Eight Pillars of Zero Evidence: The Cricket Report That Refused to Lie

It is past eleven at night. In the workroom in Brisbane, under a single lamp, I opened a file. Eight chapters. Beneath each, a table; in each table, columns; in each column, an assessment — format, player, team, league and commerce, governance, risk, public narrative, industry transmission. The structure was immaculate. Then I began reading the cells. First cell: insufficient information. Second cell: insufficient information. Third, fourth, fifth — the same sentence, the same spelling, more than sixty times in a row. Six risk categories, three scenarios, eight dimensions, two decision layers. Template completion rate: one hundred per cent. Evidence rate: zero. A cricket report in which not a single ball is mentioned. I once counted zero attendance, seat by seat, until absence itself became a statistic. That time there was a match; only the crowd was missing. This absence runs deeper — the match itself is not there. Eight years ago, in 2026, past fifty, I first understood how loudly an empty cell can shout. The World Cup was on in Russia and I was doing live data analysis for a digital sports platform out of Brisbane. My model said Australia's expected goals were 3.2 while they had scored only 2, and their passes per defensive action stood at 10.4 — pressing intensity so low that the back line would lie open like a book in front of Peru's set pieces. They lost 0-2 to Peru and went out. I spent three weeks re-watching every tape, cross-checking against Opta, and wrote a four-thousand-word autopsy. From that day a rule settled into place: no claim without a ten-match sample. Two years later, in 2026, that rule took me inside one hundred and twenty matches played behind closed doors. Home advantage fell from 0.45 goals per game to 0.18; referee bias dropped twelve per cent. I spent six weeks checking every variable and published a methodology note with confidence intervals. I was the only analyst in Australia to do that work, and this is not a boast — it only shows how few people sit down to read an empty room. In January 2026, evaluating Azzedine Ounahi for Brisbane Roar, I applied the same method. Progressive carries of 8.2 per ninety, defensive duel success at 43 per cent, an xG chain of 0.18 — the numbers said he would add weight rather than lift it. The club did not sign him. He went to Marseille. My twelve-page report carried the ledger lines of fifteen comparable midfielders. Those three pieces share one formula, and the formula is plain: behind every sentence there must sit a countable object — a delivery, a seat, a ledger line. The rest is ornament. The file in front of me is the second stage of a two-stage analytical chain. The first stage extracts from a document: title, source, publication date, article type, one-sentence summary, author stance, purpose, a list of information points, entities involved, time sensitivity, source quality, domain label. The second stage arranges those information points into eight dimensions. The chain's rule is strict — the only evidentiary base for stage two is stage one's list of information points. On this run, that list is empty. No title. No source. Article type unclassified. Summary blank. No author stance. No purpose. Zero information points. Time sensitivity never assessed. Source quality undeterminable, because there is nothing to determine it against. Only one field is populated: the domain label, which reads cricket_asia. Populated labels beside blank values — that pairing is the most important clue here. A document that is genuinely contentless usually leaves the field labels empty too. Here the labels are set out neatly and only the values are missing. That is the behaviour of a template, not of a document. The failure therefore probably sits at the acquisition layer, not the analysis layer. Three possibilities follow. One, acquisition failure — the document sat behind a paywall, or the parser returned an empty body from a JavaScript-rendered page without raising an error. Two, a classification-only pass, in which the information-point decomposition step never ran. Three, a non-cricket document routed into this pipeline by mistake. The first two are moderately likely; the third is not. Why this emptiness is especially dangerous in cricket is the report's most useful contribution. In other sports, an unknown format makes analysis difficult; in cricket it makes analysis impossible. Average, strike rate and economy rate run on entirely different benchmark scales across Test, ODI and T20. Apply one format's benchmark to another and what you get is cross-format contamination. With the format unknown, that contamination is unavoidable. Here the analysis stopped in exactly the right place. The same holds for venue, toss, weather, innings state, a Duckworth-Lewis-Stern revised target — none of it is recoverable. Result and process therefore cannot be separated. A team's genuine trend and a one-off outcome are distinguishable only through sample, and here the sample is zero. Zero sample is not the same as small sample. From a small sample you can at least extract a directional hint and then attach a confidence interval to mark it weak. A zero sample contains not even that hint; whatever remains is guesswork, and when guesswork is placed into a table cell it ceases to be guesswork — it becomes a claim. Team assessment carries the same trap. Cricket's home-away differential is among the widest in world sport — set India's spin dominance at home beside its record of adapting to pace abroad and the point is obvious: team assessment without venue context is close to meaningless. Here there is no venue, and no team. In the governance section there is one sentence I want to hold apart: the absence of an integrity allegation in the input does not mean the absence of integrity risk. In lower-tier or emerging leagues, anti-corruption monitoring exposure is structurally elevated. But there is no fact here to support such a claim, and so the claim is not made. That restraint is the report's strongest ground. The vocabulary deserves attention too. The domain is written as cricket_asia, whereas the specified classification should have been simply Cricket, with region carried in a separate field. A label is never evidence. Cricket_asia says the subject leans toward some Asian cricket event — a bilateral series schedule, Asian Cricket Council tournament logistics, or board-level administration. Those are directions of inference, not a factual base. Taken whole, the report makes one honest confession about itself: in the final row of the risk matrix it writes — analytical risk, level high, likelihood high, already realised. The reason is plain: the pipeline produced a structurally complete report written in a confident register with nothing inside it. In an analytical chain this is the most dangerous failure mode, because template completeness can be misread as evidentiary substance. This is where I part company. What the report did is methodologically correct — stopping rather than estimating. But the form in which it did so is a lie of another kind. Eight dimensions, star ratings, confidence levels, a transmission map with arrows, and a heading that reads Comprehensive Assessment — this form performs authority. The sporting director who sits down to decide will not read eleven pages of insufficient information; he will read the heading, see the stars, and then tell the meeting that the data says so. I know where that pressure comes from, because I have sat in that room. Had my Ounahi report of 2026 been built in this shape — twelve pages of insufficient information — someone at the club would have filled the empty cells with his own gut feel, and that filled-in cell would have passed the meeting as a data-driven decision. Guesswork never signs its own name on paper. And the pressure is structural rather than personal. The cricket press market pays by volume, not by restraint. Editors do not pay for reports that say insufficient information. So the empty cells get filled — not by scoundrels, by deadlines. Here the difference between two cricket cultures surfaces again. In Dhaka, silence is filled with emotion — a single empty seat can be spun into a large story. In Brisbane, silence is filled with models — a well-set table and three decimal places. Both manufacture false authority; only the typeface differs. Neither is honest by default; only the ledger is, if it is left open. A neatly formatted cricket report whose every cell stops at insufficient information is really a warning — beside every number sits an empty chair. I have kept scorebooks for more than sixty years. From that memory bank I could have written this entire piece: a series name, a bowler's figures, a reason for a defeat, all invented. That would have been the easiest work available. I did not, because memory uncorroborated by a record is not evidence, only habit. So the forward signal needs stating. From now on what to watch is not the conclusion but the information-point count. A report with confidence levels populated and evidence cells blank — that is the red flag. Second signal: the health of source acquisition. Third: the distance between label and proof. The question stays open here. How many cricket analyses are circulating right now whose foundation is exactly this empty — except that nobody stopped at insufficient information, and a plausible number was slipped into the gap? The empty report we can see is the safe one. The dangerous report is the one that looks full.