The Report With Not a Single Fact: Silent Failure in Cricket Analytics Pipelines and a Lesson in Chain-of-Evidence

Late last week, at two in the morning at my Barishal desk, I opened a file. Eight dimensions, eight analytical frameworks, every heading correct, every table neatly arranged. Yet in every cell the same sentence returned: "Insufficient information, assessment not possible." For more than eighteen years I have read injury ledgers, medical reports, scans and board statements. But I had not seen such an honest document in a long time. An analyst who does not know has written: I do not know. The danger is not there. The danger is that a document saying "I do not know" is something many people want to turn into a dramatic story. That file was the output of the second stage (Stage-2) of a cricket analytics system. The first stage (Stage-1) was supposed to extract information from the source article — title, source, summary, author stance and, most importantly, a list of atomic facts called Information Points. What Stage-1 returned was entirely empty. No title, no source, no summary, not a single information point. Only one tiny trace survived — the domain label cricket_asia, which is itself a taxonomy artefact, not information. Every analytical conclusion was required to be tied to an information point. When the information points are zero, that chain breaks, and only one honest answer remains — there is no information, so there is no assessment. What lies behind this failure touches a question larger than cricket journalism: how much do we trust our data, and where do we keep the evidence of that trust? Modern cricket analysis is no longer confined to a single journalist's pen. Today it is a two-stage system. The first stage gathers raw material — extracting verifiable information points from an article, a match report, a board statement. The second stage builds an eight-dimension analysis from those information points — format and match, player technique and data, team and ranking, league and commerce, rules and governance, risk, public narrative, and industry transmission. The foundation of the whole building is the first stage's information points. When the foundation is hollow, no vast edifice stands — only a beautiful mould. This is not merely a technical incident. In the regular season, viewers watch every match, yet most analysis is now produced automatically. In such conditions, a silent pipeline failure means the signals beneath the table — fitness, fatigue, selection pressure — go undetected, while everyone believes they know everything. As the speed of analysis rises, verification of its source falls behind — that gap is my real concern. I am not coy about this chain, because my profession taught me so. In 2026, at forty-five, I sat in the Old Trafford press box and watched Zlatan Ibrahimovic rupture his ACL. He wore number 9 and had scored 28 goals in 46 games. A colleague dismissed my question about landing mechanics. I returned to Barishal and re-watched all 46 matches, logged 312 aerial duels, and found he landed on his right leg 73 percent of the time. I began the Zlatan audit exactly where the highlight reel ended — at the first twitch. That audit taught me a hard rule — I do not print a return date without three independent medical sources. In 2026 I followed Mohamed Salah's shoulder injury under the same rule. He was struck by Sergio Ramos's tackle in the 30th minute of the Champions League final; Egypt's number 10 had scored 44 goals in 52 games for Liverpool. Combining reports from Liverpool, Egypt and UEFA, I predicted he would not start against Uruguay. He did not. My editor wanted a quick "hot take"; I refused until three sources converged. From that rule my weekly "Injury Ledger" was born, tracking 50 players, syndicated in three countries. And in 2026, during the Covid hiatus, I analysed Virgil van Dijk's ACL tear through the question of the empty stadium. Combing through 120 post-restart Premier League matches, I found ACL injuries rose 40 percent in empty stadiums. I wrote "The Empty Stadium Knee." A TV producer called it "too academic." That day I decided every piece would begin with a methodology note. All three experiences teach one lesson — without process, talent is not reliable, and the first condition of reliability is knowing where the stream comes from. If you do not know where an analysis came from, it is not analysis but guesswork. Before the tackle became a talking point, it was a joint, a load and a millisecond — just so, before an empty payload becomes a talking point, it is only a missing link. The empty file is not the analysts' failure; it is a warning — and that warning can be read at three levels. First, it identifies the type of failure. Losing title, source, summary and information points all at once is not a simple parsing error. With paywalls or JavaScript-rendered pages, the title at least usually survives. All fields going blank at once is rather a sign of a hand-off fault — meaning the source article either never reached Stage-1, or was a video, image or live-score widget with no prose at all. The distinction is not small: a parsing error is fixed in the scraper, a hand-off error is fixed in the whole pipeline's connection. Second, it speaks of taxonomy drift. The spec required the domain label "Cricket", but "cricket_asia" arrived. This mismatch is not itself information, but it hints that the label vocabulary between the first and second stages does not match. If a label goes down the wrong path, the analysis is filed in the wrong ledger — and data filed wrongly can later be miscounted in some fan-token or ranking-commerce accounting. Third — and this is most important — a zero input creates its biggest risk in the analyst's mind, not in the data. Given an empty mould, the easy temptation is to fill it with imagination. Eight dimensions, eight headings — all ready. Invent players, a match, a score, and the document looks "complete." In the age of artificial intelligence this temptation is not merely theoretical, it is institutional. If an automated system reads "empty" not as "failure" but as "must be filled," it will spin a cricket story out of itself. And that is journalism's greatest sin — dressing an evidence-free claim in the clothes of evidence. This is where the blockchain question arises, and it is structural, not metaphorical. The idea of a chain of evidence sits at blockchain's core too — every transaction timestamped, hash-linked to the previous one, and no one able to quietly rewrite an old page. My "three sources" rule is in fact a human-made form of consensus — only when three independent parties reach the same truth is it accepted; just as a block is not valid until a majority of the network accepts it. In cricket data this has a direct application: an immutable injury and workload ledger, in which the time of every hand-off, the hash of every payload and the source of every information point are permanently recorded. Had every exchange between Stage-1 and Stage-2 marked that ledger, the empty payload would not have needed an analyst to wait until two in the morning — it would have been caught instantly. Here I want to make my hesitation clear, because the risk is the real point. Blockchain enthusiasts often think immutability means reliability. That is wrong. A ledger records lies too, just as carefully as truths. Put false data on-chain and it is not purified — the falsehood becomes permanent. "Garbage in, garbage on-chain" — in injury data this is especially dangerous. If a club understates a player's workload figure and it locks onto the chain, some future reviewer will treat that error as evidence. Immutability does not protect truth; it merely makes lies harder to catch — if no one bothers to catch them. So the real reform is not in technology but in process. The cheapest and most effective defence against an empty payload is a "null-guard" — a validation step that blocks hand-off to the second stage whenever the information-point list is empty. It can be placed in a smart contract on-chain, but it must also be placed as a human rule. Because however good the technology, the question "who is accountable" must be answered by people — who will warn, who will decide to re-extract, and who will own the failure. My biggest complaint about my industry is here: we roar about fan tokens, digital cards and broadcast rights, yet stay silent about the cleanliness of inputs. Blockchain's greatest promise is not mere glitter but verifiability. Yet before we use verifiability, we must decide — what do we actually want to prove, and to whom are we accountable. The empty file can be deleted, but the question remains. This failure is less a failure of technology than of governance — a pipeline's silence spreads into every document until someone consciously catches it. And the most dangerous failure does not shout; it stays quiet, and waits for someone to turn it into a story. The question is simple: next time an analytics system returns zero, will we accept it as "we do not know," or dress it up as "we know"? And if our injury data truly goes on-chain one day, will we still keep evidence, or merely pretend to keep trust? A ledger that cannot recognise emptiness will not recognise evidence either.