Monday, April 11, 2011

Governing the Magic Circle panel 4

Ben Duranske - World of Lawcraft: Virtual Lawyers' Misadventures in Forming
Attorney-Client Relationships with Avatars

He’s in charge of virtual currency FB credits, used in Facebook games developed by third parties, such as FarmVille. CityVille is leading game, with 98 million monthly active users. A small fraction spend money in the game, which is a significant amount.

Monday Beam, Esquire in Second Life: this is an avatar who’s gone to great lengths to separate the avatar from an identifiable person. What are the ethical considerations? Model Rules of Professional Conduct: rules may vary by states.

Can’t necessarily take rules about email and apply them to virtual worlds.

Who’s the client/prospective client? Dealing with an avatar, you don’t necessarily know it’s controlled by one person. Many avatars, especially those controlled by corporations, are available to multiple people. Alternatively, one person may have multiple avatars—if someone says “I’m the alt of X” they may or may not be. For whom is the avatar speaking? The person who purports to be behind the avatar or the corporation who employs the person who purports to be behind the avatar?

Another rule: meritorious claims and contentions. Ozimals: animals that could be bred; if you treated them right, they’ll make more. Amaretto Breedables: sells horses etc. in Second Life. Sales are not insignificant. The gist of the copyright dispute is whether virtual horses infringe on virtual bunnies. One has to wonder whether the plaintiff questioned the wisdom of this; some judges and law enforcement agencies have expressed skepticism about claims he thinks are quite valid from a virtual world perspective. Virtual property: Bragg v. Linden Research. Mark Bragg, an attorney, was banned from Second Life for doing something, initially unclear. He was manipulating the Second Life auction system, which had preloaded parcels. He bid on parcels other people couldn’t see so he bought at extremely low prices. Second Life didn’t like that. Many interesting legal theories: an easement by necessity to get access to your “property”—is that a credible argument or are you risking sanctions? Another theory: maybe there was a right to ban him, but he still had a right to access the virtual goods/property he purchased, as a person would have access to his gym clothes even if he got kicked out of a gym. The case settled.

Duties to prospective clients: a lawyer can’t represent a client with interests adverse to a prospective client. But in a virtual world you may have a firehose of incoming communications. Linden Labs: an attorney who might someday represent Linden should be extremely careful acting in Second Life; people message him weekly outlining in some detail situations that could result in a lawsuit against Linden Lab. Could disclaim in communications, but various state bars say it’s not effective to put a little disclaimer at the bottom of the email; likewise, it may not be sufficient to put a disclaimer in your profile since most people don’t visit your profile to contact you. California Bar rule: disclaimer on website is only effective if it’s sufficiently plain (language and prominence) to defeat the visitor’s reasonable belief that the lawyer is consulting confidentially with the visitor. Simply having a visitor agree that an attorney-client relationship is not formed would not defeat a visitor’s reasonable understanding that the information submitted is subject to confidentiality.

Avoiding conflicts of interest: his general policy is to return unsolicited contacts saying “I haven’t read your email,” but he’s not sure that’s sufficient. Not sure it’s a good idea to hold yourself out as an attorney in virtual worlds.

Confidentiality of information v. Second Life terms of service: Linden Lab may observe and record your interaction with the Service. So you shouldn’t be giving legal advice in Second Life. Move the conversation quickly to traditional forms of communication.

Unauthorized practice issues come up—not sure of geographic origin of the person contacting you. Rule of thumb: are you confident that you’re not representing somebody in a jurisdiction you’re not allowed to practice in if all you have is as contact info?

Advertising: MPRC says any communication must include name/physical address of at least one lawyer or firm responsible for its content—how does that apply to an avatar?

Then you have the problem of proving yourself to the client. One of the only formal acts of the Second Life Bar Association: a verification program.

Joshua Fairfield - Avatar Experimentation: Human Subjects Research in Virtual Worlds

IRBs may not understand what you’re trying to do—one reaction: “you don’t need permission to do research on fake people.” But his position is that these communities care deeply about their communities, property, identities.

Researchers love virtual worlds because they allow researchers to do things they wouldn’t ordinarily do, like having people drive cars at each other to play chicken, and also to get real reactions—real monetary, dignitary, community interests, so you can find out how people really act. Unthoughtful experimental design can do harm.

Three legal interests at the top: property, community, individuality. Virtual worlds are evolving. There’s a reason 98 million use CityVille and 1/10 use WoW: greater computing power to use the nicer environments. Graphics downgrade as real-life social interactions upgrade (integrated with Facebook).

Problems that researchers face: (1) primary v. secondary research—the truck v. the clipboard. There is a loophole the size of a truck in the law governing informed consent in virtual worlds. It used to be, because of the abuses of the Nuremberg trials etc., we said you need real informed consent. Clicking “I agree” on a EULA is not enough.  But there’s a difference between primary (ethnography, or gathering your own data with bots or scrapers) and secondary (where the goods are, since developers collect everything—Second Life does record everything). A secondary researcher who received data collected by a primary researcher was ok, but we assumed the primary researcher had informed consent. Now, the secondary researcher is getting the info but the primary data comes only from the EULA. Any time anyone has sex in Second Life may be available to researchers. Best researchers won’t touch private information between players. Norms arise to handle this problem.

Problem of the clipboard: difference between qualitative and quantitative. What if researcher observes other people engaging in intimate activities without identifying herself as a researcher? Sometimes the community wants you to fit in; other times you may need to stick out—if the community is having a party and asks you to put on a costume, go along. But if you follow a subject to a sex club, probably should identify.

Law: three basic sources of law. Belmont Report—flagship of American regulation. Basic principles: respect full informed consent, beneficience (minimize harm, maximize benefit), and justice (equitable distribution of benefits of research). Especially for children, you can’t just do no harm—you have to do good.

COPPA, another source of law.

Federal Common Rule governs a pile of different organizations. If your university gets any money from the federal gov’t, all can be threatened by human subjects violation, increasingly aggressive in recent years. Sanction: defunding university.

Applies to all data either collected when subject has reasonable belief not being observed or reason to believe they won’t be identifiable. This includes names/info that could be used to contact the person in real life.

Terms of Service/EULAs also matter. You do agree the info can be passed on to anyone. EULAs are coercive from the human subjects perspective.

Note: my husband, Zach Schrag, has blogged about this paper at the Institutional Review Blog, which covers IRB issues in the social sciences.

TL Taylor, Associate Professor in the Center for Computer Games Research at the IT University of Copenhagen - Battles on the Field: Institutional Governance in E-Sports

Different ways rule systems can be sliced. We still think of game rules as the definitive place in which these things are sorted out, and that’s simply not the case. Game rules are not the final word on how the game is played. Is that true of high-end, professional computer gameplay as it is in MMOs? Yes, same messiness even with contracts and big money at stake. Actors, human and non, mediate how games are actually played. Referees, spectators, broadcast requirements shaping tournament rules, ping times, emergent play/discovered tactics/creative player performance.

People call e-sports a new industry. Commercial interests beyond those of the game developer reshape and structure actual play. Not just developers and player communities—other stakeholders like NGOs, teams, leagues. South Korea’s KeSPA, launched 2000 with support from Ministry of Culture and Tourism: Korean e-Sports Professional Association: cable channels support 24 hours of gaming/day. Gaming leagues are supported by banks and mobile phone cos. Culture, government policy, infrastructure affect gaming. Thousands of people go to stadiums to watch people gaming.

April 2010: Blizzard broke with KeSPA as partner, which was a massive change. Alleged that KeSPA illegally sold broadcasting rights to Blizzard’s StarCraft. Blizzard says it’s not about the payment itself: they wanted the property/IP rights acknowledged. Express concerns for treatment/payment of star players. Have an idea about how game should be played in tournament settings. Frown upon add-ons at events because they take away from the skill level. Didn’t want to be FIFA but understood themselves to have some interest.

KeSPA doesn’t see itself as just licensing: building a new sports industry. Examine and evaluate titles to see if it will be an official e-sports title, allowing developers to host league games—switch of power authorization. Want to provide refs, qualify winners. If a game achieves success as an iconic e-sports game, and developer pursues profits by declaring copyright covers sports industry, that’s an obstacle for the industry. KeSPA says it’s responsible for growing a scene that’s more than any single title. If we have to go back each time for permissions, it won’t work. What happens when artifacts become important beyond the intent of the original author? See Rosemary Coombe on celebrity.

Blizzard went to another organization, and KeSPA’s future is up in the air. Complex co-creation to produce e-sports. Narrow formulations of IP don’t do justice to co-creation.

Ripe for future problems: broadcasting/licensing. A tried and true issue in sports law. Clearly big stakes. People are currently rebroadcasting their gameplay experiences on various user-generated content sites. When a fan rebroadcast a game on Ustream, KeSPA had that player pulled immediately from the stage to keep control of its content. Can a company really assert the right to roll up the field and take it home?

Modified rulesets/interfaces—also common in e-sports. Will game developers object to rulesets? Has happened in Blizzard, which broke the modding system a few years ago.

Patents on human action? Routine human action v. expressive and creative human action—patents on yoga (unsuccessful), on putting a golf ball (successful). Relationship between computation and expressive human action: how does that affect how we understand player performance and potential patentability of tactics, strategies, etc.?

Rights of publicity. Current e-sports league contracts don’t talk about right of publicity. Underexplored. What about the right of publicity of avatars as expressive human actors?

Q: two themes from the day—people were getting tired of virtual worlds a couple of years ago. Now we see that Facebook and e-sports and other things are alive/vital.

When you talk about sports, that’s an environment where we’re really good at the magic circle: we know the line gets blurred all the time, but we still create and enforce the magic circle. That’s the debate on e-sports now.

Duranske: it would be easy to look at Second Life/WoW as flat, but with 98 million people playing CityVille, it happened in a way that we didn’t expect. Second Life relies on difficult tech, but we’re still seeing growth in virtual worlds. Magic circle is there to protect play as play. Game company wishes to exempt experience from real world, the way Eve does, or the NFL, where it’s reasonable to run up and knock someone over with your body, the magic circle has value.

Taylor: we sometimes misunderstand how rules work. Follow the debate in the NFL about concussions—rule systems are contingent. Magic circle becomes easy way to sidestep how contingent rules are and how much we don’t know in advance. Remains the dream that computation will take over players’ interpretive work; this isn’t true.

(My reaction: yes, concussions are a great example of how the magic circle is not so magic, even as a boundary from the rest of the world. Circle is used to deny responsibility.)

Fairfield: circle is cyberseparatist concept, and the good guys lost that battle. (I don’t see how this fits with the regulatory aspects of the paper.) We have seen virtual worlds triumph, not triumphant—virtual worlds pushing into every aspect of real life. Legal regimes used to govern dragons will govern your iPads and cars, governing what roads you can drive on as they govern what cellphone providers you can use.

Q: does professionalization of gaming lead to player organizations that are more organized/assertive? Players just want to play games right now.

Taylor: there are many conversations about that in e-sports. People now are often grateful to get to play at all; churn rate is high, and they go when they want to start families/have more financial stability. Things are moving towards professionalization but language, structures, mentorship aren’t quite there yet. Dirty secrets: lots of players don’t get paid their prize money.

My reactions to Fairfield: no mention of unchecking the box! Also, there’s an interesting contrast with the position taken here about obligations to research subjects and the debate over behavioral advertising and what companies can do with the information they have about users. Risk: Unilateral disarmament of researchers/people who do not have a vested interest in making money off users versus what the companies themselves know about their users.

Fairfield: it’s smart to uncheck the box, but many institutions don’t. Increasing divergence between people who take federal money and are tied to tight standards v. broad industry standards of data mining. That’s disastrous. What to do though? Re-consent with a follow-up survey, possibly. Another expensive move, removes quantitative value—won’t gather data on people they can’t meet in person.

Rosenblatt: the people with an interest are already mining the data. So is there a deep ethical problem, since we’ve already given consent to people with arguably more malicious intent to do this research?

Fairfield: commercial may not be malicious. One answer: we decided we wanted those restrictions on experiments by people in lab coats. (RT: Are these research projects experiments by people in lab coats?) And commercial uses don’t bother us that much yet. (RT: The decisionmakers here are very different in terms of time and knowledge.) Because there’s an “ew” factor up front, you don’t want to poison the well. There were early examples of people publishing about support groups, for example, and that felt like harm to members. That’s an ongoing issue for virtual worlds researchers. They can be distrusted as keyloggers, scammers.

Duranske: any company using the data to market is obligated to disclose that in, arguably, relatively clear terms in a privacy policy. (Duranske represents FB but is not speaking for it.) Relatively unappealing to add a line saying that they’ll be handing that info off to graduate students—aligned with government interest as articulated here. (I don’t think that last part follows. Especially with big data dumps, the research is not going to reveal individual information; I sincerely doubt the EULA restrains FB in any way from passing anonymized data to anyone it wants—even accepting that there have been challenges to whether or not the data can be de-anonymized. Thus, it seems to me that, on Fairfield’s argument, the virtual world might be willing to give you mass data that an ethical researcher should still consider not looking at.)

No comments: