Technology and Market Structure of Virtual Network Games

We cannot see the future, of course, but therein the Dungeon of Befallen, but if tens of
are a number of technological innovations that arethousands of us go there to hunt them, none of
relevant to gaming, that are also fairly easy tous will have a good time. Sometimes the only
see coming. Currently, access to gaming involvesway to reduce congestion is to add content, but
some sort of access to computing technology,this, again, is labour intensive. There will also be
and access to gaming that can earn moneycongestion effects related to connection speeds
involves access to a shared, persistent, physicaland bandwidth.
computing environment, specifically a virtual worldA third reason that the market will probably not
. The technology supporting virtual worlds isbe dominated by a few companies can be found
advancing so quickly that it would be foolish toin the many competitive strategies that are
describe the next generation in any detail. Sufficeavailable even now, but have not yet been
it to say that there are large, lucrative industriesexploited by new entrants. For example, the
working energetically on different dimensions ofcurrent set of developers have managed to
the environment that virtual worlds thrive in.impose huge switching costs on players by
These industries produce three items of interest,structuring gameplay around the time-intensive
namely, connections, interface and content.development of avatar capital. A player starts the
Developments in connections include the internetgame with a weak avatar, but gameplay gives
and, increasingly, wireless communications.the avatar ever-increasing powers. As power
Development of interfaces includes voiceincreases, the avatar is able to take more
command, head-up displays and body motionadvantage of the game world, to travel farther,
detection (computer-controlling gloves, gazedo more things, see more people. A person with
readers). Developments in content include thea high-level avatar then faces a high hurdle in
supply side of the market for games, whereswitching games, because in the new game he will
annual revenues have grown beyond Hollywoodstart out poor, defenceless and alone again. This
box office revenues. All three industries aresituation definitely locks in the game's player base,
expanding at a rapid rate. Whatever emotionalbut it is also open to defeat by any number of
experiences people seek, it may become possible,schemes to reduce the switching costs.
in the near future, to effortlessly connect to aSurprisingly, no competitor to a current game has
virtual world that provides that experience atoffered new players the opportunity to start their
fairly low cost. Kurzweil argues that the explosionavatars at a higher level of wealth and ability if
of computing power alone may be sufficient tothey can provide evidence of a high level avatar
change the daily course of life.in another game. On the other hand, two games
Since these developments all involve networks,(Ultima Online and Dark Ages of Camelot) now
they may seem to suggest a monopolisticoffer methods to effectively start out ahead: in
market structure. If economic life online involvesUltima, you can directly buy your levels; in
getting your email and hanging around withCamelot, you can start a new avatar at level 20
friends, there will be positive externalities withif you have already gotten one to level 50. These
respect to the sheer size of the virtual world onestrategies help companies discourage the buying
visits. If I spend my time on Rubi-Ka, while youand selling of avatars outside the game, perhaps
spend your time in Albion, we cannot talk to oneat a cost to the atmosphere within the world. In
another, and we cannot do things together. Thus,sum, what appear to be strong lock-ins and
our time in virtual worlds is more valuable ifswitching costs in the game market today may
everyone we know is in the same world.not be as strong as they seem; when savvy
Moreover, if two worlds compete and one hascompetitors appear, the player bases will generally
more players than another, wouldn't everyonebe at risk. For example,some network games
have an incentive to join the larger world, so assuch as lotro gold,runescape gold,guild wars gold
to enjoy the larger network of society,etc.
communication and entertainment that it affords?A final argument against a monopolization
Might such network externalities lead to atendency comes from the nature of the content
domination of this market by one player?Foritself. Games are art, for the most part, and
example,some network games such as lotromarkets for artistic output exhibit a great deal of
gold,runescape gold,guild wars gold etc.churn due to herding effects and the star
There are reasons to expect, however, that thisphenomenon (MacDonald, 1988). If a company
market is not likely to be monopolized. First, theredesigns a better game, it will attract players. And
seems to be a great diversity of tastes for thewhile it is true that development costs can be
different features of a world. Mr. Bird may wantsignificant, it will always be possible to produce a
to be on Pluto, while Mr. Castronova prefersfun virtual world for a tiny amount of money and
medieval Britain. One of the major attractions ofthen scale it up as it becomes more
life mediated by avatars is the anonymity itpopular.Whatever network externalities,
affords, and anonymity requires a person to havesupply-side returns to scale, and barriers to entry
exit options: other worlds to escape to if one'smay exist in the market for virtual worlds, they
reputation in this one gets unpleasant. Perhaps aseem insufficient to produce domination by a
savvy game developer could make a world sosingle company. The distribution of populations in
large and varied as to provide the essentialvirtual worlds is perhaps less like a natural
minimum level of entertainment and anonymity tomonopoly market than a club goods market.
a sufficiently large number of people, so thatPopulations will sort according to the services,
membership in that one world becomes optimalambience, and fees of the various worlds. Virtual
for all. This seems unlikely, however, given thatworlds will compete, as clubs do, but their size will
there is a marginal cost to creating andbe limited by congestion effects and by the
maintaining game content. Moreover, there are nomarginal cost of increasing the scale of the world.
economies of scale on the supply side to matchThis analysis allows a tentative answer to the first
the increasing returns on the demand sidequestion of the study: in the medium-term future,
(Liebowitz and Margolis, 1994). Production of gamethe online multiplayer gaming market will probably
content and its maintenance are bothconsist of a number of large, densely populated
labour-intensive activities. One could perhapsworlds, with varying degrees of portability
increase production of content by allowing otherbetween them. The worlds will generate large
producers (say, by opening game code to therevenue streams and will occupy many hours of
public), but continued control of the world beinghuman time, some of it considered play, some of
created would be problematic. On the whole, itit considered work. The hours that people devote
seems very unlikely that one developer couldto games will result in the accumulation of stocks
produce a world big enough to monopolize theof digital capital goods. These objects will have
market.considerable economic value. Given the expected
A second reason involves congestion. Virtualgrowth in connectivity, interface technologies and
worlds are virtual because they are online, butcontent, there is reason to believe that this digital
they are worlds because there is some physicalitycapital stock may eventually become quite large.
to them. Avatars take up space. If a world has aThese considerations then lead to the next set of
certain amount of entertaining content in it, thatquestions: If virtual worlds do become more
content will almost always be subject to someimportant, how will this affect the real Earth
kind of congestion effect. The cool monsters areeconomy?