Drive simulators using serial robots, such as the KUKA robot “Robocoaster”, are becoming attractive for situations in which the workspace of traditional Stewart platforms is not suited to accommodate large target rotations, allowing for a wider range of possibilities. Nevertheless—even when using serial robots—the exact target motion can often not be exactly reproduced. In these cases, motion cueing algorithms (MCA) are used to produce a motion which feels as realistic as possible while remaining in the robot acceleration workspace. This paper analyzes the numerical properties of all currently existing MCA (classical, adaptive, optimal, and model predictive control) and selects the most suitable MCA using objective criteria. It also introduces a new procedure for tuning the optimal MCA such that it behaves as good as and even better than much more involved techniques based on the model predictive control (MPC). The new algorithm, termed ZyRo-K, shows best properties for reproducing the desired linear specific force while reducing the rotational false cues. While the work shown in this paper is restricted to numerical evaluation using state-of-the-art “goodness” metrics, the application and test of the algorithms for human passengers on a Robocoaster is currently being prepared and will be published in the near future.