December 1502. The little Adriatic port of Sinigaglia has surrendered to the condottieri — the mercenary captains who were, until recently, Cesare Borgia's own generals and are now conspiring against him. Only the citadel holds out, and its commander says he will hand it over to nobody but Cesare in person.
Cesare sends word that he will arrive tomorrow.
"which was just what the condottieri wanted to hear."1
Their plan is complete. He will be caught between the citadel and their forces ringing the town, with only a small force, since the French troops have left.2 All he has to do is come.
He came.
The condottieri's confidence rested on one estimate: that the departure of the French had left Cesare with a small force.3
It had not. According to Machiavelli, Cesare left Cesena with ten thousand infantry and three thousand horse — and "taking pains to split up his men so that they would march along parallel routes before converging on Sinigaglia."4
That detail is the operation. The army was not hidden; it was distributed, and distribution is what made it invisible to observers counting columns on any single road. Nobody had to be deceived about a fact. The fact was simply never assembled anywhere it could be seen.
The rest was foreknowledge, and Cloulas states its source plainly: "he knew, from a confession extracted from Ramiro de Lorca, what the condottieri had up their sleeve."5
Ramiro de Lorca had been executed days earlier, his mutilated body left in the square at Cesena as Cesare marched south.6 The chapter reports the execution and the confession in separate sentences and does not join them.
Join them and the sequence is: extract the plot from a subordinate, kill him, leave the body in public, then walk into the plot he described.
The public corpse is not incidental to the deception — it is part of it. A man executed in a town square is a man who is finished, and the conspirators had no reason to think he had first been drained. The visible punishment concealed the invisible extraction.
Cesare's decision, in Cloulas's phrase, was to "turn their own trap against them" — the manoeuvre the historian Paolo Giovio called "the magnificent deceit."7
The mechanics are worth having exactly, because each piece does one job.
At dawn on 31 December, Cesare reached the outskirts. His advance guard of two hundred lances, led by Michelotto Corella, took up position on the canal bridge — and "this control of the bridge effectively prevented the conspirators' troops from withdrawing."8
Read that. The first move of the ambush was not an attack. It was closing the exit — before anybody knew a fight was happening.
Then Cesare "greeted the condottieri effusively" and invited them inside the Palazzo Bernardino, which Michelotto had prepared for his use. "Once indoors the men were quietly arrested by guards who crept up from the rear."9
Then, simultaneously, an attack on Vitelli's and Orsini's soldiers in the outlying areas. And that night, while their troops were being destroyed outside, "Michelotto throttled Oliveretto and Vitelli in the Bernardino palace."10
"At one fell swoop, [Borgia] had got rid of his former generals and worst enemies."11
The single most instructive feature of Sinigaglia is the timing, and it is the part that makes it Law 15 rather than a story about a clever ambush.
Cesare could have arrested the captains and dealt with their troops afterwards. He did not. The arrest indoors and the destruction outside happen at the same time, and the bridge was taken before either.
Each of those choices removes one recovery route.
The bridge removes withdrawal — the troops cannot become a surviving force elsewhere. Simultaneity removes reaction — the soldiers outside never learn their captains are taken, so there is no moment where a leaderless army chooses to fight or flee. And the throttling that night removes the captains themselves, so there is no prisoner to be ransomed, exchanged, or rescued.
What is being crushed is not the men. It is every path by which the situation could continue. That is the difference between this and Hsiang Yu's banquet, where the killing was the whole plan and every other route stayed open.
Worth recording plainly: this passage is the only material in Law 15 with a named, dated source. Ivan Cloulas, The Borgias, 1989.12 It also names Machiavelli as the authority for the troop numbers and Paolo Giovio for the phrase "the magnificent deceit."
It sits in the marginal column, beside body text — Hsiang Yu, Empress Wu — narrated verbatim with no citation of any kind.
That is the sixth instance in this build of the same pattern: the margin is where Greene put the material he had actually read. It is now the working hypothesis rather than a curiosity, and Law 15 is a clean example, because the difference in evidentiary quality between the column and the body is not subtle.
Somebody has arranged a situation and invited you into it.
The invitation is plausible — there is a real reason for you to be there, a genuine thing that needs your presence, and the person asking has a legitimate claim on it. That is what makes it work. Cesare's citadel commander really would only surrender to him.
Here is the move worth taking from this, and it is not refuse the invitation.
Cesare went. What he changed was not whether he arrived but what was true when he arrived — the force distributed along parallel roads, and the knowledge of the plot bought days earlier and paid for in a way nobody could see.
So the question when you are being invited somewhere on someone else's terms is not is this a trap. You often cannot know. The question is narrower: what do they believe about my position, and is it current? The condottieri's whole plan rested on one stale estimate about French troops. They never checked it, because it had been true when they formed the plan.
And then the second question, which is about your own exits rather than theirs. Before you walk in, know which bridge you would leave by — because the first thing a competent opponent does is not attack. It is quietly take the bridge, while everyone is still greeting each other effusively.
The strongest evidence here is provenance rather than internal consistency: a named modern historian, citing Machiavelli for the numbers and Giovio for the contemporary characterisation. Nothing else in Law 15 is supported this way.
Tension: the case illustrates a different law from the one it sits in. Law 15 is about completeness — do not stop halfway. Sinigaglia's lesson is about sequencing and simultaneity — the order in which routes are closed. Cesare's completeness was a consequence of his sequencing, not a separate resolve. The chapter's frame cannot see that, because it has only one dial.
Tension: the extraction is reported and not examined. "A confession extracted from Ramiro de Lorca" is the hinge of the entire operation, and neither Cloulas's excerpt nor Greene's placement asks what an extracted confession is worth. Cesare acted on it and was right. That the plot turned out to be real does not make extraction a reliable instrument, and the case's success invites exactly that inference.
🚩 SECONDARY WITHOUT PRIMARY does not apply here in the usual form — the source is named and dated. But the passage is an excerpt Greene selected and placed, and the surrounding chapter supplies no context for it. [POPULAR SOURCE] on Greene's framing; the Cloulas material itself is [PLAUSIBLE — needs corroboration] pending a check against the book.
Open questions. How much did the public execution of Ramiro contribute to the conspirators' confidence — did they read a dead informant as a closed leak? And whether Cesare's parallel-column march was a deception measure or ordinary logistics that happened to conceal: the excerpt says he took pains to do it, which implies intent, but a march that size may have had no alternative.
Against Hsiang Yu, the contrast is exact and the chapter does not draw it. Hsiang's banquet and Cesare's palazzo are the same setup: invite the rival to a friendly gathering, kill him indoors. One fails, one succeeds, and the difference is not resolve.
Hsiang's plan had one step — the sword dance. Everything else was left open: Liu's route out, his army, his ability to negotiate. Cesare's plan had one killing and three closures around it, and the killing came last.
The transferable rule the pair produces: a plan whose success depends on a single decisive moment is a plan that hesitation can destroy. Hsiang hesitated for one beat and lost an empire; there was nothing else holding the outcome in place. Cesare's outcome was held by a bridge, a simultaneous attack, and a distributed army — so his personal resolve at any one instant mattered much less. That is a better lesson than be ruthless, and it is available in Greene's own two cases.
Behavioral mechanics — Foreknowledge Has Only One Source
Law 14's unstated epistemology: foreknowledge cannot come from spirits, gods, precedent or astrology — only from "men who know the enemy situation."
Sinigaglia is that doctrine executed and it exposes the doctrine's cost, which Law 14 never prices.
Cesare's foreknowledge came from exactly the prescribed source: a man who knew. It was obtained by extraction, and the man was then executed. The doctrine says buy the future from people; the case shows what the purchase can cost the person sold.
And it sharpens the doctrine's real weakness. Law 14's page argued that the operative variable is not source-type but cross-checkability. Cesare had one source and no way to verify it — and the reason he could act on it confidently is not epistemic at all. It is that Ramiro was speaking under conditions where lying had no upside, which is a claim about incentives, not about knowledge. So the two pages together locate what actually made this intelligence usable: neither its human origin nor a second channel, but the fact that the informant's situation made the truth his only available move. That is a third validity criterion, and neither Sun-tzu nor Greene has it.
Psychology — False Confession Psychology
That page's uncomfortable finding: psychologically ordinary people confess to crimes they did not commit, in ordinary interview rooms — and the conditions that most reliably produce a confession are the same conditions that most reliably degrade the truth of what it contains.
Now read Cloulas's clause again: "a confession extracted from Ramiro de Lorca." Extracted, from a man who was executed immediately afterwards.
The pair produces the case's real epistemic problem, which the history conceals by ending well. Cesare acted on that confession and it was correct — the conspirators were waiting exactly as described. But that outcome is not evidence about the method. Under the interrogation literature's finding, a man in Ramiro's position would produce some account regardless, and the account would be shaped toward whatever his interrogator appeared to want. Cesare already suspected the condottieri. He received confirmation.
Which means Sinigaglia is a case where the extraction happened to be right and could not have been known to be right in advance — and it is remembered as brilliant intelligence work precisely because it worked. The successes of coerced extraction are visible and its failures are the ones where the army marched somewhere nobody was waiting, and those do not get called "the magnificent deceit" or get written down at all. That asymmetry is why the technique retains a reputation the evidence does not support, and it is the same selection effect this build has now logged three times in Greene's own sorting of figures by outcome.
Sharpest implication. What made Sinigaglia complete was not ruthlessness applied harder but routes closed in the right order — bridge, then arrest, then simultaneous destruction, then the killing last. Completeness was an output of sequencing. A law that only knows how far to go, and not in what order, cannot tell the difference between this and a banquet with a botched signal.
Generative questions.