Forty-six laws tell you what to do when things are going badly, or when someone is manoeuvring against you, or when you need something from a person who has no reason to give it. This one is about the afternoon after you won. Nothing has gone wrong. Nobody is plotting. You are in the best position you have been in for years — and Greene says this is the moment you are most likely to destroy yourself.
His sentence for it: "The moment of victory is often the moment of greatest peril."1
That is a strange thing to believe, and most people do not believe it, which is the point. Danger is something you scan for when you feel threatened. After a win, the scanning stops. The law is aimed exactly at the gap between when you stop looking and when the thing you should have been looking for arrives.
The Judgment gives a mechanism, and it is worth separating from the general warning about arrogance that surrounds it: "In the heat of victory, arrogance and overconfidence can push you past the goal you had aimed for, and by going too far, you make more enemies than you defeat."2
More enemies than you defeat. That is a ledger, not a feeling.
The claim is that advance has a crossover point. Up to some line, each further push subtracts an opponent. Past that line, each further push subtracts one and adds several — because the people watching revise what they think you are. They stop reading you as someone settling a dispute and start reading you as someone who does not stop. That reclassification is free for them and expensive for you, and it happens whether or not the specific advance succeeds.
So the law is not don't get carried away. It is: there exists a point at which your victories start manufacturing opposition faster than they remove it, and that point arrives before you feel any resistance — because the new enemies are not yet acting. They are only deciding.
The ledger claim has a hole in it, and it is worth opening because the hole is where the law's difficulty actually lives. Greene says you make more enemies than you defeat. He never says who.
The defeated are countable. They have names, they were in the fight, you can list them. The manufactured ones are by definition people who were not in the fight — bystanders, allies of the loser, neutrals with no stake, parties two industries over who watched and drew a conclusion. You cannot enumerate them because the category is everyone who observed and revised.
This asymmetry is the whole reason the law is hard to obey. One side of the ledger is a list; the other side is a population. Nothing in your reporting will ever show the second column filling up, because the entries in it are not doing anything yet — they are only deciding what you are.
And they are deciding a single thing, which is worth naming precisely: not is this person dangerous to me but does this person stop. Someone who stops can be negotiated with, allied with, waited out. Someone who does not stop has to be pre-empted, and pre-empting is what turns a neutral into an enemy.
So the practical form of the law is not make fewer enemies. It is: be visibly the kind of party that halts. The stopping is the signal, and a stop nobody observes buys you nothing.
The Authority is a long passage from the Discourses, and it is more precise than Greene's own framing: "Princes and republics should content themselves with victory, for when they aim at more, they generally lose… when this false hope takes possession of the mind, it makes men go beyond the mark, and causes them to sacrifice a certain good for an uncertain better."3
A certain good for an uncertain better. That is the error stated as a decision, not as a temperament — you are holding something realised and trading it for something probabilistic, at odds you are not calculating because you have just been shown that things go your way.
And Machiavelli names a form of the error that Greene's chapter drops entirely: "The use of insulting language toward an enemy arises from the insolence of victory, or from the false hope of victory."4 Overreach has a verbal version. You gain no territory by it, risk no army, spend nothing — and manufacture the enemy anyway. It is the cheapest possible way to trip this law, and by that fact the most common. The chapter is entirely about advancing too far physically; its own Authority says the mouth gets there first.
Flag the split rather than resolving it. Greene's law is about taking too much. Machiavelli's passage covers taking too much and saying too much, and only the second is available to someone with no army.
The Image is Icarus: "Elated by the triumphant escape and the feeling of flight, he flew higher and higher, until the sun melted the wax of his wings and he fell to his death."5
Read the myth rather than the paraphrase and something shifts. Daedalus told him. The instruction was specific and given in advance: not too high, not too low. Icarus did not lack information. He had it and disregarded it in a state where it felt irrelevant.
Greene's Image attributes the fall to elation — an internal weather condition. The myth attributes it to a disregarded instruction — an event with a timestamp.
This matters practically, because you cannot reliably detect your own elation. It does not feel like elation from inside; it feels like clarity. But "someone told me a limit and I have decided it does not apply today" is a discrete, noticeable act. It has a moment. You can catch it.
⚠ And the chapter's other framing device has the same structure. Queen Tomyris warned Cyrus, in writing, before he crossed the river.6 Both of Law 47's anchoring stories are cases where a specific warning was issued and set aside, and the chapter reads both as emotional overreach. That is a consistent misreading of its own material, and correcting it changes the practical instruction: don't monitor your mood, monitor your treatment of the last person who told you a limit. (Two instances — [n=2], not a pattern claim.)
The Judgment's closing instruction: "There is no substitute for strategy and careful planning. Set a goal, and when you reach it, stop."7
Notice the tense. Set a goal — past, before the campaign. When you reach it, stop — present, in the heat.
The law is therefore not really executable in the moment it applies to. In the moment, everything argues for continuing: you are winning, the resistance is collapsing, the cost of one more push looks trivial against what it might return. There is nothing available in that state that could tell you to stop, because the state itself is the thing distorting the estimate.
What can tell you to stop is a mark set by a version of you who was not winning yet. The law is a pre-commitment device wearing the costume of a self-control instruction. It works entirely through the goal set beforehand, and for anyone who began without naming a stopping point, it supplies nothing — you cannot pass a mark that was never placed.
That reframing also explains the otherwise odd sentence about strategy and careful planning. The failure the law describes is not a failure to plan. Cyrus planned. It is a failure to be bound by a plan once the conditions that made you cautious have gone. The planning matters not because it produces a better route but because it produces a number you can be held to by someone who cannot argue back.
The Reversal supplies the boundary, and it is narrower than the law's rhetoric: "Power cannot expand indefinitely, but must stop at a certain point… the only real danger is going past your mark and creating so many enemies that you are destroyed by them."8
So the prohibition is not on continuing. It is on continuing into new opposition. Finishing the enemy in front of you is a different act from acquiring the next one — which is why this law does not contradict the total-victory instruction it appears to reverse. See Crushing Enemies Completely, where Law 47's Reversal supplies the boundary that chapter never states: completeness applies to the enemy you already have; restraint applies to acquiring further ones.
Two operations, two different rules, and the confusion between them is what the phrase going too far usually conceals. Ask which one is in front of you: am I finishing, or am I starting?
The thing has worked. Better than you projected. Someone says the word momentum and everyone in the room nods.
Before you agree to anything, find the number. You had one — the thing you said you'd be happy with when you started, back when you weren't sure this would work at all. Say it out loud, exactly as you set it, without the upward revision your face wants to give it.
Now check which side of it you are on. If you are short, keep going; the law is not about you yet. If you are past it, you are in the chapter.
Then ask the sorting question — not is there more available, because there always is. Ask: does the next move finish something, or start something? Finishing is completing what you already engaged: the same opponent, the same field, the same dispute. Starting is anyone who was not in this fight until you decided they should be.
If it starts something, stop today and do it Tuesday, deliberately, as its own decision with its own mark. It will still be there. What won't still be there is the state you are currently in, and that is the entire benefit.
And before you leave the room, run the cheap check. Who did you talk about, and how? Machiavelli's version of this failure costs nothing to commit and cannot be withdrawn. If you said anything today you would not have said last week, that is the sun on the wax, and you noticed it in time.
Evidence quality: [POPULAR SOURCE] throughout. Greene's Judgment, Image, Reversal and Authority are all quoted directly.12578 The Machiavelli passage is quoted at length from the Discourses but Greene supplies no book or chapter reference — 🚩 SECONDARY WITHOUT PRIMARY for the attribution, though the sentiment is consistent with Machiavelli's documented position on limited war aims.
Tension: the law's mechanism (a ledger — more enemies made than defeated) and the law's illustrations (an internal state — elation, arrogance, insolence) are not the same theory. One is about how observers reclassify you; the other is about how you feel. A cold, unelated actor can trip the ledger perfectly well by rational expansion, and the chapter has nothing to say about that case.
Open question: the law assumes the crossover point exists and can be located in advance. Cyrus's mark, on Greene's own account, was the conquest of Babylon — but that is identified after the fact, from the outcome. What identifies it beforehand, other than a number you happened to write down?
Within Greene's own corpus this law sits directly against Law 15. That chapter's language is annihilation — no mercy, the venom grows, more is lost through stopping halfway. This chapter's is content yourself with victory. Greene never places them side by side, and the Reversal is the only text that reconciles them: the distinction is between the enemy engaged and the enemy created.
The convergence with the corpus's Fortune material is closer. Law 47's Napoleon marginal — "The greatest danger occurs at the moment of victory"9 — states in six words what Greene elsewhere builds pages around: that good fortune removes the vigilance that produced it. See Good Luck Is More Dangerous Than Bad Luck, the same claim developed as psychology rather than instruction.
The unresolved point across both books is whether Greene thinks the stopping point is knowable or merely assertable. In Mastery and in The Laws of Human Nature he generally holds that the operator can read the field. Here the field is unreadable by construction — the new enemies have not yet moved — and the only available instrument is a commitment made earlier. That is a different epistemology and he does not mark the change.
Hubris and Atē: The Two Falls — the Greek tragic sequence runs hubris → atē → nemesis, and the middle term is the one that matters here. Atē is not punishment; it is the blindness that arrives between the overstep and the consequence, the interval in which the actor cannot see what they have set moving. Greene's law describes exactly that interval without naming it: the new enemies exist, they have reclassified you, and none of them have acted yet, so nothing in your environment reports the change. What the tragic frame supplies that Greene does not is the reason the interval is structurally invisible rather than merely overlooked — the same success that produced the overstep is the evidence the actor uses to judge that things are going well. The feedback signal and the error share a source. That yields something the behavioral frame alone cannot generate: the interval is not a lag in information, it is a period of actively wrong information, which is why a warning delivered inside it (Tomyris's letter, Daedalus's instruction) reads as noise rather than data. It also explains why the law's remedy has to be a pre-set mark. Any instrument that reads the current field is reading the field during atē. The tragic sequence and the ledger mechanism arrive at the same operational conclusion — bind yourself in advance — from opposite ends, one from the phenomenology of the blinded actor, the other from the arithmetic of the watching third parties.
The Kingdom of Wu and the Cost of Overreach — the historical case supplies the variable Greene's chapter suppresses: time. Wu's expansion did not fail at the moment of the overstep; it failed later, and the interval was long enough that the connection between the two was not legible to the people making the decisions. Greene's Cyrus case compresses this — the transgression and the severed head arrive within a paragraph — which makes the law look like it operates on the timescale of a campaign. The historical record says otherwise, and that changes the practical claim considerably. If the enemies you manufacture act within weeks, the law is a caution about a single battle. If they act within a decade, the law is a claim about the accumulated slope of a career, and no individual decision along it will ever look like the one that crossed the line. The insight neither frame produces alone: the ledger metaphor implies a countable moment of crossover, but on a historical timescale there is no crossover event to detect — only a rate. That reframes the mark-setting instruction from identify the line to cap the rate, which is a different and more tractable discipline, and one that survives the absence of any identifiable moment at which you went too far.
The law's testable core is the reclassification claim: that past some point, further advance converts neutral observers into opponents faster than it removes existing ones. That is a statement about third-party perception, and it should be visible in cases where the advancing party gained materially at every step and still ended worse off — which is what makes Cyrus a good illustration and a poor proof, since he lost the battle too.
The open edge is the timescale, raised in the Wu handshake above. Greene's cases all resolve fast. Whether the mechanism operates on a slope rather than at a moment is untested here, and the two readings imply different remedies — a stopping rule versus a rate cap.