The RCP average just before the primary only had Gingrich +5, he won by 12.6%.
That's a pretty large miss in my book.
The RCP average is frequently wrong since it includes old polls in it’s average. It cannot reflect late swings as old data holds the average down
There have been two different sets of data...two polls showing 15% leads, and two polls showing 5-7 % point leads. It needs to be determined which polls are older. This particualr poll was 1/28 and 1/29, I believe. How ‘bout the others?
I would add that RCP had Mitt up by +8 and +10.3 three and four days before SC. And as you pointed out Newt won by +12.6.
So to conclude about RCP predictive capabilities, they were 23 points off four days before SC.
The Insider poll called the surge just before the primary vote.
They were the only one that did.
FL is an unknown, but it is surely not as far apart as the press propaganda states.
Go Newt!