What do the "Trump Tape" mean for November 8th?

As you probably know, a video (but the video part didn't show Trump) of Trump was released with crude comments about women.  They have been dubbed by social media and new outlets as the "Trump Tape".  Trump is no stranger to controversy, but this is creating a line in the sand for many republican leaders. Here is CNN's article of the status of republican congress members, and leaders: http://www.cnn.com/2016/10/08/politics/donald-trump-video-women-remarks-republicans/index.html . There is this suggestion that you can write in Pence instead of voting for Trump. However, write-in-laws are complicated and vary from state to state. In some cases, like Texas, if they didn't file paperwork saying they want to be a certified write in candidate, then the vote doesn't count and it's like you didn't vote at all. I would recommend looking  at ballotpedia.org for your state's information.  Now if you live in Texas and  voted for a certified write in candidate like Evan McMullin, that vote would count. Writing in someone who is not a candidate, has less of an effect on the election outcome, but if you vote consistently for one party or the other your absence of a vote for them has an (small) effect on  the outcome.  If people don't vote Trump and instead consider writing someone in or going third party, it may increase Hillary Clinton's chance of winning but that is complicated.  Now, if the third party and write in candidates DON'T change how a non-swing state votes, like Trump wins Texas, or Clinton wins California, the third party candidates don't impact the overall election much.  Now in swing states even a small percent of support for third party candidates can change the election.  If 10% of the people who voted for Romney in 2012 in North Carolina (about 5.1% of the overall vote) vote for Johnson, and everyone else votes for the same party  (which probably won't happen but this is an example) then the total would be 48% Clinton, 44.9% Trump, and 5.1% Johnson.  This would mean that North Carolina (a swing state) would flip sides from 2012. This doesn't mean you shouldn't vote for whoever you want. You can vote for whoever you want, but you can consider these things.  However if you look at my post from yesterday, you would notice if Trump got more than half  of the Johnson vote in Iowa, he could win it. In some cases like Michigan all the third party support couldn't help Trump win. So in my opinion going for a third party isn't going to change much, especially considering that some of the people voting third party this year are doing so because they don't want to vote for Hillary or Trump (I personally am making this decision). This is a continuing process and these things can change.  I also think Johnson may get a spike in support in the next couple of polls due to some republican-leaning votes dumping Trump. So yes, you can write in Pence, or vote Johnson if you want to dump Trump, and it won't necessarily win Clinton the election.  I will provide an update on this in about a week when post debate polls come out. I think Donald Trump will take a small hit in National polls, but in swing states it could have a larger effect.  I don't think all the voters who leave Trump will vote for Hillary, and I think most will go either back in undecided or the other category.

Disclaimer:  I am currently leaning towards voting for Johnson.  But you can vote for whomever you want.  

First Projection of the Election

Here is a current projection of the Swing States.  There is only one poll (a national poll) including data from after the Friday Trump video and WikiLeaks of Clinton's speeches.  If it is too close to call that means the confidence intervals overlap.  If it is outside the confidence interval, there is no overlap, and it is highly unlikely that the other candidate has a lead in the polls. We are still 29 days out, so anything can happen, but today it looks like Hillary will win.

Here is My Calls as of the data available today.

1. Wisconsin - Hillary, Outside of the confidence interval
2. Iowa - Too close to call, less than 1% of lead for Trump
3. Colorado - Hillary, Outside of the confidence interval
4. Michigan  - Hillary, Outside of the confidence interval
5.  Florida - Hillary, Outside of the confidence interval
6. Pennsylvania - Hillary, Outside of the confidence interval
7. North Carolina - Leans Hillary, but technically inside the confidence interval
8. Ohio - Too close to call, less than 1% of lead for Hillary
9. Virginia - Hillary, Outside of the confidence interval
10. Nevada - Hillary, Outside of the  confidence interval
11. New Hampshire - Hillary, Outside of the confidence interval

Assuming all other states with a winner-take-all system vote as usual, and excluding the two proportional states (Nebraska, Maine) to make this analysis easier. The electoral count would be 306 Clinton and 186 Trump.  Here is  a map I generated from 270towin.com:http://www.270towin.com/maps/wyoGy .

I will try to do this kind of analysis a few times before election day.


Full Results for those who care:

Note: I am using a bayesian gaussian conjugate analysis.  Technically it isn't the margin of error but rather outside the bayesian credible interval. These were all based on National Polls that included Johnson.  Other represents Jill Stein, and write in candidates.  The data is not fully normal yet, but it is nearly normal and approaching normal.

Wisconsin based on National
Hillary's Mean: 0.470415003827
Hillary's Std: 0.0034495460031859076
Hillary's CI: (0.46365401789813493, 0.47717598975665182)
Trump's Mean:  0.407319262907
Trump's SD: 0.007396991745285362
Trump's CI: (0.39282142549266735, 0.42181710032206615)
Other Mean: 0.0425827118082
Other Std:  0.0110470868148955
Other CI: (0.020930819516899922, 0.064234604099464881)
Johnson's Mean: 0.0796830214571
Johnson's SD: 0.009832086818359407
Johnson's CI: (0.060412485400202137, 0.098953557513913021)

Iowa based on National
Hillary's Mean: 0.440170105509
Hillary's Std: 0.0020669512929852073
Hillary's CI: (0.43611895541737877, 0.44422125560147774)
Trump's Mean:  0.447913044097
Trump's SD: 0.008354263417823627
Trump's CI: (0.43153898868117718, 0.46428709951376684)
Other Mean: 0.0386983448956
Other Std:  0.008928684250168961
Other CI: (0.021198445335939797, 0.056198244455262156)
Johnson's Mean: 0.0732185054975
Johnson's SD: 0.008249544533061003
Johnson's CI: (0.057049695323840029, 0.089387315671157763)

Colorado based on National
Hillary's Mean: 0.45891815616
Hillary's Std: 0.001846776782303797
Hillary's CI: (0.45529854017900528, 0.46253777214060565)
Trump's Mean:  0.413193888437
Trump's SD: 0.008570580462531743
Trump's CI: (0.39639585940368099, 0.4299919174700107)
Other Mean: 0.0449939206856
Other Std:  0.008684975201715829
Other CI: (0.027971682083646452, 0.062016159287619485)
Johnson's Mean: 0.0828940347177
Johnson's SD: 0.008056175488558737
Johnson's CI: (0.067104220907006185, 0.098683848528425183)

Michigan based on National
Hillary's Mean: 0.501309366174
Hillary's Std: 0.003442976812697726
Hillary's CI: (0.49456125562122444, 0.50805747672621249)
Trump's Mean:  0.40665815805
Trump's SD: 0.010910122711601732
Trump's CI: (0.38527471046871625, 0.42804160563202004)
Other Mean: 0.0305658890533
Other Std:  0.009116008330993849
Other CI: (0.012698841041813007, 0.048432937064843071)
Johnson's Mean: 0.0614665867226
Johnson's SD: 0.008396610702088301
Johnson's CI: (0.045009532154288892, 0.077923641290882181)

Florida based on National
Hillary's Mean: 0.471000847818
Hillary's Std: 0.000989203137339085
Hillary's CI: (0.46906204529560452, 0.47293965034076174)
Trump's Mean:  0.449741932093
Trump's SD: 0.005515633760495897
Trump's CI: (0.43893148857032033, 0.46055237561529067)
Other Mean: 0.0273230872086
Other Std:  0.005083334417496808
Other CI: (0.017359934828954079, 0.037286239588287355)
Johnson's Mean: 0.0519341328804
Johnson's SD: 0.00494773265990459
Johnson's CI: (0.042236755061845217, 0.061631510698936333)

Pennsylvania  based on National
Hillary's Mean: 0.493603584814
Hillary's Std: 0.0010456577937836015
Hillary's CI: (0.49155413319794855, 0.49565303642988751)
Trump's Mean:  0.421488143274
Trump's SD: 0.005738549894200958
Trump's CI: (0.41024079215788078, 0.43273549439012082)
Other Mean: 0.0292716638062
Other Std:  0.005377160946381598
Other CI: (0.018732622012242614, 0.039810705600209099)
Johnson's Mean: 0.0556366081059
Johnson's SD: 0.005217412872650881
Johnson's CI: (0.045410666782984027, 0.06586254942872681)

North Carolina  based on National
Hillary's Mean: 0.464364093623
Hillary's Std: 0.0011323467803349592
Hillary's CI: (0.46214473471590578, 0.46658345253083855)
Trump's Mean:  0.454520203388
Trump's SD: 0.0054298785069183
Trump's CI: (0.44387783707417394, 0.46516256970214998)
Other Mean: 0.0277225171224
Other Std:  0.006087250197773754
Other CI: (0.015791725969859047, 0.039653308274900802)
Johnson's Mean: 0.0533931858661
Johnson's SD: 0.005858329379282931
Johnson's CI: (0.041911071273118425, 0.064875300459053292)

Ohio  based on National
Hillary's Mean: 0.456654533478
Hillary's Std: 0.001440496286259149
Hillary's CI: (0.45383121263720361, 0.45947785431906685)
Trump's Mean:  0.44892501008
Trump's SD: 0.005785961682916272
Trump's CI: (0.43758473356569932, 0.4602652865945886)
Other Mean: 0.0325710680011
Other Std:  0.0063368481254245925
Other CI: (0.02015107389981715, 0.044991062102481862)
Johnson's Mean: 0.0618493884406
Johnson's SD: 0.006079794988665313
Johnson's CI: (0.049933209229400359, 0.073765567651742595)

Virginia  based on National
Hillary's Mean: 0.509817492716
Hillary's Std: 0.0015233871311132062
Hillary's CI: (0.50683170880445305, 0.51280327662744052)
Trump's Mean:  0.383083512701
Trump's SD: 0.008130523801764756
Trump's CI: (0.36714797887416767, 0.39901904652797687)
Other Mean: 0.0368006824832
Other Std:  0.007828013398694671
Other CI: (0.021458058151221614, 0.05214330681509869)
Johnson's Mean: 0.0702983120998
Johnson's SD: 0.007358225605696471
Johnson's CI: (0.055876454922535256, 0.084720169277106266)

Nevada  based on National
Hillary's Mean: 0.459625107351
Hillary's Std: 0.0021089007542614336
Hillary's CI: (0.45549173782530528, 0.46375847687594884)
Trump's Mean:  0.433940580643
Trump's SD: 0.004902596091165791
Trump's CI: (0.42433166887329182, 0.44354949241215541)
Other Mean: 0.0370111979379
Other Std:  0.006325662019835855
Other CI: (0.024613128200687256, 0.049409267675189592)
Johnson's Mean: 0.0694231140687
Johnson's SD: 0.006069913618124712
Johnson's CI: (0.057526301987917139, 0.081319926149504423)

New Hampshire  based on National
Hillary's Mean: 0.464219636203
Hillary's Std: 0.001973036141367581
Hillary's CI: (0.46035255642532358, 0.46808671597987628)
Trump's Mean:  0.420693735144
Trump's SD: 0.009608656541361667
Trump's CI: (0.40186111438313665, 0.43952635590490474)
Other Mean: 0.0400485340181
Other Std:  0.009615724105246576
Other CI: (0.021202061086498983, 0.058895006949612833)
Johnson's Mean: 0.0750380946353
Johnson's SD: 0.008782222995839079
Johnson's CI: (0.057825253859279432, 0.092250935411367543)

VP Debate Response/ 2nd Debate Preview

The VP debate was a little surprising because the polls showed a Pence win. I think this may be the contrast Pence provides to the Trump campaign. Pence is a 'typical' republican VP candidate when Trump is anything but 'typical'. Kaine did interrupt a little too much but he didn't do anything horrible. Kaine didn't have to win his goal was to because not totally screw up. The recent release of audio where Trump used sexually explicit language to describe an encounter with a woman (here is a story on it from CNN: http://www.cnn.com/2016/10/07/politics/donald-trump-women-vulgar/index.html ) will be brought up on Sunday. At the time of this posting Trump had recently released a video acknowledging the audio was true and apologizing for his actions. He also mentioned that Bill Clinton was involved in this kind of activity before. I think Trump may bring up Monica Lewinsky, and Gennifer Flowers in the debate.  Still, when you are already behind you can't afford to have something like this come out.  This debate will be crucial to Trump if he wants to be President.  He needs to show he can be presidential. I found his response to be professional and he apologized for once.  Still I think most American would find his comments offensive (but not everyone may agree with this). This next debate will definitely be interesting.

I am suspending using my twitter account

I have a twitter account for posting some blog posts and following comments, and the occasional straw poll (not for official use but for more of a feel).  I posted two polls tonight about who won the debate, one with  just Pence and Kaine and another with Pence, Kaine and no one (to provide another option).  This poll began to attract negative and unfair comments from anonymous accounts that have no relation to the poll itself.  At time the posting the poll showed 96% and 97% for Pence.  I believe that this is because numerous accounts with words that suggested strong political views.  I am not deleting the account but I am no longer using it to post, because I don't think anything beneficial can come of it.  If you want to see my account for reference my handle is @balexanderstat .  I am turning off comments on this post.   I am attaching screenshots of some of the posts for record. This not mean that I will let this stop the work I am doing. I will just stick to this blog.

VP Debate preview

What this debate means for the Trump Campaign:

Trump hasn't been doing well in the past week.  Another year of a partial tax return (from 1995 state returns from New York, Connecticut, and New Jersey) was released by the New York  Times (here: http://www.nytimes.com/2016/10/02/us/politics/donald-trump-taxes.html?_r=0) . It's been confirmed or denied by the Trump Campaign.  No tax returns have been officially released by the campaign (but the official preparer listed on the form thinks the returns are accurate).  The Trump Foundation can no longer collect donations in the state of New York (An article is here: https://www.washingtonpost.com/politics/trump-foundation-ordered-to-stop-fundraising-by-ny-attorney-generals-office/2016/10/03/1d4d295a-8987-11e6-bff0-d53f592f176e_story.html).  And Trump said in a statement at a fundraiser: "When you talk about the mental health problems, when people come back from war and combat and they see things that maybe a lot of folks in this room have seen many times over and you're strong and you can handle it but a lot of people can't handle it. They see horror stories, they see events you couldn't see in a movie, nobody would believe it,", (Source: http://www.cnn.com/2016/10/03/politics/donald-trump-ptsd-suicide/). The campaign created a website under the name clintonkaine.com attacking Clinton and Kaine.  Tomorrow Mike Pence needs to help convince voters that Trump is not corrupt, or a bigot. Looking at post debate polls on Pollster it looks like Trump right now isn't going to win most of the swing states.  The data is very limited but there were 10 post debate polls on Pollster in both the Trump vs. Clinton and Trump vs. Clinton vs. Johnson category.  National poll aren't good predictors of election outcomes, but it provides a relative measurement. Trump lead in 1 of those polls but it was in the margin of error.  Clinton lead the other 9, and 5 polls were outside of the margin of error.  This pretty similar to before the debate but Clinton appeared to be winning before debate.  I think Pence acts more calm and sensible than Donald Trump, so I think he may help the cause.  But is it too late? He can still win but will he? Probably not. Pivot, a online market for predictions used by enthuasists puts the odds at 77/23 Clinton/Trump.  Pivot hasn't been scientifically tested but provides a rough guess of the odds right now.

What this debate means for the Clinton Campaign:

I have thought that Clinton would win for a couple months now. I don't support her, but she has an easier road to the White House than Trump because of Obama's high favorability and general support of her party.  If she can win all the blue states she only has to win Pennsylvania, Virginia, Wisconsin, New Hampshire, and Michigan to win. I say that because right now they feel "safe" and I think she could even win all the swing states (but it would be difficult).  It's not over until the polls close on November 8th. But the Clinton campaign appears to have  a lead  or be within the margin of error (unscientific just based on recent polls) in most of the swing states. Kaine could bomb (which I don't think he will) and it wouldn't matter.  The leaked audio of Clinton suggesting that Bernie Sander's supports are baristas that live with their parents, could hurt her but young voters have low turnout rates and she may not need their votes.  It's more of a war of attrition, if you let Trump be Trump and continue to sit on your lead, you can win.  

Do endorsements matter to Trump Supporters?

USA Today came out with an editorial declaring Donald Trump unfit for the presidency.  USA Today is known to be one of the least biased media organizations. Many other local and national newspapers have called out Donald Trump.  No former presidents have endorsed him.  Very few senators and congressmen have endorsed him.  But yet Donald Trump could still pull off at least 40% or even 45% of the overall vote in November. The question is how is Donald Trump surviving as a candidate when so many people publically denounce him as "unfit for the presidency" or "inexperienced" or "unprepared".  I am going to look at this poll from The Economist/YouGov about Trump supporters opinions on the media and government. I  choose it because it is more comprehensive than most polls and has demographic breakdowns of every question.  In my analysis a Trump supporter is someone  who will likely vote for Trump in November.  And I am assuming that most Trump supporters is voting for him because they like him as candidate in general regardless of their opinions of the other candidates (88% of polled Trump supporters had a "very favorable" or "somewhat favorable" view of him).

The data is here: https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/q8r0rkibs1/econTabReport.pdf

Since this is one poll you can't use these results as exact percentages of all Trump supporters but you can learn some things about the typical or average Trump supporter.

The Media

72% of polled Trump supporters (remember there is going to be some variability in this number) think the media is done a "worse than usual job" covering this election than previous elections. A 95% confidence interval would be (67.6%,76.4%) so we could safely say it is very likely that most Trump supporters do not think the media is doing a good job with this election.

For comparison 50% of Clinton supporters (and 58% overall) think the media has done a "worse than usual job")

The Government

6% of polled Trump supporters "Strongly approve" or "Somewhat approve" of Congress's job performance (the percentage for all responses was 13%, and 12% of republicans).  18% of polled Trump supporters were neutral.
A 95% confidence interval would be ( 3.8%,8.22%) for "strongly approve" or "somewhat approve" of the performance of congress.


95% of polled Trump supporters say the country is on the wrong track.
For comparison this number was 65% overall.

The  top 4 most important issue  (listed in order) of the Trump supporter in this poll are: Terrorism, Social Security, Immigration, The economy.

For comparison for all respondents the four top issues were : Social Security, the economy, Terrorism, Health Care.  For polled Clinton supporters they were : Social Security, the economy, health care, the environment/gun control (these were tied).

Conclusion

The typical Donald Trump supporter doesn't like the media coverage of the election, direction of this country, and Congress, more so than other people.  Because of this they may not care about endorsements as much as the usual voter.  Another interesting thing is that 51% of polled Trump supporters said they felt enthusiastic about Donald Trump, but only 38% of polled Clinton supporters felt enthusiastic about Hillary Clinton.   Since this is one  poll from a few days it's not going to give us the answers to the opinions of all Trump supporters but it can help to explain  Trump's consistent support regardless of his actions. These opinions may also explain why Trump won the primary with very few endorsements from politicians (compared to previous years).   Trump's supporters (especially those who have been with him since the primary) are not going to change their mind at this point.

First Debate Response

I watched the debate and various pre-debate analysis. I am watching the post-debate analysis.  Here are my initial thoughts on what tonight means.

I don't think anyone won.  Trump and Clinton both acted as I expected.  Trump was open about his opinion and wasn't afraid to call out Clinton but he didn't do it in a professional way.  Clinton is the better debater in my opinion, but as much as she said she prepared I was unimpressed.  I thought she seemed canned and rehearsed. It looked like she planned every statement she said about Trump and didn't do a good job improving.  If there is a shift in the polls, it will probably be small.  I will be monitoring the response of polls  and researching the statements asserted by both candidates.  I noticed a few things that I knew where not true from both sides, but I don't know a enough about either candidate to know if most of the statements tonight were factual.