Terrorism is a problem. But the solution to that is not racial intolerance and kangaroo courts for teenaged soldiers recruited by extremists. I mean, if someone is willing to kill themselves in order to do something, how much of a deterrent will any sort of punishment be?
The above paragraph, and I suspect the objection to the ad about loners, relies on the assumption that when the media talks about "terrorists" what they really mean is "Muslims". It turns out that when you run the numbers, that doesn't make any sense..
Let's do some more math. Bayes' rule is pretty much the most important thing ever. It tells you what the likelihood of something is based on prior knowledge. In this case, let's calculate out the probability that someone is a terrorist, given that we know they are Muslim.
Using that link with graphs about terrorism and assuming that there are about 1 million Muslims in the US and about 300million people, we can calculate that
P(Terrorist | Muslim) = P(Muslim | Terrorist) * P(Terrorist) / P(Muslim)
We can get the first term, P(Muslim | Terrorist) = 0.06, from the pie chart at Loon Watch. The second P(Terrorist) = 318 / 300M = 0.0000011. comes from there, too, and our assumption about population. And finally, P(Muslim) = 1M/300M = 0.0033.
So, P(Terrorist | Muslim) = 0.06 * .0000011 / 0.0033 = 0.00002 = 0.002%
That means that out of every 50,000 Muslims you meet, 1 will be a terrorist. Compare this with the fact that of every 50,000 people you meet in NYC, 290 of them will be a violent criminal. Maybe it's time to ban I < heart > NY shirts from public places.
* The curtains keep the temperature of the apartment down on sunny days.
No comments:
Post a Comment