Issue
Article
Vol.29 No.4, October 1997
Article
Issue
Issue |
Article |
Vol.29 No.4, October 1997 |
Article |
Issue |
Every year, hundreds of thousands of people spend money from their own discretionary entertainment budgets just for the opportunity to experience the user interfaces of software applications sold as games. Much of this experience involves difficult problem solving as well as tactical and strategic reasoning and performance, that often exceeds the demands of typical tasks at work. Yet, games are fun and work often is not. Why?
Carroll and Thomas (1988) suggest that the CHI community only has a partial understanding of how to analyze software usability, with the notions of "fun" and "user-friendly" being poorly understood. "Fun" and "easy to use" are often confused and conflated. Being easy to use does not necessarily imply being fun or vice versa, although it may; the authors suggest fun requires sufficient complexity to be interesting while still tractable. Something that is fun is also intrinsically motivating. Malone (1981, 1984) discusses ways in which games promote intrinsic motivation to learn and to continue playing through challenge, fantasy, and appeal to the user's curiosity. Carroll (1982) similarly analyzes feedback in the adult game Adventure, suggesting that users are intrinsically motivated to play by features like spatiality, increasing challenge, rich feedback for user actions. Fun is still difficult to study, however, while ease of use may be measured by performance time, errors, and time to learn a task. Along with those difficulties, Carroll and Thomas suggest that researchers studying fun may not be taken seriously as researchers, and point out that in fact the development world (and marketing world) have paid far more attention to the concept than serious computer scientists or usability experts.
Games are fun because game designers make them fun. In the game community, game designers spend a year or more designing, tuning, and balancing the game play of one of these titles. They playtest relentlessly during this year because games are iteratively evolved from an initial design. As in most other fields, game designers have their own meetings, conferences, lore, and craft. Little of this field's expertise has seeped into the HCI design community, yet some have recognized its pertinence (e.g., Laurel 1990, 1991; Malone 1982; Carroll 1982).
This workshop was a step towards opening up a dialogue among CHI researchers and between CHI researchers and the game community about the issues they have in common. We sought diversity of experience and perspective. Some of the topics we were interested in discussing in advance included:
The participants were a mix of UI professionals, artists, students researching interactive narrative, and usability experts interested in online entertainment applications. Unfortunately, professionals from the game development community were not sufficiently attracted to the conference itself, and the requirement that they register for the main conference as well was a significant deterrent for the game designers we invited to participate.
William Cockayne is from Pumpkin and associated with the Naval Postgraduate School. Bill researches and designs virtual environments. He is interested in interaction methods (currently two-handed interaction methods) and the causes and effects of immersion. Virtual environments may be used for simulation and training, as well as entertainment. Bill believes games are good sources of insight into how immersion is achieved and interaction methods are communicated quickly to the user.
Thom Gillespie teaches at Indiana University. Thom teaches digital storytelling and computer game design, and is involved in a startup which is trying to design and market digital stories and computer games. Thom also taught a tutorial at CHI 97 on digital storytelling.
Lars Erik Holmquist does research on interactive narratives and user interfaces at the Viktoria Institute in Gothenburg, Sweden. He is amazed at how computer games manage to be at the same time complex, stressing, nerve-wracking and FUN, while most "serious" applications always seem to get to the first three points of that list, but completely miss the fourth. John Meech and Ella Tallyn are at the University of the West of England. John has a background in human factors research and co-founded a company to develop virtual reality technology to provide immersive training and entertainment systems (he still wonders why this turns out to be so hard). Ella Tallyn is conducting doctoral studies into the use of narrative in virtual reality.
Eric Jones, Director of Restaurant Systems for Pepsico, has led development efforts in designing business systems that take insights from games.
Maribeth Back, of Xerox PARC, is a sound designer and audio engineer. She argues that sound in a game model can help with: the internal construction of a constrained universe with specific laws (sense of place), user expectations guided by context (what happens next), virtually instant multimodal feedback to user's actions. She's interested in ways in which sound can give clues to users about the genre of a computer application, which she suggests might be a game, toy, instrument, or tool.
Matt Belge, of Vision and Logic, has been a UI designer for 10 years and is an artist who creates "light installations" -- environments of light and shadow that viewers are invited to walk around and immerse themselves in. He's interested in the emotional responses users have to applications, and finding arguments to get support for incorporating game principles and concern about emotional responses into interface design.
Kevin Keeker, of Microsoft, has been researching "appeal" in the context of usability testing Microsoft Network Shows, which are web-based programs designed to be entertaining.
Jeff Johnson has done evaluations, user-testing, focus-groups, and design at a variety of companies including FirstPerson, Xerox, HP. Jeff is interested in cross-fertilization between the practices of game design and good design of productivity software. For example, how can productivity applications be made more satisfying and captivating, and how can games be made less nerd-oriented and arbitrary?
The organizers were Chuck Clanton, Erik Ostrom, and Lynn Cherny. Chuck has been a UI designer and application architect for almost 20 years. With Emilie Young he gave a tutorial at SIGCHI and SIGGRAPH for years called "Film Craft in UI Design" in which they tried to raise the awareness of the UI community about the vast body of relevant knowledge and literature in the film community. He has worked in the computer game industry for the past several years.
Lynn is currently at AT&T Labs--Research working on UI and interaction design for chat services and studying attributes of computer-mediated conversation and community. She's interested in gender and games, as well as interactive narrative techniques. Erik is a graduate student at Northeastern University who is working in an HCI Research department at AT&T Labs--Research. He has been working on and with MUDs, text-based multi-user environments, for the past 6 years.
Before the workshop, we had a lively email discussion in which we introduced ourselves and our topics of interest. Some of this discussion helped the organizers in selecting final topics for the workshop, and in choosing games to demonstrate and discuss.
We discussed the difference, if there is one, between UI, game mechanics, and "game play". Chuck Clanton suggested the following terms for the exchange:
game play = strategy and tactics for accomplishing the goals in the game
game mechanics = physics of the world which set the constraints on how we conceive of our ability to navigate and manipulate the world
user interface = input and output mechanisms that connect us to the world
Tetris, as an example, has a game play involving filling in all of the spaces with falling bricks. The game mechanics are that the user can rotate the falling objects until they hit something when she tries to turn them, and the objects stop when they touch anything. The UI is the keys the user uses to operate the mechanics of the world to accomplish the goal.
Chuck suggested that game mechanics and UI are both designed to support the game play, and when they don't, this results in a problem with the game. One participant suggested that these were not distinguishable components, that the game is the interface, and vice versa. If a game is hard to play, the game is bad, qua game. Chuck suggested that the reason that they are so unified in many people's minds may be just that they should feel that way if the game is well designed.
A slightly less severe distinction between game mechanics and UI was proposed as well. UI designers organize the concepts of a user interface so that they make sense, which may involve mechanics. What are the objects that a program manipulates, and what are the actions on those objects? Jeff Johnson suggested a well-designed UI breaks down into the conceptual level (e.g., deposit funds, reconcile account with bank statement, withdraw funds), the functional level (e.g., select a transaction, delete it), and the physical level (e.g., keystrokes for input, images on screen for output).
For the sake of discussion, Chuck proposed a distinction between stories and histories that he learned from adventure game designer Hal Barwood: a story has an overall arc with structure, while a history (e.g., a soap opera) uses all of the storytelling devices, but there is no overarching story structure. Action games (like Quake) give you the opportunity to create a history, not a story. The difficulty of creating good interactive stories rather than histories was discussed, with Chuck proposing Gabriel Knight 2: The Beast Within, as an example of a game with a good story. Chuck also suggested there are other models for story telling such as the children's game Orly Draw-a-Story, where the interactivity relates to the narrative in a different way than in an adventure game. We demoed and discussed Orly in the workshop.
Using other media models for story telling may not be appropriate for interactive media. Ella Tallyn suggested that if we want to create interactive experiences that have story structures we have to look at existing interactive media, that tend to be more about "showing," like performance arts that include audience participation, and also perhaps improvisational art like oral story telling and commedia del arte.
The workshop was a two-day workshop with a very large number of demos and presentations on PC (displayed on the overhead viewer), television with Sony Playstation and Nintendo 64, video, and slide projector. Jeff Johnson and Chuck Clanton opened with discussion of a video-on-demand movie-finder application they and their colleagues had designed as part of a Sun venture (Green), using an animated world with an adventure game flavor. They found an age difference in the appeal of the cartoon world; potential users younger than 30 found it attractive but older people tended to consider it tedious. Jeff described an experiment he conducted to determine whether users should pan the touch-screen display by pushing the background or pushing the camera. Pushing the background won clearly (Johnson 1995).
Jeff continued with discussion and demonstration of a computer game project he had worked on recently, A Fork in the Tale; he helped design interaction methods for the fast-moving full motion video game. He designed the icons that offer users choice points during the game's speedy action, including a set of speech act descriptors for use during dialogue sequences. The game has not been very successful on the market; the problems with game play that hurt this game were described by Noah Falstein, a noted game designer, in a recent Computer Game Developers Association newsletter (Falstein 1997). Falstein notes that the UI was improved by the hiring of a UI designer, but that the design process should have incorporated better game play concepts from the beginning, rather than a post hoc fix of the UI. Although it's clear game designers know a lot about fun and learnability that would be valuable to the UI community at large, UI professionals certainly know a lot about usability that game design can profit from when the design process is appropriately holistic and interdisciplinary.
Eric Jones presented some of his design process for computer systems in restaurants which have a high employee turnover rate. The employees in his user population are usually young kids with relatively little education who are often video game literate, so games were a major inspiration for his design. The assumptions behind his design were assumptions shared by many video game designs: learning to use the system must be simple, users learn by playing or doing rather than in training sessions, and language isn't primary. Touch screens rather than mice are the interaction method in a restaurant context because the mouse gets dirty easily and requires training to use. Eric raised some questions about how international games are, and how much redesign is needed for systems in the other cultures (for example, cultures which do not attach meaning to colors).
Kevin Keeker presented research on the usability and appeal of Microsoft Network Shows (web-sites). He noted that MSN shows are produced via a model that is similar to that used in other media such as television shows or magazines. He theorized that there are five major ways to influence whether or not a web site gets used: provide relevant high-quality content, make it easy to use, promote effectively, make the experience unique to the medium and evoke emotion. He briefly gave examples of ways to evoke emotion (challenging the user, the use of plot, strong characters and pacing) and ways to design for the web medium (personalization, focus on user, community and refined content).
Lynn Cherny and Matt Belge facilitated a discussion of what they called "Computers as Co-creators", illustrated with demos of the games SimCity, The Incredible Toon Machine, and Creatures. These are games which present users with a series of constraints and responses to user actions that are complex and unpredictable, allowing the user to create often surprising, engaging situations in conjunction with the computer. These games are infinitely replayable and only loosely goal-directed, unlike many other games (e.g., games which the player can "win" or "lose" or must complete specified tasks). Such games have occasionally been classified as software "toys" rather than games, with the difference potentially being the presence of a strong organizing narrative: these "toys" may be said to have a narrative that is less structured and more open to user co-authoring. Matt suggested that the principles in these software toys may be most easily transferable to non-entertainment software applications, perhaps creativity tools like Photoshop or Word. Instead of a dumb authoring toolset, such applications might include an inspirational mechanism which employed "game-like" (or "toy-like") simulation theories to inspire users to more creative possibilities.
On day 2, Matt Belge presented some of his artwork and reflections on the use of space and light to create emotional response. He presented us with some useful excerpts from Gaston Bachelard's The Poetics of Space and Christopher Alexander's The Timeless Way of Building, which teach us how to read spaces and understand them emotionally. In many good games, this has been done implicitly by the designers. Myst, for example, provides a rich, emotionally complex space. It is not "happy" or "sad" but rather intriguing, mysterious, stimulating, inviting, and relaxing all at once. This is achieved through a variety of mechanisms that are described in these books. For example, the lighting is not uniform, but suggests fog, dappled shadows, deep forests, and open beach sunlight. Each of these has a different emotional impact. Similarly, Matt points out that the fact that Myst occurs on an island is significant. The phenomenology of an island is that one knows its limits, its borders. A person is protected from outside interferences. One need only master a very limited space, the island itself, to feel secure. In this way, the phenomenology of the island is somewhat like the phenomenology of a cocoon or the womb.
We looked at the children's game Orly Draw-A-Story, which prompts children to illustrate characters and objects which are then embedded in an animated story; like Myst, the game's action takes place on an island, a tropical island which is bright, cheerful, intensely colorful. Matt suggests that in this island cocoon, the real world is held at bay. This setting encourages children's fantasy and escape. It is a magic place, enhanced by the idea that the world can't get at you because you are protected by the water surrounding the island. This is exactly the kind of space one wants to set up for kids to give them permission to let their imaginations run wild.
We also examined the interesting lighting in the tomb spaces in Tomb Raider, a game that combines adventure game elements with action game elements as the main character explores a ruined tomb and caverns under the player's control. Matt suggested that the space the user/character explores is presented as emotionally ambiguous. There was a slight feeling of apprehension given off by the space, as well as one of intrigue and stimulation. This was accomplished by using bright but modulated lighting in an underground space. The lighting gave a feeling of comfort and intrigue by being both bright, but modulated into places of light and shadow. The fact that the space appears to be underground brought up another set of emotions. Typically, the underground is thought to be scary. Bachelard notes that the cellar is scarier than the attic, because it is associated with the unconscious and with death.
Maribeth Back discussed some of her audio performance art and the importance of sound design in games and UI generally. Although the graphics in Tomb Raider are evocative, Maribeth found the footsteps unconvincing: they sounded like a sample of a beanbag played over and over. She then played us three components of the sound of a breaking glass, showing that like other types of narrative, sounds tell stories with their constituent parts, and have a beginning, middle, and ending. The constituent parts of her breaking glass auditory narrative came from three completely different sources, and had been edited together to produce the final effect.
We continued some of the discussion from the mail list about narrative in games with short presentations by Chuck, Lars, and Ella. Some of the games demoed in this section include Orly, Silent Steel, King's Quest VI. Silent Steel is a full-motion video adventure game set on a submarine, which features several unfortunate sections in which the user gets no interaction options for long periods. Lack of sufficient interactivity is one criticism raised by many gamers against many full-motion video games (Falstein 1997).
Lars then presented his work on "the influence engine", a mechanism to provide loose scripts to direct action for multi-player, possibly networked, role-playing narratives. Some of the players in his scenarios may be computer agents rather than human actors. The scripts loosely describe character motivations and responses as sets of preconditions and actions, according to some constraints: too much script control takes away the interactivity and hence the fun, too little script control leaves actors unsure of what to do at a particularly important story juncture, and may render the overall experience boring.
We closed with some discussion about the need for an annotated bibliography as a general resource to jump start research in the area of "entertainment" and "fun."
Thanks to the members of the workshop who helped reconstruct the events, particularly Matt Belge, Jeff Johnson, Lars Erik Holmquist, Kevin Keecker, Maribeth Back.
Alexander, C. (1987) The Timeless Way of Building. Oxford: Oxford University Press.
Allen, R.B. and Breckler, S.J. (1983) Human factors of telephone-mediated interactive electronic games. ACM SIGPC/SIGSMALL Conference Proceedings (San Diego, Dec. 1983), 200-205.
Bachelard, G. (1994) The Poetics of Space. Beacon Press.
Back, M. (1996) Beyond bandwidth: sound design in interactive systems for multimedia and the Net. Invited paper in Proceedings of the Audio Engineering Society (AES'96), Los Angeles, CA.
Back, M. (1996) Sound design at the interface: narrative techniques in nonspeech audio. In W. Gao (Ed.), Proceedings of the First International Conference on Multimodal Interface (ICMI'96). Tsinghua University Press, Beijing, P.R. China.
Back, M. (1996) Micro-narratives in sound design: context, character, and caricature in waveform manipulation. In S. Frysinger (Ed.), Proceedings of the Third International Conference on Auditory Display, October.
Back, M. (1996) Reconceiving the computer game as an instrument for expression: narrative techniques for nonspeech audio in the design of multimodal systems. Thesis, for D. Des. 1996, Harvard Graduate School of Design.
Carroll, J. (1982) The adventure of getting to know a computer. IEEE Computer 15(11), 49-58.
Carroll, J. & J. Thomas. (1988) Fun. SIGCHI Bulletin, 19(3), 21-24.
Cohn, L. B. (1995) Violent video games: aggression, arousal, and desensitization in young adolescent boys. University of Southern California, Dissertation (Ph.D).
Crawford, C. (1990) Lessons from Computer Game Design. In Laurel (1990).
Crawford, C. (1984) The Art of Computer Game Design. Osborne McGraw-Hill.
Don, Abbe. (1990) Narrative and the Interface. In Laurel (1990).
Falstein, N. (1997) Speaking with a Forked Tale. The CGDA Report, 4 (1), 19-20.
Johnson, J. (1995) A comparison of user interfaces for panning on a touch-controlled display. Proceedings of ACM CHI 95, 218-225.
Johnson, J. (1990) Modes in non-computer devices. International Journal of Man-Machine Studies 32, 423-438.
Johnson, J. (1991) Effect of Modes and Mode Feedback on Performance in a Simple Computer Task. HP Labs Technical Report HPL-91-167 [also presented as a poster at CHI'89].
Johnson, J. & G. Engelbeck. (1989) Modes survey results. SIGCHI Bulletin, April.
Laurel, B., (Ed). (1990) The Art of human-computer interface design. Reading, Mass.: Addison-Wesley Pub. Co., c1990.
Laurel, B. (1991) Computers as Theatre. Menlo Park, CA: Addison-Wesley.
Lepper, M.R. & Malone, T.W. (1987) Intrinsic motivation and instructional effectiveness in computer-based education. In R.E. Snow and M. J. Farr (Eds.), Aptitude, Learning and Instruction III: Conative and Affective Process Analyses. Hillsdale, N.J.: Erlbaum.
Malone, T. M. (1982) Heuristics for designing enjoyable user interfaces: Lessons from computer games. In John C. Thomas and Michael L. Schneider (Eds.), Human Factors in Computing Systems, Norwood, NJ: Ablex Publishing Corporation.
Malone, T.W. (1981) What makes computer games fun? Byte, 1981, 6, 258-277 (Reprinted in Computers in Education (U.K.), 1982, 4, 14-21; and in D. Peterson (Ed.), Intelligent Schoolhouse. Reston, VA: Reston Publishing Co. (Prentice-Hall), 1984. Abbreviated version reprinted as: Guidelines for designing educational computer programs, Childhood Education, 1983, 59, 241-247.)
Malone, T.W. (1981) Toward a theory of intrinsically motivating instruction. Cognitive Science 4, 333-370 (Reprinted in D.F.Walker and R.D. Hess (eds.) Instructional Software,Wadsworth Publishing Co., 1984).
Malone, T.W. & M.R. Lepper. (1987) Making learning fun: A taxonomy of intrinsic motivations for learning. In R.E. Snow and M.J. Farr (Eds.), Aptitude, Learning and Instruction III: Conative and Affective Process Analyses. Hillsdale, N.J.: Erlbaum.
Nelson, G. The Craft of the Adventure. 5 articles on adventure game design, available at ftp://ftp.gmd.de/if-archive/info.
Polti, G. (1977) The thirty-six dramatic situations. Boston, The Writer, 1977.
Chuck Clanton has been a UI designer for the past 20 years, and recently worked in the game industry. He ran a popular CHI tutorial on "Film Craft in UI Design" with Emilie Young for several years.
Lynn Cherny is currently at AT&T Labs -- Research working on UI and interaction design for chat services and studying attributes of computer-mediated conversation and community.
Erik Ostrom is a a graduate student at Northeastern University who is working in an HCI Research department at AT&T Labs developing GUI toolkits for chat systems.
Chuck Clanton
Aratar, Inc.
222 Downey Street
San Francisco, CA 94117 USA
+1 415 372 3559
chac@aratar.com
Lynn Cherny, Erik Ostrom
AT&T Labs-Research
600 Mountain Avenue
Murray Hill, NJ 07974 USA
+1 908 582 {6183, 7168}
{cherny, eostrom}@research.att.com
Issue |
Article |
Vol.29 No.4, October 1997 |
Article |
Issue |