Our understanding of animal mating systems has changed dramatically with the advent of molecular methods to determine individuals' reproductive success. But why are older behavioral descriptions and newer genetic descriptions of mating systems often seemingly inconsistent? We argue that a potentially important reason for such inconsistencies is a research trajectory rooted in early studies that were equivocal and overreaching, followed by studies that accepted earlier conclusions at face value and assumed, rather than tested, key ideas about animal mating systems. We illustrate our argument using Anolis lizards, whose social behavior has been studied for nearly a century. A dominant view emerging from this behavioral research was that anoles display strict territorial polygyny, where females mate with just the one male in whose territory they reside. However, all genetic evidence suggests that females frequently mate with multiple males. We trace this mismatch to early studies that concluded that anoles are territorial based on limited data. Subsequent research assumed territoriality implicitly or explicitly, resulting in studies that were unlikely to uncover or consider important any evidence of anoles' departures from strict territorial polygyny. Thus, descriptions of anole behavior were largely led away from predicting a pattern of female multiple mating. We end by considering the broader implications of such erratic trajectories for the study of animal mating systems, and posit that precise definitions, renewed attention to natural history, and explicitly questioning assumptions made while collecting behavioral observations will allow us to move towards a fuller understanding of animal mating systems.