Immutable Error: The Dhaka Desk File Inside Cricket's Data Chain

I opened the Dhaka desk file, and the first column was already arguing with me.

The file belongs to 25 August 2026: Rawalpindi, the first Test, Bangladesh's maiden Test win on Pakistani soil. Three separate records of that match landed on my desk — the scorer app export, the broadcast feed's ball-by-ball log, and the sheet I wired by hand in Mirpur. Three sources, one match. In Mushfiqur Rahim's 191, the dot-ball count reads three different ways: 85, 87, 89.

A two-to-four ball gap looks trivial. By the end of an innings it shifts the second decimal of a strike rate, and that decimal gets quoted on social media the next morning. Data does not break in large figures; it accumulates in small disagreements.

That evening a colleague called. "If the data were on-chain," he said, "we would not be arguing at all." I put the phone down and looked at the dashboard. The dashboard did not shout; it quietly rearranged what I thought I saw.

Because the gap is not a problem of truth. It is a problem of collection protocol. Three people counting balls in three places; none of them lied. One tapped a scorer app, one took notes off a broadcast camera, one watched a stream in Mirpur. An immutable ledger would preserve all three versions forever. It would not reconcile them.

The three columns that fight each other

Every Test hides at least four data layers. The stadium scorer writes ball by ball. The broadcast tracker measures pace, line and pitch point from camera frames. A central stats provider builds the official scorecard. And the team analyst, lacking all three, runs his own sheet.

In Bangladesh these layers did not grow together; they grew one after another. The BPL began in 2026 and brought television production and early ball-tracking with it. The board's own archive and club-level paper have run in parallel ever since. The result: one domestic match can generate three official-looking datasets that do not agree.

Blockchain has walked straight into that gap. After 2026, cricket saw a wave of NFTs and fan tokens — platforms such as Rario and FanCraze, supporter votes modelled on football's Socios and Chiliz, smart-contract player payments, blockchain ticketing. The advertising line is always the same: because the ledger is immutable, the record is now trustworthy.

I read it differently. When the 2026 stadiums went silent, the home-advantage columns began to confess. Mirpur's edge did not shrink in an empty ground; it simply stopped being attributable to crowd noise. The advantage reappeared elsewhere — in pitch preparation, travel fatigue, familiarity with the ball. The data had been true all along. The noise had covered it.

Where the protocol breaks

Over six months I placed raw sheets from four BPL matches beside the official scorecards. The same four fault lines return.

The first is judgement. Leg-byes, wides and byes are one person's eyes. Whether a ball that clips the pad is a dot depends on who is scoring. One ball can shift a bowler's economy, a team's dot-ball percentage and a matchup model.

The second is timestamping. Which ball belongs to which over is stitched together from human memory and log times. Rain stoppages, Duckworth-Lewis recalibrations and review delays all soften the seam.

The third is the pitch, and here I refuse to compromise. Every sheet of mine carries four variables: pitch type (turning or two-paced), session, humidity, and rainfall over the previous seven days. Bangladesh's home Tests are decided disproportionately in the third innings and after tea, and the surface then is not the surface of six hours earlier. A ledger that records the innings but not the pitch preserves the number and loses the cause.

The fourth is the most uncomfortable: the revision log. A ball is corrected at eleven at night, the scorecard changes, and nobody knows who changed it, why, or against which source. Correcting a record in secret leaves the number accurate and the truth incomplete.

From ball-tracking I lean on false-shot percentage — the share of deliveries on which a batter misreads length or pace. It tells you more than the scorecard, because twenty runs out and twenty runs survived are not the same twenty runs.

On my own desk I keep an index I call the Dot-Ball Pressure Index, or DBPI: dots plus forced false shots in an over, divided by the pitch-type baseline for that over. At Mirpur's two-paced surface the DBPI rises for home bowlers; at Chattogram and Sylhet it flattens. The number is true. It is still an indicator, not a verdict.

What the chain solves, and what it does not

A blockchain is a notary. It hashes, it timestamps, it links each block to the last. Rewriting one figure requires rewriting the whole chain, and that gets noticed. Immutability guarantees exactly that much.

So which cricket problem does that actually cover?

Fan tokens promise supporters a say in decisions. In practice clubs run public polls while the consequential questions — match fees, selection, contract length — stay behind closed doors. Cricket NFTs are the same story in another costume: the existence of a digital collectible is provable, its connection to the game is not.

Smart-contract payments are a genuinely useful idea. If a player's fee releases automatically when conditions are met, the middleman vanishes. But automation only works when the fixture list, the days played and the match records are already standardised — that is, when the third data layer is itself reliable.

Ticketing and integrity monitoring are where the advantage is clearest. A booking ledger makes duplicate tickets and supply chains auditable, and unusual trading patterns can be preserved immutably. Cricket's anti-corruption work leans on data discipline more than it did a decade ago, and that is where the most usable application sits.

My reading stops here. A chain does not make data true; it makes data irrevocable. And irrevocable data, when wrong, is wrong in a way nobody can ever wash out. Suppose the Rawalpindi ledger says 87 dots. The chain will guard it. It will not ask why the Dhaka sheet says 89.

The row that refuses to fit the story

Now let me warn against my own biases. Organisations that adopt data tools usually hire analysts, buy camera rigs and sign tracking contracts in the same package. Decision quality improves — and the credit goes to the tool. Correlation is not causation. If a broadcaster keeps an innings-by-innings notebook and publishes it the same night, that is not the benefit of immutability; that is newsroom persistence. The analyst cohort moves outcomes far more than the ledger, and it never appears in the layer diagram.

The second discomfort is the colonial reflex. I was born in England and trained on European data models, and my early discipline was built inside that grid. But roughly a third of Bangladesh's Test cricket is disrupted by rain, and on high-humidity days my sheets show bowlers in a completely different pattern. Imported models tolerate local baselines only when the baseline is stated out loud. A report that does not open with the Mirpur surface and the monsoon air is worth less for the rest of its length.

The third is governance. Blockchain's appeal is decentralisation; cricket's structures are centralised. If the board or league owns the data, it usually owns the ledger too — and immutability under centralised ownership simply makes the error permanent. Tamim Iqbal finished with more than 8,000 ODI runs, around 8,300, the most by any Bangladesh batter. No immutable ledger protected those runs from the politics around the team. Nor was Bangladesh's first Test win — by 226 runs against Zimbabwe in Chattogram in 2026 — ever written to a chain. The record survived through newsrooms, archives and stubborn desk workers.

A warning is due here. My professional risks are two: dashboard worship, because PPDA-style tables are clean, fast and authoritative; and forced counter-intuition, because this brand rewards the surprising finding. So I now write my hypothesis down before I write the piece, then try to falsify it myself. What this sheet gave me is not surprising: three sources disagreeing, as they did before and will again. A boring result is still a result.

What I will watch in the next round

From now on I want three things with every match report: the collection protocol, the full revision log, and the identity of the scorer responsible for each delivery. If a board launches an immutable ledger without publishing those three, understand that the paper has only been given a new cover.

My next file is from September, the second Rawalpindi Test, where Bangladesh chased 185 and won by six wickets. One row in that sheet does not match the others. I have learned to trust the row that refuses to fit the story.