Sit down and think through how the next big thing goes.
You will find that it goes well. Not suspiciously well — realistically well, with a couple of setbacks in it for texture, both of which you handle. There is a scene where the difficult conversation happens and you say the right thing. There is a moment near the end where it becomes clear it has worked.
Now notice what you did not do. You did not think about the version where the difficult conversation goes badly and the person tells four other people. You did not price the eleven months of nothing in the middle. You did not name anyone whose interests your success damages.
You have been imagining, not planning. And the giveaway is that it was pleasant.
Most people believe that they are in fact aware of the future, that they are planning and thinking ahead. They are usually deluded: What they are really doing is succumbing to their desires, to what they want the future to be. Their plans are vague, based on their imaginations rather than their reality. They may believe they are thinking all the way to the end, but they are really only focusing on the happy ending, and deluding themselves by the strength of their desire.1
This is the distinction that makes the idea useful, and it is easy to lose.
Optimism is a belief about probability — you think the good outcome is more likely than it is. It is corrigible in principle. Show someone base rates and they can update.
What Greene is describing is not a probability error. It is a substitution of one mental activity for another. The person has not overestimated the odds of the happy ending; they have never generated any other ending to assign odds to. Their plan contains exactly one branch and they experienced building it as diligence.
That is why the failure is invisible from inside. Overconfidence feels like confidence and you can at least be asked to justify it. This feels like having thought about it, because you did think about it, for a long time, in detail. The detail is the trap: a richly imagined single future is more convincing than a sparsely imagined set of several, and it is the one your mind will produce if you do not force it to do otherwise.
Greene's phrase for the driver is "the strength of their desire."1 The wanting is what selects the branch, and then the branch gets elaborated until it looks like analysis.
The worked case is Athens, and it is chosen well.
The Athenians attack Sicily "believing their expedition would bring them riches, power, and a glorious ending to the sixteen-year Peloponnesian War."2 A glorious ending. That is the film they were watching.
What they did not consider, in Greene's list:
The Sicilian expedition was a complete disaster, leading to the destruction of one of the greatest civilizations of all time. The Athenians were led into this disaster by their hearts, not their minds. They saw only the chance of glory, not the dangers that loomed in the distance.2
The four omissions have a shape worth naming. Not one of them is obscure. Distance, defenders fighting at home, enemies coalescing, forces stretched — these are the first four things you would list if someone asked you to list them. They were not missed through ignorance. They were never requested, because nobody in the room was running the query. They were running the ending.
⚠ This same expedition is used at Law 36 to prove a different point. Both halves belong in the vault; the tension is documented at Law 36 rather than resolved here.
Halfway through the Bismarck section, in the margin, with no comment before or after:
He who asks fortune-tellers the future unwittingly forfeits an inner intimation of coming events that is a thousand times more exact than anything they may say.3
That is Walter Benjamin, sitting inside a law entirely about foreseeing the future, arguing that the pursuit of foresight destroys a better faculty you already have.
And it is not a soft disagreement. Benjamin's claim is that the seeking is the damage — that the act of going to be told the future costs you an intimation more accurate than any answer you could receive. Applied to this chapter, it says the reader who sits down to plan all the way to the end is trading something for the plan.
Greene never answers it. He never acknowledges it. It is printed and left.
What makes this more than a curiosity is that Benjamin's position and Greene's diagnosis are closer than either would like. Both say that what people call thinking about the future is a substitute for something better. Greene says the substitute is desire wearing the costume of analysis. Benjamin says the substitute is an oracle wearing the same costume, and that the thing displaced is a form of knowledge that does not present itself as prediction at all.
Where they split is the remedy, and the split is total. Greene's fix is more deliberate planning, harder and further out. Benjamin's implication is that deliberate planning is itself the oracle-consulting behaviour, and that the fix is to stop asking. On Benjamin's account, a reader who follows Law 29 diligently forfeits the faculty that would have served them better.
One of the two is wrong, and the chapter that prints both does not notice there is a question.
Between the Pacific and the beheading, in the margin, one more:
Look to the end, no matter what it is you are considering. Often enough, God gives a man a glimpse of happiness, and then utterly ruins him.7
The Histories, named as a work — which is rare in this book, and characteristically it happens in the margin rather than the body.
The sentence contains a claim the chapter's psychology cannot hold. Greene's mechanism is internal: the happy ending is something you generate, out of your own desire, and the fix is to survey harder. Herodotus's is external and adversarial: the glimpse is given, and it is given by something that intends the ruin. The pleasant vision is not your production. It is bait.
Which turns the same advice inside out. On Greene's account, look to the end means look further along the branch you are already on. On Herodotus's, it means treat the good feeling itself as evidence of hostile attention — the very pleasantness that Greene's method uses as a diagnostic is here the point of contact with whatever is arranging your defeat.
Greene keeps the imperative and drops the theology, which is reasonable. But the theology was carrying the operative half: it explains why the vision arrives, and Greene's version has no account of that at all beyond wanting. A man who wants something does not thereby see it in detail. Something has to supply the detail.
Clausewitz, also marginal, near the top of the chapter:
There are very few men—and they are the exceptions—who are able to think and feel beyond the present moment.4
Read against the Keys, this is a stronger claim than Greene is making, and it is doing something to the chapter's premise.
Greene's argument is that foresight is a trainable advantage — rare, therefore valuable, therefore worth acquiring: "Because most people are too imprisoned in the moment to plan with this kind of foresight, the ability to ignore immediate dangers and pleasures translates into power."5 The whole value proposition depends on the scarcity being a skills gap you can cross.
Clausewitz's phrasing — very few, and they are the exceptions — reads as a claim about distribution, not about training. Exceptions are not people who did the exercise.
Greene resolves this by not addressing it, and by immediately reframing the capacity as something you can train oneself into.5 But the marginal he chose to print says the opposite, and the difference matters to anyone deciding whether to attempt the thing.
The chapter opens the Keys with the Greek cosmology: the gods see everything to come, men are "trapped in the moment and their emotions," and heroes like Odysseus who plan several steps ahead "seemed to defy fate, to approximate the gods."6
It is a good image and it does a specific piece of psychological work. It tells you that the faculty is the same faculty, differing only in degree — gods have more of it, Odysseus had some, you can have some. Foresight becomes a quantity.
The Athenian case sitting a few paragraphs later says something the frame cannot accommodate. The Athenians did not have less foresight than the gods. They had a full, vivid, detailed vision of the future, held with great confidence. What they had was the wrong kind, not a smaller amount — a produced future rather than a surveyed one. Between watching a film and running a query, there is no continuum. You are doing one or the other.
The godlike framing therefore makes the actual instruction harder to find. If the deficit is quantity, you respond by thinking about the future more, which is exactly what the Athenians did.
You have twenty minutes and something you have been turning over for weeks.
Write down how it goes. All of it, to the end. You will find this easy and it will take four minutes, and when you read it back you will feel slightly better than you did before you started. Note the feeling. That is the tell.
Now the part that is not pleasant. Underneath, write the same story again — same start, same first move — and this time it does not work. Not catastrophically. It just fails in the ordinary way things fail: slower than expected, one person turns out to be against it, the money is late. Take it all the way to the end too. What are you holding in month fourteen of the version that does not work?
Most people cannot finish this. The pen stops around the third obstacle and the mind slides back to the good version, and it slides so smoothly that you will not notice you have switched unless you are watching for it.
If you get to the end of the second story, ask the four questions Athens did not: How far from home is this? Who is defending? Who else joins in against me? What else am I holding up at the same time?
Then read both stories and notice that only one of them cost you anything to write.
Strongest evidence. The diagnosis is unusually falsifiable for this book: it predicts that people's plans contain one branch, that the branch is desirable, and that they will report having considered alternatives. Anyone can check it on themselves in ten minutes, and the Athenian case supplies a public instance where the omissions were obvious rather than technical.
Tension — the chapter prints its own refutation and leaves it. Benjamin at l.4309 against the entire law.3 Documented above.
Tension — the marginal Clausewitz undercuts the trainability premise. "They are the exceptions."4 Greene needs the capacity to be acquirable and prints a claim that it is distributional.
Tension — the same Sicilian expedition proves the opposite thing at Law 36. Flagged, not resolved; the do-not-merge pair is built at both ends.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — the Athenian account cites no Thucydides, no source of any kind, in a chapter whose whole argument rests on it. 🚩 REPLICATION RISK does not apply directly, since Greene cites no psychological research; the mechanism as described has clear affinities with the planning-fallacy and motivated-reasoning literatures, and none of that is invoked here.
Open questions. Is the substitution avoidable, or only detectable? Every remedy on offer — premortems, second stories, adversarial review — works by importing an outside process, which suggests the faculty cannot correct itself from inside. And: if the pleasantness of the exercise is the reliable tell, what is the tell for a plan that is genuinely good and therefore also pleasant?
Greene has a whole book on this, and it is the later one.
The Laws of Human Nature's treatment of shortsightedness is developmental and physiological — the mind's default time horizon is short because it was built for short horizons, and the corrective is a practice. The 48 Laws version is moral and competitive: most people are deluded, and their delusion is your opportunity. Same phenomenon, two registers. See The Success Delusion, which is the closest neighbour and takes the third angle again — what winning does to your estimate of the material's resistance.
Set those three beside each other and something falls out that none of them states. The Success Delusion is the Happy Ending Delusion arriving later, after the first win, when the block of marble has apparently shrunk. The mechanism is identical — a desired reality substituted for a surveyed one — and the only difference is that success supplies the evidence that makes the substitution feel warranted. Which means the two are not separate failure modes. They are one failure mode at two points on a career, and the second is harder to catch because it comes with a track record.
Against the vault's Chanakya material, the contrast is sharper still. Realistic Thinking: Beyond Optimism and Pessimism rejects the optimist/pessimist binary outright and installs a third register. Greene never offers the third option in this chapter; his opposition is deluded-desire versus godlike-vision, which is the binary with the mortal end relabelled. The Chanakya frame is the more usable one precisely because realistic is a stance a person can occupy, and godlike is not.
Behavioural mechanics — the delusion, operationalised as a target profile. (Mandatory psychology → behavioral-mechanics handshake: this page describes a mechanism with a direct tactical counterpart.)
The Decoyed Object of Desire is the technique side of what this page describes as a private cognitive failure. In 1711 Marlborough needs a French fort destroyed and cannot destroy it himself without announcing which road he wants. So he takes it, garrisons it, holds it as though it matters to him — and when the French retake it they demolish it themselves, reasoning that the duke must have wanted it for something. He got his enemy to do the work and to feel clever about it.
Nothing false was ever stated. The French inference was sound; the premise had simply been arranged. If a person's plan is generated by their wanting rather than by their surveying, then installing the wanting installs the plan — you do not need to deceive them about any fact, because they will supply the missing branches themselves, favourably, and experience it as their own judgement.
That is the handshake, and it changes the defensive picture. As a psychology page, the happy-ending delusion looks like a self-inflicted problem with a self-administered fix: run the second story. As a behavioural-mechanics problem it looks entirely different, because the operator's move is upstream of the fix. Someone who has arranged what you want has already shaped what your careful, honest, second-story exercise will produce — you will imagine failure modes of their scenario, which confirms the scenario is the thing under consideration.
The insight neither domain generates alone: the standard defence against manipulation is scepticism about claims, and the standard defence against self-delusion is rigour about scenarios. Neither touches this. The vulnerability is not in what you believe or in how carefully you plan; it is in who set the agenda you are being careful about. The only question that reaches it is whose interest is served by my wanting this, and that question is not a planning question at all — which is why a chapter about planning to the end cannot produce it, and why the reader who follows Law 29 diligently is a better-prepared mark, not a worse one.
Story craft — the delusion built on purpose, as a load-bearing beat. False Victory Midpoint Architecture names a structural move screenwriters make deliberately: halfway through, the protagonist obtains what they appeared to want, feels at rest, and the audience signs on to that obtaining as the victory — Jack and Rose in the steamed car, with the iceberg six minutes out. The beat works because "what they obtained was not actually what the story needed them to have."
That is the happy ending delusion, reverse-engineered by people who have to produce it on demand. And what the craft version knows that the psychology version does not is where in the sequence it goes. It is not a beginner's error and it does not sit at the start. It is a midpoint phenomenon: it requires a genuine prior success, real effort already spent, and a plausible reason to believe the hard part is behind you. Athens had won for sixteen years before Sicily. Balboa had crossed the isthmus. Neither delusion was available to a person who had not already achieved something.
The insight neither domain produces alone: Greene diagnoses the happy ending as a failure of rigour, correctable by planning harder. The craft frame says it is a failure of position — it happens at a specific structural moment, to people whose recent evidence genuinely supports it, and rigour applied at that moment will be rigour applied to the wrong question. Which supplies something the psychology page cannot generate on its own: a schedule. If the false victory is a midpoint beat, then the second-story exercise is close to worthless when you are anxious and starting out, and near-mandatory the week after something works. Nobody runs it then. That is when it is the only thing worth running.
Sharpest implication. The reliable signal that you are not planning is that the exercise was pleasant and quick. A real survey of the future is unpleasant, because it requires generating outcomes you do not want and then staying with them long enough to price them. If your plan came easily and left you feeling better, you produced a film. The four Athenian omissions were not hard questions; they were questions nobody asked, because asking them would have ruined the afternoon.
Generative questions.