Empty Cells, Full Caution: Why Format Is Cricket Analytics' First Variable

Half past eleven at night on the Barishal desk. The monsoon air is making a wet sound against the window glass, and I have opened a return file that contains eleven cells — all empty. No title, no source, no information points, no entities identified, no time-sensitivity assessed. In the world of sport this is not a rare sight; the rare sight is someone admitting it. Because two kinds of people sit in front of an empty cell and do two kinds of work — one says there is no information, therefore there is no decision; the other says there is no information, therefore let me fill the cell with imagination. In the history of cricket writing, the second group's team total is far higher.

On the field I recognise this moment. When a batter stops before releasing the bat, looks at the pad, looks at the umpire — he is not making a decision; he is pushing the weight of the decision onto someone else's shoulders. The empty cells of a data desk behave exactly the same way. They state no truth, but they force us to answer: you do not know, and yet what will you do?

My way of working is simple, though explaining it is hard. I do not merely watch matches; I break a match into layers. The first layer is description — who scored how many, who took how many wickets, what happened in which over. The second layer is explanation — why it happened. The third layer is inference — what is the probability of what comes next. These three layers can never be blended, because the gap between description and explanation is precisely where the real work of analysis lives.

This whole structure, however, stands on one foundation, and that is input. The quality of an analysis can never exceed the quality of its material. A model may be as clean, as elegant, as confident as you like — if its base is empty, the model is decoration.

In Bangladesh's cricket ecology this input problem cuts deeper. Data collection in our domestic structure is never perfect. One Dhaka venue has ball-tracking, another does not. The camera angle in Sylhet and the camera angle in Chattogram are not the same, so the same catch becomes two different stories in two places. Upstream talent-supply information — age-group cricket, academy records, district-level performance — is still largely on paper and in envelopes. Yet it is exactly this information that decides how deep the national bench will be six years from now.

I grew up in Australia, where the data infrastructure was solid, professional, broadcast-ready. In 2026, at forty, sitting in Barishal, when I started a bilingual blog called Expected Goal, my head was full of European football, Ronaldo's 12 goals against an xG of 10.4 — an environment where every shot event is logged. In Barishal I learned that a spreadsheet can be a monastery. Coming to Bangladesh I learned something more: the first question is never what the metric says; the first question is whether the metric even exists.

That question is what today's empty file has handed back to me. The eight pillars of analysis — format, player, team, league economics, rules and governance, risk, public narrative, and industry transmission — each require an information point before anything else. Without information points the pillars stand on nothing, and analysis standing on nothing is not analysis.

In cricket, format is the first-order variable. I say this emphatically, because this is where the most errors occur. Test, ODI, T20 — these are not three forms of the same game; they are three separate systems. A conclusion reached by placing a batter's Test average beside his T20 strike rate may be arithmetically valid, but it is cricketing nonsense. Leaving the ball is a skill in Test cricket; in T20 it is self-destruction. The middle overs in an ODI are a game of control; in a Test they are a plan to put the opposition to sleep, and it works.

I learned this lesson in the language of football in 2026, when I analysed all 64 matches of the Russia World Cup remotely. France's PPDA was 14.8, meaning they had almost withdrawn their press and sat deep. The low block, the controlled risk, Mbappe's 32.4 km/h top speed — read those three facts separately and it is easy to write France off as lucky. The 2026 PPDA map was not a chart; it was a confession. Read the scoreline without matching format and phase, and analysis is almost compelled to go the wrong way. Without this format-first discipline, no analysis of any sport survives.

The second pillar is the player. Here role comes first, metric second. Opener, anchor, finisher, pacer, spinner, all-rounder, wicketkeeper — each is judged on a different metric. A finisher's value sits in strike rate, a Test opener's in leave percentage and patience at the crease. Anyone who discusses averages before identifying the player is using a role-neutral number to make a role-dependent judgment. In the Bangladesh context this is even more complicated, because our young cricketers often enter one format carrying another format's record, and that record does not match what the system demands. Age curves, form trends, condition-based splits — none of these can be measured, because the role was never identified before measurement began.

The third pillar is the team, and here the question is bench depth. ICC rankings are format-specific — before citing a number you must state which format it belongs to. More importantly, a ranking position and a squad's actual construction never tell the same story. A team may sit mid-table in the Test rankings, but if the age structure of its bowling attack is dense above thirty, the risk of the next two years will not appear in the rankings; it will appear six months later on the injury list. This is where environmental context becomes a first-class variable: travel, humidity, soil, and even whether a stadium has a crowd.

I know that mentioning empty stadiums makes some people roll their eyes. But it is not sentiment; it can be measured. Home advantage shrinks in a crowdless environment, because a large part of home advantage is really the interaction of umpiring pressure and crowd noise. When the stadiums emptied, home advantage became a ghost in the machine — present, but impossible to catch. Writing that kind of sentence requires venue data, dates, and a minimum record of attendance. Today's file does not even have that much.

The fourth pillar is the league and commercial ecosystem. The IPL, the BBL, The Hundred, the PSL, the SA20, the ILT20, the MLC — each has a different broadcast value, franchise valuation and salary structure. Read a contract figure without comparing it to a player's sporting value, and the analysis turns into a rumour. I do not chase transfers; I audit the panic behind them — that is, how much of a team's willingness to pay was cricketing need and how much was the broadcast cycle's deadline. There is another layer here that is often skipped: league-versus-national-team schedule conflict. When the calendar swells, what is called load management is often really a way of making room for commercial tours. To prove that, you need specific tour dates, travel miles and rest intervals — none of which exist today.

The fifth pillar is rules and governance. Power distribution, DLS, the umpire's call in DRS, anti-corruption investigations, NOCs, central-contract discipline, political interference — each has a different context and precedent. My position on the review system is clear: the technology has not reduced controversy, it has moved controversy from the field to the review room and the grey zones of the rulebook. But writing that still requires a specific incident, a date, a decision. Rule analysis on zero input amounts to generic commentary, which is not this column's job.

The sixth pillar is risk. This is where I spend the most time, because listing risks is easy and sizing them is hard. Sporting risk, personnel risk, commercial risk, rules risk, public-opinion risk, structural risk — each needs its probability and impact measured separately. But the one thing I can state with certainty today is not cricket risk; it is data-pipeline risk. And this risk is more dangerous than all the others because it is silent. A model is a vow: simple rules, repeated until they confess — but with an empty input, the model outputs its own imagination instead of a confession.

The seventh pillar is public narrative. This is where most accidents happen. The gap between market expectation and objective assessment can be measured — in team results, in player performance, in auction prices. But when a narrative is built without input, it stops being analysis and becomes a current of emotion. In Bangladesh's cricket conversation this habit is familiar: after one innings, the greatest of an era; after the next match, finished. Between the two sentences, nobody asks about sample size. The crowd sees drama; I see the columns breathing underneath.

The eighth pillar is industry transmission. From upstream to midstream, from midstream to downstream — that path is visible only when the event itself is known. A league's broadcast deal, a board's central-contract restructuring, cricket's entry into the Olympics — these interlock to create new markets. But tracing a transmission path from an unknown event is impossible. Writing insufficient information at each position is then the only honest answer.

Which brings the question: is empty information actually bad information? No. And this is the most important distinction of the night. Absence is not evidence — I write it repeatedly and forget it repeatedly. An empty cell does not say nothing happened; it says I do not know what happened. The distance between those two statements is the boundary between professional analysis and social-media comment.

In my experience the biggest temptation comes from the deadline. With a deadline on your neck, an empty cell is unbearable; the mind says, add an argument, the audience is waiting. But what gets written then is not an explanation of information; it is an explanation of the gap — and it is almost always a rerun of an old narrative that only sounds new.

There is a simple reason for this in my own local reality. Outside Dhaka, at the mofussil grounds, venue-level information is often kept informally — in notebooks, on phones, in someone's memory. The difference in pitch character at the same venue between two seasons may be written nowhere, and yet that is the real key to batting and bowling decisions. The easy escape from this emptiness is to tell a story. The hard road is to leave the cell empty and say: my data is not here, and I will not break it open.

I archive the noise until it becomes a signal worth trusting — and signals take time to arrive. Not one series, two or three; not one venue, a cluster of venues. That waiting is not laziness, it is method. An analyst who cannot wait is not an analyst; he is merely fast.

So tonight is not a defeat. The eight pillars have returned empty, but the framework is intact — and an intact framework means the analysis is ready the moment input arrives. The question now is technical: whether the source text actually entered the system, whether the title and source were captured, whether the format was tagged. Until a name, a number and a date arrive, I will not write a single sentence about cricket — because analysis written on wrong information is far more harmful than an empty cell.

Over the coming weeks I will watch three things: whether format tagging lands correctly, whether entity extraction works, and whether source quality can be graded. If those three cells fill, all eight pillars come alive in a single pass. And if they do not, this column has done its job: it has told you where the information is not, and why that space cannot be filled with guesswork.