The Empty Scorecard: The Cricket Data That Was Never Recorded

Three thirty-two in the morning. A Ranji Trophy match is playing on a laptop screen in a flat in Liverpool — the stream is murky, the commentary audio runs three seconds behind, and when I open the scorecard panel, three places show only a dash: "—".

That dash is not a server fault. It is the product of a fifty-year-old arrangement: runs get recorded, balls get recorded, wickets get recorded, and the things that actually form the fabric of a match never get a column of their own. Last week a page like that landed in front of me: headings printed, field names printed, and empty space after every name. My first instinct was that my own tool had broken. Then it became clear this is not a cricket problem but a recording problem — and that problem decides which questions about the game we can ask and which we cannot.

Cricket is perhaps the most carefully written-down sport in the world. The habit of keeping a scorebook has run since the middle of the nineteenth century, Wisden shelved it, and ESPNcricinfo moved it onto servers. But a record has three separate layers and we treat them as one. The first layer is the score — runs, balls, wickets, overs. That layer is nearly complete; the first Ranji Trophy season was played in 2026-35 and its final scorecard is still findable today, even though not one ball of that match exists on video. The second layer is ball-by-ball, which began to be filled systematically in Indian domestic cricket from the last decade of the twentieth century and still does not arrive from every ground. The third layer is the one modern analysis stands on — the length of each delivery, the line, the outcome of the shot, the position of the fielder, the phase of the innings. That layer is the thinnest, the most uneven, and the one that makes the loudest claims.

This difference between layers is not trivial. About a match with no ball-by-ball record you can ask how many runs a batsman made; you cannot ask how he made them in the powerplay, which bowler he struggled against, or how his strike rate moved under pressure. The scorecard tells you what happened. It never tells you how. And the understanding of the game, to my mind, is built from the second, not the first.

I am not writing this about one match. I am writing about the empty space we all see and never name. Over the past few weeks I turned over the three layers of my own archive — more than a thousand scorecards, sixty-odd matches manually tagged, and the notebooks I kept in India and did not throw away when I moved to Liverpool. What came out clearest in that shuffle was not a number. It was that I do not lack information. I have a lack that looks like information.

Absence comes in three kinds. The first: what was never recorded at all. The second: what was recorded but never published. The third: what was published and nobody read. All three look alike — all three leave a hole where a figure should be — but all three need a different cure, and treating them as one is the biggest mistake available.

In the first category sits everything cricket never built a column for. How many metres a fielder ran to stop a ball in a Ranji match is written nowhere. Whether a catch was dropped or not does not exist in the scorecard's language — there is only "catch" or "no", no middle grade. Bowler speeds in domestic cricket are not recorded on any official sheet; you cannot find a speed-gun report for any Ranji season, because it was never made. That absence is not neutral. Its direct consequence is that fast bowlers in India are identified by eyes rather than data — and eyes always look where the camera is.

The second category is trickier. A great deal is recorded and stays in a cupboard. Much of women's cricket sits here. Scorecards from women's Test matches in the seventies and eighties were printed in old annuals, but they were not digitised with the same labour as the men's game. The consequence showed up when India won the 2026 Women's World Cup and every outlet suddenly had to write the history of women's cricket: many people did not have fifty-year-old numbers, because what existed on paper did not exist on screen. Smriti Mandhana and Shafali Verma are in every feed today; finding even one line of a scorecard belonging to the women who built that road before them still takes effort.

The third category does the most damage, because the fault belongs to us and not to a system. What was printed was not read. In my archive I found a series of batsmen who had one outstanding season and were playing for a different team the next, because nobody looked at the numbers from the domestic circuit where that season was made. This is cricket's largest inequality — not of talent, of attention. A player whose numbers no editor picked up never reaches a selector's desk either, however good those numbers are.

One thing needs saying plainly, because the risk of this piece is that it reads as a complaint. I am not saying the data is lying. I am saying the data is silent, and silence gets mistaken for clearance.

The spreadsheet had a name, and the name had a season. When I was twenty-three and hand-tagging three hundred and eighty Championship matches at a small data startup in Liverpool for twenty-one thousand pounds a year, Leeds United fell to me. Chris Wood scored twenty-seven goals that season. My shot-quality model priced him at 19.4 expected goals — a twenty-nine per cent overperformance. I wrote it as a warning, not a highlight reel. In August 2026 Burnley bought him for fifteen million pounds regardless. The co-founder put the file in the client deck and moved on. I spent the next season quietly tracking whether the warning had been right. I trust the pattern most when it survives a cold Tuesday in Burnley.

That episode is not cricket, but its lesson is. What the file held was not only a number — it also stated what the model could not measure. The model was clean, but the player kept arriving late for the press. That is the gap between a number and a story, and it is the gap that goes missing from our cricket reporting every single day.

At the 2026 World Cup I got to watch sixty-four matches in thirty-two days from a studio in Salford on a remote data feed, logging PPDA and shot maps live. On the night of 30 June in Kazan, France beat Argentina 4-3, and a nineteen-year-old scored twice and won a penalty. My one-page pre-match note had flagged that his shot volume was the highest of any teenager in Europe's top five leagues. Nobody read it until the 64th minute. I watched sixty-four matches in thirty-two days, and the table began to breathe.

So when that empty page appeared in front of me last week — headings present, fields present, information zero — I did not try to fill it. The template-reporting we read about cricket has one standing disease: the blank has to be filled. No information about a team? Write "subcontinental conditions favouring spin". No league figures? Write "the financial expansion of the IPL". That is not filling, that is inventing — and the second line between the two has grown very thin in sports journalism.

The deepest irony of record-keeping is that even for the most heavily measured player you cannot answer ten questions, because the things measured and the things that matter are not the same list. Take an example. A bowler concedes four in a dead over; his economy looks good, the scorecard calls it a successful over. In that same over he bowled two balls the batsman missed entirely, and one catch went down that was logged as a misfield. Three months later, when selection is discussed, the economy will speak and the catch will not. This is not bias, it is a defect of format — the columns we built became our memory.

The same applies to the argument that resurfaces every tournament: can numbers explain a result? My answer is — up to a point, and that point must be stated. When a team loses three games and its expected-score model stays high, we say "the process is fine". But the model does not know what is happening in the dressing room, whose form is cracking, whose bowling has lost the pace it had last year. The model has no answers to those questions — and that is not its failure, it is the limit of its input. The trouble starts when we forget that limit and treat model output as the final truth about the game.

Now the least comfortable part of this piece.

An empty space is not neutral, and neither is the way we read it. When a report says "no risks were found", a reader takes away "no risks exist". Those are not the same sentence. This is the great trap — a shortage of data on a player does not prove weakness, and a surplus of data does not prove strength. Correlation is not causation, and where there is no information at all there is not even correlation.

There is a second aspect of this that I rarely hear discussed. Making a record is itself an editorial decision. Which catch was a "catch" and which was a "half chance" is not a natural truth; it is the opinion of the person filling the scorebook. Whether a delivery counts as a wide or a dot changes a bowler's economy. Whether a dismissal is logged as a run out or a fielding wicket changes two people's numbers. We treat the scorecard as a mirror. It is in fact a document somebody chose.

That is why I hold domestic numbers at a double distance. I believe the matches I watched myself, and I also believe the matches I did not watch — provided I know what, of those, I do not know. Numbers are not the opposite of stories; they are stories with their shoes tied. When the laces are loose, the whole claim of the race collapses.

Now the place where the number must be allowed to hurt.

In 2026, furloughed at sixty per cent pay, I built a dataset of three hundred and seventy-two behind-closed-doors matches — the Bundesliga, Premier League and La Liga restarts. Home win share fell from 45.6 to 38.1 per cent; home goals per game dropped by 0.22. I did not memorise those figures, they live in my fingers. But the dataset carried a weakness I did not write up at the time: it did not know which match mattered how much to which team. The weather was empty, the stadium was empty, the pressure was absent — and I had no column for pressure. The model was clean, but the player kept arriving late for the press.

This is why, when someone says Indian domestic numbers are enough to find talent, I say they answer one specific question and leave the other nine unanswered. They tell you who made the runs. They do not tell you who bowled quick, who took spin under pressure, who ran in on the fourth-morning pitch. And those are precisely the things Test selection rests on.

There is a political side to this too, and it would be wrong to look away. A system with a monopoly on the record also has a monopoly on memory. What happened in the ICC's 2026 revenue restructuring was not only a division of money; it was also a decision about which boards would have the capacity to hold documents, build data and push them out to the world. A board with a full ball-by-ball library will tell the story of history the way it prefers. Those with only a scorecard can only give you the score.

Somewhere between the heat map and the hymn, a supporter is still standing. He wants to tell the story of a match he saw in the ground, and he has no data for it — only the memory that a certain fielder did something that day which never made a column. Our entire analytical apparatus does not take that memory seriously, because it cannot be measured. And that is exactly where the crack between data and the game shows widest.

So what should we watch next?

I have a concrete answer, and it is not about numbers but about the description of numbers. The next time you read a match report or watch an analysis, look for one thing: the sentence that tells you what the writer does not know. A report containing only what is known is incomplete. A report that also says which figure did not come from which ground, whose numbers were borrowed from another format, which result rests on a sample of two matches — that is worth reading.

Three things should happen in the next domestic season, and none of them happens on the field. First, uniform ball-by-ball data from every Ranji ground — what arrives now depends on the host association's goodwill. Second, the selection process should record why a player was left out — until the reason for omission is written down, it can never be reviewed. Third, digitise the old scorecards of women's domestic cricket — and that work has to happen between seasons, because a search that begins after a trophy is lifted is always late.

I watched sixty-four matches in thirty-two days, and the table began to breathe. But at least eleven of those sixty-four I still cannot say anything about, because their data never reached me. Those eleven have no name, no graph, no thread. They are the most important part of this article.

An empty scorecard is not a failure. It is a question cricket asks itself and avoids answering: what do we choose to remember, and who does the choosing?