Reuters cooking their poll again

http://www.realclearpolitics.com/docs/2016/2016_Reuters_Tracking_-_Core_Political_9.27_.16__.pdf ^

[Viper652]

Sad

[MichelleWSC3]

I did notice at bottom of page 14 under I-13% there were 11% “None/DK” I’m assuming this is either no affiliation or don’t know what affiliation. Could be undecided.

[NYC-RepublicanCT]

Actually, you’re wrong... I just plugged it into a spreadsheet and it matches the D+8... NOT the D+12.

[notdownwidems]

-—2010 census Female pop = 50.8% Male pop = 49.2%

What about all the 'transgenders'?

[NYC-RepublicanCT]

Do the math...

Hillary - 84(.49) + 8(.34) + 20(.14) = 46.5

Trump - 7(.49) + 78(.34) + 29(.14) = 35.8

It's D+8

[rb22982]

If you take 1,041 likely voters and plug 42% at D Leaning, 34% and R leaning and the rest as independent, you get 437 Ds, 354 Rs, and 250 likely voters. If you take that matrix and multiply by the middle of page 10, you do indeed get Clinton leading by 6 (6.4% to be exact), however the %s don’t match up 44/38/19 but rather 42.6/36.4/20.8. So again, it still doesn’t match up. Not to mention page 14 clearly says 1,705 adults.

[rb22982]

46.5% and 35.8% is not what the poll results were though (44/38)

[CyberAnt]

Yep, the old “stolen election” narrative .. so what else is new ..??????

[NYC-RepublicanCT]

I give up my friend. I just proved the math. I’ve been checking this data weekly and they’ve always done it he same way. Using your 12 point gap derives a 12 point gap. They clearly lay out the breakdown that’s used on page 14. I derived the same 4 point gap using those numbers... Whatever floats your boat, I just wanted to get the accurate data out there

[NYC-RepublicanCT]

Exactly! I was using your D/R/I split to prove your numbers wrong

[NYC-RepublicanCT]

This uses your 49/34/14 split.

[rb22982]

Sigh. 49/34/14 isn't +12 but D +15. Doesn't matter. Pg 14 matrixed with page 10 voting by party ID does get you +6 *but* not the 44/38/17 they show on pg 10 but rather 42.8%, 36.4%, 20.8%. So the proportions are fairly right but the amount is wrong so basically what that means is Pg 2, 10, and 14 all use different basis.

(Doing it the way you are trying to do it would yield 46.9% vs 36.4% for a D+12.
Hillary - 84(.49) + 8(.37) + 20(.14) = 46.9
Trump - 7(.49) + 78(.37) + 29(.14) = 36.4)

[FR33DOM4ME]

This poll was referenced on thehill.com earlier this afternoon. The article referencing the poll has since been pulled off the site. I think there are legitimate concerns about its veracity.

[God luvs America]

Direct link to polling data?

[NYC-RepublicanCT]

I wrote 49/34 (type) but the numbers reflect 49/39/12 to get the 46.5 vs 35.8.

By the way, Hillary is up 4, not 6, in the 4 way... He's up 6 in the head to head, so you're comparing apples to apples.

The 42/34/24 works in both the 2 way and 4 way.

[NYC-RepublicanCT]

All the way at the top, in the original post

[luv2ndamend]

And you know what, there ain’t a thing that is going to stop them!

[rb22982]

Again - you are missing my point - it works on the delta but not on the actual reported % with the 2 way.

Two way using your methology with 42% Dems (with leaners) 34% Rs (with leaners) and the rest independent yields:
84*(0.42)+8*(0.34)+20*(0.24) = 42.8 Clinton (vs 44% reported)
=7*(0.42)+78*(0.34)+29*(0.24) = 36.4% Trump (vs 38% Reported)

on the 4 way
82*(0.42)+6*(0.34)+20*(0.24) = 41.3% Clinton (vs 42% reported)
7*(0.42)+76*(0.34)+27*(0.24) = 35.3% Trump (vs 38% reported Trump).

Not to mention this poll has 2-3% of "likely voters" self reporting they won't vote, which seems odds.

[LS]

USC poll: Trump back to 17 % among blacks.

[Billthedrill]

Trump back to 17 % among blacks.

What's the magic number on that, do you think? 20% and blow Taps for Hillary?

[LS]

No Dem has won with less than 90%