Online reader polls may be a thing of the past for The Australian after the Holt Street 2.0 heavies conceded it was likely the results of two recent polls run on its website were hacked.
Eyebrows were raised when a poll on The Oz website suggested on Wednesday there had been 70,000 votes (more than 97% of respondents) against the use of taxpayer funds to fly relatives to Sydney for an asylum seeker funeral. By Thursday morning, the poll had been updated to record more than 100,000 votes against the decision.
A couple of Crikey commenters questioned the high set of figures — three times more than a similar poll on The Sydney Morning Herald website — and the overwhelming flow of votes in one direction. And they pointed to another poll run on the website which recorded more than 263,000 people (99.17% in the affirmative) voting on Julie Bishop’s position as deputy leader of the Coalition.
The average amount of votes to be recorded on Australian polls over the last week (excluding the asylum seeker funerals and Bishop outliers) was 5627.
Nic Hopkins, online editor of The Australian, concedes it’s likely the two lopsided polls have been manipulated.
“We removed the polls from the homepage when we became aware of the results and are in the process of removing them from the system altogether,” he told Crikey. “It isn’t the first time this has happened and shows how easily they can be distorted.”
According to Nielsen figures, this week the average amount of daily unique visitors to The Australian’s homepage is 209,000. In other words, on face value half of all unique browser voted on the asylum seeker poll — an extraordinary amount of interaction between website and audience which any media boss would be rapt to see.
Get Crikey FREE to your inbox every weekday morning with the Crikey Worm.
Editor-in-chief Chris Mitchell was also miffed by the results, warning Crikey not to read too much into them: “These are not scientific polls like Newspoll. I would just as soon we never ran them online and I pay them no attention.”
Hopkins backed Mitchell’s assessment, noting the easily manipulated online reader poll may be on its way out: “Chris is right in that he has tolerated us using this type of polling on our site as a low value way of building engagement. But it’s not at all scientific and I suspect his patience with them may have just evaporated.”
Each day The Australian runs a reader poll on the hot button issue of the day. In one poll, 68% of 15,000 voters said Tony Abbott’s “shit happens” remark was appropriate, while another recorded 75% approval (7079 votes) for Abbott’s budget cuts over Julia Gillard’s proposed flood levy. Back in December, in the wake of the Christmas Island boat tragedy, The Australia ran a story off the back of poll results on News Ltd websites saying Australians were “implacably opposed to allowing asylum seekers into the country”.
So how did so many votes get shunted in one direction in the two manipulated polls? One obvious explanation — unless the votes are legit, of course — is there has been a concerted effort by a group to hack the poll.
There are a number of ways a poll can be rigged. Typically an online poll won’t let you vote twice, but nothing can stop a reader clearing their browser’s cookies or using proxies to mask a user’s URL — allowing a reader to vote to their heart’s content. In the US this method has been referred to as “freeping”, a term used to describe readers of conservative website Free Republic who hijack online polls for their own political gain.
Over at News Junkie Post they give multiple examples of political polls in the US being freeped. In one particular example a couple of years ago, 78% of votes cast (863,000) in an NPR poll were in favour of supporting Fox News over Barack Obama.