Governing the Magic Circle: Regulation of Virtual Worlds, UCI-Irvine

Panel 1

Sal Humphreys with Melissa de Zwart, U South Australia, Griefing, Massacres, Discrimination and Art: Exploring the Limits of Overlapping Rule-sets in Online Games

Underlying question with little consensus: what kind of space is this? Enclosed/hived off by magic circle, or permeable? Important when determining who should define the rules, implement them, reap the benefits. Magic circle: all governance issues to be resolved by service provider/players. Many students of games have dismissed this because of gray markets, other interpenetration between gamespace and elsewhere.

Foucault’s heterotopias: interaction between legal rules, game rules, and social/cultural contexts. Space acquires definition of meaning through context. Thus gamespace can’t be nailed down to one meaning. Heterotopias are between real space and utopian (placeless) space: other sites are simultaneously represented, contested, and inverted: like a mirror. A sort of shadow that enables me to see myself where I am absent. Partly a reflection of the self, partly a new mediated construct: the game self is partly physically located (because you are) and partly in the wires.

Game space has explicit rules. Spaces outside of everyday. Rules are hardcoded by developers as much as possible to enforce them. But there are also tacit rules, in the emergent space of gameplay—community norms, appearing over time. Broader than that: legal systems and civil rights/broader community norms.

Almost no players read the EULA. Informed consent is not what it achieves. More protection of risk for publisher. Ill-informed consent for publishers to control the space. Also generally fails to provide meaningful conduct guidelines, and is rarely used by community managers or users when negotiating behavior. Condemned to uselessness by its legalese.

Actual incidents: David Myers and Twixt, playing against social rules while complying with game rules. Myers engaged in behavior that had been modified (considered unfair) by social norms of the game. The social norms made City of Heroes more a society and less a game. He perceived this as a repression of his playing. Became subject of “kill Twixt” campaign. Players don’t have access to the same sanctions as programmers. People resorted to threats, public humiliation, exclusion from formal group structures.

WoW funeral massacre: a player died in real life; her guild held an online funeral and a competing guild broke in and massacred them. Lack of consensus: was this griefing or legitimate game play? If games are a space of inversion and transgression, how far can bad behavior go? (I’m reminded of xkcd.) Some said the mourning group broke the roleplaying codes by bringing real-world issues without the consent of other players, expecting other players to conform to the change.

Consensus is not to be found in million-player games full of players from different cultures, unlike consensus on a field.

GLBT Guild recruitment ad in WoW by Sara Andrews. Recruiting ads are seen as legit ways to increase numbers; she advertised that her guild was GLBT-friendly, which is important in an often hostile world. Customer service manager on duty sent her a message that it was against the harassment policy to use language about sexuality, and that if she did she’d be banned; also argued that an openly GLBT guild would invite harassment. Blizzard initially supported the manager. Lambda Legal sent a letter to Blizzard outlining potential action on antidiscrimination grounds. Characterized WoW as a public accommodation; unclear on the law. Blizzard eventually reversed and apologized.

Interweaving of different levels of rules and norms. Explicit rules: designed to prevent harassment. In practice: manifest failure; LGBT harassment is common and customer service doesn’t behave as if it’s a problem. But when these rules were used to justify silencing, game players sought outside legal help. Accountability within the game is virtually impossible, so help had to come from outside.

Twixt: rule adherence is trumped by social norms. WoW massacre: whether out of game cultural norms should trump game rules. GLBT: relation between outside civil rights law and game. Interpenetration shows how games are heterotopias. An empirically descriptive theory about rules, laws, norms, and the operations of power within countersites.

Ted Castranova – A Regulatory Model for Virtual Worlds and Other Shared Media Environments Virtual World Regulation
Supply and demand works when regulators are trying to determine quantity—making more increases marginal cost, depletes demand. A media product isn’t like that—you decide on how much of a given feature to provide, and if you make it too intense you reduce users, and then there’s a separate layer of monetization. Features attract eyeballs: people want the environment to have enough monsters to be fun, but not so many you get killed as soon as you log on. Eyeballs then generate benefits. Features also cost money: it’s not easy to make monsters.

Feature intensity is an optimal combination of eyeballs, cost, and profit. Anything can be a feature: how cute Barbie is in Barbie World—if it goes to far she becomes cloying, so you’re looking for the sweet spot of just the right amount of Barbie. The cost-benefit curves look different than physical-world cost-benefit curves. Note also that eyeballs affect eyeballs: network effects give increasing returns, though eventually there is a congestion zone. In congested circumstances, changing access price can enhance user value. Marketing can also lead to a direct increase in eyeballs, with possible positive or negative effects for other users depending on where you are on the curve. Can directly manage eyeballs (viral marketing, advertising to affect the access price), but these options for managing eyeballs are also features (people can like them or they can be annoying).

Production decisions are about intensity of features, not quantity of output, because the marginal cost of distribution is zero. Managing features through costs is fairly predictable, but the effect of changing access prices is janky https://secure.wikimedia.org/wiktionary/en/wiki/janky.

Kristin & Trey Hickman - The Myth of the Magic Circle: Rejecting a Single Governance Model

KH: What to do with instances that don’t quite fit a single-metric definition of a virtual community—e.g., when the system is closed v. when you can cash out in real cash; there are often instances that don’t quite fit either. Which metric you use may depend on which behavior you are trying to regulate. Thus legal scholars have been unable to achieve consensus. Need a more complex and comprehensive framework for defining the magic circle to apply real world laws with any kind of legitimacy.

TH: What is a virtual world? Three elements—interactivity (users communicate), physicality (interface allows existence and interaction with “you”), and persistence (possessions are owned; user’s spaces can be used even when the owner goes away).

Is Farmville a virtual world? Yes. What about Facebook? Interactivity yes; physicality—navigation through environment by visiting other pages, and the avatar is one’s own representation of oneself, and may have no basis in reality; persistence—yes, the actions and events you inspire persist over time. So is Netflix a virtual world? Communication is based on Netflix assets—movie reviews, preferences, ratings; physicality--your presence is defined by the info you provide; the point of the Netflix system is to have persistence so that it can react to your preferences.

So are we really talking about online communities? Social interaction is the point of the virtual world. Given this range, from Netflix to Second Life, you can’t say these two things can be analyzed along the same linear axis. Few scholars agree on what the “magic circle” really is, but the idea of play means different things to different individuals. We’ve seen real life murders associated with theft of virtual items; real life weddings and divorces result from things that happen in virtual worlds. Where’s the magic circle in Netflix?

Modern Activity Theory: individuals treated as subjects as part of a social community; related to Human Computer Interaction field for past two decades. Allows us to visualize what happens in virtual worlds. Subjects interact with tools, objects, rules, community, division of labor—all of which also interact with each other—resulting in an outcome. This allows us to contrast aspects—Facebook is self-life oriented, whereas WoW is player-goal oriented, and so on.

So what? They provide six axes of analysis of self-life/player-goal type; there is too much differentiation in virtual worlds to say comprehensively what is subject to regulation.

From a regulatory perspective, simple universal standards are necessary for successful regulation. Preliminary conclusion: leave regulation to internal mechanisms and markets. This means that real problems will be left unsolved. We don’t doubt that there is a magic circle, allowing a theoretical definition of what should be regulated, but real world application of existing laws may not be practically definable/operationalized to distinguish in-game from out-of-game effects.

My Q: I am unclear what they mean by “regulation.” Can a traditional commercial transaction taking place on FB be regulated under standard contract/fraud principles? What about defamation via twitter? Aren’t you just pushing the question to what is “regulation”?

KH: There is a point of real-worldness, where you’re not regulating what’s going on inside the community so much as regulating the publisher. Tax perspective: at some point you actually have cash value income. There’s also a point at which we have to say what happens inside the world stays inside, and we’re best off rather than trying to apply external defamation law to a community that may not have the same consensus about what constitutes defamation; allow the internal community to police itself. Or allow the publisher to police that—ban a bad actor. That’s more effective/legitimate. (I don’t think this answers the criticism; first of all, defamation has plenty of contextual standards to figure out whether the people receiving the allegedly defamatory statement would perceive it as a statement of fact, etc., and being said in WoW would matter to the application of the law of defamation. More generally, I still don’t know why this model tells me to leave defamation via twitter alone versus defamation via traditional newspaper, or indeed whether it does. Also, given the harassment just described by Humphreys in WoW, where WoW will act just enough to avoid being sued for discrimination itself, I don’t find self-policing necessarily any more persuasive than it is in the physical world.)

Eric Goldman engaged Castranova on eyeballs and ads. Q: could there be a situation where increasing attention decreases revenue because users are so focused on the game they don’t see the ads?

A: Hadn’t really considered that. There’s a difference between the eyeball curve and the benefit curve—with no ads, there’s no revenue; with ads, you drive people away, but make money. It’s not orthogonal.

Laura Heymann for Hickmans: how do we value the concerns of the outside world?

TH: Our position is that people self-select the appropriate virtual worlds. Voting with your feet is the ultimate remedy. This is becoming more difficult as people become invested in particular worlds, but even though a shutdown is traumatic, in the end it’s just a virtual world. Since you’re self-selecting you must know what you’re getting into whether you read the EULA or not. If you have a problem with the norms of the virtual world being bullying or harassing, then don’t be in that virtual world. (Ah, exit. If you can’t stand the heat … Except that as Humphreys points out expectations differ. It’s not neutral to say “leave them alone.”)

KH: Ex post problem. If you go in thinking you know the rules, it’s different. Rules of sport may differ from rules of larger society; saying “that’s bad behavior” after the fact instead of working for changes in the rules of the game is fundamentally problematic.

Humphreys: this is about tacit rules, and how they vary; people think that they’re in a shared world but they actually have varied expectations; it’s about competition between tacit rules.

Q: Regulation makes markets possible. CDA §230, immunizing service providers from bad acts committed by users in virtual worlds. So when you talk about the “market,” that variation is generated by the providers’ legal indifference to players’ speech. Without significant regulatory protection we don’t see that market.

KH: there is a fine line. Blizzard is subject to SEC regulations. Regulations can protect or destroy markets. But mostly leaving internal conduct up to the world is more likely to serve the right regulatory function without destroying the market. Not ideal, but second-best. At some point regulation is too intrusive into the community. (I am left unclear what she is arguing against, even though we probably agree on a lot of specifics.)