Quote:
Originally Posted by Slava
I had a hard time with this explanation. I think its an attempt to say "I wasn't really wrong" when in fact he wasn't anywhere close to right. It's more than just 1/100 here and there. Its those voters being in states that matter and as a result flipping those contests. (Unless I'm missing something?)
|
I think he's trying to say polling is a blunt instrument and the polling error here was not remarkable.
I object to the notion that his forecast was wrong as I said in the other thread he got way to much credit for winning coin flips previously.. In my opinion he was the most accurate Moddle as he predicted exactly how Trump would win if he did and that the firewall was not that strong in a close national race. And gave trump a 30% chance of winning. In a world of people telling him he was Sand bagging Clinton, where the RNC thought they were going to lose he was the one saying hey wait it's not over 30% is a big chance.
So the guy (Northrop) people are heralding as calling the race predicted Trump would get 52.5% of the two party vote. So that's a 5 point two party spread. When the ballots are counted the result will be Clinton plus 1.
So in terms of predicting the popular vote 538 was off by about 2- 2.5 whereas as Northrop was off by 6. So who made the more accurate prediction? Is the fact that the error was across where the race was decided meaningful?