See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/220061084

Artificial Intelligence for Computer Games

Article in International Journal of Computer Games Technology · January 2009

DOI: 10.1155/2009/251652 · Source: DBLP

CITATIONS READS

9 1,109

3 authors, including:

Kok Wai Wong

Murdoch University
220 PUBLICATIONS 2,347 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A Structural Equation Modelling Approach for Adoption of e-book Amongst Mathematics and Statistics (MAS) Students at Higher Education Institutions in Libya View
project

Sugeno-Yasukawa qualitative modeling View project

All content following this page was uploaded by Kok Wai Wong on 10 April 2014.

The user has requested enhancement of the downloaded file.

Hindawi Publishing Corporation
International Journal of Computer Games Technology
Volume 2009, Article ID 251652, 3 pages
doi:10.1155/2009/251652

Editorial
Artificial Intelligence for Computer Games

Abdennour El Rhalibi,1 Kok Wai Wong,2 and Marc Price3

1 School
of Computing and Mathematical Sciences, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, UK
2 School
of Information Technology, Murdoch University, 90 South Street, Murdoch WA 6150, Australia
3 BBC Research, Kingswood Warren, Tadworth, Surrey KT20 6NP, UK

Correspondence should be addressed to Abdennour El Rhalibi, [email protected]

Received 15 December 2008; Accepted 15 December 2008

Copyright © 2009 Abdennour El Rhalibi et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

Artificial intelligence (AI) in computer games covers the challenges as they have to satisfy rigorous requirements
behaviour and decision-making process of game-playing which involve combinatorial computation and performance.
opponents (also known as nonplayer character or NPC). In this paper, the authors investigate the performance and
Current generations of computer and video games offer behaviour of existing moving target search algorithms when
an amazingly interesting testbed for AI research and new applied to search-and-capture gaming scenarios. As part
ideas. Such games combine rich and complex environments of the investigation, they also introduce a novel algorithm
with expertly developed, stable, physics-based simulation. known as abstraction MTS. They conduct performance
They are real-time and very dynamic, encouraging fast and simulations with a game bot and moving target within
intelligent decisions. Computer games are also often multia- randomly generated mazes of increasing sizes and reveal that
gents, making teamwork, competition, and NPC modelling abstraction MTS exhibits competitive performance even with
key elements to success. In commercial games, such as large problem spaces.
action games, role-playing games, and strategy games, the The second paper is proposed by Julio Clempner and
behaviour of the NPC is usually implemented as a varia- is entitled “A shortest-path lyapunov approach for forward
tion of simple rule-based systems. With a few exceptions, decision processes.” In this paper, the author presents a
machine-learning techniques are hardly ever applied to state- formal framework for shortest-path decision process prob-
of-the-art computer games. Machine-learning techniques lem representation. Dynamic systems governed by ordinary
may enable the NPCs with the capability to improve their difference equations described by Petri nets are considered.
performance by learning from mistakes and successes, to The trajectory over the net is calculated forward using a dis-
automatically adapt to the strengths and weaknesses of a crete Lyapunov-like function. Natural generalizations of the
player, or to learn from their opponents by imitating their standard outcomes are proved for the deterministic shortest-
tactics. path problem. In this context, the authors are changing the
In this special issue, we introduce a number of interesting traditional cost function by a trajectory-tracking function
papers contributing to a wide range of these topics and which is also an optimal cost-to-target function for tracking
reflecting the current state of AI for Computer Game in the net. This makes an important contribution in the
academia. A total of 20 papers have been submitted to conceptualization of the problem domain. The Lyapunov
this special issue, of which 9 high-quality papers have been method introduces a new equilibrium and stability concept
accepted after the peer review process. in decision process for shortest path.
This special issue starts with the first paper “Performance The third paper is entitled “Fractal analysis of stealthy
simulations of moving target search algorithms” by Peter pathfinding aesthetics” authored by Coleman Ron. In this
Kok Keong Loh et al. In this paper, the authors focused on paper, the author uses a fractal model to analyze aesthetic
the design of moving target search (MTS) algorithms for values of a new class of obstacle prone or “stealthy”
computer generated bots. MTS algorithms pose important pathfinding which seeks to avoid detection, exposure, and
2 International Journal of Computer Games Technology

openness in videogames. This study is interesting since in in DEFCON, a nuclear war simulation strategy game released
general the artificial intelligence literature has given relatively in 2006 by Introversion Software Ltd, Surrey, UK. They
little attention to aesthetic outcomes in pathfinding. The describe how the AI-bot operates, and the experimentation
data reported, according to the fractal model, suggests they have performed in order to determine an optimal
that stealthy paths are statistically significantly unique in configuration for it. With this configuration, the proposed
relative aesthetic value when compared to control paths. AI-bot beats Introversion’s finite state machine automated
The author also shows that paths generated with different player in 76.7% of 150 matches played.
stealth regimes are also statistically significantly unique. The seventh paper is “Enhancing artificial intelligence on
These conclusions are supported by statistical analysis of a real mobile game” by Fabio Aiolli et al. Mobile gaming
model results on experimental trials involving pathfinding in represents a killer application that is attracting millions of
randomly generated, multiroom virtual worlds. subscribers worldwide; yet, several technical issues in this
The next paper is proposed by Frank Dignum et al. context remain unsolved. One of the aspects crucial to the
and discusses “Games and agents: designing intelligent commercial success of a game is ensuring an appropriately
Gameplay.” Multiagent system research offers a promising challenging artificial intelligence (AI) algorithm against
technology to implement cognitive intelligent NPC’s. How- which to play. However, creating this component is partic-
ever, the technologies used in game engines and multiagent ularly complex as classic search AI algorithms cannot be
platforms are not readily compatible due to some inherent employed by limited devices such as mobile phones or, even
differences of concerns. Where game engines focus on real- on more powerful computers, when considering imperfect
time aspects and thus propagate efficiency and central con- information games (i.e., games in which participants have
trol, multiagent platforms assume autonomy of the agents. not a complete knowledge of the game state at any moment).
Increased autonomy and intelligence may offer benefits for a In this paper, the authors propose to solve the imperfect
more compelling gameplay and may even be necessary for information game issue by resorting to a machine learning
serious games. However, it raises problems when current algorithm which uses profiling functionalities in order to
game design techniques are used to incorporate state of infer the missing information, and making the AI able to
the art multiagent system technology. In this paper, the efficiently adapt its strategies to the human opponent. They
authors focused on three specific problem areas that arise studied a very simple and computationally light machine
from this difference of view: synchronization, information learning method that can be employed with success, enabling
representation, and communication. They argue that current AI improvements for imperfect information games even on
attempts for integration still fall short on some of these mobile phones. They present results on a simple game called
aspects. They show that to fully integrate intelligent agents Ghosts which show the ability of their algorithm to quickly
in games, one should not only use a technical solution, but improve its own predictive performance as the number of
also a design methodology such as OperA, that is amenable games against the same human opponent increases. A mobile
to agents. phone-based version of the game has been also created which
The fifth paper presents “A multiagent potential fields can be played either against another player or against the AI
based bot for real-time strategy games” authored by Johan algorithm.
Hagelback et al. The paper discusses bots for real-time The eighth paper is “Breeding terrains with genetic
strategy (RTS). A bot controls a number of units that will terrain programming—the evolution of terrain generators”
have to navigate in a partially unknown environment, while by Miguel Frade et al. Although a number of terrain
at the same time avoid each other, search for enemies, and generation techniques have been proposed during the last
coordinate attacks to fight them down. “Potential fields” is few years, all of them have some key constraints. Modelling
a technique originating from the area of robotics where it techniques depend highly upon designer’s skills, time, and
is used in controlling the navigation of robots in dynamic effort to obtain acceptable results, and cannot be used to
environments. The authors present a multiagent potential automatically generate terrains. The simpler methods allow
field based bot architecture which is evaluated in two only a narrow variety of terrain types and offer little control
different real-time strategy game settings and compare it, in on the outcome terrain. The Genetic Terrain Programming
terms of performance, and configurability, to other state-of- technique, proposed, based on evolutionary design with
the-art solutions. The authors show that the solution is a genetic programming, allows designers to evolve terrains
highly configurable bot which can match the performance according to their aesthetic intentions or desired features.
of traditional RTS bots. They also show that a multiagent This technique evolves terrain programmes (TPs) that are
potential field-based bot is highly competitive in a resource capable of generating a family of terrains - different terrains
gathering scenario. that consistently present the same morphological charac-
The next paper is “Combining artificial intelligence teristics. This paper presents a study about the persistence
methods for learning bots in a real time strategy game” of morphological characteristics of terrains generated with
authored by Robin Baumgarten et al. The authors describe different resolutions by a given TP. Results show that it
an approach to simulate human game-play in strategy is possible to use low resolutions during the evolutionary
games using a variety of AI techniques, including simulated phase without compromising the outcome and that terrain
annealing, decision tree learning, and case-based reasoning. macrofeatures are scale invariant.
They have implemented an AI-bot that uses these techniques Finally, the last paper is “Fine-tuning parameters for
to form a novel approach to plan fleet movements and attacks emergent environments in games using artificial intelligence”
International Journal of Computer Games Technology 3

authored by Vishnu Kotrajaras et al. This paper presents the

design, development, and test results of a tool for adjusting
properties of emergent environment maps automatically
according to a given scenario. Adjusting properties for a
scenario allows a specific scene to take place while still
enables players to meddle with emergent maps. The tool uses
genetic algorithm and steepest ascent hill-climbing to learn
and adjust map properties. The authors shows that using
the proposed tool, the need for time consuming and labour-
intensive parameter adjustments when setting up scenarios
in emergent environment maps, is greatly reduced. The tool
works by converting the paths of events created by users
for a map to the properties of the map that plays out the
scenario set by the given paths of events. Test results show
good properties preservation.
Abdennour El Rhalibi
Kok Wai Wong
Marc Price

View publication stats