Wireless sensor network is considered as the most applicable standards for the monitoring system. The optimization that relates to wireless sensor network planning, design, deployment, and operation has to consider many parameters so that there is no conflict among the sensor especially in a heterogenous network with various platforms. In such a network, energy consumption determines the system stability. Clustering mechanism in a wireless sensor network can also simplify network management process. Distributed energy efficient clustering (DEEC) is used as a clustering protocol. This protocol work based on residual energy. This protocol is robust enough, but in a certain condition, it has a pit of weakness. Game theory is being proposed as an optimization algorithm of the clustering process. It is used to adjust the probability of node to become a cluster head based on residual energy to prolong cluster lifetime. Threshold value can be more accurate by using selfish behavior game theory algorithm. The weighted factor is a factor that makes the resolution of probability values more real.