📊 統計・データ

パレート分布

ぱれーとぶんぷ

少数の要素が全体の大部分を占めるべき分布。所得・資産・都市人口など、正規分布に従わない現象のモデル。

1 分で読める

Definition and the 80/20 Rule

The Pareto distribution is a power-law probability distribution where a small number of elements account for the majority of the total. Named after economist Vilfredo Pareto, who observed that 80% of Italy's land was owned by 20% of the population, it describes phenomena where outcomes are extremely unequal. Income, wealth, city populations, and website traffic all follow approximately Pareto-distributed patterns.

How It Differs from Normal Distribution

Unlike the normal (bell curve) distribution where most values cluster near the mean, Pareto distributions have heavy tails - extreme values are far more common than a normal distribution would predict. The mean is a poor summary statistic because it is pulled upward by the extreme right tail.

This is why average income is always higher than median income: the distribution is Pareto-like, with a few very high earners pulling the average above what most people actually earn.

Implications for Rankings

In Pareto-distributed data, small differences in rank near the top correspond to enormous differences in absolute value, while large rank differences near the middle correspond to tiny absolute differences. Moving from the 50th to the 60th percentile in income might mean a modest raise, but moving from the 99th to the 99.9th percentile represents a massive jump in earnings.

Recognizing Pareto Patterns

Whenever you encounter a ranking where the top performers are orders of magnitude above the median, you are likely looking at Pareto-distributed data. In such cases, comparing yourself to the average is misleading - the median is a far more relevant benchmark. Understanding the shape of the underlying distribution is essential for interpreting what any given rank position actually means in practical terms.

関連用語

関連記事

この用語解説は役に立ちましたか?