BahasaMelayuMalaysiaEnglish (UK)

Automated RPG Map Synthesis Using Evolutionary Optimization Algorithms

Authors:  Alester Leongpuh and Jason teo Tze Wi


Role-Playing Games (RPG) genre has emerged over the last ten years to become one of the most popular and widely played genres of computer game. At the
moment, RPG maps are usually hard coded and can be easily memorized over time resulting in lack of replay ability for RPG players. Moreover, such maps are very
expensive and time consuming for game designers to code and develop for games companies. An automatically generated map would not only reduce the
development costs for the games but at the same time allow different maps, levels or dungeons to be generated at run-time which gives the players a new playing
experience each time the game is played, thereby significantly improving the RPG game experience and replay value. Procedural Content Generation (PCG) is the
process of automating the construction of media types for use in game development and other creative fields which allows real-time generation of game
content instead of relying on pre-generated ones. The implementation of a procedural generation method for 2D isometric maps in the FLARE game is
reported in this paper. This work uses a technique known as Cellular Automata (CA) that allows random generation of cave-like dungeons to be created
automatically in the FLARE game. The Depth First Search (DFS) algorithm is used to detect and navigate the playability of the automatically generated dungeon areas
in the map. An evolutionary optimization algorithm in the form of a Genetic Algorithm (GA) is used to tune the parameters of the CA in order to produce and
evolve the CA rules to generate playable dungeon areas in the map of the FLARE RPG game.


Download: pdf