Going Nonconcurrent

A little heat, and less light, was generated this past week by a fluff piece in GamesIndustry.biz — ostensibly an interview with Linden Lab’s CEO Rod Humble and Emily Short, founder of LittleTextPeople (which gaming company LL recently nommed); ostensibly about “one of three new Linden Lab products to be announced this year”.

Yes, I know — and agree! — that the Lab should be fixing what’s broken in the product they’ve had for ten full years (Steller Sunshine’s 10th Rezday was 13 March, as, reportedly, was Philip “Who?” Rosedale-Linden’s). Maybe they will… and maybe the horse will learn to sing. Regardless, it may also be a rare-for-the-Lab example of business sense to build new baskets to put some of their eggs in: products which aren’t mired in 10 years of legacy code, patches upon patches upon patches, and which might attract the kind of users who need their hand (or paw) held more than Second Life provides; products which are more ~ahem~ game-like.

Most of the talk about that “article” on my Twitterstream, however, was about the new signup statistics Rod dropped into the middle of it:

“We managed to grow the new users significantly: they bumped up by well in excess of 40 per cent, and over the holiday period we had over 20,000 new people sign up a day. Now, that’s not Facebook numbers, but 20,000 a day…. that’s a lot, right?”

… coupled with an observation I first saw made by Botgirl: concurrency has not gone up as a result of the noob surge. The best source I know for those statistics is a pair of ancillary pages at Tateru Nino’s blog, Dwell On It. Let’s have a look:

First thing we notice is a natural rise and fall of concurrency vs time of day, as “prime-time” hours for most users creep westward across the face of the planet. I think it’s safe to assume that reported times are in SLT, a.k.a. Pacific; the peaks centered on 1600 hours reflect the Western Hemisphere membership (4 pm Pacific = 7pm Eastern, right after dinner) coming online as the Europeans begin to say good night (1600 PST = midnight GMT).  This, along with the changes in slope of the curve at different times, illustrates that median concurrency is the best number to look at, as it takes in the entire planet without respect to timezones… and Tateru has a number of different visualizations to aid us.

That’s an interesting one: median concurrency by day over the past year. You can even see the lack of logins on Christmas Day — but the feature I want to emphasize is that, in the last year, only twice has median daily concurrency reached 54,000.

Tateru also has, on the page labeled “(testing)”, longer-term looks at concurrency that span 6 years — three of them, at various grades of analysis (day/week/month). Here’s the weekly:

Notice that the center of the median band hovers at around 40,000, and has done so for approxmately two years. It had already slid down to that level from the all-time peak (around January 2009), when I mentioned it in July, 2010.

Now let’s look at those signup numbers. Rod was correct that the “20,000 per day” occurred during December 2011 — a qualification that seems to have been lost in the Twittersauce, so to speak:

And there was another spike right around Hallowe’en, but look how they’ve fallen off since. Vampires are sooooo last year…

One more graph from Tateru — total sign-ups over the last six months:

A fairly steady increase from ~25.5 million on October 1, 2011, to ~28.25 million on March 1, 2012; the bump in October/December is barely noticable at that scale. For the sake of keeping the math simple, let’s call it 3 million over six months, or 500,000 new accounts per month.

And yet, concurrency hasn’t moved much either way from 40,000 in the last two years or more.

Where are those new avatars???

Most of them, probably, are brand-new accounts frustrated with the complications of navigating inworld, or disappointed with the results of whatever enticed them to join, or who just don’t find Second Life to their liking once they’ve been there. So much for the alleged improvements in the New User Experience…

Some of them are probably alts of established Residents, created for any number of reasons: replacements for Resi’s who felt the need for a fresh start with a new persona; mannequins for clothing designers who need an opposite gender to try on new creations; “downline” for pyramid-scheme games like Bloodlines and Tiny Empires; RP characters kept separate from the “main”; or just for fun… Some of them might be “scripted agents”, the Lab’s euphemism for “bots”.

Some might be replacing me, and others like me: people whose priorities have changed, and don’t log in as often, or for as long, as they used to; people who have moved partially or totally to other grids; people who (unlike me) have ceased logging in altogether: avatars who have “gone nonconcurrent”.

Bottom line: The rate of replacement has done no better than equal the rate of attrition, with no change in total numbers of avatars inworld at a given time, for two years and more.

Is it possible that the niche market that is Second Life has been saturated? If so, it’s a good thing for the Lab’s investors — who, after all, are still looking for ROI — that Linden Lab is developing new products for different audiences. The audience they have now isn’t growing.



2 responses to “Going Nonconcurrent

  1. I left a comment about this on another page that wrote about it: http://www.eurogamer.net/articles/2012-03-15-pioneering-mmo-second-life-plots-a-comebackthis is what i said there:i think you should take these numbers with a grain of salt. firstly, we don't know what period these numbers are from. secondly, what is the retention? if all of these new signups don't log back in, second life has not made any net gains and the fact of the matter is that concurrency (a measure of currently logged in users) has been flat for quite some time now. linden lab has recently put the skids on publishing their economic stats and the numbers of users that have logged in once in the last sixty days (a metric used to judge the popularity of MMOs.) as to "profitability," one would expect the CEO to be a cheerleader for the company he works for. the $75 million "profit" number is an old one, mentioned some months ago and refers to gross revenue, not profit.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s