Logged but Never Audited: The Empty Seat in Cricket's Data Chain

On the evening of 18 June 2026 at The Oval, I opened my notebook before I looked at the scoreboard. Pakistan had made 338 for 4 — Fakhar Zaman 114 off 106 — and Mohammad Amir's 3 for 16 had bowled India out for 158 in 30.3 overs. I was 44, split between a Lancashire coaching consultancy at Old Trafford and a Saturday radio shift, and I spent that final logging every Pakistan delivery by length alone: 300 balls, three columns, no sentences. When the log closed the next morning the number was plain. Fifty-eight per cent of those deliveries had landed in the five-to-seven-metre slot.

India's scoreboard was the consequence of that slot, not the cause of it. That sentence changed how I write. The graphics that ran on television that evening were correct by the standards of their own feed. What nobody measured was the gap between my log and that feed, because the instrument for measuring it did not exist. Three weeks later a fourteen-minute breakdown video of mine took 400,000 views in a week, and that Friday I resigned the consultancy. I did not stop coaching because I had lost. I stopped because I had begun watching only the spaces — and because I finally understood that the game's real record sits not on the screen but on the scorer's sheet that nobody opens.

Cricket's data is a product, and every product has a supply chain. An international delivery passes through at least six pairs of hands: the ground scorer, the ball-by-ball collector, the aggregator, the rights holder, the broadcaster, and finally the commentator who gives it a voice. At each hand something is added and something is dropped. The scorer records a length; he does not record how far that length was from the batsman's body. The collector drops it into a bucket. The aggregator turns the bucket into a percentage. The broadcaster sticks the percentage on a strip. The commentator turns the strip into a sentence. The newspaper turns the sentence into a headline. What survives to the seventh hand is not the game. It is a translation of the game.

One link is missing from that entire chain: audit. In any other manufacturing process you can demand a trace from raw material to finished goods. Cricket's numbers carry no such trace. When a graphic announces that a fast bowler's "length percentage" is 41, nobody is told how many deliveries sit behind that 41, which camera measured them, how many were hand-coded, and how many vanished because the camera was behind somebody's head. The viewer receives the number. The viewer does not receive the error bar.

The money sits in that gap. In the June 2026 e-auction, the Indian Premier League's media rights for the 2026–2027 cycle were settled at ₹48,390 crore. The data right inside that package is not published as a separate line item; it travels inside the bundle, under terms that do not leave the room. The consequence is direct. The body that owns the feed is also the final judge of the feed's accuracy. For a sport in which selection, contracts and analysis all rest on the same single input, that is an unusual arrangement.

At international level the picture looks better, because a camera is present. Hawk-Eye for DRS, front-foot ball tracking, release point and seam angle all exist wherever a broadcaster's budget exists. The same technology is absent from domestic cricket — and domestic cricket is precisely where selection decisions are made. No public ball-by-ball dataset exists for the Ranji Trophy. What exists is the scorecard: runs, wickets, overs, economy, average. Those numbers tell you how large a batsman is. They do not tell you where he was standing when the ball arrived.

That single asymmetry distorts the whole domestic selection process. A selector holds seven hours of match reel and a one-page card. A seamer's 28 wickets may be the story of twelve good days; a rival with 24 may have bowled four months without breaking a partnership. Both are scored on the same instrument, though their seasons were not the same reality. I drew the shape of that collapse three weeks before a domestic scorecard confirmed it — a seamer whose 71 per cent of deliveries landed outside sixth-to-eighth stump, whose economy read 2.8, and who was not taking wickets because batsmen had simply stopped playing him. The card called him ineffective. The card was not lying. The card had no language for space.

At Edgbaston in August 2026 I charted 1,240 deliveries across the first three Tests — release point, seam angle, pitch map. India lost by 31 runs, Virat Kohli made 149, and Sam Curran's 4 for 74 turned the second innings. One number surfaced in my chart. India's seamers were landing 6.4 metres from the stumps on average; England's were at 7.1. Fuller, therefore driveable, therefore 4.6 an over. I published the ledger rather than the argument. The piece ran 4,000 words and almost nobody read it. Within a month two county analysts were citing the dataset. The ledger was right; nobody read it until someone needed a name to blame.

Here a second crack opens, and it runs deeper than the first: the betting seam. Live data, moving at three to eight seconds of latency, leaves a broadcast graphic and becomes a market. Courtsiding is a fifteen-year-old problem in cricket — men sitting in stands, typing ball-by-ball by hand, feeding a system the screen never mentions. The same feed from which I build a length table is, for someone else, money three minutes early. That is an ethical problem, but it is first a technical one: a single feed gives one customer news and another customer an edge. The moment a feed is sold into two markets, its truth becomes a shared asset and its timing becomes private property.

Now the part where I depart from the popular reading. The most common charge against cricket coverage of the last decade is that the game has drowned in data, that analytics has displaced coaching and commentary, that numbers are killing the story. Having read that charge in full, I would say it misreads where the power actually sits. The volume of published numbers has risen; that is true. The share of numbers that are checked has fallen over the same period. Twenty years ago two analysts would spend a fortnight building their own tables for one Test and then set them against each other. Now forty graphics come off a single official feed and all forty sit on screen at the same time. Multiplication has increased. Verification has not.

The expansion that enriched cricket writing has also made it single-sourced. Pick up coverage of any international match: three newspapers, three languages, the same three statistics, and the same one origin that is named nowhere. I spent the lockdown counting numbers, and the most honest thing I found was neither the press box nor the commentary table — it was a middle-aged man sitting beside a scorer at a domestic ground, thinking three times about whether a delivery was a bye or a wide. He wrote in pencil, because pencil can be erased and because he had nothing but his own accuracy. Nobody upstream of him needs to erase anything, because nobody upstream is asking.

The effect on players is the least measured and the most endured. A domestic spinner whose field map was never recorded carries an economy rate into a franchise auction without a single piece of context attached to it. He has four expensive matches and his season ends. I would not call that injustice; I would call it a sequencing failure. A system that assesses three deliveries with three different grades of instrument and then drops all three into one table cannot verify its own decisions. And a system that cannot verify its decisions slowly forgets that it ever had grounds for making them.

One detail complicates the standard lament. Cricket is the one sport where the five-day format still fills grounds in India, at a time when most markets have lost the Test audience. The lazy claim that T20 killed Test cricket does not survive the attendance record. What has eroded trust in cricket is not the format. It is the intermediation. Spectators still travel. They simply do not know how much checking went into the screen in front of them.

None of this is an argument against a name. Nor is it an argument against the ledger I build myself — my ledger is incomplete too. There is camera error in it, my own fatigue in the coding, and the same twelve per cent manual correction that every collector carries. My argument is only that the incompleteness should be named. A graphic needs one line beneath it: source, sample size, ratio of manual correction, and which formats had tracking available at all. That line is not for the viewer, and not even for the analyst. It is for a system that currently does not know on what basis it chose whom.

Watch three things in the next cycle, and all three concern documents rather than numbers. First, whether ball-by-ball public data reaches the Ranji knockouts — because while domestic cricket stays invisible, India's selection ledger will always be written in the one corner the data economy never visits. Second, whether the next media-rights tender separates the data right into its own line and attaches a declared audit obligation to it. Third, whether any broadcaster adds that provenance line first, and whether anyone then cares. The day a graphic starts stating its own sample, cricket commentary will be as honest as that scorer's pencil. Until then, every scoreboard is a claim, and every graphic is a hope.