New World Notes

Great Web Resource For Second Life Economic Data

I've mentioned the valuable work of Resident data guru Tyche Shepherd, who tracks in-world "incident" reports, SL land ownership, and other metrics. Now she has a whole website offering that data and much more at Gridsurvey.com, taking available data put out by Linden in a human-readable form. (Much of that information is still available on the official Second Life site, but in this raw format.) Above for example is the number of Residents logging in over the past 30 days, charted back to 2006, which seems to have plateaued at around a million through 2009. Much more here. Image copyright Tyche Shepherd. Hat tip: Daniel Voyager.

Second Life Stat of the Day: Only 6% of SL Land Is "Adult" Rated For Pornographic Content

Awhile ago I mentioned the invaluable work of Tyche Shepherd, a Resident who meticulously tracks in-world Second Life data such as "Incident" reports. In a recent report posted on the SL Universe Forum, she has exhaustive details on the full scope of Second Life land, summarized according to content rating (PG, Mature, and Adult, a relatively new designation for SL content with graphic sex and violence). Here they are:

• Estate - Adult: 1303
• Estate - Mature: 19264
• Estate - Offline: 9
• Estate - PG: 3222
• Linden - Adult: 348
• Linden - Mature: 3835
• Linden - PG: 1422

"Adult regions," Tyche notes, "grew by 27 to 1651 (5.6%), PG Regions grew by 5 to 4644 (15.8%)."

New World Newsfeed: How To Misinterpret SL User Stats

Social media publicist Brian Solis has a widely-discussed post on Second Life user stats, which he dubs "The Second Life of Second Life". It mostly seems to be a re-hash of a Linden press release from last month, so little of it will surprise regular NWN readers. The analysis misses some important context, however, leading to serious misinterpretations of the user data Solis cites: Specifically, how many people regularly use Second Life? Fortunately, the Lindens have also published that data, along with user hours. Here's some recent data I wrote about last August, taken from a June 2009 sample:

This represents usage data of 742,000 unique returning users for June, along with their total user hours for the month. As you can see, 401K Residents logged into SL three hours or less, while on the opposite side of the spectrum, 133K Residents were responsible for almost 90% of total user hours.

That context in mind, you can see some of Solis' misinterpretations in sharp relief:

NWN Readers by Operating System: 26% Mac Users, 1.25% Linux (Probably True of Second Life User Base Too)

The Google Analytics chart above lists the readership of New World Notes, based on the operating system and browser you use. Over 1 in 4 of you are Mac users, which is significantly higher than average (total Mac OS market share is less than 5%), with Linux users slightly higher than average (1.25% of NWN readers, .95%, among all operating systems.) According to surveys I've conducted, nearly all of this blog's regular readership are active Second Life users, so this spread of operating systems is probably representative of SL's user base in general. (To my knowledge, the Lindens haven't published data on Resident operating systems, and if that's the case, this will have to do for now.) Much thanks to Troy McConaghy, who suggested that I look to Google Analytics for this OS info, after Opensource Obscure posed the question in an NWN forum thread.

Crime Scene: In Linden Incident Reports, Content Theft Violations Hardly Register

What you're looking at above is from an astounding compilation: It's the top twenty types of Resident violations of Community Standards/Terms of Service rules, as noted on the Lindens' Incident Report page. It covers virtual crimes from June 2007 up to now, with the total number of each kind counted above right. We have this information thanks to SL analyst Tyche Shepherd, and an anonymous volunteer. "I have a program which crawls the Incident Report page and stores the results in a database," Tyche tells me. "A lot of the back data came from someone who has been collecting the RSS feed from that page back to June 07."

Tyche has the full results with exhaustively compiled graphs available on the SL Universe community forum, and they count a total of 21,665 published incidents over a period of just over two years. I'd estimate about two-three million people have been in Second Life during that time frame, which means less than 1% the total user base for that period perpetrated these metaverse misdemeanors and felonies. As you can see above, most of the common violations fall in the garden variety griefer category; no surprise there. What is surprising, especially given recent controversies, is how few content theft-related incidents were reported.  You have to scroll down for those:

• Copyright Infringement 17
• Terms of Service: Trademark Violation 14
• Failure to Comply with Intellectual Property Notification 1

So over two years, just a few dozen content theft violations among tens of thousands. Several possible interpretations to that sparse number: Content theft incidents are under-reported, or under-enforced. Or, of course, content theft as an official, community-reported and Linden-enforced violation, are relatively rare.

Other interesting data points: Ageplay violations, a controversy which exploded in 2007, are also extremely rare: 182 total. Incidents of Camping Chairs, a more recent controversy, rarer still: 63. Strangest incident category: "Second Life: Respect, Pets 5". (Sion chicken slaughter, perhaps.) What's your read on this data?

Image credit: Tyche Shepherd. Hat tip: Daniel Voyager, who has more analysis here.

Second Life Users Chat and IM 600 Million Words... DAILY

The Lindens are reporting that SL Residents have since the world's 2003 launch amassed a billion user hours and a billion US$ in user-to-user transactions, but regular NWN readers will not be surprised by those numbers. What did surprise me is this stat on in-world text messages sent via IM/Chat:

Approximately 1,250 text-based messages are sent every second in Second Life, and more than 600 million words are typed on an average day.

To give you a sense of just how large that is, that's like writing 1,200 novels the length of War & Peace... a day. (And because there's roughly 400,000 daily log-ins, that means about 1,500 words per user per day, the equivalent of a mid-length magazine article.) If you could aggregate all that text, what do you suppose the most common proper nouns would be?

August's Top Traffic Sites in Second Life (Says Mappa Novus): Jazz, Clubbing, Sex, Romance, Beaches, Gaming

According to Avarie Parker of SLEntrepeneur Magazine [Warning: Some NSFW pics at that link] these are the ten Second Life sims with the heaviest genuine traffic last month, as compiled by SL analyst firm Mappa Novus -- genuine because they exclude locales with bots, camping chairs, and other free money giveaways. They do include two jazz nightclubs (Phat Cats, Frank's Place), a beach-themed hangout (Sexy Islands), a GBLT social space (Lesbian Paradise), two modern clubs (The 7th Element, SLeek), a romantic getaway spot (Midnight Reflection), two explicit adult-content sites (Laguna Bay Nude Beach, Chained Lust), and one roleplaying game-themed area (Purgatorio, for the Kingdom of Sand RPG.) Much more info and SLurls here, and archived reports here. Longtime NWN readers will remembers this blog once compiled bot-free locations, thanks to Tateru Nino, in something we called True Community Rank, and I'm glad someone else has taken up the call. Looking at the last list we posted, November 2007, we see Phat Cat's and Laguna Bay among the upper ranks even back then. Hat tip: The all-seeing eyes of Mal Burns.

Second Life Median Concurrency Trending Down Since March (Evidence of Traffic Bots Going Into Exile?)

Tateru Nino has some very interesting charts on her blog, tracking median concurrency numbers in Second Life, which she has been compiling (along with many other stats) since 2006. While we generally analyze an online service's daily peak concurrency rates, the median number is also a valuable data point, because it better suggests average usage throughout the day. (Compare to the peak concurrency numbers above right, from T. Linden's most recent economic report.) According to Tateru's figures, Second Life's median concurrency, which spiked to 78K in March, has been trending downward since.

But why since March?

Thousands of SL Groups on Flickr! (Way More Than WoW)

Here's a handy list of Flickr groups related to Second Life, listed on the official wiki. That's roughly 700 groups, but if you do a group search on Flickr itself, you come up with 2,235. (To be sure, some of those listed are false positives, but on a quick survey, most are clearly SL-related.) By comparison, a search for Flickr groups with "Warcraft" in the description (as in "World of") brings up just 329 groups. So roughly seven times the number. This is yet another example of Second Life's pervasiveness on Web 2.0 systems outside the world itself, and why in-world usage numbers only tell part of SL's story. About 750,000 returning users log into Second Life on a monthly basis, but how many more are also engaged in SL-centric activity on other online sites and systems? The numbers are enormous. Last August, for instance, the CEO of Japan's largest Second Life web/blog portal told me they got 15 million page views a month. Just on that one site alone.

Stat Shock: Just 133K Second Life Users Responsible For Nearly 90% Total Monthly SL User Hours!

Linden Chief Product Officer T Linden just released the latest Second Life economic stats, and they're worth a look. (Short version: In-world economy strong, but Lindens' anti-bot policy is slowing user growth stats.) The most striking chart to me, however, is what you see above: That's June's in-world activity rates among unique logins, with some annotations based on another chart, where T reports that June saw 742,000 unique users logging into SL.  (Slightly down from a high of 752K in May.) As he notes with emphasis, "the 18% of monthly users who spend more than 50 hours per month in Second Life drive 87% of the user hours." Or to interpret these numbers in another way with equal emphasis:

In June, about 133,000 Second Life users were responsible nearly 90% of all in-world Second Life activity.

On the very high end, that's 22K people who are in-world 300 hours or more per month -- i.e., on average, ten-plus hours a day. But the low end is what concerns me far more: 401K unique users are not even staying in Second Life more than three hours total. As T notes with wry understatement befitting a green frog, "This also points up the opportunity to increase the engagement of new users." To me, it's yet more evidence that an achievement system is needed to increase engagement rates. At this point, less than 350K Second Life users are logging in long enough to be a regular part of the in-world economy and the community.

Note, by the way, I keep saying "in-world" as a qualifier, because I don't think it's necessarily true that those 401K extremely light users are casual noobs.