Something worked. You have a method, and it has a result attached to it, which is more than can be said for any of the alternatives you might have used. Greene's position is that this is the moment to stop using it: "you owe your success to a pattern that you are apt to try to repeat. You will try to keep moving in the same direction without stopping to see whether this is still the direction that is best for you."1
Take that seriously and notice what it asks. Everything you know about what works came from watching things work. The instruction is to treat the strongest evidence in your possession as the least trustworthy — not because it is wrong, but because of when you acquired it.
Most advice about learning from success says: extract the principle, discard the specifics. This says something harder. It says the extraction itself is unreliable, because the thing doing the extracting has just been rewarded and is no longer a neutral instrument.
The mechanism is stated once, in the Interpretation, and then abandoned: "You come to believe your character is more responsible for your success than your strategizing and planning."2
That is the whole engine and it deserves more than a clause.
Strategy is a variable. It has inputs, it can be wrong, it invites revision — asking did that work because of the plan is a normal question with a normal answer. Character is not a variable. It is a constant you carry between situations, and a constant does not get audited after a win. If your success came from your judgment, your nerve, your read on people, then there is nothing to update. The situation changed; you did not; you are still the person who wins.
So the pattern does not persist through stubbornness. It persists because a successful outcome quietly reclassifies what kind of thing you did — from a choice that could have been otherwise into an expression of who you are. And you cannot revise an expression of who you are by looking harder at the evidence, because it has stopped being the kind of claim evidence bears on.
The practical marker is verbal and easy to catch. Listen for the shift from we tried X to this is how I operate.
Greene's frame for power is tempo: "Those who succeed at the game are the ones who control the patterns and vary them at will, keeping people off balance while they set the tempo. The essence of strategy is controlling what comes next."3
Set that beside the repetition warning and two distinct failures appear, running together under one heading.
Fit failure. The situation has changed and the method no longer suits it. The remedy is to check the situation, and the signal that you need to is external — resistance, a worse result, something not landing.
Predictability failure. The method still works perfectly and has become legible. Anyone who has watched you three times can now stand where you are about to be. The remedy is to vary while winning, and there is no external signal at all, because the results have not deteriorated yet. By the time they do, the counter is already built.
These are not the same instruction. The first says adapt when it stops working. The second says change it while it is working, which is a far stronger and far more costly demand, and it is the one that most people never execute because nothing in their environment ever asks them to.
Greene needs the second and argues mostly for the first. The distinction is worth holding: ask whether you are repeating something that no longer fits, or repeating something that fits so well it has become an address.
Greene names two ways elation breaks tempo, and the second is usually read as a footnote to the first: "success tends to go to your head and make you emotional. Feeling invulnerable, you make aggressive moves that ultimately undo the victory you have gained."8
The Interpretation is more specific about the emotion, and the specificity matters: success "makes you feel invulnerable, while also making you more hostile and emotional when people challenge your power."2
Hostile to challengers. Not merely confident — actively unpleasant to the people whose job is to tell you the pattern has stopped fitting.
That closes the loop, and it is why the two arms are not independent. The first arm makes you repeat the method. The second arm punishes the person who says so. An organisation around a winner learns within about two attempts that raising the question is expensive, and after that the winner has an environment that supplies only confirmation — not through conspiracy, just through ordinary people declining to pay a cost for no benefit.
So the repetition trap does not have to survive scrutiny. It has to survive one uncomfortable conversation, once, early, and then it is unopposed. Cyrus "paid no heed to the warnings of oracles and advisers";4 the more interesting question is how many stopped warning him, and when.
Law 47's Transgression is usually read as method-repetition, and Greene reads it that way: Cyrus "thought each new war would bring the same result as the one before as long as he used the methods he knew: ruthless force and cunning."4
But the methods worked. The banquet trick at the Araxes — bait the camp, withdraw, return at night — was a complete tactical success against the Massagetai. He got exactly the result the pattern promised.
What failed was not the method. It was the assumption that the result was worth having, and no amount of adaptation would have caught that, because adaptation operates on how you do things. The question the pattern suppressed was upstream of technique: what is this victory for.
Which gives the trap a third face. The pattern that worked does not only carry a method forward. It carries forward the goal structure the method was built to serve, unexamined, because the method's success is taken as confirmation of everything attached to it. See Cyrus and Queen Tomyris for the full sequence.
Greene's remedy is oddly content-free, and the emptiness is the point: "they step back and look where they are going. It is as if their bloodstream bore a kind of antidote to the intoxication of victory, letting them control their emotions and come to a kind of mental halt when they have attained success."5
A halt is not a choice between options. It does not tell you what to do next. Its entire value is in the interruption — in the fact that the interval between the win and the next move contains something other than the momentum of the win.
This is why "simply preaching moderation" fails, which Greene concedes elsewhere in the same chapter.6 Moderation is a position, and positions can be argued against by anyone who has just watched you succeed. A halt is a procedure. It has no content to attack.
And the closing image is exact: "you have to be able to control yourself before you can control the horse."7 Not instead of — before. A sequence, with the self-management as the precondition rather than the substitute. The halt is what buys the ordering.
Vary your patterns, change course, improvise, keep people off balance. Greene treats this as pure gain. It is not, and the cost falls in a place the chapter never looks.
Your opponents read your pattern. So does everyone else. A reputation is a pattern — the thing people rely on when they decide whether to commit resources to you before seeing the outcome. An operator who is genuinely unpredictable is unpredictable to allies, suppliers, staff and anyone considering a long-term arrangement, and those parties respond by shortening their exposure.
So the real instruction has to be narrower than the text: vary the patterns opponents use to anticipate you; hold constant the patterns allies use to rely on you. Those are usually different behaviours — tactics versus commitments — and the chapter's undifferentiated vary at will would burn the second to protect the first.
Greene has the material to make this distinction and does not. It is the same blind spot as elsewhere in the corpus: advice written for a solitary operator, applied by people who need other people to show up on Tuesday.
It worked, and the obvious next move is to do it again, slightly bigger.
Before you do, say out loud what you think made it work. Then listen to the grammar of your own sentence. If it came out as we did X and Y happened, you are still holding a variable and you can carry on. If it came out as I'm good at this or this is just how I work, you have converted a method into a trait, and traits do not get checked.
Then run the two tests separately, because they have different answers. First: has the situation changed — different people, different stakes, different information on the other side? That is fit, and if the answer is no you are cleared on that count.
Second, and this is the one people skip: how many times have they now watched me do this? Three is usually the number at which a method becomes an address. If you are at three and it is still working, that is not a reason to continue. It is the last moment at which changing costs you nothing.
Last, separate what you are varying from what you are promising. Change the route, the timing, the sequence, the approach. Do not change the things people have arranged their own plans around — that is not unpredictability, it is unreliability, and it is spent from a different account.
Evidence quality: [POPULAR SOURCE]. All claims drawn from Law 47's Keys to Power and Interpretation, quoted directly.1235 No empirical support offered for the tempo model; it is asserted.
Tension: the chapter's remedy (vary your patterns) and the chapter's diagnosis (you repeat what worked) are aimed at different problems — one at legibility to opponents, one at fit to circumstance — and the text does not separate them. The distinction changes the timing of the remedy completely.
Open questions: if the attribution shift from strategy to character is the mechanism, is the halt actually sufficient? A pause creates space for reflection, but the shift has already changed what the reflection is about. Greene offers no procedure for restoring the earlier framing, only for creating the interval in which it might occur.
Greene's corpus is itself a pattern library, and this law is where that becomes awkward. The 48 Laws teaches methods; Law 47 says the methods that have worked for you are the likeliest thing to destroy you. The book's structural answer is that its content is principles rather than moves — but the chapters are overwhelmingly built from moves, and the reader's takeaway is a repertoire.
In The 33 Strategies of War he confronts this more directly, where the tempo material is developed at length and variation is treated as the strategic primitive rather than as a corrective. Here it arrives as an antidote to an emotional state, which is a weaker framing of the same idea.
The convergence worth noting is with the corpus's general suspicion of self-knowledge. The attribution claim — that success rewrites what you think caused it — is the same structure as Greene's treatment of vanity elsewhere: the mind's account of its own operation is treated as an output of the situation rather than as a report on it. Law 47 applies it to the one case where you feel most entitled to trust your own account.
The Explore/Exploit Tradeoff in a Creative Career — the formal frame supplies what Greene's tempo language gestures at without stating: repetition of a successful pattern is exploitation, and it is not an error. It is the correct move whenever the environment is stable and the horizon is short. What makes it a trap is a specific asymmetry the tradeoff literature names precisely — the returns to exploitation are immediate and legible, the returns to exploration are delayed and mostly negative, so the two arms are not evaluated on comparable evidence even by an actor who is trying to be fair. That is the insight neither frame reaches alone. Greene attributes the persistence of the pattern to intoxication, an emotional distortion that a sober person could avoid. The tradeoff frame says a perfectly sober actor, updating rationally on the evidence available, will still over-exploit — because the evidence available is systematically biased toward the arm being pulled. Which relocates the remedy. If the cause is emotional, a halt is enough. If the cause is structural, you need an allocation rule set in advance — a fixed share of moves committed to exploration regardless of how well the current pattern is performing — and that is a different instrument entirely, one that survives the good mood the halt is supposed to interrupt.
Conditioning Is the Enemy of Wonder — the creative-practice material describes what a working method does to perception, which is the half of the mechanism Greene's strategic account cannot see. A pattern that has succeeded does not merely get selected more often; it changes what registers as a situation in the first place. The practitioner with a reliable approach stops encountering raw material and starts encountering instances — the thing in front of them arrives pre-sorted into the categories the method operates on. Greene's Cyrus reaches the Araxes and sees another army to be tricked, because tricking armies is what his apparatus renders. The insight the pairing produces: "stopping to see whether this is still the direction that is best for you" presupposes that stopping restores an unfiltered view, and the conditioning account says it does not. What the halt buys is time, not fresh sight. To actually see the situation differently requires an input the method cannot process — an outside reading, an unfamiliar frame, someone who does not know how you usually do this. That converts the remedy from an act of self-discipline into an act of recruitment, which is both less flattering and considerably more likely to work.
The testable claim is the three-repetitions threshold implied by the predictability failure — that a method becomes reliably anticipable to an attentive opponent after a small, countable number of observed uses. Greene never states a number; the case material suggests one is available and nobody has looked for it.
The open edge is the allocation question raised in the explore/exploit handshake. If over-exploitation is structural rather than emotional, Law 47's entire remedy is aimed at the wrong cause, and the correct instrument is a pre-set exploration budget rather than a moment of reflection after a win.