Weapons Generated for Surprise

In this project, evolutionary search is applied to the constrained problem of generating balanced and efficient sets of weapons for the Unreal Tournament III shooter game. The proposed constrained surprise search algorithm ensures that pairs of weapons are sufficiently balanced and effective while also rewarding unexpected uses of these weapons during game simulations with artificial agents. The weapons evolved in this fashion exhibit original patterns, such as the 'mine layer': its bullets are extremely slow, with a large blast area (explosive, high collision radius) and they can also bounce onwalls or the level's floor.

Relevant Publications

  • Daniele Gravina, Antonios Liapis and Georgios N. Yannakakis: "Constrained Surprise Search for Content Generation," in Proceedings of the IEEE Conference on Computational Intelligence and Games (CIG). 2016. PDF BibTex