Evaluating the Play Quality of a Game-balancing Simulation using Monte Carlo algorithm

A central aspect in game development is the user experience, which—unlike other systems—must evoke emotions related to entertainment and enjoyment. A game that is perceived as fair, with an appropriate level of difficulty and reasonable chances of winning, is considered balanced and contributes to a more rewarding experience. Monte Carlo is a simulation tool used for game balancing, enabling the simulation of matches with bots that make "intelligent" or higher-difficulty decisions. In this study, we evaluated the decisions made by bots in an Monte Carlo simulation to balance a cultural heritage game, using expert evaluations from the game’s development team. The results suggest that the appropriate number of simulations per decision depends on whether randomness is incorporated into these simulations. In the tests conducted, it was notable that when the bots performed 1000 simulations per decision while shuffling the deck after each simulation, the judges evaluated their decisions as more appropriate compared to other configurations. The findings are discussed in terms of the tool’s usefulness during the early stages of game development, as well as the implications of its application for subsequent decision-making and its impact on resource allocation throughout the development process.