handbook of digital games

Angelides and Harry Agius Electronic and Computer Engineering, School of Engineering and Design, Brunel University, Uxbridge, Middlesex, United Kingdom In 1978, in his now classic Platon

Free ebooks ==> www.Ebook777.com

www.Ebook777.com

Free ebooks ==> www.Ebook777.com

www.Ebook777.com

Handbook of Digital Games

IEEE Press

445 Hoes LanePiscataway, NJ 08854

IEEE Press Editorial Board

Tariq Samad, Editor in Chief

George W Arnold

Dmitry Goldgof

Ekram Hossain

Mary LanzerottiPui-In MakRay Perez

Linda ShaferMengChu ZhouGeorge Zobrist

Kenneth Moore, Director of IEEE Book and Information Services (BIS)

Free ebooks ==> www.Ebook777.com

Copyright © 2014 by The Institute of Electrical and Electronics Engineers, Inc.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey All rights reserved

Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form

or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee

to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/ permissions.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts

in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose No warranty may be created or extended by sales representatives or written sales materials The advice and strategies contained herein may not be suitable for your situation You should consult with a professional where appropriate Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic formats For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data is available.

9781118328033

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

Contributors    ix

Marios C Angelides and Harry Agius

Part I Gaming Techniques and Tools

vi     Contents

Jon Lau Nielsen, Benjamin Fedder Jensen, Tobias Mahlmann,

Julian Togelius, and Georgios N Yannakakis

11.  Rated A for Advertising: A Critical Reflection on In-Game  

Laura Herrewijn and Karolien Poels

Part II Game Play

Ana Belén García Varela, Héctor Del Castillo, David Herrero,

Natalia Monjelat, and Mirian Checa

Free ebooks ==> www.Ebook777.com

www.Ebook777.com

Harry Agius, Brunel University, Uxbridge, Middlesex, United Kingdom

Marios C Angelides, Brunel University, Uxbridge, Middlesex, United Kingdom Chris Bateman, University of Bolton, Bolton, Greater Manchester, United

Kingdom

Robin Baumgarten, Imperial College, London, United Kingdom

Tilde Bekker, Eindhoven University of Technology, Eindhoven, The Netherlands Tom Betts, University of Huddersfield, Huddersfield, Yorkshire, United Kingdom Johannes Breuer, University of Münster, Münster, Germany

Cameron Browne, Imperial College, London, United Kingdom

Quinn Burke, College of Charleston, Charleston, South Carolina, USA

Paul Cairns, University of York, York, North Yorkshire, United Kingdom

Shannon Campe, Education, Training, Research, Scotts Valley, California, USA Mirian Checa, University of Alcalá, Alcalá de Henares, Spain

Francesco Collovà, Telecom Italia, Naples Italy

Simon Colton, Imperial College, London, United Kingdom

Michael Cook, Imperial College, London, United Kingdom

Sara Cortés, University of Alcalá, Alcalá de Henares, Spain

Anna Cox, University College London, London, United Kingdom

Fabrizio Davide, University of Rome Tor Vergata, Rome, Italy

Damon Daylamani Zad, Brunel University, Uxbridge, Middlesex, United Kingdom Mark de Graaf, Eindhoven University of Technology, Eindhoven, The Netherlands Celso M de Melo, University of Southern California, Los Angeles, California, USA Héctor Del Castillo, University of Alcalá, Alcalá de Henares, Spain

Jill Denner, Education, Training, Research, Scotts Valley, California, USA

Arjan Egges, Utrecht University, Utrecht, The Netherlands

Malte Elson, University of Münster, Münster, Germany

Contributors

ix

x     Contributors

Cathy Ennis, Utrecht University, Utrecht, The Netherlands

Stefano Ferretti, University of Bologna, Bologna, Italy

María Ruth García-Pernía, University of Alcalá, Alcalá de Henares, Spain

Ana Belén García Varela, University of Alcalá, Alcalá de Henares, Spain

Jeremy Gow, Imperial College, London, United Kingdom

Lindsay D Grace, American University, Washington, DC, USA

Jonathan Gratch, University of Southern California, Los Angeles, California, USA Barbara Grüter, Hochschule Bremen, Bremen, Germany

Orlando Guevara-Villalobos, University of Edinburgh, Edinburgh, United Kingdom Nassrin Hajinejad, Hochschule Bremen, Bremen, Germany

D Hunter Hale, University of North Carolina at Charlotte, Charlotte, North Carolina, USA

David Herrero, University of Alcalá, Alcalá de Henares, Spain

Laura Herrewijn, University of Antwerp, Antwerp, Belgium

Sheng-yi Hsu, National Chiao Tung University, Hsinchu City, Taiwan

Peter Jamieson, Miami University, Oxford, Ohio, USA

Benjamin Fedder Jensen, IT University of Copenhagen, Denmark

Yasmin B Kafai, University of Pennsylvania, Philadelphia, Pennsylvania, USA Pilar Lacasa, University of Alcalá, Alcalá de Henares, Spain

Janne Lautamäki, Tampere University of Technology, Tampere, Finland

Tobias Mahlmann, IT University of Copenhagen, Denmark

Natalia Monjelat, University of Alcalá, Alcalá de Henares, Spain

Ivan Mosca, University of Turin, Turin, Italy

Jon Lau Nielsen, IT University of Copenhagen, Denmark

A Imran Nordin, University of York, York, North Yorkshire, United Kingdom Eloy Ortiz, Education, Training, Research, Scotts Valley, California, USA

Ana Paiva, IST—Technical University of Lisbon, Lisbon Portugal

Karolien Poels, University of Antwerp, Antwerp, Belgium

Thorsten Quandt, University of Münster, Münster, Germany

Henrik Schoenau-Fog, Aalborg University, Copenhagen, Denmark

Ben Schouten, Eindhoven University of Technology, Eindhoven, The Netherlands

Contributors    xi

Iaroslav Sheptykin, Hochschule Bremen, Bremen, Germany

Helen Stuckey, Flinders University, Adelaide, Australia

Sybren A Stüvel, Utrecht University, Utrecht, The Netherlands

Chuen-Tsai Sun, National Chiao Tung University, Hsinchu City, Taiwan

Melanie Swalwell, Flinders University, Adelaide, Australia

Carl Therrien, Université de Montréal, Montreal, Canada

Julian Togelius, IT University of Copenhagen, Denmark

Stefano Triberti, Università Cattolica del Sacro Cuore, Milan, Italy

Juha-Matti Vanhatupa, Tampere University of Technology, Tampere, Finland Hao Wang, National Chiao Tung University, Taiwan

Linda Werner, University of California, Santa Cruz, California, USA

Georgios N Yannakakis, IT University of Copenhagen, Denmark

G Michael Youngblood, University of North Carolina at Charlotte, Charlotte, North Carolina, USA

Marios C Angelides and Harry Agius

Electronic and Computer Engineering, School of Engineering and Design, Brunel University, Uxbridge, Middlesex, United Kingdom

In 1978, in his now classic Platonic book The Grasshopper: Games, Life and Utopia,

the philosopher Bernard Suits wrote of a future in which the only human activity is game playing and where the human race has developed what he calls the lusory effect, a psychological attitude required of a game player entering into the play of a game In complete contrast to the perception on working life and Wittgenstein, Suits argued that game playing is a voluntary attempt to overcome unnecessary obstacles and that playing games is a central part of the ideal of human existence; thus games belong at the heart of any vision of utopia By the time of his death in early 2007, Suits should have been able to witness proof of a Darwinian evolution of the gaming utopia he foresaw 30 years before in the social fabric of modern life, which is driven

by social connectivity, shared experiences, and collaboration, whether in real or virtual worlds Be it political games, military games, business games, or recreational games, the boundaries between what is real and what is virtual are fused All are but clues to our future, Suits argues, concluding that cultivation is our salvation.Since Suits outlined his original vision, the cultivation of digital games has seen them grow at a phenomenal rate into a multi-billion-pound/dollar industry with a strong market share in the entertainment industry and heavily reliant on the rest of the industry for development and promotion The digital game culture has gradually been shifting from the pull culture of the arcade to the push culture of the mobile device The digital game culture began to shift in the early 1980s, along with the average age of the gamer, neither in the juvenile nor in the early adolescent range, but in the mid-30s range Digital games are now played on a broad range of portable and fixed game consoles rather than being limited to a single technical platform, including desktop and tablet PCs, dedicated consoles, and mobile phones We now have a popular game culture; games are imitating art and vice versa, game genera-tions are fast producing game decay or bit rot, and game preservation has taken on

Handbook of Digital Games, First Edition Edited by Marios C Angelides and Harry Agius.

© 2014 the Institute of Electrical and Electronics Engineers, Inc Published 2014 by

John Wiley & Sons, Inc.

1

2     Introduction

the role of saving our game history Nowadays, it is the norm to expect a popular game to become a movie and vice versa

Regardless of device and virtual environment, social connectivity allows people

to share ubiquitous experiences anywhere, at any time, on any device, and over any network, often adapted to their individual needs and desires The Web has been a huge contributor to the growth of digital games particularly via massively multi-

player online games (MMOGs), such as World of Warcraft (WoW) and The Sims,

and social networking sites, such as Facebook Social networking sites in particular have had a huge effect on the gaming community in a relatively short space of time, not least by raising the potential and actual audience of players that a game has to serve, even eclipsing those hitherto held by massively multiplayer online role-playing games (MMORPGs) Nevertheless, digital games critics argue that playing games is at best recreational and at worst desensitizing and degenerate and subse-quently no match for the education and literacy that comes from reading physical books, despite their lack of interactivity, lack of fellow readers to share the experi-ence with while consuming the book, and absence of ubiquity Despite the polemic, digital games have justifiably earned their place high up on the list of “new” com-putational media

All new media that led to the creation of lasting academic and/or industrial communities, when they emerged, shared two properties: They were mass media and they told stories that allowed us to reflect on what it means to be human The cinema is one such example When it emerged as new media in the early twentieth century, it gave birth to a lasting community of scholars and numerous industry-based communities Computer games, which are interactive, have complex rules and intri-cate real-time computer graphics have the ability to tell rich stories and provide

social commentary With games such as SimCity, which requires public transport systems to achieve large-scale cities, Civilization, which provides a technology- driven view of the march of history, or Grand Theft Auto: San Andreas, with its

graphic representation of street crime, it is clear that games tell stories and exhibit

a kind of rhetoric based on the ideas baked into the underlying computational cesses Both cinema (older media) and games (newer media) currently have research and industry-based communities focused on exploring the sociological and human-istic elements of each media As a consequence, modern computer games are the product of multidisciplinary research and development that exhibit constant technical advances and innovations in their game engines across many disciplines: computer graphics, AI (artificial intelligence), HCI (human–computer interaction), databases, network technologies, arts, social sciences, and the humanities, to name just a few

pro-It is not surprising that games resulting from multidisciplinary developments vary widely across communities However, six common and potentially definitive

characteristics define most games, if not all: rules, variable and quantitative comes, valorization of outcome, player effort, player attachment to outcome, and

out-negotiable consequences Furthermore, gamers, in their attempts to categorize games

by their structure and content, build a game landscape that converges on four

clus-ters: strategy games, first-person shooters (or rather games where the player controls

an avatar in the game space based on vagrant positioning and camera placement),

Introduction    3

progression and exploration games (such as exploration of story, character, or game

world), and perfect information games (where all information on the game state is

available to the player, sharing a resemblance to traditional physical games like Chess or Go)

On the contrary, research on games is largely focused on game design with the latest research efforts seeking to exploit technologies from other complementary areas of computing and other disciplines in order to enable players to enjoy a ubiq-uitous gaming experience anywhere, at any time, on any device, and over any network that is adapted to their individual needs and desires, such as through rec-ognition of their gaming prowess and effective opponent and teammate matching The increasing research attention drawn by the player experience has gradually extended to the social and cultural aspects during game play and cross-cultural

analysis of games For example, in WoW, a data-mining infrastructure gathers and

processes character data which are stored in an online character repository As a result, game design now exploits, for example, both static and dynamic representa-tion of game semantics, human and nonhuman autonomous player behavior predic-tion, team dynamics, avatar evolution, game world customization, story narratives that evolve with player behavior, and intelligent techniques for such processing

A nonexhaustive list of current research areas includes:

• Game design, which has not yet established a theoretical basis for creating

the virtual game spaces but is performed intuitively

• Game architectures and environments, which enable routine construction of

game engines and support environments Three-dimensional (3D) games are currently built on top of a game engine that provides 3D model display and animation, collision detection, effects, AI, level design, and so on

• Intelligent narrative technologies that deliver coherent, linear, and customized

story flows that afford a player a high degree of agency in the world, that is,

to move freely and perform actions as they wish This involves developing representations of story structures that can be reasoned over and planned from

to deliver a customized story experience to the player which requires a variety

of techniques ranging from game design to HCI to AI

• Procedural content generation spanning very large virtual spaces which are

not completely generated by human authors This requires game level, player, quest, background history, and asset design which in turn require computer graphics, animation, databases, and intelligence techniques

• Interfaces which create novel game play experiences The success of forms such as Wii and games such as Guitar Hero yields nonstandard inter-

plat-faces to the general public These by and large require HCI techniques

• Real-time computer graphics that are capable of execution in real time on

current graphics hardware and techniques for animating players, performing lighting of game worlds, nonphotorealistic rendering, and collision detection

• Databases which are capable of supporting large numbers of simultaneous

users interacting in MMOGs

4     Introduction

• Networking that is capable of supporting multiplayer play while maintaining

the consistency of the game world for all players in the face of rapid ment and frequent interactions This requires techniques for dead reckoning and determining a player’s line of sight

move-• Games for learning and education that develop new theories for exploiting

game rules and worlds for enabling learning This is gaining momentum for teaching traditional school subjects At present, developing a game experience

to teach specific kinds of knowledge is largely a game of skill and drill repetition

• Nonhuman autonomous players that are able to interact with human players,

express emotion, react in appropriate ways, and take effective action during game play This requires AI techniques, including natural language processing, animation, and representations of nonhuman autonomous player profiles

• Player recognition that tracks and records player actions for the purpose of

individualizing the game experience This requires AI techniques and sentations of human player profiles

repre-• In-play player impact assessment that assesses the impact of in-play game

design changes on players This requires game play metrics, AI techniques, and either static representations or dynamic generation of game semantics profiles

• Platform recognition that enables games to become platform aware in order

to match players on similar platforms to one another based on their device capabilities This requires AI techniques and device representations

CHAPTER SUMMARIES

This handbook comprises chapter contributions from leading researchers and opers worldwide which are grouped into three broad parts spanning the research

devel-areas, both classic and emerging, outlined above: gaming techniques and tools, game

sections and therefore we have attempted to map the chapters to a single part based

on their primary focus The primary audiences of the chapters are game industry professionals and the growing interdisciplinary body of university academics and researchers who work in the digital game area as well as areas associated with digital games, such as game studies and design, social media, and all aspects of game development A secondary audience is professional gamers and informed consumers seeking a deeper technical understanding The parts and their chapters are now sum-marized in turn

Part I: Gaming Techniques and Tools

The chapters in this first part are concerned with a diverse range of techniques and tools for digital games, encompassing adaptive and procedural content generation,

Introduction    5

automatic narratives, collision detection, simulation of crowds, network issues such

as synchronization, sharing of social information, collaboration, advertising, and the use of AI techniques for simulating game play

Chapter 1

In this opening chapter, which is the first of three chapters on content generation, researchers from Imperial College, London, explore methods for automatically gen-erating game content and games that are adapted to individual players through the modeling of player needs They identify and discuss three main aspects: generation

of new content and rule sets, measurement of this content and the player, and tion of the game to change player experience Various types of games are presented

adapta-to illustrate their approach

Chapter 2

In the second chapter on content generation, the author from the University of Huddersfield, United Kingdom, surveys the state of the art in the increasingly impor-tant area of procedural content generation (PCG), whereby algorithmic methods are used to produce game content in order to satisfy the demand for complex detail and behavior in digital games He discusses common areas of PCG implementation such

as fractal terrain, RPG (role-playing game) loot generation, enemy placement, and resource distribution as well as more diverse areas such as mission objectives, dialog trees, character profiles, and behavior patterns and even emergent areas such as AI behavior and dynamic autonomous environments

Chapter 3

In the third and final chapter on content generation, researchers from Tampere University of Technology, Finland, argue that procedural content generation is not well explored in browser environments and therefore utilize content generation methods to create content for a multiplayer browser-based fantasy game, where all the quests are generated dynamically at run time based on quest templates They identify problematic areas of game design where PCG can offer valuable solutions, consider active-versus-preparatory PCG, describe common PCG content types and their production, and present the limitations and potential for PCG in game design They show that their approach can supplement precreated content, expand overall content, and increase replayability

Free ebooks ==> www.Ebook777.com

6     Introduction

MMORPGs because they tend to spend much longer in these games than others and because allowing players to live their own stories through decisions and actions constrains the capacity for delicate story authoring To tackle these issues, he pres-ents some narrative intelligence techniques that can be used to address these prob-lems to some extent He also proposes methods for creating player memorials of in-game actions, such as video clips and comics based on game logs, so that players can remember and reminisce about their achievements

at run time Their results demonstrate a 50% decrease in collision detection time between dynamic objects in comparison to k-d trees and show that navigation mesh accelerated collision detection outperforms spatial hashing accelerated collision detection across all tests

Chapter 6

A research team from Utrecht University, The Netherlands, reveals how ments in gaming hardware and realism have made it possible to populate virtual worlds with high numbers of characters such that background crowds are able to give the player an increased sense of presence Consequently, they examine the origins of crowd simulation, look at academic research approaches, and give practi-cal guidelines on how to create crowds in virtual environments so as to minimize the resource expense while maintaining the sense of realism for the player A com-pilation of metrics and results from perceptual studies forms usable guidelines for optimizing crowd behavior for a particular game

develop-Chapter 7

In this chapter, the author from the University of Bologna, Italy, overviews some of the main issues and proposed solutions for synchronizing distributed multiplayer online game nodes in a responsive and reliable way, catering for different dis-tributed architectural solutions, such as client/server, peer-to-peer, and distributed (mirrored) game server architectures He argues that, since multiple nodes may be employed to manage the same, redundant, portion of the game state, a high reli-ability and fault tolerance is ensured, but this is at the cost of requiring consistency management algorithms to be executed by these nodes Since MOGs have strict responsiveness requirements, it is not possible to resort to traditional synchroniza-tion algorithms

www.Ebook777.com

Introduction    7

Chapter 8

Users play together to achieve goals or to conflict with each other and beat their opponents This is the basis of digital social games and, in this chapter, a research team from Italy spanning the University of Rome Tor Vergata, Università Cattolica del Sacro Cuore, Italy, and Telecom Italia analyzes the key concepts of social infor-mation, that is, the primal material of social game interactions They introduce a classification of current digital social games and consider social information and its presence in different games They then elaborate on three case studies of social games and provide schemes of how social information is implemented and used in a digital gaming context A model of social information exchange is introduced and developed

at the level of scenarios, communication process, and most relevant messages

Chapter 9

Continuing the exploration of collaborative gaming, a research team from Brunel University, United Kingdom, considers the development of multipurpose collabora-tive games which integrate both lusory and ludic dimensions They propose a frame-work incorporating both dimensions and present the implementation of a collaborative supply chain game They demonstrate that such collaborative games are both resolute and entertaining

Chapter 10

In contrast to the previous two chapters, researchers from the IT University of Copenhagen, Denmark, consider the use of AI-infused players They argue that, while strategy games are closely related to classic board games such as Chess and

Go, there has been little work on the use of AI for playing strategy games They therefore consider how to create AI that plays strategy games through building and comparing AI for general strategy game playing

Chapter 11

Moving away from game mechanics, research from the University of Antwerp, Belgium, considers the emergent area of advertising within digital games They discuss the history and taxonomy of the use of advertisements in games, showing how brands can be integrated into digital game environments and how the phenomenon has evolved throughout the years They consider the current effectiveness and future prospects of the advertising medium, discussing and juxtaposing research concerning people’s awareness and evaluation of in-game ad placements and examine how gamers really think and feel about commercial practices inside their favorite games

Part II: Game Play

The second part of the book brings together chapters that are focused on various aspects of game play, accommodating immersion, player experience, game aesthetics,

8     Introduction

mobile game play, meaningful gaming for education, and retrospective examinations

of gaming and game play

Chapter 12

In the opening chapter of this part, researchers from the University of York and University College London, both in the United Kingdom, review immersion in rela-tion to other concepts that are used to describe gaming experiences These include concepts that are not specific to games such as flow and attention; generic concep-tualizations of the gaming experience of which immersion may form a part, such as incorporation; and specific concepts around immersion, engagement, and involve-ment such as presence and other formulations of immersion They describe an experiment that positions immersion in relation to presence, thereby providing an empirically founded understanding of these rich, subjective experiences

Chapter 13

This chapter argues that studying the experiences of game players is a nontrivial undertaking due to the dynamic, interactive, and complex nature of the media Consequently, researchers from the University of Münster, Germany, propose an integrated model of player experience (IMP) that distinguishes between the preuse (choice), use (play), and postuse (effects) phase and accounts for personal (player traits and states), media (game characteristics), and contextual (setting and social environment) variables Based on the IMP, they provide an overview of available means to study player experience and describe how they have been and can be used and what advantages and disadvantages they have Their purpose in doing so is to guide each step from formulating research questions and hypotheses, the operation-alization of variables, and the selection of suitable research methods when carrying out user-centered game studies

Chapter 14

In this chapter, the author from Aalborg University, Denmark, explores player engagement by investigating how it is described in relation to digital games, how it can be evaluated, and how it can be used in the design of gamelike applications In particular, he focuses on the willingness of the player to continue playing, termed

“continuation desire.” He argues that this is an essential consideration when ing and evaluating digital games and interactive narratives Consequently, he pro-poses a model of continuation desire, which is based on an empirically identified range of causes of the desire to continue playing, and various methods that can be used to assess and evaluate continuation desire are described in order to illustrate how it is possible to assess the levels of continuation desire experienced by players while playing and when returning to play a game The practical use of the continu-ation desire model and the evaluation methods are demonstrated through a case study

design-of an interactive storytelling application: the “First-Person Victim.”

Introduction    9

Chapter 15

In this chapter, the author from the University of Bolton, United Kingdom, considers games from the point of view of them being aesthetic objects and therefore argues that they can be understood using theories originating in the philosophy of art, and the players of such games can be studied by empirical investigations into the aes-thetic values that influence their choice of game Thus, by tracing a series of player satisfaction models and positioning these in the context of other work in the area,

he extrapolates answers to the questions of how and why people play games He argues that understanding aesthetic preferences for particular kinds of play in terms

of the underlying neurobiological substrates associated with the emotions of play may provide the basis for establishing an empirically derived trait theory of play

Chapter 16

A research team from Hochschule Bremen, Germany, considers mobile games, such

as Angry Birds, Pig Rush, or Tiny Wings, as rule systems based on the physical

movement of a player in a world merging the real world with virtual dimensions They argue that the changing context of play transforms the play experience and opens up new design possibilities and consequently the chapter focuses on gaining

a comprehensive understanding of mobile game play and the particular way of playing mobile games, uniting both traditional and novel facets of gaming and play

At the center of their study are games as systems, the contexts of play, and the ties of the players By analyzing relationships between game systems and contexts

activi-of play, they conclude that mobile game play is a moment activi-of everyday life activity

technolo-Chapter 18

Another research team from the University of Alcalá, Spain, continues the digital games for education theme by exploring the different roles that commercial video

10     Introduction

games can play as integrated learning tools in primary and secondary education Their goal is to design innovative educational contexts which contribute to creating responsible citizens who possess a critical awareness of the new communication scenarios provided by today’s technology Furthermore, they hope to understand how commercial video games can inspire a motivation to learn and develop thinking skills From their data, they observe how certain commercial video games allow a hidden curriculum to arise, making it possible to develop specific thought processes and skills acquisition while promoting positive attitudes such as a respect for the environment or collaboration with others Their chapter is based on the data collected during a large research project, the aim of which is to explore how commercial video games provide innovative educational opportunities in the classroom that bring children and adults into the new digital scenario They examine relationships between real and virtual universes as situated cognition processes involving game situations based on different simulation videogames, such as SimCity

Chapter 19

Researchers from Flinders University, Australia, report on over a decade of work to research, collect, and preserve the production and reception histories of local digital games in 1980s Australia and New Zealand “Play It Again” is a collaborative project between researchers at several Australasian universities and three cultural institu-tions, the ACMI (Australian Centre for the Moving Image), the New Zealand Film Archive, and the Berlin Computerspiele Museum, where engagement with retrogam-ing and other communities is central to the approach of the project

Chapter 20

In the final chapter of this part, the author from the Université de Montréal, Canada, argues that, as digital games have evolved, game designers have sought to create more complex experiences without alienating potential players Consequently, he presents a retrospective study that focuses on the rise of the cooperative mode of address in game design, which he defines as the way game publishers and designers have addressed potential players in a more inclusive way than the competitive para-digm associated with the early days of gaming in arcades It is discovered that this paradigm encompasses the overt address to players in promotional material as well

as the implicit address inscribed in the various systems that take part in the sive experience

immer-Part III: Game Design and Development

The final part draws together chapters that focus on the design and development aspects of digital games The chapters address a range of topics: emotion in games, spatial game structures, ontological analysis of digital games, entertainment software design theory applied to human computation games, gender differences in game

in players, simulate emotion in nonplayer characters, and interpret the players’ tions In doing so, they review relevant psychological theories of emotion and com-putational models of emotion and discuss their implications for games.

emo-Chapter 22

An important aspect of digital games is its spatial structure Researchers from the National Chiao Tung University, Taiwan, argue that this is determined by the system architecture and program code and consequently analyze the deployment of puzzles and quests in varied spatial structures and the mechanisms for players to balance their skills with the current challenges so as to sustain their gaming flow They reveal how other gaming factors, such as storylines, resource allocation, and reward sys-tems, match such spatial structures in game design to provide coherent gaming experiences Categorizing spatial structures into three basic types (ladder, maze, and grid), they map their association to game genres and corresponding design principles and introduce a hierarchical architecture for hybrid games, which maintain the play-ers’ sense of balance and fairness in terms of game task arrangement

Chapter 23

The author from the University of Turin, Italy, argues that designers and mers constantly use nạve interpretations about games and therefore proposes a social ontology of games that can lead designers and programmers to develop games without subjective points of view, but rather with an objective knowledge of the fundamental game properties He argues that the ontology of games is always social and examines three examples of ontological analysis of games: the structure of gaming interaction, the role of rules in digital games, and their simulation attempt

program-Chapter 24

Players are capable of solving difficult problems through human computation games while ubiquitous gaming provides opportunities to solve those problems Con-sequently, in this chapter, researchers from Miami University, Florida, outline the

12     Introduction

characteristics of human computation games and ubiquitous games in a variety of disciplines, describing they key components of such solutions and articulating their distinguishing characteristics from other types of entertainment software They focus

on entertainment software design theory as applied to human computation games, outlining the fundamental characteristics of such games and offering approaches for applying human computation games to promote player engagement and adopt application

Chapter 25

In this chapter, researchers from ETR Associates and the University of California, Santa Cruz, argue that the essentialization of gender apparent in the stereotypes found in the most popular video games is off-putting to many girls Consequently, they present their study aimed at better understanding the role of gender stereotypes

in the gaming preferences of girls and boys and the conditions under which they vary Analyzing the content and game mechanics of 231 games made by middle school girls and boys in the United States, they found that girls’ games were more likely to focus on the players’ experience and to engage the player in a storyline, social relationships, conversations, and problem solving for the social good, while boys’ games were more likely to include violence against nonhumans and objects and focus on victory, competition, or conquest and take place in larger-than- life settings However, their further analyses revealed that these gender differences were better explained by prior computer and gaming experience and whether or not students made their game with a partner or alone

Chapter 26

Focusing on children’s learning through building video games, researchers from the College of Charleston and the University of Pennsylvania ask the question, what kind of building and what kind of learning is going on in making games? In answer-ing this question, they investigate a decade’s worth of research surrounding chil-dren’s learning through designing and building their own digital games: first by examining the sharp spike in various software applications specifically geared to allow youth to create their own video games and second by exploring children’s learning as they interact online and the overall nature of game-making communities

in fostering creative collaboration among youth They highlight successes and propose curricular and pedagogical recommendations for a more seamless incorpo-ration of game-making technologies and approaches into schools

Chapter 27

In this chapter, researchers from Eindhoven University of Technology, The Netherlands, examine how various properties of play have inspired and can inspire new design directions for digital games and intelligent play objects Play theories from a child development perspective are described and related to concepts from

Introduction    13

game design, such as game mechanics and dynamics They also discuss how ent properties of play relate to children practicing social, emotional, physical, and cognitive skills in a playful and fun context A well-known model of digital game design is the mechanics, dynamics, and aesthetics (MDA) model, which attempts to bridge the gap between game design and development, game criticism, and technical game research, and the authors present an adapted version of the MDA model as a tool that supports considering the richness of play design opportunities when creat-ing dynamics, mechanics, and aesthetics for diverse forms of play from a designer’s and a player’s perspective They illustrate the application of the adapted model by describing four design case studies of tangible intelligent play concepts for different contexts of use related to different properties of play: an interactive storytelling mat for young children, an intelligent ball pit for young children, a system that supports children sharing the use of bikes during school play time, and intelligent play objects for a trading game with a design intention of supporting social interaction

he attempts to describe and reflect on the daily work practices of independent game developers, with special emphasis on a local network in Cambridge, and to provide

an understanding of the meanings and functions of the microsocial relationships that structure their process of game making, both spatially and procedurally He con-cludes that local regular activities where independent developers participate work as spaces of learning, practice, and informal transactions that can help technically, creatively, and motivationally those interested in game development Within these relationships, developers meet the needs of knowledge-based, artistically driven, and entrepreneurially oriented capitalist enterprises such as digital game production

Gaming Techniques and Tools

Part I

Toward the Adaptive

Generation of Bespoke

Game Content

Cameron Browne, Simon Colton, Michael Cook,

Jeremy Gow, and Robin Baumgarten

Computational Creativity Group, Department of Computing, Imperial

College, London, United Kingdom

In this chapter, we explore methods for automatically generating game content—and games themselves—adapted to individual players in order to improve their playing experience or achieve a desired effect This goes beyond notions of mere replay-ability and involves modeling player needs to maximize their enjoyment, involve-ment, and interest in the game being played We identify three main aspects of this

process: generation of new content and rule sets, measurement of this content and the player, and adaptation of the game to change player experience This process

forms a feedback loop of constant refinement, as games are continually improved while being played Framed within this methodology, we present an overview of our recent and ongoing research in this area This is illustrated by a number of case studies that demonstrate these ideas in action over a variety of game types, including 3D action games, arcade games, platformers, board games, puzzles, and open-world games We draw together some of the lessons learned from these projects to comment

on the difficulties, the benefits, and the potential for personalized gaming via tive game design

adap-1.1  INTRODUCTION

Personalization of games for individual players is seen as a significant future ing factor for games and is currently a major driving force for improved game design,

market-Chapter 1

Handbook of Digital Games, First Edition Edited by Marios C Angelides and Harry Agius.

© 2014 the Institute of Electrical and Electronics Engineers, Inc Published 2014 by

John Wiley & Sons, Inc.

17

18 Chapter 1 Toward the Adaptive Generation of Bespoke Game Content

which will ultimately lead to better games and happier and more engaged, and entertained customers Within this scope, there is a particular nirvana wherein games automatically adapt before, during, and after being played to take into account the style, experience, and personality of each player Of course, games have always had

a simplistic adaptive element, whereby stronger players progress to play more ficult levels to keep them interested However, this type of adaptation only takes into account their skill level at that particular game and ignores other information such

dif-as their likes, dislikes, temperament, current mood, and overall ability Such tion can in principle be gathered through game play, sensors, surveys, and other routes and will be used in adaptive gaming technologies of the future to generate bespoke games that truly change to fit an individual player, greatly enhancing their playing experience

informa-The automatic adaptation of games to players is also a major force for applied artificial intelligence (AI) research In particular, as a research group, in addition to the long-term goal of improved games, we are also interested in studying games from the perspective of the subfield of AI known as computational creativity research [17] In this area, we study how to engineer software which can take on some of the

creative responsibility in arts and science projects In this context, games, video games in particular, can be seen as a “killer domain” for creativity research This is largely because generating a game requires the generation of all the types of artifact

we usually produce individually, including audio (sound effects, music), graphics (characters, backdrops), text (dialogue, plotlines), and concepts (puzzles, rule sets, interaction schema, game mechanics) However, there are many other advantages to working with games as a medium within which to study computational creativity These include (a) the fact that the output is entirely digital and the audiences are entirely online, hence requiring no exhibitions, concerts, readings, publications, or demonstrations in order to get culturally relevant feedback; (b) a general acceptance

of automated processes as being valuable, which is not always true in more tional artistic circles; (c) a requirement to model and ultimately alter both positive and negative emotions; (d) an interesting balance between the entertainment value and the intellectual value of games; and (e) explicit requirements to incorporate user engagement and interaction in the generated artifacts

tradi-As a group of computational creativity researchers and avid game players, over the last five years, we have eagerly investigated the potential for automating pro-cesses related to game design, with the specific long-term goal of adaptive game generation in mind We see adaptive systems in games—also known as AI directors

or game masters—as a form of procedural content generation which aims to enhance the players’ gaming experience by delivering personalized game content When thinking about such adaptive systems, it helps to consider various aspects such as

the type of player data, the types of decisions to be made about game content (the content output space ), how the latter is computed from the former (the adaptive mechanism ), and the desired effect on player experience (the adaptation require- ments) We have studied the potential for adaptive games with a shotgun approach, that is, numerous projects involving games of various genres, which address all the above aspects We present here an overview of some of these projects in order to

Methodology 19

highlight the lessons learned, difficulties encountered, and huge potential for tive game technology to both help produce next-generation games and stimulate research in computational creativity

adap-In Section 1.2, we describe an overall methodology within which content eration for adaptive games can take place This centers around a cycle of generation, measurement, and adaptation, and we expand each of these aspects further With respect to generation, we place this in a context of search-based procedural content generation and focus on two types of evolutionary search With respect to measure-ment, we split this into measuring the game, measuring the player, and measuring the adaptations Finally, we place adaptation into a broader context of improving player experience and cast it as a machine learning problem In Section 1.3, we describe various projects where we have studied aspects related to automating adap-tive game design, with respect to the methodology given in Section 1.2 These projects cover different genres of games with which we have experimented, includ-ing 3D action games, platformers, arcade games, board games, puzzles, and open-world games In the final section of the chapter, we take an overview of these projects and draw conclusions about the prospects for personalized gaming through adaptive game design

gen-Note that it is beyond the scope of this chapter to cover all the work done in the area of adaptive content generation, and we only present background material which is direcly relevant to the projects we describe Each of those projects is covered by various of our research papers which we cite in the chapter and which can be referenced for further literature reviews

1.2  METHODOLOGY

Given the need for the adaptive generation of bespoke game content, this section

describes how this can be achieved for digital games We focus on the processes that

we have used for projects ourselves but which have broader application to other domains In each case, the process involves three fundamental steps, summarized below:

1 Generation of new content and rule sets

2 Measurement of the generated content and target players during adaptive

generation or as part of system design and evaluation

3 Adaptation with the aim of changing a target player’s gaming experience

These steps are summarized diagrammatically in Figure 1.1, where the arrows cate the order of operation In the following sections, we consider each of these steps

indi-in detail

1.2.1  Generation

The first step in the cyclic adaptive process is the generation of novel game content and game rules This may be achieved through fully automated means, although a

20 Chapter 1 Toward the Adaptive Generation of Bespoke Game Content

significant amount of our research also investigates the use of the computer as a

creative collaborator that assists the designer by taking on some creative bilities In this section, we describe the generational methods most commonly used

responsi-in our work

1.2.1.1  Procedural Content Generation

The exponential growth of digital games in recent years means that there are now hundreds of millions of people playing games every day, wanting new and interesting content [29] However, the related production costs and requirements for specialized manual labor to develop content to satisfy this demand have also increased expo-nentially, and the industry is now facing serious scalability issues Games are becom-ing larger and more complex, with virtual worlds that are open, massive, and ongoing, which puts impossible demands on designers and artists alike and creates

a content creation bottleneck.

Procedural content generation (PCG)—the automatic creation of content through algorithmic means—offers a potential solution to this shortfall between consumer need and industry output and is becoming an increasingly important field

of research for digital game design for both the artistic content of games and for

game play itself Content in the context of digital games may refer to any of the

following:

• Rules that govern the gameplay

• Challenges that define initial states posed to players

• Resources that define the game’s look, theme, feel, and so on

PCG is a difficult task for creative domains such as game design, as the matically generated content must satisfy the constraints of the designers and artists

auto-as well auto-as the (often poorly defined) needs of the end users However, it offers the promise of handing at least some of the creative responsibility to the computer, and

we are now seeing an increasing amount of procedurally generated content in mercially released games

com-Search-based procedural content generation (SB-PCG) is a particular type of PCG in which a test function grades the generated content for fitness and guides the

Figure 1.1 Overview of adaptive game generation process.

Measure

Generate

AdaptContent

Player

Personality Behavior Experience

Free ebooks ==> www.Ebook777.com

Methodology 21

search for new content accordingly As depicted in Figure 1.2, Togelius et al [58] distinguish SB-PCG from other forms of PCG as follows:

1 Search Based Content is iteratively generated according to a fitness

func-tion that guides the search

2 Constructive Content is directly generated according to certain rules with

strict validation

3 Generate and Test (G&T) Content is iteratively generated according to

certain rules and filtered for fitness

SB-PCG is an ideal mechanism for adapting games and game content on-the-fly, in response to players’ needs, as the system can learn and improve its output the more

it is used See [29] and [58] for further details on PCG and SB-PCG for games The two main SB-PCG mechanisms we have used in our projects are evolution and coevolution, as described below

1.2.1.2  Evolution

In traditional evolutionary systems, a population of possible solutions to a particular problem are evaluated for “fitness” (some numerical value indicating how well they solve the problem) and recombined to produce hybrid solutions that hopefully inherit positive traits from the previous population For a generative task such as those in procedural content generation, the task at hand is to produce a piece of content to meet certain quality or player-specific targets, and a solution is a piece of finished content that can be evaluated against those targets The process of iterative

Figure 1.2 Main forms of PCG (from [58]).

Search based

Constructive

Simple G&T

Initial Rules Construct

Rules Construct Done? Y

N

Y N

Selection

Variation

Result

www.Ebook777.com

22 Chapter 1 Toward the Adaptive Generation of Bespoke Game Content

evaluation and recombination is repeated until some stopping condition is met, which may involve measurements of the content produced

Evolutionary algorithms are used in a wide variety of applications, including across the games industry Evolution is particularly useful where (a) only general

criteria for a solution can be stated—such as Paul Tozour’s City Conquest (Intelligence

Engine Design Systems, forthcoming), which used computational evolution to stress test the game for balance issues—or (b) the space of possible solutions is so large

that searching it using other methods is too difficult—for instance, the Starcraft II

(Blizzard, 2010) community was upset by a genetic algorithm that could optimize complex build orders and discovered exploits unknown to even the best human players

Evolutionary algorithms tend to perform best when the fitness functions and the representation of a solution are relatively simple For larger problems, where solu-tions may be very complex and fitness evaluations include many competing estima-tions of quality, evolutionary algorithms are harder to design optimally and take longer to produce good solutions They also lack a guarantee of robustness: Due to the random nature of the generation and recombination processes, even the best designed evolutionary systems may produce bad or severely suboptimal solutions This issue is one reason that evolutionary algorithms are more commonly used in preproduction to generate content that can be curated before inclusion This problem can often be mitigated by building additional systems to perform quality checks or adjust evolutionary parameters, and many applications of content generation come with the expectation that the system may occasionally produce curios or eccentric output

1.2.1.3  Coevolution

Cooperative coevolution (CCE) is a type of evolutionary algorithm that helps solve larger problems by decomposing them into smaller tasks that can be solved individu-ally In their paper proposing the algorithm [47], Potter and De Jong say that in order

to evolve more complex structures, explicit notions of modularity need to be duced in order to provide reasonable opportunities for complex solutions to evolve These modules are called “species” and are structured as self-contained evolutionary systems, with a population and a fitness function of their own

intro-The difference between a species and an ordinary evolutionary system is that a fitness function evaluates a member of its population in the context of the original

design problem That is, if we have a problem P decomposed into n evolutionary

combine them with s to make a solution to the original problem P The fitness tion then evaluates s both on the quality of it as an individual solution and the quality

func-of its cooperation with the other n − 1 subproblems.

Cooperative coevolution offers many benefits when building content creation tools Each module can easily encapsulate a particular design task, such as level design, which helps conceptually separate the different elements of content creation

Methodology 23

It is also easily amenable to mixed-initiative design, where a human contributes to the content generation process alongside an AI system Because fitness functions react to the context provided to them by the other species, we can remove a CCE species and replace it with a static, human-generated piece of content, and the CCE system will design and adapt its other species to the content provided For example, consider a puzzle game designer that conceives of a rule set in one CCE species and designs a set of levels with another CCE species In normal execution, the evolution

of the rule set will interact with the evolution of the level design, and over time the two will cooperate and complement each other However, we might want to develop

a particular kind of puzzle game If we replace the rule set species with a static rule set that represents the mechanics we want to use, the level-designing species will design levels tailored to the human-designed rule set This idea has enormous poten-tial for improvisational game design tools, where software and designer play off one another’s ideas

Video game design represents a particularly complex problem, being comprised

of many different components (such as levels, mechanics, artwork, narratives, and music) all of which have different estimations for their fitness and depend on each other for their definitions of quality; a “good” level for a set of mechanics like those

in Pac-Man (Namco, 1980) is very different from a good level for a game such as

account a vast array of quality estimations that would change while the evolution was still taking place, but CCE allows us to subdivide and specialize these design tasks to better deal with each individually For a description of a system employing CCE for the procedural generation of content, see Section 1.3.3

1.2.2  Measurement

The second step in the adaptive process is the measurement of the generated content This involves:

1 Measuring the quality of the generated artifacts (game content and rules)

according to specified criteria or desired aims This can be achieved, for instance, through automated self-play

2 Measuring how the player plays the game, taking external observations

during play, and measuring other contextual factors such as personality or stored profiles

Specific applications have often only used one approach, measuring the content

Measure-ment can play three distinct roles in adaptive game generation systems as follows:

Adaptive Measurement The system measures aspects of the generated content and target player to deliver content adapted to that player

Formative Measurement The system designers test the quality of generated content and player’s reaction to content and/or adaptation in order to inform

24 Chapter 1 Toward the Adaptive Generation of Bespoke Game Content

the design of the system For example, this can include gathering player feedback to train a learning algorithm

Summative Measurement The working system is evaluated in terms of the quality of generated content and a player’s reaction to content and/or adaptation

Adaptive measurement is autonomous, carried out by the bespoke game design system, whereas formative and summative measurements are human guided In

human-guided measurement, we are often able to exploit information not accessible

to autonomous measurement by the game during normal adaptive play, for instance, verbal feedback or physiological measurements However, it is possible for such data

to be measured autonomously during play by a sufficiently sophisticated system In the sections below, we consider some different approaches to measuring content and players

1.2.2.1  Measuring the Game

We understand the quality of a game to mean the potential for the game to engage

players: the capacity of that game to interest players and to keep them in that state Gauging the quality of a generated game or piece of game content can be difficult,

as the notion of quality can depend on the context and vary from player to player

One approach is to define quality metrics which provide a computational

assess-ment of an aspect of game quality Such metrics can be used to automatically guide the search during SB-PCG but can also be useful during system design and evalua-

tion An alternative is to evaluate game quality by play testing, where explicit

feed-back is gathered from players This is typically used as formative or summative measurement but could also form part of the adaptive process in systems which directly solicit players for feedback to guide adaptation

Most research into game quality metrics has been done in the context of board games, where games of any significant depth tend to involve mechanisms and strate-gies that emerge during play and which may not be obvious from their rules alone For this reason, it is generally more reliable to measure board games for quality via the playing of games, rather than from the rules alone This can be achieved by conducting series of self-play trials between artificial players Many of the metrics can, in principle, be generalized to evaluate the quality of video games AI players,

called bots, can be used to automatically test generated video game content and

gather metric data, although for more complex games, creating an AI bot may be a very time-consuming task

Browne [9] describes 57 aesthetic criteria for empirically measuring the quality

of board games, mostly from trends observed during self-play trials These include interpretations of the following four key features of abstract games, outlined by Thompson [57]:

Depth The capacity for a game to be played at different levels of skill and to reward continued study

Clarity The ease with which players can understand the rules and plan moves

Other useful metrics include uncertainty [34], balance [31], and game length

[1] Game length has proven to be a particularly effective indicator of flawed games,

as it quickly detects trivial games that end within a few moves, as well as cally flawed games in which players can defend indefinitely with optimal play, and logically flawed games in which the goals simply cannot be reached using the speci-fied rules Player testing with people is preferable if an appropriate quality metric

strategi-is difficult or impractical to implement In such scenarios, player testing can be more reliable, as the player is the end user that the game design process is ultimately trying

to satisfy Further details of measuring player experience are described in Section 1.2.2.4

1.2.2.2  Autonomous Player Measurement

To deliver bespoke game content, generation methods need to be based on data about

a particular target player or players This player data can be collected before play,

to generate new content for the next game, or during play itself, in order to adapt upcoming content in the current game Data on multiple players can be collected in order to generate common content for that group, either because they are playing a multiplayer game or because they are being collectively targeted with the same content, for example, as members of the same age group

The player data most easily gathered by digital games are game play logs, which

contain a record of in-game states and events, and player actions, from which summary player features may be computed, for instance, as in [7, 28, 52] However, other forms of player data can be gathered, such as demographic data, motion, posture, physiological signals [41], visual appearance [3], retail activity, social media activity [49], and direct player feedback on experience These data sources are not always available, but as mainstream gaming hardware develops (e.g., in the motion-aware Wii and Kinect consoles) and social media and gaming become more inte-grated, there is a growing commercial interest in exploiting these resources, as in [2] Note that even when the available data are restricted, a content generation algorithm can be still be informed by other kinds of formative player measurement during design or training, for example, feedback on player experience during testing [53, 63].Player input data are often reduced to a set of categorical and/or scalar features This provides a simplified input to the subsequent content generation stage and allows the use of standard machine learning techniques Given the range of possible player inputs, this feature data can measure any aspect of the player, his or her activ-ity, and the context of play As examples, we can take measurements of current situ-

ational intensity—as in Left 4 Dead (Valve, 2008) [7]—weapon use [28], summary

statistics for a individual combat [25], a single level [52], or an entire video game [23] An alternative to scalar and categorical features is to use structured player data, such as paths, sequences, trees, or graphs To date, structured player data have been

Free ebooks ==> www.Ebook777.com

26 Chapter 1 Toward the Adaptive Generation of Bespoke Game Content

relatively unexplored, although there has been a growing interest in the research literature [22, 27, 49]

One consideration when measuring in-game player behavior is whether the game play is uniform or divided into several distinct modes of play Much work on adaptive content generation has looked at games where the player is engaged in roughly the same continuous activity, for example, simple platform games For such uniform game play, player features can be given a consistent interpretation In other games, game play is structured as a series of distinct and possibly overlapping activi-ties Different player features, such as the rate of weapon fire, can have very different

interpretations between activities For example, Pac-Man (Namco, 1980) involves a

ghost avoidance and a ghost-hunting phase, and a “distance-to-ghosts” feature has

a different meaning in each mode Comparing such features across players may not give us a clear picture of individual differences unless the activity context is taken into account One approach is to segment game play logs into distinct activity types and measure these separately [25]

1.2.2.3  Player Models

Player input data can be passed directly to a content generation system or instead

be first converted to a more abstract player model, that is, a representation of the

player designed to be more appropriate for subsequent content generation In general, any representation (e.g., first-order logic) that raw player data are converted to in a preprocessing step could be considered a player model Typically, this will be a feature-based model, which describes players in terms of a small number of features representing significant characteristics

A feature-based player model consists of scalar traits and categorical types,

following the terminology of personality psychology A type-only model is known

as a player typology [4, 5] Trait-based models are regarded as a more accurate

representation of individual differences than discrete typologies, although in some cases a typology may be more convenient to work with for a game designer or content generation system

Providing that they capture the relevant aspects of the player data with respect

to the game, player models can provide a simpler and more convenient representation

of the player, considerably reducing the dimensionality of the input data for quent adaptive content generation Ideally, translating player input to a model will highlight relevant variations between players and filter out irrelevant data If machine learning is used to train the content generation system, working with low-dimen-sional data can increase learning performance Another advantage of player models

subse-is that they provide a simple representation of the player that can, in some cases, be transferred between gaming contexts, presented to designers and players, or reasoned about by AI agents

To employ a player model, one must first be created or selected from a set of existing models Second, a mapping from the player input data to the model must

be defined Finally, the use of the model will need to be evaluated in the current gaming context Using the wrong player model may lead to useful information about

www.Ebook777.com