Sestak
jk2020
Atlas Icon
![*](https://talkelections.org/FORUM/IMG/star.gif) ![*](https://talkelections.org/FORUM/IMG/star.gif) ![*](https://talkelections.org/FORUM/IMG/star.gif) ![*](https://talkelections.org/FORUM/IMG/star.gif) ![*](https://talkelections.org/FORUM/IMG/star.gif)
Posts: 13,284
![](./avatars/Other/O_MA.gif)
|
![](https://talkelections.org/FORUM/IMG/post/xx.gif) |
« on: February 02, 2018, 03:38:46 PM » |
|
Oh, come on, Bojicat, this falls squarely under Nate Silver's First Rule of Polling Errors: the polling error usually occurs in the opposite direction from what the media expects.
Brexit: Polls show narrow Remain Media expects wide Remain Result: narrow Leave
US presidential 2016: Polls show narrow Clinton Media expects wide Clinton Result: narrow Trump
France 2017: Polls show very wide Macron Media expects a close race Result: even wider Macron
Virginia Governor 2017: Polls show narrow Northam Media expects an extremely close race Result: Comfortable Northam
Similarly, Czechia Presidential 2018: Polls show tossup/narrow Drahos Media expects comfortable Drahos Result: narrow Zeman
In all cases, it wasn't that the polling was widely off; the error was small. It was the media interpretation that was way off.
|