The Testimony of Empty Cells: Chain-of-Custody in Cricket Data and the Blockchain Promise

The question arrived one evening in the Rajshahi xG Circle. A member wrote: “Brother, how do we know the numbers we argue about so much are real?” I set down my cup of tea. I have spent forty-nine years sifting through cricket scorecards, yet the question stopped me. Because that same day I had been staring at an analyst's worksheet whose one column was entirely blank. The numbers were not wrong — there were no numbers. The data never arrived. And those empty cells showed me how quickly an analysis can turn false. An empty cell is never neutral; it either speaks the truth or invites someone to invent it. That day I decided to write about those empty cells.

Cricket data does not fall from the sky. Every ball has a birth certificate. A scorer sits in the stadium, noting runs, balls, no-balls and wides. Ball-tracking cameras around the ground measure trajectory, bounce and spin revolutions. A vendor then buys that raw information, cleans it and pours it into a feed. Television, fantasy sites and newspapers all eat from that same feed. From the raw ball to the graphic on screen, the data changes hands at least four or five times. At every handover something is lost, something is altered, something falls into the wrong box.

In 2026, covering the Wills Cup in Dhaka for Prothom Alo, I first understood that the scorecard and the event on the field are not the same thing. The book said “four”, but I had watched a fielder stop the ball, only for the throw to reach the boundary — so was that run from running, or from the rope? The scorer decided one way, I had seen another. Back then I thought it was a small matter. Now I know it is enormous. Because the entire future of an analysis rests on that single cell.

The problem is not only the scorer. Suppose a dataset arrives labelled “Asian cricket”. But what is inside? Test, ODI, T20 — which one? Which season? Which ground? If the label is wrong, the analyst receives a file with a title but no body. And if someone draws a conclusion from the title alone, that is the greatest danger of all.

I can say with my eyes closed that I have seen many such empty files in my life. So the question is not whether data goes wrong — data will go wrong. The question is whether we can catch the error, and whether we can keep our mouths shut until we do.

Layer one: the empty cell. In an analysis, the most dangerous thing is not a wrong number but a missing one. A wrong number at least catches the eye. An empty cell sits quietly, and the reader assumes the writer deliberately left it out. I made this mistake in the early days of the Rajshahi xG Circle. In 2026 I logged the xG of every Real Madrid goal in their Champions League run. Cristiano Ronaldo scored 12 goals from an xG of 10.1 — a difference of +1.9. In my first post I gave only that one number and left everything else blank. Three hundred comments arrived within a week. Some said Ronaldo was clutch; others said it was pure luck.

Those three hundred comments taught me that an empty cell never stays empty — the community fills it with its own stories. Ever since, I open every analysis with a question from the group, and I write a “what the fans saw” section before the numbers. Before the table speaks, let the sample size breathe. I still do not break that rule.

Layer two: identity conflict. Cricket has many players with the same name. Two bowlers called the same thing, three men sharing a surname. When a vendor's feed conflates two of them, one player's runs leak into another's record. This error raises no flag, sends no warning. One number simply shifts a little. Yet from that slightly shifted number someone may decide whom to pick. Identity resolution, then, is not a luxury for an analyst but a discipline.

Layer three: the wrong letter in the wrong box. A dataset carries a label — what kind of cricket, which region, what sort of information. If the label is wrong, then even when everything inside is correct, the analysis lands in the wrong place. Running Test data through an ODI framework is wrong, and dropping limited-overs data into a Test sample is wrong too. I like to sit the eye test and the model together, but the eye test and the model must sit together in the right room, or both go blind.

I have a simple test. When a dataset reaches my hands, I ask three questions first. Who wrote it, and who checked it? How much sample, over how much time? And most importantly — what is not written? The third question is the hardest, because missing information does not speak for itself. From years of watching cricket at the ground, I can say the scorecard rarely lies, but it is very often incomplete. Nothing is more dangerous than an incomplete truth.

On the night of the 2026 World Cup final in Russia, I was live-posting France's pressing data. France beat Croatia 4-2. I wrote that France's PPDA was 14.3, and that N'Golo Kanté had covered 6.9 kilometres before his 55th-minute substitution. Three hundred comments again. The argument rose over whether Kanté was overrated. I realised then that raw statistics do not land without a human story. The Kanté question was never about one man; it was about how we measure quiet work. Placing PPDA and distance covered side by side, I showed that the man who barely touches the ball does the most work in midfield.

And then May 2026. The Bundesliga returned, but the stands were empty. I kept track of home advantage. Before lockdown, in the 2026-20 season, the home side was winning 43.3% of matches. Over the first three empty-stadium rounds, that rate fell to 33.3%. The number was clear. But the number alone said nothing. When the stadiums emptied, the numbers confessed something we had ignored. Many in the group said they felt isolated. So I organised Zoom watch parties for twelve fans. The data said the game had changed; the community said we needed connection. Two truths — both equally important.

Now to the blockchain. In the world of cricket data, the real promise of blockchain is not generating numbers but preserving their birth certificates. Once each data point — which ball, who logged it, when, which vendor — enters a block, it can no longer be quietly altered. What is the gain? The gain is that when an argument erupts over a suspicious run-out or a disputed xG, we can prove what the original feed said. The chain of custody of the data becomes visible.

But there is a clear limit here, one that a cautious man like me must repeat every day. Blockchain makes data immutable; it does not make data true. If the scorer recorded an error and that error is welded into a block, the error sits there sanctified forever. Immutable does not mean accurate. A ledger is a mirror, not a microscope. Verifying the source and verifying the result are two different jobs.

Here is my contrarian angle. In conversations about data integrity, we easily assume that blockchain, or any decentralised system, solves every problem. I do not accept that. Connectivity is not truth, and disconnection is not falsehood. Blockchain gives evidence about data, but it does not judge. Take a small sample — three matches, forty balls. Even if it is stored flawlessly in blocks, you still cannot conclude that “this bowler is in form”. Blockchain does not enlarge the sample. The reverse is more dangerous: an immutable wrong data point makes people more confident in a wrong decision, because it carries a “verified” stamp.

I have seen a World Cup rewrite what we thought we knew. Russia 2026, the empty stadiums of 2026, Ronaldo's xG in 2026-17 — each event taught me the same thing: however solid the structure of data, humans do the interpreting. Blockchain can strengthen the structure; it cannot take over the responsibility of interpretation. Anyone who thinks the technology will make the judgement for us is on the wrong road.

What helps more are three simple habits. One, write the sample size beside every number — how many matches, how many balls, how many seasons. Two, when you suspect an error, say it openly rather than hide it — I do not know yet, I will check tomorrow. Three, keep an alternative explanation beside every claim — for instance, home advantage is not only about the crowd; it is also about travel fatigue and the umpire's psychology.

Rajshahi taught me that a circle of analysts can be a sanctuary. We do not work like a secret commission; we catch each other's errors. When a member of the group asks, “Where did this number come from?”, he is not insulting anyone — he is protecting everyone. This habit of catching errors is the real blockchain — a human blockchain.

My conclusion comes with limited confidence: the integrity of data begins with technology and ends with habit. Blockchain can provide a strong foundation, but the person standing on that foundation must be answerable. With an empty cell in front of us, the next question should be: which piece of data do we now retire, and which do we re-examine? Because no system will tell us that — we have to say it ourselves.