The Immutable Error: Domestic Cricket's Birth Dates and BCB's Blockchain Ledger

At Khulna's Sheikh Abu Naser Stadium last season, a left-arm spinner bowled 32 overs in a first-class match, eleven of them in a single unbroken spell. When the scoresheet was filed, his age appeared in two places with two different values. On the BCB central registration list he was 18 years and 4 months old. On his school leaving certificate he was 22 years and 9 months. The gap was 4 years and 5 months. For a cricketer that gap is not an emotional matter but a measurement one. If age is a column in a dataset, and roughly 18 percent of that column is wrong, then every conclusion resting on it is wrong too. In the 2026 Under-19 registration pool, across 342 files submitted to the Bangladesh Cricket Board, I found exactly that rate: sixty-one files carrying two different birth dates across two or more official documents. The numbers were not lying; they were waiting for a better question.

The National Cricket League is the most played and least watched competition in Bangladesh. Eight divisions, twenty-eight first-class matches a season, nearly ninety overs a day. From 2026-23 through 2026-25, I hand-coded 84 of those matches, more than 3,100 ball-by-ball events. Complete digital scorecards survive for only 61 of the 84. The rest have not a single ball of footage. Cameras sit at Mirpur; they do not sit at the Sheikh Abu Naser Stadium in Khulna.

Mehidy Hasan Miraz, who came through Khulna Division's domestic seasons into the national side, or Taijul Islam, who bowls for Rajshahi Division in the NCL, both have documented pathways, which is why their peak curves generate little argument. The question belongs to the files that were never recorded anywhere.

In 2026, as one of three BCB advisors overseeing digital and media affairs, I was asked to choose between two proposals for domestic registration. The first was a centralised database, the second a permissioned blockchain ledger. I argued for the ledger, and the reason was audit, not cricket. Every registration event sits on the ledger with a timestamp and a hash; who changed which field, and when, stays in the log. Nobody can quietly repair a scorecard after the fact. The pilot runs across five districts — Khulna, Rajshahi, Bogra, Barishal and Dhaka Metro. At the last verification on August 13, 2026, the ledger held 96 players and 1,412 entries.

Now the actual data. Discrepancy rates by district: Khulna 22.4 percent, Rajshahi 19.6, Barishal 17.1, Bogra 15.8, Dhaka Metro 8.1. The difference is not talent; it is paperwork. Where school registration and birth registration happen in separate offices, at separate times, in separate hands, a gap opens between the two documents — and at the moment of entry into an Under-14 or Under-16 squad, the temptation to close that gap opens with it.

The shape of the sixty-one flawed files says more. Thirty-eight of them show a gap of under twelve months: the one-year shift, the most common form. The mean gap is 1 year and 9 months. Six cases exceed four years, and four of those six were in first-class squads during the 2026-25 season.

What breaks when the column is wrong? The peak curve imported from international cricket — quicks peaking between 27 and 31, spinners between 28 and 33 — is applied to Bangladesh's domestic circuit almost unchanged. But if a squad's "nineteen-year-old" seamer is actually twenty-one, his rapid development is not rapid at all; it is delayed normal development that a corrupted column renders early. When I re-ran the peak window for twelve frontline spinners using corrected ages, the mean peak age fell from 27.4 to 25.9. The spike got spiked, but the pattern stayed in the data.

Workload lands in the same place. On corrected ages, twelve leading NCL spinners bowled an average of 4,200 overs before turning twenty-one, more than a quarter of them in spells above 25 overs across four-day matches. That is a twenty-four-year-old's load inside a nineteen-year-old's body. When the age column is wrong, that load is recorded nowhere, because the damage surfaces later and the age surfaces first.

What a blockchain ledger does here is bookkeeping, not magic. Three things. First, every entry is bound to the identity of the person who made it. Second, no later correction can erase an earlier entry; it appends a new block with a reason and a signature. Third, and most valuable to me, an entry that should have existed but never did becomes visible. The boy in Khulna who was bowling well enough for an Under-16 spot but whose name is not on the ledger now occupies a negative result. In Khulna, I learned that silence is also a dataset — but holding that dataset requires a structure, which is what a ledger supplies.

This is where a false assumption enters, more damaging than any of the earlier errors. Immutability and truth are not the same thing. If the field is already wrong before it is hashed, the blockchain simply makes the error permanent. In the BCB pilot I therefore argued two clauses into the framework: a separate correction block carrying cause, document and signature; and at least one direct verification each year for every registered player. Because without a path to correction that does not erase, a ledger becomes a life sentence — and the sentence would fall on the boy who never changed a single date.

The second gap belongs to the system itself. The ledger holds only registered players. The boy from Batiaghata in Khulna, raised without a birth certificate, absent even from district school lists, whose leg-spin every ground in the neighbourhood knows by name, is not on the ledger. The ledger therefore carries the limit of its own completeness: it accounts for everyone inside the system and no one outside it. The dashboard then becomes the new heatmap — a polished interface concealing whose hand typed the entry.

What to watch next season is not block height. Watch two things. How many correction blocks get approved per 100 registrations — if the number is zero, the ledger is not protecting the players, it is protecting itself. And the distance between players on the ledger and players who turn up at registration camps, which across Khulna's five upazilas currently stands at 11 percent. Every model is a prayer until the data says otherwise. And this data says there is no point hashing a date that was never right.