Romney Looks Like the Next Pres

pspr · Sep 18, 2012

Quote from Lucrum:

Maybe you should start your own thread. Something like: "Odumbo looks like the next dictator".

Oh, I thought you said, '...the next dickeater.'

mrbill · Sep 18, 2012

Quote from AK Forty Seven:

Sounds like you and Rasmussen,but Rasmussen is often wrong

Ouch! Good one AK. Do you do research for a living? How in the heck do you find all these old charts? This is great.

Of course everyone loves the charts that show things in their favor. But, to see so much evidence and still try to refute is a bit odd.

jem · Sep 18, 2012

Quote from Ricter:

You can keep saying this but it doesn't make it true. To prove that a polling agency's sample of a population is invalid or unreliable you will need the services of a polling agency.

wrong... that b.s. sort of worked on me a few months ago, but I have done some research.

the 2010 electorate... was D 35 / I 30 / R 35.

So the question becomes... what turnout are you expecting now?
If you expect a balanced turnout you should have a balanced sample of likely voters.

so your poll should sample at D 35/ I 30 / R 35.

If you prefer slanted polls then you should be expecting a slanted turnout.

Its very simple really.

pspr · Sep 18, 2012

Quote from jem:

Its very simple really.

That's where you lose ricter.

AK Forty Seven · Sep 18, 2012

Quote from jem:

wrong... that b.s. sort of worked on me a few months ago, but I have done some research.

the 2010 electorate... was D 35 / I 30 / R 35.

So the question becomes... what turnout are you expecting now?
If you expect a balance turnout you should have a balanced sample of likely voters.

so your poll should sample at D 35/ I 30 / R 35.

If you prefer slanted polls than you should be expecting a slanted turnout.

Its very simple really.

Kerry got 59 million votes in 04

Obama got 69 million votes 08

Dems got 42 million House votes in 06

Dems got 38 million House votes in 2010

You can not use non presidential election numbers to predict a presidential election turn out because dems have much lower turn out in non presidential election years.Its that simple really

jem · Sep 18, 2012

Quote from AK Forty Seven:

Kerry got 59 million votes in 04

Obama got 69 million votes 08

Dems got 42 million House votes in 06

Dems got 38 million House votes in 2010

You can not use non presidential election numbers to predict a presidential election turn out because dems have much lower turn out in non presidential election years.Its that simple really

1. so what were the R turnouts like in those same races... I supect 2010 shows just how motivated Rs were then and they are even more motivated now.

2. this is like trading with analogs... you are trying to make your best guess at a future pattern so you look at previous ones and make adjustments.

you prefer 2008 as your analog... but I know that the dems have lost millions of registered voters... and the polls show that dems have low enthusiam.

As of now, I am pretty sure Rs are going to come out in very strong numbers.

so I think the electorate will be more like 42 R and 35 D... but for the sake of good polling concepts I choose to use the 2010 electorate as my analog.

mrbill · Sep 18, 2012

Quote from jem:

wrong... that b.s. sort of worked on me a few months ago, but I have done some research.

the 2010 electorate... was D 35 / I 30 / R 35.

So the question becomes... what turnout are you expecting now?
If you expect a balanced turnout you should have a balanced sample of likely voters.

so your poll should sample at D 35/ I 30 / R 35.

If you prefer slanted polls then you should be expecting a slanted turnout.

Its very simple really.

Actually, I agree with your premise. Is there anyone here who actually does polling? My understanding is that they make 10,000 calls for example, randomly, and ask questions. How has that changed over the years? I really don't know. Of course we can call around the Stanford zip code, and then the Berkley zip code, and try to skew the results. Is that what you're saying? When I googled random sampling, I saw what seemed to be a straight forward explanation, much as from college. Is there something you think that can help Biased Pollsters get the results they are trying to skew? And, how does coming up with skewed polls help the pollsters. Seems like it would be bad for business.

AK Forty Seven · Sep 18, 2012

Quote from jem:

1. so what were the R turnouts like in those same races... I supect 2010 shows just how motivated Rs were then and they are even more motivated now.

Bush got 62 million votes in 04

McCain got 59 million votes 08

R's got 35 million House votes in 06

R's got 44 million House votes in 2010

AK Forty Seven · Sep 18, 2012

Quote from jem:

2. this is like trading with analogs... you are trying to make your best guess at a future pattern so you look at previous ones and make adjustments.

Yes but you have to compare apples to apples/ presidential election turn out to presidential election turn out not presidential election turn out to mid terms turn out

AK Forty Seven · Sep 18, 2012

Quote from jem:

1. so what were the R turnouts like in those same races... I supect 2010 shows just how motivated Rs were then and they are even more motivated now.

2. this is like trading with analogs... you are trying to make your best guess at a future pattern so you look at previous ones and make adjustments.

you prefer 2008 as your analog... but I know that the dems have lost millions of registered voters... and the polls show that dems have low enthusiam.

As of now, I am pretty sure Rs are going to come out in very strong numbers.

so I think the electorate will be more like 42 R and 35 D... but for the sake of good polling concepts I choose to use the 2010 electorate as my analog.

So if Obama wins will you admit that your line of thinking was wrong and that Rasmussen is a garbage pollster ?