The difficult math of catching the likes of Farouk Abdulmutallab | Business

A passage from chapter 36 of Sam Savage’s excellent book The Flaw of Averages:

So how many true terrorists do you think are currently in the United States? … I have no idea myself, but for the sake of argument, suppose there are 3,000. That is, in the total U.S. population of 300 million, one person in 100,000 is the real deal.

Now consider a magic bullet for this threat: unlimited wiretapping tied to advanced voice analysis software on everyone’s phone line. The software can detect would-be terrorists within the utterance of three words … Assume that the system is 99 percent accurate; that is, if a true terrorist is on the line, it will notify the FBI 99 percent of the time, whereas for nonterrorists, it will call the FBI (in error) only 1 percent of the time. …

When the FBI gets a report from the system, what is the chance it will have a true terrorist?

a. 99%

b. 98%

c. 66%

d. 33%

e. 1%

f. One chance in 1,000

The answer is f: the system would turn up 2,970 terrorists and 2,999,970 false positives. The point that tightening up security procedures enough to catch the likes of Farouk Abdulmutallab would net lots and lots of harmless people has been made before, but Savage does demonstrate it quite elegantly, doesn’t he?

Stocks Rally Further in Run-up to EU Summit

MOSCOW — Global stocks enjoyed one of their best days in weeks on Tuesday ahead of a summit of European leaders that’s expected to be dominated by calls to boost economic growth.

Europe remains the focus of attention across all financial markets in the run-up to the June 17 Greek election that could go a long way to determining the country’s membership of the euro as well as the future of the single currency zone.

harryfox

Martin Hellman uses a similar example of how, in real world situations, even 99.9% accuracy may not be all that good:

Suppose you are a stunt pilot, and you are going to perform a maneuver that is 99.9% safe — but “unsafe” means you crash and kill yourself. If you perform this maneuver 100 times, there is a 10% chance you will crash and kill yourself. Not very appealing, huh? So 99.9% safe on a per-try measure is really not all that safe if the try will be repeated many times.

Martin uses this example as a lead-in to talking about the dangers of nuclear proliferation — if there are lots of nuclear weapons storage depots in the former USSR, Pakistan and similar places, then having each facility be 99.9% safe against “losing” a weapon is not very comforting.

Ah well, a nice thought to start your week.
geaugailluminati

if there have been as many as 3000 terrorists in the US, we’d have had a lot more damage by this time…

we’d be better off checking for the bogeyman under the bed…
waltwriston

The kind of offical numbers come from the risk in terrorism insurance.

See: http://www.insureagainstterrorism.org/PWGReport.pdf
http://curiouscapitalist.blogs.time.com/2010/01/07/time-to-revisit-the-terrorism-futures-market/ Could Robin Hanson have stopped Umar Farouk Abdulmutallab with his terrorism futures market? – The Curious Capitalist – TIME.com

[...] (0) • Related Topics: efficient markettechnology, terrorism, underpants I've already shared Sam Savage's take on the failure to keep the underpants bomber from getting on that plane to Detroit: that it's just [...]