Determining rank hierarchies
An approach that was originally developed to calculate the relative ability of chess players seems to provide more accurate results for the calculation of rank hierarchies. So far, behavioral scientists have created rank orders solely from dyadic, aggressive interactions between the individuals of a group. Hereby the temporal information of these winner-loser relationships was lost, dynamic changes of the hierarchy were thus easily overlooked. Contrastingly, Elo-ratings and Bayesian inference take temporal information into account and are suitable to detect social dynamics. Prof. Julia Fischer explains the parallels between the skill level of chess players and the rank hierarchies of chacma baboons in the Methods.blog: https://methodsblog.com/2018/10/29/elo-rating-bayesian-inference/