Automated video game parameter tuning with XVGDL+

Usually, human participation is required in order to provide feedback during the game tuning or balancing process. Moreover, this commonly an iterative process which play-testing as well interaction for gathering all important information improve and tune components’ specification. In paper, a mechanism proposed accelerate reduce significantly costs of it, contributing with solution perform parameter using search algorithms artificial intelligence (AI) techniques. The executed fully automated way, just requires specification written particular video description language. Automated play-testing, game’s analysis, are related parameters’ balancing, leading offer problem optimizing Recently, XVGDL, new language specifying games based on eXtensible Markup Language (XML), has been presented. This paper uses XVGDL + , extension lan- guage that incorporates components specify, within specification, desirable goals requirements be evaluated after each execution. A prototypical implementation Game Engine (termed XGE ) was also engine not only enables execution but provides once finished. demonstrates combination   offers powerful helping solving AI research problems, case, parameters, respect initial optimization goals. These goals, one particularities proposal presented here, included minimizing input As practical proof two experiments have conducted optimize via hill climbing local method, way.