Saturday, November 03, 2012

Saturday Politics: Nate Silver Predicts Obama Win


Nate Silver, statistical prediction maven who runs the FiveThirtyEight blog at the New York Times has come under fire from mainstream journalists (and Republican/conservative political pundits and even the public editor of the New York Times itself) for rejecting the media meme that the presidential race is "too close to call." Silver uses statistical and mathematical modeling techniques to generate probabilities of election results. His current estimate is that President Obama has a probability of 0.837 of winning the electoral college. This means that using the sum total of all the public polls released in the states over the campaign season and considering past historical trends Silver runs computer simulations of the elections and in 83.7% of them, the Obama-Biden tickets gets more than 270 electoral votes and retains the presidency.

Silver is so confident of his techniques that this indicates a likely Obama win that he tweeted a bet to Joe Scarborough of MSNBC that if Romney wins Silver would donate $1,000 to the Red Cross and if Obama wins Scarborough would donate $1,000 instead. It was the bet that got Silver in to hot water with "serious journalists" but they have been resentful of Silver's brand for years.

The problem is that many, many people (including political pundits) are functionally innumerate and do not really understand what probabilities are and have no idea of the techniques Silver uses to compute his calculations. A probability of 83.7% does not mean that it is impossible for Mitt Romney to win the presidency. In fact, the New York Times lists more than a half-dozen times when Silver made a political prediction and was defied by the actual results (most notably saying that in 2009 Maine would enact marriage equality by passing Question 1 when it failed).

Anyway, Silver summarizes his position in an article posted today which strikes back directly at his critics:
The FiveThirtyEight forecast accounts for this possibility. Its estimates of the uncertainty in the race are based on how accurate the polls have been under real-world conditions since 1968, and not the idealized assumption that random sampling error alone accounts for entire reason for doubt. 
To be exceptionally clear: I do not mean to imply that the polls are biased in Mr. Obama’s favor. But there is the chance that they could be biased in either direction. If they are biased in Mr. Obama’s favor, then Mr. Romney could still win; the race is close enough. If they are biased in Mr. Romney’s favor, then Mr. Obama will win by a wider-than-expected margin, but since Mr. Obama is the favorite anyway, this will not change who sleeps in the White House on Jan. 20. 
My argument, rather, is this: we’ve about reached the point where if Mr. Romney wins, it can only be because the polls have been biased against him. Almost all of the chance that Mr. Romney has in the FiveThirtyEight forecast, about 16 percent to win the Electoral College, reflects this possibility. 
Yes, of course: most of the arguments that the polls are necessarily biased against Mr. Romney reflect little more than wishful thinking. 
Nevertheless, these arguments are potentially more intellectually coherent than the ones that propose that the race is “too close to call.” It isn’t. If the state polls are right, then Mr. Obama will win the Electoral College. If you can’t acknowledge that after a day when Mr. Obama leads 19 out of 20 swing-state polls, then you should abandon the pretense that your goal is to inform rather than entertain the public.
And this last paragraph is precisely the point. Election coverage happens in media channels which are part of corporate conglomerates that are out to make a profit. They have a financial incentive in promoting the notion that the election result is a 50-50 proposition and that either team has an equal likelihood of winning, but just because there are only two possible outcomes on Tuesday does NOT mean they are equally likely! (If you understand this concept then you are on your way to understanding probability. That being said, it is still then a very difficult question to estimate the probabilities of the two possible outcomes, and this is what Silver succeeds at.)

Silver;s main points is that to assert that the two outcomes on Tuesdays are equally likely and thus "we have no idea what could happen on Tuesday" in order to raise ratings and increase traffic to political websites is about entertaining the public, not informing the public.

No comments:

LinkWithin

Blog Widget by LinkWithin