Thesis

Current data collection methods for video-games focus on specific events or complete recordings of the gameplay sessions. The purpose of this thesis is to provide a novel method of data collection in video-games, specifically the utilisation of animation data for recording and categorising player behaviour. This alternate method is meant to be fast, with low overhead on the machine that runs it. Through this work, we conclude that animation collection is a beneficial method for data collection, and that it can be applied to future games through the development of an add-on for popular game engines.

Animations are a core component of video games. Animations typically require dedicated animators and are relatively inflexible, making it extremely difficult to animate a character without enforcing strict restrictions on the virtual world where it is placed. We propose a framework to explore the procedural generation of animations in arbitrary tridimensional virtual worlds, by using neuroevolution to create and evolve neural networks that output forces at a creature’s skeleton joints in order to produce motion and movement that are credible and physically coherent with the virtual world topology and the creature's state. Evaluation was done by selecting tasks for which our framework was able to generate animations that are different between each other but that achieve the same result proving that neuroevolution can offer different solutions to animation problems even when using different body topologies which were also able to achieve the same task using the same fitness method.

This document describes the application and development of a TORCS robot that, on a racing scenario, follows a determined trajectory (referred as racing line) calculated with the K1999 algorithm and, in case of overtaking, one that is traced by a Rapidly-exploring Random Tree based algorithm called Adapt and Overtake-RRT (ADOVER for short), working in two different modes that will later be compared. It is meant to compete against other robots and also humans, having as a requirement maintaining an acceptable performance throughout its execution. After some testing with static opponents, the robot was unable to perform the desired task. On the other hand, it showed promising results in terms of speed and efficiency. Possible improvements are discussed in the last segment.

Technology is in constant evolution, and the gaming industry is not an exception. The latest developments have been in the Virtual Reality (VR) field, with powerful hardware being developed and release by some of the world’s biggest companies, but big gaming studios have not been following the same path. VR games have the potential to take the gaming level one step forward by putting the player in the game’s virtual world and creating endless possibilities to be explored by developers’ creativity. However, in VR users can feel isolated from the people and the world around them, so shared VR experiences play an important role in the success of VR. In order to better understand the inherent challenges of the creation of VR content, we propose the development of a VR game. This game should be played by two people, stressing the need of exploring VR as a way of interaction and communication between people, instead of delivering isolated experiences. It consists in a football game with the players facing each other. It is not supposed to be a game on which players run through the pitch; instead they are in their respective goal, kicking the ball in order to score on the opposite goal, and trying to defend the other player’s kicks. By making this game a shared experience, we intend to create and strengthen ties between people through a method so propitious to this, which is simply having fun with another person.