I have been following US presidential elections closely since late 1987 when I was 22 and started to read the Economist weekly. It is only with my discovery of Nate Silver when he started projecting the US election on the Daily Kos in late 2007 that I finally found someone that could give me the data and analysis I was looking for to understand what is going on in the US election. In 2008 he got 50 or 51 right, he only missed Indiana
I won a few bets on the US election about who would win - the last ones I took about three or four weeks ago I was agreeing to four to one odds that Obama would win. The only reason I had any confidence in making those bets was because of all the data Nate Silver provided.
Nate Silver had the states breaking out 332 Obama and 206 Romney - he got all 51 correct this time. In 2008 he missed one stated. He projected a popular vote of 50.8% Obama to 48.3% for Romney. At the moment we are at Obama 50.4% and Romney 48.1%. There are about 1 to 1.5 million votes still to come in and mainly in locations that are pro Obama. At the end of all the counting the vote percentage should be 50.6% Obama and 47.9% Romney. Nate Silver looks like he will end up being marginally high for both candidates.
There was a single state where the winner and loser were within one percentage point, Florida. This was also the only state at the end of the election that Nate Silver had as actually competitive.
For the last week or so I have been pointing out to people that this was not a close presidential election and that Obama had a lock on the electoral college. In the worst case scenario it would be Obama 281 to Romney at 257, but they seemed overly pessimistic. I was leaning towards Obama 303 to Romney 235 until last Friday when I became convinced from the Silver data that Obama would win Florida.
Nationwide polls are useless in predicting the US election, something I had realized years ago. Over the last week people were listening to nationwide polls that showed Obama and Romney close to each other, what they were not listening to were the polls showing Obama overwhelmingly winning all the swing states and even threatening to win North Carolina.
The closest Romney ever came in Nate Silver's projection of winning was a 1/3 chance on October 14th - the notional electoral college was Obama 285.6 to Romney 252.4 on that date and vote percentage of 49.9% Obama and 49.0% Romney.
The only surprise in this election was the Democratic win in the North Dakota Senate seat - admittedly it was a close race and there were few polls to go on. He did also lean towards the Republicans winning the Montana senate seat but called it a close race and the Democrats won.