Technology Trees and Tools: Constructing Development Graphs for Digital Games

In the recent years, digital games have solidified their role as important parts of life for a considerable portion of the population. Game development has become an extremely important industrial branch with a great deal of competition between developers and publishers. There is only a limited amount of resources to put in the development of a game, but the modern customers expect high quality.

Taking these constraints into account, this dissertation focuses on developing implementations of a structure that is used widely in different games: technology trees (TTs). This term covers here also so-called skill trees, talent trees, perk trees, and other such structures used to limit and guide in-game development and define development possibilities. The aim is to propose methods and usage of tools helping to achieve high TT quality, simultaneously facilitating the actual development process and reducing human workload.

The main contributions of this dissertation consist of ideas, models, methods, and software tool prototypes constructed during the research work. The significance of the thesis is amplified by the fact that there are only very few previous academic studies focusing on TTs.

The thesis proposes a generic approach to implement TTs. The design and implementation work are facilitated by tool support and automated code generation. The central prototype tool, Tech Tree Tool (TTT) is introduced, first in its core form and then as improved by TT measuring (and limited automatic adjusting) capabilities. The challenge of modifying TTs during runtime is addressed, also taking advantage of related improvements on TTT. Because TTs are often operated by artificially intelligent entities, discussion on a generic artificial intelligence approach and related tools is included. Moreover, contemporary real-life TTs are analyzed and generic TTs characterized.