Thesis

Diplomacy is a strategy board game in which 2 to 7 players compete for their supremacy over Europe in the turn of the 20th century. It is a game where luck or randomness has no role and players negotiate with each other in order to gain advantage. The board is divided into 75 provinces where 34 are called supply centers. Each turn is divided in phases dedicated to negotiation, the movement of pieces and board adjustments. At the end of each turn, each player player adjusts the number of pieces to the number of supply centers they have. The player who, by the end of a season, detains 18 out of the 34 supply centers available in the board wins the game. In this work, it was created an automated player (bot) dedicated to play the game presented above. Tagus is a bot developed for DAIDE's game platform, using the offered development kit. By negotiating with other players and using the opening libraries built in it, the bot gains advantage in the game, always with the goal of winning. The type of negotiation is simple which included only peace treaties and alliances based upon the tension and trust with other players. The experiments made reveal that negotiation brings advantage to the involved players. It was also clear that, in some cases, the opening libraries strengthen its users at the start of the game.

Procedural content generation is a technique used in many games. But if the game generates too much content this way, it can be difficult to test all possible scenarios with players before launching the game. In these cases, an AI agent is used to test the whole content, but it only performs basic validations and does not gives us the subjective feedback of a player. In this situation, it seems there is no viable way of testing large amounts of generated content. We created a methodology that uses an affective agent, to test a game using the personality and skill of a player to give us some emotional feedback on the playtesting session. The affective agent is a combination of an affective agent architecture and a personality model. We present PLEASED, an implementation of our methodology that was used as an example of our approach. To evaluate how effective is our approach, we designed and developed a game to be evaluated by PLEASED. The effectiveness of our system was measured by comparing feedback from PLEASED with players' feedback. Our results suggest that PLEASED is more suitable to simulate “casual" players than “hardcore" players and that “hardcore" players follow different criteria to evaluate a game when compared to casual ones.

Graphical Processing Units (GPUs) have evolved at a large pace, maintaining a processing power orders of magnitude higher than Central Processing Units (CPUs). As a result, the interest of using the General-Purpose computing on Graphics Processing Units (GPGPU) paradigm has grown. Nowadays, effort is being put to study probabilistic search algorithms like the Randomized Search Algorithms (RSA) family, which have good time complexity, and thus can be adapted to massive search spaces. One of those algorithms is Rapidly-Exploring Random Tree (RRT) which reveals good results when applied to high dimensional dynamical search spaces. This work consists in the design, exploration and study of the use of GPGPU-based parallelization techniques in order to improve the application of RRT to racing videogames. To approach such study, a new variant of the RRT algorithm called Iterative Parallel Sampling RRT (IPS-RRT) was developed and a bot for the TORCS open source racing game was built. The results show that, although accesses to the GPU’s memory present high latency, the use of GPGPU-based techniques like the one of this work can still improve not only the planning computational efficiency but also the quality of the returned solutions, as GPU IPS-RRT achieved temporal improvements in big problem sizes (when generating 6400 states) and lap time reductions of around 19%.

World War Online is a massive multiplayer strategy browser game made by Chilltime. This work's objective was to improve the previous battle system by reworking it completely both in gameplay and user interface, so that players would feel like they had more control over the outcome of each battle, reducing the random factor, and also to increase the number of players actively playing the game. We start by analysing the previous battle system in detail and then we analyse other games' with similar concept but with more robust battle systems. Afterwards, we describe the usability issues found on the previous battle system through user testing. As for our solution, we start by describing in detail its implementation and how it came to be by detailing the prototype evolution and its testing with users. Finally, we present the results obtained from this work for this new battle system.