Our brains are pattern-detectors, not probability calculators

Mlodinow argues that human cognition evolved to detect patterns and causal relationships quickly, a skill that was survival-relevant for spotting predators or edible plants, but this same tendency misfires badly when applied to genuinely random processes, where it invents patterns and causes that don't exist. We are, in his phrase, pattern-seeking animals confronting a universe that frequently offers noise instead of signal.

This mismatch explains why people see "hot streaks" in basketball shooting that turn out, on statistical analysis, to be indistinguishable from what random chance alone would produce, or why investors attribute skill to fund managers whose results are statistically consistent with luck. The brain's need for a satisfying causal story is stronger than its appetite for the more accurate but less satisfying answer: sometimes there's no story, just variance.

Mlodinow doesn't argue skill never matters; he argues we systematically overestimate how much of an outcome skill explains versus chance, especially in domains involving many trials or many competitors.

Takeaway: the compelling story your brain generates about why something happened is often less accurate than the boring statistical explanation you're inclined to dismiss.