Bowerbirds (Ptilonorhynchidae) are distributed in Australia and New Guinea and are best known for the fact that males build complex structures, known as bowers, for the purpose of attracting females. Bowerbirds are unique among birds in that the male’s mating success is primarily based on the quality of this structure – an externalized secondary sexual characteristic – rather than on aspects of the male himself (such as plumage, song, or display). Male bowerbirds often maraud (destroy) the bowers of rivals to increase their own mating success. However, a male’s decision to spend time away from his own bower can be costly, as he may miss opportunities to mate or defend his bower. In this study, we simulated interactions among males during a mating season and their decisions between mating seasons regarding how frequently to maraud and whether to relocate their bowers based on prior performance. We built a continuous-time agent-based model to represent a series of consecutive mating seasons (referred to as generations), allowing underperforming males to make stochastic adjustments between seasons. In our iterated simulations, we kept either marauding strategy (which is a continuum of rates of marauding behavior) or bower position fixed among all males and allowed the other attribute to be adjusted. We found that interactions between males alter the population’s levels of bower dispersion and marauding relative to our null model (in which each male’s mating success is random). These findings could inform hypotheses about the rise of marauding and of uniform bower dispersion, which might be further tested through modeling or fieldwork.