traffic and what my ticket result was

Do the tests and find out. You'll find that, for 60 minutes, the traffic number will be 60, and so on. But be careful not to straddle the moment when the day's traffic is taken or your tests will appear to produce inconsistent results, even though they don't.

Every Linden does not know everything about everything, and some/many don't know how traffic is calculated so all they've got are the 'official' documents. Of course they are going to think the documents are right, and reply accordingly - the documents are supposed to be right.

That was probably to increase the count. According to the wiki, it's possible to reach above the maximum by relogging into a sim. I.e. a single av on the parcel 24/7 will produce 1440 traffic points per day, but apparently it's possible to get more than 1440 by moving out of the sim and into it again - relogging into the sim.

All the KB article says is:

"Traffic is a number for each parcel which is based on the amount of Residents who visited, and the time spent on that parcel out of their total time inworld that day. It's calculated using a complex algorithm."

Well it says a bit more than that but that's the important part and it doesn't really say a fat lot. That paragraph suits both the dwell description and the one minute per point description.

If I'm on for 100 minutes and I spend 60 of them on one parcel, that parcel would get 60 of my points under the latter definition, the article is very vague.

I am interested in this idea of swapping bots though, I wonder if that does have an effect, even if it's only slight.

ROFLMAO!!

FWIW, I raised the topic at the Doc Team's office hours:

So, maybe there's more under the surface, or maybe there's not. Whatever it is, now we know that if the information fell into the wrong hands it would be curtains for the free world.

Thank God! Can you imagine what would happen if people were specifically gaming the traffic stat?

Luckily we've managed to avoid that nightmare scenario!

ROFL

So I have this quiet little 512 tucked away in a scenic area.
There's a little house over it at 2000m.

I put an alt up there.
The traffic on the parcel was 3 at the time.
It won't matter is any others avatars happen to wander in and out during the experiment. I'm looking for a ball-park. If traffic is avatar minutes on the parcel, then I know what the minimum will be.

Traffic gets updated at intervals, and not in real-time, so I check at intervals to see if the score has changed from 3.
Thirteen hours later, I notice that the traffic score on the parcel has changed.

Okies! Due to RL I hadn't checked for the past 2 hours, so let's say that the alt was certainly there contributing to the count for 11 hours.


Sooooo

The new traffic score should be at least 11 x 60.
It could be more, but not less - apparently.


And it was ......













0

as in nada

zero



Now that's what I call a complex formula!


I'll leave the bot there over (my) night.


Let's see what a full 24 hours on the parcel will generate.


.... and then maybe try another 24 hours with the alt at a lower altitude.


There are other things I could vary - like shoe size - or maybe the thin client.
Maybe SL was just having a bad day.

The traffic figure is updated every 24 hours - not every few hours - and we don't know if it's snapped at the time the number changes or some time before the number changes, so we don't know exactly when the number is snapped for the previous 24 hours. Leave it there for 2 days. That way, you will be sure to get a full 24 hours of traffic from the alt - the full 1440 plus maybe a few for other people. Leave it there for several days and, if nothing goes wrong (if it doesn't get logged out), you'll get the 1440 every day plus a few for other people.

The 0 figure could be from the previous 24 hours, before you put the alt there, or it could be one of those occasional things that do happen from time to time with the traffic numbers. Also, you can't say that it was contributing to that count for 11 hours. If the number was taken before you put the alt there, then it didn't contribute anything at all to the new number.

Leaving it there for only 24 hours will split the day's 1440 between two 24 hour periods, and you'll be no wiser. You may even think that "av minutes on the parcel" is wrong when it's actually been split between two 24 hours periods. It needs enough time to be certain that it's there for the full 24 hours of the traffic period.

A lower altitude won't make any difference at all.

I have an approximate fix on the measurement interval for the parcel.
The Traffic changed from 3 to 0 , eleven hours (at least) after I put the alt on the parcel
On the 24-hour anniversary of the "11th hour", I should see at least the effect of an avatar day.

Tune in tomorrow.

There is enough unknown about the detailed timing of data collection that one could imagine flitting bots back and forth between parcels might make them count as traffic on both parcels somehow. Discovering this empirically would likely take a lot of work, though, and with Traffic's future up for grabs, it's unlikely to be worth it (except perhaps as a debating point in the protracted argument we might expect once the draft policy is blogged).

If enough people believed that Flitting increased the effect then the bot armies would start to register more on the server stats as they TPed or relogged.

Alternatively, it might be simply that traffic-botting is becoming more and more common, with more and more people complaining about their experience being affected by bots taking up sim slots - particularly on Mainland.

"Jack Linden: And with people investing large sums to move up the rankings, well it's something we need to look at"


I happened to notice one traffic_injection_for_hire operation that sold competition voting services. Maybe many other operators use bots that way too. They have to 'flit'.

"Jack Linden: We especially want to look at the people who are selling traffic via piles of bots. That is clearly not cool."


There have been suggestions that we provoke a revolution by having as many people as possible run as many bots as possible.
It may be that the bot runners are doing that job anyway, saving us the bother.


Too many cooks pissing the the broth causes froth.

I've just started a little test again. I've blocked public access to an out of the way piece of land. It currently shows traffic of 1, but that's from the previous 24 hour period. I don't know if any count has been recorded for the current period, of course.

I'm standing on it, and I'll stay here for a short time. I'll make a note of the minutes I am here (somewhere between 30 and 90 minutes). When the traffic number changes, it should include the minutes I'm here. The last time I checked, the change occured during my evening and, if it still changes then, I'll get a result today. If there is any significant already recorded traffic for this period, I'll leave the ban lines up and do the same after the change, so that only I should be counted in tomorrow's change. The chances of anyone else spending any time at all hovering above the ban lines is very slim.

When I've finished standing here, I'll go somewhere else so that it's not the only place I've been in the period.

I'm inclined to think that's true, and it may be largely due to the "traffic for sale" businesses that Jack mentioned. The reason that I'm inclined to think it's true is because, in the last few months, the number of logged-in avs when I start my day (when the U.S. side of the world is asleep) has increased from the low 30ks to the high 40ks, and I'm inclined to think that most of the increase is not due to real people.

I've been trying to follow this thread because 'traffic' is something I really need to learn a lttle more about. But, the numbers name my brain hurt. Come to think of it, it doesn't work well to begin with.

As far as 'good traffic' numbers are concerned, whats the target number for a business?

I guess the real question is, what are the numbers for, say:

Bad Traffic:
Moderate to Decent Traffic:
Good to Great Traffic:

For the purposes of the thread


Bad Traffic: Not high enough to get you on page 1 or 2 of Search Places for search terms related to your parcel

Moderate to Decent Traffic: High enough to get you at the end of page 1 or top of page 2

Good to Great Traffic: High enough to keep you at number 1 to 2 on page 1


The actual traffic number might be 999999999 (very roughly) or 1

It depends on the type of business - what rankings it gets you against the competition, as Sling pointed out. For instance, a club that has one or two events a day can outdo a store that has one or more people in it, and selling stuff, all day long, which is one of the reasons why simply banning traffic bots won't produce good results - many clubs have shops attached that hardly anyone ever goes in and sell next to nothing, but the club's traffic will cause it outrank many good shops that people do go to and buy from.

The best traffic of all is no traffic for any parcel - not for rankings, anyway.

A little bit of completely unscientific data to muddy the waters further:

Yesterday, I spent most of the day on my own 512 parcel breaking scripts. Today, my traffic count (usually 0) is 27. I am pretty sure this is all down to my presence. I can't imagine what any visitors would find to do there and didn't notice any dots on the map. I was there for somewhere between 12 and 15 hours altogether with an occasional TP away and relog. Make of this what you will - I have no idea what it means.

Sorry, lack of sleep...

So, the traffic number is only a portion of the things that go into making your business show up in the Search?

One possible explanation is that the first 27 minutes of your time was the last 27 minutes of a 24-hour time window. This is assuming that traffic is simply bot-botty-minutes.
That should mean that when the count changes next day, you would see 11.5 to 14.5 x60 as the traffic count.

Or not

Traffic is the only ranking factor in the Places tab search. If your place is set to show in search, it will be ranked in the Places tab - the higher the traffic, the higher the ranking. But it's a lousy search system because it only lists places that have the exact words, in the exact order, in either the parcel's Title or its Description, that someone typed into the searchbox. It requires an exact match to be listed in the results.

The All search is a much better system. Traffic for those rankings plays a tiny part, but so tiny that it isn't worth bothering to even think about.

Unless the tiny effect of the traffic on All Search would get you up a position or two in page 1 of All Search - Maybe from #2 to #1
Given that the forged traffic would cost effectively nothing - even if using Premium bots, it could be worth bothering about.


I mention "Premium bots" only because I have seen suggestions that traffic contribution should be weighted by account type. The suggestion was that the cost would make bots uneconomic. However, given the weekly L$300 stipend and the ability to lease out the 'free' 512m of tier to a land dealer, the nett cost is very low.

Great Answers! Thank you!

The maximum value of traffic in the All search is 12 IBLs - the same as 12 Picks - and almost all places don't get near those 12 IBLs. It's much easier to go out of your way to get an extra 12 Picks if they are necessary than mess about with bots. I say, "if they are necessary" because at the top of the rankings, the chances of those few extra IBLs making a difference is relatively small.

There's no right or wrong answer for a business. If your sales to visitors ratio is good, then that's more important, that's the base to work from. I've sold more on parcels with two digit traffic than I have on parcels with five digit traffic because all parcels aren't equal and what attracts people to a parcel isn't equal either.