Non-Gallup/Rasmussen tracking polls thread

[J. J.]

I'm saying the subsample is bad, but a bad subsample doesn't invalidate the poll.

So, in other words, you are maintaining that it's a one-in-several-million event, and not a flaw in the methodology.  Why?

No, I'm saying that the poll isn't primary designed to accurately measure a subsample.

Yes, but at some level you have to question the methodologies, especially if the topline result isn't in line with other polls.

The subsample can reveal the methodological weaknesses of a poll.  If a subsample shows 500% growth among elderly voters, then perhaps the poll is doing something funky.  It's wrong to pick apart moderate subsample weaknesses (say McCain winning 15% of blacks, a common subsample weakness with about the same sample size as Youth), but at some point we got to take a step back and say that this poll is probably doing something wrong.  The odds are simply one in a trillion that this poll doesn't have an overarching flaw.

At what point do you, J.J., question a sub sample?  Is a one and a gazillion bazillion chance not enough for you?  Does McCain need to win 100% of Blacks in Mississippi?

The problem that I'm having with this entire argument is that, even with a bad subsample, the poll is conforming to the other polls, or at least was until today.

[Alcon]

The problem that I'm having with this entire argument is that, even with a bad subsample, the poll is conforming to the other polls, or at least was until today.

So it doesn't matter how invalid the methodology of the poll is, as long as it comes within close range of other polls?

hellooo arg

[J. J.]

The problem that I'm having with this entire argument is that, even with a bad subsample, the poll is conforming to the other polls, or at least was until today.

So it doesn't matter how invalid the methodology of the poll is, as long as it comes within close range of other polls?

hellooo arg

No, it means that there probably is not a problem with the methodology that affects the result.

[Alcon]

The problem that I'm having with this entire argument is that, even with a bad subsample, the poll is conforming to the other polls, or at least was until today.

So it doesn't matter how invalid the methodology of the poll is, as long as it comes within close range of other polls?

hellooo arg

No, it means that there probably is not a problem with the methodology that affects the result.

So, you think that it was just a one-in-multi-million error?  You think that is more likely than a sampling error?

Thanks for confirming that.

[J. J.]

The problem that I'm having with this entire argument is that, even with a bad subsample, the poll is conforming to the other polls, or at least was until today.

So it doesn't matter how invalid the methodology of the poll is, as long as it comes within close range of other polls?

hellooo arg

No, it means that there probably is not a problem with the methodology that affects the result.

So, you think that it was just a one-in-multi-million error?  You think that is more likely than a sampling error?

Thanks for confirming that.

No, I'm saying that whatever the problem, it doesn't seem to effect the result.  Please don't put words into my mouth.

[Alcon]

No, I'm saying that whatever the problem, it doesn't seem to effect the result.  Please don't put words into my mouth.

I'll let you put words in your own mouth, then.  There are three options:

1) The poll's sample is not as "off" as it seems. Obama is performing worse than Kerry among youths -- probably much worse.

2) The one-in-many-million event occurred, and the poll was just off by chance.

3) The poll had a methodological flaw.

You need to pick one.  There are no other possible options here -- either the poll isn't off, it is off by random chance, or it is off by error.  Which option are you arguing?

[J. J.]

No, I'm saying that whatever the problem, it doesn't seem to effect the result.  Please don't put words into my mouth.

I'll let you put words in your own mouth, then.  There are three options:

1) The poll's sample is not as "off" as it seems. Obama is performing worse than Kerry among youths -- probably much worse.

2) The one-in-many-million event occurred, and the poll was just off by chance.

3) The poll had a methodological flaw.

You need to pick one.  There are no other possible options here -- either the poll isn't off, it is off by random chance, or it is off by error.  Which option are you arguing?

4)  The poll is not designed to accurately measure the subsamples.  That isn't a methodological flaw because it's not designed to get that type of information.

[Alcon]

No, I'm saying that whatever the problem, it doesn't seem to effect the result.  Please don't put words into my mouth.

I'll let you put words in your own mouth, then.  There are three options:

1) The poll's sample is not as "off" as it seems. Obama is performing worse than Kerry among youths -- probably much worse.

2) The one-in-many-million event occurred, and the poll was just off by chance.

3) The poll had a methodological flaw.

You need to pick one.  There are no other possible options here -- either the poll isn't off, it is off by random chance, or it is off by error.  Which option are you arguing?

4)  The poll is not designed to accurately measure the subsamples.  That isn't a methodological flaw because it's not designed to get that type of information.

That makes no sense at all.  How can the poll be overall sound, while its sub-samples can readily fall ridiculously outside the MoE?  That means it's not an accurate sample.  Margin of error applies at low samples, too, just more liberally.

How exactly does one design a poll to not accurately measure a subsample within MoE, without having the poll sample incorrectly?

[Tender Branson]

Zogby

Obama: 51 (-1)
McCain: 41 (+1)

[Sam Spade]

Zogby

Obama: 51 (-1)
McCain: 41 (+1)

51.3 to 41.0 to be exact.  Whatever, Zogby's worth the entertainment.

[Sam Spade]

Rest of the early morning released trackers...

Diageo Hotline Poll - 10/24
Obama 50% (+2)
McCain 43% (nc)

GWBattleground - 10/24
Obama 49% (nc)
McCain 46% (+1)

[Sam Spade]

Daily Tracker Table - October 24, 2008

Poll Name	Obama	McCain	Margin	Change	Observation
Zogby	51.3%	41.0%	O+10.3%	M+1.6%	Who knows - it's Zogby!
Rasmussen	51.80%	44.80%	O+7.00%	O+0.17%	Good Obama sample fell off replaced by another good Obama sample.  Good for Obama.
Battleground	49%	46%	O+3%	M+1%	Hard to tell.  I think there may be a pro-McCain sample in the middle weighting it down a little (1-2 pts max), but who knows.
Hotline	50%	43%	O+7%	O+2%	Obviously, a good Obama sample came on today.
R2000/DKos	52%	40%	O+12%	O+2%	Good Obama sample came on today.
NOT UPDATED
Gallup
Expanded	51%	44%	O+7%	O+1%	I think a fairly decent Obama sample moved on.
Traditional	50%	45%	O+5%	O+1%
IBD/TIPP	45.8%	42.3%	O+3.5%	O+2.4%	Good Obama sample moved on.  Of course, this could really be just movement around the center.
ABC/WP	53%	44%	O+9%	M+2%	Hard to tell how the samples work here.
POLLS AVERAGE	50.43%	43.20%	O+7.23%	O+0.50%	Slight Obama gain, probably, from yesterday.  Regardless of what the numbers say, I still see this as a 5-7 point lead right now.

[Alcon]

Thanks for the daily tracker...tracker, Sam.

[ChrisFromNJ]

Daily Tracker Table - October 24, 2008

Poll Name	Obama	McCain	Margin	Change	Observation
Zogby	51.3%	41.0%	O+10.3%	M+1.6%	Who knows - it's Zogby!
Rasmussen	51.80%	44.80%	O+7.00%	O+0.17%	Good Obama sample fell off replaced by another good Obama sample.  Good for Obama.
Battleground	49%	46%	O+3%	M+1%	Hard to tell.  I think there may be a pro-McCain sample in the middle weighting it down a little, but who knows.
Hotline	50%	43%	O+7%	O+2%	Obviously, a good Obama sample came on today.
R2000/DKos	52%	40%	O+12%	O+2%	Good Obama sample came on today.
NOT UPDATED
Gallup
Expanded	51%	45%	O+6%	M+2%	Looks like Gallup midweek movement to me.
Traditional	50%	46%	O+4%	M+1%	Ditto.
IBD/TIPP	44.8%	43.7%	O+1.1%	M+2.6%	Looks like a good McCain sample fell on.
ABC/WP	54%	43%	O+11%	NC	No real change as far as I can tell here.
POLLS AVERAGE	50.82%	42.96%	O+7.86%	O+0.31%	Only includes polls updated so far today.

Thanks for this, Sam!

[Lief 🗽]

Thanks for the daily tracker...tracker, Sam.

^^^ Ditto, it's very convenient.

[Moooooo]

Thanks for the daily tracker...tracker, Sam.

^^^ Ditto, it's very convenient.

I agree... very nice

[J. J.]

Thanks for the daily tracker...tracker, Sam.

Absolutely!  Thank you.

[TheGlobalizer]

Thanks for the daily tracker...tracker, Sam.

^^^ Ditto, it's very convenient.

Concur.

[Rowan]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

[J. J.]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

Probably a bad sample dropping.

[Rowan]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

Probably a bad sample dropping.

Yeah, a good Mac day fell off.

[J. J.]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

Probably a bad sample dropping.

Yeah, a good Mac day fell off.

Probably not the methodology.

[Alcon]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

Probably a bad sample dropping.

Yeah, a good Mac day fell off.

Probably not the methodology.

Is that your unsubtle substitute for responding to my last post? 

[J. J.]

IBD/TIPP
Obama 45.8%(+1.0)
McCain 42.3%(-1.4)

Probably a bad sample dropping.

Yeah, a good Mac day fell off.

Probably not the methodology.

Is that your unsubtle substitute for responding to my last post? 

Well, since you've declined to respond to a few ... .

In this case, looking at the rest of the polls, along with TIPP's record, it was fairly clear it wasn't methodology.

[Alcon]

Well, since you've declined to respond to a few ... .

What posts were those?  It was not intentional.  I will respond to any post you link me to.

I would appreciate the same in return.  Here is the post in question.

If you say it's "fairly clear that it [wasn't] methodology," you're saying it's fairly clear that it was a one-in-many-million event.  It must be one, the other, or an accurate observation.  I am still waiting for you to pick one of the three.  They are the only three options.