Three weak-position moves that turn out to be one principle, and the specific way each of them dies.
In 2001, a political scientist named Ivan Arreguín-Toft published the results of an unglamorous project: he had gone through roughly two hundred years of wars in which one side was overwhelmingly stronger than the other, and counted who won. The headline number was strange enough on its own: the materially weaker side won about thirty percent of the time, which is a remarkable amount of winning for actors who by every conventional measure should never win at all.
But the number wasn't the finding. The finding, published as "How the Weak Win Wars" in International Security and expanded into a 2005 Cambridge book, was about when the weak win, and it is one of those results that reorganizes how you see everything adjacent to it: independent of regime type and weapons technology, when both sides fought the same kind of war (direct against direct, guerrilla against counter-guerrilla) the strong side tended to win. When the weak side fought a different kind of war than the one being waged against it, the odds flipped. Same actors. Same weapons. Same century. The outcome moved with one variable: whether the weaker side agreed to fight on the stronger side's terms.
Hold that shape in your head, because once you have it, you start seeing it everywhere: in how protest movements outlast police states, in why a joke can do what an army can't, in how a startup nobody has heard of gets its first enterprise customer, and in exactly how each of those moves dies when it's used wrong.
The claim of this essay is that the three classic weak-position moves (the underdog playbook, comedy as a weapon, and borrowed credibility) are not three separate tricks. They are one principle wearing three costumes, and the principle is about cost, not strength. Strength is a stockpile. It sits in one currency: divisions, market share, headcount, budget. Every asymmetric move that actually works does the same thing: it finds a dimension on which the cost ratio runs opposite to the power ratio (where acting is cheap for you and responding is ruinous for them), and then it forces the contest onto that dimension, where the stockpile doesn't convert.
That framing also predicts, with uncomfortable precision, how each move fails. Which is what separates a framework from a slogan.
Arreguín-Toft's result is the underdog playbook stated as a falsifiable finding. The weak actor's job is not to fight harder; it is to refuse the contest as offered. Every hour a guerrilla army spends not massing into conventional battalions is an hour the incumbent's artillery is a warehouse expense rather than an advantage. The strategy is cheap for the weak side (irregular war needs little capital) and hideously expensive for the strong side, which must convert its entire apparatus into something it wasn't built to be, and pay in legitimacy as well as money while doing it.
The civilian version of this finding has an even sharper dataset behind it. In Why Civil Resistance Works (Columbia University Press, 2011), Erica Chenoweth and Maria Stephan compared violent and nonviolent resistance campaigns across the twentieth century, and the summary Chenoweth herself gives is this: between 1900 and 2006, campaigns of nonviolent resistance were more than twice as effective as their violent counterparts. Not marginally better. More than twice.
Their explanation is a cost argument, which is why it belongs at the center of this essay rather than in a footnote. Nonviolent campaigns win because they present fewer barriers to participation (moral, physical, informational) and participation is the resource that actually decides these contests. Anyone can join a strike; joining an insurgency costs your job, your family's safety, and possibly your life. A regime can buy soldiers. It cannot buy a hundred thousand of its own citizens standing in the street, because mass participation is precisely the thing that is cheap for a legitimate grievance and impossible to counterfeit.
One update, and it matters: in "The Future of Nonviolent Resistance" (Journal of Democracy, July 2020), Chenoweth reported that the success rate of mass nonviolent campaigns has declined markedly since 2010, even as the technique became more popular than ever, and even before the pandemic emptied the streets. Her diagnosis names two causes: states responded more savvily, and the movements themselves changed in structure and capability. Read through this essay's lens, both causes are one event: the cost ratio moved. If digital tools made participation cheaper while making surveillance and pre-emption cheaper faster, and made it possible to assemble a crowd that had paid almost nothing to exist, then the signal mass participation once carried got cheaper to fake and cheaper to disperse, which is this frame's prediction of exactly the decline she measured. The playbook didn't stop working. The dimension shifted under it, and some movements kept fighting on ground that no longer favored them, which, you'll notice, is Arreguín-Toft's failure mode arriving on schedule.
In 1998, a group of Serbian students founded a movement called Otpor! ("Resistance!") and set about opposing Slobodan Milošević with, among other tools, ridicule. Street theatre. Absurdist stunts. Mockery deployed with logistical seriousness. Otpor was central to Milošević's fall, and its co-founder Srdja Popović went on to establish CANVAS in Belgrade, which has since worked with activists from forty-six countries, teaching what Popović calls "laughtivism." With Sophia McClennen and Joseph Wright, he gave the core mechanism a peer-reviewed formulation in the Journal of Democracy: the dilemma action.
A dilemma action is engineered so that every response available to the powerful costs them more than the action cost you. Crack down on a joke, and you are a state that deploys riot police against a punchline: brutal and ridiculous, a rare combination in that both halves recruit for the other side. Ignore the joke, and it stands, and spreads, and everyone watches you absorb it. The trap is built from the regime's own capabilities: its capacity to arrest, censor, and respond is the very surface the joke attacks. This is the purest expression of the cost-ratio principle anywhere in the trio: the joke costs nothing to make, and there exists no response that costs the target less than it cost you.
Honesty requires a label here, and the label is itself instructive: comedy is the weakest evidentiary leg of the three. There is no Chenoweth-grade dataset for humor in resistance: no rigorous effectiveness study, not even a good backfire literature. What exists is a well-theorized mechanism and strong cases, largely narrated by practitioners describing their own victories. The way to hold the claim honestly is to route it through the evidence that does exist: Popović's own stated ingredients for a successful movement are cool factor, memorable branding, and humor, and all three of those are recruitment mechanisms. The joke is not the weapon. The joke is a reduction in the cost of joining: it converts fear into permission, makes the movement legible and survivable to its potential members, and participation, per the actual dataset, is the weapon. Comedy is how the crowd gets cheap. The crowd is what wins.
The third move is the one every weak actor reaches for first, and it is different from the other two in a way that should make you slower to reach.
The mechanism is real and well documented. Third-party trust transfers: a 2008 study in Psychology & Marketing (Jiang and colleagues) found that a familiar certification seal raised consumers' initial trust in a retailer they had never encountered, standing borrowed, literally, from someone else's reputation. And the reason it works was formalized by Michael Spence back in 1973, in the signaling work that earned a share of the 2001 Nobel: a claim is informative only when it is costly to fake. Your self-praise is free, so it carries nothing. An independent endorser, by contrast, stakes their own reputation on you: the endorsement is cheap for them if you're genuine and reputationally expensive if you're not, which is exactly what makes it worth something.
You can test that rule on the wild, by the way, with a statistic that circulates through marketing blogs: that people trust third-party endorsements "92% more" than brand content. Chase it and you find no study, no methodology, no accountable author: an unfalsifiable number that lives on the websites of agencies and award programs whose product is third-party validation. A credibility claim with no cost behind it, deployed to sell credibility: this essay declines to cite it, and the declining is the lesson. (A sibling essay in this corpus, on the peacock's tail, asks when a costly signal can be trusted. This is the inverse case, and it answers itself.)
But here is the turn, and it is the most practical paragraph in this piece. The underdog playbook and the dilemma action are things you do. Borrowed credibility is a position you hold in someone else's balance sheet, and that difference is structural, not stylistic. You do not control the collateral: your endorser can fail, defect, or be discredited entirely without you, and your standing falls with theirs. The downside is unbounded and unmanageable from your seat: a failed flanking move costs you a round, a joke that lands badly costs you a news cycle, but a borrowed reputation that collapses can cost more credibility than you ever owned outright, because you were trading above your own book. It is, in the exact financial sense, leverage, and leverage is the instrument that converts a small adverse move into a total loss.
Notice what that means in the terms this essay has been building. The whole trio exists because the weak actor refuses to hold what the strong actor holds: refuses the stockpile contest. Borrowed credibility is the one move that quietly re-imports the stockpile: an asset you did not build, cannot control, and must defend. A borrowed tail is not a handicap you paid for. That is precisely the problem with it.
Every move in the trio fails the same way: the cost ratio inverts, and usually it inverts because you succeeded.
The underdog case is the documented one: it is Arreguín-Toft's own central finding run forward. The weak actor starts winning, grows, holds territory, fields formations, and one day discovers it has become conventional enough to fight the strong actor's war. Similar strategies favor the strong; the insurgent who builds battalions has volunteered for the dimension where the other side keeps its stockpile. The comedy case is, say it plainly, theorized rather than documented, since the backfire literature doesn't exist: but the logic says the joke dies when answering it gets cheap, when power learns to laugh along or floods the zone with its own mockery, or when the movement has grown serious enough that its own gravity was the asset and the punchline now spends it. And the credibility case is written into the trust-transfer literature itself: transfer runs both directions. The seal lends you its standing and you lend the seal your conduct, and loans get called at the worst moment on the calendar.
One adjacent essay in this corpus argues that some power (electricity is its example) cannot be stored at all, and has a different physics because of it. This piece is about the complementary trap: power that stores beautifully and then fails to convert. The incumbent's stockpile survives every asymmetric encounter fully intact. What the trio attacks is the exchange rate.
For a reader who builds things (products, teams, companies) the trio compresses to four working rules.
Pick the dimension where the stockpile doesn't convert. The incumbent's feature matrix, enterprise sales force, and brand budget are divisions of a conventional army. Competing there is agreeing to their war. The question that opens every asymmetric play is Arreguín-Toft's: what contest is cheap for me and expensive for them because of what they are: their pricing structure they can't cannibalize, their release cadence they can't accelerate, the customers they can't afford to serve?
Engineer dilemmas, not attacks. The dilemma action generalizes: the best competitive moves are the ones where every available response hurts the responder. Before spending on a move, price your opponent's replies. If one of their options is cheap for them, you haven't built a dilemma; you've built a request.
Route soft power through participation. Charm, humor, brand voice: these are not the weapon, and treating them as the weapon produces mascots. They are cost reductions on joining you. Measure them the way Otpor did: not "did they laugh" but "did laughing make it cheaper to walk in the door."
Audit your borrowed credibility like debt, because it is. List every logo on your homepage, every "as seen in," every platform whose algorithm delivers your audience, every partner whose name does your vouching. That is your leverage ratio. Some borrowing is rational (early on it's often the only bridge across the cold start), but know the size of the position, and pay it down: convert borrowed trust into owned trust (shipped work, kept promises, your own track record) the way a sane company converts short-term debt into equity.
And the tell, the one to write somewhere you'll see it: the moment these moves start working, you will feel a pull toward the conventional: the bigger org, the direct comparison, the prestige asset held for its own sake. That pull is not maturity. It is the ratio flipping, felt from the inside. The weak actor's discipline is not cleverness in the opening. It is refusing, even mid-victory, the flattering invitation to become the thing your whole strategy was built to fight.
Sources: Ivan Arreguín-Toft, "How the Weak Win Wars: A Theory of Asymmetric Conflict," International Security 26:1 (2001), and the 2005 Cambridge University Press book of the same name; Erica Chenoweth and Maria J. Stephan, Why Civil Resistance Works (Columbia University Press, 2011) and Chenoweth's research summary; Erica Chenoweth, "The Future of Nonviolent Resistance," Journal of Democracy 31:3 (July 2020); Srdja Popović, Sophia McClennen and Joseph Wright, "How to Sharpen a Nonviolent Movement," Journal of Democracy; Jiang et al., "How third-party certification programs relate to consumer trust in online transactions," Psychology & Marketing (2008); Michael Spence, "Job Market Signaling," Quarterly Journal of Economics 87:3 (1973).
Borrowed credibility is a position in someone else’s balance sheet. Paying it down means owning a record of your own.
The Agent Rating Protocol is our attempt at the reputation half: portable, checkable ratings for agents, so a counterparty can consult a track record instead of a logo wall. It does not manufacture trust and it will not save a bad actor. It changes who is holding the collateral.
pip install agent-rating-protocol · npm install agent-rating-protocol
See a verified chain · Hosted CoC, for the shipped-work half of the same problem