The New York Times in the last week has taken an interest in not just one but two virtual worlds. Both of them are something other than a reskinning of Zynga games that many former virtual world designers seem to have concluded is their inevitable future. Both of them have their problems, both as games and as worlds, but if you'll forgive me taking another ride on my long-established hobbyhorse, both of them also underscore how powerful "world-making" has been and could yet be as a design niche in interactive media.
Dwarf Fortress is pretty much the most notorious, extreme example of a sandbox game that prioritizes simulation and complexity over visual design or interface. As the excellent Times profile puts it, it's "willfully noncommercial", which is putting it mildly. When you've played it enough that you can actually see the ASCII graphics for what they represent, you've performed the gamer equivalent of being able to open up Finnegan's Wake and read it straight from the page.
I think there's more in the creative space that Dwarf Fortress defines, however. Minecraft and Terraria have a strong familial connection both in their homebrewed development process and in their open-ended world-building. Wurm Online is another member of this kin group, as demanding and frustrating in its way as Dwarf Fortress can be. Wurm also reminds again that Eve Online still stands as a commercially successful example of this kind of design direction, even as many World-of-Warcraft imitating rollercoaster-ride MMOGs have failed.
Sandboxes of this kind have their own serious discontents and shortcomings, and most of them aren't really about graphics. The more they involve the simultaneous interaction of many players, as Wurm and Eve do and Minecraft can, the more dependent they are on either finding congenial social clusters of like-minded creators or on accepting a brutally Hobbsean social order. Dwarf Fortress and Minecraft, however, are maybe the most successful examples of sandboxes with substantial emergent or procedural elements, where the underlying rules governing the generation of gameworlds and the behavior of elements within those worlds produce surprising or unanticipated effects and structures. (The Times article gives a couple of great examples of interactions that surprised even Dwarf Fortress' two developers.)
That kind of unpredictable, emergent action is the aspect of "worldliness" that terrified most developers of commercial massively-multiplayer online games, that they worked hard to control, streamline or eliminate. But in many ways, that kind of activity is not just what makes Dwarf Fortress and Minecraft so interesting, it is what has lent a good deal of social media and gaming technology its creative, inventive energy. What is powerful about Twitter or Facebook is less what their owners think that users will do with the tools and platform and more what users discover to be possible to do almost in spite of what the owners imagined. What is turning out to be interesting about Kinect is less the few poor, dull implementations of Kinect in commercially available games for the XBox and more what various hacks have uncovered as its artistic, expressive potential.
Shadow Cities, celebrated by Seth Schiesel in his characteristically chirpy, promotional style, is more disappointing by far in its actuality, yet still suggestive in terms of what might come into being in more imaginative hands. It's an iOS app that promises a virtual world generated by real-world locational data.
Schiesel eye-rollingly terms it the most "innovative, provocative and far-reaching game" in the world, which is about ten degrees of hyperbolic overstatement from the mark. It's a free app with a monetized sideline (you can pay for mana potions rather than grind for them, and some voluntary activities almost require paying for some or having sufficient patience to wait days or weeks to acquire enough through normal gameplay). You play a mage in one of two factions. The gameworld is an overlay over real maps, and uses your GPS data to locate your avatar in the world. The principal action is PvP against your opposing faction, with an intermittent PvE grind against "spirits" that is pretty much a requirement if you want to level up your character.
If you've played even one of the major commercial MMOGs from the original Everquest right up to the current generation of free-to-play or subscription games, you will immediately see a whole bunch of tediously familiar design problems, some of which are almost unforgiveable at this point in the history of the genre. Even Schiesel notes, for example, the frustrating way that the game awards credit for banishing opposing players or destroying their structures: the player who makes the kill gets all of the reward. I've seen this movie before and I didn't like the ending the last ten times I saw it. The predictable consequence of that design mechanic is that a game whose fundamental premise is cooperative play compels players to be suspicious of their own allies, to set up controlled cross-faction farming, and tons of accusations about kill-stealing.
What disappoints me most is not the repetition of design mistakes, however, but the unimaginative usage of the possibilities of the platform and gameworld. Other alternate-reality game designers have recognized that the real point of geolocating a game is the use of massive amounts of player-generated metadata to create the kind of emergent, unpredictable game environment that otherwise has to be created through painstaking top-down effort by programmers.
Shadow Cities' use of the world we live in is bland and monotonous, nearly random. It's exciting the first time you have a "friended" ally who gives you access to New York City, Tampa, or Helsinki. What you quickly find out is that in Shadow Cities, everywhere really is the same. Some maps have chokepoints that make for good gameplay, but not because the actual topography of the lived city has generated those chokepoints, simply because an automated relation between the map and the population density has created a random arrangement that happens to be tactically interesting.
Imagine instead if any number of methods for importing player-generated metadata about locations were used to put some of the emergent complexity of the real world into the game's fantasy environment. People already tag and describe places in any number of existing social media: there is plenty of information available that could be used with even simple semantic parsing to generate appropriate bonuses, tactical quirks or distinctiveness to gameworld locations in Shadow Cities. Locations in significant "natural" places could generate bonuses to one faction, while locations in important technological or industrial places could give bonuses to the other.
If that prospect seems less appealing than what Shadow Cities' developers seem to be doing instead, which is designing various new game mechanics that will in time probably generate predictable kinds of player angst about balance issues (and probably goading more purchasing of monetized services), that's not surprising. That would be the same timidity that drove many commercial developers away from sandbox-type design in the first place, because to go that direction is to accept that the content of the gameworld will no longer be authored or controlled by the developer. That anxiety doesn't just affect game designers: it haunts the owners of social media, as well. But some things are only possible if you're willing to be delighted rather than horrified when your newly introduced carp start eating your dwarves.
Comments