|'Rollercoaster' sheriff poll|
|The Press - Opinion|
|Written by Joel Addington|
|Thursday, 01 November 2012 14:27|
Among the many kernels of wisdom fed to young journalists is this — don’t become the story.
As the theory goes, when members of the media go from being the ones gathering the news to the people in the news, they inevitably lose their objective perspective because they are now inside the fish bowl instead of outside. And as such, they can no longer be trusted with a story about themselves.
This week — thanks to complaints about our online poll, how our poll functions and the general hysteria that accompanies all elections, and worsens as the election nears — The Press came close to becoming the story.
I don’t mean in the literal sense that would imply there was a chance that other news organizations (or us for that matter) would be reporting on what we do here from our Macclenny office. But rather in the sense that our actions caused what could be described as “news.”
Let me explain.
As you can see on the front page, this week’s online poll shows that incumbent Democrat Joey Dobson led his Republican challenger Cameron Coward 55 percent to 43.8 percent, with 1.2 percent of respondents indicating they were undecided about who they’re supporting for sheriff.
As of press time, the poll tallied 685 votes. Admittedly, three of the undecided votes were cast by yours truly while testing the poll’s settings early this week. Over the weekend, we received complaints via our Facebook page about people not being able to cast votes.
“I’ve had four friends call me saying they can’t vote ... They said the site says, ‘You have already voted.’ Is this poll rigged?!” wrote George Doran, a supporter of Mr. Coward’s.
That comment was followed by this one from Kathleen Johnson: “I tried to vote and couldn’t ... gawd ... please don’t tell me our local paper is rigged!”The poll was certainly not rigged to favor either candidate.
But there are settings that determine how the poll works. One setting we generally keep in place, and the setting that remained in place until Monday afternoon, prohibits multiple votes from the same IP address.
Typically, each computer connected to the Internet has a unique IP address, so the restriction is meant to limit each computer to one vote in the poll. But, sometimes, multiple computers on the same network can share the same IP address.
The effect is that after one person votes from one computer on a shared network, then nobody else using a computer on the network can cast a second vote.
While this system of polling does not ensure “one person, one vote,” and lets one person vote from multiple computers, I believe it’s still a good tool for measuring public opinion.
For instance, a person semi-interested in whatever the poll topic happens to be will likely vote only once and be interested to see the overall results in the next edition of the newspaper.
A person who is really interested in the poll topic, say when it’s about who has more support in the race for sheriff, well, they may vote multiple times: once at work, once on a home computer and once on a cell phone.
To me, voting more than once simply shows that some voters care more than others about the poll’s result. And because everybody can do it, the playing field is largely level, so to speak, and the poll remains a fairly accurate gauge of opinion, and arguably, enthusiasm, too.
That said, on Monday, October 29, after fielding more complaints, this time from staffers at the Baker County Sheriff’s Office saying they couldn’t cast votes at work, we had our tech guy look into it. I didn’t find out until later that the sheriff’s office computers all share the same IP address.
Our tech guy addressed the complaints by changing the settings to allow multiple votes per IP address and sent us a message explaining that Internet service providers like NEFCOM could have customers using shared IP addresses, thereby blocking groups of customers from voting if somebody with the same IP address had already cast a vote.
We soon posted to Facebook that the poll problems were addressed and everybody should be able to vote, and sent the same message to the sheriff’s office.
And boy, was that a mistake. We neglected to fully understand how wacky local politics can get eight days before a big election. Forgive us, it’s just another week here at The Press.
After some research, here’s what I determined happened.
Almost immediately after the setting was changed, and for about an hour following the change, 290 votes came pouring into the poll, many from the same IP addresses.
During the previous three days, the poll logged only 223 votes. That bears repeating: for one hour when the IP restriction was turned off, we got 290 votes. The Friday, Saturday and Sunday before the change, we received 223 votes.
Even more interesting is where the flood of votes came from.
Further analysis showed that 90 percent of them carried the sheriff’s department IP address. And of those 261 votes from the headquarters of BCSO, 65 percent were cast for Mr. Coward and 35 percent for Sheriff Dobson.
As the votes for Mr. Coward climbed, the sheriff took notice. He called our office and said his staff had figured out how to “hack” the poll and vote repeatedly, and since it’s been proven unreliable, we should take it down.
Sheriff Dobson posited that Mr. Coward likely had somebody casting vote-after-vote for his opponent, skewing the results. And at the time, it seemed plausible. After all, the sheriff was leading the polling 62 percent to 38 percent through the first three days and the gap was closing.
But my response was basically a shoulder shrug. There was no way we would pull down the poll, especially not at the request of an elected official. First, what would we put in that little space on the front page where reader’s expect to see the poll? And why does the sheriff think we would even consider letting him set our agenda?
I explained that if Mr. Coward’s got supporters willing to waste their day away clicking on our poll, then they must be pretty dedicated. And just as easily, the sheriff’s supporters could be doing the same. The sheriff agreed, saying he could make up ground in the poll by having his employees sit at their computers and vote over and over.
Fair is fair, right?
Well, it wasn’t long after that we realized what was happening Monday afternoon, and we reversed the poll setting back to one vote per IP address.
And guessed what happened? Voting slowed down considerably, and through Tuesday evening , Sheriff Dobson began getting three votes for every one vote of his opponent’s.
So rest easy sheriff, your reign appears intact.