This election will be very close. I will be very surprised if the difference in the popular vote is greater than 2 percent. And there are plausible scenarios for each candidate to win. But the President is a clear betting favorite, despite consistently trailing in the Likely Voter model of the two most well known daily tracking polls.
Nate Silver's forecasting model is the best evidence for why the President remains the betting favorite. Silver has become quasi-famous in this cycle and his track record justifies his status as the go-to prognosticator for political forecasting. In 2008 his popular vote prediction was off by just 1 point. And he correctly predicted every jurisdiction except for Indiana and Nebraska's 2nd congressional district. In the last 2 congressional races he has correctly predicted all but one Senate race.
Silver is a Democrat and he has (predictably) become a target of some right wing bloggers who claim bias in his forecasting. But what's most interesting about his track record is that almost all of his incorrect predictions were very close and none exhibit a bias against Republicans. In 2008, he thought Barack Obama would win the popular vote by 6.1 points. He won by 7.2. And Indiana and NE-2 were both carried by the President. The only Senate race he has gotten wrong was Alaska in 2008. There he predicted the Republican nominee to win, but the race ultimately went to a more moderate Republican winning as a write-in candidate. So none of his (very few) mistakes involved mistakes in favor of the Democratic party.
Silver currently projects that the President will win the popular vote by a point and a half. He projects that the President will lose Indiana, North Carolina, Florida and Nebraska-02. Every other locality will be carried by the same party that won them in 2008. This would result in an electoral college tally of 303 for Obama, 235 for Mitty Romney.
So how does Mitt get to 270? Well, it would take a number of upsets in specific states. Here is a breakdown of the states where Romney is most likely to score an upset:
1. Colorado, 9 Electoral Votes, 42.3% chance of a Romney win. This gets him to 244.
2. Virginia, 13 EV, 40.2%. These two get him up to 257.
3. New Hampshire, 4 EV, 29.5%. Now he's up to 261.
4. Iowa, 6 EV, 27.7%. Up to 267....still needs one more upset.
5. Ohio, 18 EV, 26.5%. Now he crosses the line with 285 Electoral Votes.
If Romney pulls all five of these upsets, and avoids being upset in Florida (36.2% chance of an Obama win) he will be the Next President of the united states. Put differently, he would need to win every state that he has at least a 25% of winning in order to be elected.
The odds of pulling all five upsets are quite long. If you multiply each of these probabilities, you find that there is a less than one percent chance of pulling all 5 upsets. But this is a somewhat misleading number because the outcome of these states are not completely discreet events as the voters there will be influenced by similar forces, such as the tenor of the news for the next nine days, the jobs report that comes out on Friday and of course, the weather, which could drive turnout up or down significantly.
And Romney could also win Wisconsin (10 EV, 13.7% chance for Romney), which would allow him to lose several of the above states and still get to 270, assuming he doesn't also lose North Carolina. (13 EV, 17.5% chance for the President).
So the President is not 99 to 1 favorite. But he is about a three to one favorite.
No comments:
Post a Comment