Estimating task difficulty in educational games

This paper presents a method for estimating difficulty of game tasks found in educational games. The method uses techniques from the field of socio-cognitive ICT, a branch of cognitive infocommunications. Our aim was to create an adaptive gaming experience for users of educational games. Adjusting the difficulty of game tasks according to the mental state of users is required for this approach. For some game tasks, however, difficulty is not known in advance. Our paper proposes a solution to this problem by presenting a model of an algorithm for game task difficulty estimation based on Bayesian probability theory and existing research on human intelligence. The paper also presents a simulation algorithm which is used to analyze the validity and efficiency of the estimation algorithm. An example of a real-world application of our method-an educational game made for university students of architecture and civil engineering-is also part of the paper.