Reuters/Ipsos: (Most) States

[Ebsy]

How does Reuters consistently manage to produce ridiculous results across the board?

your hero 'Nate Silver(liberal)' rated it as A- Pollster



 

I think Nate Silver is rather stupid and said so multiple times this past year, but you were probably too busy unskewing polls with bogus Ohio voter registration numbers to be aware of that.

[ApatheticAustrian]

maybe they are A pollsters but even those don't seem to get "multi-polls" straight.

while i...as a liberal hack...have accepted the whole rust belt to be competetive this cycle there are just too many strange numbers to believe anything at all.

CO may be a sleeper trump state if poll after poll finds this result but at the same time PA won't be to the right of OH and NM absolutely won't have the same numbers as CO.

and don't even get me started about trump clearing maine but losing NC like that.....i guess some are right, some are wrong but since i can't say which numbers are which ......i guess those polls are only useable for aggregators like 538 and not simple folk like us.

[Mallow]

New numbers:

Pennsylvania: 46% Clinton, 44% Trump
Ohio: 47% Clinton, 44% Trump
Florida: 50% Trump, 46% Clinton
Colorado: 43% Trump, 40% Clinton
New Mexico: 43% Trump, 40% Clinton
Nevada: 41% Trump, 38% Clinton
Michigan: 44% Trump, 44% Clinton
Wisconsin: 43% Clinton, 40% Trump
Maine: 41% Clinton, 40% Trump
North Carolina: 46% Clinton, 44% Trump
Iowa: 49% Trump, 41% Clinton

Apparently, Trump is leading in Vermont. lol

Trump: 243 EV
Clinton: 242 EV

Did they not poll Latinos?

This is pretty much exactly what I thought as well. It seems like these polls are actually very consistent, if you HEAVILY undersample the Latino vote. That would account for NM (the most Hispanic state) being where it is, as well as NV, CO, and even FL.

[Seriously?]

Reading this correctly, it's Battleground Michigan for all the marbles.

[Ozymandias]

This is pretty much exactly what I thought as well. It seems like these polls are actually very consistent, if you HEAVILY undersample the Latino vote. That would account for NM (the most Hispanic state) being where it is, as well as NV, CO, and even FL.

Would also explain Trump being up 7 in Arizona and 22 in Texas...

[‼realJohnEwards‼]

New numbers:

Pennsylvania: 46% Clinton, 44% Trump
Ohio: 47% Clinton, 44% Trump
Florida: 50% Trump, 46% Clinton
Colorado: 43% Trump, 40% Clinton
New Mexico: 43% Trump, 40% Clinton
Nevada: 41% Trump, 38% Clinton
Michigan: 44% Trump, 44% Clinton
Wisconsin: 43% Clinton, 40% Trump
Maine: 41% Clinton, 40% Trump
North Carolina: 46% Clinton, 44% Trump
Iowa: 49% Trump, 41% Clinton

Apparently, Trump is leading in Vermont. lol

Trump: 243 EV
Clinton: 242 EV

Did they not poll Latinos?

This is pretty much exactly what I thought as well. It seems like these polls are actually very consistent, if you HEAVILY undersample the Latino vote. That would account for NM (the most Hispanic state) being where it is, as well as NV, CO, and even FL.

Yeah, they estimate Latino turnout at 30%, which is just absurd. Ipsos trying to unskew themselves, lol.

[Seriously?]

Wow@that CA margin.

I'd say that's very very bad for Clinton and the Democrats. Trump can run up the score elsewhere.

[Dr. Arch]

New numbers:

Pennsylvania: 46% Clinton, 44% Trump
Ohio: 47% Clinton, 44% Trump
Florida: 50% Trump, 46% Clinton
Colorado: 43% Trump, 40% Clinton
New Mexico: 43% Trump, 40% Clinton
Nevada: 41% Trump, 38% Clinton
Michigan: 44% Trump, 44% Clinton
Wisconsin: 43% Clinton, 40% Trump
Maine: 41% Clinton, 40% Trump
North Carolina: 46% Clinton, 44% Trump
Iowa: 49% Trump, 41% Clinton

Apparently, Trump is leading in Vermont. lol

Trump: 243 EV
Clinton: 242 EV

Did they not poll Latinos?

This is pretty much exactly what I thought as well. It seems like these polls are actually very consistent, if you HEAVILY undersample the Latino vote. That would account for NM (the most Hispanic state) being where it is, as well as NV, CO, and even FL.

[Skye]

This also has WV as Trump +10, for what it's worth.

[mencken]

This also has WV as Trump +10, for what it's worth.

Probably best to treat with significant skepticism subsamples with lower than traditional sample sizes (I wish people would apply this logic to demographic crosstabs...)

[afleitch]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

[Erich Maria Remarque]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

Yes, and your point is?

[afleitch]

For the record, this is their turnout projection....



So yeah. If we adjust it to the 2000-2012 average, then it's 2012 less Vermont (ugh), Iowa, Florida, Nevada, New Mexico and Colorado.

In the PV it's 46-40 to Clinton.

If you reduce turnout to 0 for everyone and up to 100% for Hispanics, then Reuters have insufficient data for all states bar California, Texas, Colorado, Mississippi, New Jersey and Massachusetts.

[afleitch]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

Yes, and your point is?

The point is child, that in the past two weeks out of 4,899 respondents, only 162 respondents are Hispanic. That's 3%. That sample isn't accurate enough to give you an accurate national poll, never mind states ones.

[Erich Maria Remarque]

For the record, this is their turnout projection....



So yeah. If we adjust it to the 2000-2012 average, then it's 2012 less Vermont (ugh), Iowa, Florida, Nevada, New Mexico and Colorado.

In the PV it's 46-40 to Clinton.

Unskewers detected...

If you reduce turnout to 0 for everyone and up to 100% for Hispanics, then Reuters have insufficient data for all states bar California, Texas, Colorado, Mississippi, New Jersey and Massachusetts.

Yes, so? They have low sample. Is it so difficult to understand, Red Hacks?

For instance in NM, that you referenced to, it was totally~120. Is it so strange?

[Erich Maria Remarque]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

Yes, and your point is?

The point is child, that in the past two weeks out of 4,899 respondents, only 162 respondents are Hispanic. That's 3%. That sample isn't accurate enough to give you an accurate national poll, never mind states ones.

Yes, Hispanics are undersampled in most polls (not just IPSOS, not just this year). So are those under 34 years-old (not just IPSOS, not just this year).

BREAKING NEWS, lol

[afleitch]

How is pointing out that a sample size is too low 'unskewing'? i said nothing about what the sample says, what the results are, just simply that is too low. Would I be 'unskewing' a poll if 25% of the respondents were women and i thought it should be closer to 50%

[Erich Maria Remarque]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

Yes, and your point is?

The point is child, that in the past two weeks out of 4,899 respondents, only 162 respondents are Hispanic. That's 3%. That sample isn't accurate enough to give you an accurate national poll, never mind states ones.

By the way, where did you find those numbers? Ipsos doesn't make it simple to find it

[Erich Maria Remarque]

How is pointing out that a sample size is too low 'unskewing'? i said nothing about what the sample says, what the results are, just simply that is too low. Would I be 'unskewing' a poll if 25% of the respondents were women and i thought it should be closer to 50%

You must be logged in to read this quote.

[afleitch]

How is pointing out that a sample size is too low 'unskewing'? i said nothing about what the sample says, what the results are, just simply that is too low. Would I be 'unskewing' a poll if 25% of the respondents were women and i thought it should be closer to 50%

You must be logged in to read this quote.

That's not unskewing the sample. Again, I have no issue with the numbers they are getting. It's instead looking at how they are applying a LV model.

[Fmr President & Senator Polnut]

Affie - don't bother. Any poll, no matter how dodgy, needs to be accepted and ANY attempts to try to make sense of them is unskewing.

[Erich Maria Remarque]

Lol, neither you or I were talking about the sample.

You tried to "adjust" their turnout model. How is that not an unskewing?

It reminds me of this

[JerryArkansas]

Lol, neither you or I were talking about the sample.

You tried to "adjust" their turnout model. How is that not an unskewing?

It reminds me of this

Again, why cite Nate Silver?  He is pretty bad.

[Erich Maria Remarque]

Affie - don't bother. Any poll, no matter how dodgy, needs to be accepted and ANY attempts to try to make sense of them is unskewing.

Lol, Mr. Unskever.

You still has not answered me, what change in CBS methodology (compared to 2012/08 etc) you doesn't like

You must be logged in to read this quote.

So you were complaining about CBS/NYT methodology back in 2012? If not, what did they change in their methodology, that you don't like?

So tell me, what oddities in their method you have found, but are so unwilling to reveal for us?

Or you can admit, that you don't like the results

[Seriously?]

Go to the turnout model and reduce everyone to 0 but Hispanics to 100. I can't do that right now but last time ir showed that there was not enough in the sample to measure for states like NM

Yes, and your point is?

The point is child, that in the past two weeks out of 4,899 respondents, only 162 respondents are Hispanic. That's 3%. That sample isn't accurate enough to give you an accurate national poll, never mind states ones.

Not if you reweigh those Hispanics to the equivalent of your suggested turnout model, which it appears that Reuters has done. So Reuters in essence takes the 162 Hispanics and extrapolates them to the real number which is likely about 400 or so.

Whether they did it on a micro level for each state is the only real question that you have when attempting to reskew these polls.

These numbers fundamentally are fine and in line with most state polls within the margins. Of the four companies doing these 50-state polls (Morning Consult, Google, Survey Monkey and Reuters), the way Reuters is doing it makes the most sense to give you the most up-to-date polls possible.

It's the states with 6 or fewer EVs where you start to run into issues methodologically.