questproquo

Part 3: Squaring the Circle

If you’re reading this you probably either came straight from the Introduction or were patient enough to work through Part 1 and Part 2.  For simplicity, here is the short form of my theory on the ideal MMO content:

End-game primacy is the idea that the bulk of any MMO development time needs to be spent on maximizing end-game content and that this goal is best achieved by embracing complex systems driven by player interaction, rather than static content. Or put a different way: an MMO’s real story begins after the scripted story ends.

As I discussed in part 2, complex systems can be consumed by players longer than static content, but they are also significantly harder to create well. What works in a single player game may not work in an MMO, and vice versa. It is a lot easier to account for a few players’ actions than it is to predict the impact of millions. Millions of players is millions of opportunities to undo any work developers put into the game.  Structured static content offers a lot more control for developers. Instancing, phasing, personal stories, and other tools isolate players from the larger game population, while simultaneously making it easier to tell a story to the player.  It lets developers preserve the narrative format of single-player games.   In contrast, unstructured simulations take control away from a development team.  The simulation elements are built into the game world and then left to the players’ whims on what to do with them.  This is a frightening prospect because systems not designed to scale particularly well can tend to spiral well out of control, at the detriment of other aspects of the game. For instance, if players can build structures on game landscape, the game world will almost definitely end up suburban.  If players can build in a particular spot, they will if for no other reason than they can.  Expecting that it would not happen is betting against the odds.

Qualities of Good End-Game Content:

End-game content systems needs to ebb and flow organically, self-regulating through internal feedback mechanisms.  In the above example of players building on the terrain, physical game space is limited while the numbers of players who can (and will) build houses has no theoretical limit.  Instancing off the construction projects or expanding the physical land mass using some kind of terrain generator are two ways you can address the finite space.  I’m not stupid enough to think that the second of those two options is exactly feasible.  In many cases, it may be impossible. The alternative of instancing off construction projects or imposing artificial limits on where players can build to control the urban boom misses the point of building a game with lots of players in the first place.  To make the system work, there needs to be a natural feedback mechanism in place.  For housing, if players can create them – they must also be able to destroy them.  Determining the ratio between building and destroying would require a fair amount of find tuning, however.  If it’s hard too build and easy too destroy, no one will bother building.  On the other hand, if it is to easy to build and to hard to destroy, you still end up with a sprawl.

Going back to Part 2, auction houses have natural self-regulation through the invisible hand of supply / demand, but they lack a mechanism to contribute to player narrative.  Systems that have ways to report activity and interactions back out to players are another hallmark of a good end-game system.  Players can buy and trade all day on the auction house, but if there is no mechanism to report major fluctuation in trade, players will likely feel isolated in their game experience.  Even if such a mechanism exists, it probably does not tie into other systems.  Ideally, if a market change occurs, it should be in response to in-game events.  Knowledge of these events provides players immersion, rather than just leaving them staring at a user interface spreadsheet tracking their purchases.  A good end-game content system will let developers track and aggregate player interactions within and across systems to facilitate narrative development.  In the housing example above, a mechanism to monitor and encourage players to cluster buildings together facilitates the construction of towns instead of haphazard sprawl.  For instance, perhaps building in close proximity to one another increases their natural defense.  Two houses have the defense of three;  three the defense of five; and so on.  A shack in the woods could be torn down in a few minutes, where several dozen buildings together takes over a day.  Several towns merge to form a city which takes a week to siege.  And each of these units (e.g. town, city) provides an entity that developers can track and use as a piece to a personal player narrative that is not separate from all other player narratives.

An Opportunity in Player Social Landscape:

These are simplistic examples, to be sure, but it gets to the point of what I’m trying to explain. I cannot suggest an end-game system that will work in every MMO. Each game has its own limitations based on engine, design philosophy, resources and a host of other factors. However, there is one area in many MMOs today that I believe remains underdeveloped, but which also has great potential to add depth to end-game experience and help developers create quality evolving narratives in the process.   While most developers spend plenty of time shaping their landscapes and dungeons and are loathe to let players ruin that art, social terrain is an area that does not currently have structure and therefore cannot be destroyed.

The fact that I can often only be the member of one guild and my relationships to other players are defined as “friend” or “not friend” is fairly simplistic.  This flat and binary social structure is surprising given that large numbers of players are a feature of MMOs. Certainly, social terrain is difficult to communicate meaningfully – even Facebook struggles with it – but games need more than a few binary associations to link players. In life, we play many roles and in games we do as well.  Is there a particular reason that a player can and should only belong to one guild in a game? I would argue no. Further, guilds are generally the only mechanism to permanently join players in games.  This is an artificiality that misses the point. Letting players create different kinds of associations among themselves and build on the quality of those relationships through game play would be a fantastic way to add depth to a relatively one dimensional system.

Imagine if player organizations came in many different forms, which were not binding to individual players. Guilds might still exist as the highest form of player grouping, offering resources like shared banks and chat channels. Others might exist for circles of friends to communicate.  These circles could extend between guilds and offer benefits like being able to travel immediately to your friends’ location. Still others might exist for trade groups: players who regularly share crafting resources to each others’ benefit.  Being a member of this group may offer additional crafting benefits. Player organizations like this only come into existence when the game recognizes a cluster of individual player associations strong enough to warrant it. When players “friend” one another they select the kind of relationship they want to build. Small groups of friendship-linked players might warrant a “friend circle,” clusters of adventuring-linked players might warrant a “guild circle,” and clusters of trade-linked players might warrant a “trade circle,” and so on.

Natural player clusters provide the basis for player groups of all types.

These interlocking social circles would create an inherently organic system that already lies on top of almost any game’s existing game play.  Anchoring these circles into less organic game systems could vastly improve and regulate end-game play.  For instance, if a game allows players to fight for control of game regions, a map which only has guilds will be one dimensional.  Add in trade organizations that may operate across regions and suddenly you have two dimensions with the same players.  This in itself is a story, and one that with the right tools can be communicated back out to the players.  Using the housing example above, anchoring the ability to build houses to player associations may be a way to organically limit the growth rate of construction in a game.  Player groups may be the unit to build (instead of players) and they may only be eligible while the quality of relationships between members remains high enough.  Social pressure does the rest.  Capture that narrative and push it back out to the affected players and perhaps one or two tiers out (using the same association network). Players would learn about attacks on towns that their friends live in or where their trading partners do business. The game tells a story that people care about.  By necessity, this needs to occur in near-real time, a challenge in itself, but done right could absolutely change the face of MMOs today.

Conclusion:

As you can guess, there are a lot of things to consider in anything as complex as a virtual world, so these theories are as much a work in progress as anything else.  I hope that they were, at least, somewhat thought provoking. There are plenty of technical limitations preventing much of this from happening in the near future, but adding one new end game system to new MMOs should not be out of the realm of possibility. Hopefully over time some of those will be captured as best practices and replicated out, leaving room for newcomers to add even more dimensions to virtual world game play.