Ranking athletes and teams across epochs

Abstract

Who’s the Greatest Of All Times, be it a team or an athlete, is hotly debated among sport fans. To provide a tentative answer, we apply the mathematical concept of network centrality upon the paired comparison model. We capture the twin notions of fitness and prestige across epochs in order to rank performers who never met on the field, this with a minimal transparent ad-hoc parametrization. Our gauge is the eigenvector of a large matrix of games results, nor- malized to level the playing field (à la PageRank) by applying the insight of baseball statistician Bill James. We present tables for chess, club and country football, motor racing, skying, cycling, tennis, baseball, US pro & college football, basketball, hockey, rugby, poker, cricket and fighting.

Keywords: Sports, Network, Pairwise Comparisons, Multi-criteria Decision Making

JEL Classification: C02, C43, C67, D71, Z13