There’s a nerd fight going on between Nate Silver and several other analysts about how to interpret the Presidential election polls. Silver is projecting Clinton as having about a 65% chance of winning. HuffPo and Princeton are projecting her at 98%. I have some thoughts over at my personal blog:

Put aside everything you know about the candidates, the election and the polls. If someone offered you a 50-to-1 or a 100-to-1 bet on any major party candidate winning the election, would you take it? I certainly would. I would have bet $10 on Mondale in 1984 if it was a potential $1000 payoff. And he lost by 20 points.

It seems a huge stretch to give 98 or 99% odds to Clinton, considering:

- Clinton has never touched 50% in the poll aggregates.
- There are still large numbers of undecideds and third party supporters who will doubtless vote for one of the two candidates (and Trump’s recent surge has come from fleeing Johnson voters).
- We have fewer live interview polls now than we did in 2012.
- As Nate Silver noted, the average difference between final polls and the election has been about two points.
Basically, I think Wang and HuffPo are not accounting enough for the possibility that the polls are significantly off. In the last 40 years, we’ve had one Presidential election (1980) where the polls were off by a whopping seven points. That’s enough for Trump to win easily (or for Clinton to win in a landslide).

HuffPo’s analysis seems kind of bizarre to me, actually. They currently have Clinton up 5 points in the polls. There is not a single national poll that as Clinton up by that much right now. The average polls advantage for Clinton is two points. Silver estimates that corresponds to a real advantage of three. If he’s right, Clinton has an advantage but any outcome is possible.

I don;t take much stock of polls, they are reliable at best there was a Jimmy Stewart movie decades ago, where he found a town that predicated everything perfectly, until they found out they did. polls are the same way, how many people would say they are going to vote for Trump to a pollster, do they divide up the independents up into former Democrats or Republicans. I don’t think polls are as accurate as they want you to believe.

0

That’s why I like 538 – they bake that sort of thing in.

One way I heard Nate Silver explain it that made a lot of sense to me is that there are 3 scenarios possible.

Which in a nutshell explains why he gives Clinton a roughly 2/3 chance of winning.

1

Yep I’d definitely back Silver.

0

If I could just make a statistical (non-partisan point).

If 10 polls all have a (roughly) similiar result, then the margin of error is no longer +/- 3 points. It is much less, because any statistical ‘noise’ is lessened by having polls from different organisations.

The only way that a significant numbers of polls can all be wrong

in the same directionis if there is a common across the board methodology error (which there may well be in this instance, I’m just making the general point).Therefore, if you’ve got a range of polls going from around a dead heat up to HRC +6, then the chance of the ‘centre’ being correct is greatly increased than if you just had one HRC +3 poll.

I don’t agree with the 99% chance of HRC win, but I think that there would really have to be a significant event between now and election day for the votes to go towards DJT.

The statement by the FBI on the lack of problems with HRC’s emails and the stats on early Hispanic voting also tend to indicate winds blowing in a blue direction.

1

I’ve never been less sure of something that was so heavily polled as this election.

0

Actually there’s quite a bit less polling done this election than in previous elections

0