New Second Life Statistical Release Disappoints

Linden Lab has released a new set of statistics covering the virtual world of Second Life that are slightly more granular than the information that’s been available before. Information about land mass, population and L$ transactions and trading, as well as other bits and bobs are available in an Excell spreadsheet from the official Linden blog. Still, the release is disappointing on a couple of levels. Methodology is one. The source for each stat needs to be spelled out much more rigorously than it is here. On the “L$ Supply” tab, for instance, things like “Sinks to LL” and “Other sources (sinks)” need a more complete definition than they receive here or on the economics Web pages. In addition, the new release does almost nothing to move the debate over usage forward. While new figures finally describe “unique” users that have actually logged in separately from “residents” that may be alts or may not ever have logged in (around 2 million vs. 3 million), there is no additional information about usage other than an aggregate number of hours used per month. The blog post accompanying the release notes that “Approximately 10% of unique users have logged in for 40 hours or more,” but that’s all we get. The Lindens could do better here, and at this point have no excuse for not coming up with a metric they like, even if it’s one that not everyone will be satisfied with. Because what they’re giving us now is something that almost no one who is looking closely will find sufficient. That said, the post does note that this is a preliminary report and that more information will be incorporated as things move forward. We look forward to seeing a good, deep report get designed, one that’s released on a regular schedule. Hopefully, we won’t have too long to wait.


  1. John Hurliman

    What is the argument against using aggregate hours per month in-world? That seems like a very telling statistic and would also be useful for comparing across MMORPGs or metaverses. I am planning to use that figure in my upcoming study, unless there is a skew or bias in that figure that I am missing?

  2. Mark Wallace

    There’s absolutely nothing wrong with that figure. But it would also be great to get averages and means for hours per user, as well as a range of other metrics to do with churn and various other things. That’s all I was lamenting.

  3. Mark Wallace

    Actually, what that figure doesn’t tell us is how many of those hours were spent by users who then didn’t log on again for 30 / 60 / 90 days, etc. I.e., how many of those are dips of the toe that found the water too cool. Could be a significant figure.

  4. Prokofy Neva

    The important thing to note is that just in the time Shirkey was griping about the numbers in Second Life, the landowners jumped from more than 42,000 to more than 57,000, i.e. in about 60 days. That’s pretty amazing. Those are the real numbers.

  5. Clay Shirky

    Prok, you have trotted out that same lame observation in about 30 different places now. The landowners data says nothing about user base.

    Mark is right — all we know is that 10% of unique users have logged in for at least 40 hours or more *ever*. Not in the last week, or month, or even year. Ever. This is an almost meaningless statistic, it gets more meaningless with each passing month.

    I’ve been at this almost two months now, and the core question — How many current active users are there? — has never been answered, by the Linden’s or anyone else. If fewer than 200K users have ever spent 40 hours in SL, how many active return users were there in January? (Put more technically, how many users logged in in January of 2007 whose accounts were created more than 30 days prior to login?) We don’t know, and as Mark says, the Lindens could do better here, especially if their goal is “complete transparency and openness”, as Zee Linden says.

  6. Prokofy Neva

    Prok, you have trotted out that same lame observation in about 30 different places now. The landowners data says nothing about user base.

    Clay, I’ve been forced to trot it out in about 30 places, precisely because you remain mysteriously resistant to facing the obvious about the numbers. These ARE the users — the users that log on for the longest hours, the users who spend the most money, the users that have some sort of stake, so they stay in and keep logging on throughout the 30-60-90. I don’t know why on earth that is so hard to grasp.

    I can only chalk it up to some ideological allergy to the concept of land-owning. You feel this says something about “only the rich can play,” or “that isn’t the whole membership” or “that can never be representative”. But that’s silly. Landowners means 512 m2 of land and a $9.95 subscription, less cost and less property/inventory than your typical levelling-up WoW character?

    Why this heavy ideological resistance to seeing that the Holy Grail of real, live, logging on, every-30-days people, that you have been griping and groaning about for weeks as not existing in SL and being hyped, resident *in these very numbers, which are growing*.

    And, as I’ve been having to explain over and over again precisely because you are so resistant to this, these people represent the base for the architecture of all the numbers, and how to understand them. The largest landowners in this cohort then sustain the non-landowning in the forum of rentals or entertainment sims. So for every one of these, there are some 4 or whatever renters. So the real, actual, live numbers of people who actually have a reason to stay and be there (on land) runs to 200,000 or whatever. It could be computed looking at the landowner/tenant ratios and sampling them.

    The architecture of Second Life’s numbers simply doesn’t work like a flat MMORPG where there is no real estate, no property, no land, no homes, no clubs with proprietors. Try to accept that this world is not structured in the same way, and looking for facile WoW like subscription nevers, will never, ever capture what it is. I point to land owners, because they are the log-ons of this world that are visible and present just as you claim to want them to be to be “counted”.

    It doesn’t matter whether these 57,000 who generate 250,000 tenants and guests (or whatever the coefficient is) are “only 10 percent of the users”. The fact is, we know they LOG ON. They pay tier every 30 days. They are present and accounted for.

    I’ve JUST TOLD YOU that the current active users is 57,000 times 4 or 5 to make 250,000 or 300,000 — and each time I tell you this, you bat it away with an ideological stick as long as your arm.

    The only thing you don’t know then, is whether this 250,000 of real people are people who will not return in 30 days, or people who have been there since 2003, And yet, the fact that this number has grown from 5000 to 50,000 definitely tells you there is growth.