The distribution of books ordered from Amazon has a long and heavy tail, like the distribution of videos ordered from NetFlix, and accounting for these distributions requires something like Gnedenko’s refinement of the usual central limit theorems.

Boris Vladimirovich Gnedenko was himself one of the unlikeliest characters ever to grace the field of probablity and statistics. He had the misfortune to begin his university career at the height of the Stalinist purges, and after expressing himself carelessly one day about the infinite virtues of all things Soviet, he was arrested and imprisoned with 120 other prisoners in a cell built for six.

Gnedenko was a very small fish for the secret police; the real prize was his professor, the great Andrey Nikolaevich Kolmogorov.

His interrogators constantly badgered him to “confirm that Kolmogorov was the ringleader of a group of “enemies of the people” centred in the mathematics department. Though interrogated daily over a six-month period, held in grim conditions, and promised his release if he cooperated, he refused to admit even the possibility of such an interpretation, knowing that there could be no hard evidence, and that the fate of all, himself included, depended on his resolution.” [D Vere-Jones, Boris Vladimirovich Gnedenko, 1912-1995. A personal tribute, Austral. J. Statist. 39 (2) (1997), 121-128.]

Gnedenko was unexpectedly released after six months, resumed his career with the help of Kolmogorov and Khinchin, published over 200 scholarly articles and books, and served for 30 years as head of the Department of Probability and Statistics at Moscow University.