Bot Survey, The Estates Strike Back

For two days I surveyed the estate sims for bots and campers. I chose randomly, by map location. On the second day of surveying I collected data during hours I did not collect on the previous, so this data fairly uniformly covers 24 hours, including both high and low concurrency times. The raw data has timestamps, so I could break it into low and high concurrency, but I haven't done that. Once again, I'd expect the low-concurrency percentage to be higher than the number I'm reporting and the high-concurrency percentage to be lower. The results are as follows, detailed under the cut: 524 sims 953 avatars, 481 of them bots and campers 50% bots and campers, overall For sims with 5 or fewer avatars there was an average of 13% bots and campers (42 bots out of 325 avatars). For sims with more than five avatars there was an average of 70% (439 bots out of 628 avatars). The overall percentage of bots was 50%. Some observations:

While visiting it really felt like there was going to be far fewer bots. I thought I'd be able to report less than 40%, but the large bot farms just dominate the numbers. Even if you remove the largest of them (79) from this data set, the overall percentage is still 46%. And the 79-bot farm isn't even the largest I found. One of the 50-bot farms in this survey was adjacent to a 94-bot farm and a 98-bot farm (those weren't included in this survey).

There's an even more dramatic division between low- and high-occupancy sims in the percentage of bots (13% vs 70%) than there is on the mainland (~20% vs ~62%).

More than half the estates I chose (63%) were empty.

I could not get to some estates because they do not have public access. Assuming those are not commercial they'd likely have few or no bots. I assumed any avatars in non-public sims were human. This only amounted to perhaps 20 avatars.

Data collection was cut short by about two hours because of grid borkage. I'd have to wait a whole day to resume collection, but...screw it. Close enough.

Data under the cut.

Data: The first column is the number of avatars in a sim, the second column is the number of sims that had that number of avatars. The numbers following the second comma are the number of bots found in each sim with that many avatars.

For example, a line that begins

1, 31,

means there were 31 sims with one avatar, and there should be 31 numbers that follow.

Following the data in parenthesis is the total number of bots, total number of avatars and percentage of bots for all sims with that many avatars.

Got it? Here's the data...

0, 326,
1, 63, 0 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 (7/63=11%)
2, 42, 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 2 0
 0 0 0 0 0 0 (5/84=6%)
3, 24, 2 0 0 0 1 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 (5/72=7%)
4, 14, 0 0 0 3 0 0 0 2 0 0 0 0 4 0 (9/56=16%)
5, 10, 5 0 0 0 0 3 3 5 0 0 (16/50=32%)
6, 5, 6 6 4 2 5 (23/30=77%)
7, 7, 6 7 0 6 6 0 6 (31/49=63%)
8, 5, 8 1 0 8 8 (25/40=63%)
9, 5, 0 7 0 9 0 (16/45=36%)
10, 6, 7 10 7 0 3 4 (31/60=52%)
11, 1, 10 (10/11=91%)
12, 4, 12 0 11 11 (34/48=71%)
13, 1, 13 (13/13=100%)
14, 1, 2 (2/14=14%)
15, 1, 3 (3/15=20%)
16, 2, 0 12 (12/32=38%)
21, 1, 21 (21/21=100%)
22, 1, 22 (22/22=100%)
24, 1, 23 (23/24=96%)
25, 1, 19 (19/25=76%)
50, 2, 50 25 (75/100=75%)
79, 1, 79 (79/79=100%)

Comments

[Nexii Malthus]

Interesting, it seems the attention span of people is inversely proportional to the amount of bots.

[Nuschi Martynov]

may i recommend some pictures we took in our
MAKE BOTSPOTTING YOUR HOBBY thread in a german SL froum?
http://www.slinfo.de/phpbb/viewtopic.php?t=16939&postdays=0&postorder=asc&start=240

or have a look at the tag "BOTSPOTTING" on my blog.

greetings from the bus, nuschi

[Jahar Aabye]

Just out of curiousity, you got any p values for this data?

Uh-huh.

You really should be more honest about the fact that it's really a survey of 198 sims, and stop calling it 524. A phone survey doesn't count the people who hung up or otherwise refused to answer.

[Nuschi Martynov]

Jahar...
Would it be much of a difference if we had only 25% bots?
It's still a number that is not communicated by LL - - - and 25% is the result of MY little census in summer 2008, taken in the (european) evening with ~60.000 online.
The Second Travel Team had a look at about 400 SIMs and we counted EVERY avatar. If there was only one Avatar on a SIM, we ran after it, too.
Now, we found about 600 Avatars.
150 of them were Bots.
Another 120 were busy camping. You can be sure that there was a significant number of Bots among them.
The few single Avatars we were not able to locate, we counted as "real" people. We jumped from SIM to SIM 5 times on different (randomly chosen) latitudes from far west to the last SIM in the east. On a straight line.
Different method. For you maybe still not scientific. But:

Two People independently did the same survey (Green Dot and Second Travel didn't know each other nor did we have knowledge of the other's census), and we came to very similar results.

As long as no one else is taking the pain to count again, these are the best numbers we have - and the results are clear!

Way to many Bots to just say "Ah, isolated cases, just 10%, Big HUG!"

[Anya Ristow]

Jahar, there is no expected outcome (the percentage of bots is not a known quantity or statistical function), and therefore no variance to report, and therefore no p-values. Are you sure you know what p-values are?

[Anya Ristow]

Nuschi, I was not aware of your survey. Thanks for mentioning it.

In my travels I have not encountered any so well-arranged bot farms as the couple pictures I've seen of yours, though I've seen a few that have taken the trouble to package them compactly and make the result look like technical apparatus. I'll see if I saved the snapshots.

[realperson]

I am truly fascinated by the number of detected bots. I too am very dissapointed in going to clubs to find many bots. Perhaps you could have a forum or way to post the bots we find on our travels. I didn't look at the other botspotting website yet but I will soon.

Could this bot data be somehow turned into a script that is used inworld to detect bots as we travel from location to location?

oh, here are 28 bots i found in Tuzandi (place called Dania's garden paradise). location (24,223,183). a big bubble in the sky with 28 dead avatars just sitting in the bubble in the sky.

very disapointing to see this. I don't know why people want to waste these resources to make false numbers. it really needs to be outlawed somehow as it really does wreck the second life experience.

Does it help to notify anyone of this? is this considered to be allowable, or should we report these bots to someone at LL? or somewhere?

[realperson]

I was thinking. This bot problem reminds me very much of a similar problem I had with Microsoft regarding their xbox360 backwards compatibility list. It took 6 months of forum postings to get MS to admit on their website that most of their bc games had issues. I think the same will be true of LL and the bot problem, only it will take much longer as they are obviously profitting from it. I was surprised to notice there is a jira out on this problem, however they just want to change the color of the map dots. I don't think that is good enough. Most of the bots i have seen recently do nothing other than pump the numbers up and after what I have seen, it is defiantely closer to 50% and no where near only 15%.

I think this site should combine with the other botspotting site to provide a multi-lingual database page along with the spotter names, etc.

I noticed on the other site that only a few people were getting most of the spotting. This bot spotting needs to become a sport like the previous posters suggest and a multilingual database would be a great start, and if this site could somehow merge with the other site or colloborate with it extensively.

The goal should be to find the most bots by any one person. I would suggest a prize for that but we don't want people to decide to create bots to win the prize either, so the prize has to just be name recognition or something small.

I like how the other site has the link right to where all the bots are so non-believers can log in and see them.

The first time i saw my first cluster of bots in a little bubble, it sort of really felt like the matrix when Neo woke up inside the battery cocoon pod he was in. It felt very strange.

It's one thing to see bots all over and all about sitting and standing in a regular built up sim. It is entirely different when you bump a cross a huge cocoon of them all in a small light blue bubble that blends into the sky. Kindof surreal in a way.

For those who have not seen them yet and who disbelieve the high numbers, you will eventually see them and understand why they should not be allowed.

It is definately false advertising. I wonder if ll could be sued for constantly advertising their numbers to be so high if it could be proved that their numbers were false?

And the only way to prove it to them, will be with a multilingual database webpage.

[realperson]

Not sure if you guys have seen this yet or not. It's a video someone maid with a hundred bots or so:

http://www.youtube.com/watch?v=bVzEib-uB08

Is it against ll policy for someone to move an unattended bot from 1 sim to another? just curious.

[realperson]

Found another 20 bots in a place called Currency - Harlots - High Class Brothel 212,93,200

Is this enough information, or do I need to post the landmark?

[Nuschi Martynov]

What we did, here in Germany:
We started a Forums Thread in SLinfo.de, where we posted SLurl's and - if possible - nicely made photos.
We received a HUGE echo!

Maybe, if you are registered with any Forum or so, start it there in your Home Language (or of course offer it to THIS site here! I don't want to get in trouble here!). The more people know about Bot Abuse the better!

Friendly greetings from the bus, Nuschi

MAKE BOTSPOTTING YOUR HOBBY

[realperson]

http://slurl.com/secondlife/Halfclaw/141/28/751
15 bots here.

[realperson]

For anyone who wants to move bots from 1 sim to another, are there any spells that can allow you to move yourself and a bunch of bots through a wall so that the bots can be pushed out into other sims? I've seen spells that allow you to drag a bunch of bots with you, but oftentimes walls are in the way because the bot makers store them in an area with walls on all the sides with no opening doors so that the bots are stuck in a little area and are not free to roam the sims with your helping them.

[realperson]

http://slurl.com/secondlife/Mounford/168/40/151

another dozen bots.

[realperson]

Another dozen bots, all with the same first name of Jocasta.

http://slurl.com/secondlife/Simyit/43/236/735

[realperson]

Wow. here are about 40 to 50 bots:

http://slurl.com/secondlife/Charmed%20One/74/168/105

All in the sky in a little box.

[JudgeDread]

The biggest gamer of Traffic bots has to Rebecca Vacano, that lady takes the biscuit.Not ony does she game Traffic, she games Picks Camping, she games ALL search with invisible prims in the water. Thank god she doesn't run a classified advert, I'm pretty sure there would be keywords used for products not sold there.
Lets see her Platinum Skins shows 85 dots only 2 are milling around, her Escala sim has 73 green dots of which 0 visitors and finally her Indulgence sim with 58 green dots with just 1 visitor. Thats over 200 green dots with a total of 3 genuine visitors across 3 sim. What is the point of all this? 200 bots to catch 3 genuine visitors? This does not make good business sense. Its not business full stop. People should boycott such places :(

[Tracey Sassoon]

hahahaha Linden Lab are so thick! they haven’t even noticed all the bots that I have at Alady Sim or Noob Island! Whats the point of having a policy that they don’t police!

More fool them, that I am still able to rip off other hard working designers by cheating the system with all of my bots!

I say that all the money I make goes into noob island but actually I don’t make any money because my products are complete rubbish! More fool the people that buy them as my shapes are nothing more than freebie templates!

[Discount Louis Vuitton]

Girls spent more time than boys on social networking sites, listening to music and reading.