Mobile crowdsourcing has shown great potential to address problems with large scale by outsourcing tasks to pervasive smartphone users. Smartphone users will join a crowdsourcer if they can receive satisfying rewards. In a mobile crowdsourcing market, smartphone users have free choice of crowdsourcers, and multiple crowdsourcers will interact with the rest of the market to share the limited smartphone contributions (i.e., sensed data). To better fit the gap between the demands of crowdsourcers and the capabilities of smartphone users, the underlying rationale of crowdsourcers’ behavior needs to be well understood. However, little attention has been given to this issue. In this article, we analyze and predict the behavior (i.e., adjust the price paid) of crowdsourcers. We use a dynamic non-cooperative game to formulate the interaction among crowdsourcers and extend it to a repeated game since crowdsourcers may cooperate with each other to get the optimal profit considering the long-term profit when the game is played multiple times.