2016 Postmortem

rumdude

(448 posts) Thu Nov 3, 2016, 09:30 AM Nov 2016

Why does Nate Silver's "adjusted leader" always seem

to increase Trump's poll numbers by one or two points and decrease Clinton's by the same?

WTH is that all about? Have all these polls been historically unfair to Republicans? I don't think so. In fact, in 2012 did not all the polls show a closer race than the actual final numbers? Did not Obama beat Romney by wider margin than was reflected in the polls?

Were not these polls, therefore, favorable to Republicans?

3 replies

= new reply since forum marked as read

Highlight:

Why does Nate Silver's "adjusted leader" always seem (Original Post) rumdude Nov 2016 OP

He got burned in the primaries. Bleacher Creature Nov 2016 #1

no he didn't. Adrahil Nov 2016 #2

Bleacher Creature

(11,256 posts)

1. He got burned in the primaries.

Reply to rumdude (Original post)

Thu Nov 3, 2016, 09:42 AM

Nov 2016

He basically added in a fudge factor to make sure it doesn't happen again. I think it's pretty clear he went too far in the other direction.

Adrahil

(13,340 posts)

2. no he didn't.

Reply to Bleacher Creature (Reply #1)

Thu Nov 3, 2016, 09:52 AM

Nov 2016

He called over 90% of the primary races correctly.

He IS attempting to model house effects in his model. This is the first time he's done it this way, and sometimes it seems a bit wonky to me. For example, he still maintains a Clinton lean house effect for the ABC poll, which I think is now bullshit.

Reply to this discussion