“Uh, no he wasn’t. Missed several senate races and was wrong on predicted margins. He was mostly right, but not “spot on” to a “frigging tee””
Go look again and be sure to look at the “projected vote share” and not the “chance of winning.” for each state:
http://fivethirtyeight.blogs.nytimes.com/
He was so accurate for each state it’s mind boggling. He wasn’t just aggregating and averaging polls, he was using LOTS of other data points in his model like history, census data, house effects, statistical bias history, sample size, date, weather, cell phone percentage, auto/live interviewers, and lots of other apparently relevant stuff.
This is similar modeling that he used for years on baseball players. He’s trained for this for years and although new to the political scene in 2008, kind of, he’s not new to the numbers game.
And he only got one Senate race wrong. He thought the Republicans would get North Dakota but democrat Heitkamp got it. But it is interesting to note that she won by barely 0.9%.....and that was the only thing he missed for the entire night.
Yup....some people just can’t deal with the truth of this. Sorry you took so many hits out here yourself cause you did an excellent job.
You sound starstruck. RCP got 49 out of 50 states just by mindlessly averaging poll numbers.