Good God, please tell me someone is joking here…
Defense Dept. Program Taking Terror Bets
http://www.thewbalchannel.com/news/2363276/detail.html
My guess is it’s a predictive tool based upon a stock market model. Both terrror events and the stock market respond to outside stimuli (politics, weather, actual data, etc.) but still have an element of randomness. My guess is a stock market model would mirror terror events pretty well. Someone probably said “monte carlo simulation” and the reporter thought “place your bets”.
Alright everybody, the game is Stud Poker, aces and bin Ladens wild…
Or perhaps we should play “Rummy?” 
Well I know ONE way to corner THAT market.
::wringing hands, evil grin::
Well, looks like I was wrong…some moron in the pentagon actually IS setting up a betting parlor.
Here is an update.
http://www.washingtonpost.com/ac2/wp-dyn/A64157-2003Jul29?language=printer
As crazy as this sounds, this idea comes from gaming theory, and is very powerful.
The US Navy has used it with great success finding wrecks at the bottom of the sea. You take a pool of people with some expertise in the subject matter, feed them all available information, and have them place bets on where the ship will be found on the seabed. It works, though I’m certainly not capable of explaining why.
Hey, SkyBum, wanna buy a Brooklyn Bridge? I’ve got a… :smack: Holy shit, it’s true.
The idea is to tap into unconscious information processing. It looks like black magic, but as Boyo Jim notes, the Nav uses it with a great deal of success. The Coast Guard also uses it, when they have time, to help plan searches. Oh, and it obviously is used in large financial (futures) markets, where systems that don’t work die a darwinian death. The underlying concept works quite well.
Frankly, I think most of the furror over this plan was ill-informed, prejudiced, and ill-founded. I also think the massive, ill-informed uproar has cost us access to a potentially useful tool, before we ever got a chance to see if it works or not.
Well done, folks.
I’ll second that. We should at least test it to see if it is an accurate predictor.
I’m curious as to how this thing was supposed to work, given that the government would be routinely intervening based on these investment strategies.
Say this market predicts another terrorist attack on NYC. Since these futures would be public information, every terrorist and his brother can see this, and anticipate that in will step the Homeland security forces to keep a closer eye on New York. So they switch to DC, or another out of the country embassy. And those who bet on the attack actually help prevent it, and consequently lose their bets. That’s hardly comparable to a free market system.
I seriously question whether this method can work in such an environment. One reason it works so well for searches is that the sunken ship can’t tell where the search is going to be, and go hide somewhere else.
I’m pretty light on this, but essentially, it works on the thery that people are better at processing information than they realize, and that many ‘hunches’ are in fact guided by analysis the holder doesn’t realize they have. Now,one person making a hunch is pretty hazy, but when several dozen well-informed people make the same ‘bet’, there’s a fairly good chance there’s something concrete underlying that bet.
My guess is (and mind you, this is all based on how I’d do business) that they’d not actually do any interventions based upon the ‘bets’, but would rather focus my intelligence gathering aparatus with special attention on those areas, to see if there isn’t something to the guess. Also, these ‘guesses’ would quite possibly identify unexplored areas of vulnerability too, a further benefit.
There’s a good article in Wired today about this with a link to a non-governmental site that’s been doing the exact same thing for awhile now.
I guess the fact that the government should somehow be above this sort of thing and that the media blew it out of proportion turned it into big news. Meh.
I don’t see what the point to setting up the whole thing would be if you don’t act on the data. And re-focusing your intelligence IS doing something. Can you imagine the political consequences if this predicted something, and the security forces just put up cameras to watch it, and DIDN’T act to prevent it?
Quite aside from the Yuck Factor, that is the problem with this plan; the Boyo figured that out in minutes. Did the point escape DARPA?
Oh, it works by concentrating knowledge and information, I think.
It took Boyo over eight hours.
Remember all the rumours about people who sold air-line stock short right before September 11th, and the theories that the hijackers essentially financed the attack with the attack itself?
Re-focusing one’s intelligence efforts is most certainly doing something, but it is also most certainly not an intervention. Intervention would be based upon, as I understand, the results turned-up by more traditional methods, no matter that the initial lead came from the ‘futures’ program.
Still those who ‘bet’ on an event occuring would be assisting in preventing that event from occuring.
And those who intended for an event to occur could to arrange to profit from the event.
So when a terrible event occured, the intelligence community could examine trading to see who had ‘bet’ heavily on the event.
So, one either loses money or comes under the intense scutiny of the government (said in a paranoid, conspiracy-theory-believing tone).
I do not see the most intelligent and knowledgeable people trading in this market.
I can only see this working if the bets are kept secret. Like Boyo Jim says, if everyone can see that the bets are favoring New York, then the terrorists might decide to go to Chicago and blow up the Sears Tower instead.
This is how I heard it reported over here (UK). The reporter basically said that there were two options: the obvious one that is described at the beginning of this post (ie. analysts would better identify risks) and the other, which is that terrorist groups would inadvertantly identify themselves by putting money on particular events happening, or not.
Personally I can’t see that either work for the reasons posted by j66 and Boyo. The first only works with only one varient (the prediction rather than the predicted event). The second only if people are stupid enough to identify themselves to suspected terrorist activities.