Median

The median is the value that sits exactly in the middle of a sorted dataset: half the values fall below it and half above. For an odd number of observations it is the single middle value; for an even number it is the average of the two middle values.

The median differs from the mean, the arithmetic average, in an important way: it depends only on order, not magnitude, so it is a percentile, specifically the 50th, rather than a computed sum. Because of this, an extreme value on one end of the distribution barely moves the median, while it can shift the mean substantially. This makes the median far more resistant to outliers and skew.

Practitioners reach for the median when a distribution is skewed or has heavy tails, which is common for income, home prices, session duration, and latency, where a small number of very large values would otherwise drag the mean upward and misrepresent the "typical" case. The common pitfall is reporting only the mean for such data and implying it describes a typical observation; pairing the median with the interquartile range gives a much more honest picture of where most values actually fall.