Modeling the evolution of winner and loser effects: A survey and prospectus

• Game-theoretic models of the evolution of winner and loser effects are reviewed.
• Current models have been well supported by the empirical literature for over 40 years.
• But no model predicts a winner effect without a loser effect, as recently observed.
• And none predicts two losses to make winning more likely, as recently observed.
• The need for future models to address these and other issues is characterized.

The evolution of winner or loser effects—higher probabilities of winning after winning or of losing after losing—has received remarkably little attention from theoreticians, even though such effects are widespread across the animal kingdom. We review game-theoretic models that regard such winner and loser effects as outcomes of a strategic response. We show that these models have been well supported by the empirical literature in the past, but are not designed to address some recent observations. In the light of this recent progress on the empirical front, we identify factors that newer theory must be developed to explore.