Only 17.51% of RP supporters voted in Iowa. No wonder we were 5th. Learn from stats:

[Terry82]

You're making such a fucking fundamental error in your analysis that it's laughable.

How are you determining how many supporters Ron Paul has in Iowa in total? By multiplying his polling % by what you believe is the total eligible voters? And you don't see the flaw in this?

YOU'RE COMPARING TWO ENTIRELY DIFFERENT GROUPS!

The % given in terms of polling is simply that of Ron Paul support among likely voters. What you're then assuming is that he has THE SAME support among a general population. There's no reason to assume this, and as far as I know, no such poll was ever done. He might have a lot of supporters which didn't vote, or maybe almost every Ron Paul supporter did vote. Your analysis in no way determines this though.

What you calculated is the VAR, Voter Apathy Ratio. This number will be roughly the same for almost all candidates, because it's simply the average of people who voted in Iowa and who didn't.

To recap, Zogby gave a poll with RP at 10%. Your analysis would then in turn assume RP has 112,500 supporters in Iowa. This is wrong. Ron Paul could have a million supporters, most of whom didn't bother voting, or 12,000 with almost every Ron Paul supporter voting.

 

[justatrey]

You're making such a fucking fundamental error in your analysis that it's laughable.

How are you determining how many supporters Ron Paul has in Iowa in total? By multiplying his polling % by what you believe is the total eligible voters? And you don't see the flaw in this?

YOU'RE COMPARING TWO ENTIRELY DIFFERENT GROUPS!

The % given in terms of polling is simply that of Ron Paul support among likely voters. What you're then assuming is that he has THE SAME support among a general population. There's no reason to assume this, and as far as I know, no such poll was ever done. He might have a lot of supporters which didn't vote, or maybe almost every Ron Paul supporter did vote. Your analysis in no way determines this though.

What you calculated is the VAR, Voter Apathy Ratio. This number will be roughly the same for almost all candidates, because it's simply the average of people who voted in Iowa and who didn't.

To recap, Zogby gave a poll with RP at 10%. Your analysis would then in turn assume RP has 112,500 supporters in Iowa. This is wrong. Ron Paul could have a million supporters, most of whom didn't bother voting, or 12,000 with almost every Ron Paul supporter voting.

I think he is right; I overlooked this too.

What you would really need is a poll that estimated his support among all eligible voters, not just likely voters. If a poll like this exists, someone please post it. I'd like to know what it said so can estimate turnout and see if we're improving in future states.

 

[cheese]

bump

 

[Opulen]

people of New Hampshire please get out and vote for Cong. Ron Paul. After that do the Ron Paul dance..

 

[tttar]

You're making such a fucking fundamental error in your analysis that it's laughable.

How are you determining how many supporters Ron Paul has in Iowa in total? By multiplying his polling % by what you believe is the total eligible voters? And you don't see the flaw in this?

YOU'RE COMPARING TWO ENTIRELY DIFFERENT GROUPS!

The % given in terms of polling is simply that of Ron Paul support among likely voters. What you're then assuming is that he has THE SAME support among a general population. There's no reason to assume this, and as far as I know, no such poll was ever done. He might have a lot of supporters which didn't vote, or maybe almost every Ron Paul supporter did vote. Your analysis in no way determines this though.

What you calculated is the VAR, Voter Apathy Ratio. This number will be roughly the same for almost all candidates, because it's simply the average of people who voted in Iowa and who didn't.

To recap, Zogby gave a poll with RP at 10%. Your analysis would then in turn assume RP has 112,500 supporters in Iowa. This is wrong. Ron Paul could have a million supporters, most of whom didn't bother voting, or 12,000 with almost every Ron Paul supporter voting.

I tried to duplicate the same percentage for Ron Paul, as the voter turnout stats you can see in the Turnout sheet.

On that sheet, you see the Voting Eligible Populations for each state, and then it tells you what subset of that group showed up to vote.

Finding out how many people in that VEP "support" a particular candidate before they actually vote entails that one relies on polls, and you're right that you have to make some assumptions, but I doubt the groups are "entirely different".

We know the polls before the election, and we know how many people showed up to support individual candidates after the election.

What else is there to use? You don't know of anything either, so there's no way to correct the "error". It was better than not thinking about this at all.

I defined "supporter" as loosely as any pollster would - just someone who'd told a pollster he liked Ron Paul.

If the polls we love to follow actually mean anything, they showed that there's a higher later participation of Ron Paul "supporters" than of the average for all others - but it's not high enough.

Those numbers suggested we need about 3 times as many polled "supporters" to care enough to show up in the end.

So maybe some of our attention would be better used to get existing "supporters" to go to the polls, rather than in trying to make new converts.

It's a problem of apathy, but maybe also one of changing some simple habits - very many more in New Hampshire already vote in the primaries (because they know NH COUNTS), so I'm sure something might be said to make those in other states think the same way.

Nothing to disagree with, there.

 

[RonPaulNLiveFree]

every supporter for Ron Paul better vote in the primary or kiss this election goodbye, get off your asses and vote! dammit!

 

[rs3515]

Just a number of fundamental flaws with the analysis ...

First, having been the person directly involved with the Letter Writing Campaign, I can tell you nearly 1/3 of all registered voters in Iowa are "no party". Over 700,000 people.

Second, taking the 6% figure and extrapolating it to an entire set of registered voters is an incorrect assumption. You have to realize there are many people who might be Ron Paul supporters, but have never even heard of the guy! Your analysis assumes 100% recognition, which is impossible. Hell, a poll came out today that shows 12% of Americans think Rudy Giuliani is still mayor of NYC!

Third, saying only 17% of Ron Paul supporters went to the caucus is just untrue. Directly tied with the second point above.

To me, the analysis basically says, "if we made more people aware of Ron Paul, more people would vote for him". Uh, yeah ... don't think anyone is denying that.

 

[Richandler]

Iowa caucus was only like 2 hours long. Not exactly convient.

 

[rs3515]

I defined "supporter" as loosely as any pollster would - just someone who'd told a pollster he liked Ron Paul.

The problem though is you're assuming this factor using 100% of the voting population. You make the assumption we can get 100% of potential (not confirmed) voters to support Ron Paul ... AND on top of it, you assume we can extrapolate it to the entire electorate! Just totally incorrect.

I don't care if Ron Paul can turn water into gold and is running for president, there is no such thing as 100% name recognition or 100% turnout.

 

[ssmoler]

I tried to duplicate the same percentage for Ron Paul, as the voter turnout stats you can see in the Turnout sheet.

On that sheet, you see the Voting Eligible Populations for each state, and then it tells you what subset of that group showed up to vote.

Finding out how many people in that VEP "support" a particular candidate before they actually vote entails that one relies on polls, and you're right that you have to make some assumptions, but I doubt the groups are "entirely different".

We know the polls before the election, and we know how many people showed up to support individual candidates after the election.

What else is there to use? You don't know of anything either, so there's no way to correct the "error". It was better than not thinking about this at all.

I defined "supporter" as loosely as any pollster would - just someone who'd told a pollster he liked Ron Paul.

If the polls we love to follow actually mean anything, they showed that there's a higher later participation of Ron Paul "supporters" than of the average for all others - but it's not high enough.

Those numbers suggested we need about 3 times as many polled "supporters" to care enough to show up in the end.

So maybe some of our attention would be better used to get existing "supporters" to go to the polls, rather than in trying to make new converts.

It's a problem of apathy, but maybe also one of changing some simple habits - very many more in New Hampshire already vote in the primaries (because they know NH COUNTS), so I'm sure something might be said to make those in other states think the same way.

Nothing to disagree with, there.

You can't take a poll that has an assumed population of likely primary voters and apply it to the entire population of Iowa. You might as well take a poll of Ron Paul supporters showing 100% support of Ron Paul, apply that to the entire state of Iowa and go WTH! we should have gotten 100% of the vote! While it's not that extreme, the populations of the poll and the entire state of Iowa are really that different.

The only way to know how many people support Paul is just to do a random sample poll of the entire population of Iowa - something I don't think any professional pollsters are doing though it would be interesting. I think you'd find that while most people might have an opinion on who they like, the % of the population who are RP supports and the % who showed up at the primaries would be far and away the strongest of any candidate.

What we can look at is that professional pollsters try their best to artificially create the population of people they believe will vote in the primaries. The reason why the polls are off is because no one actually knows the population of who will vote before the voting actually happens. That's the pollsters job - to try and best create the population of who will vote and poll that group. The fact that RP received 10% over an expected 6% or whatever it was shows that a lot of RP supporters fall into the category of people the pollsters expected not to vote, and thats a GOOD THING.

So to sum up, while your hard work is appreciated, your methodology is unfortunately wrong - so stop lets stop causing all this depression and anger!

 

[tttar]

The problem though is you're assuming this factor using 100% of the voting population. You make the assumption we can get 100% of potential (not confirmed) voters to support Ron Paul ... AND on top of it, you assume we can extrapolate it to the entire electorate! Just totally incorrect.

I don't care if Ron Paul can turn water into gold and is running for president, there is no such thing as 100% name recognition or 100% turnout.

I'm saying "here's what we would need".

What's actually possible, I don't know. But 50% has been possible for others in New Hampshire, so I was hoping something like that would be possible for us in other states.

 

[Paul4Prez]

The details may be off, but the point tttar is making is basically correct -- Ron Paul can win in the low-turnout states if everyone who likes him just gets out and votes. This is especially true in low-turnout caucus states like Nevada and Maine.

 

[bakinjoo]

that's just not even possible...that only 18% of people who really support RP took the time to vote for him
if that's true, then Americans ARE as dumb as the politicians they elect, and get the government they deserve...
I just don't think that's true...

 

[tttar]

Just a number of fundamental flaws with the analysis ...

First, having been the person directly involved with the Letter Writing Campaign, I can tell you nearly 1/3 of all registered voters in Iowa are "no party". Over 700,000 people.

Second, taking the 6% figure and extrapolating it to an entire set of registered voters is an incorrect assumption. You have to realize there are many people who might be Ron Paul supporters, but have never even heard of the guy! Your analysis assumes 100% recognition, which is impossible. Hell, a poll came out today that shows 12% of Americans think Rudy Giuliani is still mayor of NYC!

Third, saying only 17% of Ron Paul supporters went to the caucus is just untrue. Directly tied with the second point above.

To me, the analysis basically says, "if we made more people aware of Ron Paul, more people would vote for him". Uh, yeah ... don't think anyone is denying that.

So how else would you get at the truth of how many people of those who recognized and liked Ron Paul better than the others actually showed up, in comparison to the average of those of other candidates?

Could I have had more accurate results by some other (available) method?

Of course everyone knows greater participation leads to greater percentages. I was looking for something that more precisely backed that up, and hoped we could say "x times more people need to show up" (which might lead to better-focused strategies), instead of "we oughta try harder" (which leads to "yeah, we oughta").

 

[tttar]

that's just not even possible...that only 18% of people who really support RP took the time to vote for him
if that's true, then Americans ARE as dumb as the politicians they elect, and get the government they deserve...
I just don't think that's true...

Those who "really support" aren't the problem. For all I know, those people turned out at much higher rates.

It's what % of our lukewarm people vs. what % of their lukewarm people.

A much larger pool.

 

[spacehabitats]

I am an Iowan, and have never been involved in politics before this year. I have never been to a caucus before this year. I was contacted a week before the caucuses because I had volunteered to help on Ron Paul's main website. Until that time I did not understand what needed to be done, and how important it is for all supporters to show up and vote. Some supporters I contacted did not plan to show up for the caucuses.
We need to make it very clear to all RP supporters that we all need to talk to, e-mail etc. everyone that we know. Talk to those we meet, and make sure they understand how important it is show up and vote.
I could have done so much more to help if I knew this months ago.
I think it was taken for granted that all supporters understood what needed to be done.

 

[ssmoler]

So how else would you get at the truth of how many people of those who recognized and liked Ron Paul better than the others actually showed up, in comparison to the average of those of other candidates?

Could I have had more accurate results by some other (available) method?

Of course everyone knows greater participation leads to greater percentages. I was looking for something that more precisely backed that up, and hoped we could say "x times more people are needed" (which might lead to better-focused strategies), instead of "we oughta try harder" (which leads to "yeah, we oughta").

Was Iowa a closed-party format? If so, get a random sample of everyone registered Republican in Iowa and ask them who they support for president, with a none of the above option as well of course. Then multiply the percentage for RP over the registered republican population to get a roughly accurate hard number and compare that number with the number of people who actually voted in the primaries for Paul. The latter divided over the former will give you the percentage of RP supporters who voted out of the population who supported him and could vote.

The point is you just can't take the poll you used and apply it to the general population.

 

[rs3515]

Of course everyone knows greater participation leads to greater percentages. I was looking for something that more precisely backed that up, and hoped we could say "x times more people are needed" (which might lead to better-focused strategies), instead of "we oughta try harder" (which leads to "yeah, we oughta").

I think it would need more analysis of the category breakdowns rather than an approach that spans the entire electorate. For example, the campaign has been touting that entrance polling showed 29% of Independents supported Ron Paul. Rather than seeing this as a victory, maybe we should see it as though we could have done a lot better.

Also, Huckabee captured 80% of the vote of those who were self-declared Evangelicals. While we had a few activities attempting to reach conservative Christians, there probably should have been more "boots on the ground" courting of faith leaders. We should have reached out to Iowa Meetup members who were active church goers, and ask them to meet with their church leaders to discuss about Ron Paul.

Most all political choices ultimately depend on personal interactions, not just watching debates or seeing videos on the Internet. This is what we should be taking from Iowa.

 

[colecrowe]

bump

 

[saahmed]

I don't understand this. Are you taking the poll numbers and extrapolating them to the entire population? Those polls poll only registered Republicans. He was polling at 6% among this sample, not the entire population.