Evaluating the Benefits of Software Product Lines in Game Software Engineering

Video game development is one of the fastest-growing industries in the world. The use of software product lines (SPLs) has proven to be effective in developing different types of software at a lower cost, in less time, and with higher quality. There are recent research efforts that propose to apply SPLs in the domain of video games. Video games present characteristics that differentiate their development from the development of classic software; for example, game developers perceive more difficulties than other non-game developers when reusing code. In this paper, we evaluate if the adoption of an SPL in game software engineering (GSE) can generate the same benefits as in classic software engineering (CSE) considering the case study of Kromaia. As in other disciplines dealing with human behaviour, empirical research allows for building a reliable knowledge base in software engineering. We present an experiment comparing two development approaches, Clone and Own (CaO) and an SPL in terms of correctness, efficiency, and satisfaction when subjects develop elements of a commercial video game. The results indicate that the elements developed using the SPL are more correct than those developed with CaO but do not indicate significant improvement in efficiency or satisfaction. Our findings suggest that SPLs in GSE may play a different role than the one they have played for decades in CSE. Specifically, SPLs can be relevant to generating new video game content or to balancing video game difficulty.