Thesis

The importance of artificial intelligence in games has been growing over the years due to their continuous increasing realism and the need to keep their immersiveness while playing. Games like 8-Ball offer many interesting challenges to both communities of AI and optimization duo to the continuous and stochastic characteristics of the domain. To succeed a player must be able o plan the best sequence of shots and execute a shot with accuracy and precision, so he doesn’t lose the turn. There are already several good artificial players developed, however they tend to take more than 30 seconds to select and executed a shot. Under normal circumstances a player would give up playing the game if he had to wait that long to play. In this document I propose a real-time solution for a 8-Ball artificial player using a Monte-Carlo Expectimax hybrid search algorithm with raytracing techniques.

As the gaming industry develops, competition has become a today’s reality for most game developers and publishers. New ways to attract player’s and improve their gaming experience are being researched and developed, one of them being the adaptation of game content during gameplay. This adaptation is often based on player behavior or on player personality. In this work our objective is to improve the players gaming experience by helping him make decisions within the game. To make this possible, a game was created featuring a companion that would assist the player whenever he asks. This advices where based on the players personality using the Conscientiousness dimension and its corresponding facets of the Five Factor Model. After the evaluation of our solution we could not conclude that this approach would work for all the spectrum of the Conscientiousness dimension. Most of the subjects evaluated did not ask for a recommendation. However, for the participants that did asked for advice on which quest they should take, we got positive results for people that scored low on this dimension.

Over the years video game creators have tried to create narrative systems where the player could have the power to change the world. To have agency, and face the consequences of his choices. These systems took a variety of shapes, and a range of solutions were adopted. Regardless of the chosen approach, the deciding factor on whether these systems succeed or fail at their main goals lies in how engaged they can make the player feel with their world and their personal stories. This work addresses the issue of complex game mechanics and their relation to the player experience while engaging with an emergent narrative single-player game. In this document we describe the tests made in order to try and discern a possible correlation between the two, as well as the systems implemented to be able to carry out the tests themselves. In order to achieve this, we present the components of player experience we desire to test, such as Interest, Perceived Competence, Agency, Effort and Tension, and the testing scenarios we built to test these, using Crusader Kings 2 as a base framework. Furthermore we extensively describe our testing methodology as well as our statistical analysis of the data recovered, and how we've only found significative differences in the competence dimension. Finally we comment on the results extracted from this experiment and our suggestions for further work and exploration into this topic.

Serious games have an entirely different purpose than just entertainment: they are meant to educate, make plans, help in the decision making process or, as we will see in this dissertation, enable ill and elderly people recover from their illness and disabilities. These games’ goal is to help solve problems through interactive and fun activities. They should have a benefic effect on the players, having them make a noticeable progress throughout the entire treatment. Our aim was to develop a serious game for Parkinson’s disease patients. In this dissertation we explore the use of serious games and technologies to aid patients, by improving their lifestyle, delaying or reducing drug use, while still maintaining or improving function. We also describe the disease and its normal treatments for comparison with our solution. Our methodology is based on a game developed with Unity and using Kinect, where players have to perform a series of meaningful tasks (challenges) that aim to have the benefits previously referred, while being monitored by the therapists and doctors, who also control the parameters on which the game runs, tailored to each specific patient. The main goals were achieved, with proven satisfaction and great feedback both from the patients and doctors/therapists.