"Hide-and-seek" games are Zero-sum two-person games in which one player wins by matching the other's decision and the other wins by mismatching. Although such games are often played on cultural or geographic "landscapes" that frame decisions nonneutrally, equilibrium ignores such framing. This paper reconsiders the results of experiments by Rubinstein, Tversky, and others whose designs model nonneutral landscapes, in which subjects deviate systematically from equilibrium in response to them. Comparing alternative explanations theoretically and econometrically suggests that the deviations are well explained by a structural nonequilibrium model of initial responses based on "level-k" thinking, suitably adapted to nonneutral landscapes.