A common saying goes that if you get something for free, you’re not the customer, but rather, the product being sold. Ripard Teg had raised this topic recently and some of the comments are quite insightful. But let me ramble on about it in greater depth, because it’s quite relevant to the whole thing.

Just what is it that a MMO sells, and to whom?

While it may be tempting to just assign the labels of "product" or "customer",1 the actual relationship between the users of a virtual world and the company that runs it is typically more symbiotic than anything else, regardless of the methods used to extract the actual revenue. It can be more or less parasitic,2 but parasitic can go both ways, and there are multiple levels and types of "products" served to different "customers" as part of the whole relationship.3 Thinking of "product" and "customer" simplifies things, which may be useful for campaigning (justified or not) but isn’t very useful for analysis.

In a MMO, what the company really provides it’s users (regardless of what is seen as the actual product by the management — or whether they think of it in these terms at all) is the environment in which a relationship can exist between players — a relationship that is, in it’s nature, social, and colored by the mechanics and style of the game world that is initially defined by the company itself, with it’s design goals.4 Content is more commonly created to serve this atmosphere, rather than to be the product itself — even though a MMOs can amass enough content to exceed a typical single-player game over the years, it’s quality often doesn’t measure up, and quantity is normally watered down endlessly with grind.

That environment is the initial draw for the users, but what keeps them around once the content is no longer fresh and new (even if there’s still more of it) is the community, that emerges between them once a critical mass of users is achieved. Once the initial phase of collecting the critical mass is past, it may be tempting for the company to shift from symbiotic relationship to a parasitic one, and many certainly try. But with MMOs in particular, creating a truly parasitic setup is typically impossible, unless some kind of monopoly is involved.5 The relationship remains symbiotic, the users depend on the company to provide them with a medium for interaction, and wouldn’t find each other without it, while the company depends on the users both for the revenue and the continued attractiveness of the medium.

The specific level of importance of the community to the whole project can vary, however. The average lifespan of an account on a typical themepark MMO is something on the scale of 1.5 months. They usually expect high turnover, providing just the right amount of content to complete in a relatively short, intensive usage cycle, through grinding, and don’t aim to keep users for much longer than that — it keeps the costs down and spreads the population evenly across the server architecture. They periodically release extra content packs, prompting past users to resubscribe, grind through the next batch of content, and leave, to return when the next content pack comes in. As a result, the limit of catering to the community itself is in making sure people don’t obstruct each other too much and the new player experience is running smoothly. There’s little emphasis on the hardcore crowd and the aim is to appeal to as wide an audience as possible. It’s generally impossible for someone employed6 to sustain an intensive gaming binge for long in the first place, so more often than not this resubscribe-lapse cycle works out fine.

A complete opposite is a long term MMO, which aims to keep subscribers over as long a period as possible. There are relatively few examples of those7 — unless we count the MUSH/MUX textual environment family of yore.8 Community and player-generated content, whichever kind of player-generated content the environment can support, become the main, if not the only draw for those, and the company-provided substrate is reduced to the role of initial attractant. Content packs give way to releases that emphasise new interaction features over content as such. It is basically impossible to occupy a user with company-provided content for sufficiently long to pull off an eternal subscription, but once you give the users enough rope, they shall happily bind themselves.

It is a sliding scale, and every real game falls somewhere in between the two possible extremes of letting the users rotate endlessly by piling on company-created content, and giving the users full run of the place without ever trying to control the emergent content. Neither extreme actually exists in practice, if only because of various limiting legislation.9 It is important to note that the earliest MMOs and their direct precursors actually aimed for the longer end of this scale,10 and recent games tend to aim for the shorter end instead. It’s tempting to say that this is merely because there’s more money in the casual crowd. However, while obvious, this is clearly not the only reason — fast looping cycle involves considerably higher development and upkeep costs, support costs go through the roof, and ensuring a constant influx of new players requires a constant marketing effort. While the amount of money floating through such a project is considerably higher than that floating through a project on the opposite end of the scale, the return per dollar invested is probably not far off. Other factors are at play here.

Beyond the purely technical reasons that starting from scratch every time tends to produce a more modern and more maintainable piece of software,11 longer term games eventually run into the endgame problem — how to occupy the players who have reached maximum levels. Multiple approaches evolved to deal with this,12 and adapting to this happening more than once eventually can produce the resubscribe-loop strategy all on it’s own, but even that isn’t really it.

You probably won’t find this cited anywhere as the reason. But my best guess is that the biggest reason for this shift towards shorter term looping was, paradoxically, the user community becoming the "product". Which is generally seen as giving the company "producing" it more freedom, — and it would, if the consumer was an unaffiliated third party. With the "product" consuming itself by growing, it actually binds the company with obligations instead, becoming too volatile to handle.

Once a game’s community is set in it’s ways and a social structure emerges, the developers don’t have many ways to change it — influence, by all means,13 but direct change is only available through pinpoint account banning. Once the community grows big enough, it starts making a nuisance of itself. Producing derivative works that infringe on the company’s intellectual property because they are themselves part and authors of it, RMTing the things they spent their time and subscription money acquiring,14 and in general throwing their weight around and demanding a more active role in developing something that is already a major part of their lives. People who are aware (consciously or not) that they are the main attraction of a MMO naturally feel entitled to be treated as such, and can come into conflict with the developers because of that alone.

Should the company actually want to move with the flow, a game’s community is composed of subgroups with diverse interests, and all of them are yelling at the developers at once.15 If you take even a cursory glance at the Eve forums, you can see no end to shining examples of that. Mishandling the community can cause a localised or long term drop under the critical mass, which produces it’s own feedback loop of "Nobody’s on, so I’ll log off too." and dealing with it becomes a tightrope walk. Corporate structures tend to find that sort of thing uncomfortable — they aren’t very good at working with mobile natural disasters.

These issues with running a long-term game are inherent to MMOs as such, or at least to their current design traditions, many have no ready universal solutions, and combined, they make truly long-term games less attractive to develop, so over time, many drifted off that end of the scale, edging closer and closer to something like a cooperative multiplayer game with infinite DLC, and seeking a balance between long term job security and short term manageability.

How does this whole thing apply to Eve Online? That will be the subject of the next installment, but you can probably see where this is going.

  1. Certainly is with, say, Facebook. Incidentally, once I have quipped that Livejournal is the biggest MMO ever created — mind you, that was just months after Facebook was launched and well before I ever heard about it. The structural similarities are quite clear.

  2. Some of the revenue collection methods, like the energy mechanic, and the whole gamification business in general are inherently exploitative and prone to turning properly parasitic, but that’s beside the point.

  3. Notice that in the current web services ecosystem, many such businesses are created with the express purpose of tempting larger internet companies into bying them to add to their existing service packages, rather than actually collecting revenue directly, which makes things even more complicated.

  4. To quote myself from analysing Second Life, which I’ll mention again many times, "When you give people unrestrained freedom to exercise their imagination, you will find that 90% of them have no imagination and the remaining 10% have no taste." Insert sarcasm tags as appropriate. In fiction, collaborative work on universes without any kind of editorial oversight usually leads to unsatisfactory results, something has to set the tone at the very least. In Second Life, nothing sets the tone except the technical limitations, which were built with a very different audience in mind and now cannot be safely changed. Doing anything interesting there turns into an endless exercise in creative workarounds.

  5. Usually, on the thematic elements. As long as your game is based on a licensed property, people who enjoy the original property will have to play your game, if they want to play anything based on the property at all. If you’re explicitly a Tolkien fan, LOTRO is your only option. As long as you’re the only internet spaceship game around, there’s no other way to play internet spaceships. The inverse is also true — as long as your only audience is people who like internet spaceships, people who want elves won’t come.

  6. MMO players are in general quite a diverse sample of the population, from kids to grannies (met quite a few of the latter myself) and the average floats somewhere around 26 years of age. At least half of these people will be employed and have families. See The Daedalus Project for more information. It is, admittedly, a bit dated, the biggest difference from current data will be the narrowing gender gap. Eve Online, by being so special, has considerably different demographics, but I digress…

  7. The only one I know beside Eve that aims for an eternal subscription is the aforementioned Second Life. When my brother first saw it at my insistence, he said, "I know what this is. This is World of Warcraft for women." But it’s actually Eve Online for women. So many things are startlingly similar…

  8. These count in the multitudes, or at least did. A very different beast, now almost extinct and practically forgotten, which is a crying shame really, because since then, you pretty much can’t see roleplaying worth a dime on the Internet. That’s one more topic for a later post…

  9. Will anyone please think of the children? No? Thank you.

  10. The longer end does seem more commercially interesting at the first glance. Ultima Online in particular was the template for many, and notably included such things as player housing and other social trappings that only make sense in a very long term game. The lead designer for the project, Raph Koster, whom I shall mention many times over yet, directly cites stone age MUDs as inspiration.

  11. Well, not always. But one day, when I’m sufficiently depressed, I shall tell a horror story or two about my adventures within the Ragnarok Online client. I still shudder whenever I remember what they had instead of an item database and how they edited it. By the time I started messing with it, Ragnarok Online has already been fairly old, and so was it’s open source server emulator, which was horrible in very different ways. I ended up writing most of the scripting language manual by pecking at the source and trying to figure out what it’s doing.

  12. The aforementioned Ragnarok Online’s answer was interesting in that it introduced new classes, to get which you had to reach max level, revert to zero and go all the way up again. Which is about as smartass an answer as possible. The most survivable approach would be to dissociate player activity from advancement, but as far as I’m aware, Eve is unique in doing it that way, and Second Life just dispenses with the gaming elements altogether.

  13. There’s a lot you can do to subtly influence a community by (secretly) changing game mechanics in such a way that the change is hard to detect and exploit, and rewards the behaviour you wish to promote. Unfortunately this is done very rarely, mostly because a completely undetectable change is such a pain to make. I only succeeded in pulling something of the sort a couple of times, but with good results.

  14. RMT is an interesting subject in itself. Companies typically say they are against it because it detracts from the fairness of the playing field. But see Bragg v. Linden Lab for a precedent of how this can screw a company directly — through introducing it into legal obligations it is not prepared to serve.

  15. One of the major Soviet sociological research projects back in the 70s brought up an interesting conclusion that I find myself applying to just about everything lately: Knowing what the audience really thinks is inherently hard. The original research applied to what newspaper editors thought about the tastes of their audience, and compared it to actual data collected from the audience, which the editors were incapable of collecting themselves. The disconnect between the two was startling. Basically the only metric the editors had was the sales of their newspaper, and they operated with an imaginary reader in mind, (the logic literally was "It’s impossible that there aren’t enough people like this") but no easily quantifiable variables to tweak. The readers would just vote with their pockets, which would get delayed because of subscription cycles, and base their decisions on things completely unrelated, further muddling up the situation. The disconnect between the audience’s actual tastes and the letters to the editor was no less startling, because people inclined to write them are not by any stretch representative of the entire population simply by the virtue of being such. Ferreting the objective truth out of the whole thing every time it’s needed simply costs too much.