How steep is the rank curve?
Rank-frequency patterns can be highly uneven without a sharp cutoff.
Inside this model
Twenty ranks receive weights proportional to 1/r^s, normalized to 100%. The plot displays their relative frequencies as the exponent changes.
Out in the world
A practical use
Examine why a few words appear far more often than most others in a text corpus.
A useful lens. Not a universal law.
- The exponent varies by data set and range; a good-looking rank plot alone does not establish a universal law.
Associated thinkers
Further reading
Explore the original research or the teaching reference behind this experiment.
George Zipf — Human Behavior and the Principle of Least Effort ↗