In the heat of the moment: convergent behavior of rankings under simulated annealing

Abstract

Rankings are a recent tool being used in several fields, including in management. Their pervasive use is associated to the fields of behavior and decision making. Despite their constant usage, few research have tried to define the concept of ranking and the parameters to judge whether it is acceptable. In absence of a more precise understanding of what the term ranking means, its power is diminished as well as its purpose. Thus, in this work I present the characteristics, advantages and disadvantages of rankings. Further, I analyze patterns of behavior elicited in real rankings. To this end, I propose the use of simulated annealing to study the quality of convergence of rankings’ chosen dimensions. The graphical analyses suggest, at least, two different patterns of convergence for rankings. The categories were named after well behaved and poorly behaved rankings.