It's easy to get confused by virtual worlds metrics. After all, there are many different ways to track population. Will you choose total registered users, total active users, total concurrent users, number of hits to the launch website, number of total accounts registered... you get the idea. To make matters more difficult, companies that own virtual worlds are typically mum, or at least, very selective, about the types of data they release (recall Ryan Olson's exercise in frustration at Red Herring earlier this year).

In the comments section here at Worlds in Motion, reader "darniaq" correctly pointed out:

And yet, I wonder: what advantageous do companies have to report numbers in a consistent and comparable format? Each companies' business needs are slightly different. Their first audience for such data may be publishers, hosts, share holders, management, whoever. The data they prepare for these audiences is tailored to what is most relevant to them.

It's nice to have subscriptions and uniques and registrations in press releases. But in the end I feel these numbers matter most to folks with the least stake in the success of a game: analysts and players.

Over at Cisco's Virtual Worlds blog, Christian Renaud has some apt observations -- if we were to factor in the aggregate of all these "flexible" virtual worlds metrics for 44 of the top products that call themselves virtual worlds, we'd end up with some 465 million subscribers, more than the entire population of Mexico, the United States and Canada!

So numbers are clearly a poor way to study virtual worlds. Nonetheless, Renaud says, it is essential for the industry to start defining some kind of taxonomy:

You can slice and dice the market by 2D vs. 3D, web-based vs. client software, apples vs. oranges, but we need to find a common set of language by which to differentiate the QQ and Cyworlds from the ActiveWorlds and Kanevas from the Metaplaces and Toontowns. Until then, you have emoticon-on-steroids avatar chat in IM and Social Networking sites being compared apples to apples with narrative driven virtual worlds like World of Warcraft or Runescape. It’s not apples and apples at that point, it’s apples and orangutans.

Raph Koster agrees somewhat, irritated at seeing the user-generated worlds of Areae's Metaplace put into the same category as Disney's kid-focused minigames-for-jellybeans world of Toontown ("OK, we get the message, we’ll redo the site’s graphics!" he says). Nonetheless, Koster says, perhaps it's not sensible to differentiate them totally when they share "99% of their core architecture."

On the other side of the coin, trying to press certain products into the category of either "virtual world" or "game" doesn't make much sense either; oddly, it looks like MMO designer Steve Danuser says that World of Warcraft is the virtual world, while Metaplace is the game.

So if numbers are useless, "game" is subjective, and client display, as Koster adds, doesn't determine purpose, what factors will we use to segregate and study virtual worlds as their very own entity, in a world where "game" is a subjective term, and virtual worlds for business and retail purposes are on the rise? It's an interesting question, and one to which a clear answer has yet to surface.

[Via Raph's Website]