Thesis

The video game industry has been growing over the years due to its continuous innovations and creativity. One of the existing video game genres is the music-based video game, in which the gameplay is influenced by the music. Music-based videogames had a irregular path over the years, but nowadays they are expanding not only with traditional gameplay, but with innovative devices that improve the experience. In this work, we developed a rhythm music video game prototype, with characteristics from other rhythm music-based games, in which we try to influence the gameplay with the music that is played. For this purpose, we built a game engine to support a particle system and we created a level editor to create music levels. Finally, we evaluated the game experience with users and whether the game achieves the rhythm experience of games of this genre. From such evaluation, we can conclude that our game delivers the essential experience of a music rhythm game.

Motivation for this work comes from the belief that there is some untapped potential in having a computer work as a colleague with the videogame level designer as a source of creative stimulus,instead of simply being a tool, in order to achieve overall more creative results than those obtained from solo development. The proposed solution consists of a co-creative level design tool, focused on fostering creativity by allowing human and computer to work together in producing content using the Legend of Grimrock 2 Level Editor, exploring the digital “peer” paradigm. Its interface can be used by the designer to preview generated suggestions and orient its behavior. Suggestions are generated and iteratively evolved by two genetic algorithms and can be guided by the designer on different domains:innovation,objective and user map. Innovation seeks to generate content different from the designer’sl evel, objective seeks to generate content respecting specific layout guidelines and user map seeks to generate content strongly based on the designer’s level. Results showed this approach to be promising since it takes into account the smaller nuances of the co-creative interaction as a source of a positive influence. Outlined improvements such as a better way to support designer-specific patterns, selection context or on-demand suggestion generation helped set a direction for future work. We concluded that through an intuitive interface, flexible and adjustable behavior and an agile interaction process, we were able to provide some interesting contributions to the quality of the co-creative level design process.

Games of today are extremely dependent on procedurally generated content to engage players in replaying a game. Most endless-running platformer games tend to generate their content based solely on the duration of the current playthrough, ignoring completely the skill of the player. In order to progress, the player must adapt itself to the game, instead of the game to the player. We propose a model for progression in video games, using the level of mastery of the player when using the different mechanics and overcoming the different challenges provided by the game. The model is used to determine which features (mechanics, challenges) will be presented next to the player in order to maintain the player engaged and in flow. In this dissertation, we implement this model for an endless-running platform game. A custom editor was developed to allow for level designers to specify their own game logic and the rules for guiding the progression of the game as the mastery of the player evolves. The different game features are organised in a graph with conditional transitions specifying the threshold for the level of mastery needed to enable other features down the graph. As the skill of the player evolves, the game will adapt itself in providing features of the appropriate skill level, with the final objective of increasing the level of replayability, fun and engagement thanks to the constant adaption of the game to the skill of the player.

The aim of this work is to provide a new approach on the study of identity applied to agents, taking in consideration the strengths, reviewing some weaknesses of other novel approaches, and congregating the existing models into a robust model, with the intent of using it in (but not limited to) virtual simulations of social interactions, for example, serious games. We will develop a model for identity, based upon several aspects that constitute identity described in related investigations. The idea is to integrate this model, called IDeA Model, into an architecture of a society of agents, enabling the agents to handle and comprehend identity, providing context for the behaviour of each agent. It is our belief that games which feature social interactions such as conversations, like role-playing games or simulation games, would benefit from our model to increase the idiosyncrasy of characters. Our research focuses on the dynamism of identity through the evaluation of social contexts and how an agent can use identity processes as a tool for behaviour filtering. The agent’s active identity is dependent on the agent’s resources and goals, its known social categories and the context features in which it finds itself in. It is our intention for this model to be used in serious games of Project RAGE and, as such, the agent’s believability in its behaviour is also part of the focus of this research.