Consensus algorithms are widely used in control theory to design distributed control techniques in networked dynamical systems. One of the most conventional techniques is model predictive control (MPC) because it can handle restrictions over the states and system inputs. However, the computational resources required by the algorithm could make the controller unimplementable for some systems. In this work, we compare the MPC with a non-traditional controller based on potential games using agents with single integrator dynamics and nonlinear ground vehicles. We have found that the potential game-based controller can achieve a similar performance spending less computation time.