| Resumo: | Dynamic Difficulty Adjustment is the area of research that seeks ways to balance game difficulty with challenge, making it an engaging experience for all types of players, from novice to veteran, without making it frustrating or boring. In this dissertation we propose an approach that aims to evolve agents, in this case predators, as a group and in real time, in a way that they adapt to a changing environment. We showcase our approach after using a generic genetic algorithm in two scenarios, pitting the predators vs passive prey in one scenario and pitting the predators vs aggressive prey in another, this is done to create a basis for our approach and then test our algorithm in four different scenarios, the first two are the same as the generic genetic algorithm and in the next two we switch prey in the middle of the experience progressively from passive to aggressive or vice versa.
|
|---|