Quote:
Originally Posted by Fighter of Foo
|
Which is exactly the point of the article I posted. That 538 site gives an average to strong poll weight in its overall calculations to Survey USA, which obviously is using heavily flawed weights in its polls given previous election cycles. If a website is using flawed polls to achieve a composite result, doesn't it stand to reason that the 'crap in, crap out' data model may apply in this case? Once again, I'm not even accusing anyone of bias. I'm just saying that sites like the one you mentioned are relying on highly flawed data. It should be a red flag for them.
If you seriously think that, under the best of circumstances, 52% of all voters in PA in November will be Democrats, I've got some fine oceanfront property to sell here in Missouri to you at a deeply discounted price.