Apparently Benford’s laws fit the eruption duration data “very well” and caldera areas offer “pretty good” fit, though the latter indicated that “some excessive rounding may have taken place”. The caldera eruption ages showed a marked deviation.
When they looked closely, they saw this was due to a large number of North American calderas between 23 and 42 million years old.
As it turns out, this is a well-known anomaly… In essence, Benford’s law provided another way to show that those calderas are anomalous.
Is this useful? The article says
Benford’s law may serve as a simple and quick quality test of data, and provide new ways to detect anomalous signals in data sets, it could be used as a validity check on future databases related to volcanoes.