Read the source before the story: the probability problem behind a leaked list

A screenshot. An unnamed social media account. A list of eight names that had already left a competition, along with a prediction for the next departure and the eventual winner. Within hours it was reshared thousands of times, because people believed it. I read this description while tidying my notes for the transfer window. What made me stop was the structure. That screenshot operates exactly like the lines I read every morning: an anonymous source, a list that looks verified, and a crowd ready to treat a prediction as a result. In football, we call that a transfer rumour. Every transfer window, the same screenshot is rebuilt, with different names. Football runs on information. An unsigned contract, a release clause, an incomplete medical — every scrap of that data has a price. And because it has a price, there is a middle layer that lives by spreading it. The structure of football sourcing resembles a tower. At the top are journalists with direct club relationships, recognisable because every line they write carries a timestamp and a verification mechanism: medical completed, personal terms agreed, only the fee remains. The middle holds national outlets, slower but stable. Below is the aggregation layer, reposting without naming a source. The base of the tower is anonymous accounts, accountable to no one, with nothing but a screenshot. The paradox is speed. The base spreads fastest. The lowest tier is unbound by truth; it only needs attention. A journalist at the top weighs a decade of credibility every time they hit publish. An anonymous account loses nothing. Based on my experience following matches and transfer windows, I keep a source notebook. Whenever a rumour appears, I log three things: who reported it, when, and how specific the claim is. After a few months, the notebook paints a clearer picture than any credibility ranking. The accounts that post most are usually right least. The accounts that post least, but always with a timestamp, are usually right most. The leaked list I read is a clean example of that mechanism. Its only origin is a social media account. No newsroom, no byline, no independently verifiable publication date. The evidence chain has a single link: that account said so. Let us take it apart the way I take apart a match. The first phase is seeding. The account posts the list. It does not need to be right; it needs to spread. To spread, it needs a headline name and a claim that cannot be checked immediately. A claim verifiable on the spot dies within hours. A claim vague enough to survive the night gets a chance. The second phase is acceleration. Aggregator sites pick up the list. Each pickup replaces the faint credibility of the origin with the credibility of the reposter. By the fifth hop, readers have forgotten who said the first sentence. The third phase is the climax. A real event — a reveal, an official nomination list, a match — is used as an anchor. The real event is blended into the false story, and the false story borrows the truth to look credible. The fourth phase is verification, and it is the most essential one. If the prediction lands, the account is canonised. If it fails, the story instantly gains a new explanation: the script was changed. Those four phases are not exclusive to television. They repeat intact in every transfer window. There is a calculation nobody states. With six names on the nomination list, the baseline chance of a single prediction being right is roughly one in six. A blind guess still wins once in six. In football, the equivalent calculation is an account predicting a player will join one of five clubs. There are five plausible destinations, and people will remember the hit and forget the four misses. That is confirmation bias running like a machine: it filters in supporting evidence and deletes the rest. If you guess blindly across a transfer window with thirty major deals, you will still get a few right. Of those thirty, only five correct calls need to be remembered for you to have a legend. And that legend needs no evidence, only people repeating it. I have a habit of asking myself three questions before logging a rumour in my notebook. Does the source take responsibility by name? Can the claim be checked against a timestamp? If it is wrong, does the source lose anything? With that leaked list, all three questions return the same answer. Anonymous source, no timestamp, no risk. The three matching answers form a signal, and that signal sits at the lowest tier of credibility. A list can be retrofitted. This is the most overlooked point. When someone publishes a finishing order, people only check the part that already happened. But the part that already happened can be adjusted to fit before publication. A list matching the past says nothing about the future; it only says its author read the news. The incentive mechanism here is asymmetric. Get it right and the account gains followers. Get it wrong and attention usually moves to the next rumour, and the account suffers very little real damage. The cost of being wrong is near zero; the reward for being right is finite but repeatable. That is a bet with an edge. In football, that bet has a name: the transfer window. Recall the big deals. Some players were said to have agreed personal terms for months, then joined the right destination at the last minute. Some deals were declared done, then collapsed at the final step. Kylian Mbappe and the long saga with Real Madrid is the classic example: for over two years, every week someone predicted he would go, every week someone predicted he would stay, and in the end only one prediction came true — yet many claimed they had called it. The same mechanism applied to Neymar and the record deal of 2026. The 222 million euro fee in the Neymar deal is not the most important part. What matters is how many people got it wrong while the story was unfolding, and how many later misremember being right. So when a club pays a hundred million euro for a player who has not yet played fifty top-flight matches, remember that the money is decided on the basis of reports. Part of that may come from an internal source consisting of a single agent generating noise. The transfer market does not buy players; it buys problems. A transfer rumour is really a hypothesis about how the variables — fee, wages, age, contract length, sell-on clause — fit together. Every formation is a hypothesis, the match is the experiment. A leaked list works the same way: it is the hypothesis, and the results reveal is the experiment. The trouble is that most readers remember the hypothesis and skip the experiment. There is a striking detail in how such an article is built. The only element with verification value — the official nomination list published by the producers — sits inside a piece whose bulk is unverified rumour. The truth is used as a prop for the rumour. Readers take the whole piece as one block, and the credibility of the true part bleeds into the false part. In football, this technique is so common it is standard. A piece claims a club has made an enquiry, adds a few verifiable details — the manager once worked with the player, the club showed interest last year — then packages everything into a confident block. What readers remember is the part that cannot be verified. A rumour does not just reflect the market; it can create the market. A story that spreads far enough makes fans expect, forces clubs to respond, and sometimes makes a deal that did not exist become real. That is the most dangerous part of the mechanism: self-fulfilment. In the Vietnamese market, where football news passes through many layers of translation and sharing, this effect is stronger. An English line, through three reposts and one translation, can lose both context and origin while keeping its confident tone. What the final reader receives is an assertion, not a hypothesis. I do not want to turn scepticism into a religion. Sometimes an anonymous source is the right one. Leaks have been part of football long before social media. Agents leak to create pressure. Clubs leak to test reactions. Some people know the outcome in advance because they are genuinely in the room. If I dismissed everything without a name attached, I would miss the earliest signals of the market. The issue is not whether to believe. The issue is assigning probability. An anonymous source with an unverifiable claim gets a low probability, but not zero. A source with a track record of hits gets a higher one. A source with a timestamp and a specific mechanism gets higher still. And here is the blind spot of the over-sceptical: they confuse unverified with false. The two differ. An unverified rumour can still be true. A good reader does not say this is false; they say I lack enough evidence to lift its probability above the baseline. I do not believe in randomness; I believe in repeated passes. In football, a lucky long shot says nothing; a repeating pattern does. Rumours work the same way: a single hit says nothing; a verifiable, repeatable streak does. The most worrying thing is the self-defending structure of the story. If the prediction lands, the story is reinforced. If it fails, the story shifts to another version: someone intervened, the script was changed, the result was staged. No outcome can falsify the story. And a story that cannot be falsified is not information; it is belief. This is like how we read an offside frame. A line precise to the millimetre can be used to judge a moving attack, and while we argue about a toe, we forget how an entire counterattack unfolded. The same error: using one precise data point for a process that never stands still. Back to the television producers. The greatest risk is not a leaked result, but audiences losing faith in the integrity of the competition. In football, the equivalent risk is called a fixed match. Once faith in integrity collapses, every result is suspected, including entirely fair ones. Lost trust is far harder to regain than it was to win the first time. The same expectation mechanism puts pressure on real people. A player returning from a long injury is demanded to prove himself in his very first match, because the story was already written before he touched the ball. The only part of the story I could cross-check was the official nomination list: published by the producers, verifiable, citable. Everything else has a single link: one account's account. And the unverifiable part is the part remembered longest. The standard of good analysis is information gain: it must tell you something you did not know. A leaked list adds nothing — it merely repeats what someone wants you to believe. A probability calculation adds one thing: how to place your bet. Next transfer window, when you read a line attached to a screenshot, ask the same three questions I ask. Who takes responsibility by name? Does the claim carry a timestamp? If it is wrong, what does the reporter lose? The answers will not tell you whether the rumour is true or false. They will only tell you what percentage to stake on it. The only thing that cannot be faked on the pitch is tactics, and the only thing that cannot be faked in the market is an accountable source. Everything else is waiting to be falsified. And when the results reveal arrives, what is worth watching is not who was right or wrong, but whether the reporter answers for the story they told.