Spurious Correlations is a collection of charts pairing unrelated or unlikely-to-be-related trends that happen to move together over time. Examples include the popularity of the name Annabelle alongside UFO sightings in Maryland, butter consumption alongside biomass power generation in Lithuania, and margarine consumption alongside Maine’s divorce rate. Each chart presents two measures across a span of years, making their apparent alignment easy to see; the humor comes from how plausible-looking patterns can connect subjects with no sensible relationship.
The site also explains why these matches appear: its creator compares 25,237 variables, producing more than 636 million possible correlations, then displays striking matches found by searching across the data rather than testing a prior hypothesis. This process, called data dredging, makes it likely that large datasets will contain strong-looking correlations that arose by chance. The central lesson is that correlation does not establish causation: a shared pattern alone does not show that one measure affects the other or that a meaningful link exists. Some pairs could happen to be causally connected, but the site says it filters out obvious relationships and uses the examples chiefly to demonstrate the risks of drawing conclusions from patterns discovered through extensive searching.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.