You know how you always read those whimsical news reports about this bookie or that bookie giving whatever odds on Saddam being alive, or aliens touching down, or Michael and Latoya Jackson being the same person? It turns out that Uncle Sam wants to be that bookie, and base national policy on how the bets are going. The thinking goes, gamblers aren’t stupid. If lots of them are betting that Saddam is dead, then he’s probably dead. They prove that this (what they call “market based”) method works with this helpful graphic:
See how the “Market Method” arrow hits right at the center? You can’t lose!
Here’s the page at DARPA about the project: FutureMap
And here’s a news story about it. Apparently this idea has already secured $3 million in funding from the House, but is having trouble in the Senate.
Now before you tell me that this is somehow different from taking bets and making odds, read this summary of an earlier version of this project (search for Electronic Market-Based Decision Support):
Typically, the market maker issues a basket of contracts covering a set of events that is mutually exclusive and exhaustive. Market participants trade the issued contracts freely, buying and selling individual contracts through an electronic market. When the outcome is known, the market maker pays off only the winning contracts; before the outcome is known, the prices reflect market opinion of the probability of each outcome.
Me
After I read it here yesterday, I concluded that it HAS to be some sort of joke that I didn\’t get…but wait as they say. It\’s been a story everywhere and proves once again that the loonies are still running amuck.
steve
Yeah, every site seems to be picking up this story. I originally found it at the Agonist.
Arthur Van De Lay
There\’s a story about it on the NYtimes website (not sure if it\’s in print)
become a trader:
http://www.policyanalysismarket.org/
steve
Amazingly, in the NYT story they actually use the \’bullseye\’ image I have above!
http://www.nytimes.com/2003/07/29/politics/29TERR.html
Look in the sidebar…
Arthur Van De Lay
but what does the bullseye mean? They never really explain why the crosshairs cross at the intersection of 5% of \”probably of overt action\” and 25% \”probability of terrorist activity\” - why is that better than \”poll of experts\” which crosses around 7% and 33%….
Note that if the lower axis was actually drawn to scale, the \”poll of experts\” and \”general\” poll would look alot better (they would have been closer to the apparently randomly placed bullseye).
Arthur Van De Lay
NYT reports that the project is being cancelled, and some of the web pages are being deleted.
Was this page deleted, or am I pulling it up from my cache. It does appear that the javascript link is now bad.
http://www.policyanalysismarket.org/pam_example.htm
Glad I\’m not the King of Jordan.
steve
That page you linked is still there. Here are some images that were on deleted pages. The images are still there, for now:
http://www.policyanalysismarket.org/images/background.jpg
http://www.policyanalysismarket.org/images/background3.jpg
Arthur Van De Lay
Thanks,, I also saw those linked at slashdot.org
Look at the upper right hand corner of the graphic, on the first one it\’s marked \”global securities\”… you can bet on the number of US military deaths.
Me
Gone, gone, and hopefull buried.
http://www.sky.com/skynews/article/0,,30200-1098517,00.html
steve
The site is now deleted!
Me
no it isn\’t, your comment section cut off the rest of the site URL: paste the second line and it works.
steve
No, I mean the actual terror-betting site ( http://www.policyanalysismarket.org/ ) has been deleted.
Arthur Van De Lay
On NPR on my way home from work, a senator mentioned that DARPA has 10 other programs that were just as bad.
I think he was looking at this:
http://www.darpa.mil/iao/programs.htm
I think my favorite is Human ID at a distance. Although Babylon is nice too, since it reminds me of Star Trek.
steve
Yeah, they have some far out stuff. They also want to identify who people are based on the way they walk. I think they should try at least getting face recognition to actually work first.
Then there\’s the other thing I keep hearing about where they want to decide if you\’re a terrorist by measuring the way you shift in your seat.
I don\’t know. Maybe I\’m overreacting. Some far out wacky ideas pay off. I\’m using the results of one right now. I guess I just don\’t have confidence in DARPA lately.
Chris Carlin
The thing is, the project would have probably worked. There is hard scientific reasoning behind the proposal, some of it raising other veyr interesting questions.
This arm of DARPA is charged with finding wacko ways of doing things. This is a GOOD THING to have, as it follows that idea that without any failures there are limited successes.
It\’s a shame the media and public overreacted so greatly to the discovery of this project. Perhaps had the project had a chance to grab a bite, it might have been able to justify itself to the knee jerkers out there.
Pleae leave a comment!