Atlas Forum

Election Archive => 2016 U.S. Presidential General Election Polls => Topic started by: Tender Branson on July 16, 2016, 12:37:47 am



Title: VA: Hampton University: Clinton and Trump tied
Post by: Tender Branson on July 16, 2016, 12:37:47 am
New Poll: Virginia President by Hampton University on 2016-07-10 (https://uselectionatlas.org/POLLS/PRESIDENT/2016/polls.php?action=indpoll&id=5120160710250)

Summary: D: 39%, R: 39%, U: 22%

Poll Source URL: Full Poll Details (http://news.hamptonu.edu/release/HU-Poll%3A-Presidential-Candidates-in-a-Statistical-Dead-Heat-among-Virginians)


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Devout Centrist on July 16, 2016, 12:49:57 am
New Poll: Virginia President by Hampton University on 2016-07-10 (https://uselectionatlas.org/POLLS/PRESIDENT/2016/polls.php?action=indpoll&id=5120160710250)

Summary: D: 39%, R: 39%, U: 22%

Poll Source URL: Full Poll Details (http://news.hamptonu.edu/release/HU-Poll%3A-Presidential-Candidates-in-a-Statistical-Dead-Heat-among-Virginians)
Those undecideds.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Tender Branson on July 16, 2016, 12:51:26 am
New Poll: Virginia President by Hampton University on 2016-07-10 (https://uselectionatlas.org/POLLS/PRESIDENT/2016/polls.php?action=indpoll&id=5120160710250)

Summary: D: 39%, R: 39%, U: 22%

Poll Source URL: Full Poll Details (http://news.hamptonu.edu/release/HU-Poll%3A-Presidential-Candidates-in-a-Statistical-Dead-Heat-among-Virginians)
Those undecideds.

That's not surprising:

Voters have to choose between a total trainwreck (Hillary) and a racist fu**er (Trump) and the poll does not offer 3rd party choices.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Landslide Andy on July 16, 2016, 02:26:11 am
New Poll: Virginia President by Hampton University on 2016-07-10 (https://uselectionatlas.org/POLLS/PRESIDENT/2016/polls.php?action=indpoll&id=5120160710250)

Summary: D: 39%, R: 39%, U: 22%

Poll Source URL: Full Poll Details (http://news.hamptonu.edu/release/HU-Poll%3A-Presidential-Candidates-in-a-Statistical-Dead-Heat-among-Virginians)
Those undecideds.

That's not surprising:

Voters have to choose between a total trainwreck (Hillary) and a racist fu**er (Trump) and the poll does not offer 3rd party choices.

Don't be a salty Tender. ;)


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Ebsy on July 16, 2016, 02:35:30 am
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Erich Maria Remarque on July 16, 2016, 06:24:22 am
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.
What do you mean by "trying". They rang randomly in different counties. Or do you mean, that they should weight the responses to make crosstabs look more realistic?


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: JRP1994 on July 16, 2016, 06:29:36 am
Their last poll showed Clinton +9 back in February.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Fusionmunster on July 16, 2016, 06:31:00 am
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

54% men and its tied!?

Safe D. Virginia is solid blue.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Fmr President & Senator Polnut on July 16, 2016, 06:58:49 am
I'm not bothering with any polls until a week or so after the Dem convention finishes.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: heatcharger on July 16, 2016, 09:15:45 am
Hampton also had Warner up 25 points over Gillespie, but every pollster was was off in that one.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Pro-Life Single Issue Voter on July 16, 2016, 09:28:29 am
Why is it an Atlas tradition to try to "unskew" every poll that has a Republican doing relatively well, but to mock anyone who questions a poll that has Democrats doing relatively well?  Of course there will always be outliers on both sides, but sometimes outliers can be the truth, especially in this day of "herding".


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: pbrower2a on July 16, 2016, 10:07:42 am
Tie or leader with less than 40%? Worthless even if valid.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: michelle on July 16, 2016, 10:50:08 am
Interesting result. However, it is an outlier and only has a B grade (plus, it's also not very predictive) on FiveThirtyEight, while Marist and Fox both have A's. The Marist poll that has Clinton up seven also has a larger sample size. And it does seem to be oversampling men.

However, I wouldn't throw it out. Just put it in the average. It's only July.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: pbrower2a on July 16, 2016, 10:56:26 am
Tie or leader with less than 40%? Worthless even if valid.

::) ::) ::) ::) ::) ::) ::)

Kentucky:

()

Look at all the undecided. Because the leader has under 40%  my system calls this a tie.

So why do I call it a tie? Because nobody is close to winning.

In 2008 I saw polls of Tennessee in which Barack Obama led John McCain 39-37% or something to that effect.

Maybe I would do better with a map that distinguishes 45% and a lead from 43% and a lead. Maybe I could 'yellow' the reds to orange and the 'blues' to green. Still, in a binary race, a lead with less than 40% isn't even close to winning.

Maybe after the Conventions. 


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Xing on July 16, 2016, 11:16:41 am
Why is it an Atlas tradition to try to "unskew" every poll that has a Republican doing relatively well, but to mock anyone who questions a poll that has Democrats doing relatively well?  Of course there will always be outliers on both sides, but sometimes outliers can be the truth, especially in this day of "herding".

To be fair, some blue avatars do the same thing.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: The Other Castro on July 16, 2016, 12:01:49 pm
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

Hey, it is 2016. A lot can happen.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Kaine for Senate '18 on July 16, 2016, 01:51:23 pm
Hard to take those crosstabs seriously. Although, if it's tied with those crosstabs, that's also a good sign.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Tender Branson on July 16, 2016, 01:55:13 pm
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

While unlikely, it's not out of question:

If males are more motivated to vote this year than women, it could produce an electorate that is overall more male than female.

For example if 51% of the population is female and they only have 55% turnout, while males (49%) have 65% turnout, you could get the electorate that the poll shows.

But yeah, that's still unlikely.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: psychprofessor on July 16, 2016, 02:12:56 pm
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

While unlikely, it's not out of question:

If males are more motivated to vote this year than women, it could produce an electorate that is overall more male than female.

For example if 51% of the population is female and they only have 55% turnout, while males (49%) have 65% turnout, you could get the electorate that the poll shows.

But yeah, that's still unlikely.

With Trump and Pence, women will be motivated to vote. Especially black and latina women. I actually think Colorado and Virginia are lost causes for Trump. He needs Ohio, Pennsylvania, Michigan, Iowa, etc...I don't think he gets it, but that's where he should be focusing.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: pbrower2a on July 16, 2016, 03:56:26 pm
Why is it an Atlas tradition to try to "unskew" every poll that has a Republican doing relatively well, but to mock anyone who questions a poll that has Democrats doing relatively well?  Of course there will always be outliers on both sides, but sometimes outliers can be the truth, especially in this day of "herding".

As one of the most-likely accused...

1. Some pollsters are really awful. (These folks had Obama winning Missouri in 2012... ha, ha, ha!) Is it any surprise that I distrust outliers?

When one starts seeing the same outliers, then maybe something has changed.

2. I am as likely to go along with the common wisdom. In 2008 I could hardly believe that Barack Obama had a chance to win Virginia, a state that had gone for every Republican nominee had won after 1948 except for the LBJ landslide. The only former Confederate state that went for Ford in 1976 and a state that was going to go for Bill Clinton (apparently a good match for Clinton in culture) in 1992 or 1996, it was definitely not going to go to the most blatant d@mnyankee Democrat since George McGovern in a close election.

I changed my tune on Virginia once just about every poll showed Obama leading in Virginia.

Heck, as the 1992 election came nigh, I was extremely certain that George Herbert Walker Bush would be re-elected because Bill Clinton wasn't going to win Texas, a state that Democrats had not lost in a winning election -- ever. In 2000 I expected Al Gore to win West Virginia because West Virginia was historically the sort of state that Democrats win in elections other than 45-state blowouts for Republicans.

States trend. So if a poll of Minnesota shows Donald Trump winning Minnesota by 5% my first impression may be "Whiskey Tango Foxtrot".  But let me see other polls showing much the same, and I might recognize either a collapse of Hillary Clinton or a Republican trend in Minnesota.        

3. In the case of the infamous poll of Kansas showing a lead for Hillary Clinton, even I said that I didn't really believe it, but nothing but common sense says that this result is absurd.I dumped it as soon as a poll that made sense appeared.

Don't like a poll? Then wait for the next one.

4. Most of us can rationalize anything.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Fmr President & Senator Polnut on July 16, 2016, 11:18:06 pm
Why is it an Atlas tradition to try to "unskew" every poll that has a Republican doing relatively well, but to mock anyone who questions a poll that has Democrats doing relatively well?  Of course there will always be outliers on both sides, but sometimes outliers can be the truth, especially in this day of "herding".

It's called the context. Unskewing came from assumptions the GOP had about the electorate, this is NOT unskewing, it's psephology. If the basic elements of the poll actually make no sense (as in based on history, not assuming there's a default electorate), then it's appropriate to question it. Considering dodgy models and methodologies are the elements helping Trump and the GOP right now, best not to get too snippy.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: StatesPoll on July 16, 2016, 11:41:30 pm
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

54% men and its tied!?

Safe D. Virginia is solid blue.


Red Avatar snobbies can't even do simple math?

TRUMP vs Hillary  PPP Poll June 13-15,2016(Virginia)  1032 RV

TRUMP got 45%. and 40% of Women, 49% of Men. Just 9% difference.
(in virginia less gender gaps than other states)

http://www.publicpolicypolling.com/pdf/2015/PPP_Release_VA_61616.pdf (http://www.publicpolicypolling.com/pdf/2015/PPP_Release_VA_61616.pdf)

so even convert the Hampton University Poll as realistic gender ratio. it doesnt make that much diffrence.

if TRUMP got 43% of Men, 35% of Women at this Time.

54%(men) x 43% +  46%(women) x35% = 39.32%

convert the TRUMP's numbers as realistic gender ratio

53%(women) x 35% + 47%(men) x 43% = 38.76%

just 0.58% reduced. bah :p.  let's plus 0.58% for Hillary. Then

Original Numbers(Poll): TRUMP 39.2% > Hillary 38.8%   Spreads: TRUMP +0.4%
Fixed numbers as realistic gender ratio % : Hillary 39.38% > 38.62%  Spreads: Hillary +0.76%

soooooooooooo big diffrence. and definately 'safe D' with +0.76%ish leads.

Red avatar snobbies are dumb. they can't even do simple math.
7% overweighted men doesn't mean TRUMP gets 7% more than real.





Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Fmr President & Senator Polnut on July 16, 2016, 11:42:44 pm
Is grammar and punctuation that big of a challenge for you too?


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Landslide Andy on July 16, 2016, 11:42:56 pm
Ugh, he's back. I was hoping he was banned.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Seriously? on July 16, 2016, 11:57:25 pm
Honestly, looking at the crosstabs, they weren't even trying with this poll. Equal party ID and 54% men. That's right. 2012 exit polls had it at 53% women to 47% men, but the great pollsters at Hampton University have discovered that 14% of Virginia's women have changed sexes in the last 4 years.

While unlikely, it's not out of question:

If males are more motivated to vote this year than women, it could produce an electorate that is overall more male than female.

For example if 51% of the population is female and they only have 55% turnout, while males (49%) have 65% turnout, you could get the electorate that the poll shows.

But yeah, that's still unlikely.

With Trump and Pence, women will be motivated to vote. Especially black and latina women. I actually think Colorado and Virginia are lost causes for Trump. He needs Ohio, Pennsylvania, Michigan, Iowa, etc...I don't think he gets it, but that's where he should be focusing.
I remain amazed at how some on here think that Virginia has so radically shifted that it is no longer the same very, very slightly leaning Republican bellwether state that it was in the past three or four elections.

Even if it slightly skews Democrat this cycle, it's not skewing by 7%, no matter how much red and faux blue avatars hate Trump.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: dspNY on July 17, 2016, 10:23:45 am
54% men and it's tied?

VA's electorate is 53% female so it's probably Clinton +5


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: RINO Tom on July 17, 2016, 04:27:13 pm
LOL, a federal ban of SEMI-automatic rifles has more support than a federal ban on FULLY automatic rifles...


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: StatesPoll on July 23, 2016, 01:12:58 am
Ugh, he's back. I was hoping he was banned.

even Liberal Nate Silver Adjusted Hampton Univ VA poll as TRUMP +1%

how r u feel? ;)

http://projects.fivethirtyeight.com/2016-election-forecast/virginia/ (http://projects.fivethirtyeight.com/2016-election-forecast/virginia/)


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: StatesPoll on July 23, 2016, 01:17:43 am
54% men and it's tied?

VA's electorate is 53% female so it's probably Clinton +5

7% overweighted men doesn't mean TRUMP gets 5.4% more than real.

and even Liberal Nate Silver adjusted Hampton univ VA Poll as TRUMP +1% ;)

http://projects.fivethirtyeight.com/2016-election-forecast/virginia/ (http://projects.fivethirtyeight.com/2016-election-forecast/virginia/)


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: President Johnson on July 23, 2016, 07:41:40 am
Now let's see how the Kaine selections impacts in the next poll.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: heatcharger on July 23, 2016, 09:45:02 am
Lol at StatesPoll and muh red avatars.

But seriously (lol), why do you spend so much time unskewing? Polls have a margin of error for a reason.


Title: Re: VA: Hampton University: Clinton and Trump tied
Post by: Xing on July 23, 2016, 11:03:49 am
I'm glad StatesPoll cares about us red avatars so much, since he spends so much time thinking about us. :)