Massively Multiplayer Online Games (MMOG) keeps attracting a large number of players. To sustain the increasing number of gamers, MMOG infrastructure have switched nowadays to Cloud computing paradigm, which can offer to the game service providers as well as to the end users a scalable and an affordable access to the infrastructure required to support MMOG. While many technical challenges related to cloud gaming have recently been addressed, several open issues have yet to be investigated. In particular, ensuring an acceptable Quality of Experience (QoE) for all players is a fundamental requirement to the economical development of the cloud gaming industry. This paper aims to contribute to this effort, by proposing a mathematical modeling of the QoE for MMOG cloud-based service. Precisely, after conducting a set of experiments, we provide an expression of the response delay as a function of the number of players and the number of allocated Virtual Machines (VMs) in charge of executing MMOG server tasks. We then use this model to discuss the tradeoff between the QoE and the induced costs to the provider and to the customers.