How we find stories
Lead Stories mainly uses the Trendolizer™ engine to monitor the internet and look for newly trending content. We have created several lists of known prank, satire or straight up fake news websites that we specifically keep an eye on with Trendolizer, along with lists of automatically collected stories tagged with fake news related hashtags on social media (#hoax, #fakenews, #fake...). You can read more about how that works in this article we wrote last year.
(screenshot from our Trendolizer™ dashboard)
On a typical day we will look for the most trending stories from these lists and pick the ones that meet following criteria:
- Make a claim that can be objectively checked (i.e. no opinions or merely emotionally phrased repackaging of events from mainstream news like "Video: X Destroys Y In Speech! This Will END Y!").
- Haven't been debunked or checked yet by another reputable site.
- Are at least somewhat trending (no point debunking something if almost nobody saw it in the first place) or are likely to become trending based on our experience (for example stories from sites that often have their content copied by networks of fake news websites).
- Are likely to have an emotional impact on readers.
Generally we will prioritize stories that are most trending or which are very new and which are gaining traction fastest. We might occasionally also tackle a story that has been debunked by another site already if we have more/better information. And if we receive a tip from a reader or spot an obvious hoax making the rounds on social media we might also decide to fact check and/or debunk it if it meets the above criteria.
Our main goal is to quickly provide articles debunking trending fake stories that haven't been debunked elsewhere yet. This means we usually don't write articles to confirm something is indeed true, most of the time we will be pointing out things that are false.
Each story is different but these are the questions we try to answer before rendering a verdict and writing an article:
- Is this version of the story the original? Or was it taken from some other site, and if so, which one and when?
- Once the original source has been established: how reliable it has been in the past? Is it a legitimate news site? Are there satire disclaimers? Is the author identifiable? Is there contact information? Have previous stories from this source been debunked/checked by ourselves or other sites?
- If there are pictures or video in the story: have they been used before elsewhere? Have they been manipulated? Can they be verified?
- Do the people, places and quotes in the story really exist?
- Is there evidence the events in the story didn't take place?
Depending on the answers to these questions we will label a story "False", "Mostly False", "Unproven", "Mostly True" or "True" after laying out all the evidence for our readers. We may also opt to label a story as "Satire" in accordance with our satire policy.
Headline and supporting image policy
If we deem a story not to be true we will generally publish our article with the same headline as the original article but prefixed with "Fake News:" and with the addition of (partially) capitalized words that invert the meaning. For example if a false story is titled "This Is An Apple" our headline will read "Fake News: This Is NOT An Apple". The goal of this method is to stand out in search results when people who doubt the original look up the title through search engines. Some fact checking sites prefer to rephrase false claims as questions in their headlines but we feel this makes the false headlines look like they are answering the question in the affirmative. We believe that if search results for the query "This is an apple" look like this our headline stands out much better even if it is only in fifth place:
- This Is An Apple
- This Is An Apple!
- This is an apple - Confirmed!
- Is This An Apple?
- Fake News: This Is NOT An Apple
- Breaking: This Is An Apple
Generally the thumbnail images added to stories debunking articles will be partial screenshots of the originals with bright red words or short phrases superimposed on them that indicate our judgement. Examples of such words or phrases are "Fake!", "Prank", "Unconfirmed", "Satire!", "Not true"... This stops readers from accidentally believing the claims in the screenshot if they ever see it in isolation and also gets our fact check into image based search engines like Google images.
If new facts come to light after publication we might alter the article further according to our corrections policy.