One inference from game theory models of animal conflict is that adversaries should not inform one another about concealed components of their fighting ability. This poses a paradox for the customary ethological account of aggressive displays in that it is usually assumed that the primary function of such behavior is to make such information available. To resolve the paradox, I propose that the information in aggressive displays may not be strictly truthful, but may instead represent "optimal deceit," a balance between the advantages of deceit or bluffing and the disadvantages of selecting for skepticism in the receiver. Numerical simulation of this model was performed to examine the effects of differences in fighting ability and in the risk of injury in an escalated conflict. The model converged on an equilibrium level of deceit, even when the receiver was ignorant of the average level of deception being employed.