Shared energy storage is a key technology for energy reform, and the alliance of multiple data center microgrids and shared energy storage can realize energy complementarity and resource linkage among different entities. To improve the economic benefits of systems with shared energy storage, a two-stage robust multiple data center microgrids and shared energy storage collaborative game model considering wind power and photovoltaic uncertainty is proposed. First, an alliance is formed between the microgrids and community-shared energy storage to construct an optimized operational model for electricity trading. Then, the uncertainty of wind power and photovoltaic output is represented in the form of a box-type uncertainty set to build a two-stage robust optimal dispatch model. Nash negotiation theory is used to construct a multi-microgrid and shared energy storage multi-entity cooperation game model. The model is equivalently transformed into the subproblems of alliance cooperation cost minimization and individual entity trading negotiation. These are analyzed in a distributed manner using the alternating direction multiplier method. Finally, simulation results show that this method can effectively reduce the operational cost and calculate the optimal compensation price for data center users with different robust parameters.