Analysis of methods for simulating character encounters in a game with RPG elements

This paper investigates algorithms that predict the outcome of a duel in a game with RPG elements and determine the losses incurred. The aim is to evaluate the effectiveness of the following approaches: based on Lanchester's laws and stochastic, using the Monte Carlo method. Verification was carried out through manual gameplay and comparison of the obtained results with those predicted by the algorithms, measuring their accuracy with the MAPE. The analysis showed greater efficiency and stability of the Monte Carlo algorithm, while the Lanchester model turned out to be less reliable in one of the cases.