Skip to main content



Nate Silver and Thomas Bayes

In this article by the Guardian, we explore the usage of Bayes Theorem by celebrated political predictor Nate Silver.  During the recent Presidential election, armies of pollsters were claiming that the race was too close to call, some even giving the upper hand to Romney.  But not Nate Silver.  Nate Silver dismissed the national polls, even when they showed Romney up by a significant margin, and called the race for Obama.  He did this thanks in no small part to Bayes’ Theorem. As we learned in class, Bayes Theorem gives the relationship of conditional probabilities of A and B .  Nate Silver summarizes the rule as “a statement – expressed both mathematically and philosophically – about how we learn about the universe: that we learn about it through approximation, getting closer and closer to the truth as we gather more evidence.”

In the article, Nate Silver discusses the probabilities of 9/11.  The conditional probability discussed is whether or not the event of planes crashing into the Twin Towers was a terrorist event.  The conditional probability is how likely terrorists are to fly planes into skyscrapers in Manhattan.  Or rather the chances that event A (terrorists hijacking the planes) would occur given B (the number of planes hitting the towers).  Before the planes hit, the likelihood of terrorists hijacking aircraft and flying them into buildings in Manhattan was 0.005%.  Then, when a plane hit, the likelihood of a terror attack (A) went up to 38% because the number of planes (B) had increased.  After the 2nd plane hit the tower, the probability increased to a 99.99% chance. This is a simplified version of the entire formula in class, and a depressing example, but it gets the point across.

At the end of the article we are reminded by Silver that it is “not that much of an accomplishment to describe history in statistical terms.”

http://www.guardian.co.uk/books/2012/nov/09/signal-and-noise-nate-silver-review

Comments

Leave a Reply

Blogging Calendar

November 2012
M T W T F S S
« Oct   Mar »
 1234
567891011
12131415161718
19202122232425
2627282930  

Archives