Political Polling October 17, 2016

[StatsCougar]

This week's polling continued the trend that was seen in previous weeks, with Mr. Trump falling farther behind. HRC is now averaging 349 Electoral votes (Summary statistics below.) The best Mr. Trump did was have HRC receive 281 (The next lowest was 299). Unless about all undecided votes break from Trump, he has no chance to be President. A good last debate might help, but I doubt it.

There are currently only three toss-up states. Arizona, Mississippi and Alaska. Alaska is following the trend of many states away from Mr. Trump. While I expect on election day Alaska will still vote Red, it is closer than it should be. Arizona is dog fight. With the concerns the state has with illegal immigration, the vote should not be close. It looks like only John McCain can save Arizona, but I doubt he would try to do it for Trump. Mississippi is a toss-up due to the lack of polling information.

Texas has shift to only Lean Republican. And many of Lean Democrat state are now solidly behind HRC. Ohio is now Leaning Democrat, when two weeks ago it was leaning Republican.

There is some good news in the polls for Mr. Trump. The first is the some recent polls have Ohio going back to him. without Ohio, there would be no way for Trump to win. While North Carolina is Lean Democrat, it appears to have stalled there. There is still hope it could swing back red.

Now what some of you have waited for, how often does Evan McMullin win Utah. The answer is never. Now this is not due to him not having a change. The methodology I'm using does not allow for trending candidates like Mr. McMullin. With the steam he is gaining, his victory is in the realm of possibility. (It also helps the almost 50% of votes for Mr. Trump and HRC are voting to prevent the other major candidate from winning.) The other issue is there was a major poll released from CBS that had Trump at 37%, much higher than other recent polls. This polling is likely very wrong, or the polling was taken well before it was published. I expect more polls from Utah to be coming out soon that will add clarity to the state of the race.

HRC Electoral Count:
Min. 1st Qu. Median Mean 3rd Qu. Max.
281.0 341.0 351.0 348.7 352.0 406.0

[BroncoBot]

Interesting that the southern border states are now all trending democrat. I don't think that's simply a "we don't like Trump result". I'd bet the open immigration policy is having a larger than expected boost for the dems in those areas. yipee.

[BlueK]

This week's polling continued the trend that was seen in previous weeks, with Mr. Trump falling farther behind. HRC is now averaging 349 Electoral votes (Summary statistics below.) The best Mr. Trump did was have HRC receive 281 (The next lowest was 299). Unless about all undecided votes break from Trump, he has no chance to be President. A good last debate might help, but I doubt it.

There are currently only three toss-up states. Arizona, Mississippi and Alaska. Alaska is following the trend of many states away from Mr. Trump. While I expect on election day Alaska will still vote Red, it is closer than it should be. Arizona is dog fight. With the concerns the state has with illegal immigration, the vote should not be close. It looks like only John McCain can save Arizona, but I doubt he would try to do it for Trump. Mississippi is a toss-up due to the lack of polling information.

Texas has shift to only Lean Republican. And many of Lean Democrat state are now solidly behind HRC. Ohio is now Leaning Democrat, when two weeks ago it was leaning Republican.

There is some good news in the polls for Mr. Trump. The first is the some recent polls have Ohio going back to him. without Ohio, there would be no way for Trump to win. While North Carolina is Lean Democrat, it appears to have stalled there. There is still hope it could swing back red.

Now what some of you have waited for, how often does Evan McMullin win Utah. The answer is never. Now this is not due to him not having a change. The methodology I'm using does not allow for trending candidates like Mr. McMullin. With the steam he is gaining, his victory is in the realm of possibility. (It also helps the almost 50% of votes for Mr. Trump and HRC are voting to prevent the other major candidate from winning.) The other issue is there was a major poll released from CBS that had Trump at 37%, much higher than other recent polls. This polling is likely very wrong, or the polling was taken well before it was published. I expect more polls from Utah to be coming out soon that will add clarity to the state of the race.

HRC Electoral Count:
Min. 1st Qu. Median Mean 3rd Qu. Max.
281.0 341.0 351.0 348.7 352.0 406.0

The CBS poll in utah that has had a vastly different result than the others had a goofy methodology. It only asked between Hillary and Trump. Only if the respondent wanted to say "other" would it give them the option of selecting McMullin, Johnson or Stein. In other words, biased. It should be ignored. Dumb way to do it because it didn't match the way it will work in the voting booth. It's not like the other candidates' names are going to be hidden for the actual voter until they go through an extra step to be able to see them.

[StatsCougar]

The CBS poll in utah that has had a vastly different result than the others had a goofy methodology. It only asked between Hillary and Trump. Only if the respondent wanted to say "other" would it give them the option of selecting McMullin, Johnson or Stein. In other words, biased. It should be ignored. Dumb way to do it because it didn't match the way it will work in the voting booth. It's not like the other candidates' names are going to be hidden for the actual voter until they go through an extra step to be able to see them.

You have a point on why this is different and how it introduces bias. This, however, is a common practice in the polling world. There are several reasons for this. The first there less names for the respondent to remember. The second it that it can often inflate the number of people who say they will vote third party. Third is that this part of questionnaire is likely the one that CBS uses in all of its polling. As not all the candidates are on the ballot in every State, there is no need to list all of them. This allow for easier comparison across polls as there is no addition bias from asking different questions.

That being said, CBS was wrong in this case. McMullin is polling a high enough percent that his name should be included in the question, especially as he is now a major player in the race in Utah.

[StatsCougar]

Update: New poll came out today with Evan McMullin winning Utah. He is at 31% and Trump at 27%. This is beyond the margin of error. Running a simulation based on all the polls that have come out recently (except for the CBS poll due to possible bias), I currently have Evan McMullin winning Utah 3.1% of the time. Hillary wins 2.2%, Trump winning 94.7%.

While Trump is winning the vast majority of the time, there is a flaw in the methodology that I employ, as it does not take into account fast rising trends. I expect the trend for Mr. McMullin to continue and would not be surprised if come election day, he wins the Majority of Utah votes.

[SpiffCoug]

I this point, I'll call my shot: McMullin wins Utah.