jillian
Princess
Yes. You are missing the fact that EVERY other pollster got it grossly wrong and this guy was almost perfect.
Rassmusen and it's ilk should go out of business. They are either liars or stupid. Anyone who trusts them in the future is a damned fool.
Thanks and fair enough but that is my question -- did he get it perfect as several have said here and several in the media have said or - as you say and I agree -- almost perfect (which i did too - i made apick on florida but for romney so that almost surely will prove wrong). THAT is the question. And there is a big difference between perfect and almost perfect (which is just another way for saying imperfect).
Although the copmarison to the pollsters is apples to oranges and silver would be the first one to tell you he is not a pollster. My guess is that he probably would even say that a pollster who said it was +1 romney (as I think rassmussen did) with a 3.5 MOE didnt get it wrong (assuming obama ended up winnig within the MOE as i think he did for most polls)
Lets just get a few terms down Nate Silver, Pollster, and Real Clear aren't pollsters. They construct prediction models using many polls as data. Both Nate and Pollster, and a handful of lesser known analysts correctly predicted 51 out of 51 results. Than you had the analysts like Real Clear Politics who miscalled Florida. On the other extreme, you had unskewedpolls and analysts like Dick Morris and Karl Rove becoming a complete joke in their predictions.
Yes, Rassmussen and Gallup were among the worst poll organizations this go around. I can't say why Rassmussen was so terrible, but Gallup totally screwed up their LV algorithm, predicting that white turnout would be 78%, when it turned out to be 72%. Gallup also predicted that Republican turnout would exceed Democrats. Both of these organizations came in near the bottom of the list for accuracy.
rasmussen was horrible because they didn't survey cell phones (where a lot of people, especially younger people, now have their only phone contact; and he also decided that democrats were being oversampled even though that obviously wasn't the case.
Rasmussen miscalled SIX of TWELVE battleground states. A tie is automatically a miscall:
Ohio
Virginia
Florida
Iowa
Wisconsin
Colorado
But, Rasmussen nailed the margin in Pennsylvania, for Obama.
yep