Modeling Human Gaming Playing Behavior and Reward/Penalty Mechanism using Discrete Event Simulation (DES)

Humans are remarkably complex and unpredictable; however, while predicting human behavior can be problematic, there are methods such as modeling and simulation that can be used to predict probable futures of human decisions. The present study analyzes the possibility of replacing human subjects with data resulting from pure models. Decisions made by college students in a multi-level mystery-solving game under 3 different gaming conditions are compared with the data collected from a predictive sequential Markov-Decision Process model. In addition, differences in participants’ data influenced by the three different conditions (additive, subtractive, control) were analyzed. The test results strongly suggest that the data gathered from the model can possibly represent the ones gathered from the human participants in a practical experiment