Why Behavioral Prediction Is a Dangerous Fantasy

Why Behavioral Prediction Is a Dangerous Fantasy

We are obsessed with the idea that we can pre-identify the "broken" among us.

Every time a mass tragedy occurs, the media cycle follows a predictable, exhausting script. We hunt for the "missed signs." We look for a scientist or an algorithm that promises to spot the next perpetrator before they pull the trigger. We pour hundreds of millions into behavioral threat assessment, social media scraping, and psychiatric "red flag" profiling.

It is a comforting lie. It suggests that violence is a technical glitch we can patch.

The reality is much uglier. The harder we look for "the signs," the more we realize they don't exist in a vacuum. By trying to turn human behavior into a predictable dataset, we aren't stopping violence—we are building a panopticon that treats deviation as a crime and provides a false sense of security while the real threats remain invisible.

The Myth of the Universal Profile

Most academic research on mass violence—the kind often touted as "groundbreaking"—is built on a foundation of survivor bias and retrospective analysis.

When a scientist looks back at a perpetrator’s life, every detail looks like a clue. "He played violent video games." "He was socially isolated." "He bought a lot of ammunition."

The problem? Millions of people do those things and never hurt a fly. This is the Base Rate Fallacy.

When you have an event as statistically rare as a mass shooting, even a "prediction tool" with 99% accuracy will produce thousands of false positives for every one correct hit. In a country of 330 million people, "identifying" a potential shooter based on behavioral markers is like trying to find a specific grain of sand on a beach by looking for grains that are "slightly more tan."

If we actually acted on these "red flags," we would have to incarcerate or monitor hundreds of thousands of lonely, disgruntled young men every single day. We don't have the resources, the legal right, or the moral ground to do that.

The Logic of the "Loner" Is Dead

The competitor narrative suggests that if we just find the lonely scientist who understands "the darkness," we can solve the problem. They focus on the individual’s psyche as if violence is a tumor that can be surgically identified.

This ignores the Contagion Effect.

Research by figures like Dr. Sherry Towers has demonstrated that mass killings often follow a mathematical pattern of contagion. They aren't just isolated psychological breakdowns; they are social phenomena. When we profile the individual, we ignore the ecosystem.

Mass violence is often a performance. It is a scripted act. Perpetrators study previous events like athletes study film. By focusing on "predicting" the next person, we miss the fact that our media-saturated culture provides the script, the stage, and the audience.

I have watched school districts spend $500,000 on AI-powered "sentiment analysis" software that scans students' emails for words like "kill" or "bomb." You know what it finds? Kids talking about video games. Kids quoting rap lyrics. It misses the student who is actually planning something because that student isn't using keywords. They are using silence.

Psychiatry Is Not a Crystal Ball

We love to blame "mental health." It’s the easiest out for politicians on both sides.

💡 You might also like: The Great Firewall of Childhood

But the data is brutal: the vast majority of people with mental illness are never violent. In fact, they are more likely to be victims of violence. Conversely, if you could magically cure all serious mental illness tomorrow, the rate of violent crime in the United States would only drop by about 4%.

When we task psychiatrists with "stopping mass violence," we are asking them to do something their profession isn't designed for. Clinical practice is about helping an individual cope with distress, not acting as a pre-crime division for the state.

The "mental health" argument is a deflection. It shifts the burden from social and systemic failures onto a medical profession that can barely keep up with the demand for basic therapy.

The High Cost of the False Positive

Imagine a scenario where a high school student is flagged by an algorithm. He’s depressed because his parents are divorcing. He’s been researching historical firearms for a school project. He’s withdrawn from his friends.

Under the "stop the violence" models currently being pushed, this kid gets a knock on the door from the police. He’s put on a list. He’s stigmatized by his peers. He is treated as a monster in waiting.

What does that do to a person? It accelerates the very isolation and resentment that leads to radicalization. We are creating a self-fulfilling prophecy. By treating citizens as potential threats, we erode the social trust that actually keeps communities safe.

The Actionable Truth: Hardening vs. Profiling

If behavioral prediction is a dead end, what actually works? It’s the stuff no one wants to talk about because it’s boring and expensive.

  1. Environmental Design: You don't need to read minds if you design spaces that are difficult to exploit. This isn't just about "locks and clocks"—it's about creating visibility and reducing the "performance" value of a location.
  2. Crisis Intervention, Not Surveillance: Instead of "threat assessment teams" that hunt for bad guys, we need "resource teams" that provide actual material support (jobs, housing, healthcare) to people in crisis before they even consider violence.
  3. Disrupting the Script: We have to stop naming these people. We have to stop publishing their manifestos. We have to stop making them the protagonists of their own tragedies.

Stop Hunting for Monsters

The "lone genius" scientist looking for the "root cause" of violence is a great character for a magazine profile, but they are a distraction from the structural reality.

Violence is not a mystery to be solved. It is a byproduct of specific social conditions: extreme wealth inequality, the erosion of local communities, easy access to high-capacity weaponry, and a media machine that rewards infamy.

Predictive algorithms and behavioral profiling are the "snake oil" of the 21st century. They offer the illusion of control in an uncontrollable world. They allow us to avoid the hard work of fixing our society by promising us that we can just find and remove the "bad" people.

As long as we keep looking for "the signs" in the individual, we will continue to be blind-sided by the system.

The "signs" aren't in the person. They are in the mirror.

Stop looking for the needle. Fix the haystack.

IE

Isaiah Evans

A trusted voice in digital journalism, Isaiah Evans blends analytical rigor with an engaging narrative style to bring important stories to life.