Another Presidential Poll Late June 2016

[buckshot24]

No, go learn it yourself.

Going by raw support and not spread there are only 3 non-outliers where both values are within range.

46-41
44-39
47-42

This poll was within range for Trump but not Hillary. The Fox poll is in range for Hillary but not Trump.

 

[buckshot24]

For comparison purposes, there were no outliers for either candidate in the final 9 polls in RCP in 2012.

Went back to similar time period and took 8 polls from 2012

There was only 1 outlier for either candidate.

 

[Cerpin Taxt]

Nate Silver gave Trump a 20% chance of winning the election vs Clintons 79% chance.
This was on today's Good Morning America.

http://fivethirtyeight.com/features/donald-trump-has-a-20-percent-chance-of-becoming-president/

...but what does Nate Silver know about forecasting presidential elections?

 

[fskimospy]

Going by raw support and not spread there are only 3 non-outliers where both values are within range.

46-41
44-39
47-42

This poll was within range for Trump but not Hillary. The Fox poll is in range for Hillary but not Trump.

You do not understand how polls work. You embarrassed yourself four years ago by trying to act like you did. Don't make the same mistake.

 

[BxgJ]

No, go learn it yourself.

Well even when it's been shown to him it hasn't made a difference.

 

[buckshot24]

You do not understand how polls work. You embarrassed yourself four years ago by trying to act like you did. Don't make the same mistake.

Do the math then. How do you determine an outlier poll?

 

[buckshot24]

http://fivethirtyeight.com/features/donald-trump-has-a-20-percent-chance-of-becoming-president/

...but what does Nate Silver know about forecasting presidential elections?

I wonder if his prediction engine takes into account indictments.

 

[Perknose]

Do the math then. How do you determine an outlier poll?

Dude, it's right there in the very term "outlier." It's a poll whose results are significantly different than the results of the other polls.

You don't even need math to grasp this, except to know that, if the other polls show one candidate with, say, a range of 38-42% support but one poll has that candidate at 51%, well . . . spot the outlier!

 

[buckshot24]

Dude, it's right there in the very term "outlier." It's a poll whose results are significantly different than the results of the other polls.

You don't even need math to grasp this, except to know that, if the other polls show one candidate with, say, a range of 38-42% support but one poll has that candidate at 51%, well . . . spot the outlier!

You need to be more rigorous than that. What does "significantly different" mean? When it gets into grey areas, how do you tell the difference between an outlier and something within expected values?

I ran basic stat formulas on the numbers and I came up with 5 out of 8 polls having outliers in them. None in the 2012 final 9 polls. This poll in the OP has an outlier in it and that is with Hillary's support.

 

[nickqt]

I wonder if his prediction engine takes into account indictments.

Don't forget to factor in Benghazi!

I predict a Rmoneyslide!

 

[buckshot24]

Don't forget to factor in Benghazi!

I predict a Rmoneyslide!

Do you think Benghazi is like the FBI investigation in any meaningful way?

 

[nickqt]

Do you think Benghazi is like the FBI investigation in any meaningful way?

As a means to prevent a Democrat from usurping the White House?

Quick question: are Rice and Powell going to be investigated and "indicted" for something, too?

 

[buckshot24]

As a means to prevent a Democrat from usurping the White House?

In any meaningful way.

Quick question: are Rice and Powell going to be investigated and "indicted" for something, too?

Neither of them did what Hillary did. You should look further than www.hillaryclinton.com for your information.

 

[fskimospy]

Do the math then. How do you determine an outlier poll?

Aren't you ignoring me?

 

[buckshot24]

Aren't you ignoring me?

Mostly, yes. Now how do you determine if this poll is an outlier?

 

[umbrella39]

Now another cherry picked poll, just in, DERP DERP:

Fox News Poll: Clinton up by 6 points, 89 percent say 'hot-headed' describes Trump -

http://www.foxnews.com/politics/201...9-percent-say-hot-headed-describes-trump.html

In the electoral college where it matters, she and the dems are rolling...

 

[sportage]

It's way too early for polls.
Most people don't even realize there is an election looming until November 1st.
Maybe November 2nd.
Not to mention all the democrats that realize it on November 9th.
( I kid.... no, not really)

Trump will get a boost out of his convention. Then a week later, Hillary will get her boost from her convention.
And the debates will have a deciding factor, if they even have debates.
I still predict Donald Trump will NEVER debate Hillary face to face.
Donald will have some lame excuse not to attend.
Or a bad hair day.

We won't know who actually wins until November 8th.
And never have the polls meant so little as with this time.
Anything can create a huge swing in the polls. An terrorist attack somewhere in the world, more email scandal news, Trump insulting yet another minority class, anything can happen.

If there is a major terrorist attack somewhere in the world on November 7th, Trump could win. If not, then Hillary.
How in the hell can any poll predict THAT?

I liked the Canadian response.
After Obama gave a speech to the Canadian parliament, they all stood and cheered for Obama "4 MORE YEARS".
So whoever does win, whatever does happen, I thank god at least one country in the northern hemisphere still acts like an adult and maintains their sanity.
And they don't have a fence. (hint hint)

 

[fskimospy]

Mostly, yes. Now how do you determine if this poll is an outlier?

You used to say that your failure to ignore me was a failure on your part. What changed?

 

[buckshot24]

You used to say that your failure to ignore me was a failure on your part. What changed?

It is my failure. How do you determine what an outlier is in polls? Do you have any idea or no?

 

[buckshot24]

Now another cherry picked poll, just in, DERP DERP:

Fox News Poll: Clinton up by 6 points, 89 percent say 'hot-headed' describes Trump -

http://www.foxnews.com/politics/201...9-percent-say-hot-headed-describes-trump.html

In the electoral college where it matters, she and the dems are rolling...

According to my calculations (and they could totally be wrong) this polls was an outlier in Trump support (too low).

 

[fskimospy]

It is my failure. How do you determine what an outlier is in polls? Do you have any idea or no?

Math. Go look it up yourself.

I am looking forward to making fun of you about this.

 

[sportage]

One thing I've noticed.
This....



Feel free to share

 

[emperus]

Obama gives a good rant on Trump and his supposed "populism". It's a good listen.
http://www.cbsnews.com/videos/obama-trump-is-no-populist

 

[Perknose]

You don't even need math to grasp this, except to know that, if the other polls show one candidate with, say, a range of 38-42% support but one poll has that candidate at 51%, well . . . spot the outlier!

You need to be more rigorous than that. What does "significantly different" mean?

So . . . I take it you couldn't "spot the outlier?" D:

Dude, at best, you're overthinking this. I know you believe you're making a rigorous statistical point, but to the rest of us you're just being willfully obtuse.

I see you here as the guy at 1:14. :biggrin:

 

[buckshot24]

So . . . I take it you couldn't "spot the outlier?" D:

No I can't. I don't simply assume different data is an outlier to be ignored and discarded.

Dude, at best, you're overthinking this. I know you believe you're making a rigorous statistical point, but to the rest of us you're just being willfully obtuse.

I wish my analytical chemistry prof was as loose with outliers as you guys seem to be. Just throwing out the "its an outlier" just bothers me, maybe my non 4.0 in that lab has scarred me beyond repair. It being different wasn't a good enough reason to discard the data point and my calculations were off because of it.