Eight Sections, Zero Evidence: The Silent Failure of Cricket's Data Pipeline

At ten past two in the morning in Rangpur, a file opened on my laptop screen. It was called Stage-2 Deep Analysis Report. The document was complete. Eight sections, a table under each, a row in each table, and in every row the same answer: N/A, insufficient information. From the input-integrity diagnostic to the comprehensive judgment, not a single box was empty. And the informational value of the entire document was zero.

I read it twice. Then I pulled an old notebook out of the drawer. A page from 2026, written at Mirpur: 'Not Taskin — he hadn't arrived yet. That day a left-arm spinner bowled in the powerplay, economy 3.2. The commentary said "outstanding bowling average". Who was going to point out that the 3.2 was a six-over spell, while the bowler at the other end went for eleven in the death?'

The number was true. The sentence was false.

The document in my hand is a different kind of falsehood. It contains no numbers, so nobody will be caught out. But the damage is identical: a wrong number and an empty box do the same job. Both are made to look like knowledge. The reader sees a table and trusts it. What is missing inside the table, the reader cannot see.

I have watched this game for forty years. The spreadsheet still surprises me. But the reason has changed. Once I was surprised by what the numbers said. Now I am surprised by where the number came from in the first place.

The market for numbers, and the economics of the template

Cricket journalism is no longer just journalism. It is a supply chain. The match ends at seven in the evening. Photographs at nine. Scorecard at eleven. First reaction at one. 'Analysis' at four in the morning. Player previews for fantasy leagues at seven. A sponsor deck by nine.

Every link in that chain runs on a template. Templates are not evil. I have built them myself and put them in junior analysts' hands. Metric templates, matchup sheets, workload calculators. Nobody covers five matches a day without one.

But a template has a temperament, and nobody writes about it. A template wants to be full. It does not want to be true; it wants to be filled. When rain arrives at the Wanderers, the box cannot simply stay empty — someone drops in an estimate, someone drags in last season's figures, someone writes 'conditional'. In the final file there is not a single N/A left. The reader assumes it is analysis.

The document that reached me is the exact inverse of that habit. Nobody filled anything. Eight sections, four tables each, a risk matrix, a transmission map, star ratings — all present, with not one sentence inside them. That is precisely what makes it interesting. It is not an accident. It is a confession from a process.

The process is simple. In stage one, someone reads an article and extracts its information points, its entities, its central claim and its publication date. In stage two, an eight-dimension analysis is built on that skeleton. Now imagine stage one returns nothing. What does stage two do?

The answer is in front of me. Stage two builds the entire scaffold anyway, writes N/A in every cell, and closes with a 'comprehensive judgment' awarding one star out of five. It looks impeccably professional. It is a hollow shell in a suit.

There is a detail buried in the document itself. The domain label arrived as cricket_world, while the schema demands Cricket. It looks trivial. It is actually the clue. A system that recognised cricket could not capture the content. Which means the failure was not in the source article. It was in extraction — possibly a paywall, possibly image-only text, possibly language, possibly a governance story.

In Bangladesh this failure mode is familiar. A large share of our cricket data comes from English scorecards, a smaller share from Bangla commentary, and a substantial share from somebody's notebook — written by a person sitting in the ground. Those three sources are not formatted alike. The scorecard says 'economy 7.4'. The commentary says 'there was pressure in that over'. The notebook says 'it was the sixteenth over, the batter was new, the spinner was bowling'.

The first does not understand the second. The second does not understand the third. Unless the three layers are joined, the analysis that emerges looks complete while remaining partial. And that is the industry's central flaw: we worry about the shortage of numbers, not about their meaning.

Eight pillars, one at a time

One: the format wall

The first section asked which format. Test, ODI, T20, The Hundred? No answer.

With that one box empty, the other seven become meaningless. In cricket, format is not a division of time. It is a different sport.

Take an opener with a strike rate of 140. The headline writes 'aggressive opener'. But 140 in a T20 powerplay is devastating. And 140 in the second session of a Test innings is freakish — and possibly unsustainable. Same number, two opposite stories.

My notebook has a rule written in red ink: mixing formats means adding the averages of two different sports. It is not merely inadmissible; it is meaningless. A spinner's ODI economy and Test economy cannot be compared, because in Tests he is buying wickets and in ODIs he is buying dot balls, and those two ambitions do not live in the same delivery.

In Bangladesh this error recurs. We look at a seamer's T20 death-overs economy and conclude he cannot handle pressure, while his ODI new-ball numbers are excellent. Two jobs, two skills, two separate ledgers.

The second empty box matters just as much — venue. Nobody wrote where the match was played. Mirpur's surface is one thing in winter and another in April. Sylhet's outfield is slow, Dhaka's is quick, and dew arrives in Chattogram at about half past seven in the evening. None of those sentences fit in a table, and all three decide results.

In the 2026 Nidahas Trophy final in Colombo the match went to the last ball. Dinesh Karthik's six is often explained as batting brilliance. In my ledger it was a spin-bowling matchup — a part-timer bowling the final over because the frontline seamers had run out. The decision belonged to bowling resources, not to the batter's greatness.

Without format and venue, that distinction disappears. And when it disappears, the story attaches itself to the wrong place: to an individual's legend, instead of to a team's arithmetic.

Two: the player, and the trap of the small sample

Nobody was named in the second section.

That is the most dangerous emptiness in cricket analysis. Without a player there is no metric, and even with a player the metric frequently lies.

One thing I have watched for years and that nobody wants to admit: home numbers hide a player's real face. If a Bangladeshi batter's home average runs fifteen above his away average, his career average tells you almost nothing. He grew up at Mirpur, where the ball stays low and spin arrives slowly, and he has rehearsed that surface his whole life. Abroad he is a different person.

That is why my desk keeps three columns — Home, Away, Neutral — and never merges them. Dubai and Abu Dhabi are not the same place either.

The second trap is sample size. A bowler takes nine wickets in three matches and we write 'back in form'. Nine wickets in three matches sits comfortably inside the normal variance of a good bowler. In my own database the innings-to-innings wicket variation for a regular T20 bowler is wide enough that no five-match series proves anything. Yet the sentence is written after every series.

The third trap is role. A bowler's economy is 8.2 — good or bad? It depends entirely on when he bowls. In the powerplay, 8.2 is excellent. At the death, 8.2 is outstanding. In the middle overs, 8.2 is ordinary. One number, three verdicts.

Put all three traps together and you get something like the reaction to Bangladesh's 2026 Under-19 World Cup win. After the final, the talent analysis ran the same line everywhere: 'a new generation has arrived.' The real question was different — how many of that XI would play fifty senior domestic matches within three years? Nobody wrote that number, because it was not knowable, and the unknowable does not fit a template.

I remember writing in my notebook on the night of that final: 'Three of these players will be regulars in the national side within five years, if their domestic workload is managed. Names recorded.' I wrote the names and the date. Later I checked. Some matched, some did not. The ones that did not taught me more.

Three: the team ladder and home soil

No team was identified in the third section. No ranking, no squad.

The heart of team analysis is comparison, and comparison needs a ladder. The World Test Championship table is one ladder, the ICC rankings another, the home-away record a third. They rarely point the same way.

For Bangladesh this is daily experience. At home we can squeeze any side. Abroad the arithmetic collapses. Because the home advantage is not pace — it is spin — and the spin advantage depends on weather. Winter Mirpur keeps the ball low; April Mirpur starts turning from the second session. Two tracks, one stadium.

On squad structure, one thing is clear. Our batting depth is reasonable in the middle order and thin at the top. Our bowling depth is deep in spin and shallow in pace — the pace department still leans on one man. That sentence took me five years to write, because every series I assumed it would fix itself.

Age structure is another silent variable. An average age of 29 sounds fine. But if it is built from two 35-year-olds and five 24-year-olds, the average is useless. The mean is the most widely used metric that erases every piece of information held in the distribution.

So I do not look at average age. I look at the distribution of experience: how many have batted in a fourth innings, how many have bowled at the death, how many have played the match after a loss. Those three questions describe a squad's real ladder.

Four: leagues, auctions, and the two languages of value

No league in the fourth section. No contract figure. No auction.

This is where my own risk is greatest, because every time I see an auction price my mind starts thinking in two languages. The language of price: how many crore, how many million. And the language of cricket: how many runs, how many wickets, how many match-winning innings.

The two never converge, and that gap is the real auction story. Auctions price scarcity, not skill. Where the position is thin, a moderate player is expensive. Where the position is crowded — openers, in our part of the world — a good player goes cheap.

The premiums hide inside the price and must be separated: local-young-star premium, all-rounder premium, scarce-position premium, and panic-bidding premium, the last two minutes of filling an unfilled purse. Distinguishing the third from the fourth is hard, and it is exactly the difference between a good decision and a bad one.

One number stays with me as an illustration. After 2026, a Premier League club signed a Moroccan centre-back for eight million euros on the basis of my low-block composite. A year later his value had multiplied. The question is not whether I saw the future. The question is whether the club bought because they had watched the matches or because their data desk had built a matrix. The second, and that is the only argument my profession has.

But those successes are also my biggest trap. Once a data man succeeds, he starts to believe his model is true. The model is never true. It is an estimate in a suit.

Five: rules, power, and the risk of silence

No rule, no board decision, no sanction in the fifth section.

One point deserves to be stated plainly, because it is the most important lesson in the whole document. 'No risk' and 'risk not assessed' are not the same thing. An empty box never means safe. It means not yet looked at.

In cricket, rules and power travel together. Revenue-share disputes, the Big Three, the politics of bilateral series — none of it sits in a table as a number, and all of it changes how the game is played. The day a board cancels a series, no database has that written in advance.

Two episodes keep me careful, because both changed my method. The Cronje affair of 2026, and the Lord's spot-fixing of 2026. Since then, whenever an innings produces a 'strange' statistical shape, I hold two explanations at once — a cricketing one and a non-cricketing one. Writing only the first leaves the analysis partial. Then the 2026 IPL case, which showed that the larger the structure, the harder the monitoring.

So empty boxes in this section ring alarm bells for me. Without the rules you do not know the conditions under which the cricket was played.

DLS is the example. A batter scores fifty off thirty balls, the chase is won, everyone says magnificent. But if the match is washed out four overs later, the value of that innings cannot be judged by run rate. It has to be judged by resources. Same number, different weight.

And DRS? Umpire's call is a margin. That margin does not change a result, but it changes the story of a series. Anyone who lives in numbers knows the margin is an arbitrary line, drawn by committee rather than by nature.

Six: the risk matrix, with empty rows

Sixth section: an eight-row risk matrix, every row N/A.

There is one line in it that is the most honest sentence in the whole report: the risk is not sporting but procedural — high.

That is true. But I have a separate opinion about cricket risk, and I state it often.

In the workload era, the biggest invisible risk is the calendar, not the body. The Future Tours Programme is a document, but every box in it lands with weight on a fast bowler's knee. Tests, ODIs, T20Is, leagues — four places call, and none of the four can be refused, because three are boards and one is a contract.

There is a subtler thing that big clubs know and we do not write. The five-substitution rule rewards deep squads — everyone has written that. Nobody has written that the reward arrives in the last twenty minutes, when one team stops running and the other introduces fresh legs. Cricket is importing that logic into T20: pressure after a slow over rate, the arithmetic of bowling resources, the captain's spell as a calculation rather than an instinct.

When football returned to empty stadiums in 2026, I looked at the first fifty matches. Home win percentage dropped. Home pressing dropped. Distance covered dropped. I wrote it in the notebook that night: the stadium emptied, and home advantage walked out with the crowd. I have the receipts.

That number still earns its keep, because many clubs still treat home advantage as a property of the pitch. A large part of it was always noise made by people.

Seven: the market for stories

No narrative in the seventh section. No cycle.

Narrative is a product in cricket journalism. 'Return', 'revenge', 'farewell' — those three words sell more columns than any metric. But a narrative has a fixed lifespan. A side that wins a series is exalted for three months; the first defeat makes the same side a crisis. That oscillation has less to do with skill than with media rhythm.

My rule here is simple. When I hear an expectation, I ask how many matches the number behind it came from. Below thirty, I do not call it expectation. I call it noise.

We have a clear example at home. We win a home series and everyone says 'a new Bangladesh'. Three months later we tour abroad — same players, different results. Nobody said the home advantage had gone. Nobody said the two series were played under different conditions.

Why not? Because 'a new Bangladesh' sells and 'the conditions changed' does not.

And this is my profession's deepest discomfort. If I write only the winning story, my readership grows. If I write the conditions story, it shrinks and my work stays correct. I have chosen the second, and I pay for it every month.

Eight: how numbers travel through the industry

Eighth section: a transmission map with 'no data' in all three boxes.

What that map is trying to show matters. An event in cricket never stays in one place. An Under-19 World Cup win transmits into three: the youth supply chain, the national selection structure, and the broadcast-advertising market.

The first is slowest and most important. The second is fastest and least stable. The third is loudest and shallowest.

Because the three move at different speeds, the commonest error in cricket journalism is to read the third's volume as evidence about the first.

One boundary belongs here, because it is my ethical line. I never draw the betting arrow on that map. When a number enters a betting market its meaning changes. My job is to explain the cricket, not to issue a wagering signal.

The contrarian angle: the full sheet is the dangerous one

I have spent this much space on an empty document. The reader might assume my anger is aimed at the file — a bad artefact, a failed pipeline.

The opposite is true.

The empty document is harmless. The dangerous document is the complete one.

Suppose the same pipeline had returned one wrong fact instead of nothing. Suppose stage one had mislabelled a one-day international as a T20. What would stage two have done? It would have produced all eight dimensions, filled every box, and got every sentence wrong — while looking flawless. No N/A anywhere, so nobody asks a question.

That difference is the centre of my profession. An empty document catches itself. A full one does not.

Which is why I say: the real danger in a template is not its empty cells but its filled ones. An empty cell invites a question. A filled cell delivers an answer.

Here I have to confess a methodological sin of my own. For years I leaned on heatmaps — where the ball lands, where the fielder stands. Beautiful to look at. Then I understood: a heatmap conceals a player's role. The picture shows more fielders on the off side but never says why. Perhaps the captain knows the bowler will not bowl the slower ball, so he keeps the slip. The heatmap does not know that.

So I interrogate my own model: if I remove this picture, does the analysis still stand? If it does not, it was never analysis. It was decoration.

The second contrarian point. The most honest sentence in the report was that no model should be published without its conditions. I believe that. But the belief has a toxic form, and I have seen it in myself: turning context into an alibi.

I make a forecast, it fails, and afterwards I say the dew did it — except I never wrote 'dew' before the match. Then it is not analysis. It is an apology.

My rule, which I give to juniors: conditions must be written before the match, not after. Written before, they are inputs. Written after, they are advocacy.

The third contrarian point is the most uncomfortable. In this report seven risk rows are empty and one is full — process risk. Reading it felt like a mirror. If I am honest, several of my own published analyses were procedurally weak. The number was right, the source was right, but the conditions list was incomplete. Readers could not catch it, because the table looked full.

Completeness is not a promise of truth. Completeness is an aesthetic preference. And in cricket journalism we sell aesthetics as evidence.

The final line of the document says its own contents must not be cited. Imagine that — an eight-dimension analysis whose author closes by declaring it uncitable. In professional life such confessions are rare. What usually happens is that the empty boxes quietly fill themselves and nobody confesses.

So this document is not a specimen of failure. It is an honest failure, and an honest failure teaches more than a dishonest success.

Signal for the next ball

Three things go into my notebook, with dates, because a promise without a date is not a promise — it is conversation.

First, a minimum-input gate. Before any analysis is published, four questions must be answerable: what competition and format, who is at least one named entity, what is the central claim in one sentence, and what is the publication date. If any of the four is missing, the analysis does not run — however elegant the document.

Second, a conditions list attached to every forecast and written before the match. Dew, rain, venue, travel, rest. Written early, they cannot become excuses later.

Third, a ledger of my own errors. On my desk there is a file called the book of mistakes, containing every wrong call I have published, with dates. Nobody asks for that file. It is the only proof I have that I am estimating rather than prophesying.

And one question for the reader, which I cannot answer either.

When every cricket match is written in numbers before it is played — who wins, who scores what, who sells for how much — what will the reader come to watch? The number is already known.

My guess is that they will come for the moment the number is proved wrong. Because that is where the human being is. That is where the game is.

And to catch those moments, my notebook is still in the drawer, beside the laptop.