COMMUNICATIONS OF THE ACM January 2005/Vol. 48, No. 1 61

Story Creation inVIRTUAL GAME WORLDS

The fact that children spend considerable amounts of their time playing computer games is a

phenomenon that has not goneunnoticed by educators [5]. In addition to the obvious

motivational effects of games on young people,research has begun to look at the actual learningthat can occur in game-playing settings, for example, logical reasoning and teamwork skills [9].

Allowing young people to create computer games they will ultimately want to play not onlyoffers key educational benefits but builds self-esteem and teamwork skills.

By JUDY ROBERTSON and JUDITH GOOD

62 January 2005/Vol. 48, No. 1 COMMUNICATIONS OF THE ACM

Much of the research examining games in educa-tional settings looks at children playing gamesdesigned by professional software developers, with lit-tle focus on children creating their own games. Somework investigating this issue (for example [7, 8]),looks at how the process of game creation fosters chil-dren’s programming skills as, until very recently, it wasimpossible to create a game without some knowledgeof programming. In the case of 3D virtual environ-ment role-playing games, it has required bothadvanced programming and mathematical skills.

However, technology has reached the point wherethese skills are no longer always necessary: more com-mercial games are shipping with game-editing tools toencourage skilled members of their user communitiesto develop further (free) content.

A good example of this phenomenon is Bioware’srole-playing game, Neverwinter Nights, released in2002 with the Neverwinter Nights toolset that can beused by nonexperts to create characters, settings, andinteractive plots. No knowledge of 3D modeling isrequired, and basic tasks can be accomplished with-out any programming skills.

Given these developments, we can now begin toexplore how children use game-creation toolkits tobuild sophisticated 3D-interactive games that theirpeers can play. This new ability to produce a finishedproduct of commercial quality is important, as chil-dren have high expectations for games in terms ofquality of graphics, among other features. If we are touse games in education, and expect them to motivateand interest children, they must be of similarly highquality, otherwise their effect will be lost [4].

Additionally, with the decreasing focus on pro-gramming skills, at least for designing games within agame-creation environment, computer games can beintegrated more tightly with specific domains withinthe curriculum, for example, literacy and narrativedevelopment, the subject of this article.

Literacy Development inVirtual EnvironmentsLiteracy development, particularly the improvementof writing skills, is a high priority for many educa-tors, as a number of children lack motivation forwriting, and experience writing apprehension [10].This is unfortunate, given the central importance ofnarrative as a mode of thought [3]. Although manychildren have a wealth of creative story ideas, writtenlanguage may act as a barrier to expressing these ideas[1, 10]. If children could develop their creative ideasin a nontextual medium, they would be able to sharethem with others more readily, which could have pos-itive effects on self-esteem.

We believe the toolsets included with 3D-interac-tive role-playing games offer a unique opportunity fordoing so. Because creating a high-quality game usingthese toolsets no longer requires programming, inter-active 3D audiovisual narrative can be createdthrough user-friendly interfaces and minimal script-ing. This allows children to focus on literary conceptssuch as creating gripping plots, believable characters,and compelling settings.

Additionally, these toolsets allow children to exper-iment with a type of fiction with which they mayhave little experience, namely interactive fiction,which offers the reader a choice of paths through astory. This type of fiction has been realized in variousformats, such as early text-based computer adventuregames, and research has shown that having childrencreate these types of games has a positive effect ontheir writing skills [2, 11]. More recently, 3D virtualrole-playing games embody interactive narrative.Players choose their own path through the virtualworld by deciding how to respond to the situationsthey encounter.

This article considers the benefits to be gainedfrom enabling children to create stories in themedium of interactive 3D virtual reality computergames. We describe a study in which 10 teenagers cre-ated their own stories using the NeverWinter Nightstoolset, describing the games creation process, themain findings, and their implications.

The Game Maker WorkshopThe Game Maker workshop was a week-long sum-mer workshop open to all secondary school pupils(ages 12–15). Its aim was to:

• Give young people an opportunity to tell theirown stories through computer games;

• Explore computer games creation as a method fordeveloping literacy skills; and

• Discover the strengths and weaknesses of theNeverwinter Nights toolset with a view to developing a custom made educational games authoring tool.

The workshop participants included nine boys andone girl, and were self-selecting, since participation inthe workshop was voluntary. The workshop leaderswere the first author, a visual artist, a professionalstoryteller, and an experienced amateur gamesdesigner with considerable experience using the Nev-erwinter Nights toolset.

The workshop was organized around a series ofactivities, described here (with approximate timesindicated in parentheses).

Group discussion (30 minutes). The discussionfocused on participants’ preferred games and gamecharacters, and was followed by a general introductionto the workshop. The discussion was lively, andworked as an effective icebreaker.

Initial trial of Neverwinter Nights game (30minutes). Participants played the tutorial level ofNeverwinter Nights to become familiar with thegame. This led to a critique of the tutorial and a dis-cussion of principles for good game design.

Character design (30 minutes). The young peo-ple drew pictures of their characters’ physical appear-ance, and wrote notes describing their backgrounds,motives, personalities, and missions.

Character model making (4 hours). Once char-acters were designed, the artist explained how to make3D character models by layering plasticine onto wireskeletons prepared by the artist based on creaturesavailable in the Neverwinter Nights toolset. The artistthen facilitated the creation of three sets for story-boarding: a river, castle, and a graveyard.

Plot planning (40 minutes). On day two, half ofthe participants planned stories in the morning andcreated games in the afternoon, while the other halfdid the opposite. Each participant wrote down thefirst plot episode in a game, that included the charac-

ter they had created. Participantshelped each other brainstorm byadding plot episodes to each oth-er’s plots.

Storyboarding with digitalcameras (20–60 minutes). Character models wereposed in the stage sets to represent each plot episode.Digital photos were then taken of each scene andimported into Microsoft Powerpoint where they werearranged into a storyboard with captions explainingeach scene.

Game authoring using the Neverwinter Nightstoolset (10–13 hours). The participants explored theNeverwinter Nights toolset at their own pace, withthe games expert available to answer questions. Par-ticipants tended to first explore the facilities for creat-ing areas and characters before moving on to moreadvanced tools, such as joining two areas using portalsor doors; making interactive conversations; makingmerchants sell items to a player; and using the plotwizard to make quest storylines.

The figure shows the Neverwinter Nights toolset.The area in which the story will take place is repre-sented as a view from above, with a superimposedgrid. The game designer can choose from a range ofobjects on the right-hand side and place them in the

COMMUNICATIONS OF THE ACM January 2005/Vol. 48, No. 1 63

Inserting a campfireon an area using the

Neverwinter Nightstoolset.

area. In this case, a campfire has been placed (outlinedin light green).

Reflecting and planning (20 minutes). On daythree of the workshop, participants reported on theirprogress. Half of the participants estimated they spentthe most time creating areas, while the other halffocused on a range of other aspects.

Participants then planned out their remainingtime. All expected to spend the highest percentage oftheir total time creating areas, although some consid-ered creating conversations and characters almost asimportant. In all cases, the young people designed ontheir own, rather than in groups, although they oftenconsulted each other.

Interview AnalysisOn the last day of the workshop, the participantswere interviewed individually. These interviewslasted approximately 15 minutes and were video-taped, and later transcribed. The interviewer asked aseries of questions relating to the participants’ com-puting background, their opinions of the workshopactivities, and their experiences with the Neverwin-ter Nights toolset. We focus here on the latter twoissues, as this data is useful both for developingfuture workshop activities, and for evaluating thefeatures of the Neverwinter Nights toolset.

Workshop components: Plasticine model-mak-ing. Reactions to building wire frame and plasticinecharacter models were primarily positive. Some partic-ipants found the experience useful for visualizing theircharacters, although others were forced to abandontheir character models because physically similar char-acters could not be found in Neverwinter Nights. Oneparticipant noted that although the physical appear-ance of his model did not make it into the game, thecharacter’s mind and personality did. This suggeststhe inclusion of pregame design activities that focus onthe personality and motivations of a character.

Workshop components: Game design. Whenasked which aspect of game design they enjoyed themost, “creating the characters” was most popular,with “creating areas” being second favorite. Thisencouraging find suggests that participants enjoyedthose aspects of game design similar to aspects in thedesign of plays and other types of drama.

The plot wizard (which guides users through thecreation of the story by having them specify dialogue

between the character and player) was the leastfavorite game-design feature. Although not difficultper se, most participants felt it was tedious to plan outthe conversations, and click through all of the wizard’svarious screens.

Use of the toolset. Feelings about the toolset werepositive overall. All of the participants found thetoolset easy to use once they became familiar with it.

The most difficult part of the toolset was the por-tals, which can be placed within an area, and allowseparate areas to be linked together. The difficultywith defining portals appears to stem from the needto have a unique identifier for each portal in order tospecify which location should be reachable fromwhich portal. Furthermore, this task requires script-ing, and many participants needed help with this.They suggested the toolset could be improved by hav-ing it automatically generate unique identifiers, or byhaving a drag-and-drop interface for portal creationrather than a wizard.

One of the biggest issues targeted for improvementconcerned the characters. Specially, young peoplewant the ability to do full character design, for exam-ple, to include oneself in the game; and an increasedability to modify features of existing characters such asfacial features and clothing.

Other suggestions for improving the toolsetincluded:

• Make the inclusion of sound recordings possible;for example, include key phrases for characters;

• Increase choice of objects;• Provide more options for present day scenarios

(as opposed to “medieval” ones);• Include the toolset as part of the game itself (cur-

rently, it is necessary to quit out of the toolsetand launch Neverwinter Nights in order to viewone’s game); and

• Allow the use of icons within the plot structurerepresentation to represent things like quests orconversations.

DiscussionThe Game Maker workshop was the first steptoward understanding the educational benefits to begained from enabling young people to create storiesin the computer game medium.

The most important benefit was, as anticipated,

64 January 2005/Vol. 48, No. 1 COMMUNICATIONS OF THE ACM

The Game Maker workshop was THE FIRST STEP TOWARD UNDERSTANDING

THE EDUCATIONAL BENEFITS to be gained from enabling young people to create

stories in the computer game medium.

the strong motivational effect this workshop had onthe young people who took part. They all reportedthey enjoyed the experience and would continue touse the Neverwinter Nights toolset if they had accessto it. They became engrossed in the games design taskand it was very difficult to persuade them to stopworking and take breaks.

A related benefit of the workshop was bolsteringthe self-esteem of the participants, who were eager toshow off their games to other workshop participants,and to friends and family members at the EdinburghInternational Games festival, which took place oneweek after the workshop. One parent commentedthat her son didn’t like writing stories at school, yet hespoke with great pleasure about the game he createdin Neverwinter Nights. She encouraged him todevelop his interest in computer game design, andperhaps even study it in further education.

The Game Maker study also answered some openquestions about the feasibility of computer gameauthoring for young people. Although the Neverwin-ter Nights toolset simplifies the game design process,there are some difficult concepts and complicatedprocedures involved in creating interactive narrative.These include planning an interactive plot; under-standing how interactive dialogue is represented in theinterface; and ensuring that objects such as portalshave unique tags. In spite of the complexity of thegames-creation task, all of the young people managedto create a game to their satisfaction. The games var-ied in sophistication, but this would be expectedwhen the authors have varied computing back-grounds and are of different ages.

Although the young people were able to use the Nev-erwinter Nights toolset to express their story ideas ingame form, some toolset features were difficult to use,such as the plot wizard. In addition, the software has nofacility for visually representing multiple plot threads.

Finally, Neverwinter Nights has a number of fea-tures that make sense in the context in which theywere designed, but may not be desirable in terms ofgame design for educational purposes. For example,characters and settings are confined to the Dungeonsand Dragons genre, and have a pseudo-medievalappearance. Allowing for a wider range of settings andcharacters would similarly expand the types of storieswhich young people could create.

Additionally, the characters possess inherent fea-tures, such as skill and ability in different types ofcombat. Ideally, a game author should be able to cre-ate a character as a blank slate, and add personalityfeatures that go beyond the typical combat-orientedscenario, for example, compassion or trustworthiness.

We are currently working on a purpose-built tool

(AdventureAuthor) for creating stories in the gamesmedium, designed to directly support story construc-tion. The AdventureAuthor prototype addresses thelimitations of Neverwinter Nights by providing avisual representation to support plot planning andinteractive dialogue, as well as guides the user throughthe game creation process [6]. It will also allow for abroader range of settings and characters with morefully developed personalities.

More research is required to identify how thegame-making process could be integrated with thecurriculum in a classroom context. In addition to thegeneral storytelling skills common to all storytellingmedia, such as developing coherent plots, motivatedcharacters and convincing dialogue, game authors arelikely to develop medium-specific storytelling skills.Learning how to express the emotional content of sto-ries using sounds, music, and lighting effects is a valu-able part of becoming literate in the medium ofcomputer games.

References1. Ananny, M. Supporting children’s collaborative authoring: Practicing

written literacy while composing oral texts. In Proceedings of the Computer-Supported Collaborative Learning Conference. (Boulder, CO, Jan. 2002).

2. Bruckman, A. Moose Crossing: Construction, Community, and Learn-ing in a Networked Virtual World for Kids. Ph.D. Dissertation, 1997.MIT Media Lab, Cambridge, MA.

3. Bruner, J. Actual Minds, Possible Worlds. Harvard University Press,Cambridge, MA, 1986.

4. Elliott, J., Adams, L., Bruckman, A. No magic bullet: 3D video gamesin education. In Proceedings of International Conference of the LearningSciences. (Seattle, WA, Oct. 23–26, 2002).

5. Gee, J.P. What Video Games Have to Teach us about Learning and Lit-eracy. Palgrave MacMillan, New York, NY, 2003.

6. Good, J. and Robertson, J. Children’s contributions to new technology:The design of AdventureAuthor. In Proceedings of the Interaction Designand Children 2003 Conference. ACM Press, New York, NY, 153.

7. Kafai, Y.B. Learning design by making games: Children’s developmentof design strategies in the creation of a complex computational artifact.Constructionism in Practice: Designing, Thinking and Learning in a Dig-ital World. Y.B. Kafai and and M. Resnick, Eds. Lawrence Erlbaum,Mahwah, NJ, 1995, 71–96.

8. Kafai, Y.B. Electronic play worlds: Gender differences in children’sconstruction of video games. Constructionism in Practice: Designing,Thinking and Learning in a Digital World. In Y.B. Kafai and and M.Resnick, Eds. Lawrence Erlbaum, Mahwah, NJ, 1995.

9. McFarlane, A., Sparrowhawk, A. and Heald Y. Report on the educa-tional use of computer games. Teachers Evaluating Educational Multi-media Report, 2002; www.teem.org.uk/resources/teem_gamesined_full.pdf 27/11/03.

10. Mulholland, C. and Robertson, J. (2001). Storymakers Pilot ProjectReport—Pictures of the Present, Visions for the Future. Scottish Story-telling Centre, Edinburgh, 2001; homepages.inf.ed.ac.uk/judyr/publi-cations/storymakers.pdf.

11. Sharples, M. Cognition, Computers and Creative Writing. Ellis Hor-wood, Chichester, U.K., 1985.

Judy Robertson (judyr@inf.ed.ac.uk) is a lecturer and researchfellow in the Department of Informatics, University of Edinburgh,Scotland.Judith Good (judithg@unm.edu) is an assistant professor at theUniversity of New Mexico’s College of Education, Albuquerque, NM.

© 2005 ACM 0001-0782/05/0100 $5.00

c

COMMUNICATIONS OF THE ACM January 2005/Vol. 48, No. 1 65