- Joined
- Dec 17, 2013
- Messages
- 11,698
- Supports
- Brazil, Arsenal,LA Aztecs
It wasn't. It was empirically less accurate. You can't post a reply based off of an article about a poll which specifically cites a popular vote prediction, and then claim that your point didn't have anything to do with the popular vote.
You're just strawmanning my post at this point or you really don't understand I am making two separate points. In the case of the latter I will try to explain again.
One is in general about comparing percentages, the other is about flaws in methodology and how to correct that.
The point is the USC/Dornsife polling (different entity than the LA Times so its not "patting itself on the back") that delved into the data to a deeper level revealed more information about the Trump's prospects of winning than the other older methodology's. In other words the information value of each poll is not limited to the single reductive conclusion you keep making out.