Another datapoint follows for those tracking the rise of virtual property. A court has ordered the transfer of stolen virtual property back to the original player. Most interesting here is (1) it's a Chinese court which is leading the way, and (2) the court ordered the developer to transfer the property. This latter point merits attention that I don't have time for now, since I'm knee-deep in spackle. Will say some stuff in a bit.
[Edit: Go here for more stories. Original Reuters story that I included has been deleted from the post, since copyright is a blunt weapon and copyright plaintiffs have no sense of humor these days]
Score another one for the little people :)
Posted by: Lee Delarm | Dec 19, 2003 at 13:51
That rocks my world. That seems incredible. I don't know how truly surprised I should be though because I don't have a clue about Chinese law. Bear with my ignorance: is there a short answer that can describe the primary difference between Chinese and American legal structure? If one was paying attention, could she see that decision coming or does it strike me as so revolutionary only given American legal precedent and my insular thought (i.e., is their legal system ahead of us on this one, or structured so differently that it shouldn't mean much to the American system?) Also, what's the deal with appeals over there? Will this actually happen? No jurisdiction problems--- the company *will* have to do what they say in the end?
{Good call on linking us over to the story--- who needs a scary, sober, "humorless copyright plaintiff" chasing them around? Life's hard enough.}
Posted by: Candace | Dec 19, 2003 at 17:40
In Korea, if the reason of item loss are somebody's hacking, system bug, then gameing company recover the status of the injuried gamer.
While, if the reason are item trading, system lack, then the company refuse to help the gamer.
So, looking at this point of view, the chinese court's decision is not a sensational but quite natural ruling.
The case is about the passive right(preserving the status quo)of gamers be protected by company.
I guess the key point of the gamer's winnig is not at property(ownership or copyright) matters but at security matters.
The core question of whether gamers have full ownership including the active right to earn money by selling items is, IMHO, not fully answered by this ruling.
I woder how the chinese court will respond if a gamer sue the company for item loss by item-trading which EULA prohibites to do.
Posted by: Unggi Yoon | Dec 19, 2003 at 18:22
Thanks for the info--- I suspected it would be more of an earth-shattering deal to me, given my perspective so your comments help a lot.
Very interesting, indeed.
Posted by: Candace | Dec 19, 2003 at 18:45
"I guess the key point of the gamer's winnig is not at property(ownership or copyright) matters but at security matters."
It's about something much bigger. It's about liability.
Hmmm... There is something even deeper here. Liability towards what, and to whom? When was this liability created?
Posted by: DivineShadow | Dec 20, 2003 at 01:54
Dan Hunter>A court has ordered the transfer of stolen virtual property back to the original player.
The property was "real-world" stolen, in that someone hacked into the server and changed data outside of the context of the virtual world. It wasn't "stolen" within the context of the virtual world, which is kind of ironic given that it's only in the virtual world that the character "owns" the objects that the real-world player is claiming were stolen.
I'm not surprised by the ruling so much as I'm surprised by the fact that the operating company didn't transfer the objects back to their original character in the first place. There are different definitions of "stealing" in virtual worlds, the most prevalent of which are:
1) In-context steal. A character obeying the rules of the virtual world appropriates an object that was in the possession of some other character in that virtual world. If a court were to insist that this "theft" were returned, it would have a devastating effect on virtual world design.
2) In-context but buggy steal. The character operated within the apparent rules of the virtual world but did something that the designers hadn't foreseen and that now they have seen it they regard as a bug. Here, we have a genuine dispute that may result in compensation for one party or the other. If a court were to be brought in to rule on such an incident, it would effectively be overriding the authority that VW operators currently have to deal with these situations independently (ie. it would stomp on the EULA). This would make operating a VW difficult for the developers, and mean many more disputes would go to court.
3) Fraud. A player gets another player to hand them something for a promised reward (real-world or otherwise) that never materialises. Here, the VW operator would probably not want to intervene, believing it to be the player's fault they were suckered. This is especially likely to be true if the fraud took place concerning an activity of which the developers didn't approve (eg. eBaying). I don't think they'd necessarily care if a court ordered them to transfer "ownership" back, so long as the onus wasn't put on them to resolve these disputes themselves in future.
4) Out-of-context buggy steals. Here, some player exploits a security loophole to gain the ability to change the database in ways they're not supposed to, eg. they hack an account. I'd expect the VW operator to jump on the hacker and undo all their misdeeds if this were discovered, as it's relatively easy to show that the design of the VW wasn't "intended" to allow such things to happen, and the cracker can't claim to be playing by the rules. The developers could probably take the hacker to court in the real world over it if they felt so inclined, too.
5) Real-world theft. Someone breaks into your house, steals your computer, logs into it and transfers a bunch of things from one character to another. This is like 4), but the involvement of real-world courts is inevitable as there was real-world property theft. The courts could reasonably ask that everything done while the computer was stolen be undone as much as possible. I don't think they could reasonably ask that the virtual world developer offer any additional compensation to the victim, though.
The the situation we have in the Chinese case cited is 4), except that for some reason the VW operators declined to rectify the problem until their arm was twisted by the court. Maybe there are 20,000 other people with the same problem and they felt they couldn't address everyone's issues? If so, they're in for a nasty surprise...
Richard
Posted by: Richard Bartle | Dec 20, 2003 at 09:00
That poor gamer is *so* banned.
At least he is if the Chinese game company has a TOS (and attitude towards its customers) anything like Western game companies.
Posted by: Bryan Allman | Dec 20, 2003 at 15:51
Bryan Allman >> That poor gamer is *so* banned.
See, thats where it gets realy interesting. Now that the courts have decided that he is entitled to this virtual arms stockpile, what happens when he is banned? Can he go to court and demand he can play?
Of course, IANAL and all that, but it seems to me that this chinese jurist just made it a much pricklier situation to ban an active player. After all, the ban does the same thing the hacker did - takes away his hard earned virtual property. And the case didn't seem to be player vs hacker - it was player vs developer. I can't see that the law would make much of a distinction between the two situations.
bbc
Posted by: bbc | Dec 20, 2003 at 17:51
BBC,
While I'm sure the game's contract stated that nothing in the game has any worth, "all your base are belong to us", and there is no restitution of anything, the chinese court is saying -in spririt- that the game has to be played by the game rules, or "the social contract", and not by the letter of the contract.
There are also some marked cultural differences in the way chinese businessmen view contracts (when compared to US), which should be taken into consideration when trying to "translate" these events.
Posted by: DivineShadow | Dec 20, 2003 at 20:26
Well what about retribution for all the monthly payments that have been made? That's what I don't quite see, how does one go about banning and/or suspending someone's account without being liable for a partial or total refund during the unplayable time?
If you get banned from a game for only violating the games rules (such as exploiting) then is it really legal to be able to ban someone permanently without refunding their money for the game and all the past monthly payments put in?
Posted by: Lee Delarm | Dec 21, 2003 at 13:13
Seems to me to be more of a tort case than a property case. That is, it recognizes a developer's duty to the player to protect server integrity more than it recognizes a property right in items. (At least as far as I can tell from the news reports).
The case would be more interesting if the court awarded monetary damages measured as the market value of the items lost in spite of an EULA provision prohibiting the sale of items (as opposed to damages measured as the value of time invested based on subscription price).
However, at some point, the fraternal order of items-are-not-property are going to have to reconcile their charter with the fact that developers intend and encourage their players to treat items as property. Items are property and it's not because of anything so sexy as IP law: it's simple, basic property law a la Johnson v. M'Intosh, Pierson v. Post, Ghen v. Rich, Armory v. Delamirie and all the other Property I classics.
Jeff Cole
Posted by: Jeff Cole | Dec 23, 2003 at 09:23
Jeff Cole>However, at some point, the fraternal order of items-are-not-property are going to have to reconcile their charter with the fact that developers intend and encourage their players to treat items as property.
But what about the developers who don't encourage their players to think of object this way? Or ones who encourage players to think of objects this way in the virtual world but not the real one? Will we see a one-size-fits-all precedent? Whichever way it went, it would undermine several major, extant virtual worlds.
Richard
Posted by: Richard Bartle | Dec 23, 2003 at 10:59
Richard: "But what about the developers who don't encourage their players to think of object this way? Or ones who encourage players to think of objects this way in the virtual world but not the real one? Will we see a one-size-fits-all precedent? Whichever way it went, it would undermine several major, extant virtual worlds."
I think the issue is not whether items are property, but rather who owns them.
In America at least, I predict that items are property and the relationship with respect to items between developer/player is a bailment with the developer as bailor and the player as bailee.
Imagine I owned a paintball range on which I hid paintball guns with longer ranges and pieces of "armor" that if hit would not count as a hit (or as half a hit, whatever). I send the teams out minimally equipped and let them go at it. If they kill another player, they can take his items.
Only the most unreasonable player would assume that I intended they actually "own" the items they find and take off of there victims. Especially if part of the waiver (EULA) you signed acknowledged such a fact.
Everyone understands that I intend to retain ownership of the guns/aromr and am only letting them use the items to have fun.
While the players obtain possession, they never obtain title.
And, developers can contractually manage the bailment to suit there needs.
Jeff Cole
Posted by: Jeff Cole | Dec 23, 2003 at 11:17
Good God. "Their," not "there."
Posted by: Jeff Cole | Dec 23, 2003 at 11:18
Well put, Jeff. On the other hand, worlds like Second Life clearly are setup to allow users to retain rights to their creations (users who brought their own paintball guns to your example) and are hurt by attempts to retain IP rights to all creations.
Cory
Posted by: Cory Ondrejka | Dec 23, 2003 at 11:53
Cory,
Why do you need "the other hand"? "My analysis in no way prevents you from administrating your game (and the associated property rights) as you see fit. It merely provides a default rule that items are property bailed to the player. Contract aroundd it however you want.
I also don't address IP rights and specifically argue that items-as-property doesn't per se implicate IP at all and can be effectively analyzed under personal property law.*
Sure, there may may some IP independent of the item's status as property (e.g. texture IP or scripting IP). But I am pretty sure EULA's are not effective as an exclusive license to developers of player IP (though it is likely sufficient to assign a non-exclusive license).
Indeed, it is a testament to the immersive quality of these worlds that this issue seems so thorny.
Jeff Cole
* Welcome to Law-Dev circa late 18th Century. ;-P
Posted by: Jeff Cole | Dec 23, 2003 at 12:18
Jeff,
I may be misunderstanding bailment (I'm not a lawyer and my knowledge of Law-Dev circa the 18th Century is pretty limited) but my quick lookup said that bailment is defined as "the delivery of an asset by its owner to another person or persons for temporary care."
That sounds like a good working definition of conventional MMORPGs, where the developers create the content and then the players consume it, but doesn't seem like as good a fit for Second Life. Since SL allows users to build from scratch (I've been using the term atomistic construction for this and I talk about it at length in the paper) they are not just receiving content from the developers. Instead, they are truly creating new and interesting works so allowing users to retain IP rights to their creations is an important step.
Also, you may have misunderstood my intent since you say:
"But I am pretty sure EULA's are not effective as an exclusive license to developers of player IP (though it is likely sufficient to assign a non-exclusive license)."
Forced granting to developers of player IP is exactly what SL chose to move away from.
You use the term own. I would defintely enjoy learning more about the legal definition of own, since the virtual space seems to embody IP rights more than ownership. If I post a book to my website, do I need to care if my web host owns the bits representing the book so long as I retain all of my IP rights to it and didn't grant any to the web host?
Also, to my non-legal eye, Dan and Greg's paper on virtual property seemed to be solidly argued and that virtual property and IP rights are therefore deeply linked in cases where the user is creating it.
Cory
Posted by: Cory Ondrejka | Dec 23, 2003 at 12:41
A bailment is essentially the bailor giving to the bailee the possession of, but not the title to, property for a given purpose and with the understanding that the property will be returned to the bailor.
So, the developer gives to players items for the purpose of playing the game with the understanding that when the player is done, the items will be returned to the developer.
"Playing the game" may well include obtaining items from both NPCs or PCs. It is well established in the law that a person cannot transfer more rights in property than he himself has. So, PC's trading items are only trading possession and not title. The developer is the "true owner" and retains rights as against subsequent possessors.
By "own" I mean "have title" as opposed to merely "have possession."
I come at it from Demsetz's position that property rights are "an instrument of society and derive their significance from the fact that they help a man form those expectations which he can reasonably hold in his dealings with others. ... An owner expects the community to prevent other from interfering with his actions, provided that these actions are not prohibited in the specifications of his rights." (Emphasis, consider EULAs).
The term "property" describes, but does not confer, a set of rights. That is, players' expectations don't hinge on whether the item is property; rather, whether an item is property hinges on players' expectations.
There isn't a single player that creates a character and expects not to exercise complete dominion (including alienability) over his training sword.
Unfortunately, the "I" in "IP" has become synonymous with "intangible." IP imposes artificial scarcity on a resource that is inherently non-rivalrous. Virtual items may be intangible, but they are also rivalrous.
I have not played SL, so am not familiar with your handing of IP. But the fact that SL owns an instantiation of an item embodying player IP is not contradictory. The fact that SL owns the item does not imply that they own the IP anymore than the fact that the player owns the IP implies they own the instantiation.
A music label owns the the copyright in the particular performance embodied on a CD; but that ownership does not, by itself, give the label ownership in the copyright in the composition embodied in the performance. The subsequent sale of a CD transfers only title in the CD and not in either copyright interest.
At any rate, you can probably file that under TMI.
Jeff Cole
Posted by: Jeff Cole | Dec 23, 2003 at 13:59
Jeff,
Thanks for the explanation! It's good to see that I understood things correctly!
It's too bad that you haven't read my State of Play paper nor investigated SL further because the profound ways in which SL differs from conventional MMORPGs are clearly hard to explain.
For the record, I don't believe in too much information. The whole reason I'm spending time on TN is because of the unique opportunity to learn about new areas.
Cory
Posted by: Cory Ondrejka | Dec 23, 2003 at 15:11
FYI, for those who don't read Spanish, the trackback link at the top of this page leads to the wackiest version of this story yet, published as fact by TV Espanola's newswire division:
"Li Hongchen, a Chinese youth of 24, has won a case against online game company Arctic Ice, which must now return his stolen arsenal of virtual biological weapons.... [Li's] supremacy in the game was such that he was spoiling the excitement [for everybody else], and therefore Arctic Ice hired the "hacker" Shuiliu0011 to steal his virtual arsenal of biological weapons [!]."
No reason to think it's true, but if anyone out there is planning to re-run the case in a moot court, this definitely makes for a juicier fact pattern, no?
Posted by: Julian Dibbell | Dec 23, 2003 at 16:51
Jeff> Unfortunately, the "I" in "IP" has become synonymous with "intangible." IP imposes artificial scarcity on a resource that is inherently non-rivalrous. Virtual items may be intangible, but they are also rivalrous.
Why do you think that virtual items are rivalrous? Sure, object 0A3D23 is uniquely different from object 0A3D23, but the rivalry is inherently artificial isn't it? (Not that this affects the other parts of your observations. I'm just curious)
Posted by: Dan Hunter | Dec 23, 2003 at 20:28
Dan,
It's rivalrous because if I give you my Sword of Banishing I can no longer use it. Nor can anyone else (until you or DevGod gives it tom someone else). As a "thing" it is rivalrous (same as any real-world item).
I don't see how the rivalry is meaningfully artificial.
Jeff Cole
Posted by: Jeff Cole | Dec 23, 2003 at 22:24
[enters waving IANAL flag]
What is the legal definition of "rivalrous"?
Cory
Posted by: Cory Ondrejka | Dec 24, 2003 at 00:34
Something (item or resource) is rivalrous if its use or consumption diminishes or depletes its availability to others.
Sometimes easier to consider what is non-rivalrous: hearing a song on the radio, watching a movie or fireworks display, sharing an idea. My consumption of (by watching) the movie doesn't diminish its availability to you.
However, the copy of the DVD is rivalrous.
Closely related to excludability. "Public goods" as usually defined as non-excludable and non-rivalrous.
Jeff Cole
Posted by: Jeff Cole | Dec 24, 2003 at 01:19
Dan Hunter > Why do you think that virtual items are rivalrous? Sure, object 0A3D23 is uniquely different from object 0A3D23, but the rivalry is inherently artificial isn't it?
Jeff Cole> I don't see how the rivalry is meaningfully artificial.
What I take Dan to mean is that the rivalry is ‘meaningfully artificial’ because while indeed control of any given instance (token) of virtual object tends to be exclusive this has no impact on the type as there is effectively an infinite supply of Swords. In sort (within a given VW) Tokens are rivalrous Types are not.
Though this does bear some further analysis as when we get into virtual objects its not so clear what it is we are referring to and what it is we mean by rivalry.
Jeff, when you refer to your Sword of Banishing what is it that you are really referring too ? This Sword is a certain type within a system with a set of properties. One of those properties is that there can be many tokens of it. How much does it make sense to talk about rivalry of tokens and not types when it comes to virtual objects ?
Or to put it another way – I have a cup on my desk (a science museum Periodic Table cup as it happens (science geek or what!)). This is a token of a type of cup. It is rivalrous in a strong sense: it can’t hold my coffee and your tea at the same time, it can’t be held in the very same way by more than one person at the very same time.
But when it comes to this Sword, what is underneath is a set of code that describes the properties of the sword, including things like its visual appearance – this very same code (well ok a version per-shard probably) is used for every instance of the sword (OK each instance will probably be made up of some kind of pointer with a unique memory location back to the common code - urrg we could get into the client server split here and the db bit verses teh run time memory bits, it could get ugly). So I’m not sure it’s meaningful to say that a virtual Sword is rivalrus in the strong sense as the type is non-rivalrus in almost the same way a song is (and yes we get the same issue with MP3s).
A counter argument is: but right now I can use my Sword to do X and you can’t use it. Indeed, but I can use the very same type of Sword to do X. So where is the difference? Similarly if there was only one such sword in the whole system then this would be very ‘artificial’ to use Dan’s term.
I think that I grant that Swords are indeed rivalrus. As even if there were only one token of a sword and all members of a virtual world had access to it, control of it would ultimately come down a sequence of individual controls (instructions) possibly queued a the level of individual CPU time slices.
However I deny that this rivalry has exactly same meaning or value as it does for physical objects. Indeed here I think we have an example where we are bending an existing legal term almost completly out of shape and meaning.
Ren
www.renreynolds.com
Posted by: Ren | Dec 24, 2003 at 07:46
Oh, something that is kind of the other way round in VW’s is names (here I am really referring to avatar names). One feature of VWs is that they are name spaces – like the internet unique IP numbers, URLs, SMTP addresses. Name spaces by design cannot have two instances of the same name.
So the value and meaning of a name in a VW is somewhat different from the physical world.
Posted by: Ren | Dec 24, 2003 at 07:53
Ren,
The rivalrous nature of your cup is no stronger.
You are making the very mistake I that I rue: you are too focused on tangibility. You are equating the sword with its underlying IP.
Need me to explain the rivarousness? At the moments I am swinging my sword, on the processor that is executing the code, no other sword is (or indeed other code) is executing.
But you don't really need to put such fine a point on it. The macro-effect is still rivalrous in the same sense as your cup. We just understand the "code" behind sword; but we don't understand the code behind your cup (for all we know there may be infinite instances of your cup).
Sure, the computer is very efficient at cranking out swords and could create an infinite number of them. But so could your cup-maker. And it could do so much more efficiently than the person who handcrafts the same cups. However, certainly the rivalrousness nature doesn't turn on effeciency of production.
That said, efficiency of re-production can seem to implicate rivalrousness. To the extent that you can clone something at almost-zero or zero cost, you take the decision of production out of the control of the producer.
MP3's are a perfect example. The MP3 on your drive is rivalrous. If you fileshare and allow others to copy the MP3 from your drive you are creating other copies, each rivalrous in their own right. Even in that instance, though, it's not the individual MP3s that seems to become non-rivalrous, but the performance embodied in those MP3s.
But that would also be true of your cup if I could reproduce it at zero cost (assuming a similar distribution network for pirated cups). It's not unlike dividing by zero. Let me do that and I can prove to you that 1 = 2 (for other than extremely large values of 1).
You haven't changed (let alone dispelled) the rivalrousness of the MP3s. You just changed the market conditions: you now have an effectively infinite supply with no marginal cost, so the only people who won't have an MP3 (or cup, or whatever) are those for whom the cost of accessing the network exceeds their reserve price.
You might be tempted to call it non-rivalrous, but to do so neuters the concept. Better to consider it a failure of the market, the law, or both.
Ren: "Indeed here I think we have an example where we are bending an existing legal term almost completly out of shape and meaning."
I agree. You do bend the concept almost completely out of shape and meaning. =P
Merry Christmas Eve and Happy Holidays!
Jeff Cole
Posted by: Jeff Cole | Dec 24, 2003 at 08:56
Jeff,
I agreed that tokens of virtual swords are rivalrous, that is they have certain property and that exclusive to a given user at a given time. I suppose I read into the term rivalrous a set of implicit values that are based on certain properties of physical objects such as their cost of production, the fact that no replica is ever exactly the same etc. Virtual objects lack some of these properties – so when we are saying that a virtual object is rivalrous the implications of this are not quite the same. Now as rivalrous is a term used in law I feel we have to be careful about unpacking the motiviatoins about why we care if something is or is not rivalrous and what rights and duties we might want to ascribe to such things, as even though the term may apply to tokens of virtual swords I don’t want to jump to the conclusion that the same rights do – hence my attempting to tease out strong and weak notions of the term and foregrounding issues such as cost of production, identical replication etc.
Merry midwinter
Ren
www.renreynolds.com
Posted by: Ren | Dec 24, 2003 at 09:12
Let me just make very obvious observations, because people seem to be talking past each other.
1. Rivalrous goods are those which are depleted by use or which only a finite number of people can use at a given time. It is not a legal term of art, but it is a term used often in law and economics.
2. A VW artifact (with the exception of, e.g., a VW viral contagion artifact) is generally a rivalrous good within the game context.
3. Outside the game context, a VW artifact is not even a "good" -- in fact, it practically doesn't even exist outside the game, except as an IP right which is not contingent on any given instantiation except the one which gave rise to the right.
4. The typical VW artifact is "artifically" rivalrous, unlike physical objects, insofar as the designers have chosen to "artificially" make the VW artifact incapable of being shared and incapable of being depleted. Again, this is true only *within* the game context -- it is certain not true at the software level and is not true outside the game context, where traditional IP (& contract) control the ability to reproduce VW artifacts.
5. At some level, one could argue (as many have) that IP is "artificially" rivalrous, because *the law* "artificially" prevents certain pictures, symbols, and inventions from being recreated -- despite the fact that allowing those pictures, symbols, and inventions to be recreated would not deplete the original owner's use or deprive the original owner of use.
6. On the other hand, because the thing "owned" by an IP owner is a right to prevent infringing activity by others, I think you could make an argument that IP *is* rivalrous simply *by definition.*
Posted by: Greg Lastowka | Dec 24, 2003 at 09:44
Greg> 6. On the other hand, because the thing "owned" by an IP owner is a right to prevent infringing activity by others, I think you could make an argument that IP *is* rivalrous simply *by definition.*
Hmm... As I understood it, "Rival" is the economic concept (admittedly borne in a more innocent time when the virtual wasn't real) defining a good where the marginal cost of production of good n+1 is zero. Once the Sword of Ultimate Vanquishing is defined, one can have millions of SoUV at zero cost (at least in a Coasean world, which we're pretty close to in the virtual). Hence, I can't understand Jeff's analysis of virtual assets as rival. Neither can I see (except in the most inclusive of definitions of "rival") that one can create a rival good by creating artificial legal regimes that make it rival. At least, that's what the economic critique of some aspects of IP would suggest.
And I don't understand at all the idea that "tokens of virtual assets" are rivalrous. This essentially strips the concept of rival of all meaning. Sure, tokens are unique but that doesn't make them rival.
Posted by: Dan Hunter | Dec 24, 2003 at 10:24
There are different definitions of rivalrous goods. Dictionaries don't help much, since they define rivalrous as simply "competitive." Turning to Google, as I'm always wont to do, I can offer this to explain my understanding:
"Rivalrous goods are those that if one person consumes, another cannot, because one person's use means another will not be able to use the good."
www.public.iastate.edu/~potoski/POLS485/Fin07%20.doc
(from Matthew Potoski, Assistant Prof. of Poli Sci at Iowa State.)
So I guess the primary question is whether an economic conception of rivalry would/should recognize the practical effects of legal regimes / game rules on the (practical) availability of "goods." I understand the reason for not doing that in law & econ analysis of IP -- the law is theoretically mutable, so legal theory should generally not defer to the artificial distinctions it creates.
If you were to say "well, game design is mutable too" -- yes it is, but not in the same way. I think there is a stronger argument for recognizing the importance of game rules in determining whether VW artifacts are rivalrous, because the very existence of the artifact is embedded in the game design, without which the property doesn't exist (except as IP).
The value, utility, usefulness, social recognition of any "good" in a gamespace is highly contigent upon the regulations of the game. E.g., if you and I are playing basketball and you steal the ball from me, despite my ownership of another basketball which is sitting on the bench, I'm going to want to steal the "game important" basketball back from you, not the equally available and physically fungible one on the bench. (And note how the legal ownership of the ball is irrelevant to the idea of "stealing" on the court.)
So, as may be apparent from all that, I think debating what "rivalrous" means empirically in relation to VW artifacts is largely a distraction -- or more accurately, a restatement of standard IP debates.
Posted by: Greg Lastowka | Dec 24, 2003 at 11:18
Greg> There are different definitions of rivalrous goods
Tricksy lawyerses.
Posted by: Smeagol | Dec 24, 2003 at 12:54
Greg is right to point out that there are various definitions for 'non-rivalrous', but I think the spirit of these definitions is consistent -- namely, that an arbitrary number of people can enjoy such a good, irrespective of their wealth. (Ironically, I sent a post to the mud-dev list a few days ago where I defined 'non-rivalrous' in a way subtly different from all the foregoing definitions. Opinions, it would seem, are themselves non-rivalrous. But then -- were they ever really goods?)
Greg Lastowka> So, as may be apparent from all that, I think debating what "rivalrous" means empirically in relation to VW artifacts is largely a distraction -- or more accurately, a restatement of standard IP debates.
I don't think it's a distraction at all; in fact, I think it calls attention to some crucial properties of MMORPG, particularly as regards their intersection with economics.
Game content is the template from which game items are created. Though content certainly has a non-zero production cost, an existing unit of content can be instantiated in the game world (and thus consumed by players) almost freely; in this sense is it non-rivalrous. Game items are rivalrous because games make them so, and this is definitive because -- as Greg pointed out -- the game rules that prevent items from being shared or reproduced arbitrarily are part of the same rule set that define the items themselves; discount those rules and you discount the items' very existence. So game items are rivalrous, but the content they manifest is not.
There is an implicit tension here, between players who (like consumers) always want 'more', and game designers who (for various reasons, selfish and otherwise) don't want to give it to them. Designers impose artificial scarcity in the game world through the mechanism of it's rules, and this, ultimately, it what makes it a game. Consider: is winning non-rivalrous? Because that, in some sense, is what game items are: the MMORPG world's metric for game success. If the client simply announced 'you win' every time a player logged in, would they actually have won anything?
By way of analogy, I will also point out that the power that game items confer is very much rivalrous; if one player starts with the best equipment, but (do to some bug or exploit) finds their gear has been duped a million times, then they no longer have the 'best' equipment; they have merely the 'average' equipment. I'd like to see a judge fix that one.
As it happens, I'm working on a paper that addresses such issues from a more-or-less economic standpoint. For those who might be interested, an excerpt is available here http://www.anthemion.org/PlayTimeExcerpt.doc or http://www.anthemion.org/PlayTimeExcerpt.htm. I'd be glad to hear your thoughts on my work.
P.S. Greg Lastowka> Outside the game context, a VW artifact is not even a "good"
I have to disagree with you here. The term 'good' is typically defined as 'anything people want more of'. I think game items definitely fit this definition.
Posted by: Jeremy Neal Kelly | Dec 24, 2003 at 14:18
Jeremy,
Thanks for the thoughts and I'll take a look at your paper. On the p.s. point, all I was saying was that, as tangible material devoid of hardware mediation, VW artifacts do not exist in any meaningful way.
We actually have a Economics PhD in-house on this collablog who you're probably familiar with, but I think he's occupied with eggnog at the moment... and I'm going on hiatus now as well.
Happy Holidays, TN!
Posted by: Greg Lastowka | Dec 24, 2003 at 14:41
Smeag: too funny!
Note: I had been using "rivalrous" and "rivalrousness" to underscore context. I am going to stop, but I still intend meanings for "rival" and "rivalry" narrower than those in Websters.
Greg: we are indeed talking past each other as well as cross-assuming.
Dan,
Let me be more explicit. I am using rivalry in this sense: if I write a song and teach you the song, your teaching to others or playing the song does not diminish my ability to teach to others or play the song. Or, if you and I attend a fireworks display, my consumption of the performance does not diminish your ability to consume (though my attendance might diminish your enjoyment of your consumption).
In the case of my song. There is a marginal cost of production when I teach it to you as the n-plus-one'th person to consume it: the cost of our time. There is a marginal cost for me to provide it to you. However, there is no marginal cost associated with you playing for or teaching to others the song.
I am not familiar with a production-centric definition (I do not doubt you: I am not a trained economist). However, I have seen goods for which the marginal cost of use is zero called non-rival.
I we include the concept of exclusion (I have not, yet), then: an excludable and rival good is a private good. An excludable and non-rival good is a "club" good (e.g. a concert performance in a club-- the club can exclude you from the non-rival performance). A non-excludable and rival good is a an open access good or resource (e.g., air). A non-excludable and non-rival good is a public good.
I think that's a pretty solid economic interpretation, but I could be wrong.
So, in Everquest, the Sword of Banishing is a private good: it is excludable (my using one means you cannot use that one) and rival (the cost of your use is not zero--either you camp it or buy it). Consider the uproar that arises from server's guilds (or loose afilliations of guilds and unguildeds) when a raiding guild, frustrated with the rival nature of items, declares a server FCFS. Sounds like rivalry/excludability to me.
I am analyzing the problem under the games as implemented. A developer could indeed implement a non-rival sword (though one wonders why it would).
One of my main points is: why can't the law treat such items in the manner in which the developer implements them and subject to the EULA's? Determine the default rules and contract to taste around them.
What problems can such a system not address?
What normative problems could one possibly have with that?
I will work on this more later. Am off to partake of relativity.
Jeff Cole
Posted by: Jeff Cole | Dec 24, 2003 at 15:47
Rereading this thread and comments reinforced how much of an outlier in the online game space Second Life is, with almost every other product tightly clustered around the MMORPG model defined by Meridian and Ultima (even There.com, despite much public positioning, lies far to the "Developers create the content and users use/modify the content"-side of the space) . Clearly, if you haven't actually created something within SL (or watched someone do it), it is very difficult to wrap your mind around the potential and power of atomistic creation. It is much closer to creating something in Photoshop (or writing an application to run under someone else's OS) than it is to MMORPG crafting.
So, my question to Jeff et al is, given content created within the online space by the users, why should a EULA be the legal instrument of choice rather than real world IP laws? And, if a EULA is chosen, why true creation within an online space should be treated differently than creation within Photoshop, MS Word, or development under Windows?
Happy holidays, everyone!
Cory
Posted by: Cory Ondrejka | Dec 24, 2003 at 21:36
Actually, having though about it a bit, I am not at all sure that a developer can implement a non-rival sword. Can non-rivalry even be enforced? Can something that must be "possessed" (physically or virtually) to be used/consumed be non-rival?
Cory: I am talking about the default rules. I am not talking about IP other than to say that it doesn't apply to the meat-and-potato property analysis. IP law still applies. SL needs a license to use player IP: that is copyright law. To get that, you need to contract for the rights. Copyright (probably) doesn't give you the right to use player IP without the license.
I am arguing the basic property issue of developer-instantiated items.
Jeff Cole
Posted by: Jeff Cole | Dec 26, 2003 at 17:09
Jeff Cole> ...I am not at all sure that a developer can implement a non-rival sword. Can non-rivalry even be enforced? Can something that must be "possessed" (physically or virtually) to be used/consumed be non-rival?
Certainly. Imagine a game that gives every player a newbie sword, and automatically replaces it whenever it is lost. (I've seen games that approach this, but as I recall, they required newbies to visit some special location to have gear replaced; this introduces a measure of cost, which I suppose makes the items technically non-rivalrous.) Game items are necessarily fungible, therefore newbie swords form a non-rivalrous good, one that can also be 'possessed'. Again, this demonstrates the fact that content (in this case, the code that defines all newbie swords) is naturally non-rivalrous; it can be experienced and thus enjoyed by an arbitrary number of players. Designers introduce rivalry for various reasons, among them, to make a game of what would otherwise be a free-for-all. However, item rivalry and the scarcity that results are -- in all cases -- wholly artificial. As is, really, the entire notion of 'possessing' a game item.
Posted by: Jeremy Neal Kelly | Dec 26, 2003 at 19:01
Jeremy,
That's not non-rivalry. It's more accurately a non-excludable, rival good.
The use of a single sword prevents another from using the same sword. Sure, and infinite number of private goods might result in quasi-non-rivalry, but it isn't the same.
A concert is non-rival because many people can use/consume the same event/good.
It's not that content can be enjoyed by an arbitrary number, it's whether the marginal cost of consumption/use of the sme good is zero.
Sure, you can force the same result by giiving every character created an unalienable item. But that doesn't make the item rival. It may be the same effect, but the implications aren't the same. The developer could simply not "produce" as many swords. There isn't the same solution for a rival good.
Quests are non-rival as usually implemented (though you could certainly implmenet rival quests), though they may be excludable (based on race/faction/class/etc.)
Jeff Cole
Posted by: Jeff Cole | Dec 26, 2003 at 19:28
Ooops ... I meant, "... same soltution for a non-rival good" in the next to last sentence.
Posted by: Jeff Cole | Dec 26, 2003 at 22:38
Wow. Typo city, my apologies. And I hadn't even had a single beer.
Substantively: I also meant "But that doesn't make the item non-rival" in second to last paragraph.
Posted by: Jeff Cole | Dec 26, 2003 at 23:22
Jeff Cole> The use of a single sword prevents another from using the same sword.
What do you mean by “same sword”? When this question (in various forms) has been put to you before, you’ve referenced the physical media (CPU queue, memory) that store and instantiate content. But surely the game’s implementation is (or should be) irrelevant to this discussion? Consider the difference between a single sword and a stack of twenty swords. Where in that stack is the ‘original’ sword? The logical representation of those two numbers (00000001 and 00010100) don’t even share any bits! If the CPU writes a ‘1’ to some location in memory, is it guaranteed to get the ‘same’ ‘1’ back when it reads it? This notion of ‘sameness’ is defined in the real world because physical objects manifest persistence over time, but it is not defined in the context of a computer program, where functional equivalence is the rule.
I think Ren Reynolds’ post of December 24, 2003 04:46 AM has done the best job so far of summarizing this discussion, though I see no need to discern between ‘strong’ and ‘weak’ notions of rivalrousness; the standard definitions (all of them) seem to work just fine in the game world.
Perhaps Greg was right to call this language debate a distraction. I think the important truth is that scarcity of items in the game world is entirely artificial. I assume we all agree on that?
Posted by: Jeremy Neal Kelly | Dec 26, 2003 at 23:47
Jeremy,
You confuse the concept of rivalry. Like Dan, et. al., you mistakenly focus on marginal cost of production; and not on marginal cost of use/consumption.
We can agree a Ford pickup truck is rival and excludable. So, if Ford decided to produce a (perfectly fungible) Ford for every person in America (or, the World), then all of a sudden, the Ford pickup truck wouldn't be rival?
Consider, a radio or television broadcast, a fireworks display, or a showing of a movie ... each a classic example of non-rival goods. They're not non-rival because everyone who consumes such a good consumes a different peformance--they're all consuming the same performance.
The reason we have IP is to enforce a property right in non-rival goods. The law has to artificially impose that property right because the fact that the marginal cost of use of the same IP is zero inheres in the IP and does not result from a production decision. Non-rivalry (and non-excludability) invite market failure.
In your example of infinite swords, the only thing the developer needs to do to create a market is to stop giving every character an unalienable sword. The same solution doesn't apply for non-rival goods: you need to impose an IP property right.
You want the implementation to be irrelevant, yet your example expressly depends upon implementation (mine does not). Per your argument swords are only non-rival to the extent that the developer gives every character created an unalienable sword. If the developer doesn't, then the swords are unalienable? So, if MAX_SWORDS_PER_SERVER = 5 then those five swords are non-rival too (merely because the developer could--but doesn't--give everyone an unalienable sword)?
If you have an infinite supply (say a kiosk on which you click to receive a sword), then you have a rival, non-excludable good (a classic "resource"). So, in my argument, rival is independent of implementation: whether you have 5 or an infinite number swords, they are rival--regardless of whether 4 others or every person has one too. They're all rival.
You confuse my sword and the sword-archtype.
Jeff Cole
Posted by: Jeff Cole | Dec 27, 2003 at 00:53
I think this part of the thread is starting to repeat it itself, and my interest in it is largely exhausted. However:
Jeff Cole> You confuse the concept of rivalry. Like Dan, et. al., you mistakenly focus on marginal cost of production; and not on marginal cost of use/consumption.
As Greg pointed out (and as I reiterated), there are varying definitions of goods rivalry. I have not actually been using the ‘production cost’ definition, but even if I had been, it could hardly be called a ‘mistake’.
Jeff Cole> We can agree a Ford pickup truck is rival and excludable. So, if Ford decided to produce a (perfectly fungible) Ford for every person in America (or, the World), then all of a sudden, the Ford pickup truck wouldn't be rival?
You’re farming straw men. My example posited a good that could be freely created by players. A one-time nationwide truck endowment (God help us!) obviously doesn’t fit this description.
Jeff Cole> In your example of infinite swords, the only thing the developer needs to do to create a market is to stop giving every character an unalienable sword. The same solution doesn't apply for non-rival goods: you need to impose an IP property right.
According to the Oxford English Dictionary, ‘unalienable’ refers to “that cannot be alienated or transferred from its present ownership or relation.” I invoked the concept of ubiquity, not inalienability.
Aside from that, this is exactly the point and I (and I think Dan Hunter and Ren Reynolds) have been trying to convince you of. In the real world, laws enforce rivalry where it does not exist. In the game world, rules enforce such rivalry. Game rules have the character of real world laws (they are constructed or ‘artificial’) but, because they also define the ontology of the game world, they have the essential force of natural laws.
Jeff Cole> If you have an infinite supply (say a kiosk on which you click to receive a sword), then you have a rival, non-excludable good (a classic "resource").
If the supply is unlimited and freely available, then how can the good be anything but non-rivalrous? Such an example fits every definition of non-rivalrousness I have ever seen.
Most importantly, you haven’t answered my question about your definition of ‘sameness’, which seems to underlie your conception of rivalry in the game context.
Ultimately, I think this discussion suggests a new ontology of game items. Specifically, there are no ‘items’ in this space (and how could there be?); there are only rights or permissions to use certain elements of content. A player with one ‘item’ has access to the content behind that item. A player with several of the same item has the ability to access that content, or to transfer access to several other players. (I will forego the obvious Matrix quote.)
Posted by: Jeremy Neal Kelly | Dec 27, 2003 at 14:08
I, too, am tired of this exchange. And offer this as my last:
Jeremy:"If the supply is unlimited and freely available, then how can the good be anything but non-rivalrous? Such an example fits every definition of non-rivalrousness I have ever seen."
Then you need to see more definitions. Each and every "good" is rival and excludable in and of itself.
Ubiquity does not imply non-rivalry. Non-rivalry might lead to ubiquity, but ubiquity does not, by itself, imply non-rivalry. Not even close.
Show me one definition of rivalry (in the economic sense) that invokes production/supply and not use. There is no such definition. Rivalry as an economic concept has to do with use.
Indeed the very problem presented by non-rival "things" is that such "things" might be very expensive to initially produce (e.g. R/D assoiated with a patentable idea) and the non-rival nature of the "thing" makes it difficult or impossible for the producer to recover those costs.
Jeremy: "In the real world, laws enforce rivalry where it does not exist."
Wrong. The law enforces exclusion. The IP remains non-rival. The law provides me a cause of action against you if avail yourself of my non-rival "property" without my license. It is that cause of action that imposes on you a cost for use of my non-rival "property." Exclusion and rivalry are different concept entirely.
The goods embodying the non-rival IP (say, a CD, DVD, or patented invention) are rival and excludable. The IP embodied is still non-rival.
As for my definition of "same," I mean it in the regular sense.
Consider some classic definitions of non-rival goods: a television or radio broadcast, a fireworks display, a concert. In such cases, all users/consumers are using/consuming the same performance and not separate distinct performances.
Consider another classic: an idea. If I have an idea nad tell Greg my idea, his use of the idea does not diminish my ability to use the idea. If I give Greg a screwdriver, then my ability to use that screwdriver is diminished.
Consider a freeway. Off-peak, it is non-rival: my use doesn't diminish other's use. During rush hour, though, it becomes rival: my use imposes a cost for use on others (the additional time spent on the freeway).
Jeff Cole
Posted by: Jeff Cole | Dec 27, 2003 at 15:07
One last, hopefully illustrative comment:
Jeremy: "I think the important truth is that scarcity of items in the game world is entirely artificial."
It is precisely the fact that items are both rival and excludable that allows for such scarcity.
Jeff
Posted by: Jeff Cole | Dec 27, 2003 at 16:10
Seasons's Greeting to all!
I surfed internet and found some useful information written in Chinese characters on this case, so I'd like to share it with you.
-A chinese news that explicates this case.
http://china.sina.com.tw/games/newgames/2003/12/121910332.shtml
- Interview with scholars and lawyers in Beijing on this case.
http://61.129.65.8:82/gate/big5/game.eastday.com/epublish/gb/paper370/20031117/class037000002/hwz1333823.htm
- Feature news that analyzes this case.
http://china.sina.com.tw/games/newgames/2003/12/121910333.shtml
Posted by: Unggi Yoon | Dec 30, 2003 at 20:06
Thanks, Unggi!!!
Didn't know where to put this, so I'll put it here -- FT article on Red Moon, which quotes Ted and Beth Noveck:
http://news.ft.com/servlet/ContentServer?pagename=FT.com/StoryFT/FullStory&c=StoryFT&cid=1071251800683
Posted by: Greg Lastowka | Dec 31, 2003 at 13:43
So I have a friend who had his account hacked and lost 300 million simoleans on tso. I guess that would be roughly $15,000 at current market value. That pretty well wiped him out too. Are there any lawyers that we know of that would be willing to pursue a Chinese-style case on his behalf? Complicating the matter is that most of his income is generated by running bots over the tso money objects -- but on the other hand they were bot programs that he wrote himself.
Posted by: ludlow | Apr 26, 2004 at 00:49