2801

No, I think the confusion here is the sad poll denying going on.

2802

No, that doesn’t track at all. You’re drawing from a different demographic make-up, different employment trends, different economic factors, etc. Not to mention it’s a much smaller population in absolute terms anyway.

Party ID in polling is designed to fluctuate depending on voter attitudes. It’s not a demographic. It’s whatever respondents to a poll say it is, assuming you have a sufficiently representative sample.

2803

Just on a fairness note, you didn’t seem to take Jason to task when he did exactly what Driveled did, using national PID stats to justify the skewed sample in Ohio. In fact, Driveled’s post was in response to Jason’s.

2804

While Jason would not unfairly describe this as “conservative agitprop”, I would counter that its a partisan description of the issues with Quinn samples and results: CBS/NYT/Quinnipiac Survey Narnia, Find Obama Up

[quote="Battleground Watch

[spoiler]
Florida: The Lion

Obama leads +1 at 48 to 47 with 3% Undecided; Romney leads with Independents by 5
Party ID wasD +7 (Dem 37, Rep 30, Ind 29). In 2008 it was D +3 (Dem 37, Rep 34, Ind 29). In 2004 it was R+4 (Dem 37, Rep 41, Ind 23). Good show Quinnipiac! In a state with a GOP governor and massive increases in congressional delegations, popular GOP Senator, and strong state house swings to the GOP since 2008, you found Democrat strength equal to 2008 while Republican flight since 2004 continues unabated. You found the Democrat identification advantage in your survey more than doubles the advantage they enjoyed in 2008 despite a nearly net 300,000 swing towards Republicans in voter registrations. Your Florida poll is unassailable…at least in Narnia.
Obama job approval +1 at 49/48 … if Quinnipiac surveyed only Dade County and even there I’d double check the numbers.

Ohio: The Witch

Obama leads +5, 50 – 45 with 4% Undecided; Romney leads with Independents by 6
The party ID was D +8 (Dem 37, Rep 29, Ind 30). This compares to D +8 in 2008 (Dem 39, Rep 31, Ind 30) and R +5 in 2004 (Dem 35, Rep 40, Ind 25). There is no chance the Democrat turnout advantage will meet Obama’s 2008 best in a generation turnout which we write as D +8 based on the CNN party ID generally used. This is even though the actual 2008 party ID was really only D +5 making this D +8 that much more implausible. How many statistics on changes in enthusiasm favoring Republicans, unrealistic Democrat demographic assumptions and elimination of Obama’s early vote advantage do you need to see before they start polling an electorate dissimilar to 2008 when their dream candidate fulfilled their liberal inner guilt and healed a nation or whatever BS they were peddling at the time? Quinnipiac is not going to let silly facts get in the way of its mission to buck up the Lefties and turn this contest into a horse-race. One more piece to the puzzle before the Death Star is complete.
Obama job approval +3 at 50/47 — Can you imagine what it would be if they surveyed Ohio?

Virginia: The Wardrobe

Obama leads by 2, 49 – 47 with 3% Undecided; Romney leads with Independents by 21
The party ID is D +8 (Dem 35, Rep 27, Ind 35). This compares to 2008 of D +6 (Dem 39, Rep 33, Ind 27) and 2004 of R +4 (Dem 35, Rep 39, Ind 26). Who knew the blue wave continues so far South of the DC Beltway? Certainly not Virginia and certainly not Governor Creigh Deeds. Just because Virginia flipped its state delegation dramatically in favor of Republicans doesn’t mean the voters turned their back on Democrats, it’s just there must have been a good TV rerun of Martin Sheen spouting non-sensical liberal tripe on the Left Wing that distracted Democrats from voting. Good thing Quinnipiac found these ultra-micro-targeted hidden Democrats only Project Narwhal knows about because otherwise, without those gnomes (Step 1: Call random #s only in Fairfax County, Step 2: ???, Step 3: Obama wins!) I’m not sure we’d have a fully operational Death Star. Come November 6, we’ll see how well those gnomes delivered for this survey of a fantasy electorate.
Obama job approval flat at 49/49 — Really? 49% with a D +8 turnout in a state closer to even D/R? Suuuuuuuure.
[/quote]
][/spoiler]

2805

On prediction markets, Nate is an easy target because he’s at NYT but he’s hardly alone.

In OH at least party ID is largely meaningless. In that state anyone can e.g. vote in the primary but you need to register with that party (going from memory there, but I’m fairly certain that’s the case).

Here’s a Twitter exchange with a professor who tracks early voting:

harveer sandhu ‏@harveer_sandhu
@ElectProject http://www.politico.com/news/stories/1012/82948.html …: is it true, seems like spin “220,000 fewer Democrats have voted early in Ohio compared with 2008”
View summary
27 Oct Michael McDonald ‏@ElectProject
@harveer_sandhu pure spin. OH does not have party reg. see last weeks blog post. Reporters now know so spin discredits spinners.

Lastly the election is not decided. Voter turnout is still going to determine this.

In Fl it’s close because of data from early voting results, not polls. Going into election day in 08 Obama was down 3. He’s currently leading.

2806

Fair enough. I think it’s probably because I often have trouble parsing Jason. Just a lowly state school graduate.

Anyway, it’s not like like we have long to wait to find out. Five days and a wake-up. It’s just that all this poll denial is sounding eerily familiar. Democrats had to go through it in 2004. I wasn’t posting here then, but I was posting other places and I’m pretty sure I was saying that there was just something off about that Ohio poll average showing Bush up by 2%…it couldn’t possibly be right. Sadly, Bush won Ohio by almost exactly 2%.

2807

I think the same could be said for a few around here. Most polls are really close and within the margin of error.

Does anyone dispute that Romney is winning indys? http://www.scribd.com/doc/111619086/Quinn-10-31-FL-OH-VA?secret_password=2l8php1ck5l2c2wr8j6x

I just don’t see how he loses if he’s +5-10 indy and turnouts are somewhere between 04-08. Barring a really strong D turnout or really weak R turnout, Romney will be the next president.

2808

Thanks for comity. I certainly agree re:five days and a wake-up. If the turnout models were closer to realistic, I would hate it and hope they were wrong but wouldn’t take any comfort in it. In Ohio, D+4 Obama up by 5 and I start hoping for an elusive turnout win and looking to other paths to 270. D+9, more than doubling 2008’s margin? Without clear and obvious reasoning, that to me looks more like push-polling.

2809

I just think it’s funny that everyone was posting about polls for a while…then after the first debate, suddenly no one posted anything about polls for a while. And now, suddenly people want to talk about polls again. I guess we’ll know in a few days.

2810

…anyway, today was probably the best day the Obama campaign has had in a very long time.

On the few national trackers that are going, he’s pulled even (ABC/WashPo and Ipsos/Reuters). He had some great battleground numbers, including Ohio and a late poll in Virginia from PPP that has him +3 there and shows very strong internals. Today the myth of Michigan in play got destroyed by two separate polls that showed it is not, and a Marquette University poll showed that Wisconsin is likely also firmly Obama (+8 there.)

Perhaps more importantly, though, was Obama dominating this news cycle because of Hurricane Sandy. All 3 major nets and CNN played it leading on newscasts through the night, and Obama side-by-side with Chris Christie and comforting victims could play big with very late breakers.

Tomorrow brings another day of polling, including Gallup resuming their tracking calls. Obama may have seized the initiative at the right time here, but we’ll know more tomorrow. My hunch is that this will be close, but a win for the incumbent at this point.

2811

Let’s take Ohio from the poll results. Romney is winning independents by 5 points. Independents make up about 30% of the total PID. So, that’s…is that even a full point in the top line result? I’m not even sure how to do that math, but it’s not a whole lot. And turn-out is the X factor. We can’t really predict it, only estimate it.

2812

Isn’t the last week/10 days when polls start to actually reflect the final result?

2813

No candidate has won while losing indys by a significant margin. Bush tied indys in 04 but had record R turnout.

2814
2815

Bingo.

“Independent” in past elections has meant someone in the middle, sharing both ideologies of liberals and conservatives. The ranks of “independents” swelled in 2010 with conservatives for whom the GOP wasn’t conservative enough, and using it as an identifier of apples to apples demographics with 2000, 2004, or 2008 is going to paint a picture that just isn’t grounded well in reality.

2816

Define ‘significant’.

And once you’re digging around the internals looking for reasons that the top-line numbers are wrong, you might be entering “denial” territory. The fact that Obama is 5 points down with independents is baked into the top-line. Right?

2817

Maybe I am digging but those indy numbers are certainly real. The republican exodus is a valid point but D numbers are also down and D/R’s were even split in 2010.

2818

Viewing this from the outside, it seems pretty much exactly like a mirror of 2004 to me.

Back then - as I recall it - Democrats were also busy cherry-picking and nitpicking polls to show how the overall picture just couldn’t be right (polls were derived as skewed, Gallup biased, etc). It doesn’t make the comparison any less apt that back then the D’s were voting for a rich-man candidate from Massachusetts that no one seemed to really like very much, but felt compelled to because of their strong dislike of the incumbent; that the incumbent led the challenger strongly after the conventions, then lost the first debate to tighten the race, or that Ohio seemed likely to be and was the decisive state.

I would be extremely surprised, if the results are going to be any different this time (complete with outraged claims of electoral fraud).

2819

No. If Romney loses, the Republicans will spend 4 years promoting the idea that the Communist Dictator Obama stole the election through voter fraud. They’ve been lining up for this runway all year long.

2820

It is incredibly similar, and you’re not the first to note it in this thread, but astute nonetheless.

Back in 2004 I can tell you that Democrats talked all about poor polling models (“No one’s polling cell phones!”) and skews. The conventional wisdom among true believers was that John Kerry was being underpolled by 1-3 points in battleground states. Wasn’t true, of course; Kerry underperformed the polling in Ohio and Florida and was actually defeated more easily than anyone on the Democratic side thought.