The Posterior

Underneath its warmth, the show is one of the most disciplined dramatizations of Bayesian inference popular television has produced. The idea is simpler than the name. You hold a belief, evidence arrives, and you revise the belief by the right amount, then do it again.

Ted never asks anyone to hold a conviction immune to evidence. He asks them to hold a workable belief long enough to let the evidence come in. That is the whole difference the show runs on, and it has a technical name.

Bayesian reasoning begins with what is termed a "prior," the degree of belief you hold in some proposition before new evidence arrives. It updates that prior in proportion to the likelihood, how probable the evidence would be if the proposition were true compared to if it were false. The result is the posterior, your revised belief after the evidence, which becomes the prior for the next round. Belief, in this account, is not a fixed possession. It is a running quantity, carried forward and corrected, over and over, as the world supplies data. A person reasoning well is not a person who arrives at certainty. It is a person whose beliefs track the evidence at the right speed, moving fast when the evidence is strong and slowly when it is weak, never quite reaching the limits of zero (no belief) or one (total belief).

Everything the show calls belief is this process wearing a friendlier name.

The line Ted attributes to Walt Whitman, and that this library treats elsewhere as misattributed and therefore all the more revealing, turns out to be a near-exact paraphrase of a technical principle in Bayesian statistics. To be judgmental, in the show's usage, is to have decided in advance and closed the account, to have collapsed a person onto zero or one where no future evidence can ever move them. This is what a closed mind is, in plain terms, a belief locked at certain-yes or certain-no, sealed against anything the world might later show. To be curious is the opposite, an open mind, keeping the belief somewhere between the two extremes, where the person in front of you can still surprise you, which is to say where evidence can still do its work. Ted's ethic, delivered as a folksy imperative, is a formal rule of inference stated as a homespun sentiment. Think it possible you may be mistaken about this person. Leave a little probability on the table. It can be as small as you like, but leave it there.

The show's villains, read through this lens, are never the skeptics. Skepticism is healthy, a demand that the evidence be strong before the belief moves far. The show's villains are the people who have assigned zero or one and can no longer update. Rupert has set the probability that anyone matters, except as an instrument, to zero, and no evidence reaches him because none can. The early press corps sets the probability of Ted's competence to zero, and the interesting thing the show does is show which journalists update and which do not. A skeptic says the evidence is not yet sufficient. A judgmental person says no evidence would be sufficient. The first is doing inference. The second has stopped.

The clearest place to watch belief revised in public is the crowd, because a stadium is a population of priors being updated at once. In the first season the Richmond supporters hold a low opinion of Ted and they chant the verdict at him, wanker, the same word the show tracks with almost experimental care. That chant is a prior stated aloud. Then the evidence accumulates across three seasons, and the show does something a lesser story would not bother to make precise. It does not have the crowd flip from contempt to adoration in a single moment of triumph. It has the verdict shift by degrees, and then, in the third season, it relocates. The same chant is directed at Rupert. The crowd has revised on both questions at once, the probability that Ted was the fool driven down by the evidence of who he turned out to be, the probability that Rupert was the fool driven up by the same. A crowd that revises its verdict on evidence, and revises it against its own initial position, is a crowd reasoning correctly.

Trent is the single well-calibrated reasoner shown at close range, and the show is unusually precise about both his starting point and the moment it first moves. His prior is stated as a question at Ted's introductory press conference. Handed the floor, Trent instead asks whether the whole thing is a joke. Given the evidence available in that room, an American gridiron coach hired into the Premier League with no knowledge of the sport, a low opinion of Ted's competence is the correct reading, and a good reasoner assigns it. What separates Trent from the people who never recover is that his is not welded. He keeps it off the boundary, a low number rather than a zero, which is precisely what leaves it open to revision.

The revision begins at a specific, datable moment, the day early in the first season when Trent shadows Ted for a profile meant to be a takedown. For one day the evidence arrives in a rush. He watches Ted move through a school visit and then through a dinner where Ted eats food far too spicy for him without complaint, rather than embarrass the man who invited him in front of that man's family, and Trent cannot make that behavior more probable under his joke hypothesis than under its rival. Each observation is better explained by a Ted who is the real thing than by a Ted who is a fraud. The proof that the update is real is what Trent then publishes. The profile does not come out as the hatchet job it was commissioned to be. He writes that he still expects Ted to fail, but that he finds himself rooting for him and will not gloat if the failure comes. That printed sentence is a belief caught in the middle of moving. It has shifted off its starting value, the direction of travel is now toward Ted, and it has not yet crossed into confidence. What he put in print was a belief that had partially moved, no longer his starting certainty that Ted would fail, not yet confidence in him, which is a more honest thing to publish than either certainty would have been.

The deepest proof of what Trent has become arrives when his two beliefs collide. A journalist protects a source. Inside the profession that rule sits about as close to a certainty as a working belief ever gets, the sort of near-certainty a person treats as unrevisable. So the interesting question is why Trent, of all people, breaks it. When the panic-attack story runs on a leak from Nate, Trent tells Ted who the source was. The move is not a lapse in his principles but the triumph of his best one. His belief about Ted, built up across two seasons of evidence, has grown strong enough to override even a rule he had held at the boundary. The accumulated data about the kind of man Ted is outweighs the professional certainty, and the near-welded belief turns out, under sufficient evidence, to have been off the boundary after all. Even the sacred rule was not actually pinned to one, and the proof is that the right evidence moved it.

The self-punishment that follows settles the account. Trent does not merely reveal the source, he turns himself in, confessing to his editors that he burned it, and accepts the firing rather than let the deception stand. A man does not do that to escape a consequence. He does it to pay one. And then the update runs in the other direction. Ted receives the burned source as data, and the datapoint is enormous, a hostile journalist who gave up his profession to be honest with him. Ted's own belief about Trent, still shaded by two seasons of adversarial press, revises hard toward trust, which is the quiet precondition for everything that follows, the book, the access, the seat among the Diamond Dogs.

The show also runs an update on the audience, and the cleanest instance is the pair it introduces in the very first scene. On the flight to England the pilot hands us a tidy read of who Ted and Beard are. Beard is the brains, head down in a soccer book, the one holding the technical knowledge. Ted is the heart, sleepless and full of belief and unable to name the offside rule. It is a fast, low-information impression formed on thin first-scene evidence, which is exactly what a prior is, and the show invites us to hold it. Then it spends three seasons supplying the evidence that reverses it.

Watch the tactical axis first. The book in Beard's hands on that plane is Inverting the Pyramid, a real history of soccer tactics, and it contains a chapter on Total Football. The most sophisticated tactical idea in the series is sitting in the supposed tactician's lap on day one, coded as his department. Yet when Total Football finally arrives, in Amsterdam, it does not come out of Beard. It comes out of Ted, unbidden, from a vision, and Beard's role shrinks to recognizing the thing Ted produced and giving it its name. The tactic the brains carried as theory, the heart delivered as insight.

The psychological axis reverses in the other direction, and the two crossings form a single figure. The man introduced as heart is the one who cannot reach his own interior, and his sleeplessness on that very plane is the first data point in a case the show builds for three seasons, the panic and the buried grief he cannot face. The man introduced as brains turns out to be the steady relational ground beneath his friend, the presence that could sit beside Ted on the plane and can sit beside him still. Each man is strongest in exactly the domain the pilot assigned to the other. Hold the first-scene reading as a certainty and you never see it. Hold it as a prior, a number off the boundary, and the series becomes a long, satisfying update to its reverse.

The show also shows what happens when the reasoning breaks down, and it locates the breakdown, poignantly, in the people who assign extreme priors to themselves. Nate sets the probability that he is worth anything to zero, a prior handed to him by a father whose approval never arrived, and the tragedy the show documents is that no evidence can reach a self-belief pinned to the boundary. Ted's warmth is strong evidence. The team's acceptance is strong evidence. None of it moves him, because the belief is welded at zero, and Bayes' theorem is merciless about that particular starting point. What looks like Nate's insecurity is, in the precise technical sense, an un-updatable belief about himself. His redemption, when it finally comes, is not the arrival of new evidence. The evidence was always there. It is the moment the belief is unwelded, lifted off zero, so that what had been bouncing off for years can finally register. You cannot update someone off the boundary by arguing. You have to get them to move first, which is why the show's answer is never a persuasive speech. It is a relationship, the thing that quietly relaxes a zero into a small positive number where evidence can start to accumulate.

Rupert is the same breakdown with the sign flipped. He has assigned himself a one, certainty of his own centrality, a mind closed around itself, and a one is just as unreachable as a zero. No evidence of his own diminishment can land, because the update cannot move a one either. The chant that finally names him is not information he can use. A man at probability one has forfeited the machinery that would let him hear it. The show lets him walk out intact and unrevised, which is the accurate ending for a belief that was welded. Nothing was ever going to reach it.

Roy asks the question directly in the finale, and it is the most Bayesian question in the series. Can people change. There are two ways to hold a person in mind. One treats a person as a fixed value that more data merely pins down more precisely, and on that view the question is almost meaningless, because the value does not move. You are what the long-run average says you are. The other treats a person as a set of possibilities carried forward and revised by what actually happens, and on that view change is not only possible, it is the default condition of any system still taking in evidence.

The answer the room lands on, offered by the man named for his mother, is that people never reach perfection but can keep asking for help and accepting it, and by doing so keep moving toward better. Read literally, that is a description of a posterior that never terminates. It does not arrive. It keeps updating, moving toward better as long as the evidence keeps coming and the belief stays off the boundary. The show's final word on human change is a refusal to let the estimate harden into certainty.

Which returns us to the word over the door. The other essay in this library that treats the Believe sign calls it a placebo that announces itself as a placebo, and notes that the announcement does not weaken the effect. In these terms the paradox dissolves into sense. The sign is an instruction to set a prior above zero — to not close the account before the evidence has had a chance to come in. The show's dangerous move is always the zero end, the person who has given up, decided in advance, welded the prior to nothing. One is harder to reach than it looks; zero is available at any moment of despair. The sign does not push anyone toward certainty. It pulls them back from surrender. It does not claim the belief is already justified by evidence, which is what makes it look like a placebo. It claims only that you should hold the belief above the floor long enough to gather the evidence that will justify or refute it. A belief that survives being torn to pieces and reassembled is not a belief that was never tested. It is a posterior that has been through many rounds of updating and has not collapsed, which is far stronger than a conviction that was never contested at all. The team that pieces the sign back together during the match is not restoring a naive faith. It is confirming a belief that has been revised against three seasons of hard data and has survived the revision.

The show gave you the instruction on the first day and asked you to take it as a feeling. Take it instead as a technical directive and the whole series reorganizes into a single sustained argument. Believe. Don't close the account. Think it possible you may be mistaken about the American, about the striker, about the kit man, about the version of yourself you have already given up on. Then watch the evidence come in, and let it move you.