This paper evaluates performance of human subjects and instances of a bidding model that interact in continuous-time double auction experiments. Asks submitted by instances of the seller model ("automated sellers") maximize the seller's expected surplus relative to a heuristic belief function,...