Donald Trump’s Strongest Supporters: A Certain Kind of Registered Democrat...43% of them...

a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

originally posted by: matafuchs
43% of registered democrats 'polled' say they would vote for him.

No, you've misinterpreted yet another poll.

It's 43% of those voters who are registered Democrat but actually lean or vote Republican, a small subset of registered Democrats.

Not "43% of registered democrats" in general.

originally posted by: MystikMushroom
Did anyone go to the article? Did anyone catch this?

Hm...so on the left we have Republicans that support Trump, the darker the more support.

On the right we have areas where racially charged internet searches are most common, the darker the more common.

Anyone seeing any kind of pattern here?

What happens if you overlay population on those maps? i.e is the correlation with racially charged internet posts just to do with the number of people living in those areas?

Below is a map of population density in the US.

From the article it does appear that the right hand chart is based on VOLUME - so it looks like the NY Times journalist is either stupid or willfully misleading people (surely that could not be the case, could it? )



Be careful of partial correlations;
It's true to say that the more firemen you have called to a fire, the more deaths there are. Very close correlation, however the reason is that more firemen were called because the fire was bigger.

originally posted by: Phage
a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

You're asking Trump supporters to do math - no fair!

And as far as all these polls go another interesting analysis from Nate Silver's FiveThirtyEight:

What’s amazing is that polls at this point in the campaign are no more predictive than a January-to-June polling average from the year before the primary. The predictive margin of error is the same, as is the correlation between the national polls and the result.

fivethirtyeight.com...

originally posted by: Phage
a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

I don't know where you come from, but around these parts there is 148% in the whole.

originally posted by: Aazadan

originally posted by: Phage
a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

I don't know where you come from, but around these parts there is 148% in the whole.

Oh. Ok then.

originally posted by: Aazadan
a reply to: matafuchs

Furthermore, it's true that, Trump is popular but if you read the article you linked you'll see that his popularity is not the type that wins a national election. Pretty much the entire article explains that he's polling best with a specific group (poorly educated racists)

I believe that to be incorrect.

originally posted by: Phage
a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

Why did you selectively omit the asterisk of that data?

a reply to: Vector99
Because it doesn't explain how a total of 148% can be reached. If anything it would indicate a total of less than 100%.

But I thought you put no validity on the poll anyway.

originally posted by: Phage
a reply to: Vector99
Because it doesn't explain how a total of 148% can be reached. If anything it would indicate a total of less than 100%.

But I thought you put no validity on the poll anyway.

I don't, and you said you would prefer the raw data yourself, but you're running with it otherwise.

*Only for the 31 states where voters can choose to register with a party.

The asterisk.

a reply to: Vector99

I don't, and you said you would prefer the raw data yourself, but you're running with it otherwise.

On the contrary, I am questioning the analysis provided.


Yes. I know what the footnote was.
Can you explain how that can result in a total of 148%

originally posted by: Blackmarketeer

originally posted by: Phage
a reply to: matafuchs

It does not include people who are undecided either.

How do you know this?

I just noticed something though. For the 33% who favor Trump:
Unregistered: 40%
Republican: 29%
Independant: 36%
Democrat: 43%
See anything odd?

You're asking Trump supporters to do math - no fair!

When asked if they supported Trump this is how they responded....

40% of those unregistered said yes
29% of Republicans said yes...
etc
etc
etc

And you guys are telling me I can't add or read a poll?

a reply to: Phage

Yea, incomplete data

a reply to: matafuchs
No, you have it wrong...again.

Mr. Trump’s share of the vote among self-identified Republicans.

Any idea what 148% of 33% adds up to?

a reply to: Vector99
That is not an explanation of how a total of 148% can be arrived at.
I'm really curious about this, there could well be an explanation but that is not it.

originally posted by: Phage
a reply to: Vector99
That is not an explanation of how a total of 148% can be arrived at.
I'm really curious about this, there could well be an explanation but that is not it.

Well when you read the source

Mr. Trump’s share of the vote among self-identified Republicans

and they say

Donald Trump fares best among registered Democrats and those with little chance of voting in 2016.

and base it on

The Civis estimates are based on interviews with more than 11,000 Republican-leaning respondents since August

and source another chart with

Civis Analytics estimates for candidate support among Republican general election voters.

and they draw data by webcrawling here

Well, I am left still searching for real raw data.

a reply to: Phage



Mr Trumps share among self identified Republicans. Ok. If you read the whole article these are the categories of people they asked. Unregistered, Republican, Independent and Democrat.

This would mean there are 4 demos questioned. Each one would be 100%. So, if anything, he garnered 37% of the vote from these folks which is where he is polling in many states. 400 divided by 148 is 37%