How many bots are in-world, and how much do they artificially boost a parcel's Traffic numbers? The aggravating mystery continues, but here's another data point: Tabliopa Underwood reports that she's been covertly watching a landowner experimenting with a tent of snuggling (!) bots. "I can now say with some certainty," she writes, "that a 24/7 online traffic bot generates 1440 dwell points every 24 hours. One point a minute." Does that sound accurate?
Post a comment
Your Information
(Name is required. Email address will not be displayed with the comment.)
Yes it is a well known fact that one avatar on a sim 24 hours a day will produce a traffic number of exactly 1440. One point per minute. I proved this a long time ago on my sim before it was opened to the public by using sleek and an alt account. The numbers also jive with the numbers associated with bot farm sims.
And again bots/camping is not the problem. Having traffic be tied to search relevance is the problem. Linden Lab needs to disconnect the traffic from search. In addition the profile pick thing is also falsified because people pay for those. It really doesn't matter what Linden Lab tries to do on the topic of search. Whatever they try it will be gamed. So the real improvement will come when Linden Lab produces a TOS that covers bot use in very explicit terms and then enforces the policy.
I know plenty of people that would have paid camping if traffic was not tied to search relevance.
Posted by: Ann Otoole | Friday, January 02, 2009 at 03:27 PM
Linden Lab said it absolutely does not work this way. An official statement from the Lab last year was that each user (including bots) generates exactly one point of parcel traffic per day, which is then divided among the parcels they visit.
Actually, that's provably rubbish, because you can easily demonstrate that that is not so, which leads me to suspect that either the system is buggy (that is, it is not functioning as they intend) or they don't actually understand what the system is doing. Perhaps both.
1440 (1/min) seems to be accurate only in the case of avatars that do not move across a parcel boundary at all. Avatars that move around seem to accrue traffic at a far different (and unrelated) rate.
Posted by: Tateru Nino | Friday, January 02, 2009 at 06:17 PM
Tateru:
Considering the vintage of the dwell system upon which traffic is calculated and determined, I put it to interested parties that the system is basically a poorly (if even) documented black box whose originator no longer works at Linden Lab. A real maintenance nightmare, to put it bluntly.
Traffic needs to die in a fire preferably at the soonest feasible opportunity that Linden Lab can determine what needs to be excised from server code to remove unneeded traffic from such a change.
Posted by: Patchouli Woollahra | Saturday, January 03, 2009 at 11:28 PM
Bots should be forbidden!
Posted by: Knusper | Sunday, January 04, 2009 at 07:12 AM
Bots should not be forbidden!
Posted by: Repsunk | Monday, January 05, 2009 at 03:37 AM
"Linden Lab said it absolutely does not work this way. An official statement from the Lab last year was that each user (including bots) generates exactly one point of parcel traffic per day, which is then divided among the parcels they visit."
[citation needed] (Which is to say, I believe this is false; provide a link to this official statement if it's true. I suspect is you just happened to read last year the original description that they posted many years ago, which is by their own subsequent statements no longer accurate, and your summary of it is incorrect in any case.)
To the best of my knowledge, LL has not given any details on how traffic works since 2005, and those details are clearly stale. The original implementation was entirely too complicated and easy to game, so their original description was invalidated within months as they started fiddling with it. I believe they eventually gave up and simply decided the number of avatar-minutes was simply more intuitive and less abusable than the byzantine alternatives.
Posted by: Gaius Goodliffe | Tuesday, January 06, 2009 at 10:00 AM
I only report what I observed. The bots never left the parcel and they not even move. They just snuggle down forever in the same positions =)
Like Tateru Nino say I think is maybe a different formula for agents that move. Maybe even a difference between parcels on same sim and parcels on different sims.
Maybe be a useful test to have a bot walk from one parcel to another parcel and back again. As has been said by others is a 3 minute (or maybe 5 or something depending on who you talk to) minimum period before dwell points are counted.
Be interesting to see what happens if the agent walk between 2 parcels and stay on each for less than 3 minutes. Would either parcel get dwell points ??? Maybe they get none ??? And if so then what about the dwell points of an avatar that do same thing ??? Maybe none as well even after 2 or 3 hours inworld jumping about all over the place within the minimum window. Although I dont see how you could track the dwell if you just random jumping. Could only track that accurately really by owning the parcels and locking everyone else out. But it could be done and maybe someone into bots can test that and let us know.
I do think though that if a bot walk back and forward across a sim boundary then it maybe get logged off after a while. Like what happened in the Varvello Picconi Diot Biersack project. Although they say in their paper that their agents got logged off when they TP to many times to quickly. Maybe same thing happen in a cross sim boundary walk. Maybe. Probably. Dont know.
Posted by: Tabliopa Underwood | Thursday, January 08, 2009 at 12:19 AM
I think your blog need a new wordpress template. Downalod it from http://genericwpthemes.com . The site has nice and unique wordpress templates.
Posted by: WP Themes | Wednesday, January 21, 2009 at 11:10 PM
I must say that you provide genuine, quality information. Thanks for this!
BTW, dpn't you think your blog needs a better wordpress template?
Posted by: Applebees Menu | Saturday, February 28, 2009 at 03:12 AM