Ice Cream Sales Predict Drowning Deaths? The Stats Trap Every Student Falls Into

Hinzugefügt: 2026-05-16

[00:00:00]Did you know that ice cream sales perfectly predict drowning deaths? It's true. The correlation is almost perfect, 0.95 out of 1. A student finds this in their data set and gets excited. They've discovered something groundbreaking.

[00:00:13]They start building theories. Maybe ice cream makes people drowsy in water.

[00:00:17]Maybe sugar affects swimming ability.

[00:00:19]>> [music] >> They write pages explaining this revolutionary connection between dairy products and water safety, but they missed something obvious, hot weather.

[00:00:27]When it's hot, people buy more ice cream. When it's hot, people also swim more. More swimming means more drowning accidents. Ice cream doesn't cause drowning, summer does both. This happens to students constantly. They find strong correlations and assume they found causation. They see impressive numbers and think they've solved something important, but correlation doesn't equal causation. And big data sets create fake relationships everywhere. The number of films Nicholas [music] Cage appears in correlates with pool drownings. Divorce rates correlate with margarine consumption. These aren't real relationships. They're statistical accidents. Before celebrating your correlation, ask yourself, what's the real cause behind both variables? Your supervisor will thank you.