Silver Called It

Karl Rove’s inability to acknowledge the facts on the ground in Ohio may be belief perseverance–“maintaining your original opinions in the face of overwhelming data that contradicts your beliefs.” (All of us–not just conservative Republicans–engage in it.) Yesterday he predicted Romney would win with 285 electoral votes to Obama’s 253.  In contrast the NY Times’ Nate Silver, who conducts state-by-state statistical analysis of all the major polls, predicted Obama would win with 314 electoral votes. (With Florida’s 29 electoral votes up in the air as of this writing Obama leads the electoral college 303 to 206.) Last Friday Silver reported that Obama’s chances of winning were better than 4 in 5 (83.7%, actually):

[The polls] represent powerful evidence against the idea that the race is a “tossup.” A tossup race isn’t likely to produce 19 leads for one candidate and one for the other — any more than a fair coin is likely to come up heads 19 times and tails just once in 20 tosses. (The probability of a fair coin doing so is about 1 chance in 50,000.)

Instead, Mr. Romney will have to hope that the coin isn’t fair, and instead has been weighted to Mr. Obama’s advantage. In other words, he’ll have to hope that the polls have been biased in Mr. Obama’s favor.

Silver explains that polls might provide an inaccurate election forecast due to statistical sampling error, voters changing their mind after the “snapshot in time” the poll represents, and statistical bias (“the polls are not taking an accurate sample of the voter population”). He states

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.

In other words, either the polls on the Friday before the election show Obama is the odds-on favorite or they contain statistical bias outside the range Silver accounts for in his models.