Utterly Predictable

The Science of Predicting Exactly What You're About to Do

This is a behavioral prediction engine: feed it a crisis scenario, and it hands back group-by-group predictions of how people actually respond, built on peer-reviewed behavioral science instead of punditry. It predicts what people will do, not what they'll tell a pollster they're going to do.

Pick a Disaster. We'll Tell You What Everyone Actually Does Next.

Pick any scenario below to see behavior broken out group by group. Every analysis runs the same five-step process: classify the event, find the closest historical precedent, split the population into groups, map how it unfolds over time, and predict the behavioral output for each one.

Utterly Predictable /

60 Years of Disaster Data Say You've Got This Backwards

Every crisis in this series turned up the same handful of findings, and most of them contradict exactly what you'd guess.

Mass Panic Is a Myth

The Disaster Research Center, founded in 1963, has 60 years of data, and it says populations turn overwhelmingly prosocial under crisis. The panic that actually shows up is "elite panic" — authorities deploying the military, locking down property, restricting movement — not the public losing its head.

Political Identity Outpredicts Everything

In the modern West, political affiliation is the #1 predictor of crisis behavior, beating age, education, income, and health status. COVID mortality diverged along party lines harder than it did along any single medical variable.

The Say/Do Gap Is Universal

Preference falsification (Kuran) means public behavior never lines up with private belief. Watch the gap between what people do anonymously (search data) and what they do in public (social media, polls) — when that gap widens, a cascade is coming.

Historical Parallels Beat Theory

Tetlock's outside view says skip the first-principles modeling: find the closest historical reference class, anchor on its base rate, then adjust for what's actually new this time. Algorithms beat experts because they're noise-free.

Humans Adapt to Almost Anything

Hedonic adaptation kicks in within 3-6 months for most constant stimuli. The exceptions that never fade: chronic food insecurity, chronic unpredictability, and chronic lack of control.

25% Tips It, 3.5% Topples Governments

Centola's research shows visible behavior locks in once 25% of a population adopts it. Chenoweth found no government has survived a sustained nonviolent campaign with 3.5% active participation. Below that threshold, nothing moves; cross it, and you get a sudden phase transition.

Millions Think Radical, Almost Nobody Acts Radical

McCauley's Two Pyramids model explains why: millions think radical, but only a sliver ever act on it, and the two run on completely different mechanisms. What actually pushes someone from opinion to action is network closure and status deficit, not how intensely they believe.

Framing Beats Facts

Kahan, Feinberg and Willer found the exact same threat produces opposite responses depending on whether you frame it through the audience's values or against them. Identity-protective cognition overrides the data, right up until the material cost of denial exceeds the social cost of admitting you were wrong.

The Accuracy Ceiling Is ~60-70%

Kahneman's "Noise" research says the honest move is predicting categories and directional probabilities, not exact magnitudes, and that past 3-5 years out, accuracy approaches pure chance. That's why every prediction here carries a confidence level instead of a guarantee.

Get Notified When the Next Scenario Drops

All 12 Catastrophes, Mapped

Twelve scenarios, group-by-group behavioral predictions for each one. Pick any link below to read the full breakdown.

Related on scovert.com