Gallup Tracking Poll Thread [Obama vs McCain]

[Alcon]

lol I can't believe that. Fake poll ("air connu").

Gallup's long-established national tracking poll is a "fake poll" based on a 34% subsample that you intuitively feel is wrong?  Or is this a joke? 

[Alcon]

lol I can't believe that. Fake poll ("air connu").

Gallup's long-established national tracking poll is a "fake poll" based on a 34% subsample that you intuitively feel is wrong?  Or is this a joke? 

Gallup did crazy in 2004. Completely off. Hence, I'm cautious with this firm.

On the state level, not their tracking poll, I thought.

[Alcon]

I had a sneaking suspicion you'd be unable to resist posting this, Sam. 

[Alcon]

(its been up for about 30 minutes, at least)

I was on my way to!  Some of us don't subscribe to the damn Gallup RSS feed. 

[Alcon]

(its been up for about 30 minutes, at least)

I was on my way to!  Some of us don't subscribe to the damn Gallup RSS feed. 

I subscribe to the RSS feed? - that's news to me.  I'm too cheap to subscribe to anything except for this site.  I do know it was posted to RCP over 30 minutes ago - that's what I was reading it from.

An RSS feed is a free format, and Gallup doesn't actually use it.  I was being playful, but apparently that don't come across

[Alcon]

Amazingly stable thru the summer.  A 4 to 5 point race since the ist part of June.  Not even worth looking at the tracking polls until the conventions.  There will be no change.

Don't you keep saying that you're done with tracking polls? 

I think we're both having the same problem.

[Alcon]

and for likely voters ?  it stays the best for me. I don't care about RV who will not vote. Go in hell !

LV's tend to get more useful as election season goes on, and RV less so.  Early on, ignoring RV's in favor of LV's is potentially a bad idea.

RV's that aren't LV's are not unlikely voters.

[Alcon]

and for likely voters ?  it stays the best for me. I don't care about RV who will not vote. Go in hell !

LV's tend to get more useful as election season goes on, and RV less so.  Early on, ignoring RV's in favor of LV's is potentially a bad idea.

RV's that aren't LV's are not unlikely voters.

what then ?

Well, any voter who fails the LV screen...that doesn't mean they are unlikely to vote, necessarily.  Plenty of them will vote.

[Alcon]

This is a complete weekend sample in Gallup polling.

Unless it's a long weekend, that's never going to be the case.  Gallup is a three-day tracker.

[Alcon]

This is a complete weekend sample in Gallup polling.

Unless it's a long weekend, that's never going to be the case.  Gallup is a three-day tracker.

Saturday, Sunday and Monday, the reported on Tuesday.

Err?

It's Monday.  This poll is Friday-Sunday.  So, it's not "all-weekend."  I don't think last Friday was a holiday, was it?

[Alcon]

Obama is really running away with this race!

Feel like a clown now?

Feigned conditional outrage!  Feigned conditional outrage!

[Alcon]

Can Conservatives please explain why an all American war hero is losing in the polls to a man with a middle name Hussein?

I'm not a conservative, but I'm also not a total ignoramus, so I can.

[Alcon]

Monday, August 4, 2008

Obama - 46% (+1)
McCain - 43% (-1)

The three percentage point advantage for Obama matches the average since early June, when Obama clinched the number of delegates needed to head to the Democratic convention as the presumptive presidential nominee. Since then, Obama has never trailed McCain among registered voters, though McCain has tied Obama five times during this span, including Gallup Poll Daily tracking reports for last Friday and Saturday.

The Democratic convention is now just three weeks away, while the Republican convention will begin four weeks from today. The conventions are one of the most anticipated events of the election calendar, in part because each candidate typically receives a "bounce" in support in the polls following their official nomination as the party's presidential candidate.

[Alcon]

Oh well J. J.  Win some, lose some.

[Alcon]

Weekend bounce is back.

Sigh.

Umm... no.

There is not a weekend sample in this tracking poll.

Friday, Saturday, Sunday constitutes the weekend.

You're kinda underestimating the potential for static here, too.  Obama may do 1-3 points better on a weekend, but either candidate may do 1-3 points better randomly too.

[Alcon]

Hmm, I must say I am baffled how his lead didn't grow with Saturday polling included.

Because, for the second time, you are ignoring natural volatility/bad sampling, and assuming that this is a flat-lined race with Obama spikes on weekends.  Also, consider rounding; it's possible, maybe likely (tendency toward the mean) that the previous result was a "weak" +5.

[Alcon]

Hmm, I must say I am baffled how his lead didn't grow with Saturday polling included.

Because, for the second time, you are ignoring natural volatility/bad sampling, and assuming that this is a flat-lined race with Obama spikes on weekends.  Also, consider rounding; it's possible, maybe likely (tendency toward the mean) that the previous result was a "weak" +5.

There may have been a higher than normal sample for Obama midweek?

Very possible, but again:  Consider rounding.  Consider margin of error.  We're seeing movement here at a very low confidence rate.  People trust polls too literally, and then assail them too much because of it.

[Alcon]

Monday and Tuesday are included in these numbers.  Obama was about +1.5, maybe +2, beforehand.  Let's assume (arbitrarily) that Monday and Tuesday were +1.75 average together.  Obama is +6 today.

The Sunday sample would have to be at least Obama +14 for these numbers to be possible.

I don't think a contraction is unlikely, considering that Obama jumped up two points when the Sunday sample was included.  But, obviously his Monday sample was better than his Saturday   sample.  Does that have meaning?  Probably, but we can't be sure yet.

[Alcon]

Oh, how I really miss Mason-Dixon.

No kidding.

[Alcon]

As you can see from the most recent numbers, yes it does.

This is a fallacious argument.  This is not a static race.  There is statistical noise, and there is actual movement.  As such, simply comparing one weekend sample to one weekday sample, and determining a causation from that, is spurious.

[Alcon]

You know, at least in some ways, Having a Hurricane during McCain's convention could be more damaging than anything Obama could hit him with... Even if it just diverts some coverage.

Symbolism is a powerful thing.

Praying for victims can help.  Yes, I'm a cynic.

Praying for victims during the convention? 

[Alcon]

Yes, as in, "Let's have a moment to pray for those people in the path of Hurricane Gustav (Gustaf?)."  It was for the country, but Reagan did it in 1980.

I think Americans would see that as a nice gesture, but it hardly defuses the central problem:  the networks aren't going to pre-empt coverage of a major hurricane to re-broadcast the RNC footage -- prayer or otherwise.

[Alcon]

Yes, as in, "Let's have a moment to pray for those people in the path of Hurricane Gustav (Gustaf?)."  It was for the country, but Reagan did it in 1980.

I think Americans would see that as a nice gesture, but it hardly defuses the central problem:  the networks aren't going to pre-empt coverage of a major hurricane to re-broadcast the RNC footage -- prayer or otherwise.

I didn't say it would.  I was referring to the symbolism.  It also depends on if the storm is "photogenic," for lack of a better word.  Katrina itself wasn't, but the aftermath was.

And McCain really wants to get into a battle over photo ops regarding hurricanes?  That would present a great opportunity for stock footage.

In any case, I have to say, J. J., your ability to find optimism for Republicans in every twist and turn of this election season is remarkable

[Alcon]

The latest one-day sample is apparently Obama +15. That is where the bump is coming from.

[lololololol]

15-point bump

[/lololololol]

[Alcon]

The latest one-day sample is apparently Obama +15. That is where the bump is coming from.

[lololololol]

15-point bump

[/lololololol]

When did I ever say that?

No, no.  I was just saying that Obama is actually polling with a 15-point bump, like the McCain campaign "predicted."  Obviously just a result of a lucky sample and the convention bounce, but still

I'm just talking about the symbolic aspects and the images.  Think back to Katrina.  How much footage did you actually see of it hitting?  Huge amounts of the aftermath, but not of it hitting.

Hey, I lived through a political campaign during the Iranian Hostage Crisis.  I saw what worked.

No footage of it hitting.  But you don't think the media would run with this?



It seems to me like you'd be calling this potentially "very bad" or "disastrous" if this was Obama instead of McCain.  Maybe that's just me seeing that, though...