Analysis of 2016 GOP Governor Poll
Last week, I launched an online poll of declared and likely candidates for governor.
This was not a controlled survey, so be careful what you read into it. Instead of using the poll as a measure of breadth, I recommend looking at from a few other angles.
Suspicious Voting #
Yes, there was a lot of suspicious voting. At least one candidate received his or her strongest support from a single data center in India. Another candidate received over 50 votes from a cluster of IP addresses. That may or may not be cheating, but it’s suspicious.
Because this is a Missouri race, here’s the results filtered for votes from Missouri only:
I tried to prevent multiple votes through a variety of controls on IP address, cookies, and super cookies. But motivated people can get around all of those. (About 8,000 attempts to cast multiple votes were thwarted.)
Of 1,602 votes, I tossed out about 400 for suspicious patterns. Only 703 votes were verified from Missouri.
Geographic Distribution #
I also looked at geographic distribution of votes to see which candidates have made inroads into the widest geographic swath. I filtered for Missouri zip codes.
This zip code data might the most valuable to candidates. I was not surprised to see Brunner and Hanaway represented in many zip codes. Two surprises were Asbury, whose geographic breadth of support is impressive, and Kinder, who might have been held down by his late announcement.
Social Media Sophistication #
What this poll might demonstrate best is campaign sophistication and organization of social media. Randy Asbury wins this hands-down. Asbury used campaign websites, Twitter, and Facebook to encourage voting in the poll. Additionally, whether designed or not, his Tweets were timed for maximize impact and clickability. All other candidates have work to do if they’re to match Asbury’s lead in social media organization.
** John Brunner and Eric Greitens might have been handicapped a bit because they are still in exploratory phase. That limited their ability to get out the vote, or so I assume. ** I expected Catherine Hanaway to perform better. Her campaign has been active the longest and has plenty of money. While Hennessy’s View might not be daily reading of her crew, I would have expected better organization through social media. ** I’m got going to name the candidate who received a lot of suspicious votes, because I have no way of knowing whether the campaign itself conducted the operation.
** As a marketing strategist, I get a thrill seeing responses to polls like these. The top line numbers are pretty much meaningless, especially a year out. But analytics they provide could be valuable.
If you’d like to delve into the raw data, I’ve made it available via Google Drive.
This week’s poll asks your degree of support for Donald Trump.