Perhaps to you. But that's because you have prior cultural knowledge of chess, and what it means to "play" something. It seems to me to be perfectly reasonable for an ai agent to define "play chess" as "behave in a way that this arbiter indicates that you have done well"

Would that help? It seems like Do not cheat is a new Do not hallucinate.