Non-Gallup/Rasmussen tracking polls thread

[Verily]

Finally have the internals...

October 6-12, 825 LV

http://www.ibdeditorials.com/Polls.aspx?id=308790300872940#pollb

Small sample size per day, but there shouldn't be ridiculous movement, as this poll is highly weighted.  I think the party ID has always been 39D, 35R, 26I (2000/2004 as well).

Unless they're defining "Midwest" very strangely, I find it highly unlikely McCain is leading in the region. (Kerry won it under any reasonable definition.)

[Verily]

Ah, Battleground, you amuse me.

http://www.realclearpolitics.com/RCP_PDF/BG_101408_2-way-ballot-trender.pdf

Obama: 53 (+2)
McCain: 40 (-3)

[Verily]

What is the party ID weighting for this Zogby poll?

Weighting? Zogby doesn't need weighting.

[Verily]

Okay, Zogby must just be making numbers up. Obama leads by 16 points among independents and 4 points overall?

[Verily]

RCP (for what its worth) says -5.0

Does that include Zogby and Kos.

Yes to Zogby, no to R2000. But I bet if R2000 was more favorable to McCain and Zogby less so, they'd been amenable to switching. It's their policy with ARG to only include McCain-friendly polls, after all.

[Verily]

ABC / Washington Post tracking poll

likely voters:

Obama: 54 % (+1)

McCain: 43 % (-1)

Looks good.  It is undeniable at this point.  There can be no doubt that the race is widening.  Obama looks like he is on his way to a big landslide.

I would be looking at one poll, especially that one.  Today, surprisingly, the results have been mixed.

ABC/WaPo is a good poll, easily the best of the news organization polls. I wouldn't consider it over Rasmussen or Gallup, but probably over pretty much everything else. Of course, you're certainly right that one poll does not a trend make, but the numbers overall have suggested that the McCain "comeback" was the polling fluke, not Obama's peak numbers. Of course, I don't expect Obama to improve past his peak lead of ~8 points, and to end slightly less than that ahead on Election Day (~6 points), but it will very obvious who will win come Nov. 4 unless some unforeseen Happening happens.

[Verily]

IBD/TIPP also has McCain leading 18-25 year olds 74-22, which is just hilarious. You can't blame that on small sample sizes; that's either a disastrous likely voter model or some serious sampling problems (polling only Liberty University students, say). If that number were changed to a relatively reasonable 58-40 Obama, Obama would have a substantial lead (assuming IBD/TIPP has 18-25 year olds as about 9-10% of their sample, which was the 2004 numbers).

[Verily]

FTR, anyone wanting to calculate an MoE, it's 100/[sqrt(sample size)]. Assuming IBD/TIPP has about 9-10% of respondents as 18-25, that's around 100 respondents, for an MoE of about 10%. 74-22 McCain under such as MoE means that McCain is at at least 64% among the group--definitely not true.

In fact, 538 did a calculation to Obama being ahead by twenty points among the youth and the odds of being that far off. I think 538 might be slightly overestimating Obama's actually lead among under-25s, but he comes out to the odds of a 74-22 McCain sample being 55 billion to one. (Take an Obama lead of 9 points, Kerry's lead in the demographic, and the odds are still well into the millions to one.)

[Verily]

What type of lead among this age group are other pollsters showing?

All, that I know of, are showing at least 15-20 points.

Could this skew the top line #'s?

To some (small) degree, but I think the bigger point is that it points to a likely flaw in methodology.  That kind of result doesn't just materialize out of statistical noise.

A 70-point swing (from a 50-point McCain lead to a 20-point Obama lead) in 10% of the sample is a 7-point swing in the overall result, to an Obama lead of just over 8 points. Or, in other words, the difference between IBD/TIPP being one of Obama's worst polls overall and his best tracking poll and above his RCP average lead.

That might overstating the case a bit if you want to say that the youth vote is the 9.3% it was in 2004 (it won't be, and will probably be more like 12%) and if you want to be conservative in your estimate of Obama's lead among 18-25s (say, 12-15 points instead of 20), but the skew among youth is hurting Obama's lead by at least five, maybe as many as eight or nine points.

It's not at all irrelevant to the topline numbers even ignoring the obvious problems that must be inherent in the methodology if they're turning up such awful samples.

[Verily]

What type of lead among this age group are other pollsters showing?

All, that I know of, are showing at least 15-20 points.

Could this skew the top line #'s?

To some (small) degree, but I think the bigger point is that it points to a likely flaw in methodology.  That kind of result doesn't just materialize out of statistical noise.

A 70-point swing (from a 50-point McCain lead to a 20-point Obama lead) in 10% of the sample is a 7-point swing in the overall result, to an Obama lead of just over 8 points. Or, in other words, the difference between IBD/TIPP being one of Obama's worst polls overall and his best tracking poll and above his RCP average lead.

That might overstating the case a bit if you want to say that the youth vote is the 9.3% it was in 2004 (it won't be, and will probably be more like 12%) and if you want to be conservative in your estimate of Obama's lead among 18-25s (say, 12-15 points instead of 20), but the skew among youth is hurting Obama's lead by at least five, maybe as many as eight or nine points.

It's not at all irrelevant to the topline numbers even ignoring the obvious problems that must be inherent in the methodology if they're turning up such awful samples.

Let's examine every internal then to see where there could be flaws to add points to your candidate.

In most cases, internals do not make a big difference. Or, rather, the skew from internals is very small, say 0.2 points one way and then 0.3 points the other way and then 0.1 points back the first way from three skews in three subsamples in one poll. This poll has a single subsample which is so absurdly far off that it massively screws up the overall result as well.

But you can accuse me of being a hack all day long. It just makes you cuter.

[Verily]

IBD/TIPP
Obama 46.5%(+0.7)
McCain 43.3%(+1.4)

this poll has been all over the place.

Actually its kept it pretty stable as a close race.

It had Obama up eight at one point.

[Verily]

Nah.  That was their attempt to correct it.  Basically, they'd been getting a weirdly low number of youth voters in their polls.  They ignored it, until the sample shifted to being ridiculously McCain.  To fix this, they started applying a non-random sample to meet quotas.  Basically, he ignored a methodological flaw until it started showing up, and then applied a non-random solution.  It's shoddy pollster behavior.

What I meant was that what you quoted indicated why the numbers were screwy (in that much-highlighted chart last week) in the first place -- extremely low sample sizes.

I'm not enough of a stats guy to get into the fix side of things.

True; we were assuming that young voters were a reasonable 9-10% of the sample in our calculations when they were actually about 3%. That would increase the MoE substantially, to the point where the results within the subsample were outside the margin of error, but not radically so. But the undersampling clearly was a major problem.

[Verily]

Folks, remember what I said Friday evening.  Do not trust Halloween night polling.  Even if it is correct or shows a pro-McCain result (which it is unlikely to considering who is out).  I've been reading up a bit on Zogby's polling practices, and my sample today is partially Halloween night polling, so there.

I disagree with the assessment that Halloween would be a good night for Obama generically. A lot of adults with children will be out, but a lot of young adults will also be out (as Halloween has increasingly become a holiday for 20 and 30-somethings as well as children). Plus, Halloween means vacationing for the weekend is much less likely.

[Verily]

Considering that the McCain +1 sample falls off tomorrow, I see a Zogby surge to Obama +10 or 11 by election day.

Probably more. To be Obama +7.1 today, he must have an average of Obama +10.5 on the past two days (plus McCain +1 three days ago). But it wouldn't be Zogby without a last-minute surge. I say it will be Obama +12 tomorrow.

[Verily]

IBD/TIPP
Obama: 47.5 (+0.8)
McCain: 43.0 (-1.6)
Undecided: Obscene [9.5 (+0.8)]

[Verily]

Obama over polled on nearly every poll.

IBD/TIPP:   7.2
Gallup:     11.0
R2K:           5.0
Zogby:     11.4 of course.
Hotline:      5.0
ABC/WP:   11.0
'bots:          4.92
Average:     8.3

Nationally, it looks like 4 points.  Bradley?

What? Numbers are still coming in from the west. We won't know the real PV for awhile I would imagine.

Aye. California's only 28% in, and it's huge, plus Oregon and Washington, also barely in. Obama's lead is also up to 5 points since he posted that. (Plus, J. J. amusingly included known trash like Zogby while ignoring Rasmussen.)

In any case, the state polls seem to have understated Obama's position much more often than they overstated it. There are exceptions: Obama overpolled in Georgia and North Dakota (the only two counterexamples). But he underpolled much more severely in Pennsylvania and across the Northeast. Elsewhere, the state polls, in aggregate, seem to have been pretty much exactly on target (maybe slightly understating Obama in Florida and Ohio as well).

[Verily]

So have we figured out yet why McCain did a couple of percent better than the polls in general, apparently?  You know what?  We might not really know. Oh the horror, the horror!  By the way, Obama will be close to a 7% lead before this is all over I suspect.

Obama is now at a 6.5-point lead, and that will probably still grow a bit with Oregon, Washington, Georgia (Fulton County, a heavily Dem area), New York and Illinois still not fully counted (plus absentees and provisionals).

http://www.politico.com/electionmap2008/index.html