LIMITS OF BENFORD'S LAW IN EXPERIMENTAL FIELD

Abstract

Two of the main flaws of Benford’s law will be discussed in this article: (i) the first one, which leads the observer to consider an experimental dataset as the result of a single random variable rather than several, makes this law so mysterious; (ii) the second one is that Benford’s probabilities have long been considered as perfect values: this is obviously not the case. AMS Subject Classification: 60E05