jabowery
11-20-2007, 09:48 PM
The news that in Zogby's "blind poll" Ron Paul won with general voters but lost with "likely Republican voters" points the way to a new strategy to get the nomination:
Blind Poll Canvasing
What this means is construct a scientifically unbiased blind poll (see below) and use it to conduct phone canvasing of general voters to see which of them would be likely to vote Republican if they knew the positions of the candidates. This combines a scientific blind poll with phone canvasing to "get out the vote".
First let's first note a point about the Zogby poll for how not to do a blind poll:
Since likely Republican voters are known by pollsters to care more about immigration than the war, Zogby’s omission of immigration from the candidates’ positions must be taken to mean Zogby doesn’t have the first idea about constructing unbiased blind polls.
Moreover, the astounding performance of Ron Paul among general voters isn’t at all surprising to those who follow the prediction markets like intrade.com where people put their money where their mouths are.
Ron Paul has been leading the entire GOP pack in the likelihood of being elected president, if nominated, according to the professional gamblers:
http://blog.vdare.com/archives/2007/11/19/ron-paul-most-electable-republican-for-the-last-seven-weeks/
How can we construct an unbiased blind poll which, with as few questions as practical, identifies voters likely to vote for Ron Paul in the primary if they were fully informed, and use it to turn potential votes for Ron Paul in the general election into actual votes for Ron Paul in the primary?
This can be accomplished in 3 stages:
Construct a questionnaire by setting up a tournament with a nonrefundable entry fee of $X, awarding a first prize of $Y*8, second prize of $Y*4, third prize of $Y*2, and fourth prize of $Y, where X is sufficient to deter casual players from participating and Y is sufficient to motivate most serious "experts". The daily elimination rounds are conducted as follows:
Each contestant submits a new questionnaire of 5 multiple choice questions they think are most informative.
Judges disqualify contestants who submit frivolous or abusive questions, along with their questionnaires.
The questionnaires are made available to the contestants.
A phone survey is run during which each questionnaire, along with 5 questions selected at random from among the other questionnaires, is asked of at least one potential voter.
Each contestant receives the responses to their questionnaires and the additional questions, but not the responses to those questions.
Each contestant then tries to guess the responses they were not given, scoring one point for each response they get correct.
The contestants scoring in the upper half return for the next round and all are provided all questions and responses given by all potential voters.
Put the candidates through the winning questionnaire.
Send the Ron Paul volunteers forth to canvas the general voting public with the winning questionnaire, informing potential Ron Paul supporters why they are a likely Ron Paul supporter and offering to help them vote in the primary.
Blind Poll Canvasing
What this means is construct a scientifically unbiased blind poll (see below) and use it to conduct phone canvasing of general voters to see which of them would be likely to vote Republican if they knew the positions of the candidates. This combines a scientific blind poll with phone canvasing to "get out the vote".
First let's first note a point about the Zogby poll for how not to do a blind poll:
Since likely Republican voters are known by pollsters to care more about immigration than the war, Zogby’s omission of immigration from the candidates’ positions must be taken to mean Zogby doesn’t have the first idea about constructing unbiased blind polls.
Moreover, the astounding performance of Ron Paul among general voters isn’t at all surprising to those who follow the prediction markets like intrade.com where people put their money where their mouths are.
Ron Paul has been leading the entire GOP pack in the likelihood of being elected president, if nominated, according to the professional gamblers:
http://blog.vdare.com/archives/2007/11/19/ron-paul-most-electable-republican-for-the-last-seven-weeks/
How can we construct an unbiased blind poll which, with as few questions as practical, identifies voters likely to vote for Ron Paul in the primary if they were fully informed, and use it to turn potential votes for Ron Paul in the general election into actual votes for Ron Paul in the primary?
This can be accomplished in 3 stages:
Construct a questionnaire by setting up a tournament with a nonrefundable entry fee of $X, awarding a first prize of $Y*8, second prize of $Y*4, third prize of $Y*2, and fourth prize of $Y, where X is sufficient to deter casual players from participating and Y is sufficient to motivate most serious "experts". The daily elimination rounds are conducted as follows:
Each contestant submits a new questionnaire of 5 multiple choice questions they think are most informative.
Judges disqualify contestants who submit frivolous or abusive questions, along with their questionnaires.
The questionnaires are made available to the contestants.
A phone survey is run during which each questionnaire, along with 5 questions selected at random from among the other questionnaires, is asked of at least one potential voter.
Each contestant receives the responses to their questionnaires and the additional questions, but not the responses to those questions.
Each contestant then tries to guess the responses they were not given, scoring one point for each response they get correct.
The contestants scoring in the upper half return for the next round and all are provided all questions and responses given by all potential voters.
Put the candidates through the winning questionnaire.
Send the Ron Paul volunteers forth to canvas the general voting public with the winning questionnaire, informing potential Ron Paul supporters why they are a likely Ron Paul supporter and offering to help them vote in the primary.