July 24, 2002

5:12 AM

Is the passenger screeningless secure than purely random screening? According to the write-up in this paper, complete with probabilistic analysis and computer simulation, the answer is yes. I've hijacked the link from BoingBoing.
posted by substrate (10 comments total)

Heh. I've been saying this forever. If you have a non-random security system, then you give the terrorists ways to increase their ablity to get through.

I mean, I I rember reading a quote from some parents who were upset that their 6 year old girl had to be searched. But come on, if we took 6 year old little girls out of the pool of people who can be searched, terrorists will just start using 6 year old little girls to cary their shit.
posted by delmoi at 6:14 AM on July 24, 2002

Good article, although I'm unconvinced by his attempt to claim that terrorists have a diverse pool of potential recruits.

Helder, Kaczynski, and McVeigh were not part of the kind of terrorist cell necessary to use Carnival Booth. And there's no evidence that just because Lindh was willing to fight in an Afghan army he'd be willing to carry out a terrorist attack. (Would you assume that any given member of the US military would be willing to be a suicide bomber against, say, Iraqi civilians?)

I agree with the overall premise of this paper, that terrorist cells will likely be able to find people who can defeat profiling. But it undermines its argument by confusing the Ted Kaczynskis of the world with a group like Al Qaeda. And I don't think there's any evidence that groups like Al Qaeda are yet at the place of having recruits that look like my grandma.
posted by straight at 6:32 AM on July 24, 2002

The point is that airport screen as it is currently established provides an entirely false sense of security. I've said for months that the answer to the problem was to leave screening alone and secure the cockpits. Secure cockpits and an FAA mandate to not succumb to terrorist demands translates to no hijackings and risks only the passengers of a single plane. It's a reasonable societal risk, no more dangerous than traveling by bus or going to a public gathering. It just amazes me how people repeatedly publicly state that they "feel safe" when they should be feeling violated.
posted by shagoth at 7:02 AM on July 24, 2002

I think that there is an opportunity for El Quada or some other group to recruit from within the U.S. It's easy to write off the possibility of an American on a suicide mission because its trivial to project your own circumstances onto them. I'd imagine that if the average person in Palestine or Saudi Arabia were as affluent as the average person submitting to a weblog it would be very hard to find a willing suicide bomber. Recruiting from people who have an artificially shortened lifespan through disease, famine and war is much easier, especially if you can convince them that its the Wests fault.

I think people that fit this mold do live in the United States, and many of them are in jail. If you can recruit them before they've finished their 30 year prison terms you have the potential of producing terrorist tools that don't fit the prevailing profiles. Recruit from the drug offenders who will be returned to poverty once their sentence is up.

shagoth, I don't know that people always state that. Rather, I think the media picks and chooses interview blips that reinforce the view that it wants. On February 25th I was flying back from a Hawaiian vacation, I was interviewed by a news crew, probably selected because I was a huge white guy with a back pack and sandals. They asked me loaded questions and I responded with my actual opinions. No, I don't feel unsafe in the air because in general planes aren't hijacked. No, the enhanced security screening doesn't make me feel safer since I didn't feel unsafe in the first place and it would be easy to circumvent in any event. No, racial profiling is wrong, not all terrorists are Arabs. I figure it took them 30 seconds to decide to drop my footage on the cutting room floor.
posted by substrate at 7:36 AM on July 24, 2002

I think that there is an opportunity for El Quada or some other group to recruit from within the U.S. It's easy to write off the possibility of an American on a suicide mission because its trivial to project your own circumstances onto them. I'd imagine that if the average person in Palestine or Saudi Arabia were as affluent as the average person submitting to a weblog it would be very hard to find a willing suicide bomber. Recruiting from people who have an artificially shortened lifespan through disease, famine and war is much easier, especially if you can convince them that its the Wests fault.

I think people that fit this mold do live in the United States, and many of them are in jail. If you can recruit them before they've finished their 30 year prison terms you have the potential of producing terrorist tools that don't fit the prevailing profiles. Recruit from the drug offenders who will be returned to poverty once their sentence is up.

shagoth, I don't know that people always state that. Rather, I think the media picks and chooses interview blips that reinforce the view that it wants. On February 25th I was flying back from a Hawaiian vacation, I was interviewed by a news crew, probably selected because I was a huge white guy with a back pack and sandals. They asked me loaded questions and I responded with my actual opinions. No, I don't feel unsafe in the air because in general planes aren't hijacked. No, the enhanced security screening doesn't make me feel safer since I didn't feel unsafe in the first place and it would be easy to circumvent in any event. No, racial profiling is wrong, not all terrorists are Arabs. I figure it took them 30 seconds to decide to drop my footage on the cutting room floor.
posted by substrate at 7:37 AM on July 24, 2002

Sorry for the double post! The post process said it errored out (twice).
posted by substrate at 7:38 AM on July 24, 2002

The basic argument is obvious (everyone here understands it) without doing computer simulations. The point of the simulations is to see whether it is possible to work round the system under reasonable conditions.

In other words, the technical bit of the paper that involves computers and statistics is there to answer the hard part of the problem - whether the loophole can be exploited or not [translation to less polite terms: if you don't understand the technical details then opinions about whether or not the paper's conclusions are correct are a waste of time].

For example, it's important to understand the distinction between the distribution of terrorists and the distribution of their CAPS scores. All terrorists might be smelly, swarthy types with big moustaches, shifty dark eyes and funny foreign clothes, but CAPS scores are based on a variety of other factors (see the article) which are inherently "noisy". So even thought all the terrorists are identical (and obvious - why oh why don't we just lock these foreigners up?), their scores vary. And the carnival booth algorithm exploits this.

PS If you get a time-out error while posting it normally means that the server couldn't generate a reply, not that it didn't get your post. Sending a post to MeFi doesn't require much work on the server's part, so is likely to happen. Getting a confirmation back is more likely to fail.
posted by andrew cooke at 8:31 AM on July 24, 2002

I've said for months that the answer to the problem was to leave screening alone and secure the cockpits.

Which is exactly the opposite of what the authors of the paper are saying. Well, OK, not exactly opposite, as they're not arguing that you shouldn't secure the cockpits. But although they make a strong case that random screening is preferable to screening by profiles, they also advocate that screening of those randomly selected be made much more intense than it currently is.

...risks only the passengers of a single plane. It's a reasonable societal risk...

Whether it's a reasonable risk or not depends not only on the number of people at risk, but also the likelihood of that risk. If a plane with 200 passengers is destroyed by a terrorist, on average, once every thousand years, that's a reasonable societal risk. If it happens every week, that's not a reasonable risk. You can't evaluate whether it's a reasonable risk or not solely on the fact that only hundreds of people will be at risk in any single incident, rather than thousands.
posted by DevilsAdvocate at 10:51 AM on July 24, 2002

...risks only the passengers of a single plane. It's a reasonable societal risk...

If a plane with 200 passengers is destroyed by a terrorist, on average, once every thousand years, that's a reasonable societal risk. If it happens every week, that's not a reasonable risk.

I think you misread his quote, if the cockpit doors are secure then the plane would NOT be destroyed, even a dozen terrorists with box cutters would have a hard time manually killing each of the 200 passengers...
posted by Iax at 11:18 AM on July 24, 2002

How will securing the cockpit doors prevent terrorists from bringing bombs on planes?
posted by DevilsAdvocate at 2:31 PM on July 24, 2002