Gallup Tracking Poll Thread [Obama vs McCain]

[Verily]

Saturday 22 June, 2008

Obama - 46% (nc)
McCain - 44% (nc)

Cue Phil ... 1 ... 2 ... 3 ... 4

Gallup and Rasmussen seem to be back to their primary game of agreeing to disagree. When one has Obama's margin wide, the other has it narrow, and vice versa.

[Verily]

Saturday 22 June, 2008

Obama - 46% (nc)
McCain - 44% (nc)

Cue Phil ... 1 ... 2 ... 3 ... 4

Gallup and Rasmussen seem to be back to their primary game of agreeing to disagree. When one has Obama's margin wide, the other has it narrow, and vice versa.

Uh,

Neither Gallup nor Rasmussen has had a "wide" margin for Obama (a wide margin would be on the order of the ludicrous Princeton Associates poll).

Both Gallup and Rasmussen have had Obama consistently in the lead by between two to six points.

The problem here is how you handle respondents who really haven't made up their minds.

If you look below the toplines you will find that the softness in McCain's support is rather high, especially when compared to Obama.

So, pushing "leaner's" at this point will result in better number for McCain than is actually the true story.

You would be incorrect. Rasmussen has Obama leading by seven points today. Rasmussen pushes leaners and finds that the margin stays the same pushed and unpushed. In any case, my definitions of "wide" and "narrow" are "outside of margin of error" and "within margin of error". In other words, more than three points is wide, less than three points is narrow. Fifteen points is absurd.

You, quite simply, are a moron. And I will not be drawn into a pointless debate in which you constantly lie and distort, as has happened to Alcon in the other tracking poll thread.

[Verily]

Gallup and Rasmussen, once again, agreeing to disagree. (Which mostly just shows that there are no trendlines to be drawn, and things are steady at about Obama+4/5.)

[Verily]

I don't see how there can be movement when nothing is happening...

I think Sam thinks people drift toward Obama when no one is paying attention. May or may not be true.

[Verily]

What's with the raging discussion between Carl and Dan anyways? I'm minded to rename the thread

Dave

Angry Rants Thread (and Gallup Tracking Poll [Obama vs McCain])

[Verily]

If this is somehow an indication of undecided voters deciding a candidate it could prove lethal for McSame.  Plus the hurricane is going to sway some viewers from watching the RNC especially Florida voters. 

Not Florida voters. Gustav isn't going anywhere near Florida. It's New Orleans and Houston that are anxious. TS Hanna could affect Florida, but, if it does, that won't be until after the convention.

[Verily]

Infomania: last few days were Obama +1, Obama +7 and Obama +4.

Is that reverse order or actual order? (Was Obama +1 yesterday or Tuesday?)

[Verily]

Straightforward math would tell us today's sample was somewhere around McCain +1 to McCain +3 given the graph posted above.

More like a tie to M+2. (7+1+0)/3 = 2.333, (7+1-1)/3 = 2, (7+1-2)/3 = 1.667. M+3 would give an Obama lead of 1. Of course, allowing for the possibility that O+7 and O+1 are rounded, it could be anywhere from O+2 to M+3.

[Verily]

Straightforward math would tell us today's sample was somewhere around McCain +1 to McCain +3 given the graph posted above.

More like a tie to M+2. (7+1+0)/3 = 2.333, (7+1-1)/3 = 2, (7+1-2)/3 = 1.667. M+3 would give an Obama lead of 1. Of course, allowing for the possibility that O+7 and O+1 are rounded, it could be anywhere from O+2 to M+3.

Ya, that's right.  Anyway, please continue to be patient... 

EDIT:  If Wed is O+7 and Thurs is O+1, then M+3 on Friday, would give you 5/3 = 1.667.  So, my number is right, so long as those numbers are right.  We should have a better feeling by Monday at the earliest, Thursday at the latest, IMHO.

Wait, you're right. My math sucks. Sorry.

[Verily]

Well, I'm assuming that's the 10 point Carter bounce Gallup is referring to. Going from 29 to 39.

You're right that it's far from the same situation as this year, just pointing out that incumbent party candidates can still get big convention bounces and lead in the polls even when their party is insanely unpopular.

The thing is, neither Carter nor the Democrats were insanely unpopular at that point.  It was just a post GOP convention bounce versus the DNC bounce.

Not as unpopular as Bush, that's true. But low enough that it should've been obvious that Reagan was likely to win.

Gallup poll Carter approval ratings

Actually less popular than Bush is now, at least in terms of who'd the voter vote for.  After the RNC, it was Carter 29, Reagan 45.

Have we seen a poll of Bush v Obama? No.

[Verily]

Well, I'm assuming that's the 10 point Carter bounce Gallup is referring to. Going from 29 to 39.

You're right that it's far from the same situation as this year, just pointing out that incumbent party candidates can still get big convention bounces and lead in the polls even when their party is insanely unpopular.

The thing is, neither Carter nor the Democrats were insanely unpopular at that point.  It was just a post GOP convention bounce versus the DNC bounce.

Not as unpopular as Bush, that's true. But low enough that it should've been obvious that Reagan was likely to win.

Gallup poll Carter approval ratings

Actually less popular than Bush is now, at least in terms of who'd the voter vote for.  After the RNC, it was Carter 29, Reagan 45.

Have we seen a poll of Bush v Obama? No.

Actually we have. Rasmussen did one in July. It had Obama 54%, Bush 34%.

http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/obama_leads_bush_by_twenty_but_clinton_does_better_against_mccain

I think Bush's negatives were higher than Carter's then.

Let me get this straight.

1. Bush has a lower approval rating than Carter.
2. Bush would lose by a wider margin to the opposition candidate than Carter.
3. Therefore, Bush is more popular than Carter.

Please, tell me how you get to 3.

[Verily]

1. Bush has a lower approval rating than Carter.
2. Bush would lose by a wider margin to the opposition candidate than Carter.
3. Therefore, Bush is more popular than Carter.

I thought Bush had higher approvals than Carter, but even higher disapprovals.

Nope. As the chart on the previous page shows, Carter was somewhat above 30% approval at this point. Bush is slightly below 30% approval. (And obviously has higher disapprovals; Carter's peak was still below 60% while Bush has recorded disapprovals as high as 70%.)

[Verily]

The most hilarious thing is that McCain has eroded Obama's lead on the economy to 3-points. Ha, more of the failed same gaining ground on the economy. There's no hope whatsoever is there?

Dave

Where are you getting this from?

Here: http://www.gallup.com/poll/110170/Economy-McCain-Gains-Ground-Obama.aspx

But it pertains to the latest Gallup/USA Today poll

Dave

Oh. So it's just useless crap.

[Verily]

McCain's in trouble now!  The trend is there!  THE TREND

McCain has not gained ground in over a week. It's clear that his campaign is finished. The Republicans might as well dump him and replace him with Mitt Romney.

[Verily]

That midweek McCain bounce is clearly in full swing.

[Verily]

The Battleground poll is slow to react because it's an eight-day tracker.

[Verily]

It doesn't come out till 1 PM, where are you getting this from?

Someone at Kos changed the image url to get the image above on the previous page.

http://media.gallup.com/poll/graphs/080920DailyUpdateGraph1_jkdelad.gif

[Verily]

Wednesday, September 24, 2008
Obama - 47% (nc)
McCain - 44% (nc)

If Rowan's numbers are correct, which is undeterminable, Obama had a very good night last night as he did with Rasmussen.

Actually no.  The last number he posted above would have dropped off yesterday.  Of course, that assumes that the numbers are, in any way, correct.

I don't see how this disagrees with what I posted... for the numbers to stay constant at 47-44, last night would have to have been equivalent to the night which dropped off, which was a very good night.

No, that night dropped off in yesterday's numbers, when Obama went down a point, not today's. The numbers that dropped off today were actually around a tie.

[Verily]

Who exactly was the last candidate to lose with a 3-7% lead 1 month before the election ?

Gore, I think.

Gore was never ahead for even a whole week, just sporadically. Bush led 95% of the time, and then Gore surged at the last minute.

[Verily]

FWIW, this is the first time all year at which point all three useful tracking polls agreed with each other exactly on the margin.

[Verily]

...And it is all 6's.

Battlground is usally decent. Just because it doesn't favor your candidate, it doesn't mean it is inaccurate.

Yes, just because it's oversampling elderly voters extremely doesn't mean it isn't bad...

The fact is, the age gap has been widening at each election. Gore won young voters by only 2 points, barely more than his national margin. So Battleground did just fine. Kerry won young voters by 9 points, so Battleground overestimated Bush by a couple of points, but nothing egregious. Obama is now leading among young voters by as much as 35 points (and on average around 20 points), so clearly Battleground is having serious difficulties. The fact that their flawed model has worked in the past does not mean it should be acceptable as a flawed model, especially when the flaws are being blatantly exposed in the current environment.

[Verily]

So, as of October 8, Obama currently leads, in tracking polls*, McCain from +1 to +11

Average lead: 6% (rounded)

* Batteground (+4), Diageo-Hotline (+1), Gallup (+11), Rasmussen (+6), Research 2000 (+10) and Zogby (+2)

Those differences are just crazy. I don't know what to conclude at the moment, but since Rasmussen sits in the middle, maybe they're on to something?

It's not hard to look at these relatively objectively. First of all, toss Zogby, it's worthless. Diageo I have had some faith in, but they've been prone to big swings, so I wouldn't pay a whole lot of attention to them. While initially I was willing to give benefit of the doubt to the DKos tracker, it's clear they've been fiddling with the partisan weights, so scrap it. Battleground has had weighting problems this whole time, so I have no faith in them even though they've fixed some of their most egregious problems.

So it's really just Gallup and Rasmussen. And 8-9 points feels reasonable right now.

[Verily]

At this point McCain would probably need to be ahead by at least 2 or 3 points in the traditional voter model to account for early voting.

that brings up a question:  don't these polls query those who have already voted early?

Good point. I had been subconsciously assuming they weren't when I made that statement, but they very well might be.

They are.  They're separated in Rasmussen, and are Obama +9.  The poll closeness comes from future likely voters, who are tied.  No reason for a poll to exclude early voters; they're automatically made LVs, I imagine.

I wonder if they skip the LV screen for those who claim to be AVs.

Of course, Rasmussen wasn't very good at screening early voters for the California primary, so who knows. I think McCain has to lead by a point or two on election day to win at this point. more in states like Virginia and Ohio which have early voting (although Ohio early voting has not seemed all that spectacular).

[Verily]

Thursday, October 30th, 2008

RV
Obama: 50% (-1)
McCain: 42% (nc)

LV (Expanded)
Obama: 51% (nc)
McCain: 44% (nc)

LV (Traditional)
Obama: 50% (+1)
McCain: 45% (-1)

[Verily]

It is interesting that the traditional and expanded LV models are now merging. Is that by accident or design, as Gallup struggles to make a prediction on the actual turnout model as we get close to the wire?

For some reason, the "traditional" turnout model was coming up with more likely voters than the "expanded" turnout model for a long time. That has now reversed and may help to explain the convergence.