This research presents a methodologically consistent comparative analysis of mathematical models aimed at determining the throw distance of a moped resulting from a central collision with a motor vehicle. The analysis is based on simulation data generated using the PC Crash 9.0 software, examining nine distinct vehicle speeds ranging from 20 to 60 km/h. Several regression models are applied, including third- to seventh-degree polynomials, as well as exponential and power-law models. Each model was evaluated using multiple statistical indicators, the coefficient of determination, mean absolute error, root mean square error, maximum individual error, and mean percentage error. The sixth-degree polynomial emerged as the most optimal, achieving an R2 of 0.99999, an average error of less than 0.02 m, and a maximum deviation of under 0.4%. In contrast, the exponential and power-law models demonstrated greater deviations and reduced stability. The recommended model constitutes an analytically validated polynomial function with high predictive accuracy, enabling precise modeling of the moped’s throw distance as a function of the vehicle’s travel speed. A limitation of the present study lies in the exclusive reliance on simulated data generated through a software application, without validation from experimental or real-world incident data. Furthermore, the analysis is confined to a central collision type within a restricted range of angular and mass variations. Future research is recommended to extend the model by incorporating field-acquired empirical data and investigating additional collision configurations, including partial or oblique impacts, to enable broader applicability and validation of the model across diverse scenarios.