Determining Your World Size

Two of the really tough decisions you’ll ever make when creating a massively multiplayer game are: how many people should be in a single game world, and how much space should that world cover? Do you make one massive, seamless server? Do you make a massive, instanced server? Do you make a bunch of worlds? It’s a tough decision, and there isn’t a heck of a lot of formal research to base such a decision on.

The first part is, in my mind, the easiest. The only way to have one massive world with everyone in it is to have the right type of game. The best candidate for this is a game like EVE. Exactly like it, in fact. A game that will be for a niche market, can expand its borders without much trouble, and is designed to spread players out. There are probably one or two other criteria I’m leaving out, but that’s an ideal situation for a single world.

The trouble with making a single world is that the community becomes spread thin. It becomes difficult to find your position in the community, because there are so many people around. This will lead to the formation of smaller communities, but it tends to make you feel insignificant in the scheme of things.

There’s the variation of the “single” shard structure, which allows for a whole lot more people, but comes with its own set of issues. Specifically, the heavy utilization of instancing. Or, since these aren’t traditional instances and are instead fairly persistent (and shared), perhaps it should be called persistancing. See: Guild Wars. This game uses persistancing heavily in city areas, and uses traditional instancing for adventure zones. The main issue I have with persistancing is that it’s not unlikely that you’ll never see the same person twice. At least in massive, single world games, you can see the same people who congregate in the same areas more than once.

That leads us to what is usually thought of as the traditional MMO server model: shards. Multiple versions of the same basic world, shared by a limited set of players. This is my favorite for a few reasons. The first is the obvious development benefit: You don’t have to make a ridiculously massive world for people to adventure in. The larger the world, the less handcrafted it can be (with the same budget), so it becomes less interesting for players. Even if you have the budget and manpower to make a massive, massive world, there’s still a point at which you lose cohesion when the dev team gets too big, not to mention the more important reasons I have for liking shard systems…

The other two reasons are social. It’s much easier to form a great community with a smaller set of players than it is with a huge set of players, because you can really carve a place in the society of that particular world. And, in a massive world, the population becomes spread so thin you don’t see the same players often (so it’s hard to form communities), and it becomes very daunting to decide where you need to go for adventure.

My personal opinion, based entirely on my own observations and no research? 500-1000 players per major faction. In EverQuest II and World of Warcraft, that would mean 1000-2000 total players logged in at one time (Freeport/Qeynos and Alliance/Horde, respectively). In Ultima Online, a game with one major faction, that’s 500-1000 total. Server sizes, I have no idea. Smaller, really concentrated, handcrafted content is what I like personally, but I can’t give you a square mile or kilometer number.

Before I go on too long, I’m going to ask for your input on this issue. Does anyone know of any good studies about world size, or about the size of real world society sizes and their impact on the people in the community? What are your observations in this area?

26 Responses to "Determining Your World Size"

Comments are currently closed for a server migration!

Return to Ryan Shwayder's Nerfbat »