Cortés landed in Mexico in 1519 with five hundred men. The Aztec Empire could field hundreds of thousands.1
There is no version of that fight he wins. So he did not fight it — he spent his first months on the empire's bookkeeping.
Knowing that many smaller Mexican tribes resented the powerful Aztec Empire, he slowly worked to peel them away from their alliances with the Aztecs.1
And the specific move is the thing worth studying, because it is not persuasion:
By filling a tribal leader's ears with horrible stories about the Aztec emperor's plans, for example, he might bait the man into arresting the Aztec envoys on their next visit. That of course would infuriate the emperor, and now the tribe would be isolated and in danger—and would appeal to Cortés for protection.1
Follow the sequence, because every step is done by somebody else.
Cortés supplies a story. The chief takes an action. The emperor punishes it. And the tribe, now exposed, comes to Cortés — who has not attacked anyone, has taken no territory, and has just acquired an ally at the cost of a rumour.
On and on Cortés went with this negative version of the Alliance Game, until the Aztecs' allies had become his.1
He did not build a coalition. He transferred one.
Greene's compression of it:
Breaking your enemy's alliances is as good as making alliances yourself.1
Which is arithmetically obvious and strategically underused. A defection is a two-unit swing — one off their side, one onto yours — where a recruitment is one. And defections are cheaper, because the resentment already exists. You are not creating a motive; you are supplying an occasion.
The toolkit is short and unglamorous:
Your focus here is on stirring up mistrust. Make one partner suspicious of the other, spread rumors, cast doubts on people's motives, be friendly to one ally to make the other jealous. Divide and conquer. In this way you will create a tide of emotions, hitting first this side, then that, until the alliance totters.1
And then the harvest:
Now former members of the alliance will feel vulnerable. Through manipulation or outright invitation, make them turn to you for protection.1
Note that the alliance does not have to be destroyed by you. It has to be made uncomfortable enough that its members start looking around — and the insecurity does the rest.
The chapter's second half inverts the position: what to do when the coalition is on the other side and you cannot break it.
In facing an enemy that is composed of allies, no matter how large or formidable, do not be afraid. As Napoleon said, "Give me allies to fight."1
The reason is structural rather than psychological:
In war, allies generally have problems of command and control. The worst kind of leadership is divided leadership; compelled to debate and agree before they act, allied generals usually move like snails.1
A coalition's size is real and its tempo is terrible, and tempo is what decides most engagements. See Strategy 5: The Command-and-Control Strategy, where divided leadership is named as the cause of history's greatest defeats, and Gallipoli and the Broken Chain for what it costs.
Then the two targeting rules, and they are the most portable thing on this page:
When fighting large groups of allies, as he often did, Napoleon always attacked first the weak link, the junior partner. Collapse here could make the whole fabric of the alliance fall apart.1
He would also seek quick victory in battle, even a small one, for no force is more easily discouraged by a defeat than an allied one.1
That second one is the sharper observation. A coalition is held together by each member's belief that joining was correct. A small early defeat does not damage a coalition militarily; it damages the reasoning that produced it, and every partner immediately begins recalculating in private.
The unifying idea across both halves — peeling allies off and attacking a coalition — is that alliances fail at the seams rather than in the middle.
The Aztec tributaries resented the empire before Cortés arrived. The junior partner in a coalition is the one whose interests are least served and whose commitment was always most conditional. In both cases the weakness is not a flaw in the alliance, it is the alliance's own structure: parties with genuinely different interests, held together by a shared situation.
Which is why the technique needs so little force. You are not breaking something; you are declining to hold it together, loudly, near the parties who were already carrying the strain.
And it explains the asymmetry between the two halves of this page. Peeling an ally away is cheap and slow; defeating a coalition in the field is expensive and fast. Both are attacks on the same weakness — that the members' interests only overlap, and the overlap is smaller than the alliance's public position implies.
Which gives a way to read any bloc from outside. Ask what each member would do if the shared situation vanished tomorrow. Where the answers diverge sharply, that is the seam, and it is there whether or not anyone is working on it.
You are facing a bloc — a coalition of teams, a customer and their incumbent supplier, two parties who have agreed a position before you got in the room — and taking it on frontally loses.
Start by finding who is in it for the least. Every alliance has a member whose interests are served worst: the junior partner, the tributary, the one who joined because the alternative was worse rather than because they wanted this. That person is not disloyal; they are under-served, and they know it.
Then look for the resentment that is already there, and do not manufacture one. Cortés's stories worked because the tributaries already hated the Aztecs; a rumour that has to create the grievance as well as trigger it will not hold. You are supplying an occasion, not a motive — and if you cannot find a motive already in place, this strategy is the wrong one.
Then let the alliance punish its own member. This is the whole mechanism and the reason it is so hard to trace. You want an action that their partner will react badly to — a question asked in the wrong room, a term renegotiated, a commitment declined. The damage to the relationship is inflicted by the senior party, and your name is not on it.
Then be the available alternative at the moment of exposure, and be so without triumph. The tribe came to Cortés. Being the obvious safe harbour when someone becomes isolated requires only that you were visibly not the cause.
And if you are on the other side of this — facing a coalition rather than dismantling one — attack the junior partner and take an early, small, visible win. Coalitions are held together by each member's private belief that joining was right, and a first defeat is worth far more against a coalition than against a single opponent, because it starts every partner recalculating at once.
And notice how easily this page reads from the other chair. If you are the junior partner in something, everything above is a description of how you are about to be worked, and the tell is a well-timed story about your senior partner's intentions arriving from someone with nothing obvious to gain.
Cortés's systematic recruitment of Aztec tributary states — above all the Tlaxcalans — is the central military fact of the conquest and is not seriously contested. Napoleon's practice of attacking the weakest coalition partner first is well documented. [POPULAR SOURCE]
Greene's account of the conquest omits its actual decisive factor. The tributary alliances were essential and so was smallpox, which tore through Tenochtitlán in 1520 and killed a very large share of the population including the emperor Cuitláhuac. A chapter presenting the fall of the Aztec Empire as a masterclass in alliance-breaking, with no mention of epidemic disease, is attributing to strategy an outcome that strategy did not produce alone. This is the book's retrospection problem operating by omission rather than by interpretation — everything Greene says Cortés did, he did; it was not sufficient.
And the tributaries are not agents in this telling. The Tlaxcalans had fought the Aztecs for generations and made a considered decision about which of two dangerous powers to align with. Greene's version has them baited into arresting envoys by a Spaniard's stories, which makes them instruments rather than parties. They had their own strategy, and within a generation it had bought them very little — which is the case's real lesson about being recruited off a coalition, and the chapter does not tell it.
The technique is also stated without any mark on it. Spread rumors, cast doubts on people's motives, be friendly to one ally to make the other jealous is a description of deliberate slander among people who are not your enemies, aimed at getting them punished by their own side. The chapter's framing — that this is as good as making an alliance — is presented as efficiency. It is efficient, and it is also the one technique in Strategy 27 whose cost falls entirely on a third party who never entered the contest.
Against Strategy 17: The Divide-and-Conquer Strategy this is the same operation ten chapters later and Greene does not cross-reference it. Strategy 17's principle is that joints are the weakest point of any structure and that you attack the emotional joint rather than the argument; this page applies it to the joints between allied parties. The two chapters are one idea at two scales, and the vault's Divide and Rule and the Central Position already holds the Roman version — divide et impera against the Latin League in 338 BC.
Against The Alliance Game this is Chin's second move given its own page: ally with the distant, attack the near, and where the near has a protector, dissolve that relationship first with dissension, rumour and money. The positive and negative versions of the game are the same skill, and Greene says so — breaking your enemy's alliances is as good as making alliances yourself.
Against Cortés and the Burnt Ships and Cortés and the Exposed Salient this is the third Cortés treatment in the build, and the three together describe one continuous campaign that Greene splits across chapters on death ground, flanking, and alliances: he coerced his own men into commitment, destroyed his internal opposition through their own faction, and dismantled his enemy's coalition — all in the same eighteen months, with five hundred men.
Psychology → The grievance is already there; you are supplying the occasion. The technique fails without pre-existing resentment, which is why Cortés targeted tributaries rather than allies of the Aztecs. The vault's Family System Roles holds the general form — a system under stress allocates roles, and the members carrying the least favourable ones are held there by the situation rather than by consent.
The insight the pairing produces: an alliance's stability is set by its worst-served member's alternatives, not by its members' agreement, which means the parties who look most reliably committed are frequently the ones with nowhere else to go. This inverts how coalitions are usually read from outside — quiet compliance is evidence of constraint rather than of loyalty, and it is precisely the profile of the member who will move first the moment an option appears. And it gives the defensive move, which nobody in this chapter states: if you lead a coalition, the partner you should be spending attention on is the one who never complains.
History → Tempo is what coalitions cannot buy. Compelled to debate and agree before they act, allied generals usually move like snails. Set that beside Strategy 14: The Blitzkrieg Strategy, whose entire argument is that speed defeats mass by acting inside the opponent's decision loop.
Together they produce a claim neither makes alone: a coalition converts material superiority into decision latency at a fixed exchange rate, and the rate worsens with each additional partner. Every member added brings resources and adds a consultation. Which explains why the historical record is so unkind to grand alliances against single competent opponents, and gives a usable rule for anyone joining one: the coalition is stronger than you and slower than your opponent, and if the contest is decided by speed you have joined the losing side by joining at all. Napoleon's "give me allies to fight" is not bravado; it is a statement about arithmetic.
A defection is worth twice a recruitment and costs a fraction as much, because the resentment already exists and you only have to supply the occasion — and the punishment is administered by the target's own partner, so your name never appears in it. The same page read from the other chair is a description of how you are about to be worked.