Players perceive information about game environments through a virtual camera. While a significant discussion in the industry and in academic research circles has centered around effective camera control, it is focused mainly on occlusion free placement and smooth movement. The relationship between information communicated by the camera about game state and the selection of camera parameters has not been investigated.In this paper, we present a systematic investigation of the effect of different camera profiles on player experience in a 3D prey/predator test-bed game. We describe a constraint-based dynamic camera system that maintains the position and orientation of the camera based on the constraints imposed by given camera profiles. The impact of different profiles on the amount of game information provided to the player and the player's game challenge preferences is investigated through a user experiment. An artificial neural network model of challenge constructed using artificial evolution reveals the non-linear mapping between challenge and information features.