This project explores balance control in bipedal humanoid robots by analyzing the "balanced region"—the set of center of mass (COM) states from which a robot can recover from a push without falling. Using the Robotis-OP3 robot in the Webots simulation environment, the study tests various initial COM conditions to determine the velocity thresholds that lead to a fall. Both ankle and hip strategies are used for push recovery. The resulting balanced region can inform the development of real-time COM-based controllers, improving balance and fall prediction in humanoids. This approach also has potential applications in other legged systems, such as exoskeletons and prosthetics.