Any Statisticsfags in here? Am I messing this up? I was having a political discussion about the Florida election and while I had an epiphany. But I wonder if the overall premise of my analysis is just wrong and I'm dumb.
The person I was speaking to insisted that "People voting for Trump don't want to early vote or mail in vote and will vote in person on election day." Then they insisted that the sizeable Independent voter pool in Florida (26%) will vote Trump without citing hard facts and they didn't like polls because 2016. So I wanted to use hard data to extrapolate, based on their premise, what percent of the cryptic Independent voters were possibly voting for Biden over Trump. Here's what I came up with. If I'm wrong, please call me dumb and throw rocks until I scream and cry like a chimpanzee.
First here's my data sources:
https://electproject.github.io/Early-Vote-2020G/FL.html
https://dos.myflorida.com/elections/data-statistics/voter-registration-statistics/voter-registration-reportsxlsx/voter-registration-by-county-and-party/
Time for maths:
26% of registered voters in Florida are "Independents" aka "No Party Affiliated". If Trump voters vote in person and election day then let's look at how many "NPA/Independent" voters mailed in their ballots and voted early compared to the % total.
First in person "early voting" is at 17% of the overall total.
Second mail in ballots counted is at 20.5% of the overall total.
Therefore, based on the rates, 65% of all NPA in person early ballots were probably Biden votes.
Further, based on the rates, 79% of all NPA mail in ballots were probably Biden votes.
Is there a better way to analyze this data to extrapolate the % votes for which candidate or does this kinda work for what it is?
The person I was speaking to insisted that "People voting for Trump don't want to early vote or mail in vote and will vote in person on election day." Then they insisted that the sizeable Independent voter pool in Florida (26%) will vote Trump without citing hard facts and they didn't like polls because 2016. So I wanted to use hard data to extrapolate, based on their premise, what percent of the cryptic Independent voters were possibly voting for Biden over Trump. Here's what I came up with. If I'm wrong, please call me dumb and throw rocks until I scream and cry like a chimpanzee.
First here's my data sources:
https://electproject.github.io/Early-Vote-2020G/FL.html
https://dos.myflorida.com/elections/data-statistics/voter-registration-statistics/voter-registration-reportsxlsx/voter-registration-by-county-and-party/
Time for maths:
26% of registered voters in Florida are "Independents" aka "No Party Affiliated". If Trump voters vote in person and election day then let's look at how many "NPA/Independent" voters mailed in their ballots and voted early compared to the % total.
First in person "early voting" is at 17% of the overall total.
Second mail in ballots counted is at 20.5% of the overall total.
Therefore, based on the rates, 65% of all NPA in person early ballots were probably Biden votes.
Further, based on the rates, 79% of all NPA mail in ballots were probably Biden votes.
Is there a better way to analyze this data to extrapolate the % votes for which candidate or does this kinda work for what it is?
