Web Game Helps Predict Spread of Epidemics

An anonymous reader writes "Using data from the web game wheresgeorge.com, which traces the travels of dollar bills, scientists have unveiled statistical laws of human travel and developed a mathematical description that can be used to model the spread of infectious disease."

Re:Shades of Psychohistory

[Kuciwalker]

This article is really light on details, but the concept sounds strikingly like something that would be predictable through Seldon's psychohistory.

Yes, apart from the fact that Seldon's psychohistory is completely fictional.

The real reports

[MonkeyBoyo]

Here is the blurb in Nature [ucsb.edu], Nature's Editor's Summary [nature.com]
and here is the PDF research paper The scaling laws of human travel [ucsb.edu].

Woohoo!

[NathanBFH]

I remember coming across this site several years ago (2001? 2002?) and just for fun entered a couple bills to see how it worked. Since then, I totally forgot about it until this Slashdot reminded me! I'm very curious to find out how 'my' bills are doing these days. I do remember, however, reading somewhere that the average lifespan of a one dollar bill is less than a year or two, so the chances my bills made it past the few months I handled them may be slim.

Re:Business model

[Anonymous Coward]

http://www.wheresgeorge.com/faq.php [wheresgeorge.com]

No they don't. Thanks for playing.

Re:Business model

[OfficialWheresGeorge]

The is no "business model". The site DOES NOT SELL RUBBER STAMPS. It stopped selling rubber stamps in 2000 at the request of the U.S. Secret Service.

It's also not "these guys"... it's "this guy".

Please stop spreading this disinformation.

Re:Notes as a form of delivery device?

[Thing 1]

Frank Herbert wrote a book, "The White Plague", that was about exactly that. A researcher was vacationing with his family in Ireland, and watched (from the hotel window) his wife and kids get killed as they walked next to a car which blew up.

So he created a virus that killed only women, and released it to the world via paper money.

The only downside is the book had about 3x as many words as a gripping novel would have, or I was a bored teenager; I haven't read it in a dog's age.

The site is WHERESGEORGE, not where-IS-george

[OfficialWheresGeorge]

But if you go to the wrong URL, it will be corrected automatically.

And it was only temporarily down.. it's back up now.

Re:erm? is the data even legit? Yes, it is.

[OfficialWheresGeorge]

I spend a significant amount of time EVERY DAY to ferret out fake data. I have several automated processes that search for and remove any data that does not fit certain criteria. I take this site, and the data integrity very seriously, so I take personal offense to your offhand, unfounded, and ignorant comments.

-Hank

Re:Shades of Psychohistory

[Mr. Underbridge]

Overall statistical laws don't say much about free will or not. There are always going to be regular patterns in behaviour (caused by things like the fact that most people don't want to walk 10 miles to work every day).

Kind of like how Heisenberg's principle and statistical mechanics aren't mutually exclusive, for the physics crowd out there.

Re:Notes as a form of delivery device?

[zenray]

It's been some time since I've read the book but as I recall it was to get revenge on the men who took his wife from him - so 'I'll take your wife from you'. Since he did not know exactly who did this to him he decided to 'get everybody'. Radical isolation to keep the infection from spreading was tried. How the book ended I can't remember.

Re:erm? is the data even legit? Yes, it is.

[OfficialWheresGeorge]

So did the researchers who used your data set take this into account? For that matter, did you have a significant role in the study beyond providing the data set?


I did not take part in the study, I only provided the data. The data removed from the data set are entries that are obviously fake - for instance two entries thousands of miles apart, entered from the same machine. Or people who intentionally snail-mail bills to each other.

Re:erm? is the data even legit? Yes, it is.

[OfficialWheresGeorge]

First, only the FRB can judge if a bill is unfit to be re-issued. You can't, a bank can't, a merchant can't. Second, the website simply tracks bills by serial number, and works if bills are marked or not. The site does not encourage or endorse the defacement of currency. I am forbidden by law to sell rubber stamps that do so. What the users do and purchase on their own accord is between them and their lawyer.

euro-tracker

[wardv]

Europeans can track their euro-bills here: http://www.eurobilltracker.com/ [eurobilltracker.com]

Re:Woohoo!

[Poutsi]

Never re-visited the site, because you couldn't enter dollar bills from Europe.

You can track Euro banknotes then: http://eurobilltracker.com/ [eurobilltracker.com] :)

Re:Business model

[OfficialWheresGeorge]

As I said in my previous post, nothing on the site endorses nor encourages defacing currency. But I guess you didn't actually read what I wrote.

Re:Business model

[Anonymous Coward]

Defacing United States currency is NOT illegal. That is a common misconception.

It is your intent that is important here.

United States Code
TITLE 18 - CRIMES AND CRIMINAL PROCEDURE
PART I - CRIMES
CHAPTER 17 - COINS AND CURRENCY
  333. Mutilation of national bank obligations

"Whoever mutilates, cuts, defaces, disfigures, or perforates, or
unites or cements together, or does any other thing to any bank bill,
draft, note, or other evidence of debt issued by any national banking
association, or Federal Reserve bank, or the Federal Reserve System,
with intent to render such bank bill, draft, note, or other evidence
of debt unfit to be reissued, shall be fined under this title or
imprisoned not more than six months, or both."

...In other words, as long as the serial and bill numbers are clearly legible, you can mark your bills as you see fit.

Re:erm? is the data even legit? Yes, it is.

[OfficialWheresGeorge]

If you're trying to have a pissing contest to prove you know how to spoof the system and create bad data - congratulations - you win.

But with over 8 million "hits", and statistician will tell you that the data is statistically significant within some margin of error. Any researcher worth their weight will take bad data and outliers into consideration.

The overwhelming majority (easily > 99%) of the hits are valid. Most people (i.e. general public, not slashdotters or technical people) - who use my website don't know an IP address from a street address, and don't know cookies from brownies. I know my demographics very, very well - I know the types of people that participate on the site - YOU DO NOT. For instance, do you know how many times I've gotten an email from a user with the question "Where's the zip code on the dollar bill?" or sign on as "[e-mail]" or "hotmial.com" ?? ALOT OF THEM. We're not dealing with technical people, I'm usually dealing with people who can barely figure out how to log on and send email.

The very few people who do attempt to spoof the system are *usually* detected. I've gotten very good at the detecting the patterns of abuse after doing this for eight years. Do some slip bogus entries through - sure they do.. and that's why anyone using this dataset would take that into consideration.