Home About EIS →

Agency Reconsidered

Steven Dow, Michael Mateas, Serdar Sali, and I have an abstracted accepted for DiGRA titled “Agency Reconsidered.” We’re working on the full paper this month, and will certainly share it when available, but one of the things I value about blogs is that they provide a place to do academic work in public. So I’m posting the abstract here, along with some thoughts on where we’re going for the final paper, and I’d appreciate any ideas/pointers that people have. Suggestions and criticisms that arrive now (rather than after it’s completed) are much more likely to shape the final paper.

Agency is a fundamental concept for understanding computer games, and digital media generally. However, as with many important concepts, its initial formulations were not entirely fleshed out. Subsequent work has developed the concept further, through humanistic, design-oriented, and empirical means. This paper provides a synthesis of some of the most important work on agency over the last decade and describes how a more mature formulation can serve as a basis for future research.

The fields of game scholarship and game design have two different starting points for discussions of agency. In scholarly circles the concept is generally attributed to Janet Murray. In Hamlet on the Holodeck (1997) she writes: “Agency is the satisfying power to take meaningful action and see the results of our decisions and choices” (126). In the field of game design the idea is often associated with Doug Church, who uses the term “intention” for the concept. He writes of “allowing and encouraging players to do things intentionally” (1999) — understanding the game world well enough to make and execute a plan of action, then seeing a clear reaction from the game world. Both Murray and Church lean heavily on examples of in-game navigation, with Murray focused on Zork and Church on Super Mario 64.

A focus on navigation, however, leaves aside a fundamental issue: the motivation of players to move, or plan to move, to particular places. What creates the desire that agency satisfies? The first part of an answer is the fundamental contribution of Michael Mateas’s formulation of agency, integrated with Brenda Laurel’s theory of neo-Aristotelian drama, as a balance between material and formal constraints (2001). This moves agency from an amorphous ability to plan and carry out actions to a phenomenon in which the actions motivated by a game are matched with those it enables. In other words: Agency requires evoking the desires a work satisfies. This became an important design guideline for Mateas and Andrew Stern’s Façade project (2005).

Further work has investigated how works evoke audience desires and attempt to satisfy them. Noah Wardrip-Fruin (forthcoming) stresses the importance of audience expectation of the fictional situation, the resulting patterns of play, and the relationship to underlying computational systems. For example, Joseph Weizenbaum’s (1966) Eliza/Doctor system successfully evokes expectations of a therapeutic situation, and both accepts and responds to audience statements, yet the play patterns that emerge do not produce agency, but breakdown. On the other hand, Will Wright’s SimCity (1989) actually responds to the play patterns evoked by audience expectations of city planning. Through play, the system transitions the audience from initial expectation to an understanding of the possible audience actions and system responses. The lesson: agency requires transitioning audiences from initial expectation to an understanding of a computational system, something that can be obscured by focusing on trivial examples such as spatial navigation.

Expectation is also central to Steven Dow’s (2008) study of players in two versions of Mateas and Stern’s Façade system. One group played the original desktop version of the game, while another played a fully-realized augmented reality (AR) version (with a physically constructed set, including furniture and props, onto which the Façade characters were projected via a head-mounted display, and with support for both spoken and bodily interaction). The AR version succeeded in increasing the audience’s feeling of presence. However, perhaps counterintuitively, the audience’s experience of agency decreased. Greater presence simultaneously increased audience expectations of the resulting experience, making the distance between everyday conversational interaction and the actions supported by the Façade system a larger gulf to cross. In short, presenting an approximation of reality (given we lack the necessary elements for a true “holodeck”) can actually decrease, rather than increase, the audience’s experience of agency.

The resulting conception of agency is still one in which the audience feels a satisfying power to take actions intentionally, based on their understanding of the game world, and see results. But it now considers the source of the audience desire to take actions in a well-designed experience, the importance of building an understanding of how actions and outcomes map onto an underlying system, and the centrality of audience expectations as set both by fictional situation and interface specifics. The paper authors are now using this as the basis for an empirical study of dialogue interface designs.

While our reviews for the abstract were generally positive, one thing I’ve learned recently is that it’s always a mistake to mention empirical studies unless that’s the main thing you want people to comment on. So, while we are doing some user/player studies, probably no mention of this will make its way into the final paper.

Another thing that the final paper will need is a deeper review of the literature. While there was only space, in the abstract, to outline the main work on agency that informs the paper, there has been quite a bit of publishing on the topic. Serdar and I have been thinking about which to address, and here are some possibilities we’ve collected:

D. Fox Harrell and Jichen Zhu, “Agency Play: Dimensions of Agency for Interactive Narrative Design” (link)

Paul Cheng, “Waiting for Something to Happen: Narratives, Interactivity and Agency and the Video Game Cut-scene” (link)

Esther MacCallum-Stewart and Justin Parsler, “Illusory Agency in Vampire: The Masquerade – Bloodlines” (link)

Ulrike Spierling, “Interactive Digital Storytelling: Towards a Hybrid Conceptual Approach” (link)

Rune Klevjer, “Computer Game Aesthetics and Media Studies” (link)

I’m also likely to include some of the thoughts on the everyday world (and references to Paul Dourish’s work) found in my Ars Electronica talk from last year (streaming version). Any thoughts on other things we might include, or any responses to the overall paper concept?


About the author:  Noah Wardrip-Fruin is a Professor of Computational Media at UC Santa Cruz and the author of Expressive Processing: Digital Fictions, Computer Games, and Software Studies. Read more from this author


This entry was posted in Academics. Bookmark the permalink. Both comments and trackbacks are currently closed.

23 Comments

  1. Posted July 13, 2009 at 5:58 PM | Permalink

    I see that a lot has been published on agency and games, including “Towards a taxonomy of perceived agency in narrative game-play,” Bride Mallon, and Acting with technology: activity theory and interaction design, Victor Kaptelinin and Bonnie Nardi. Of the papers you list, I can vouch for Fox and Jichen’s as being worthwhile and relevant for your project. Good luck on it.

  2. Posted July 14, 2009 at 12:56 AM | Permalink

    The Oz project’s drama manager (as described in Peter Weyhrauch’s 1997 PhD thesis) had a direct computational implementation of a very formal-constraints notion of agency, mixed with some author-specified notion of what’s important in the story, by keeping a count of the number of major “options” available in the story (which constitutes a major option, and what constitutes it being “available” both being hand-coded by the author)— the more options not yet foreclosed, the more agency.

    Well, I don’t think Weyhrauch himself called this “agency”, but he sort of implied it, and Michael and I characterized it as such in a 2008 AAAI paper (“Another look at search-based drama management”).

  3. Posted July 14, 2009 at 1:23 AM | Permalink

    In a different direction, I’m not sure how wide a net you want to cast, but there is some interesting stuff in the interactive-cinema community that seems to be talking about similar things. This paper by Bennington & Gay, for example, discusses how giving power to the spectator-navigator to reorder/disrupt/structure an experience rather than merely viewing it linearly allows “the agency and intentionality of the spectator-navigator [to intrude] into the expressed perception”.

  4. Dakota Reese Brown
    Posted July 14, 2009 at 7:59 AM | Permalink

    Is the basic argument being made in stating:

    In short, presenting an approximation of reality (given we lack the necessary elements for a true “holodeck”) can actually decrease, rather than increase, the audience’s experience of agency.

    …that agency too suffers from an uncanny valley effect when presented to the user in such a manner that approximates, but violates, human norms?

  5. Posted July 14, 2009 at 8:58 AM | Permalink

    What do you think is the role of first-principles philosophical conceptions of agency in this question (e.g. action theory), in addition to media- and design-theoretical ones?

    One snippet of the above that intrigues me (“Agency requires evoking the desires a work satisfies”) also makes me think that deep answers cannot be found in media theory alone. This is rather the reverse direction from empirical studies.

  6. Posted July 14, 2009 at 11:39 AM | Permalink

    oh cool… I look forward to seeing how this paper develops. Since its DiGRA you should also see questions of agency (especially in reconsideration) raise the hairs on the backs of most of us social science and cultural studies types… super fun especially if you even slightly invoke empirical studies of players :)

    One thing intrigues me though by starting the discussion with Murray vis a vis design, whats the difference between “taking meaningful action and seeing the results of our decisions and choices” and having the feeling like we are taking meaningful action…? Who is attributing this ‘agency effect’ if and when it occurs?

    Hmm.. how about some more radical refs for this reconsideration? Lucy Suchman’s Human-Machine Reconfigurations

  7. Noah Wardrip-Fruin
    Posted July 14, 2009 at 2:39 PM | Permalink

    Thanks everyone for the thoughtful comments!

    Today is a day I’m spending with my daughter, so I only have time for a quick reply while she naps, then probably offline for any further discussion today. But hopefully this is a topic we can continue tomorrow.

    Dakota, I’m going to ask Steven Dow if he can address your question, because that’s a section of the paper about his research.

    Nick and Mark, thanks for the pointers! I realize, given the length constraints of DiGRA papers, we may not get in as much literature review as I’d hoped. But if all goes well there may also be a journal version of this paper, which would have space for significantly more.

    Ian, I think conceptions that flow from things like phenomenology and activity theory can inform our thinking generally in helpful ways. I tend, however, to lean on people like Dourish and Nardi — who provide good discussions of ways these can be mapped onto human/software contexts. However, those mappings are usually CSCW in flavor, so I think there’s another set of thinking to find ways they’re useful in theorizing and analysis for digital media. Once we’ve done some of that, and combined with other insights specific to digital media, I think we have some ideas that we can use in the normal way (e.g., as design guidance) and that also can help us think about what to look for in user/player studies. In our case, the discussion we’re dropping from this paper is of a study of dialogue system interfaces, which have different sorts of arguments made about them, and we see these arguments as placing emphasis in different areas of this conception of agency. Now we’re planning to try this out with some live people and see how we get surprised :-)

    Bart, I think the pointer to Suchman is a good one. Actually, I was planning to invoke her in discussing how Church’s description of intention/agency is very much in language that sounds like old-school “intelligent action flows from planning.” Obviously, agency in gameplay is a lot more improvisational than that. Do you have other thoughts about connections to Suchman?

  8. Posted July 14, 2009 at 2:51 PM | Permalink

    Not necessarily a helpful connection, but something that’s bugged me for a while that Ian’s invocation of action theory brought up: is the question of agency in games just the virtual-worlds version of the question of free will in the real world?

  9. Posted July 14, 2009 at 3:25 PM | Permalink

    Hmm, I think you may have misunderstood what I meant… I was not citing activity theory, but action theory, which is one approach to understanding agency in philosophy. But no matter really, because the more important question is…

    What is the thesis of your paper? I mean that in an earnest way… the abstract intrigues me, but I’m not exactly sure I’ve fully grasped it. I get from the final graph that it’s a definition of agency, but I’m still a little at sea vis-a-vis what precisely you want that defn to include? Or is the thesis more structural, namely that agency is more like a recipe than an ingredient, and here are some of the common processes by which it can be prepared? Or?

  10. Posted July 15, 2009 at 7:09 AM | Permalink

    On Suchman… still not sure yet what I’d suggest but your answer to Ian’s questions will help. Its just a hunch if you want to get into (or cover) more radical ontologies of distributed agency which is where I imagined the Dourish line would take you if you want head a bit away from CSCW and straight up HCI… but there are just so many trajectories to cover here. If you can find a way to turn this into a thick and rich (and chocolaty) discussion at DiGRA I am there in heartbeat.

    How about something related like Adrian Mackenzie’s Cutting Code? I’d like to convince him to come to DiGRA actually – there’s a good conversation there to.

  11. Noah Wardrip-Fruin
    Posted July 15, 2009 at 4:45 PM | Permalink

    Mark and Ian, thanks for pointing out that I misread “action theory” for “activity theory.” After Nick mentioned the Kaptelinin and Nardi book (and given that things like Vygotsky and activity theory were big in my UCSD department) I read one for the other. Anyway, on to other topics.

    As it happens, I think Ian and Mark’s comments are connected beyond pointing out my misreading. This paper is about one of the many ways people use the term “agency” — which is a specific experience with digital media (not the only important one, though it may seem that way when discussed by some) identified by authors like Murray and Church. Unfortunately, because it is often discussed as “being able to do what you want” it sounds a lot like free will. But that elides the way that games are structured microworlds, suggesting and supporting actions in ways quite different from the everyday world.

    The thesis of our paper is that we’ve been clarifying elements of thinking about agency which, even though we’re focused on the same digital media phenomenon that attracted the attention of Murray and Church, should now be combined to replace the loose, early formulations of Murray and Church. It’s rather modest, I guess, in that we’re not trying to make the leap to big picture questions of agency and related topics (Wendy Chun, for example, has suggested connecting with the argument of Butler’s Bodies that Matter). That’s an enticing next step. But for now, like Murray and Church, we’re talking about the way digital media authors and interpreters conceive the experience of agency within the fictional world of a game. I haven’t tried to formulate it in these terms before, but perhaps our thesis is still that agency is “the satisfying power to take meaningful action and see the results of our decisions and choices” but that we more effectively design toward agency (and understand its experience) when we see how internal fiction and surface presentation combine to produce expectations that its systems are designed to satisfy.

    Or that’s what I’m thinking at the moment. We’ll see what my co-authors think when I try slipping that into the draft…

    But, more seriously, Bart, Ian, and Mark, does that address what you were looking for in the way of clarifications?

  12. Posted July 15, 2009 at 5:39 PM | Permalink

    Noah: that clarifies some of my questions, thanks!

    In retrospect, the conflation with free will wasn’t totally apropos, since the problems are different: the no-free-will issue is usually the inability to make choices at all (if they’re causally determined by the external world), while the no-agency-in-games issue is that, while you can provide different input, in some important sense the game just ignores it and does what it was going to do anyway.

    Would this be a fair reading of what you’re proposing? —> You initially stay with a fairly convention definition of agency as “the satisfying power to take meaningful action and see the results of our decisions and choices”. But you propose to unpack “satisfying” and “meaningful” in terms of the game’s interpretive layer, in particular with reference to the player’s expectations of which of their choices should matter, and how they should impact the visible outcomes. This could be contrasted with a more formalist definition of agency that identifies it with the literal ability to produce different outcomes by providing different input. That is, you’re arguing (?) that if someone were to write out a story graph—all possible sequences of outcomes given different player choices—that would by itself not be sufficient to determine the extent to which the game supported player agency, despite the fact that it’d let you figure out how linear vs. dependent-on-player-input the narrative was.

  13. Posted July 15, 2009 at 6:20 PM | Permalink

    Ok, I’m going to try what Mark tried, namely restating what I think your position is. Then we’ll take it from there.

    Conceptions of agency in digital media (e.g. Murray, Church) are helpful and correct in the abstract, but they are too loose and abstract to remain useful for design or analysis. There are many refinements needed, but one involves the coupling between agency and a game’s fictional world.

    In that regard, one important way that agency takes form is not in what players feel able to do, but in how the coupling between fiction and presentation makes those actions evident, as well as the effects such actions will have (both immediate and emergent).

    Am I close?

  14. Noah Wardrip-Fruin
    Posted July 15, 2009 at 10:22 PM | Permalink

    I think you’re both getting at important parts of what we’re arguing.

    Mark, I definitely agree that showing the possible player actions is insufficient for trying to figure out what’s going on in terms of agency.

    Ian, I agree that the coupling to the fictional world is key, but only one element of what we’re after — and it’s as important how the fiction suggests actions as how it helps make them evident.

    Here’s another attempt at a summary of what I see as the paper’s argument: Murray and Church are pointing toward an important part of game experiences. But it’s also necessary to consider the way the game’s world/fiction suggests certain actions (the things it would be “meaningful” to do) which, if matched with those actions supported by the game, produce satisfying results (Mateas). And, in fact, what a player who experiences agency is actually doing is operating a software model, so a key aspect of this is transitioning them from their initial expectations (derived from the everyday world, other fictions, previous software experiences, etc) to an understanding (perhaps implicit) of what is supported by this particular piece of software (Wardrip-Fruin). Further, expectations for how closely the operations of the software model will match those of its referents in the everyday world are shaped by other factors in the experience, such as the interface, and when those expectations are strongly violated the experience of agency is diminished (Dow).

  15. Posted July 16, 2009 at 7:41 AM | Permalink

    ooh okay – this is starting to sound a bit like the prescription, delegation, scripting and descripting stuff from actor-network theory (Madeline Akrich especially but also Latour and co.) Very big from my time at UCSD (which was before your time there but I too remember all that activity theory keeness)

    You seem to be after some kind of notion of agency as a matter of translation or transposition from normal/everyday contexts to gameworld contexts… in which case knowing how software acts on players to produce a transposed universe of expectations that produce a feeling (I still say agency is just a feeling one has) of agency is the key… ugh that was a yucky mouth-full.

  16. Posted July 16, 2009 at 9:10 AM | Permalink

    Ok, we’re getting somewhere. Here’s what I extracted from your last comment as the key idea:

    Agency is successfully operating a software model.

    That’s obviously a simplification, but perhaps that’s where my feedback is leading: it shouldn’t take this much effort to identify the thesis of your argument, so maybe you want to start with something oversimplified and provocative, and work from there into the nuanced detail you obviously have.

  17. Noah Wardrip-Fruin
    Posted July 16, 2009 at 10:09 AM | Permalink

    Ian, I think the issue may be that this is inherently an “integration” paper. It’s about taking the ideas on agency that Michael, Steven, and I have suggested and integrating them into something we hope can be of use to the field. So, the provocative statement you extract presents my part, but not Steven or Michael’s, and our argument is that the three together are what we need to move toward. But maybe I can make the combination short and provocative? Here’s a shot:

    The feeling of agency in a game’s fictional world arises from successfully operating a software model — specifically, one that supports the actions suggested by the world, that helps players understand how their actions have impact on the model, and that is presented within an interface appropriate to the model.

    Hmmm… lost some punch, I guess. Maybe something more provocative is:

    Agency, as outlined by Murray, is a thought-provoking but underdeveloped concept. We integrate three bodies of work, focused on how games suggest and support actions, to take the discussion of agency to the next level.

    Better?

    Bart, yes, I think this is part of it (again, mostly my part, which is probably an outgrowth of me being the one posting/commenting here). The idea is that, for actions from immediate navigation through long-term simulation cultivation, the feeling of agency is helped both by being able to do what the fictional world (and interface) suggest would be appropriate and by transitioning from one’s initial expectation of how that would be done (and what it’s impact would be) to what is actually present in the software model.

  18. Posted July 16, 2009 at 11:11 AM | Permalink

    The second summary is punchier but doesn’t say as much. But I have more clarity now, which is good!

    I’m not sure you really want to locate agency in the fictional world? This is something like what Bart said above about agency being in the head, but you’re sort of going for an ANT-like “agency is a network” sort of thing.

    If you wanted to go that direction, you end up with something like:

    Agency is a network of: a software model, that model’s player operator, the possible actions that model’s representation suggests, the interface that presents those actions.

  19. Noah Wardrip-Fruin
    Posted July 16, 2009 at 2:57 PM | Permalink

    Glad to be finally communicating the argument! I think this process is definitely going to improve the readability of the final paper.

    As for the above summary, maybe what I really should have written is agency in relation to the game’s fictional world? I’m trying to make it clear that we’re not talking about any imaginable kind of agency in relation to the digital object. For example, games made in Kodu have an ever-present prompt suggesting that one open up, examine, and perhaps alter the code. We’re not talking about that, or many other possibilities.

    As for whether to throw ourselves in with a particular theoretical paradigm (like ANT) I’m not sure we’re ready to do that as a group. Sounds more like future work, for us or others.

  20. Posted July 16, 2009 at 9:48 PM | Permalink

    Ok, so now that I have some understanding of what’s going on, I feel compelled to point out a possible gripe, namely that someone might observe that you don’t really have a theory of agency, but a bunch of threads from your various individual research trajectories that deal with agency, which you’re attempting to marry in a cohesive (but not gift-wrapped) whole.

    There’s nothing wrong with that if the result is a complex and layered dish, rich in surprising but compatible flavors. But there might be something wrong with it if the result is a plate with a number of individually tasty flavors that don’t go together. (Can you tell I’ve been watching a lot of Food Network?)

  21. Posted July 24, 2009 at 4:17 PM | Permalink

    Awesome discussion. Noah, great idea to open up our discussion to others! I’ve been traveling the past two weeks so I am a bit behind, but I’m going to comment on a few statements that struck me throughout this thread:

    Dakota: “…that agency too suffers from an uncanny valley effect when presented to the user in such a manner that approximates, but violates, human norms?”

    This of course this depends a great on how we define agency. In the past I’ve defined agency as a sense of empowerment over events in a world. Agency is an experiential pleasure. As such, it can fade in and out; it can fail altogether. Agency is not automatic, and so simulated environments should be cleverly constructed to help users/players get there. Designers can, of course, choose not to focus on empowering players. Or, they may deliberately disrupt a sense of agency. I don’t see it as an ultimate goal.

    Another distinction: is agency experienced as a result of your actions or also from the actions of other agents? Ken Perlin talked about feeling agency in linear media by living vicariously through a character’s challenges (e.g., he experiences Harry Potter’s agency). Here we may need a different categorical descriptor to refer to the perception of agency in others. I’ll say self-agency and external agency.

    The uncanny valley typically refers to artificial appearance and behavior on external agents (e.g., robots, and I suppose digital characters). I believe the uncanny feeling arises from a mismatch in expectations—people expect something that looks and behaves so close to “human norms” that every nuanced violation stands out. There’s a relationship perhaps with self-agency, but I think the uncanny valley maps better onto external agency.

    Bart: “whats the difference between “taking meaningful action and seeing the results of our decisions and choices” and having the feeling like we are taking meaningful action…? Who is attributing this ‘agency effect’ if and when it occurs?

    I think that’s a subtle but important distinction. In my notion of agency as an experiential pleasure, I attribute agency to the “feeling”. In that sense, the feeling of empowerment can be an illusion.

    Ian: “is the thesis more structural, namely that agency is more like a recipe than an ingredient, and here are some of the common processes by which it can be prepared?”

    If we say that agency is an experience or feeling, then it’s the pleasure you get from eating the meal. Our thesis, on the other hand (and I need to talk to Noah in more detail about it), offers a recipe for how to prepare users to enjoy the meal (i.e., we need to set people up to enjoy agency).

    Noah: “we more effectively design toward agency (and understand its experience) when we see how internal fiction and surface presentation combine to produce expectations that its systems are designed to satisfy.”

    Of course, designing towards agency is not a foregone conclusion. Recent works by Mateas, Pheobe Sengers, Alex Taylor, Bill Gaver, etc. would suggest that merely satisfying expectations is dull. Violating expectations may make artificial characters more engaging. Following on my prior distinction, violating self-agency has a very different experiential effect than violating expectations on external agents. We want to talk about self-agency, I believe.

    Noah: “As for whether to throw ourselves in with a particular theoretical paradigm (like ANT) I’m not sure we’re ready to do that as a group. Sounds more like future work, for us or others.”

    I agree. This gets into much deeper theoretical waters. I’m only marginally knowledgeable of Andy Clark’s recent argument—see “Supersizing the Mind”—which delves into the relationship of cognition and agency. I believe he claims the “mind” resides outside of the body—in networks of brains, machines, etc.—but agency is bounded to single living embodiments. Of course, one could argue that agency cannot be experienced without a world to act on. Maybe action network theory is appropriate.

    Ian: “someone might observe that you don’t really have a theory of agency, but a bunch of threads from your various individual research trajectories that deal with agency.”

    Someone could argue that we are pulling together disparate threads. More importantly, does our theory make sense? Is it valid? Does it move the discussion forward? I think Noah’s arguments represents the crux of our paper’s central thesis. Here’s my own version:

    Players experience agency in simulated worlds when they feel empowerment over events. Designers of simulated worlds can cultivate this pleasure by transitioning players from initial expectations to an understanding of supported actions.

    This “transition” requires that a simulated world builds on and does not violate expectations from derived from the everyday world (the physical world, other fictions, previous software experiences, etc.,) and that it suggests/motivates world-specific actions which are balanced by corresponding system reactions.

    Should we be saying more about how the transition happens?

  22. Posted July 24, 2009 at 4:36 PM | Permalink

    @Steven: If agency is solely the experiential pleasure, then would you argue that there’s no agency-relevant difference between an experience in which the player has a strong feeling of having taken meaningful action, but in fact there was little to no impact on the trajectory of the experience; and one in which the player’s actions did have actual impact on the outcomes?

  23. Posted July 25, 2009 at 10:07 AM | Permalink

    Mark, I interpret agency as a psychological state. As such, I believe players can feel agency even when the system does very little to respond to a player’s actions. We witnessed indications of this in the final stages of Façade. When most players realize their input is having little affect, a few players talked about how Trip and Grace were too caught up in their own drama to really listen to them. They continued to console, argue, and engage the characters. They believed the illusion of having agency, even though their actions had very little real impact at that point. On the flip side, a system might go far to give players the ability to make real impact and this doesn’t guarantee they will feel a sense of agency.

    This does not imply that the design of a simulated world has no effect on player agency; quite the contrary. Much like a well-orchestrated magic show, an interactive system must perform subtle tricks to keep the audience in suspension of disbelief. We argue this means understanding initial expectations, establishing motivation for action, transitioning one’s beliefs about actions onto a current software model, and following through with system reactions. We are saying (I believe) that the structure of the software model can affect (in non-obvious ways) whether (and how many) people believe they have agency. Noah, we should talk more about this part.

3 Trackbacks