Predictions for the 2016 Election

[Nous]

I account for them with logic and common sense. A few thousand max on landline calls doesn't equate to the millions who will be voting. One can have confidence in these polls, but they aren't the real deal.

What do you mean that these polls are not “real”?

Like all other scientific studies, a poll cannot and is not supposed to be a clone of the election itself, but, rather, uses a (hopefully representative) sample of voters.

Obviously, even if all the polls used in the RCP average are wrong, they definitely don't lead to the conclusion that Trump will win the election.

I believe personally that this is one of the most unpredictable elections.

Why?

 

[Nous]

Well within the margin of error.

(1) No margin of error is calculated for the RCP average of poll results. How would one calculate a margin of error for such an average of poll results? The polls had a variety of sample sizes--one was 447 likely voters and another near 3,000. These polls used different methodologies.

(2) Clinton's average point spread of 5.5 is not within the margin of error of any of the polls used in the average.

Given the unpredictability of this race, who knows what will happen before the election? Who knows if there are significant numbers of people who will vote Trump but won't admit it publicly to pollsters? Who knows about the turnout? Who knows about how many will vote for 3rd party candidates?

Why do you believe that the polls are less accurate in this election than they have been in past recent elections?

And why do you believe that are or may be lots of Trump supporters who "won't publicly admit to pollsters" who they will vote for?

In any case, I did say that one is welcomed to change one's prediction as the circumstances dictate. After all, for all we know now, Trump may drop out before the election.

There is also a reasonable possibility of a Trump win.

From what fact(s) do you draw that conclusion?

 

[jonathan180iq]

Hillary will win before polls close in the Midwest. And although the news outlets will spend all night attempting to portray the race as an actual thing, highlighting how Trump is closing the gap as results pour in from a few midatlantic states, the outcome will already be known.

If NC goes Blue it will be over before Western Tennesseans stop voting... This election is done. Trump is a joke and the Republicans missed an easy opportunity to win the White House. Any of the mainline GOP candidates would have been a slam-dunk win. The electorate is foolish, undereducated, and delusional.

 

[jonathan180iq]

I find it odd that polls were a big talking point for Trump, a topic to brag about for most of the primary season - but once the national shifts came (as was to be expected) they're suddenly faulty, don't represent the majority of people, and the elections are rigged...

Are we really so blinded by our biases that we fail to hold people accountable to their own logic?

 

[sun rise]

(1) No margin of error is calculated for the RCP average of poll results. How would one calculate a margin of error for such an average of poll results? The polls had a variety of sample sizes--one was 447 likely voters and another near 3,000. These polls used different methodologies.

(2) Clinton's average point spread of 5.5 is not within the margin of error of any of the polls used in the average.

Why do you believe that the polls are less accurate in this election than they have been in past recent elections?

And why do you believe that are or may be lots of Trump supporters who "won't publicly admit to pollsters" who they will vote for?

In any case, I did say that one is welcomed to change one's prediction as the circumstances dictate. After all, for all we know now, Trump may drop out before the election.

From what fact(s) do you draw that conclusion?

Those tracking polls note the number of undecided voters is higher this time around so they've stated that this makes the results more uncertain. I think that's a fair conclusion.

 

[Nous]

Hillary will win before polls close in the Midwest.

Wow, that would seem to be a prediction of an very unusual election, no? I say that because of all of the "solidly red" Southern states.

Trump is a joke and the Republicans missed an easy opportunity to win the White House.

I agree. We can never know, but I suspect that just about any of the elected officials who ran in the Republican primaries would be polling higher than Clinton at the moment.

We can only thank God that Republicans were so crazy in the primaries of this highly critical election (by which I mean, where the make-up of the Supreme Court hangs in the balance).

 

[Nous]

Those tracking polls note the number of undecided voters is higher this time around so they've stated that this makes the results more uncertain.

I haven't examined any of the polls used in the RCP average. But, if true, yours is an interesting point. I am not sure what to make of the fact of larger numbers of undecideds at this point in this election.

 

[jonathan180iq]

Wow, that would seem to be a prediction of an very unusual election, no? I say that because of all of the "solidly red" Southern states.

I agree. We can never know, but I suspect that just about any of the elected officials who ran in the Republican primaries would be polling higher than Clinton at the moment.

We can only thank God that Republicans were so crazy in the primaries of this highly critical election (by which I mean, where the make-up of the Supreme Court hangs in the balance).

Yep - it's funny that Supreme Court selections seem to be the only rallying cry of the GOP to get Trump elected at the moment, since it would have been just as true a year ago as it is today. If that was really the motivating factor behind their presidential vote, why didn't they chose someone who had a chance of actually winning?

I mean, these outcomes are not (or at least should not) be surprising to be people who actually pay attention... Trump never had a chance in hell of winning a national election. Did people REALLY not see this coming?

 

[sun rise]

I haven't examined any of the polls used in the RCP average. But, if true, yours is an interesting point. I am not sure what to make of the fact of larger numbers of undecideds at this point in this election.

The folk at 538 took a stab at it

http://fivethirtyeight.com/features/election-update-where-the-race-stands-with-three-weeks-to-go/

2. What’s the degree of uncertainty?

Let me approach this question in two ways. First, there’s uncertainty as our model attempts to define it. The most important factors in that calculation are the number of days until the election and the number of undecided and third-party voters. Obviously, we’re getting closer and closer to Election Day, with early voting already underway in many states. But the number of undecided voters remains fairly high (although it’s declined slightly). In national polls, about 85 percent of the vote is committed to Clinton or Trump, as compared with around 95 percent that was committed to President Obama and Mitt Romney at this point in the campaign four years ago. Those unpredictable undecided and third-party voters are why our models show both a better chance of a Trump victory than most of our competitors and a better chance of Clinton winning states like Texas.

 

[Nous]

Yep - it's funny that Supreme Court selections seem to be the only rallying cry of the GOP to get Trump elected at the moment, since it would have been just as true a year ago as it is today. If that was really the motivating factor behind their presidential vote, why didn't they chose someone who had a chance of actually winning?

I mean, these outcomes are not (or at least should not) be surprising to be people who actually pay attention... Trump never had a chance in hell of winning a national election. Did people REALLY not see this coming?

It is just so very funny and wonderfully ironic. The President is in so many ways a figurehead, but not in his/her ability to get someone on the Supreme Court, and thereby affect things for decades to come. So Republican voters nominated the one candidate who never had a chance--the one candidate for which the release of such vulgar tapes could have been foreseen! If this election were a movie, it would be unbelievable.

 

[Nous]

The folk at 538 took a stab at it

http://fivethirtyeight.com/features/election-update-where-the-race-stands-with-three-weeks-to-go/

2. What’s the degree of uncertainty?

Let me approach this question in two ways. First, there’s uncertainty as our model attempts to define it. The most important factors in that calculation are the number of days until the election and the number of undecided and third-party voters. Obviously, we’re getting closer and closer to Election Day, with early voting already underway in many states. But the number of undecided voters remains fairly high (although it’s declined slightly). In national polls, about 85 percent of the vote is committed to Clinton or Trump, as compared with around 95 percent that was committed to President Obama and Mitt Romney at this point in the campaign four years ago. Those unpredictable undecided and third-party voters are why our models show both a better chance of a Trump victory than most of our competitors and a better chance of Clinton winning states like Texas.

Thank you. I haven't looked at 538's models in any detail--though I do get the impression they are more sophisticated than any poll average. Clinton winning Texas seems to be a very strange outcome.

 

[jonathan180iq]

Thank you. I haven't looked at 538's models in any detail--though I do get the impression they are more sophisticated than any poll average. Clinton winning Texas seems to be a very strange outcome.

http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo

88% chance of victory for Clinton.

 

[columbus]

Thank you. I haven't looked at 538's models in any detail--though I do get the impression they are more sophisticated than any poll average.

The main reason I like 538 is that they have a range of polling models and they will explain what the different ones mean until your eyes glaze over.
They are the most transparent political pundits on the planet.
Tom

 

[Adramelek]

Donald Trump will win this presidential election in an upsetting landslide just as Ronald Reagan did in 1980 c.e., and the "Brixit" movement this year in the UK. Just like with all things in the cosmic order, the political world is cyclical.

Hillary Clinton is a two faced, lying, corrupt, incompetent, power and money hungry, conniving, bought and paid for establishment political whore. A curse upon her candidacy! And after she is defeated in this cycle may she and her political whore of a husband go into hiding; never to be seen or heard from again!

 

[Augustus]

(1) No margin of error is calculated for the RCP average of poll results. How would one calculate a margin of error for such an average of poll results? The polls had a variety of sample sizes--one was 447 likely voters and another near 3,000. These polls used different methodologies.

All likely between 4-2.5% stated margin of error

(2) Clinton's average point spread of 5.5 is not within the margin of error of any of the polls used in the average.

Yes it is. Would fall within MOE for a 3% poll.

True margin of error has historically been larger than 'textbook' margin of error anyway.

Given recent polls have ranged from Trump +2 to Clinton + 11 that's pretty clear.

Why do you believe that the polls are less accurate in this election than they have been in past recent elections?

And why do you believe that are or may be lots of Trump supporters who "won't publicly admit to pollsters" who they will vote for?

This election has dealt with more unpredictable events and volatility in poll results.

Some people might not want to admit they will vote for him given his controversial nature. Anonymous polls have tended to report higher numbers for Trump than those carried out personally which might suggest there is something to it.

From what fact(s) do you draw that conclusion?

Most results fall within the margin of error, and there are still 3 weeks to go so he has a reasonable chance, even if the odds are against him winning.

 

[Augustus]

88% chance of victory for Clinton.

These % figures are not particularly useful, just look at how the figures have changed.

Clinton has gone from 77 to 49 to 80 to 53 to 88.

They are 'if the election was held today' figures, not the actual percentage chance of winning the election. The 77% didn't help you predict the 49%, and the 49% didn't help you predict the 88%, so the 88% doesn't help you predict what happens next.

 

[Papoon]

More shocking revelations on both sides as the Mossad and FSB conduct a Wikileaks War. General confusion and distress in The Herd. Domestic terror reaching shocking levels with a little help from aforesaid agencies...Martial law in time for global meltdown accelerated by failure of all major US banks as the derivatives market collapses. Merry Christmas.

 

[meghanwaterlillies]

More shocking revelations on both sides as the Mossad and FSB conduct a Wikileaks War. General confusion and distress in The Herd. Domestic terror reaching shocking levels with a little help from aforesaid agencies...Martial law in time for global meltdown accelerated by failure of all major US banks as the derivatives market collapses. Merry Christmas.

So who would you vote for than?

 

[Lighthouse]

What do you mean that these polls are not “real”?

Like all other scientific studies, a poll cannot and is not supposed to be a clone of the election itself, but, rather, uses a (hopefully representative) sample of voters.

Obviously, even if all the polls used in the RCP average are wrong, they definitely don't lead to the conclusion that Trump will win the election.

Why?

Just as you said, it isn't a clone to the final one.
I don't think anyone or any polls can already conclude anything weeks in advance. It can take a small sample size which could lead to predictions and assumptions and allow people their freedom to use that in whichever way they might.

 

[Nous]

http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo

88% chance of victory for Clinton.

Thank you. I notice their "polls-plus" forecast is a healthy 85% to 15%.