A Survey of Procedural Content Generation for Games

Abstract

With the continuous development of video game, the complexity of video game is gradually increasing. Besides, the cost of manually creating video game content keeps increasing although it is already very expensive. Nowadays, the cost of Game development is gradually prohibitive. In the process of a 3A game development, tens of millions development budget is spent on creating game content. In order to curb the increasing consumption of time and money, Procedural Content Generation for Games (PCG-G) has become a popular way to solve these problems by automatic game content generation. This paper reviews the PCG-G research. Firstly, we introduce what PCG-G is and its development. Then we classify PCG-G into three groups which include traditional methods, search-based methods and machine learning methods. Finally, we will show cases of PCG-G application in video games, introduce what effect PCG-G has on the game player experience and some problem of PCG-G.