Greene's image for this whole technique is a sheepskin.
A sheep never marauds, a sheep never deceives, a sheep is magnificently dumb and docile. With a sheepskin on his back, a fox can pass right into the chicken coop.1
Notice what the fox is not doing. He isn't hiding. He isn't creating a distraction at the far end of the field. He walks in, in plain view, and is looked at — and the looking finds nothing worth reporting.
That's the whole difference between this and everything in Part I. A decoy makes you look somewhere else. A smoke screen lets you look right at the thing and see nothing.
Greene marks the distinction himself, and it's the most useful sentence in Part II:
In the decoy and red herring devices discussed in Part I, you actively distract people; in the smoke screen, you lull your victims, drawing them into your web. Because it is so hypnotic, this is often the best way of concealing your intentions.2
Two different verbs and two different economics.
Distraction is active and expensive. You have to build the decoy, maintain it, and keep it interesting enough to hold attention. Marlborough had to garrison a fort. Bismarck had to give a speech he'd have laughed at.
Lulling is passive and nearly free. You supply nothing. You simply present as ordinary, and the target's own pattern-recognition does the work of classifying you as not-worth-watching, then stops updating.
And the second one is more durable, because a decoy can be seen through in a moment of insight, while a lull has to be interrupted — which requires something to arrive from outside and jolt the observer's category.
The most counterintuitive claim in the law, stated flatly:
If you believe that deceivers are colorful folk who mislead with elaborate lies and tall tales, you are greatly mistaken. The best deceivers utilize a bland and inconspicuous front that calls no attention to themselves. They know that extravagant words and gestures immediately raise suspicion. Instead, they envelop their mark in the familiar, the banal, the harmless.3
Then the operational rule:
The grayer and more uniform the smoke in your smoke screen, the better it conceals your intentions.2
The popular image of the deceiver — the charming rogue, the elaborate liar — is a description of the ones who got caught. Elaboration is memorable, and memorable material gets rehearsed, compared, and eventually contradicted. A person who tells you a fascinating story has given you something to check.
Blandness gives you nothing to check because it gives you nothing to remember.
The mechanism underneath is a claim about how attention works, and Greene states it plainly:
This derives from a simple truth: people can only focus on one thing at a time. It is really too difficult for them to imagine that the bland and harmless person they are dealing with is simultaneously setting up something else.2
The second half is the load-bearing part, and it isn't about attention capacity — it's about imaginative capacity.
The target isn't failing to notice. They're failing to entertain the hypothesis. Holding "this dull person is running an operation" requires actively maintaining a second model of someone whose first model is already filed as uninteresting, and nobody has the budget for that against everyone they meet.
So blandness doesn't defeat observation. It defeats the allocation of observation, which is upstream and much cheaper to attack.
Greene's sharpest and least intuitive claim comes out of the Selassie–Balcha case, and it deserves isolating:
The paranoid and wary are often the easiest to deceive. Win their trust in one area and you have a smoke screen that blinds their view in another, letting you creep up and level them with a devastating blow.4
Work through why this should be true rather than backwards.
A wary person is not scanning everywhere. Wariness is expensive, so it gets spent on a specific threat model — the thing they've decided to worry about. Balcha knew Ethiopian history and knew banquets were traps, so he brought six hundred armed men and watched the food, the drink and the exits.
His vigilance was correct, well-informed, and pointed at the wrong three miles.
And here's the part that makes it worse than ordinary inattention: once the wary person's specific fear has been addressed, they experience relief, and relief is a stronger analgesic than ignorance ever was. Selassie let him bring the bodyguard, deferred to him, gave him top billing in the songs. Balcha didn't merely stop worrying — he concluded he was winning.
Greene's list of the best instruments here is worth keeping:
A helpful or apparently honest gesture, or one that implies the other person's superiority—these are perfect diversionary devices.4
An unguarded person has diffuse, low-grade attention everywhere. A guarded person has concentrated attention in one place and nothing anywhere else — which is a better target, because you know exactly where the attention is and therefore exactly where it isn't.
Set beside the Selassie case, in the marginal column, Greene reproduces a line from the Thirty-Six Strategies via Thomas Cleary:
This means to create a front that eventually becomes imbued with an atmosphere or impression of familiarity, within which the strategist may maneuver unseen while all eyes are trained to see obvious familiarities.5
This is worth flagging as more than ornament.
The corpus-wide 🚩 SINGLE SOURCE flag on this book is serious — Greene's historical material is retold without citation from popular anthologies. But here a Chinese strategic tradition, transmitted independently and reaching Greene through a translator he names, arrives at the identical mechanism and the identical phrasing: a front, an atmosphere of familiarity, and manoeuvre unseen because all eyes are trained on the obvious.
That's genuine corroboration, and it's the strongest available answer to the single-source problem anywhere in Law 3. It doesn't validate Greene's anecdotes. It does mean the mechanism itself was arrived at twice, in unrelated places, by people with no access to each other.
You want to know what's actually happening in your organisation, and the official channels keep telling you things are fine.
Start from the constraint rather than the technique: your attention is already allocated, and someone benefits from where it's pointed.
So audit the allocation instead of the information. Write down the three things you've been worried about this quarter. Then ask who knew you were worried about them. If the answer is everyone, because I said so in a meeting, you have published your threat model, and everything not on that list has been running unobserved with your blessing.
Now go looking in the boring places. Not where the drama is — drama is where attention already goes and therefore where nothing hides. The unexciting recurring process. The vendor nobody asks about. The person whose updates are always fine and always brief.
And when relief arrives, treat it as data. The moment a specific fear you'd been carrying gets resolved — the reassuring answer, the deferential gesture, the concession you didn't expect — is precisely the moment Balcha was riding back to his camp planning his triumph. Relief is when the second look is cheapest to skip and most valuable to take.
One rule for the other direction, if you're the one who needs room to work: don't build a cover story. Just be uninteresting on the specific axis people are measuring, and genuinely so. A cover story is a decoy and it can be checked. Being ordinary can't.
Greene names the price and it's not what most readers expect:
It takes patience and humility to dull your brilliant colors, to put on the mask of the inconspicuous.6
Humility is the operative word, and it's a real cost.
The technique requires you to be consistently underestimated, by people who are wrong about you, for as long as the operation runs — and to not correct them. Every meeting where someone talks over you, every time you're the least impressive person in the room and could easily not be, is the price being paid.
Most people cannot do this, which is why the technique keeps working. The desire to be seen accurately is stronger in most of us than the desire to be positioned well, and Greene's consolation — "it is often your unreadability that draws people to you"6 — is thin, because unreadability that draws attention has stopped being grey.
Strongest support. Two fully-narrated cases with different targets and identical structure (Weil's mark, Selassie's warlord), plus an independent traditional source naming the same mechanism, plus a stated cognitive constraint that is at least plausible on its face.
Tension — the technique contradicts its own advertised payoff. Greene closes by suggesting unreadability "makes you appear a person of power."6 But the whole method requires appearing bland, harmless and unremarkable. Being read as powerful and being read as negligible cannot both be the goal, and Greene wants both in the same paragraph.
Tension — against Law 6. Court Attention at All Cost will argue that being noticed is the primary currency and obscurity is death. This law argues that being unnoticed is the master technique. Documented pair for the tension log; Greene never reconciles them.
Tension — the paranoid claim is asserted from one case. "The paranoid and wary are often the easiest to deceive" is a strong generalisation resting on Balcha. It's plausible and it's supported by exactly one anecdote, retold without source.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — applies to the anecdotes, though the Thirty-Six Strategies corroboration partially answers it for the mechanism. [POPULAR SOURCE].
Open questions.
Greene's The Performance of Power — one of the superseded 2026-04-24 pages, pending rebuild — treats power as something continuously enacted and displayed.
The smoke screen is the direct negation, and the two sit in the same book.
Displayed power requires being read as powerful; the smoke screen requires being read as negligible. And Greene supplies no rule for which régime you're in, which is the same gap that leaves Law 3 and Law 6 pointed in opposite directions.
The nearest thing to a resolution is sequential rather than logical: be grey while acquiring, be luminous once secured. Selassie was mild-mannered and obsequious right up to the afternoon Balcha's army was bought out from under him, and thereafter he was the Emperor. Greene never states it, but every case in this law fits it — the smoke screen is a technique of the approach, and nobody in the book maintains it after arrival.
→ psychology · Habituation and Deprivation
The psychology page holds that a constant stimulus stops being processed — the nervous system's basic economy is to allocate to change and to write off the steady state.
That's the smoke screen's actual engine, and it reframes the technique from a social manoeuvre into an exploit of something involuntary.
Greene attributes the effect to imagination — people "cannot imagine" the bland person is doing something else. Habituation says it's cheaper and more brutal than that: after enough exposure to an unchanging signal, the signal is no longer reaching the place where imagining happens. The target hasn't considered and dismissed the hypothesis. They have stopped receiving the input that would prompt it.
Which yields a prediction Greene's version can't make and which the technique's own history supports: the smoke screen should strengthen over time, and it should be defeated by interruption rather than by scrutiny. More careful looking at a habituated stimulus doesn't help, because the stimulus has been filtered upstream of looking. What breaks it is change — the grey person doing one uncharacteristic thing.
That's why Selassie's cover held for a career and Balcha's for an afternoon. And it's the practical defence: don't audit the boring channel harder. Perturb it and watch what moves.
→ history · Secrecy and the Formlessness Doctrine
The Chinese strategic material treats formlessness as the absence of a shape for the enemy to plan against — you cannot be countered if there is nothing to counter.
The smoke screen is a different solution to the same problem, and setting them side by side sharpens both.
Formlessness offers no shape. The smoke screen offers a wrong shape — specifically, a familiar and reassuring one. And the wrong shape is operationally superior in one important respect: no-shape is itself detectable. A person who cannot be read reads as unreadable, which is information, and an alert opponent will treat it as a threat signal. Balcha would have been on guard against a cipher.
He was not on guard against a mild, deferential man who sang his praises at dinner.
The insight neither tradition states alone: concealment has two failure modes, and they pull opposite ways. Reveal too much and you can be countered. Reveal nothing and your opacity becomes the signal. The smoke screen threads it by transmitting a full, coherent, entirely misleading picture — which costs more to construct than formlessness and is much harder to detect, because there is no gap where information should be. Greene's grey man is not hiding. He is broadcasting, on a channel nobody bothers to decode.
Sharpest implication. Wariness does not protect you — it relocates your blind spot and tells everyone where it went. Balcha's threat model was historically informed and entirely correct about banquets, and it functioned as a published map of the three miles he wasn't watching. Which means the security you feel after a specific fear is addressed is the single most dangerous moment in any interaction, and it feels exactly like winning.
Generative questions.