Second Life data update

Good news:

Second Life total signups is back on the air. The feed lit up again on 4 May. The weekly signup chart will come back online as soon as there's enough data. The figure presently reads 15,348,641. Pre-purge that was 16,753,668. How many accounts were purged is unknown, but shouldn't be hard to estimate with some reasonable accuracy.

Not so good news:

Linden Lab is only providing updated concurrency data twice per hour. Once at 5 minutes past the hour, and again five minutes after that (ten minutes past the hour). Then nothing for another 45 minutes. That means the 30-minute deltas are pretty useless. There's not enough data to chew on.

Technology is awesome.

Anyway, right now you'll see a lot fewer red sections in the charts. There have been a few internal structural changes as well, but they likely won't affect chart viewers in any significant way (except to save you a little bandwidth).

Concurrency feeds running again

Second Life concurrency data has just flickered back to life. Looks like it was a bug after all. The Lab apparently set to fixing the problem just as soon as the issue was discovered. The charts should start updating within the next quarter hour or so.

Concurrency stats offline

The live data feed for Second Life concurrency is off the air. It went dark on the afternoon of Thursday, 30 April, and hasn't been updated since.

Normally, I'd think of this as just another glitch in the statistical feeds. Heaven knows, there have been plenty of those -- having some missing data is not uncommon.

The timing got me thinking, though.

You see, Linden Lab purged a whole bunch of inactive, never-logged-on accounts (somewhere around the 800,000) mark. And just prior to doing so, the feed for the total number of accounts went dark - and has remained that way ever since.

30 April, coincidentally marked the day that the Lab was going to start taking action against traffic bots. Everyone, of course, was expecting to see concurrency take a bit of a hit, though it wasn't clear how much of a hit. Hard to tell, though, because the concurrency stats went dark that very day.

I'd love to think this was a coincidence, but the available history makes that an unreasonable conclusion.

Live feeds/dead feeds

No word about the broken statistical feeds. On the JIRA side, Coco Linden reopened an old JIRA ticket about it. On the PR side, still waiting for a response.

The end of April is coming along, so I'm going to assume that the feeds are shut down as a matter of policy (rather than say incompetence). They're not getting fixed, and they're not getting talked about, even though statistics are quite the PR focus at the moment. Even if they were a lower priority than the slightly balky wheel on Harmony's chair, you'd have at least expected to hear about that by now.

What's left, but to assume that the change is intentional? (Or that the breakage wasn't, but leaving it broken is?)

I've updated the charts page, moved live charts towards the top, and marked the defunct information more clearly.

CORRECTION: The Last-60-days active user data sputtered erratically back to life a few days ago, and is reporting again. My scripts, alas, ignored it because the signup data itself had ceased. That graph is working again - though the feed has been silent for almost all of Q1 2009.

Some statistical feeds continue offline

I've been getting a lot of email since January about the statistical graphs, particularly the users-logged-in-over-the-last-60-days figure and the total signups. Fact is, that those parts of the statistical feeds from Linden Lab were frozen in January (around the 13th/14th of that month) and have not been updated since, though the data is still flowing for other parts of the feeds.

I've made some queries about it, but there has been no response. It's anyone's guess as to why the data is unavailable, especially as it happened shortly before an interview with Mark Kingdon, where he was talking about the quality and quantity of the available statistical data.

It's been more than three weeks of downtime on this data, so if it isn't an intentional outage, I guess it isn't a priority item to fix.

Mixed Reality Headcount, Week ending 11 November, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 11 November, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	70	42	11,772 (stable)
Gossip Girl	56	6.29	9,408 (new!)
IBM	46	60	7,764 (up 11.17%)
I Am Legend	37	31	6,304 (down 4.60%)
The L Word	26	32	4,524 (stable)
Pontiac	26	28	4,368 (stable)
Greenies Home	24	34	4,176 (down 3.87%)
Playboy	23	34	3,936 (up 19.71%)
The Weather Channel	17	18	2,988 (up 5.51%)
Virtual Holland	17	24	2,928 (down 2.79%)
Nissan	15	8.57	2,604 (up 5.85%)
Sun Microsystems	11	13	1,932 (down 13.44%)
Microsoft	10	9.71	1,800 (up 11.11%)
ABC Island	10	7.43	1,740 (down 2.03%)
AOL Pointe	4.07	2.86	684 (up 14.00%)
STA Travel	3.57	4.29	600 (down 19.35%)
Cisco Systems	3.29	3.14	552 (up 12.20%)
Manpower HQ	2.58	2.86	434 (down 34.04%)
Coldwell Banker	2.57	3.71	432 (down 5.26%)
NBA	2.21	1.14	372 (down 11.43%)
Dell	1.86	1.14	312 (down 7.14%)
Reebok	1.57	2.29	264 (stable)
Comcast	0.86	0.86	144 (up 33.33%)
Armani	0.71	0.29	120 (down 16.67%)

The top ten, with commentary is posted to New World Notes.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day, seven days per week for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour (essentially, every 6 hours). For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 4 November, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 4 November, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	69	38	11,664 (up 6.81%)
IBM	41	39	6,984 (down 3.96%)
I Am Legend	39	10	6,608 (new!)
The L Word	26	30	4,500 (down 8.98%)
Greenies Home	25	36	4,344 (stable)
Pontiac	25	33	4,296 (down 7.73%)
Playboy	19	26	3,288 (down 10.75%)
Virtual Holland	17	26	3,012 (down 17.43%)
The Weather Channel	16	18	2,832 (down 4.84%)
Nissan	14	14	2,460 (down 7.24%)
Sun Microsystems	13	16	2,232 (up 11.38%)
ABC Island	10	10	1,776 (up 2.07%)
Microsoft	9.64	10	1,620 (stable)
STA Travel	4.43	4.0	744 (down 12.68%)
Manpower HQ	3.92	4.86	658 (up 56.67%)
AOL Pointe	3.57	2.29	600 (down 5.66%)
Cisco Systems	2.93	2.86	492 (up 2.50%)
Coldwell Banker	2.71	3.43	456 (up 5.56%)
NBA	2.5	2.0	420 (down 16.67%)
Dell	2.0	1.71	336 (down 20.00%)
Reebok	1.57	2.29	264 (down 8.33%)
Armani	0.86	0.86	144 (down 14.29%)
Comcast	0.64	1.14	108 (up 12.50%)

The top ten, with commentary is posted to New World Notes.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 28 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 28 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	65	32	10,920 (stable)
IBM	43	48	7,272 (up 10.58%)
The L Word	29	33	4,944 (down 4.41%)
Pontiac	27	23	4,656 (up 9.30%)
Greenies Home	25	31	4,344 (stable)
Playboy	21	32	3,684 (up 35.24%)
Virtual Holland	21	37	3,648 (up 41.40%)
The Weather Channel	17	21	2,976 (up 5.08%)
Nissan	15	12	2,652 (up 6.76%)
Sun Microsystems	11	15	2,004 (down 10.22%)
ABC Island	10	7.71	1,740 (up 13.28%)
Microsoft	9.71	11	1,632 (up 13.33%)
STA Travel	5.07	7.71	852 (up 51.06%)
AOL Pointe	3.79	2.57	636 (down 3.64%)
NBA	3.0	2.0	504 (up 55.56%)
Cisco Systems	2.86	1.71	480 (down 9.09%)
Useful Technology	2.71	0.29	456 (down 13.64%)
Coldwell Banker	2.57	3.71	432 (up 5.88%)
Dell	2.5	2.29	420 (up 34.62%)
Manpower HQ	2.5	2.0	420 (down 3.23%)
Reebok	1.71	3.14	288 (up 4.35%)
Armani	1.0	1.43	168 (stable)
Comcast	0.57	1.14	96 (up 14.29%)
Coldwell Banker:$3m house	0.08	0.29	14 (stable)

The top ten, with commentary is posted to New World Notes. Native sites aren't in this chart, and will be posted separately.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Native Site Community Headcount, Week ending 21 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 21 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
NORSIM MultiSIM (Goth RPG)	298	361	50,160 (up 5.66%)
Avilion Grove (Fantasy Roleplaying)	149	172	25,072 (stable)
Phat Cat's Jazzy Blue Lounge (PG Social/Shopping)	140	156	23,556 (up 15.54%)
Blackheart's Cafe (PG Social)	93	105	15,760 (up 3.47%)
Nude Island (Mature Social)	92	69	15,600 (down 34.25%)
Laguna Nude Beach (Mature Social)	91	111	15,344 (up 15.13%)
Bound and Determined (BDSM Social)	90	104	15,248 (stable)
Heavenly Rose Gardens (PG Social)	85	132	14,400 (up 11.52%)
Lost Gardens of Apollo (PG Social)	82	108	13,860 (stable)
Independent State of Caledon (Steampunk Roleplay)	81	92	13,740 (down 3.13%)
Toxian City (Goth RPG)	80	112	13,584 (up 42.69%)
Hedonistic Isle (Mature Social)	80	93	13,584 (stable)
Midian City (Goth RPG)	79	97	13,344 (up 2.21%)
Capture Roleplay HQ (BDSM Social)	78	90	13,216 (stable)
Bare Rose (PG Social)	78	82	13,168 (stable)
City of Lost Angels (Goth RPG)	75	80	12,684 (up 5.07%)
New Citizens Incorporated (Noob Resident Welcome/Education)	71	90	11,988 (up 4.17%)
Icedragon's Playpen (Casual Gaming/Social)	67	80	11,312 (up 4.43%)
City of Midgaard (BDSM Social)	58	69	9,888 (up 5.28%)
The Shelter (Noob Resident Welcome/Education)	54	67	9,084 (up 2.02%)
Isle of Lesbos (Mature Social)	52	73	8,796 (up 2.52%)
Midgar (Final Fantasy VII RPG)	36	42	6,060 (down 11.40%)
London Commercial Sims (UK urban community/Social)	26	36	4,464 (stable)
Ballers City (Urban basketball Social)	20	18	3,432 (down 22.07%)
Svarga Svarga (Artificial Ecosystem)	18	27	3,108 (stable)
New Brussel Mega Mall (PG Social/Shopping)	3.24	5.14	544 (down 20.93%)

The top ten, with commentary is posted to New World Notes. The list is by no means exhaustive. There are thousands of sites - far more than we can track.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 21 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 21 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	64	34	10,824 (down 7.68%)
IBM	39	37	6,576 (down 8.05%)
The L Word	30	35	5,172 (up 4.61%)
Greenies Home	26	45	4,416 (down 4.91%)
Pontiac	25	28	4,260 (up 7.90%)
The Weather Channel	16	14	2,832 (down 3.67%)
Playboy	16	16	2,724 (down 8.84%)
Virtual Holland	15	18	2,580 (up 27.22%)
Nissan	14	9.71	2,484 (down 4.61%)
Sun Microsystems	13	14	2,232 (up 43.08%)
ABC Island	9.14	6.57	1,536 (down 12.33%)
Microsoft	8.57	8.29	1,440 (down 2.44%)
AOL Pointe	3.93	2.86	660 (up 17.02%)
STA Travel	3.36	2.57	564 (up 2.17%)
Cisco Systems	3.14	2.0	528 (up 4.76%)
Useful Technology	3.14	0.29	528 (up 76.00%)
Manpower HQ	2.58	1.43	434 (up 14.81%)
Coldwell Banker	2.43	2.57	408 (down 17.07%)
NBA	1.93	1.43	324 (down 28.95%)
Dell	1.86	0.57	312 (up 23.81%)
Reebok	1.64	2.86	276 (down 11.54%)
Comcast	0.5	0.86	84 (stable)
Coldwell Banker:$3m house	0.0	0.0	0 (down 100.00%)

The top ten, with commentary is posted to New World Notes.  Native sites aren't in this chart, and will be posted separately.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Native Site Community Headcount, Week ending 14 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 14 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
NORSIM MultiSIM (Goth RPG)	282	354	47,472 (up 7.71%)
Avilion Grove (Fantasy Roleplaying)	147	176	24,704 (down 8.43%)
Nude Island (Mature Social)	141	144	23,728 (down 10.34%)
Phat Cat's Jazzy Blue Lounge (PG Social/Shopping)	121	123	20,388 (down 15.30%)
Blackheart's Cafe (PG Social)	90	90	15,232 (stable)
Bound and Determined (BDSM Social)	90	103	15,120 (up 2.13%)
Independent State of Caledon (Steampunk Roleplay)	84	108	14,184 (up 6.10%)
Lost Gardens of Apollo (PG Social)	82	100	13,836 (stable)
Hedonistic Isle (Mature Social)	80	92	13,584 (down 9.09%)
Laguna Nude Beach (Mature Social)	79	96	13,328 (stable)
Bare Rose (PG Social)	79	82	13,328 (up 6.19%)
Capture Roleplay HQ (BDSM Social)	79	88	13,280 (down 2.90%)
Midian City (Goth RPG)	77	90	13,056 (down 11.81%)
Heavenly Rose Gardens (PG Social)	76	132	12,912 (stable)
City of Lost Angels (Goth RPG)	71	78	12,072 (down 3.64%)
New Citizens Incorporated (Noob Resident Welcome/Education)	68	70	11,508 (up 2.46%)
Icedragon's Playpen (Casual Gaming/Social)	64	76	10,832 (up 9.83%)
Toxian City (Goth RPG)	56	79	9,520 (down 12.74%)
City of Midgaard (BDSM Social)	55	72	9,392 (down 12.47%)
The Shelter (Noob Resident Welcome/Education)	53	77	8,904 (up 4.95%)
Isle of Lesbos (Mature Social)	51	81	8,580 (stable)
Midgar (Final Fantasy VII RPG)	40	60	6,840 (up 7.75%)
Ballers City (Urban basketball Social)	26	33	4,404 (down 2.39%)
London Commercial Sims (UK urban community/Social)	26	38	4,392 (down 6.15%)
Svarga Svarga (Artificial Ecosystem)	18	28	3,156 (stable)
New Brussel Mega Mall (PG Social/Shopping)	4.1	4.86	688 (up 56.01%)

The top ten, with commentary is posted to New World Notes. 

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 14 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 14 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	69	43	11,724 (stable)
IBM	42	42	7,152 (stable)
The L Word	29	36	4,944 (up 19.77%)
Greenies Home	27	38	4,644 (up 8.10%)
Pontiac	23	21	3,948 (up 2.17%)
Playboy	17	33	2,988 (up 19.71%)
The Weather Channel	17	22	2,940 (up 4.70%)
Nissan	15	12	2,604 (up 7.43%)
Virtual Holland	12	8.86	2,028 (up 3.05%)
ABC Island	10	10	1,752 (up 7.35%)
Sun Microsystems	9.29	14	1,560 (down 11.56%)
Microsoft	8.79	11	1,476 (up 21.78%)
AOL Pointe	3.36	2.0	564 (down 6.00%)
STA Travel	3.29	1.43	552 (up 21.05%)
Cisco Systems	3.0	1.71	504 (up 5.00%)
Coldwell Banker	2.93	3.71	492 (up 2.50%)
NBA	2.71	2.57	456 (up 18.75%)
Manpower HQ	2.25	2.0	378 (down 23.90%)
Reebok	1.86	3.71	312 (up 23.81%)
Useful Technology	1.79	0.29	300 (down 24.24%)
Dell	1.5	0.86	252 (down 22.22%)
Comcast	0.5	0.57	84 (down 50.00%)
Coldwell Banker:$3m house	0.08	0.29	14 (stable)

The top ten, with commentary is posted to New World Notes.  Native sites aren't in this chart, and will be posted separately.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 7 October, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 7 October, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	68	33	11,568 (stable)
IBM	42	42	7,104 (up 2.07%)
Greenies Home	25	36	4,296 (down 2.19%)
The L Word	24	23	4,128 (down 21.46%)
Pontiac	23	20	3,864 (stable)
The Weather Channel	16	17	2,808 (up 2.63%)
Playboy	14	25	2,496 (up 10.05%)
Nissan	14	12	2,424 (up 2.54%)
Virtual Holland	11	8.0	1,968 (down 7.34%)
Sun Microsystems	10	18	1,764 (up 24.58%)
ABC Island	9.71	8.57	1,632 (up 11.48%)
Microsoft	7.21	4.57	1,212 (down 13.68%)
AOL Pointe	3.57	2.86	600 (down 15.25%)
Manpower HQ	2.96	2.86	497 (down 20.93%)
Coldwell Banker	2.86	4.29	480 (down 6.98%)
Cisco Systems	2.86	2.0	480 (down 21.57%)
STA Travel	2.71	2.0	456 (down 32.14%)
Useful Technology	2.36	4.0	396 (up 57.14%)
NBA	2.29	2.57	384 (up 3.23%)
Dell	1.93	2.86	324 (down 18.18%)
Reebok	1.5	2.29	252 (down 8.70%)
Comcast	1.0	1.14	168 (down 41.67%)
Coldwell Banker:$3m house	0.0	0.0	0 (down 100.00%)

The top ten, with commentary is posted to New World Notes.  Native sites aren't in this chart, and will be posted separately.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 30 September, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 30 September, 2007:

Site	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
The Pond	68	28	11,436 (down 3.15%)
IBM	41	43	6,960 (down 5.54%)
The L Word	31	38	5,256 (up 6.57%)
Greenies Home	26	40	4,392 (up 3.10%)
Pontiac	23	19	3,912 (stable)
The Weather Channel	16	16	2,736 (down 18.28%)
Nissan	14	11	2,364 (down 9.22%)
Playboy	13	21	2,268 (down 3.08%)
Virtual Holland	12	10	2,124 (stable)
ABC Island	8.71	7.43	1,464 (down 10.29%)
Sun Microsystems	8.43	16	1,416 (up 4.42%)
Microsoft	8.36	6.29	1,404 (up 9.35%)
AOL Pointe	4.21	4.0	708 (down 29.76%)
STA Travel	4.0	6.86	672 (up 19.15%)
Manpower HQ	3.74	4.29	628 (up 10.26%)
Cisco Systems	3.64	4.57	612 (down 10.53%)
Coldwell Banker	3.07	4.57	516 (stable)
Ratepoint	2.64	8.0	444 (down 27.45%)
Dell	2.36	2.0	396 (down 31.25%)
NBA	2.21	1.14	372 (down 3.12%)
Comcast	1.71	0.57	288 (down 7.69%)
Reebok	1.64	3.43	276 (up 27.78%)
Useful Technology	1.5	3.14	252 (up 40.00%)
Coldwell Banker:$3m house	0.09	0.0	15 (down 50.00%)

The top ten, with commentary is posted to New World Notes.  Native sites aren't in this chart, and will be posted separately.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 23 September, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 23 September, 2007:

Site (* Native reality site)	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
* Phat Cat's Jazzy Blue Lounge	141	153	23,772 (stable)
* Lost Gardens of Apollo	81	100	13,692 (down 5.00%)
* City of Lost Angels	76	73	12,828 (stable)
The Pond	70	34	11,808 (down 3.62%)
* New Citizens Incorporated	69	76	11,688 (down 7.33%)
* Independent State of Caledon	68	87	11,496 (down 11.21%)
* The Shelter	52	68	8,808 (down 4.68%)
* Isle of Lesbos	49	72	8,256 (up 14.67%)
IBM	43	45	7,368 (down 2.38%)
* Midgar	41	52	7,020 (stable)
* London Commercial Sims	31	43	5,208 (down 2.91%)
The L Word	29	30	4,932 (stable)
Greenies Home	25	38	4,260 (down 4.57%)
Pontiac	23	20	3,912 (down 31.80%)
* Ballers City	21	24	3,684 (down 10.76%)
The Weather Channel	19	21	3,348 (up 7.31%)
* Svarga	19	25	3,336 (stable)
Nissan	15	9.14	2,604 (down 7.26%)
Playboy	13	14	2,340 (down 5.34%)
Virtual Holland	12	10	2,148 (down 11.39%)
ABC Island	9.71	13	1,632 (up 37.37%)
Sun Microsystems	8.07	12	1,356 (up 22.83%)
Microsoft	7.64	6.86	1,284 (up 2.88%)
AOL Pointe	6.0	7.43	1,008 (up 18.31%)
Cisco Systems	4.07	6.0	684 (up 32.56%)
Ratepoint	3.64	12	612 (up 240.00%)
Dell	3.43	5.14	576 (up 14.29%)
Manpower HQ	3.39	3.43	570 (down 4.88%)
STA Travel	3.36	3.43	564 (down 2.08%)
Coldwell Banker	3.07	4.29	516 (down 8.51%)
NBA	2.29	1.14	384 (up 3.23%)
Comcast	1.86	2.0	312 (up 13.04%)
Reebok	1.29	2.29	216 (up 20.00%)
Useful Technology	1.07	2.0	180 (down 40.00%)
Coldwell Banker:$3m house	0.17	0.29	29 (down 60.00%)
Adidas	0.0	0.0	0 (down 100.00%)

The top ten, with commentary is posted to New World Notes.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 16 September, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 16 September, 2007:

Site (* Native reality site)	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
* Phat Cat's Jazzy Blue Lounge	142	159	23,964 (up 3.47%)
* Lost Gardens of Apollo	85	109	14,412 (up 2.56%)
* Independent State of Caledon	77	99	12,948 (up 2.18%)
* City of Lost Angels	75	87	12,636 (down 6.07%)
* New Citizens Incorporated	75	89	12,612 (stable)
The Pond	72	34	12,252 (stable)
* The Shelter	55	67	9,240 (up 7.24%)
IBM	44	46	7,548 (up 7.34%)
* Isle of Lesbos	42	56	7,200 (down 11.76%)
* Midgar	41	49	6,924 (down 8.12%)
Pontiac	34	42	5,736 (up 26.79%)
* London Commercial Sims	31	44	5,364 (down 4.49%)
The L Word	29	35	5,016 (down 6.49%)
Greenies Home	26	38	4,464 (up 7.51%)
* Ballers City	24	27	4,128 (stable)
* Svarga	19	27	3,288 (up 3.01%)
The Weather Channel	18	16	3,120 (up 4.84%)
Nissan	16	16	2,808 (up 18.18%)
Playboy	14	21	2,472 (up 25.61%)
Virtual Holland	14	16	2,424 (up 3.06%)
Microsoft	7.43	8.0	1,248 (down 9.57%)
ABC Island	7.07	6.57	1,188 (down 26.67%)
Sun Microsystems	6.57	8.86	1,104 (down 5.15%)
AOL Pointe	5.07	4.86	852 (up 2.90%)
Manpower HQ	3.57	3.71	599 (down 2.38%)
STA Travel	3.43	3.43	576 (up 11.63%)
Coldwell Banker	3.36	4.86	564 (down 2.08%)
Cisco Systems	3.07	2.29	516 (down 10.42%)
Dell	3.0	3.14	504 (up 2.44%)
NBA	2.21	1.14	372 (down 6.06%)
Useful Technology	1.79	0.86	300 (down 53.70%)
Comcast	1.64	0.86	276 (down 28.13%)
Ratepoint	1.07	1.43	180 (down 74.14%)
Reebok	1.07	1.43	180 (up 150.00%)
Coldwell Banker:$3m house	0.43	1.14	73 (up 400.00%)
Adidas	0.21	0.0	36 (stable)

The top ten, with commentary is posted to New World Notes.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.

Mixed Reality Headcount, Week ending 9 September, 2007.

Here is the full ranked table of tracked sites for the week ending Sunday 9 September, 2007:

Site (* Native reality site)	Est avg hourly visits	Est avg hourly visits (peak hrs)	Estimated total weekly visits
* Phat Cat's Jazzy Blue Lounge	137	158	23,160 (stable)
* Lost Gardens of Apollo	83	114	14,052 (down 3.06%)
* City of Lost Angels	80	87	13,452 (stable)
* Independent State of Caledon	75	112	12,672 (stable)
* New Citizens Incorporated	74	89	12,588 (stable)
The Pond	72	40	12,252 (stable)
* The Shelter	51	59	8,616 (down 6.75%)
* Isle of Lesbos	48	68	8,160 (up 3.50%)
* Midgar	44	63	7,536 (down 4.99%)
IBM	41	41	7,032 (up 2.09%)
* London Commercial Sims	33	47	5,616 (up 7.59%)
The L Word	31	36	5,364 (up 3.23%)
Pontiac	26	27	4,524 (down 20.46%)
* Ballers City	25	30	4,212 (down 6.15%)
Greenies Home	24	32	4,152 (stable)
* Svarga	19	21	3,192 (stable)
The Weather Channel	17	18	2,976 (down 12.68%)
Nissan	14	8.86	2,376 (down 6.16%)
Virtual Holland	14	12	2,352 (down 7.55%)
Playboy	11	5.14	1,968 (stable)
ABC Island	9.64	9.14	1,620 (up 7.14%)
Microsoft	8.21	6.0	1,380 (stable)
Sun Microsystems	6.93	8.57	1,164 (down 22.40%)
AOL Pointe	4.93	4.57	828 (up 16.95%)
Ratepoint	4.14	12	696 (up 18.37%)
Useful Technology	3.86	10	648 (up 42.11%)
Manpower HQ	3.65	3.14	614 (down 6.67%)
Coldwell Banker	3.43	4.29	576 (down 17.24%)
Cisco Systems	3.43	4.29	576 (stable)
STA Travel	3.07	2.29	516 (down 18.87%)
Dell	2.93	4.29	492 (stable)
NBA	2.36	1.43	396 (up 3.13%)
Comcast	2.29	4.29	384 (down 17.95%)
Reebok	0.43	0.29	72 (up 20%)
Adidas	0.21	0.0	36 (stable)
Coldwell Banker:$3m house	0.09	0.0	15 (down 75%)

The top ten, with commentary is posted to New World Notes.

Methodology

Mixed reality sites in this headcount are selected for their prominence, either from publicity or real world name recognition. Sites with consistent low traffic (500 or less weekly) may be dropped in future Headcounts in favor of other sites. We do not count sites with camping chairs, or visitors in the orientation sims, as there seems to be little evidence to suggest that they will become visitors to the parent site - and if they do, we catch them when we headcount the site anyway.

We collect data four times per day for each site at 2am, 8am, 2pm and 8pm (times in SLT/US Pacific) plus/minus 1 hour. For each sample we count the number of people at the site at the time. We average those samples across the week, and then assume that average to hold constant, with each visitor spending a half hour on-site. This methodology does not necessarily include one-time events that generate high traffic missed by our sampling, which we'll make note of whenever possible. Headcounts do not factor in returning visitors, so assume that the total number of unique Residents are likely to be significantly less than the estimated total visits.