Chapter 10 — A Real Method

Chapter 10 — A Real Method

I have spent nine chapters taking Sherlock Holmes apart, and it would be a cheap and cowardly book that left him in pieces on the floor without offering you something to put in his place. So here it is: the method. The actual, usable, do-it-at-home discipline of honest inference — everything the great detective pretended to have, reassembled into something a real person can practise on a real Tuesday. It will not make you feel like a genius. That is the first sign that it's working.

Think of it as the Baker Street method with the five bugs of Part I surgically removed and the five engines of Part II installed in their place. It has six moves. None of them is difficult. All of them are unnatural, which is exactly why they work — they are braces against the specific ways your mind bends.

One — Observe what isn't there

Holmes observes brilliantly, and everyone remembers it, and everyone learns the wrong lesson from it. The lesson is not "notice more details." Your pattern-drunk brain already notices far too many details and spins each into a story. The real discipline, the one that separates the honest reasoner from the conspiracy theorist, is to notice what is missing — the evidence that should be there if your theory were true and isn't. Remember WYSIATI, the mind's fatal contentment with whatever happens to be in view. The corrective question is not "what do I see?" but "what would I expect to see if I were right — and do I?" The dog that did not bark in the night was Holmes's one genuinely great insight, and it is great precisely because it is an observation of an absence. Train yourself to feel the shape of the missing evidence. It is the single most powerful habit in this book.

Two — Hold a crowd of suspects, never one

The instant you find yourself with a single theory, you are in danger, because from that instant your mind stops investigating and starts defending. Confirmation bias needs only one hypothesis to feed on; starve it by keeping several alive at once. Force yourself, always, to generate at least three explanations for anything that matters — the innocent one, the boring one, and the one you'd prefer not to be true. The open window was left open, blew open, or was forced open; keep all three in the room. This is not indecision. It is the deliberate refusal to fall in love before the evidence has spoken, and it is the habit our storytelling instincts fight hardest, because a single vivid narrative is so much more satisfying than a committee of maybes.

Three — Ask how common it is before you ask how clever you are

Before you get excited about your clue, consult the base rate — the boring background statistic, the prior. How common is the thing you're leaping to, in the general run of the world, before any evidence? The tan line thrills you toward "soldier"; the base rate reminds you that London is full of gardeners and holidaymakers. The strange symptom frightens you toward the rare disease; the base rate reminds you that common things are common. This is the move that Meehl's index card makes and the confident expert skips. Whenever a single striking detail is about to convince you that you have understood a stranger, a situation, or a lump on a scan — stop, and ask how many other people share that exact detail without sharing your conclusion. The base rate is boring, and it is almost always right, and it is the exact thing every failed inference in this book forgot.

Four — Go hunting for the "no"

This is the hardest one, the one that feels physically wrong, and therefore the one that matters most. Your mind, left alone, seeks evidence that confirms. So you must override it, deliberately, and go looking for the evidence that would prove you wrong. Not to be noble — to be right. Ask of your favourite theory: what would I expect to see if this were false, and have I actually checked? Wason's little number game undid almost everyone because they tested 8-10-12 to confirm and never tested 1-2-3 to break. The whole of real science is just this move, formalised and institutionalised: the active, disciplined, unnatural hunt for your own error. In your own life it means seeking the friend who disagrees, the fact that doesn't fit, the disconfirming test — and thanking the evidence that wounds your theory, because a theory that survives an honest attempt to kill it is the only kind worth trusting.

Five — Move in inches, not leaps

Holmes vaults from ignorance to certainty in a single bound; it is his signature and his poison. The honest reasoner does the opposite — updating belief in small increments as each piece of evidence arrives, a few percentage points at a time, the way the superforecasters do and the way Bayes demands. Resist the two great temptations: the leap to certainty when a vivid clue lands, and the stubborn refusal to budge when an inconvenient one does. Neither the tan line nor any single fact should move you all the way; it should move you some. Think in probabilities you can adjust, not verdicts you must defend. Belief is a dial, not a switch, and the whole skill is in the turning.

Six — Keep score, and hold loosely

Finally, do the thing Holmes never does and could never bear to: find out how often you are actually right. Write your predictions down before the answer comes — actually write them, because memory is a liar that edits you into a genius after the fact, dissolving your misses and gilding your hits exactly as Watson does for Holmes. The only cure for that flattering forgery is a written record reality can check. And carry every conclusion loosely, with a grip proportioned to the evidence, ready to drop it the moment the world says otherwise. This is the calibrated humility of Chapter 9, brought home: not a lack of confidence, but confidence that has earned its size by keeping count.

The method, walking around

Let me show you the whole thing in motion, in a scene with no corpses in it, because this is where you actually live.

Your partner comes home quiet and distant. Your inner Holmes fires instantly — a single vivid theory, fully formed, emotionally gripping: they're angry with me. Feel how complete it is, how it explains everything, how badly you want to act on it. Now run the method. Observe the absence: if they were angry with me, I'd expect a cold remark, a pointed silence, some aim to the mood — is it actually aimed at me, or just... heavy? Crowd the suspects: they're angry with me; they had a brutal day at work; they're worried about something that has nothing to do with me; they're coming down with something. Base rate: across all the times they've come home quiet, how often was it actually about me? (Honestly? Rarely. Quiet usually means work.) Hunt the no: what would I see if it weren't about me — and could I just ask, instead of interrogating my own anxiety? Move in inches: the quietness nudges "something's wrong" up a little; it does not justify a confrontation. Hold loosely: so I ask, gently, and let their answer move the dial — instead of prosecuting a case I invented in the hallway.

That is the Science of Deduction, restored to what it should always have been. Notice that it made you less certain, slower, more tentative — and far more likely to be right, and far less likely to start a fight over a theory you conjured from a mood. Notice, too, that it is unglamorous to the point of embarrassment. There is no thunderclap. Nobody gasps. You will never narrate it to an admiring Watson. You will just, quietly and unremarkably, understand the people and problems of your life a little more accurately than the person who chose to feel like a genius instead. That is the whole trade the book offers: the flourish, for the truth. Take it. It's the better deal, even though it's the quieter one.

But there is a shadow over this gift, and an honest book has to walk into it. For a method this powerful does not stay politely inside your own skull and your own kitchen. Turned outward — turned upon strangers, citizens, whole populations, by states and corporations armed with the machine of Chapter 8 — the science of reading people becomes something else entirely. It becomes the most formidable instrument of power ever built. And that is the subject we can no longer avoid.