A warlord rides back from a banquet with six hundred bodyguards, pleased with himself.
He has spent the afternoon being deferred to. The songs at dinner honoured only him. The new leader seemed nervous, almost obsequious. Balcha has decided he'll be the one calling the shots from here — within weeks his soldiers will march through the capital and this pretender will be in prison or dead.1
Then he comes in sight of his camp.
Where there had been colourful tents stretching as far as the eye could see, there is nothing. Doused fires. Smoke.
Greene gives him one line of interior monologue: "What devil's magic was this?"2
No battle had been fought. Nobody had been killed. In the hours he spent at lunch, an army had walked up a side route he hadn't watched, carrying baskets of gold and cash, and bought every weapon his ten thousand men owned.3
This case only works as an argument if you take Balcha seriously, and Greene does.
In the mid-1920s the Ethiopian warlords were realising that a quiet young nobleman — Ras Tafari, later Haile Selassie — was outcompeting them all. Most "could not understand how this wispy, quiet, mild-mannered man had been able to take control."4
By 1927 Selassie was summoning them one at a time to Addis Ababa to declare loyalty. Some hurried, some hesitated. Only Balcha "dared defy Selassie totally" — a blustery man, a great warrior, who considered the new leader weak and unworthy, and pointedly stayed away.4
When finally commanded to come, he didn't refuse. He did something cleverer: he came at his own speed and with an army of ten thousand, stationed in a valley three miles from the capital, waiting "as a king would." Selassie would have to come to him.5
Then the banquet invitation, and here is the crucial detail:
Balcha, no fool, knew history—he knew that previous kings and lords of Ethiopia had used banquets as a trap. Once he was there and full of drink, Selassie would have him arrested or murdered.6
He knew. He had the correct historical precedent, correctly applied.
So he agreed to come — with six hundred of his best soldiers, armed. He warned them not to drink. He told Selassie that if he wasn't back by nightfall his army had orders to attack the capital.7
Every one of those is a good move. He was defending, competently, against the exact attack that had killed men in his position for generations.
Greene's summary is the sentence the whole page turns on:
Selassie played on the man's wariness, his suspicion that the banquet was a trap—which in fact it was, but not the one he expected.8
The banquet was a trap. Balcha was right about that. He was wrong only about its mechanism, and being right about the category is what made him wrong about everything else — because a correctly identified threat, once defended against, produces the feeling of safety.
Watch what Selassie does with each of Balcha's defences.
Balcha demands to bring six hundred armed men. Selassie answers "with the utmost politeness that he would be honored to play host to such warriors."7 The demand is granted instantly and warmly, which converts Balcha's precaution from a confrontation into a courtesy extended to him.
Balcha threatens to have the capital attacked if he isn't back by dark. Selassie reacts "as if hurt by his mistrust."1
At the traditional singing of songs honouring Ethiopia's leaders, Selassie "made a point of allowing only songs honoring the warlord of Sidamo."1
None of this is resistance. Every single move gives Balcha exactly what he asked for and slightly more — and the cumulative effect is stated plainly:
It seemed to Balcha that Selassie was scared, intimidated by this great warrior who could not be outwitted. Sensing the change, Balcha believed that he would be the one to call the shots in the days to come.1
The six hundred bodyguards weren't overcome. They were relocated — moved three miles from the thing that mattered, by a man who was delighted to have them at his table.
The operation three miles away is the part worth working properly, because every element of it is chosen.
Selassie's ally marched on the encampment "by a side route he had not seen." The army did not come to fight:
Knowing that Balcha would have heard a noisy battle and hurried back with his 600-man bodyguard, Selassie had armed his own troops with baskets of gold and cash.3
That clause contains the whole design. The constraint was noise. A battle would have been audible from the capital, and Balcha's return with six hundred men would have turned a clean operation into a running fight with an unpredictable outcome.
So the method had to be silent, and money is silent.
They had surrounded Balcha's army and proceeded to purchase every last one of their weapons. Those who refused were easily intimidated. Within a few hours, Balcha's entire force had been disarmed and scattered in all directions.3
Three things follow that Greene doesn't spell out.
The soldiers were never the enemy. Balcha's ten thousand men had no stake in whether Balcha or Selassie ruled Ethiopia. They had rifles and no particular reason to keep them if someone offered enough gold. Selassie identified the actual interest of the people holding the weapons, and it turned out to be purchasable — which it would not have been for the man commanding them.
Disarmed is better than defeated. A defeated army has survivors, grievances, and a story. A bought army has scattered men who each individually made a reasonable trade and have nothing to organise around. There is nobody to avenge.
The bodyguard was the price of the deception, and it was cheap. Selassie gave up six hundred armed men in his own capital for an afternoon in exchange for ten thousand disarmed. Balcha experienced that concession as his own victory.
Then the endgame, which Greene renders as a board position: Balcha tried to march south to regroup and found the same army blocking him; the other way out was the capital, which Selassie had garrisoned. "Like a chess player, he had predicted Balcha's moves, and had checkmated him."9 For the first time in his life Balcha surrendered, and agreed to enter a monastery "to repent his sins of pride and ambition."9
Not one bullet fired.
Greene widens it, and the widening is the reason this case sits in Law 3 rather than in a chapter on tactics:
Throughout Selassie's long reign, no one could quite figure him out. Ethiopians like their leaders fierce, but Selassie, who wore the front of a gentle, peace-loving man, lasted longer than any of them. Never angry or impatient, he lured his victims with sweet smiles, lulling them with charm and obsequiousness before he attacked.8
The smoke screen here is not a costume put on for one afternoon. It is the man's entire public identity, maintained for decades, and it worked because it ran against the local expectation of what power looks like. Ethiopians expected fierce. A mild man read as weak, and reading him as weak was the error every rival made.
Balcha's own verdict, given as he entered the monastery, is the epitaph Greene chooses to close on:
Do not underestimate the power of Tafari. He creeps like a mouse but he has jaws like a lion.10
You're going into a negotiation with someone who has more leverage and knows it, and you've prepared your defences.
Before anything else, write down what you're defending against. Be specific — the clause they'll push, the number they'll anchor, the concession they'll demand late.
Now assume you're right about all of it, and ask the Balcha question: if they give me every one of these, what have I stopped watching?
Because that's the move. A skilled counterpart will not fight your prepared positions. They'll concede them, quickly and graciously, and the concessions will feel like wins because you prepared for a fight and didn't get one.
Three specific signals from this case.
Immediate agreement to your protective demand. Balcha asked for six hundred armed men at the table and got a delighted yes. When a precaution you expected to have to argue for is granted warmly, the precaution has been priced and found cheap.
Deference that exceeds the situation. Only songs honouring you. Selassie's apparent hurt at your mistrust. Someone treating you as more powerful than you are is giving you something free, and nothing in a negotiation is free.
The feeling that you're now in control. This is the reliable one. Balcha rode home planning his triumph. If a meeting you entered from weakness ends with you feeling you've taken charge, something happened that you didn't see, and the honest move is to go and physically check the thing you left unattended.
And from the other side: the reason the money worked is that Selassie asked who actually held the weapons and what they wanted, rather than who commanded them. That question is nearly always available and almost nobody asks it. The people executing your opponent's position usually have interests that are not your opponent's.
The vault's practice is to document rather than smooth, so: this is told as a masterpiece, and it is one.
It is also the consolidation of a state by a man who would rule Ethiopia until 1974, and Greene's frame — "the gentle Selassie" destroying his enemy "without firing a single bullet"11 — treats the bloodlessness as the moral of the story. Balcha's ten thousand disarmed men and their scattering in all directions are counted as elegance.
No judgement is required for the technique to be worth understanding. But the technique's admiring presentation is part of what this source is, and it goes in the record alongside the mechanism.
Strongest support. A dated, named, fully-narrated operation with a clear constraint (silence), a clear method (purchase), and a documented outcome — plus, unusually, a quoted verdict from the defeated party.
Tension — the case proves the opposite of Part I's rule. Law 3 Part I insists all is lost if the target has "the slightest suspicion." Balcha was maximally suspicious, correctly, from the start. His suspicion is what Selassie used. The law's stated theory and its best case are in direct contradiction.
Tension — "the paranoid are the easiest to deceive" is doing a lot on one case. It's the generalisation Greene draws in the Keys to Power, and it rests substantially on this anecdote. Plausible, unproven.
Tension — against Law 2's own Emperor Sung. Sung disarmed his generals by making them rich and comfortable openly, at a banquet, with the offer stated out loud. Selassie disarmed an army by buying it secretly during one. Same objective, opposite methods, and both are presented as exemplary. The variable — whether you can afford to be seen doing it — is never named.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no citation for the banquet, the side route, the baskets of gold, or Balcha's closing line. Greene's bibliography for this period is popular anecdote collections. [POPULAR SOURCE].
Open questions.
Greene's Emperor Sung and the Banquet of Resignations — built one law earlier in this same cluster — is the closest structural twin in the book, and the pair is more instructive than either alone.
Two banquets. Two rulers facing armed subordinates who could depose them. Both solve it with money rather than execution. Both hold for a long time afterwards.
And the mechanisms are exact opposites.
Sung's generosity is offered in the room, out loud, to the men themselves. He names the danger, names the remedy, and lets them accept. The transaction is visible and its visibility is what makes it binding — the generals resigned publicly and retired as nobles, so everyone knew the terms.
Selassie's generosity is spent three miles away, in secret, on other people entirely. Balcha is never offered anything. The men who take the money are not the man being defeated.
Same problem, same currency, opposite geometry — and Greene, who tells both stories within a hundred pages, never sets them next to each other. The variable that decides which you can use is whether the subordinate's power is personal (Sung's generals commanded loyalty) or contractual (Balcha's men held rifles for pay). Nobody in the book states it.
→ behavioral-mechanics · The Smoke Screen
The mechanism page holds that the bland, familiar front works because attention is single-channel and the target cannot maintain a second model of a person already filed as harmless.
Balcha sharpens that into something more specific and more useful.
He had not filed Selassie as harmless — he had filed him as a specific kind of dangerous, namely a man who poisons people at banquets. That is a richer model than "harmless," and it made things worse rather than better.
So the smoke screen's real target isn't the absence of a threat model. It's the presence of a wrong one, and a detailed wrong model is far more useful to an attacker than no model at all. It tells you precisely where the sentries are posted.
The insight neither page produces alone: vigilance is not a scalar you can raise. It is a distribution you publish. Balcha's six hundred armed men, his instruction not to drink, his nightfall deadline — every precaution was a public statement of where his attention lived, and Selassie read the statement and walked around it. Which means the counter-technique is not to be more vigilant but to keep the shape of your vigilance unreadable, and almost nothing in Greene addresses that.
→ business · Manufactured Scarcity
The business page treats scarcity as something constructed to move a decision — the supply constraint is real to the buyer and manufactured by the seller.
Selassie ran the inverse operation, and the inversion is what makes the pairing pay.
He didn't manufacture scarcity. He dissolved it. Rifles in an Ethiopian valley in 1927 were scarce, valuable, and the whole basis of Balcha's position — and Selassie's baskets of gold converted them, in an afternoon, into an ordinary commodity with a market price.
The general form: any power resting on a scarce asset is only as durable as the scarcity, and scarcity is a fact about a market, not about the asset. Balcha's men held something valuable right up to the moment someone arrived with enough cash to make it merely expensive.
Which supplies the defence the marketing frame implies but never states. Against manufactured scarcity, the counter is to find the wider market. Against Selassie, the counter would have been to ensure his men held something that couldn't be bought — loyalty, land, a stake in the outcome — which is precisely what Emperor Sung gave his generals and what Balcha never gave his. He paid his army in wages, and anything bought with money can be outbid.
Sharpest implication. Balcha did everything right. He knew the history, identified the correct threat category, brought proportionate force, refused the drink, and set a deadline — and his competence is exactly what defeated him, because a well-constructed defence is a well-drawn map of everywhere you are not. The feeling he rode home with, that he had won and would now be calling the shots, is the technique's signature: the smoke screen doesn't produce inattention, it produces satisfaction.
Generative questions.