“$1.6 Million US Dollars Spent,” and It’s Fake (ok, what else is being faked?)
by Urizenus Sklar on 10/11/06 at 9:19 am
I was impressed when I check the LL homepage yesterday and saw that it was announcing $1.6 Million in “US Dollars spent” in game. But then I checked my email and found an anonymous letter from a Herald reader claiming that one million of that amount was genererated in one and half hours by a script s/he wrote in which money was passed back and forth between two avatars (letter below). Nothing was bought and nothing was sold, that is, not one penny was “spent”. But now if it is that easy to fake these much-hyped numbers, how do we know that someone at LL isn’t doing precisely the same thing? Or generating accounts? Or regularly logging in dummy acccounts? Is there any independent auditing on these claims or are we just supposed to believe the Linden numbers because … um because…(I’m drawing a blank here). Folks, a lot of money and investment is coming into SL because of these numbers.
date Nov 9, 2006 6:00 PMsubject US$ Spent in last 24 Hoursmailed-by gmail.com
I just wanted to let you know that I ran a little experiment today. I passed a fixed amount of L back and forth between two accounts via a scripted system that was able to pass the money back and forth repeatedly. When I started the test, the 24 hour total was approximately 486,000 USD. After about an hour and half, my system pushed the total to over 1.5 million. What this means is that Linden Labs does in fact include transactions that net a zero sum. After all the testing was complete, I had the same amount of L as when I started. I have the transaction history for both accounts involved, as well as the scripts used for the test.
Regards,
Anonymous Sl User
Darren Herman
Nov 10th, 2006
Second Life’s $1.6mm Spent Today…
I was reading my RSS feeds this morning and saw a very interesting one from the Second Life Herald about the total amount of dollars spent in Second Life during the last 24 hours. The number seems extremely high to
TP
Nov 10th, 2006
This would be a good way for “The first PR Firm” to generate numbers to show to potential corporate entities in bid attempts for that new 30 sim island pre-fab advertising campaign.
Hiro Pendragon
Nov 10th, 2006
I was going to do something similar to this once, until I found out that it had already been done… about 18 months ago. Apparently the person who did it felt bad as the number skyrocketed, and so he contacted Linden Lab and told them how to remove it from their database.
Well, I guess I’d assumed the Lindens would have put in place some checks to make sure that didn’t happen, but I guess not.
The “US$ spent” number has always been a wishy washy number. There’s dozens of different valid reasons people might exchange money and not actually spend it – from people gambling for extended periods and taking little net loss or gain, to people testing out vending machines to see if they work, to passing money to alts.
Heck, if Anshe is passing her money to an alt and selling from one central location, all of her money could be doubled in the transaction value.
I’d love to see this statistic disappear from the homepage. It’s basically meaningless. A much more interesting number is L$ sold for US$, which is showing actual commerce.
Urizenus
Nov 10th, 2006
Agreed, Hiro. L$ sold is much more meaningful.
Hiro Pendragon
Nov 10th, 2006
I did some calculations. If I can move 50,000L every 10 seconds, then in 24 hours at the going exchange rate, I can single-handedly do $1.6mil US in transactions.
If I gave 5 of my friends the code to do that — which is all of a permission request and a payment to a key of an alt, and vice versa — that’d be 8 million US$ in 24 hours.
What I want to know is if Linden Lab would suspend me if I did it? There’s nothing expressedly in the TOS or CS against rapid transactions, and once every 10 seconds could hardly be considered a DOS attack. Is there any clause against, like, “screwing around with Linden Lab’s statistics”?
Anonymous
Nov 10th, 2006
Actually Hiro,
It took a lot less than that. The test conducted yesterday passed 10,000L back and forth between 2 accounts. A rough calculation of about a full cycle per second was achieved. Note that a full cycle was from user1 to user2 back to user1. With the 10K being passed it was registering as 20K per loop. This ended up netting just over 1 million USD on the front page in somewhere between an hour and a half and an hour and forty five minutes. I also did not see any specific TOS or CS violations by using the 2 person test system I set up.
I, too, agree the L$ to US$ is a far more accurate number to watch. The only problem with that is that it’s not the number they show front and center. One must actually be logged in to their account (via the web log in) to view additional statistics. The simple fact that they continue to display completely useless numbers on the front page for the world to see is rather troubling to me. It speaks to the overall methods they are willing to use in their advertising campaigns.
I know what I did yesterday was nothing new, and that this isn’t any big surprise to many, but I wanted to verify the fuzzy math for myself first-hand.
Regards,
Anonymous
wbmny
Nov 10th, 2006
$1.6 Million US Dollars Spent, and Its Fake (ok, what else is being faked?)
I think it might be unfair to say it is faked, because a transaction is a transaction. It would be the same thing if a friend and I invoiced a million dollars worth of fabricated services.
In real life it wouldnt happen because youd have …
Anonymous
Nov 10th, 2006
The amount of money bought and sold on the LINDEX is not really any better. You could play the same game, except for one thing: The 5% Linden commission would reduce your money to 0.
I would argue that the solution to this “gaming” the system is simple: Re-establish the sales tax. As long as there are any transactions that are 0-sum, then it does not seem possible to distinguish “real” trade from fake trade.
Testing for a zero-sum transaction would be silly. If my friend buys one of my items, and then I go hear her play live music and giver her a donation … this isn’t really zero sum! Even though the total amount of money changing hands adds up to zero, an economic good was created, and we both benefited.
By putting back the sales tax, buyers and sellers would have to think carefully: How *badly* do I want to make this transaction? Is the good I’m getting worth more than 105% of the asking price?
I’m sure this post will get flamed out of existance, but …
Viva la sales tax!
Clickable Culture
Nov 10th, 2006
Second Life TransactionsPerDay Stats Rigged?
The Second Life Herald broke startling news today alleging that Linden Lab is publishing inflated statistics on the home page of its virtual world platform Second Life. According to the Herald, stats that showed $1.6M USD were transferred in a 24h…
Urizenus
Nov 10th, 2006
Well, the TOS still allows Linden Lab to suspend/terminate an account “for any reason and no reason”, and I think this falls under that. It won’t surprise me if your account gets killed, Anonymous. We will hold a funeral if it does.
Anonymous
Nov 10th, 2006
Urizenus,
Given where this is all headed, I wouldn’t be surprised if the accounts are closed as well. We’ll have to see if I can fair the stormy weather ahead.
niko donburi
Nov 10th, 2006
Since we are talking questionable numbers, what does the “553,825 Logged in Last 60 days” reflected in the screen shot at the outset of the article mean?
Is it:
(1) In the last 60 days, 553,825 DIFFERENT residents logged in; or
(2) In the last 60 days, there have been 553,825 log-ins by residents
As with the $$ spent, knowing this would certainly shed some light on the real number of users in SL. If (1) is correct, then LL could safely say there are half a million active users. True, there are some alts that are increasing the number but presumably not that many.
However, if (2) is correct and each time I log in to SL the number increases by one, the active user base is woefully overestimated. I log into SL every evening, so I would count at 60 of those “logged in”. But then think about how many times during an SL session you get booted and have to relog. Does each of these count as another log-in as well?
Any thoughts?
Prokofy Neva
Nov 10th, 2006
Boys and girls, you are not understanding the essence of Second Life. Second Life is all about the creativity and the valuation of creativity. You may think that passing money back and forth between avatars and jacking up dollar amounts is “pouring from the empty into the void,” as Georgy Ivanovich inimitably put it, but in fact, it is a form of commercial activity that the system faithfully recorded.
The system of SL values the creativity of someone who makes a script and puts out 2 avatars and has them play see-saw for two hours. It values that creativity so much that it shows a half million or even a million dollars U.S. The system is neutral as to morality or art — if your labour away, and pay even your worthless alt money because you think it’s worth it, then the system rewards you labour.
Yes, many people have to switch money among alts for all kinds of money and no doubt that gets recorded twice.
It’s like the land that you have to switch around from different groups and sell for $0 or $1 — all that is reflecting somewhere as land transactions and driving down the average cost of land per meter, which is why it always has this number of “$5/m” when you have to buy it in real virtuality for $10/m.
Artemis Fate
Nov 10th, 2006
I’m surprised anyone even pays attention to those numbers. I always figured they were likely padded.
Shep Korvin
Nov 10th, 2006
Was there ever any doubt – from anybody but the most ultra-naive – over what this figure represented?
Gotta love the anonymity thing though… I mean, it’s true that we sometimes have doubts over the linden’s technical prowess, but surely you’ve just left a L$1,000,000-sized neon sign in the SL database pointing to exactly who you are
Matt MIhaly
Nov 10th, 2006
Niko wrote:
As with the $$ spent, knowing this would certainly shed some light on the real number of users in SL. If (1) is correct, then LL could safely say there are half a million active users. True, there are some alts that are increasing the number but presumably not that many.
No, LL could safely say that half a million people have logged in in the last 60 days. Given all the media fluffing, my bet is most of those people came, saw, and left.
A better metric would be everyone who has logged in in the last 2 weeks AND (who has either spent money or spent a total of at least 5 hours in-world).
–matt
Urizenus
Nov 10th, 2006
Suppose someone had, let’s say, 20 accounts. Would it be possible to write a script that would repeatedly log them in and then out as rapidly as possible, and could we use this to probe whether the alleged half million logins was a measure of total logins (including multiple logins by the same accounts)?
niko donburi
Nov 10th, 2006
Matt,
Thanks for the correction!
Here is a link from the beginning of 2006 discussing similar issues. It is worth a look at:
http://www.secondlifeherald.com/slh/2006/02/secondlife_reac.html
Niko
Random Writer
Nov 10th, 2006
Something interesting to note:
Those accounts logged within 60 days, look padded to me too. If we even go by them logging in for just 1 single hour, that means that 800,000 never even logged on. Period. As for alts, I know people after the creation deal that went from 1 alt to 7.
With a good comp, you can run 2-3 SL’s at once. I’d like to see numbers on unique IP’s logged in 60 days.
Random Writer
Nov 10th, 2006
Redoing my post cause a sentence got lost. Silly me. *blush*
Something interesting to note:
Those accounts logged within 60 days, look padded to me too. If we even go by them logging in for just 1 single hour, 370 unique accounts are logging in each hour. 800,000 never even logged on. (We are starting to see the stagger, now that 60 days is no longer pre 6/6/6.) Period. As for alts, I know people after the creation deal that went from 1 alt to 7.
With a good comp, you can run 2-3 SL’s at once. I’d like to see numbers on unique IP’s logged in 60 days.
Cocoanut Koala
Nov 10th, 2006
Come to think of it, Niko, you’re right.
I hope someone will test this, as Uri suggests.
As for transaction taxes, I would hate to be taxed whenever I use my alt to test my products by buying and deploying them.
coco
mrlk
Nov 10th, 2006
im reading further and further down and coming to the conclusion…. who the fuck cares?
the bottom line is that if a corporation (cos thats what YOU lot are essentially talking about) is thinking of coming into sl and is using those figures as the basis of wheher they will or wont then they deserve everything they get
i know people around here dont have a lot of time for rl businesses coming in but does anyone honestly really think that the financial controllers, accountants and bean counters arent going to want to see some decent figures other than whats on the front page?
as for everyone else its absolutely meaningless and all its doing is giving you lot something to gas about
Tom
Nov 10th, 2006
I don’t think the 60 day logins number is actually padded. I think it is what it claims to be — the number of unique avatars that have logged in at some point over the past 60 days.
Now, what that number *doesn’t* tell you is this:
Accounts logged in over past 60 days: 533,852
Accounts registered over past 60 days: 646,059
Registered/Logged in: 1.21 (i.e. over the past 60 days, 21% more users registered accounts than actually logged in)
So, at least 18% of new account registrations will not log into SL even once.
Digging deeper, we can determine that if 80% of all registered accounts *do* log in at least once, then only 4% of accounts that existed 60 days ago have actually logged in over the past 60 days.
In fact, in order to get even 30% of 60-day-plus old accounts logging in over the past 60 days we need no fewer than 50% of all new registrations abandoning SL without ever logging in.
So, what we have in actuality is either extremely low user retention or an excessive amount of junk accounts. Or both.
mrlk
Nov 10th, 2006
4 words
lies, damn lies and statistics
again – so what? we all know that on average there are only around 12-14000 logged in at any one time…
“or both”
sounds good to me…..
Hiro Pendragon
Nov 10th, 2006
Right now the homepage says: US$ Spent Last 24 Hrs: 634,228
Linden Lab does correct known errors. The problem, of course, is that there’s so much room for little errors.
In this case, this is hardly a case of Linden Lab gaming the number, but it’s probably more of a case of Linden Lab not fully thinking through the meaning of the number. Remember how long it took Linden Lab to post “past 60 days” stats — I remember at least a year of discussion on the forums before they started actively releasing that. This is likely the same case; we just need to give them more feedback on what’s valuable and what’s not. Maybe it just needed a story like this to really focus attention on it…
Urizenus
Nov 10th, 2006
Hiro, I think the problem self-corrected in the following way. 24 hours have passed since Anonymous ran his experiment. If you look at the graph on the SL Reuters homepage, the spike lasted 24 hours.
But yeah, I agree about your point. We are all still beta-testing as it were to see what numbers are meaningful.