347
Comments (54)
sorted by:
You're viewing a single comment thread. View all comments, or full comment thread.
10
WU_HAN_FRU [S] 10 points ago +10 / -0

Methodology:

In 2016, Hillary's Election Day morning polls averaged 7% inflation in her support.

On the assumption that this inflation is very likely to be even worse for Quid Pro Joe, I slashed 7% support from his totals, gave 6% to Trump, and threw away the odd 1%.

I also took the undecided totals and gave those to Trump.

My baseline polls were taken from Dave Leip's Atlas of Presidential Elections, which takes an average of the 3 most recent polls in each state. It reflects the current publicly available suppression polls.

Needless to say, this method produced some surprising results. Is it too optimistic? What factors did I fail to consider? I'm eager to hear criticism of all kinds. Except the faggoty kind, obviously.

7
Pepe 7 points ago +7 / -0

the thing i could see happening is that they targeted their poll manipulation this time, so your 7% might not pan out where it counts.

but i pray you're right...

5
WU_HAN_FRU [S] 5 points ago +5 / -0

Meaning they would target it AWAY from the large swing states, like Penn.?

3
Pepe 3 points ago +3 / -0

something like that... i could see

  • Play up trump in trump-heavy states to reduce trump turnout
  • Play down trump in biden-heavy and/or purple states to make Trump look like a doofus and increase Biden enthusiasm, make him appear more legitimate nationwide

the latter is what i think was employed in huge margins (as you say, on avg 7%) in 2016 for Crooked Hillary... but i could see them being more careful with it this time around

4
WU_HAN_FRU [S] 4 points ago +4 / -0

Other than these published polls, literally EVERYTHING points to a Trump win.

There is no way that Slow Joe outperforms Obama and Bubba, which he would if these polls were accurate.

3
JimmyJam 3 points ago +3 / -0

it doesn't really work that way - you have to dive deep into the poll methodology - CA for example wasn't 7% too high for HC in 2016 - you can't just give a 7% haircut to all states. PA for instance, in 2016 pollsters inaccurately weighted non-college graduates. Today, they are under sampling "did not previously vote" as likely voters. It is a very state specific thing.

1
WU_HAN_FRU [S] 1 point ago +1 / -0

Well, take Virginia, for example: I'm definitely not making any allowances for federal government workers. I'm also not accounting for a massive shift with black voters everywhere.