Assessment of fighting ability in animal contests

Selection should favour accurate information gathering regarding the likely costs and benefits of continued conflict. Here we consider how variation in the abilities of contestants to assess resource-holding potential (RHP) influences fights. This has been examined in various game theory models. However, discriminating between assessment strategies has proven difficult and has resulted in confusion. To add clarity, we group existing models into three main types that differ in the information about RHP that contestants are presumed to gather: (1) pure self-assessment, (2) cumulative assessment and (3) mutual assessment. Within this framework we outline methods advocated to discriminate successfully between the three main assessment models. We discuss support for each model, before highlighting a number of conflicting and inconclusive studies, leading us to consider alternative approaches to investigate assessment. Furthermore, we examine support for newly emerging concepts such as ‘varying degrees of assessment’, ‘switching assessment’ strategies and the possibility of contestants adopting different assessment strategies within a fight involving distinctive roles. We suggest future studies will benefit by judicious use of a battery of techniques to determine how animals settle contests. Finally, we highlight difficulties with current game theory models, and raise concerns regarding the use of certain behavioural criteria to accept or reject a model, particularly since this may conflict with evidence for a given assessment strategy. Furthermore, the failure of existing models to account for newly emerging concepts points to limitations of their use and leads us to challenge game theoreticians to develop upon them.