"Hot Hand": A Critical Analysis of Enduring Rivalries

Statisticians refer to discernible patterns of events that are nevertheless unrelated as a "hot hand." An observer watching a gambler on a winning streak may suspect that the game is fixed, but it is not necessary to depart from standard probability theory to account for series of low-probability events. Enduring rivalries research often uses frequency counts of militarized disputes to identify rivalries as "enduring." Nevertheless, discernible patterns in dispute data could be hot hands. Using the same datasets and definitions as the original researchers, we test the hypothesis that the number of enduring rivalries reported in the literature is significantly different from the number generated by a model in which disputes are randomly distributed. We compare the null model to existing definitions of enduring rivalry using a Poisson process. We find that we cannot reject the hypothesis that dispute series identified as enduring rivalries result from a stochastic process.