On Continuous Time Markov Games With The Expected Average Reward Criterion