently, and the error is never signed with its name.

Strike rate, economy, expected runs — all measure process, not decisions. Who moved the fielder, who changed the bowler, who tore up the plan after the toss: none of that enters those numbers.

One name sticks in my radar: a nineteen-year-old data kid in Barishal who keeps his own book of a mofussil tape-ball tournament — who bowled how many overs, how many runs in which over, who dropped whose catch. He has no software, only a phone and a pen. But his book holds something the big platforms do not: when he does not know a cell, he leaves it blank and puts a question mark beside it.

A skateboard teen in an empty stand taught me what loyalty actually costs. This boy taught me the same lesson in harsher language: a system that hides its own ignorance by manufacturing numbers will betray its own players too.

Now to where I could be wrong.

It is possible the empty report was a routine pipeline fault and I am inflating it into a crisis of principle. It is possible automation is the answer — video tracking, chip-in-ball, Hawk-Eye — leaving less room for human handwriting and therefore less room for blanks. And it is possible I am over-weighting the labour story because the balcony view is emotionally seductive and hardship photographs well. That is an old weakness of mine.

Still, I will put the counter-fact in the same breath: the more technology entered, the more scoring controversy grew — DRS, no-balls, run-out frames, over rates, net run rate. Every new instrument added a new argument. The machine does not remove the blank cell. It moves it somewhere else, and nobody takes responsibility.

And if my whole argument is wrong — if the empty report really is neutral — then someone should answer one question anyway: who left that cell blank, and what are they being paid to sit there? Answer that and I will go quiet and rewrite my checklist.

In the next six months, across the coming auction and selection cycle, at least one franchise or board will make a decision that was actually built by reading a blank as a zero. That is not a gambling prediction. It is a testable one — because if you ask your own city's scorer for his book the night before a match, you will see which cell is genuinely empty, and who has already counted it.