population density vs. time data wanted

Hi all,

Does anyone have data (or an informed guess, at least) as to what percentage of the SL population tends to be logged in to the grid at any given time, within general times of day, days of week? I'm trying to work out a business model for a service on a purchased island, and before I go any further, I need to know how many customers it's reasonable to sell monthly access to (if we go that way), so that the number of them likely to log in at any given time won't exceed the magical number of 35.

For example (all times PST, assuming most of the customers are in that time zone, even though I'm not):

Weekday during PST workday:
Weekday evenings 5-9 PST:
Weekday Night 9pm-1am PST:
Weekday "graveyard": 1am-7am PST:

Weekend workday:
Weekend evening:
Weekend night:
Weekend graveyard:

Any advice?

Thanks,

neko

PS - I plan to post this to Linden Hotline, as well, but I value the input from this group.

This is probably commercially sensitive data. I wouldn't expect a response mate!

You bet it's commercially sensitive -- to me as well as to the Lindens. The reason I posted it in both the Hotline and here is in hopes that someone with more experience than I would be able to give advice, Linden or not.

If the Lindens want to encourage people to develop for-fee service areas (as is implied by the ability to sell passes to parcels of land), and there are sharp limits on the number of avs that can be in a parcel at a time, they need to provide this data so prospective developers can plan. If they won't (which wouldn't be surprising), I'm hoping someone else who has a high-traffic area and visitor tracking will be willing to help me out.

It would be reasonable to charge me for the data, of course. If someone believes they have good reliable data and wants to set a price, IM me and we'll talk. I probably won't pay much for a SWAG -- but if you've got aggregated visitor tracking data spanning several months that you'd be willing to sell, I think we can work something out.

neko

Neko,

Ezhar Fairlight has some interesting data that might be of use to you here:

http://sl.sdfjkl.org/userdata/

It's not as granular as you're looking for, but maybe if Ezhar is hard up for cash these days you can convince him to parse it for you.

Whats' the service you're planning?

Thanks, Walker. Ezhar's data did help some. I'll try to contact him to see if he is interested in providing more detail with respect to time breakdowns. It's clear that he has the data, but it would be extra work for him to parse it.

The service is, as noted constantly in my sig a biofeedback-based game. See the link below for a general description. In addition to this page, I have a cost-income model I've sent to a couple of people, but some of it depends on how many people we could reasonably hope to subscribe, and I hadn't considered the upper bound of how many avs can be logged into the same sim at once. The other technical issues are relatively minor and mostly resolved at this point. But if the peaks of online activity are too sharp, the model shows it will be difficult to pay development costs off in a reasonable period of time. (Even if I couldn't get the funding to pay developers real-world prices up front, I'd like to at least commit to pay "volunteers" back proportionally if the service generates income.)

If I can get the data together to finalize the model, and get enough backing to at least make the SL land and other related server arrangements, I'll be posting here again and also in "Help Wanted". I'll be looking for a number of scripters, builders, texturers, etc.

neko

This may be over simplistic, but have you thought of counting them?

The in-world map takes a snapshot of online users, hover your mouse over a sim and it tells you how many are in it. I know there are a lot of sims to add up, but a good estimate could be made by sampling say 20 and extrapolating.

You could deliberately avoid the sims with clubs and the empty sims as well to get a more acurate result.

Surina,

It's a good idea. If I don't turn up someone with better data, I'll try it. There are an awful lot of sims up now, but you're right, random sampling of 20 or so should answer the question reasonably well.

neko