Empty File, Empty Stands: When the Cricket Story Isn't There

The file opened and there was nothing in it. N/A where the title should be, N/A where the source should be, and where the summary should be, just a blank. In sixty years I have seen many empty stands in stadiums — at a Ranji knockout there were three people on the third tier while the microphone warmed up in the commentary box. But the emptiness on my screen that evening in Hyderabad was not the emptiness of a stand. It was the emptiness of a story. I have said since 2026 that I learned more from the empty seats than the full ones, and I stand by it. But nothing can be learned from an empty file. An empty file says only one sentence, and that sentence is: right now, you have nothing. This report is about that sentence. It is not a match report, not a transfer story, not a retention-list analysis. It is the report of the moment when a beat writer is handed an open screen and told to write — and decides, before writing, whether there is anything worth writing. The first check was clear. The Stage-1 extraction came back empty. No title, no source, article type unclassified, one-sentence summary blank, author stance undetermined, purpose undetermined, the list of information points empty, and the instruction for entities — "identify from the information points above" — became a joke, because there were no points to identify. Time sensitivity could not be assessed. Source quality could not be weighed because there was no source to weigh. One signal survived: the domain label, cricket_world. The system knew only that this had something to do with the world of cricket. That is like being told cricket was played in Hyderabad yesterday evening and being asked to write the story of that match. Which match, whose match, which series, which ground, what score — nothing. There are thousands of cricket stories in the world. If you have no basis for choosing one, you are not choosing a story; you are inventing one. I carry a notebook that has travelled with me for three decades. In 2026, when I joined The Indian Express in Madras as a trainee, the first lesson was this: write what you saw, write what you heard with the source attached, and say plainly when you do not know. For about twenty-five years I also wrote film reviews for the same paper. Film criticism had one convenience — if you had seen the film, you had every right to an opinion. Cricket does not give you that convenience. In cricket you must first establish that you watched the match, or that the person you asked was actually there. In 2026, when I was living at the team hotel with Sunrisers Hyderabad and writing the twelve-part series "The Locker Room", I learned that the biggest story in a dressing room is often the one nobody says. What I wrote about David Warner's captaincy was not built on press conferences; it was built on who was last to leave the ground after training and who was first onto the bus. Bhuvneshwar Kumar's Purple Cap season — 26 wickets — I understood not from the scorebook but from the rhythm of his bowling, slow for three overs and suddenly quick for the last two. Writing that required getting up at six in the morning with the team. After every post I read more than 200 fan comments. I had a WhatsApp group of 150 fans, "SRH Beat". I would ask them: did I get the tone right? That question became a habit. And that habit is what protects me in front of an empty file today. Because if I sent that group a message saying "nothing came in today, so I made something up", 150 people would say in one voice: then don't write it. In 2026, when the IPL was in the UAE and the stadiums were empty, I wrote "The Silence of the Stands". Five thousand words, 300 fan voice notes, built around T. Natarajan's 16 wickets and Sunrisers' third-place finish. The hardest part of that piece was not the cricket. The hardest part was deciding whether the exhaustion on a player's face inside the bio-bubble was a story or not. I decided then that you protect the person and scrutinise the decision. That is still the first rule of my work. At the 2026 mega auction in Bengaluru I live-tweeted Sunrisers' buys — Washington Sundar at 8.75 crore, Nicholas Pooran at 10.75 crore, and Mumbai Indians' record 15.25 crore bid for Ishan Kishan. That day I ran a 2,000-vote fan poll on Sunrisers' best buy and spoke to three fans who cried when Pooran was sold. One fan accused me on Twitter of a "cold" analysis of a released player. I rewrote the piece to lead with the player's family reaction. I am recalling all this because there is only one honest way to fill a file that is empty today — to go back and find where the material fell. And that check offers four clear routes. First, the input never reached the extractor. The body was empty, or what arrived was a paywall or login page. This is the oldest mistake of my newspaper years — the reporter thinks he has filed copy when he has only filed a link. Second, a schema mismatch. The model answered in prose, the parser dropped it, and every field fell back to N/A. The check must be run on the raw output, not the parsed object. Third, the input was not an article at all. A scorecard, a live-commentary feed, a video page, an index page. These have no stance by construction, so the empty fields are not wrong — they are exactly right. And the correct response is for the pipeline to refuse to proceed, not to try to fill the blanks. Fourth, the domain label may have come from a fallback. It must be confirmed whether cricket_world came from classification or from a default. Because if it came from a default, we have nothing at all — not even a label. Here an old habit returns. At the auction table, every pause has a price. When the auctioneer raises a hand and stops, that pause tells you whom the room trusts and whom it does not. In the same way, the pause in this empty file is telling us a price — and that price is trust. If we fill this emptiness by inventing something, we are spending the reader's trust. The minimum needed to proceed is clear: the raw article text, or populated information points each carrying a claim, a source, a timestamp and an evidence type, or at least a title, source and publication date — enough to weigh source quality and route the work. I would also want to know what the downstream consumer needs — fact-check, summary, entity graph, or a bias read. Each task has a different shape, and I would rather not produce the wrong shape twice. One thing needs saying plainly, because I know the newsroom will ask it. The question will come: you have the domain label cricket_world, so write something. Three hundred words, anything, the reader will read it anyway. I take that question seriously, because it touches the oldest weakness of my trade. A newspaper has space to fill. The page cannot close. And when the hurry to fill space grows larger than the truth, journalism slowly becomes a trade in guesswork. But one thing I have learned in forty years. A wrong number in cricket never travels alone. It carries other wrong numbers with it. If I print a wrong auction figure today, tomorrow that figure is in a fan's tweet, then in a video, then in a child's school project. Ten years later some journalist will treat it as a source and write from it. I have seen that chain of error many times in auction reporting — one cold remark, one misread stance, and then for months that remark travels attached to a player's name. So when the file is empty, the most honest story is that the file is empty. That is not weakness. It is the discipline that makes a beat writer a journalist rather than a publicist. An empty extraction and an empty stand are not the same thing, but both ask one question: why is what is missing, missing? When the third tier looks empty at a Ranji knockout, that emptiness tells you about ticket prices, about broadcast timing, about a city's habits. That emptiness is information. In the same way, the emptiness of Stage-1 is information — but it is information about our system, not about cricket. And if we cover that information with a story instead of reading it, we have erred twice: once in the pipeline, and once on the page. In April 2026 I was sitting in the Sunrisers dugout when they made 287/3 against RCB, with Abhishek Sharma's 69 off 28 and Travis Head's 102 off 41 in front of me. That evening I ran a fan poll: is this fearless batting good for cricket? Seventy-eight per cent said yes. When data analysts called it unsustainable, I asked the fans to defend it. I still think that was one of my better decisions, because that day I let the story become the room's voice rather than mine. But there was one thing I did not do that day either. I did not say 287 would be made on every ground. I did not say this was the future of cricket. I wrote only what happened in that dugout that evening and how the people in that room felt it. That is the difference. Experience can be turned into news. Extrapolation cannot. And here another old habit comes back, the one I value most now. In 2026, when the players looked tired inside the bio-bubble, I asked fans to send encouragement videos and I played them in the team room. There is no statistic for that. There is no scorecard for it. But looking at this empty file today, it seems the most necessary part of journalism is the part outside the numbers — deciding who gets to speak and who does not. Now to the most uncomfortable thing in this whole episode, which I will write despite the room's displeasure. One section of the newsroom will say this is all technical, that the reader has no stake in it. Their argument will be clean: the reader wants a story, not an internal pipeline report. And I concede the argument has force. The reader should not have to care about schema mismatches and parser failures. But here I would ask them one thing. If the reader does not care about the pipeline, why does the reader care about the news that comes out of it? Why do we grant ourselves permission not to inspect the thing that made the news? I remember a fan writing to me on Twitter after the 2026 auction that my analysis was cold. It stung at the time. Later I understood he was right. I had written about a released player's price and forgotten his family. I had the data right and the human being wrong. Since then I open my "Fan Auction Tracker" before every transfer story — the spreadsheet where fan sentiment sits opposite a player's price. The same applies to this empty file. The right thing is not to invent a story to please the reader. The right thing is to inspect and repair the pipeline, because the next time it brings the news of a player's most fragile day, we need to be sure that what is written is true. One more thing, for those who think this is a technical glitch that will fix itself. It will not fix itself. Because on the day a system has the label of the cricket world and no story, that system also has something else — room to invent. And room always gets used, if nothing stands in its way. I have seen many times in forty years how a small error grows. Once a local paper printed a wrong score for a Ranji player. The next week that score was on a radio bulletin. Then it reached the player's relative, who called to ask what had been written. It took three weeks to correct one wrong digit. And all of it happened because someone guessed instead of opening the scorebook. So when I say nothing can be written on an empty file, I am not delivering a moral sermon. I am telling you what one guess costs. It costs three weeks, or a player's family their sleep, or a reader their trust. Now, forward. For those repairing this pipeline, I have one suggestion, and it comes from my trade, not from technology. Keep an "emptiness reason" field with every extraction. If Stage-1 returns empty, let it say why — empty body, paywall, schema drop, or input that was never an article. We do not have that information today, and its absence leaves us in the dark among four possibilities. Second — never leave a domain label alone. A label is a direction, not a story. If there is no information point beside the label, treat the label as empty too. Third, and most important — put a gate in the pipeline that can say plainly: we cannot go further from here. In my trade that gate is called an editor. A good editor is not the one who prints every story but the one who says, this is not a story, go and look again. I think every system needs such an editor. And the last word, which I will give to the notebook that has been with me for three decades. Many pages in that notebook carry only a date, with nothing beneath it. Those blank pages are my greatest asset. They record that on those days I wrote nothing, because on those days there genuinely was nothing. And on the days when there genuinely was something, I wrote all of it. The same should hold for this file. Today it is empty. Tomorrow, if raw article text arrives, or populated information points with claims, sources and timestamps, I will run the full analysis. Until then, this emptiness is the most honest news in this report. And the next time someone asks me to just write something because we have the cricket_world label, I will say what I said in that WhatsApp group in 2026. Drop it, fix the tone, then come back. Because the locker room was quiet before the noise arrived. And what was not there in that quiet, I will never write.