In recent years, the spread of fake news stories (designed to drive traffic to fake news-generating sites) by inducing social media users to share links has been increasing in a manner that appears almost exponential. (We have an entire category here devoted to the most infectious fabricated news stories.) The old adage about a lie traveling the world faster than the truth can get its boots on seems more apt than ever, and it often seems temptation to share

intriguing claims can outweigh user incentive to determine whether a particularly interesting story is indeed cut from whole cloth.

Most of the time, fake news stories come and go (presumably with some of the fleeced never the wiser for having believed the tale). In some cases self-styled “satirical” sites have caused localized panic or harmed businesses, but mostly fake stories tend to annoy more often than they truly inconvenience.

Over time, Facebook has tracked the pattern and spread of such content on that social network. In a note published by Facebook Data Science on 29 April 2014 titled “The strange truth about fiction,” Snopes.com was referenced by the authors of a corresponding study as a reference point in the overall movement of misinformation on social media platforms:



Tracking rumors on Facebook requires two types of information: a corpus of known rumors, and a sample of reshare cascades circulating on Facebook which can be matched to the corpus. The website Snopes.com has diligently documented thousands of rumors, and provides the starting point for our analysis. To match known rumors to this anonymized set of reshare cascades, we identify uploads and reshares that have been snoped — someone linked to a Snopes.com article in a comment. Those comments are posted by people to either warn their friends that something they posted is inaccurate or to the contrary, to validate that a rumor, though hard to believe, is in fact true.

That research led to a number of interesting discoveries about the overall impact of rumors and false claims on social media sites, among them that users are more than four times more likely to delete an untrue or inaccurate post when they are “snoped.” Later that year Facebook stated some posts would be labeled as “satire” to inhibit the spread of misinformation, but details of that initiative were not fully disclosed.

On 20 January 2015, Facebook issue a press release titled “News Feed FYI: Showing Fewer Hoaxes” in which the social network announced a more direct approach to false information and users’ News Feeds:



We’ve heard from people that they want to see fewer stories that are hoaxes, or misleading news. Today’s update to News Feed reduces the distribution of posts that people have reported as hoaxes and adds an annotation to posts that have received many of these types of reports to warn others on Facebook. We are not removing stories people report as false and we are not reviewing content and making a determination on its accuracy.

The note defined what constituted a hoax for the purposes of the announcement and referenced the earlier findings about deletion of rumors and false information:



Hoaxes are a form of News Feed spam that includes scams (“Click here to win a lifetime supply of coffee”), or deliberately false or misleading news stories (“Man sees dinosaur on hike in Utah”). People often share these hoaxes and later decide to delete their original posts after they realize they have been tricked. These types of posts also tend to receive lots of comments from friends letting people know this is a hoax, and comments containing links to hoax-busting websites. In fact, our testing found people are two times more likely to delete these types of posts after receiving such a comment from a friend.

According to Facebook, more direct or humor-based satire content won’t be affected by the update: “publishers who are frequently posting hoaxes and scams will see their distribution decrease,” but content deemed humorous by users won’t be affected by the change. Links flagged for containing false information might be appended with a warning to users attempting to share the content.

Last updated:   20 January 2015