Ball-by-Ball Audit Trails: Why Cricket's Ball-Tracking Data Now Needs Blockchain-Grade Provenance

The ball-tracking projection stayed on screen for roughly three seconds at Nassau County Stadium in New York. Then the feed cut back to the live camera, and the impact point — a few centimetres inside off stump — existed nowhere. That day I was not thinking about the verdict. I was thinking about the question. If a projection for a single delivery disappears three seconds after it appears, who verifies that delivery five years later, when the same bowler's career has to be reconstructed? Cricket's data economy has not answered that. In football I was used to a system where the number was not mine, but the number survived. In cricket the number does not survive; only its photograph does.

When I started the blog "Expected Truth" in Rajshahi in 2026, the ambition was modest: every match claim should be testable afterwards. After Abahani Limited Dhaka beat Sheikh Jamal Dhanmondi Club 2-0 in the Bangladesh Premier League, I worked out an xG of 1.4 against 0.6 and a PPDA of 8.2. The scoreline flattered Abahani. That thread reached twelve thousand readers and was quoted by a Dhaka sports outlet. It taught me one thing. Football handed me a number that anyone, sitting anywhere, could rebuild from scratch. Cricket handed me a projection graphic that vanished after three seconds.

Treating the data layer as one object is a mistake. At least three layers run at once. Layer one is the sensor: stump cameras, Hawk-Eye or Virtual Eye, UltraEdge. Layer two is the model: trajectory fitting, spin estimation, pitch-bounce variables, the probability attached to an impact point. Layer three is presentation: the small animation that floats on screen. Layers one and two sit inside board and broadcaster contracts; layer three sits in front of everyone. The viewer sees layer three. Trust breaks in layer two, and nobody publishes layer two. In 2026 Nassau County was a brand-new venue, with drop-in pitches shipped from Australia and a camera rig that matched neither Lord's nor Eden Gardens. If the calibration file is not version-controlled, who declares the reliability of a projection at a new venue?

This is why blockchain enters my argument as infrastructure rather than crypto excitement. Cricket's problem is not a wrong verdict; the problem is that the evidence behind the verdict is destroyable. An append-only ledger does nothing miraculous. It simply says: this impact point was produced by this camera configuration, this calibration version, this software build, at this timestamp. Once that is known, a dispute stops being a contest of belief and becomes a contest of measurement. That contest has not even started in cricket.

I stopped watching deliveries and started reading the space before them, because that is where the value of a decision hides. On 10 June 2026 in New York, Bangladesh lost to South Africa by four runs. My notebook holds the result of that match and no proof of any single delivery inside it. Across the tournament, every review I watched followed the same structure: the projection appears on the third umpire's screen, the decision is announced, the feed moves on within seconds, and the data is gone. If I want to write about that delivery a month later, all I own is memory.

Umpire's call is not a decision problem; it is an unpublished variance report. Every supporter knows the verdict wobbles when the ball clips half the stump. How much uncertainty, in which direction, and why — none of that is released. If the model published a confidence band per delivery, the probable range of the true path, umpire's call would stop being a mysterious ruling and become a probability: inside the band, the decision stands; outside it, the decision falls. Without a published band, cricket treats its audience as people who must accept an overruled decision while being refused the evidence.

Calibration drift is the more uncomfortable question. The same bowler hitting the same length in Chennai may produce a trajectory that does not match Durban, because camera height, angular distance, light and pitch friction differ at every venue. Football has fewer venue-level variables, since the ball travels on grass in a consistent direction. In cricket the ball changes direction once it hits the pitch, which makes the sensor's job far more fragile. A system that does not publish its own calibration version cannot prove that it erred, and cannot prove that it did not. If both are impossible at once, the output has stopped being data and become belief.

In Bangladesh the stakes are higher. Most venues hosting the BPL and the Dhaka Premier League carry no ball-tracking at all, so scouting runs on video, scorecards and a coach's eye. That information vacuum has a quiet economic consequence: whoever holds the private ledger — board or franchise — is the one pricing a young player's future. When an agent in England, Australia or India negotiates with a proof file in hand, our domestic players sit down with handwritten notes. The transfer market is a story the market tells about its own fear, and in a market without evidence, the monopoly on telling that story belongs to whoever owns the sensor.

Here the cross-sport translation earns its place, not as decoration but as an argument-changer. In football, xG entered the public domain; Opta and StatsBomb publish their model documentation. Baseball went further: Statcast pitch tracking is open, and pitcher evaluation was rebuilt because of it. Both sports saw the same shift — when the data opened, the language of decision-making opened with it. Cricket's ball tracking never walked through that door. That is why arguments about impact points in international cricket are not arguments; they are embarrassed silences.

Let me bring my own model into the dock, because there is something to confess. After 2026 I began transplanting football's pressing vocabulary into cricket, treating PPDA-style pressure as a proxy for dot-ball pressure, and I assumed Bangladesh's slow middle overs came from pressure. At the 2026 World Cup the assumption failed. The dot-ball rate was high, but the cause was risk management, not pressure. Batters were declining risk deliberately because a wicket would worsen the position. My model could read the situation; it could not read intent — and the T20 dot ball remains the variable where the model is most blind. I am timestamping that admission so nobody later rewrites it as foresight.

So what would a blockchain-anchored structure actually do? Three things. First, timestamp every telemetry frame at the moment of release and anchor its hash in a ledger, so that any later alteration of the record is detectable. Second, publish a Merkle root per over, so that a full over's data set can be verified as a single number. Third, version-control the calibration file, so it is knowable which sensor configuration produced which projection. These three steps do not prove that data is true; they only make the history of the data unalterable. An unalterable history at least narrows the room for fraud.

Let me state the limit clearly, because data is a monastery: you sweep the floors before you see the vision. The oracle problem is old and unresolved. If a sensor generates a wrong number, the ledger preserves it perfectly — forever, immutably, with flawless fidelity. Immutability is not accuracy. Blockchain will give cricket durable verdicts, not true ones. The real reform is model transparency: a published calibration report per tournament, and a right of re-adjudication when later data proves a ruling wrong. Without those two, a ledger merely immortalises old errors.

The contrarian point sits here. Everyone assumes transparency produces fairness. My reading differs. Transparency does not produce fairness; transparency produces better arguments. And cricket's resistance is commercial, not technical. Ball-tracking data now sits on the asset side of a board's balance sheet. Whoever releases it releases market power. Remember also that a blockchain bolted onto centralised ownership will dress centralisation in decentralisation's clothes and strengthen it. The World Cup did not create value; it simply turned the lights on. The value the 2026 New York leg revealed — the South Asian diaspora market — had existed for years, unseen because nobody had switched the light on. Now the question is accounting: whose ledger holds that light, and who can switch it off.

The signal is patient; the noise is always in a hurry. Next round I will not count how many reviews happened or how many umpire's calls were made. I will count one number: thirty days after a tournament ends, how many ball-tracking projections can still be retrieved in public. The current answer is zero, and my confidence band around that figure is narrow. The day that number rises above zero, cricket's data economy changes — and that will not be cricket reforming itself, only cricket doing its accounts.