Polls (Page 11)

Even taking all the averages at face value this race has been swinging wildly the last couple weeks and I have no reason to believe the swings are going to stop.

As you mention Trump is at worse low single digits away in a handful of swing states.  One bad wikileak or a stock market crash/world event can easily swing those numbers.

Here's a CNBC Oped talking about the wild poll swings.

You ever watch the twitter feed of the guy who runs Monmouth?  He's a raging Trump hater and it's no surprise that his bias makes it into his "scientific" polls.

As for the others, they're honestly just guessing on the turnout.  They have no idea what it is because it's anybodies guess.  Even with the events of the last couple weeks there can be massive swings in either direction on turnout that aren't being reflected in these polls.  That's why I always say to just watch the overall trends for each poll to get a flavor of things going up or down.

I read something earlier today that talked about the polls for the Brexit vote.  They were in the 8%-10% range of staying in the EU, but the pollsters were incapable of capturing the turnout of such a populist movement.  The end result ended up with the leave vote at +4% I believe. 

Just out of curiosity what do you mean by "correct for their internal bias"?

The irony is that a drastically biased poll in favour of Clinton could end up electing Trump.  If the polls show she has a >10% lead just before election day, the Clinton voters will think she has it in the bag and the panicked Trump voters will show up in record numbers.

I do agree with you there and as much as I want Trump to win I don't want it to be through people not voting because they think it's in the bag.

The media has a fine balance between disheartening Trump voters (which is what I feel their mission is) but not convincing Hillary voters they don't need to vote. 

I guess my question is more towards adjusting the bias.  Obviously we can take the raw data and apply it to the exact turnout of 2008 or 2012 and get a number, but I think you and I both could agree that this year is "different".  This is why I believe the pollsters have a really tough job, because they are being tasked with figuring out what the real bias is going to be based on pollster responses.   IMHO the polls that are using historical models will likely be found to be the least accurate. 

Yeah, I see your point.  No question the media is skewing the election though

There is no doubt that Trump's hardcore fans greatly outnumber Hillary's hardcore fans. But you can't underestimate people voting against a candidate. Tony's pointed out plenty of times in the past how Obama had the rockstar appeal and nobody got excited about Romney. But Romney received the fourth most votes of any presidential candidate ever (losing to Obama twice and Bush's second term). A lot of people unenthusiastic about Romney were still very enthusiastic about voting against Obama. 

So there's more people that love Trump than love Hillary, but I also think there's way more people that hate Trump than hate Hillary. How that'll play out in the election? ¯\_("/)_/¯

I think Hillary is going to win, but I won't be shocked if Trump does. I will be shocked if this election doesn't set the record for voter turnout. Nobody is sitting this one out. 

New WaPo poll released for some battlegrounds.