In the spring of 1862 the Confederacy was close to losing the war. It had lost a series of important battles, its generals were squabbling, morale was low and recruits were scarce. A large Union army under General George McClellan was moving on Richmond, and if Richmond fell the South would have to surrender.1
Stonewall Jackson had 3,600 men in the Shenandoah Valley and a modest job: defend the valley.
Over the next three months he diverted more than 60,000 Northern troops, bought the South enough time to organize Richmond's defense, and altered the course of the war.1
Robert Greene uses him as Strategy 10's first case because Jackson had no strength to speak of and won anyway — on perception alone.1
The lever was psychological and it was personal. Jackson had been McClellan's classmate at West Point, and he knew that underneath the brash, talkative exterior the man was basically timid — overly anxious about his career and about making any mistake.1
McClellan had 90,000 men ready to march on Richmond, nearly double the available Confederate forces. But Jackson knew this cautious man would wait until his army was overwhelming, and specifically that he wanted the extra troops Lincoln had promised him.1
That gave Jackson his target. He could not beat McClellan. He didn't need to. He needed Lincoln to see danger somewhere else, because Lincoln would not release those reserves while a threat existed near Washington.1
On 22 March Jackson's spies reported that two-thirds of the Union force in the valley was heading east to join McClellan, with another army near Washington under McDowell soon to follow.1
Jackson marched north immediately and attacked the Union troops still in the valley at Kernstown. The fight was fierce and by day's end his men were forced to retreat. Outnumbered nearly two to one, they had taken terrible casualties. To his own soldiers it looked like a defeat, even a disaster.1
Greene's line: Jackson, always a hard man to figure out, seemed oddly satisfied.1
Days later the news he'd been waiting for arrived. Lincoln had ordered Banks's army back to the valley and McDowell's army to stay put.1
The mechanism is the chapter's first method — the bold maneuver — operating exactly as designed.
Kernstown got Lincoln's attention and made him worry. Only a little, but enough. Lincoln did not know what Jackson was up to or how large his force was, and he wanted the Shenandoah pacified before releasing anyone.1
Greene identifies the precise inference Jackson exploited: Lincoln and McClellan could not imagine that anyone would be so stupid as to send only 3,600 men against a Union stronghold.1 The attack was therefore read as evidence of hidden strength. The audacity was the signal, and its cost — real casualties, a lost field — is what made it credible.
McClellan, wanting certainty, agreed with the logic even though he had the men to march on Richmond immediately.
Jackson then retreated south and lay low for weeks. By early May, believing the valley secure, Lincoln sent McDowell toward Richmond and prepared to move Banks.1
Jackson was ready with the chapter's third method. He marched his army in a completely bizarre fashion — first east toward McDowell, then back west into the valley. Greene notes that not even his own soldiers knew what he was doing.1
Mystified, Lincoln imagined — but could not confirm — that Jackson was moving to attack McDowell. He halted McDowell's march south again, kept half of Banks's army in the valley, and sent the other half to defend against a threat that did not exist.1
The Union's plans, which had looked perfect, were suddenly in disarray with troops too scattered to support each other.1
Now Jackson went in. He linked up with other Confederate divisions and on 24 May marched on the divided, diminished Union force remaining in the valley, maneuvered onto its flank, and sent it in headlong retreat north to the Potomac.1
The pursuit sent a wave of panic through Washington. This now-dreaded general, commanding forces that seemed to have doubled overnight, appeared to be heading straight for the capital.1
Secretary of War Stanton telegraphed Northern governors to muster troops for the city's defense. Reinforcements rushed in. Lincoln, determined to eliminate Jackson permanently, ordered half of McDowell's army west to destroy him and the other half back to Washington.1
McClellan could only agree. Jackson retreated again, his work done.
The reading is stated as a simple truth: what matters in war, and in life generally, is not necessarily how many men you have or how well supplied you are but how your enemies see you.1
If they think you're weak and vulnerable they act aggressively, which by itself puts you in trouble. If they suddenly think you're strong, or unpredictable, or have hidden resources, they back off and reassess. Getting them to change their plans and treat you more carefully can by itself alter the war.1
And the practical corollary for anyone without resources: in any struggle some things are outside your control — you may not be able to assemble a large force or defend every weak point — but you can always affect people's perceptions of you.1
Greene's closing image: like a conjuror, Jackson created a bogeyman out of an army that was, in essence, laughably small.1
Enrichment added 2026-08-09 from the 48 Laws of Power re-ingest. Greene uses this campaign in two books and assigns it two different mechanisms, without noting in either that he has done so.
At Law 17 of The 48 Laws of Power — Keep Others in Suspended Terror: Cultivate an Air of Unpredictability — the same three months are told in five sentences, and the framing is not perception management at all.
Greene's account there: Jackson and "a force of 4,600 Confederate soldiers" were tormenting the larger Union forces in the valley while McClellan, "heading a force of 90,000 Union soldiers," marched south to lay siege to Richmond. Jackson repeatedly led his soldiers out of the valley, then back to it, and "his movements made no sense." Was he defending Richmond? Marching on an unprotected Washington? Heading north to cause havoc? "Why was his small force moving in circles?" The Union generals delayed the march while they tried to work it out, the South poured reinforcements into the town, and "a battle that could have crushed the Confederacy turned into a stalemate."2
And the quotation Law 17 hangs it on, in Jackson's own voice: "Always mystify, mislead, and surprise the enemy, if possible… such tactics will win every time and a small army may thus destroy a large one."3
Two things to record, and the second is the substantial one.
The numbers do not match. This page, from The 33 Strategies of War, gives 3,600 Confederate soldiers. Law 17 gives 4,600. Neither book cites a source. 🚩 SINGLE SOURCE applies to both; [CONTESTED] on the figure.
The mechanisms are different, and they are not two descriptions of one thing. Strategy 10's account is targeted: Jackson knew McClellan personally from West Point, knew he was timid and career-anxious, and aimed a specific perception at a specific third party — Lincoln, who would not release the promised reserves while a threat sat near Washington. The manoeuvre has an addressee and a desired inference.
Law 17's account is untargeted: the movements "made no sense," nobody could work out what he was doing, and the delay came from the resulting confusion. There is no McClellan psychology in it at all, and no Lincoln.
Those imply different skills and different preconditions. Perception management requires knowing your opponent's specific weakness and manufacturing a legible false signal. Unpredictability requires only that the signal be illegible, and works against an opponent you have never met. The first is much harder and much more precise; the second is available to anyone outnumbered.
The campaign supports both readings, which is exactly why the divergence matters — Greene has taken one three-month episode and extracted from it two techniques that require opposite things of the operator, and a reader encountering either book alone would take away a general lesson the other book contradicts.
Filed as a documented divergence, not resolved. See Unpredictability as Terror for the Law 17 side, and Unpredictability as a Subordinate Liability for the closing warning that makes Jackson's rank part of why the illegible version worked for him.
You're in a position with no leverage — a small team against a much larger competitor, an individual against an institution, a junior person whose objections can simply be overruled — and the conventional advice is to build strength you don't have time to build.
Jackson's move is to stop trying to change your actual position and start working on what the other side believes about it. That requires first knowing, specifically, what they're anxious about. Jackson's whole campaign rested on knowing McClellan feared career-ending mistakes and that Lincoln feared an undefended capital. Neither fear was about Jackson.
Then act in a way that is more audacious than your position could rationally support. The credibility comes from the implausibility: nobody believes you'd take that risk without something behind it. Jackson attacked and lost at Kernstown, and the loss cost him casualties and gained him the entire campaign.
Then become hard to read. The bizarre marches worked because uncertainty is more paralyzing than a known threat — a decision-maker who cannot model you must hedge against everything, and hedging against everything means committing to nothing.
The honest caveat is that this is a bluff, and bluffs have a shelf life. Jackson's worked for three months, which was exactly as long as the South needed. It is a way of buying time, not of changing your actual strength, and it fails the moment someone calls it.
The Valley Campaign is among the most studied operations in American military history, and Jackson's diversion of Union forces from the Peninsula Campaign is standard historiography rather than Greene's construction. [POPULAR SOURCE] for the psychological framing.
Open tension: Greene attributes the campaign's success almost entirely to perception management, and Jackson was also, by most accounts, executing genuinely superior operational maneuver — his marches were fast enough that his men were nicknamed "foot cavalry," and he repeatedly achieved local numerical superiority against divided Union forces. The campaign was not only a conjuring trick; it was also good soldiering, and the chapter's framing needs it to be primarily the former.
A second tension, unexamined: Kernstown is presented as a deliberate strategic sacrifice, and Jackson's satisfaction afterward is read as evidence he intended the loss. Some historians read Kernstown as a genuine tactical error — Jackson attacked on faulty intelligence believing the Union force was much smaller — that happened to produce a favorable strategic result. If that's right, the chapter's cleanest illustration of deliberate costly signaling is partly a mistake reinterpreted after the fact.
Set this beside Robert the Bruce and Tit for Tat, the next case in the same chapter, and the two deterrence methods separate cleanly.
Jackson deterred through uncertainty — nobody could tell what he had or what he intended, so the Union hedged against every possibility. It worked fast and it was a bluff; his actual force never grew.
Robert the Bruce deterred through reliable cost — the English knew exactly what would happen if they invaded, because it had happened every time. That took twenty years and it was not a bluff at all.
What the pairing produces: these are opposite information strategies aimed at the same result. One works by being unreadable, the other by being perfectly predictable. Greene lists them as two of five methods without noting they're contradictory in mechanism — and the choice between them turns on whether you have time. Uncertainty buys months. Reliable retaliation takes decades and then holds indefinitely.
Business → The Startup That Appears Larger Than It Is. Small companies routinely manage perception to affect competitor behavior — announcing roadmaps broadly, appearing in markets they've barely entered, signalling capabilities still under construction.
The insight the pairing produces: Jackson's version specifies the condition that makes this work rather than backfire. He was not managing the perception of customers, who eventually check. He was managing the perception of a decision-maker who had to allocate resources under uncertainty and whose cost of guessing wrong was asymmetric. Lincoln could not afford an undefended capital, so any nonzero probability of threat forced a real allocation. Perception management works on people whose downside from being wrong is severe, and fails on people who can afford to wait and see.
Psychology → Ambiguity Aversion. Decision research consistently finds that people prefer known risks to unknown ones, and will pay a real premium to avoid ambiguity — often behaving more cautiously against an unquantified threat than against a quantified larger one.
The insight: this explains why Jackson's incomprehensible marches were more effective than his actual attack. A known enemy of 3,600 men is a problem to be solved with arithmetic. An enemy whose size and intentions cannot be established is a problem no allocation fully resolves, so the rational response is to over-hedge in every direction. Jackson's most powerful weapon was that his behavior did not admit an estimate.
Jackson lost the battle at Kernstown and it was the most valuable thing he did all spring. A defeat that is inexplicable given your apparent resources is read as evidence of resources you don't have — which means, against the right opponent, the cost of the loss is the price of the signal, and it can be worth paying.