Wednesday, December 21, 2005

Supply-Side Content Generation in MMORPGs


"There's not enough content!" How often have developers heard this lament from their most active players of world-y, "live-in" games like MMORPGs (e.g., Star Wars Galaxies, World of Warcraft) and dungeon crawls (e.g., Oblivion, Two Worlds, Fallout 3)?

Let's start with a couple of simple observations:

1. Content (things to do) is a crucial feature for attracting and retaining players.
2. Players will always consume content faster than you can create it (the "content race").

Many of the most consequential game design decisions in games lately seem to be aimed at trying to invalidate that second observation. How do you keep ahead of your players as they burn through content? What kind of design features help to mitigate content consumption?

This is a supply and demand problem, and as such it can be attacked from one or both of those aspects:

Demand-side:

  • Regulate access to content.
Supply-side:

  • Hire lots of talented artists and writers.
  • Use the work of lots of unpaid fans
  • Design a system for autogenerating content.
  • Provide players with ways to allow them to create content for each other.
There may be some exceptions, but based on what they actually do, most gameworld designers these days seem to believe that "regulate access to content" to be the only realistic option to hiring lots of talented artists and writers. Other supply-side approaches, when suggested, are dismissed as leading to "sandbox"-type games (as though the Grand Theft Auto series hasn't been freakishly popular).

I suspect that this preference for demand-side regulation of access to content is why we continue to see gameplay mechanisms such as the following:

  • Levels -- some content can be accessed only by characters within a relatively small range of power
  • Experience points (XP) -- actions generate XP which must be accumulated to reach the next level
  • "Grinding" -- some content may be (or must be!) repeated to generate XP
  • Zones -- level-based content grouped geographically
  • Race -- some content can be accessed only by characters of a specific race
  • Faction -- some content can be accessed only by characters liked sufficiently by a specific faction
These mechanisms and others like them do address the problem of players burning through content faster than a limited stable of writers and artists can crank it out... but why are designers so intent on framing the larger issue of the content race solely as a demand-side problem? Why not make a serious attempt to design systems that make good content easy to create instead of trying to ration content by regulating player access to it?

One objection to eliminating rationing mechanisms like character levels is that players "need" to feel that their characters are "advancing" somehow. I've heard it said by many people that "without levels, or a skill-based system which is similar, you must still allow the player to feel that his character is developing over time." Why? Says who? Where is there any evidence for this other than "because everyone believes it"?

Character development over time, to me, is just another way to say "XP-based levels." It's just another rationalization for using a comfortable demand-side mechanism to regulate the player's access to content. Yes, it may be a mechanism that has become familiar to players (and harried developers), but is familiarity really going to be what drives game design from now on? There was once a time when none of these mechanisms existed; they had to be imagined and implemented. So why can't we keep trying new approaches to content provision?

OK, maybe they won't prove to be as popular as the now-familiar demand-side mechanisms. But how can we know that unless someone gives these alternatives or others a serious tryout?

[2008/04/24 update: Apparently there is now going to be one game that does try to solve the content-production problem by creating content generation systems that allow any player to make aesthetically pleasing content... and the name of that game is Spore. Suppose this attempt is successful: will other game developers try it in their games? Or will they beg off, claiming that supply-side content generation doesn't fit "their" game?]