The Lesson of an Empty File: Saying 'I Don't Know' Is the Hardest Job in Cricket

That afternoon at Chepauk still sits on page seventeen of my notebook. It was a Tuti Patriots net session, Washington Sundar was seventeen, and he had just taken 2 for 19 in four overs. He stayed back. Thirty carrom balls, same spot, same angle, and I watched for twenty minutes to see which way the ball turned in the air. Finally I asked why he had shortened his grip. He answered. The camera was running. The clip did 4,200 views on a local YouTube channel. Four thousand two hundred. Cricket's biggest stories start from numbers nobody counts.

Now, in the same city, another screen. A terminal is open and a file has come back. 'Article title: N/A.' 'Source: N/A.' 'Information points: none.' 'Entities: underivable.' The whole payload is empty. No player named, no match reference, no format, no date. Only one label stuck to it — cricket_world, and even that does not sit right in its own taxonomy.

This is where the real work begins.

Cricket coverage now runs on a two-tier pipeline. The first tier breaks an article down — title, source, facts, units, date. The second tier builds an argument on those facts. When the first tier returns empty, the second has nothing to write. The problem looks technical. The consequence is professional.

Three things were conspicuously missing from that payload: format, names, and date. Without all three, not one sentence about cricket can be verified. No format marker to say Test or T20, no team, no board, no auction, no rules dispute. What remained was an empty file and a question this trade asks far too rarely: when you know nothing, what exactly do you write?

Cricket has become the most data-soaked sport on earth. Ball tracking, shot maps, over rates, fielder sprint data. The more data arrived, the less room there was for 'I don't know.' A commentary box cannot show a blank space; every gap has to be filled. The data desks won the argument years ago that technical numbers change the story. I write the ground-level version of that win — stopwatch and pocket notebook.

The nets tell you who

The nets tell you who. What happens in the forty minutes before a match says more than any post-match quote. Who bowls to whom, who is kept away from the new ball, who stays for throwdowns and who leaves early. Since 2026 I have carried a notebook to every session. Run-up length, grip, over-by-over intent. This is not data, it is a log — and a log cannot lie, it can only be incomplete.

In November 2026 the stadiums were empty and match-day access was gone. I went to Tamil Nadu's Syed Mushtaq Ali camp. N. Jagadeesan took forty-five minutes of extra wicketkeeping drills; six net bowlers sent down eighteen overs in 34°C heat. I worked out the over rate — roughly 1.9 overs a minute, fast for that humidity. The 2,000-word camp piece ran before Tamil Nadu won the title. It was not a match preview. It was a process file.

Even so, there is a rule I do not break: no 'ready' or 'finished' verdict on a player before five sessions. Daily access breeds false certainty. The man who was standing there feels he saw everything. So: five sessions, and the bowling coach's account set against my own read, because a reading has to collide with someone else's testimony.

Counting, wrists, and the margin of error

I counted the seconds between his trigger movement and the bowler. By hand. With a watch, cross-checked later frame by frame. Best case, ±0.2 seconds of error. One count from one spell cannot carry a paragraph. So the method is: at least three spells, two venues, and one trial against the bowler who actually got him out. Otherwise the number is poetry, not measurement.

On 30 June 2026 I ran the live blog from a Chennai desk for France against Argentina — twenty-seven timestamped updates in ninety minutes, and I flagged Mbappé's second goal before the TV feed cut to the replay. That taught me to write in short bursts. I applied the same habit to net sessions: one observation every ten minutes rather than saving everything for a long piece. It is my default on deadline.

Without a format, every number is a rumour

This is where an empty file hurts most. A Test strike rate belongs to a different world from a T20 strike rate. Economy scales differently, the new ball matters differently, death-overs pressure is a different animal. Dropping a label like 'cricket_world' into a database is easy, but that label cannot tell you whether the innings in front of you was World Test Championship cricket, a bilateral, or a warm-up. The weight of the result differs in all three. The moment the format marker disappears, every comparison collapses — average, strike rate, economy, ranking.

If the name drops, the analysis drops

In cricket the unit is the player. No name means no role — opener, anchor, finisher, pace, spin, all-rounder, keeper. No role means no benchmark. No benchmark means no comparison. And what survives without comparison is opinion, not analysis.

There is a system-level question here, bigger than any one match. If a name drops once, it may be a fetch failure — paywall, JavaScript, anti-bot block. If it drops every time, the parser is broken, and the results from earlier batches become suspect too. This is not a cricket problem, it is an information-supply problem — and information-supply problems are the most expensive kind, because they spread quietly.

The small-sample trap

Declaring a player 'ready' or 'finished' off two T20 league highlights is the easiest thing in the world, and the most dangerous for me, because I have daily access. The second trap is home data. A spinner's average at Chepauk tells a different story from his average away; home numbers hide weakness by default. The third is age. If the last twelve months are not set against the career curve, you mistake old greatness for current form.

Price and form are not the same thing

An auction price is not proof of sporting value. A high bid shows the gap between sporting worth and market worth — sometimes a local-young-star premium, sometimes an all-rounder shortage, sometimes panic bidding. Writing the two on one line is the oldest error in the trade. Broadcast cycles, franchise valuations and salary structures move at three different speeds. None of that was in the empty file, so one sentence about it would have been invention.

Where conclusions are forbidden

Nothing about betting or fixing can be written without evidence. Boards, tribunals, anti-corruption investigations, NOC disputes, central contracts — without confirmation of any of these, a sentence like 'fixing risk' can wreck a life and a livelihood. The answer here is clean: unverified. That is not a weak answer. It is the professional one.

The popular read says an analyst who writes N/A or 'unverified' is hiding incompetence, and that the real reporter always finds the story. My experience is the reverse. The reporter who fills the blank damages the record; the one who shows the blank protects it. A pipeline that reads 'no risk found' out of an empty file is more dangerous than one that publishes nothing at all, because the correct translation of an empty file is 'zero knowledge,' not 'zero risk.'

I also have to guard against my own contrarian reflex. The reversal is my identity, but a reversal is only valid if next match's data can prove it wrong. So the rule is: write the counter-intuitive claim into the notebook first, report it second. A claim that cannot be cut down by the next ball is not a claim, it is a currency.

The most useful output of this whole episode is its own failure. An empty payload is a clean signal: re-run extraction, spot-check entity recall against the raw source, fix the label mapping. And if the second run still returns a title of N/A and zero information points, the right move is to drop the item from the analytical queue rather than force an eight-dimension report into existence.

Three things to watch. The re-extraction should return at least five information points, keep at least one named player, and carry the format marker — only then does any cricketing judgement exist. Until then the only true sentence available is that we do not know yet. The question is not when the story arrives. The question is who should be holding the pen when the file comes back empty.

— Chepauk net diary, the Beat Keeper