Recently, game-playing agents based on AI techniques have demonstrated super-human performance in several sequential games, such as chess, Go, and poker. However, in many real-world human-agent interactions that are repeated over time, humans may adopt suboptimal strategies and drop out from the repeated interaction if they are not sufficiently engaged in the competition. For instance, this may happen when they are losing too much against the agent, as well as when they are winning too often sin…