There was lots of interesting discussion on twitter and in the comments of this post about the rather obvious poll stuffing by both Place 1 campaigns over the last few days. That discussion has also spread to Harold Cook's Letters from Texas where he humorously offers possibly the best take on the situation.
So my poll, identical to BOR's, is on top of the right hand sidebar. Bend it 'til it breaks. Knock yourself out. Let 'er rip, tater chip. The one with the most votes by Friday afternoon may or may not win the race, but they'll win the "my supporters can figure out how to vote more than your supporters in a meaningless blog poll" contest. Make the meaningless mean something.
That's pretty much what the poll we had on Burnt Orange Report has been reduced to. Let me take a brief moment to walk through some slight differences in how we approach our polls here.
First off, by no means are our online polls scientific nor have we argued that they were. The ones you find on our site are meant to get the pulse of the BOR Community, the regular readers and activists who take the time to bother to register as users here so that they can contribute to the discussion in our comments and in adding their own posts. Given that there are close to 5000 registered users on our site and that most polls are lucky to gather a couple hundred votes that's about 5% of the userbase weighing in.
Often times we are a bit adrift for what to run polls on, but they are useful as a tool for the staff to see where the readership is various races. It helps to get a quick check of whether our own thinking is in line or out of line with that of our readership which is useful in directing some of our content decisions. For instance, a poll last fall asking when our users planned to vote showed that indeed, our readers are a bunch of crazy vote-a-matrons. A recent one about the Austin Mayor's race confirmed our feelings that our readership is largely for Lee Leffingwell.
There are only a couple reasons that sites like ours use polls.
- To get a read of the readership. Closed registered voting helps protect this to a degree.
- To generate user accounts & community. Registered users are more inclined to comment and author new content, which increases community and pageviews.
- To list build. We could use 3rd party polling services to run controversial polls in order to capture email addresses.
- To drive new traffic. Like Harold Cook's approach, set up the poll in order to specifically draw out poll stuffing to drive pageviews (and/or ad revenue).
Bunrnt Orange Report's site polls are centered primarily around purpose #1 with a very auxiliary benefit of #2-4. We should probably think about doing some in the future that are geared toward the other goals but that's a discussion for the next generation of BOR which I am now actively working on behind the scenes here.
So on that note, we were truly interested last week to get a sense of what the exit of Rick Cofer did to the Place 1 race. I wish we had run one prior to his departure so we could have gotten a sense of where his active supporters went to (and that might be the poll I replace the current one with). In any case, both us, and I think the general readership would have been interested in this.
Most polls inspire a certain number of people who are regular readers to finally create a user account and vote. That's fine, expected, and doesn't alter the polls to any large extent.
But now we have no idea because first someone sent a facebook message to 350 Perla Cavazos supporters in her group urging them to vote in the poll and then some subset of Chris Riley supporters got an email pointing them to the poll to vote as a response to Perla's lead in it.
On Friday and Saturday, BOR reported the highest percentage of new registered users as compared to daily site traffic... EVER. Each day, 1.5% of all visitors created user accounts which is a lot when the average is about a tenth of that. About 100 new user accounts since the poll went up when normal traffic would have generated less than 10.
Now, to be fair, I recognize a number of the emails and people. This isn't fake email address vote stuffing like the Watts campaign used on the Democracy for Texas U.S. Senate endorsement poll in 2007. I recognize a number of the people and actually believe that a decent portion of them are aware of Burnt Orange Report and might even be semi-regular readers who hadn't registered before. If those 100 users start getting involved and active in our comments or in offering up new content in writing Journals of their own, awesome! I hope that happens.
But it still doesn't change the fact that it's now altered any ability for our regular readers or the staff to glean anything from the poll (as well as altering the short term 'electorate' so that we can't just re-run the poll). That sucks and I'm aware that there is nothing that prevented this from happening and to a certain degree, it's a sign that we've effectively engaged and made campaigns aware of this space.
I recognize that is a double edged sword and I can't expect the "usual" to simply keep occurring if I keep writing about technology use in campaigns which is intended to actively encourage new behavior. That's a failure that could only be prevented by altering the poll process somehow, like having a couple day waiting period for new users to vote in a poll, which while technically possibly, would require a change in the Soapblox base code. And right now, I'm more concerned with working to keep the site up, working, and running with Soapblox until we migrate to a new server so the site stops going down randomly.
So I guess that's all I have to say on the matter. Other than to tell y'all to go vote in Harold's poll now.
P.S. I might not need a poll for this but I'll ask. I'm looking to get a new USB drive since they are so cheap on Amazon and the one I have right now is just 250MB. Should I get this SanDisk 16GB or this Kingston 8GB one?
|
