Why Averages Lie — Visual Guide
Companion guide to: Why Dating Apps Feel Broken — Explained With Data
Averages are useful when the group is balanced and the question is simple. They become dangerous when the distribution is skewed, when outliers dominate the result, or when the median person is far away from the mean. This guide shows why income, dating outcomes, housing prices, and creator audiences often need medians, percentiles, or full distributions instead of one average.
The core move is simple: before asking what the average is, ask what shape the data has. If the top end pulls the mean upward, the average may describe the arithmetic while failing to describe the typical experience. That is why a claim can be numerically correct and still misleading.
Mini exercise: write down the mean, median, and 90th percentile for a small dataset. Then explain which number you would use in a headline and which number you would use if someone asked what most people experience.