lol. I think I just realized he doesn't know there are significantly more registered Democrats than Republicans, as a whole, in the country.
Good gravy.
You DO know that the entire point of likely voter models is to account for such variance? Among registered voters, Obama is up by 2 points.
You're double-correcting.
And even so, it doesn't matter what the national numbers say. Only the state polls.
Well, except for trend. And there, Pew shows a 4-point swing among LV towards Obama.
You're a dishonest beautiful being.
He's right, though. You calling this as a possible "landslide" for Romney is absurd. That would represent an unprecedented failure of polling.
A win, perhaps. A landslide for either candidate is off the table at this point.
It doesn't look like they count weighting when figuring out who is a likely voter vs registered voter. Although I don't think they weight by expected turnout but by landline/cell phone demographics.Good gravy.
You DO know that the entire point of likely voter models is to account for such variance? Among registered voters, Obama is up by 2 points.
You're double-correcting.
I said "near landslide". I think a ten point win is considered a landslide?He's right, though. You calling this as a possible "landslide" for Romney is absurd. That would represent an unprecedented failure of polling.
A win, perhaps. A landslide for either candidate is off the table at this point.
It doesn't look like they count weighting when figuring out who is a likely voter vs registered voter. Although I don't think they weight by expected turnout but by landline/cell phone demographics.
In any case the unweighted numbers of likely voters would give the numbers I gave.
That's a lot of words to not address the math or what Gallup was finding in their poll. It also is ignoring the last two elections. Bush won the independents by +1 over Kerry and Obama won them by 8 points according to cnn exit polls.^ This just comes down to basic lack of education and/or information. It's well known Independents as a group have voted more Republican than Democrat for years and years. It's also well known that there have been more (a lot more, actually) registered Dems than Repubs for years. Elections are still close because a lot of independents were formerly registered Republicans but still prefer to self-identify and/or self register as independents. It's also well known that Republican's base turn out in larger numbers because their voters more reliably get to the polls, for years now.
It is not surprising that people who don't know anything about polls are surprised or unaware of this, and attempt to do their layman version of adjusting polls to reinforce their "common sense gut feeling", as it were. Meh.
I said "near landslide". I think a ten point win is considered a landslide?
Likely voters polled 1495Sorry, I don't know what you are talking about, honestly. I don't get what numbers you are using or what you are doing with them.
I don't see how you can take that poll and turn it into a 2.1 advantage for Romney. Maybe you could try to explain again using smaller words for me.
I was basing this on Gallup's party identification numbers where there is a +1 Republican advantage and the independent gap for Romney. No sense in quibbling about what a landslide is so I'll change the term from "near landslide" to "comfortable win".Short of one of the candidates being caught with the proverbial dead girl or live boy, neither is winning by anything approaching ten points.
The question isn't "are you a registered democrat/republican". Party registration has nothing to do with it.lol. I think I just realized he doesn't know there are significantly more registered Democrats than Republicans, as a whole, in the country.
Likely voters polled 1495
522 Republicans at 92-6 Romney
509 Democrats at 94-5 Obama
414 Independents at 48-40 Romney
Turns out to be a 2.1 lead for Romney.
You must not have done the math. My calculation has nothing to do with gallups party id numbers. Do the math I posted and you end up with a 2.1 point advantage for Romney. It also has nothing to do with registered vs likely voters. I'm not saying my calculations are right or anything like that but you've made them a bigger issue by trying to debunk them incorrectly twice now.What you're missing is that those numbers are after the likely voter screen is applied, which means they've already taken out more Romney supporters than Obama supporters.
They started with 570/560/491 D/R/I. That means they are tossing out 7% of Republicans, but 11% of Democrats. And they are still tied after that.
You can't take those numbers and then just apply party ID splits from a different polling form. It's apples and oranges -- even Gallup says so.
Furthermore, you can't assume that independents who have made up their minds will split the same way as those who have not.
And how can I take Gallup seriously when they show Obama's approval swinging 24 points in a 5 day span? Honestly.
You must not have done the math. My calculation has nothing to do with gallups party id numbers. Do the math I posted and you end up with a 2.1 point advantage for Romney. It also has nothing to do with registered vs likely voters. I'm not saying my calculations are right or anything like that but you've made them a bigger issue by trying to debunk them incorrectly twice now.
Take each groups percentages and multiply them out and add them up Romney has 2.1 points more than Obama.
I wouldn't say it doesn't mean anything it is a direct calculation before weighting and that isn't "nothing". It can help us see how much they rely on weighting instead of contacting the right people in the first place.Yes, I understand. The thing is -- what you're doing doesn't mean anything. It's the statistics equivalent of "If the queen had balls, she'd be the king".