Regenerative braking in a hybrid energy storage system for a rubber-tyred gantry crane (RTG-HESS) requires a trade-off between energy recovery and battery charging stress. This study develops an energy management strategy using game theory (EMS_GT) in a MATLAB/Simulink–AVL Cruise co-simulation model comprising a diesel range extender, a lithium-ion battery, and a supercapacitor. An auxiliary finite zero-sum matrix evaluates candidate power allocation policies under competing objective emphases. The matrix is solved by linear programming, and the resulting probabilities are reused heuristically as weights in a deterministic constrained allocation cost. This mapping does not establish a Nash-equilibrium guarantee for the physical controller. Under a shared 100 A battery charging limit, EMS_GT achieves a recovery efficiency of 0.914, compared with 0.890 for the rule-based strategy (EMS_RB), corresponding to an absolute increase of 0.024 and a relative increase of 2.70%. Its normalized battery-degradation indicator is 0.641 relative to the constrained EMS_RB reference of 1.000, and its supercapacitor utilization is 0.817. Additional ECMS, parameter sensitivity, Monte Carlo, and ablation results characterize performance within the tested conditions. These co-simulation results support a trade-off between recovery and modeled charging stress; they do not constitute measured battery-lifetime extension or comprehensive braking-safety validation.