This thesis investigates the use of machine learning techniques in computer games to create a computer player that adapts to its opponent's game-play. This includes first confirming that machine learning algorithms can be integrated into a modern computer game without have a detrimental effect on game performance, then experimenting with different machine learning techniques to maximize the computer player's performance. Experiments use three machine learning techniques; static prediction models, continuous learning, and reinforcement learning. Static models show the highest initial performance but are not able to beat a simple opponent. Continuous learning is able to improve the performance achieved with static models but the rate of improvement drops over time and the computer player is still unable to beat the opponent. Reinforcement learning methods have the highest rate of improvement but the lowest initial performance. This limits the effectiveness of reinforcement learning because a large number of episodes are required before performance becomes sufficient to match the opponent.

We consider the problem of strategic adversarial planning in a Real-Time Strategy (RTS) game. Strategic adversarial planning is the generation of a network of high-level tasks to satisfy goals while anticipating an adversary's actions. In this thesis we describe an abstract state and action space used for planning in an RTS game, an algorithm for generating strategic plans, and a modular architecture for controllers that generate and execute plans. We describe in detail planners that evaluate plans by simulation and select a plan by Game Theoretic criteria. We describe the details of a low-level module of the hierarchy, the combat module. We examine a theoretical performance guarantee for policy switching in Markov Games, and show that policy switching agents can underperform fixed strategy agents. Finally, we present results for strategy switching planners playing against single strategy planners and the game engine's scripted player. The results show that our strategy switching planners outperform single strategy planners in simulation and outperform the game engine's scripted AI. / Graduation date: 2013

Thesis (Ph. D.)--University of Washington, 2006. / Vita. Includes bibliographical references (leaves 149-159).

<p>The topic of this thesis is “Gender Differences in Computer games/ Video games Industry”. Due to rapid development in technology and popularization of computers all around the world, computer games have already become a kind of common entertainment. Because computer games were designed especially for boys at the very beginning, there are still some remaining barriers when training female game designers and expanding game markets among female players.This thesis is mainly based on two studies which have enormous contributions to gender issue in computer games area. A simple model is established by summarizing factors mentioned and discussed in those two books. The main purpose consists of two comparisons under Gender Differences: one comparison is between the current data with the previous one, in order to check whether there are any changes during the past 10 years. The other one compares the young people in two regions, Sweden and China, in computer games perspective.Model designing, test, questionnaire and interview methods are used in this paper aiming to collect and categorize the data, which facilitates to analyze the results of the comparisons. The results reflect that although computer becomes a familiar “friend” in modern daily life, there are not obvious changes of girls‟ perspectives in computer game industry. Certainly, there are some differences between the young people coming from two regions which will be expounded in the thesis.</p>

Thesis (Ed.D.)--University of West Florida, 2007. / Title from title page of source document. Document formatted into pages; contains 133 pages. Includes bibliographical references.

Thesis (M.S. in Computer Science)--Naval Postgraduate School, June 2004. / Thesis advisor(s): Cynthia Irvine, Paul Clark. Includes bibliographical references (p. 189-190). Also available online.

Not available / text

The focus of this thesis is the representation of football, as sport and culture, within the medium of the digital game. Analysing its three sub-genres - the televisual, the extreme and the management genre. This dissertation outlines how each configures, incorporates, rejects or ignores certain conventions, values and practices connected to the socio-cultural system of the sport, and the wider sport-media complex in its representational and ludic system. Though the marketing literature surrounding these games often makes claims to their simulational quality, these products will be shown to present very particular, dissimilar and above all ideologically loaded versions of the sport they claim to impartially represent. In an evolution of Baudrillard's (1983) concept of the simulacrum, these digital games will be revealed as euphemisms for sport, as inoffensive representations of something that may in reality be considered to offensive, or too harsh for inclusion into a marketable product. Also central to this thesis is the increasingly trans-medial nature of sport consumption, as the post-modern consumer does not simply distinguish between the various media he or she consumes and the 'reality' of the mediated object; indeed it is the mediation of the sport that now constitutes the dominant interpretation of its conventions and values. To quote John Fiske, the media no longer supply 'secondary representations of reality; they affect and produce the reality they mediate' (1994, p.xv).

With the digital revolution within video games, the need for Digital Rights Management (DRM) has increased significantly, alongside with the increasing problem of copyright pirates. To counter pirates, DRM was created to prevent illegal copying of software, this to ensure that the Distributors received an income for their work. DRM has, since the start of its use, been getting, a lot of bad criticism from the users of the software protected by DRM.  The main function of this paper is to describe the creation and development of DRM by analysis of the vision of different groups on this phenomenon. The main questions are as follows, is it possible to define the very reason for why DRM was created and if so, can its development through time be defined too? What differences in opinions are there when it comes to DRM, counting the two major groups of creators, sellers, distributors (referred to as distributors) versus individual users (referred to as consumers)? In what way will the research results suggest that the future DRM will develop? The development has gone from solving puzzles in a handbook to start the game each time the user wants to play, to serial numbers that is needed during the installation of the game. Even more extreme measures has been taken, consisting of the installation of an external software to verify the license key and ensure that no illegal actions were taken.  Distributors have shown through the years that the use of DRM is a must to protect their games from piracy. With the years that gone by, the DRM-system has developed into a more advanced software protection system and with this more problems have begun to emerge affecting the legal consumers, like errors preventing the users from playing the game. At the same time Distributors show little interest to remove or lower the usage of DRM. Users believe that the removal of DRM is the perfect solution, but discard the fact that a software without any kind of copy protection would risk not to generate any income at all for the developers.  When we consider now and then, we can see distinct patterns of continuing development of DRM-methods that do not create the same amount of issues for the consumers. The problem however previously addressed by DRM to stop illegal copies has now shifted to whether the consumers have the right to modify or change their purchased games.

No description available.