Dynamic Game Difficulty Balancing in Real Time Using Evolutionary Fuzzy Cognitive Maps

Players may cease from playing a chosen gamesooner than expected for many reasons. One of the mostimportant is related to the way game designers and developerscalibrate game challenge levels. In practice, players havedifferent skill levels and may find usual predetermined difficultlevels as too easy or too hard, becoming frustrated or bored. The result may be decreased motivation to keep on playingthe game, which means reduced engagement. An approachto mitigate this issue is dynamic game difficulty balancing(DGB), which is a process that adjusts gameplay parametersin real-time according to the current player skill level. Inthis paper we propose a real-time solution to DGB usingEvolutionary Fuzzy Cognitive Maps, for dynamically balancinga game difficulty, helping to provide a well balanced level ofchallenge to the player. Evolutionary Fuzzy Cognitive Maps arebased on concepts that represent context game variables andare related by fuzzy and probabilistic causal relationships thatcan be updated in real time. We discuss several simulationexperiments that use our solution in a runner type game tocreate more engaging and dynamic game experiences.