A 119-Word Local Crime Brief Became Facebook’s Most-Shared Story of 2019

Facebook? Slate? AFAIU, advertisement margins don't scale linearly with popularity, so viral content probably brings in more money even if the total number and exposure time of advertisement impressions remains constant in the aggregate.

What? There has to be a 'most shared story', so something does have to win, unless you're going to get nitpicky about 2 stories that each had 49,994 shares or something.

It looks like the original headline was “Texas Rangers Search Here for Human Trafficking Suspect” which would have at least suggested that it was taking place somewhere in Texas. The radio news reporter that reposted it just changed it to “Suspected Human Trafficker, Child Predator May Be in Our Area."

> Newspaper stories used to have a “dateline” which gave the place-of-origin and the date the story was written, usually followed by the writer’s name.

Actually, the dateline is far more important than that. I didn't always know this, but the dateline is actually a contract telling you where the reporter was in person, and is the difference between the "hard" journalism of yesteryear and the "soft" web journalism of today.

Read this investigation [0] by the NY Times into the story of Jayson Blair, whom they fired after discovering he was faking sources for his articles. It talks a lot about journalistic integrity and was the first time I learned that datelines weren't just where the article happened but also the journalist giving you their word that they were there, on the ground.

Once you learn this, you start to see it everywhere. For instance, a recent article about the deep freeze we had here in Chicago (we hit around minus 50 degrees F with windchill for a couple of days) talked almost exclusively about Chicago, a little bit about the Midwest in general, and a single sentence towards the end about Michigan. The byline? Detroit, MI. Now you know why!

[0]: https://www.nytimes.com/2003/05/11/national/times-reporter-w...

I suppose a picture with a hundred fingers might come off as a bit creepy.

John Mulaney has a skit in his I'm New In Town special that I think sums this phenomenon up.

> there’s a hierarchy in the New York Post [...] different people that they don’t like. [...]And if you pay attention, you can start to identify some of the rankings that they have.

> [...]Um, the number one thing that you can be in the eyes of the New York Post is an angel. An angel is a child who has died. That is the best thing that you can be in the eyes of the New York Post.

> [...]Down towards the bottom of the spectrum, there are pervs. Pervs touch tots

People are drawn to news about "bad guys" and for some reason, sexual predators seem to be the most compelling kid of bad guy. Which likely also explains their popularity in crime drama plots.

There were some pretty interesting things in the sharing patterns, which probably warrant more analysis. (Short version: the post produced lots of comments, lots of users directly tagging one another, and a high "user share" to "site/brand share" ratio.) At a glance, all of those things imply that in addition to rewarding personal/organic discussion, Facebook is implicitly rewarding inflammatory, false, or satirical stories; scary crime stories and outrageous jokes are the sort of things that make people tag each other instead of just liking a post.

I'm irritated, though, that the article can't offer any support for any of those claims. I understand that Facebook isn't forthcoming about their algorithms, but there are several hints that this piece is trying to build a "scary evil social media" narrative rather than simply seeking answers. In particular, the line "While no sources I talked to suggested this..." looks like journalistic shorthand for "No matter how hard I tried, I couldn't get a sourced quote with which to claim that..."

This story is definitely more interesting than a simple clickbait headline, but I'm reluctant to trust the specific theories on offer here.

Well, if you actually read Slate's "privacy policy", it's pretty horrific too. It lists vast amounts of data they collect, both personal and (what they consider to be) non-personal (which can include your geolocation and browser fingerprint, apparently), and lots of ways they use it and share it; but it's not at all clear how I'd be able to opt out of having them track me, share my history with all their favourite partners, etc.

There's no way simply reading an article creates a "legitimate need" for all that.

They're being more honest. The Yahoo! security breaches combined with the takeover by Verizon mean that they're trying to transparently follow the letter of the law almost to a fault.

The security breaches meant that Yahoo felt they needed to regain trust. One method of building trust is transparency and honesty. Of course, transparency doesn't work well if you're exposing behavior that is inherently untrusthworthy, but like everybody else Yahoo's bread & butter is advertisement and tracking.

This is compounded by the fact that Verizon typically works in highly regulated, anti-competitive environments where you can do whatever you want without consequence provided you strictly follow the rules. This explains Yahoo's inability to appreciate why their transparency is ill conceived; why their disclosures are so matter of fact and tone deaf.

The proper way to run a media company is "see no evil, hear no evil". Give your engineers a mandate to collect everything and anything, but don't tell management and legal the details so there's plausible deniability when people figure things out. Disclosures should be vague because (1) legal doesn't know anyhow and (2) you're only doing the minimum necessary to avoid jail time and class actions--there's no regulatory agency to worry about and very little civil liability in general.