I recently re-read the Eighth Circuit’s 2005 decision addressing online gaming, reverse engineering and interoperability (Davidson & Associates v. Jung, 422 F.3d 630 (8th Cir. 2005)). The case raises the usual questions regarding open- versus closed-system competition, but it adds a twist and it is that twist that I want to focus on. As usual, I will come out on the side of evil, or at least evil as most of the copyright world usually sees these things.
Blizzard produces a variety of war games, including StarCraft and WarCraft. In 1997, Blizzard created an online gaming environment, Battle.net. At Battle.net, you could test your WarCraft skills against other users. Today, this is a common move in the game industry taking what had been a freestanding product—a game DVD or CD—and turning it into a full-blown service and a multi-player environment.
A small group of programmers and game players decided that they would like to create an alternative online gaming environment, a competitor to Battle.net. We might characterize this group as entering the online gaming environment market. To enter, the group needed to reverse engineer the protocol language required to interact with the free-standing Blizzard CDs. That protocol operates as a technological barrier to entry, and, in reverse engineering it, the entrants did nothing more than what was required to make possible entry into the gaming-environment market.
Blizzard would undoubtedly see the case otherwise. This is a systems competition case, meaning that we have two or more components that work together. Here those components are the free-standing physical games and the Battle.net online environment. The stronger the ties between the physical games and the environment the more that Blizzard can choose a sensible mix of prices between the games and the environment. If there is simply free entry in the online environment market, Blizzard will have to charge a relatively high price for the freestanding games.
These are standard systems competition issues and I will not belabor them. The interesting wrinkle has to do with the relationship between the online environment and attempts to control illegal copying of the physical games. Imagine this copy-control system: each time you play the game, the game has to be authenticated over the Internet. That might not prevent physical sharing of a single copy of the game, but that isn’t necessarily what the game maker is concerned about. Instead, authentication online will prevent simultaneous use of multiple copies of the same game. That is precisely the sort of indiscriminate copying and sharing that the game maker would like to try to prevent and that online authentication would address.
Of course this presents a number of issues. Many games are played off-line; our Xbox at home for example is not on the Internet. Moreover, if we required the game to phone home with each use, this could raise privacy concerns, though we are seeing and will see much more of this. I just set up two new Dell laptops at my house and I am not sure exactly how many phoning-home technologies I turned on. At least one for Windows Update, another for McAfee for viruses and another for Dell itself. Maybe more.
What is the alternative to mandatory online authentication? Honey, honey in the form of Battle.net. Far better that your users want to go online. Turn a freestanding product into a service and use that ongoing relationship as a way to monitor copying of the physical media. This is exactly what Blizzard was doing, as the Eighth Circuit’s opinion makes clear.
Now we see the problem of the competing gaming environment. That environment made no effort to assess whether the games that were going online were legitimate or illegitimate. The emulator simply granted access regardless of whether the user going online could present a valid CD key. Indeed, one suspects the part of the attraction of creating an alternative online environment was precisely that it would make possible multi-player participation for those who do not have legitimate copies.
In this case, the link between the online and the offline—the online gaming environment and the physical CDs—operates as a way of policing impermissible offline copying of the CDs. A competitor in the online gaming market will have no incentive to respect the copyrights in the games, indeed, just the opposite, as it can offer the illegitimate copy holder a place to play.
We have choices. We can bar the reverse engineering and thereby prevent entry of a competitive online gaming environment for WarCraft (or, more precisely, entry without Blizzard’s permission). That was the result in the Eighth Circuit which upheld a contract that waived the right to reverse engineer. Alternatively, we can allow reverse engineering and lose the copy control check that Blizzard created. We might order a series of forced transactions by requiring Blizzard to unbundle copy authentication from the online environment and by forcing entrants to buy authentication services from Blizzard (at what price?).
What choice do you like?
It seems like the best answer depends on the market involved (which is why these issues seem so difficult).
For example, if reverse engineering is allowed, how much will the cost of the software go up for users that pay for it? What will that do to the "illegitimate copy" market?
More important, does copy control really have that much to do with it in reality? If a competitive environment came along that included copy control, would that affect the price of its offline software? That's a question that would require some good data on different online gaming systems and pricing associated with them and their offline components.
I would consider a fourth option (or maybe this is a subset of your third): Allow entrants to create competitive gaming servers, but force them to include offline authentication (at zero price) at Blizzard's request. Blizzard cannot claim loss of control, and competition is free among online systems. It wouldn't surprise me if we started seeing large product purchase discounts in exchange for longer term service agreements, similar to those you see with computer purchases.
Of course, this doesn't consider the widespread effect of a rule that bars "no reverse engineering" agreements.
Posted by: Michael Risch | July 31, 2006 at 03:00 PM
"The stronger the ties between the physical games and the environment the more that Blizzard can choose a sensible mix of prices between the games and the environment. If there is simply free entry in the online environment market, Blizzard will have to charge a relatively high price for the freestanding games."
I'm pretty sure the price of Battle.net access was zero...
Posted by: Tim Lee | August 01, 2006 at 07:33 AM
Advertising embedded in the online environment? The court of appeals opinion makes an obscure reference to this--at least to me--seemingly suggesting that the official version contained embedded advertising--another way of charging--while the unofficial version might have had none or different advertising. Yes? No?
Posted by: Randy Picker | August 01, 2006 at 08:49 AM
That's a good point. Battle.net does display advertising. I would be surprised if the ads generated very much revenue compared with game sales, but it's probably does increase the profitability of the game.
Posted by: Tim Lee | August 01, 2006 at 10:13 AM
There's an important point that you are missing: first amendment issues.
Software is speech. If BLizzard wants to make its games networkable, it has to accept the possibility that someone will listen to the packets being emitted (listen to their speech) and possibly talk back to them. They can make this more or less difficult -- by talking in a language no one else knows (an undocumented protocol) or by speaking in code (using strong encryption). Nevertheless, a person who is restricted from listening and talking to the software that he has legitimately installed on his computer has lost his voice.
Posted by: -dsr- | August 01, 2006 at 11:24 AM
"As usual, I will come out on the side of evil...." Randy, you're a regular manservant Hecubus. (Reference explained here: http://en.wikipedia.org/wiki/The_Kids_in_the_Hall#The_Pit_of_Ultimate_Darkness_.28Simon_and_Hecubus.29 )
Your third choice (and Michael Risch's fourth choice) sounds like the content protection model used for DVDs and other audiovisual media -- the content protection technology is licensed by some entity to third-party device makers (including software makers) for some sort of fee. There was a lot of debate between Philips and the "4C" and "5C" companies during the Broadcast Flag technology certification proceedings over exactly what those fees should be and how they should be determined -- but even if they are truly priced "at cost," cost for large corporations like Sony and Matsushita is *way* beyond the means of your average geeky programmer/online wargamer designing his or her own matchmaking service. I think the fees tend to run in the thousands of dollars, minimum. Licensing a content protection technology is going to be difficult or impossible for someone attempting to build their own matchmaking service at home, that operates at no charge to users.
I tend to side with "evil" myself on these issues, so I don't think that ends the discussion, but it is further evidence that the fuzzy lines between geeky consumers, "pro-ams," small businesses, and large businesses are really gumming up the works and making these problems incredibly difficult. The third option works well for handling the large players, but if you believe that hobbyists, or "tinkerers" as Ed Felten would call them, should have a role to play in these markets (essentially by disrupting them free of charge to the end user), both the first option and the third option are problematic. The difficulty is that hobbyists now in many ways have the disruptive impact that previously only large for-profit companies could have had.
Posted by: Bruce | August 02, 2006 at 12:17 AM
Bruce:
Thanks for recognizing that geeky tinkerers exist, and MAY a role to play in tech markets. That's more than I've seen out of Randy.
Posted by: Doug Lay | August 02, 2006 at 09:14 AM
Isn't there another point we are missing. Namely, Battle.net has the right incentives to get the question of having multiple competing online gaming sites correct. If they believe that the value of the game increases by having competing online venues then maybe they would purchase the competing environment or simply allow it to exist. However, if they find that splitting users among multiple sites drives down the value to all users of having a single point of online entertainment they can shut down access to the alternative online environments. Of course, either option could be correct, but shouldn't we believe that Battle.net is more likely to get the question answered correctly and thus we should defer to their choice. While anyone can second guess, as a policy matter, it seems allowing the copyright holder to choose is best.
Posted by: Ian | August 02, 2006 at 11:32 AM
Two notes from someone who followed these events fairly closely:
1. The bnetd team offered to have their software authenticate the player's software ID, thus mooting the issue of Battle.net being copy protection software (bnetd being the name of the third-party software at issue). I don't beleive the 8th Cir. mentioned this fact in their opinion, but I'm pretty sure it was in the briefs. Such a solution might seem strange, but it is in fact quite succesful in real life (See E.G. Return to Castle Wolfenstein multiplayer). It was this fact that led many to beleive that the issue in this case had nothing to do with the economic analysis given here, and everything to do with retaining control over multiplayer usage of the games.
2. It is asserted that " If there is simply free entry in the online environment market, Blizzard will have to charge a relatively high price for the freestanding games." Accepting the proposition that current and past game prices have not been relatively high, this proposition is not borne out by the facts. Nearly all games save for massively multiplayer ones offer free online play, even those games which are designed primarily for online play (See E.G. Unreal Tournament, Quake III Arena).
I'd strongly urge anyone seriously interested in the case to read the case documents on the EFF's website in order to get a fuller picture of the facts and arguments in this case.
Posted by: Zvi Rosen | August 02, 2006 at 04:45 PM
Ian brings up a good point--but as I remember the facts of the case (perhaps Zvi can confirm?) it ends up cutting the other way. If memory serves, the (purported) reason that Jung et al created their own version of the battle.net server was that the official server was having stability issues. It was frequenty inaccessible, so some users decided to bypass the problem so that they could use the program as it was meant to be used.
I'm not quite sure what to do with this point--should we decide whether to allow users to bypass access controls based upon whether they are effective period, effective as implemented, or implemented competently regardless of whether they are effective?
--Gus
Posted by: Gus | August 02, 2006 at 05:35 PM
Gus is exactly right...they created it because the Battle.net servers couldn't handle the traffic load when the service was rolled out. However, this wasn't the only reason (somewhere in the supporting docs seven reasons are given, none having anything to do with piracy). I'll try to post it tonight.
Posted by: Zvi Rosen | August 03, 2006 at 02:57 PM
Gus and Zvi,
I think the thrust of my argument has been missed. Namely, the particular factual circumstance of the case is meaningless. We have to answer the question without our bias of whether we believe that the company acted rationally. Once we accept that Battle.net (or the originating company) is most likely to get the question right (due to a review of the incentives of the different parties involved) as a policy matter (or as a matter of law) we defer to the party with the best incentives. Basically, I think this should be a question of law and not of fact. The facts shouldn't matter.
Posted by: Ian | August 04, 2006 at 02:37 PM
抛光
磨料
Posted by: 磨料 | September 15, 2006 at 03:58 AM
Tim said:
"If BLizzard wants to make its games networkable, it has to accept the possibility that someone will listen to the packets being emitted (listen to their speech) and possibly talk back to them."
by "has to accept", do you mean that by law, they can't forbid it? This raise an issue to me. When someone use a program to talk back to the server, Blizzard BAN de Cd key and Close the used account (losing all active character). is there any pratical law that could prevent them to act like that ? If someone using a pirated copy of Windows has his disk formated when logging into windows update, Microsoft would probably be sued. What make it different for Blizzard ? last month they closed 30 000 account and ban cd key for the *talk back* program used. can you make reference to any law that allow (or shouldn't allow) them to act like that ?
Posted by: Patrice Villeneuve | October 08, 2006 at 10:20 PM