Clustering illusion

Hypothesis assessment biases biases

The clustering illusion is the tendency to see meaningful patterns—like streaks or clusters—in random data. We overinterpret small samples and treat coincidence as evidence of a real effect. Clustering illusion and the Texas sharpshooter fallacy are the same idea: one is the tendency to see patterns in randomness, the other is the argument that treats such a pattern as proof. Related: Texas sharpshooter.

Examples

  • You notice that accidents seem to happen in threes, and conclude there is a pattern—when in fact random events will sometimes cluster by chance.

  • A sports fan sees a "hot hand" in a player who has made several shots in a row, even though streaks of that length are common in random sequences.

  • You think a certain street is "unlucky" because you have heard of several incidents there, without comparing how many incidents occur on similar streets.

  • A trader sees a run of winning trades and believes they have found a pattern in the market, when the run could easily be random variation.

← Back to Biases