We experimentally investigate people's preferences towards different sources of information on a prediction market. The incentivized experimental task is to predict an unknown second-mover's behavior in an earlier hidden game experiment. In Experiment 1, we vary the source of information about that second-mover (picture, neutral video, video containing strategic content). Observed prediction accuracy rates serves as a measure of the empirical value of each source of information. In Experiment 2, we elicit the subjective value of information using the stated preferences method (WTA). The main result is that preferences towards information in a strategic setting are rational in the sence that they are closely aligned with its empirical value.