The Trap of Pattern Recognition: Why Experience Can Be Deadly

Why You Trust Patterns That Don't Exist During the 2018 Camp Fire in Paradise, California, incident commanders had a problem. The fire was moving fast—far faster than any historical pattern their experience had taught them to expect. They had decades of collective knowledge about how wildfires behave in that terrain, under those winds, in November. Every previous fire had followed a certain envelope of speed and spread. This one did not. By the time command adjusted its mental model, the fire had already overrun the town. Eighty-five people died. The gap between what the commanders expected and what the fire did is not a story about incompetence. It is a story about induction—the cognitive habit of assuming that what has happened before will happen again. The philosopher David Hume identified this problem in the 18th century: no amount of past observations can logically guarantee that the future will resemble the past. The sun has risen every morning of your life, but there is no deductive proof it will rise tomorrow. Hume wasn’t being pedantic. He was pointing out that induction is a psychological habit, not a logical necessity. And that habit, when it operates outside its safe zone, can kill. The Camp Fire was a novel event. The combination of extreme drought, record winds, and a century of fire suppression created conditions that had no precedent in the region’s recorded history. The commanders’ pattern recognition—honed by hundreds of smaller fires—told them they had time. They didn’t. The very expertise that made them confident was the same mechanism that made them wrong. This is the crisis decision-maker’s dilemma. Induction bias is not simply a cognitive error you can think your way out of; it is a design flaw in how we train people to make high-stakes judgments. We reward pattern recognition. We promote people who can size up a situation quickly based on what they’ve seen before. We call it experience. And most of the time, it works—because most situations are not novel. The problem is that when a situation is novel, the pattern-seeking brain does not know to switch modes. It keeps feeding you the old map, even when the terrain has changed. You see the same mechanism at work in domains far from wildfire. In investing, the induction bias leads people to extrapolate recent market returns indefinitely—buying at the top because “stocks always go up,” selling at the bottom because “this time is different.” In relationships, we assume that a partner who has never broken trust will never break trust, ignoring the fact that people and circumstances evolve. In medical diagnosis, a clinician who has seen a hundred cases of viral pneumonia may miss the one case of a novel coronavirus because it doesn’t match the pattern. The pattern itself is not the enemy. Pattern recognition is one of the most powerful tools the human brain has. It lets you walk into a room and instantly read social dynamics, or glance at a spreadsheet and spot an anomaly. The danger comes when you treat the pattern as a certainty rather than a hypothesis. Hume’s insight is that induction gives you no guarantee—only a probability based on past frequency. And when the underlying conditions shift, that probability becomes worthless. So what do you do? The first step is to acknowledge that your pattern-recognition system operates automatically and unconsciously. You cannot simply will yourself to be more skeptical in the moment. You need structural countermeasures. One of the most effective is the pre-mortem. Before you commit to a decision based on a pattern, assume that the pattern has broken. Imagine that you have already failed—and then work backward to figure out what could have caused that failure. This forces your brain to actively search for disconfirming evidence, which is the exact opposite of what induction does naturally. Another technique is to explicitly ask: “What would have to be true for this situation to be fundamentally different from every previous one?” The Camp Fire commanders might have asked: “What would it take for this fire to move three times faster than any we’ve seen?” The answer—extreme drought, unprecedented winds, decades of fuel buildup—was already in the data. But they weren’t looking for it because the pattern said otherwise. A third, simpler practice is to keep a decision log. Write down the pattern you are relying on and the evidence that supports it. Then write down one counterexample—even a hypothetical one. The act of externalizing the reasoning makes it easier to see the gap between the pattern and the evidence. None of these techniques will eliminate induction bias. It is built into the architecture of the brain. But they can slow you down enough to notice when the pattern might be lying to you. The pattern always breaks eventually. The question is whether you will see the break before it breaks you. Hume’s problem of induction is not an abstract philosophical puzzle. It is a practical warning about the limits of experience. The Camp Fire killed eighty-five people because a group of highly trained professionals trusted a pattern that no longer held. The same logic plays out every day in smaller ways—in the stock you hold too long, the relationship you assume is stable, the diagnosis you don’t question. The future is not obligated to repeat the past. The only way to survive that truth is to build habits that force you to check whether the pattern is still real.