Trials, Triumphs & Trivialities #143:

If You Build It They Might Come: Critical Mass & Other User Buzzwords

by Shannon Appelcline
2004-03-04

So, you've got your brand new multiplayer game ready to go. You're ready to turn it on and then, of course, the entire world will flock to your game, at levels appropriate to the genius you've instilled within.

Or so you hope.

As I've said many times before, completing your game is just the first step in the process of creating an online game. Next you're going to have to figure out how to get players there and keep them. I wrote a whole series on this topic a couple of years back called, "If You Build It They Might Come", covering Attention (#85), Discouragement (#88), Encouragement (#89), and Assistance (#90). If was a mini-opus on player attraction and retention.

What I didn't discuss at that time was the fact that player attraction and retention can actually be a science all its own. There are a number of postulates (or theorems or corollaries, if you prefer) that you need to understand in order to manage the continued growth of a user community. Today I'm going to discuss four of the big ones, some of which I've probably touched upon before in this column without properly defining them. So, without further ado: critical mass, user plateaus, user churn, and the networking effect.

Critical Mass

There's a monthly event in nearby San Francisco called "Critical Mass", wherein hundreds of bicyclists join together in a group 3-4 blocks long and stream through the interesections of The City. The name itself is drawn from Ted White's documentary, "Return to the Scorcher", where it refers to China's intersection crossing etiquette: cross traffic waits until it has sufficient mass of cars, then it starts pushing into the intersection, at which point the original direction of traffic stops until it regains sufficient mass.

Critical Mass.

The idea in each situation (and in reactors, where you can need critical mass of nuclear material in order to start fission) is that if you have just a little bit of material (bicyclists, cars, Uranium-238), nothing happens, but when you step over a certain magical breakpoint, everything happens at once.

The same applies to online games, and it's the first thing you need to understand related to users. Until you have enough users for your game to be "populated", it won't be much fun to play.

The problem, as often as not, is that without sufficient players there won't be a lot to do in your game (or perhaps in a tabletop game like Gang of Four, your game will be utterly unplayable). This means that players will stick around for much shorter times, then leave, making the game that much more empty for the next player and the next and the next ...

When you're at 99% of your critical mass, your game might only be 10% as fun as when you get that extra 1%. Suddenly there are enough people for the game to be interesting, and so everyone stays around for longer times and thus there's enough people when the next person logs in and the next and the next ...

Trying to hit that first critical mass is going to be the first user-related challenge for any game. There are many possible solutions for getting there, including: heavy staffing, interesting single-player puzzles, and constrained hours of initial operation. The exact solutions which are helpful will partially be determined by what type of game you have (and perhaps there's another column in the future just on this topic).

User Plateaus

I talked above about the "first critical mass", above, because social theory actually predicts that there will be many of them. After your game is up and already successful you'll every once in a while hit what I call "user plateaus", but really they're just additional critical masses.

Social theory says, you see, that certain organization sizes work well and certain ones work poorly. This theory is used most frequently when discussing businesses (because that's where there's money to study the effects), but really it applies to any social groupings, including online games.

I've worked in a lot of companies, spanning from 3 or 4 employees to hundreds or even tens of thousands. And, I'd agree that the size of a community can make a real difference as to how well it works. Take, for example, a teeny company of 6 or 7 employees. You probably have one "boss" who can see the whole picture of the company and 5 or 6 highly empowered employees who each have absolute control over a specific area of expertise and feel very confident making decisions related to their expertise. It's a social "sweet spot" in the community-size range.

Now, increase that company up to 15 or 20. Your 4 or 5 areas of expertise suddenly have 2 or 3 people each in them. Either you have absolute anarchy or else one of your employees in each area is forced to step up and manage the group. Given that the company is small enough that you still need highly empowered employees, there's often friction between the employees and the managers. On the other side, things have now gotten big enough that the company boss can't keep track of everything on his own, and so his managers have to keep him apprised of status and help make decisions. You've increased your overall company size from perhaps 7 to 20, but in the process you've lost a third of your company to infrastructural (management and reporting) needs where they're nonetheless underutilized. It's entirely possible that once you pushed up above some magic level, perhaps 12 or 13 employees, each additional employee effectively cost you an employee as well--as a current employee moved up to infrastructural tasks to help run the now-larger department.

The same thing happens in online communities. At certain sizes (I don't know the exact breakpoint numbers, but I suspect one of them is around 175-200; there are probably several in the ranges of thousands or tens of thousands as well; and where these numbers lies probably varies from one game type to another) a community gets large enough that every new player who joins will in a very real way cost you an old player. This is partially because of sandboxing wherein there's only a certain number of "positions" within a game for people to fill, but it's also partially because of attention requirements, wherein all players need certain types of attention from plotters, game systems, or whatever, and when you hit these user plateaus every new player will steal away attention than an old player was enjoying.

This means that every once in a while you'll see your community size stagnate and at that point you must figure out how to get past this new problem of critical mass, and get your community growing again.

User Churn

It might seem a little odd for your community size to stagnate when you're actually still adding new players, but it will happen and the reason is: user churn.

Whenever you develop a new game community, it should flat-out grow for a while. But, eventually, you'll start having players leave too. For a multiplayer game, the user longevity ratings can be quite good. For example, our average Castle Marrach player sticks around for over 15 months and our average The Eternal City player plays TEC for 19 consecutive months. But, when those players starting leaving you have "user churn".

"User churn" is typically defined as "a number of users leaving an online service that are in turn replaced by new users." Thus, for example, if we had 70 new users every month and 50 users leaving, it'd look like we only had 20 new users a month because we had a user churn rating of 50.

To a certain extent, churn is inevitable. Players don't stay interesting in the same thing forever. They're going to move on to other games or other obligations within their life as a whole. But, not every lost player is inevitable, and thus user churn points to a way to quickly grow your user base.

If you can figure out why players are leaving you might be able to dramatically crank the growth of your community. In my above example, if I can figure out ways to cut down on 40% of users leaving (20 out of 50) I'd actually double the growth rate of my community, from 20 users a month to 40. So, to deal with user churn, quite simply: figure out why people leave. (Here at Skotos we're moving toward developing an "exit poll" which we hope will define why people leave.)

Here's another truism about user churn: user churn will move toward equilibrium. In other words, this is another reason that user growth can stagnate, when users leaving equal users arriving (and this is an expected balance point, because users leaving is usually a consistent percentage of your community while users arriving is probably a logarithmic rate that increases more slowly the more resources you throw at it, due to the law of diminishing returns).

The Networking Effect

A lot of what I've written this week talks about problems that you might have in growing a community. I'd like to end on a more positive note by saying there are some ways that you might get magical help in increasing your community numbers too.

"The Networking Effect" is really a fancy way of saying "word of mouth". If you manage to create a cool enough game that's unique and interesting, the players of your game will help spread the word. And, the more players you have, the more networking that will occur.

How can you boost the networking effect? That's a trickier question.

One way is to make sure that you have lots of players at the early stage, through an advertising blitz or what not. (I think that fact that Castle Marrach was free for the first 6 months helped not just to create critical mass, but also to create a networking buzz, because more players were around for longer.)

Second, consider ways to encourage networking. We've done that a little here at Skotos with our Story Points, though I think we could do much, much more.

Conclusion

Adding users to your game won't be a simple and steady climb. You'll hit various user plateaus along the way where you must figure out how to burst through critical mass barriers. In addition, churn will be an increasing problem the more users you have in your game. Conversely, networking effects might help you to increase your usership even further.

The bottom line is, you need to be aware of all this. You'll need to watch your user numbers as they grow, stall, or drop, and figure out ways to get around the tyranny impossed by the science of user statistics.

And that's it for this week. In upcoming weeks I have, ever so vaguelly scheduled: auction games, copyrights, trademarks, patents, and probability. In other words, the usual chaos of thoughts that I have on game design and administration.

[ <— #142: The Disconnect Dilemma | #144: Strategic Insights: Auction Games —> ]