"One of the most common [features of legal realism] is faith in masses of figures as having significance in and of themselves. . . . If, for example, we are studying congestion of the criminal dockets in certain parts of the country, and its effects upon the enforcement of a particular law, and find that in those places the percentage of sentences to imprisonment runs from 4 to 6, whereas in the country at large it is 41, these figures throw much light upon the workings of 'bargain days' and 'cafeteria courts.' Of themselves they mean nothing. They get their significance from the connection in which they were sought for and the conclusion, probably reached in the first instance on another basis, which they confirm. Masses of figures do not make a piece of work scientific. But a scientific inquirer may have an idea which he can fortify or confirm by knowing where and how to find a mass of figures significant for his purpose. Very little experience of using current official statistics is required to convince that statistics gathered for no purpose beyond filling a report with impressive tabulations are seldom valuable for anythign else."
Roscoe Pound, 44 Harv. L. Rev. 697, 703.