National Tracking Poll Thread

[J. J.]

The 'bots seem to be capturing something; Romney's up by 4.

I doubt it will last.

By "'bots" do you mean "only Rasmussen"?

Gallup hasn't reported yet.  Yes, Rasmussen = the 'bots. 

[MorningInAmerica]

The 'bots seem to be capturing something; Romney's up by 4.

I doubt it will last.

By "'bots" do you mean "only Rasmussen"?

Gallup hasn't reported yet.  Yes, Rasmussen = the 'bots. 

Rasmussen On Monday August 27th:
Obama: 47%
Romney: 44%

Rasmussen Today:
Romney: 48%
Obama: 44%

Thats a net 7 point shift to Obama. I'd say that's a pretty fair bounce.

Reuters on Monday, August 27th
Obama: 46%
Romney 42%

Reuters today:
Obama: 44%
Romney: 43%

That's a net 3 point shift.

Still waiting on Gallup. But 2 of the 3 tracking polls show a shift.

[Eraserhead]

Too bad Reuters shifted two points away from Romney yesterday.

[Likely Voter]

Sunday:

Rasmussen (LV): Romney +4  
Obama     44
Romney     48 (+1)

Gallup (RV): Obama +1
Obama     47
Romney     46


1 week ago

Rasmussen: Obama +2
Obama     47
Romney     45

Gallup: tied
Obama     46
Romney     46

[Devils30]

Only reason Romney gaining in Rasmussen is b/c they changed their formula to make it even more republican. If you really believe the electorate will be 38% GOP, 33% Dem and 29% indy then Mitt may really win easy!

[MorningInAmerica]

Today on Rasmussen (w/ leaners):
Obama - 45%
Romney - 49%

9/2/2008 (w/ leaners)
Obama - 51%
McCain - 45%

9/2/2004 (w/ leaners)
Bush - 51%
Kerry - 47%

[MorningInAmerica]

Too bad Reuters shifted two points away from Romney yesterday.

...and today it shifts back towards Romney: http://www.reuters.com/article/2012/09/02/us-usa-campaign-poll-idUSBRE87U1CJ20120902


Obama - 45% (-)
Romney - 45% (+1)

Same poll on Monday, before the RNC.
Obama - 46%
Romney - 42%

So there's been a net 4 point swing to Romney on Reuters, and a net 7 point swing to Romney on Rasmussen. No bounce from Gallup.

[MorningInAmerica]

Reuters Update: 9/2/2012

http://www.reuters.com/article/2012/09/02/us-usa-campaign-poll-idUSBRE87U1CJ20120902
Obama - 45% (-)
Romney - 45% (+1)

Same poll on Monday, before the RNC.
Obama - 46%
Romney - 42%

There's been a net 4 pt swing to Romney on Reuters, and a net 7 pt swing to Romney on Rasmussen. No bounce from Gallup.

[Frozen Sky Ever Why]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

[MorningInAmerica]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

As far as I can tell, around here, the only polls that are "crap" are the ones you disagree with.

[Starbucks Union Thug HokeyPuck]

Traditionally I track the race through the combined signs the state polls are pointing towards, CBS/NYT national polls and ABC/washington post nat'l polls. 

I dont care what your political philosophy is, the electorate in a Presidential election is not going to be GOP +4, so no, I'm not paying attention to the Rasmussen tracker.

[後援会]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

As far as I can tell, around here, the only polls that are "crap" are the ones you disagree with.

Basically. I've never seen an actual criticism based on methodology. I've only seen "oh, that poll sucks because it has to be wrong".

[後援会]

You must be logged in to read this quote.

This is not a very helpful poll.

[Sbane]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

As far as I can tell, around here, the only polls that are "crap" are the ones you disagree with.

Basically. I've never seen an actual criticism based on methodology. I've only seen "oh, that poll sucks because it has to be wrong".

The post above yours is criticism based on methodology.

[Sbane]

Too bad Reuters shifted two points away from Romney yesterday.

...and today it shifts back towards Romney: http://www.reuters.com/article/2012/09/02/us-usa-campaign-poll-idUSBRE87U1CJ20120902


Obama - 45% (-)
Romney - 45% (+1)

Same poll on Monday, before the RNC.
Obama - 46%
Romney - 42%

So there's been a net 4 point swing to Romney on Reuters, and a net 7 point swing to Romney on Rasmussen. No bounce from Gallup.

Rasmussen has shown the race to be about tied usually, hasn't it? Odd that Obama led by 3 just before the convention. In any case a 3-4 bounce seems to be showing up. Gallup won't show it for a while because they have a 7 day sample, so we have to be patient there.

[後援会]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

As far as I can tell, around here, the only polls that are "crap" are the ones you disagree with.

Basically. I've never seen an actual criticism based on methodology. I've only seen "oh, that poll sucks because it has to be wrong".

The post above yours is criticism based on methodology.

Not really. The argument is "There's absolutely no way the electorate is R+4. So the poll is wrong. So I can ignore it."

Now, don't get me wrong. I don't actually think the electorate is R+4. But you don't toss out a polls that you think are wrong. That's not how statistics work.

My attempt to explain why this poll may seem to not match what you would expect is that I suspect the same "convention bounce" phenomenon that causes more people to say they're voting for a candidate may also cause more people to identify with that's candidate.

I think there's this assumption that "oh, Rasmussen will have their bounce bring Romney up to +4, and then they'll adjust it to +8 with their R+4". I find that a very unlikely outcome.

[J. J.]

I really don't know why everyone is attacking the polls.  There normally is a post convention bounce; we've all been expecting one.  It isn't gigantic. 

Now it's starting to show up.  If it's like the rest, it won't be around for long.

J. J.'s First Rule of Elections:

If a candidate that say something like “I don't look at the polls,” or “The only polls that matter are the ones on Election Day,” that candidate will lose.

That could apply in this case.

[Lief 🗽]

When did Barack Obama or Mitt Romney say either of those things?

[Sbane]

Ok, so if Rasmussen and Gallup are both crap (which would also screw up the average) what polls am I supposed to follow the race with?!?

As far as I can tell, around here, the only polls that are "crap" are the ones you disagree with.

Basically. I've never seen an actual criticism based on methodology. I've only seen "oh, that poll sucks because it has to be wrong".

The post above yours is criticism based on methodology.

Not really. The argument is "There's absolutely no way the electorate is R+4. So the poll is wrong. So I can ignore it."

Now, don't get me wrong. I don't actually think the electorate is R+4. But you don't toss out a polls that you think are wrong. That's not how statistics work.

My attempt to explain why this poll may seem to not match what you would expect is that I suspect the same "convention bounce" phenomenon that causes more people to say they're voting for a candidate may also cause more people to identify with that's candidate.

I think there's this assumption that "oh, Rasmussen will have their bounce bring Romney up to +4, and then they'll adjust it to +8 with their R+4". I find that a very unlikely outcome.

It's still a criticism of the methodology, which you said nobody was doing. Of course you don't toss it out though. You probably weren't here for it, but people were doing the same with polls back in June and July and just saying they weren't going to look at it or consider it due to Dem friendly samples. I think a criticism of a overly GOP or Dem electorate is fair, but like you said there is no reason to throw out the poll. Just add it to the mix.

I really don't know why everyone is attacking the polls.  There normally is a post convention bounce; we've all been expecting one.  It isn't gigantic. 

Now it's starting to show up.  If it's like the rest, it won't be around for long.

J. J.'s First Rule of Elections:

If a candidate that say something like “I don't look at the polls,” or “The only polls that matter are the ones on Election Day,” that candidate will lose.

That could apply in this case.

LOL where were you when people were discounting polls earlier this summer that showed Obama up big? You're such a lovable hack.

[Teemu]

I dont care what your political philosophy is, the electorate in a Presidential election is not going to be GOP +4, so no, I'm not paying attention to the Rasmussen tracker.

If we look at the Rasmussen party ID poll history (2004,2006,2008,2010) for the 3 pre-election months August, September and October, on only 2008 August and September, the Rasmussen monthly party ID gap was more Republican than the eventual election exit poll party ID gap. The 2008 August and September numbers didn't have the full impact of the economic collapse yet, and the October party ID (D+7.1) was the same as the exit poll party ID.


rasmussenreports.com/public_content/archive/mood_of_america_archive/partisan_trends/summary_of_party_affiliation

Rasmussen August Party ID gap, [Rasmussen October party ID gap], (exit poll of the November election night)

2010: D+1.2 [D+2.9] (D=R)
2008: D+5.7 [D+7.1] (D+7)
2006: D+5.3 [D+6.2] (D+2)
2004: D+2.6 [D+1.5] (D=R)

*the poll sample that had more favorable party ID gap for Republicans than the eventual exit poll on bold

So historically, Rasmussen party ID gap has been more favorable to the Democrats than the exit poll party ID gap. Maybe this year is the first time they got the party ID horribly wrong in favor of Republicans on election year, despite of their 15k sample size for the monthly party ID poll, maybe not.

[Teemu]

Also for the first time Republican party has better pre-convention net (un)favorables (44/50 {-6}) than Democratic party (43/52 {-9}) in Gallup polls, Gallup's numbers begin from 1992 convention.

gallup.com/poll/156959/gop-favorability-matches-2008-pre-convention-level.aspx

So it's not like Rasmussen is the only pollster spotting something going on with parties that hasn't been seen for a while.

[MorningInAmerica]

September 3, 2012

Rasmussen: Romney +4
Obama : 44% (-)
Romney: 48% (-)

Rasmussen w/ leaners: Romney +4
Obama: 46% (+1)
Romney: 50% (+1)

[J. J.]

When did Barack Obama or Mitt Romney say either of those things?

Just a reminder, Lief.  Just a reminder. 

[J. J.]

I really don't know why everyone is attacking the polls.  There normally is a post convention bounce; we've all been expecting one.  It isn't gigantic. 

Now it's starting to show up.  If it's like the rest, it won't be around for long.

J. J.'s First Rule of Elections:

If a candidate that say something like “I don't look at the polls,” or “The only polls that matter are the ones on Election Day,” that candidate will lose.

That could apply in this case.

LOL where were you when people were discounting polls earlier this summer that showed Obama up big? You're such a lovable hack.

What polls showed Obama up big?

Right now, I think it is too early to use the polls to make a prediction about November, but that doesn't mean that I think the polls are bad, especially the main ones.

[Likely Voter]

Monday:

Rasmussen (LV): Romney +4  
Obama     44
Romney     48

Gallup (RV): Obama +1
Obama     47
Romney     46



1 week ago

Rasmussen: Obama +3
Obama     47
Romney     44

Gallup: Romney +1
Obama     46
Romney     47