Towards Causal Models for Adversary Distractions