Facebook’s Widely Viewed Content Report is strange | Hacker News Reader

Facebook’s Widely Viewed Content Report is strange

ethanzuckerman.com

162 pointsanigbrowl59 comments

The question I’ve been asking myself since skimming through the report earlier is this: Why should I take Facebook at their word for anything?

If the data is publicly available, then it can be verified. But until then, I don’t believe Facebook have the credibility to be taken seriously.

This. Facebook has a proven track record of saying whatever the hell people want to hear to feel good about Facebook. To say nothing of the fact that this report seems to omit ad views and those are occupying an increasingly large portion of peoples' feeds

What would they be hiding?

If they wanted to hide any data, why not simply not post it?

It’s not necessarily about the hiding as much as it controlling the narrative. If you publish reports but leave out one or two things here and there you can paint an entirely different picture. You don’t get to paint that picture if someone else is telling it.

"controlling the narrative" is absolutely the line of defense here.

The first couple paragraphs of the linked article lay out pretty clearly what they are hiding: the "facebooktop10" Twitter account publishes a list of the ten most shared posts every day, and it's always dominated by Dan Bongino, Ben Shapiro, and Tucker Carlson.

Facebook doesn't want to be seen as a platform for sharing bigotry and misinformation, so they are releasing this report to counter the data aggregated by facebooktop10.

That account posts the top 10 links shared and interacted with by (what it considers as) US pages which is quite different to the most seen by all users. As the report states pages (let alone just US pages) contribute less to what you see than either friends or groups.

I'd argue that pointing to what gets shared, rather than what reaches people and actually gets viewed, as proof that Facebook is a hotbed of right-wing content and that the idea they discriminate against it is a lie is actually outright misinformation. The thing Facebook's algorithms control is which widely-shared and widely-interacted content actually shows up in people's feeds, and if you ignore that you're ignoring Facebook as a company's entire role in influencing what their audience sees.

Pages can share whatever they want. I can create 5000 pages tomorrow to only share Colbert. That's something that Facebook has less control over than what gets seen, and is less relevant at that.

>The thing Facebook's algorithms control is which widely-shared and widely-interacted content

This disproves your point though. According to facebooktop10 most of what is shared is 'right-wing content' yet most of what people actually see isn't. So really, it does kind of look like they might 'discriminate' against right-wing content if they show it less despite being so widely shared.

Yes. That’s the point. They are publishing this list, because the other list makes them look like a cesspool of forwards-from-racist-grandpa and MAGA-chums.

Both lists are probably correct. (It’s not like—completely hypothetical—view counts of videos which are sort-of like the the double-slit experiment for Facebook.)

Which one is more meaningful? It probably wasn’t entirely accidental that they started with measuring engagement, and everything they (and others) do is intended to raise engagement.

The pivot to view counts is motivated only by their increasing fear of nor just increasing regulation. They are simply bleeding users, especially the most lucrative groups that are young, educated, and international, who are leaving the country club of social networks because they see these lists-even though their own feeds have maybe slowed down but not changed much otherwise.

Are they even losing users overall? This[0] suggests MAU is still increasing.

0. https://www.statista.com/statistics/264810/number-of-monthly...

By that measure, I haven’t changed my Facebook habits. But I’ve gone from an hour per day to checking it twice a month, as have many people I know.

BUT: I wouldn’t be surprised if there are difference between age groups, social classes, and regions of the globe, and that it may even be possible they are still growing their audience.

> As the report states pages (let alone just US pages) contribute less to what you see than either friends or groups.

Facebook wants to control the public narrative about Facebook, and the narrative that they want seen is that Facebook is a positive place to connect with friends and form communities and they're doing everything they can to keep it that way. There's enough data out there to call that narrative into question, and Facebook's doing what it can to limit "transparency" to a window that only shows what they want people to see.

IMHO, any report Facebook releases that supports their preferred narrative and disconfirms more critical ones is unbelievable unless accompanied by enough (verified) data that a skeptic can recreate their analysis and be satisfied. Otherwise, it's like deciding a trial based on only the defense's case.

> If they wanted to hide any data, why not simply not post it?

The article already answers this question. They're publishing data to make it look like they're transparent while not being transparent at all, i.e. "transparency theater."

> If they wanted to hide any data, why not simply not post it?

The article answered that question in the beginning. It seems like the real content that is most popular, based on other analyses, is right-wing talking heads and conspiracy theorists.

If you don’t believe summary statistics of the data, why would you believe the data?

If you think Facebook is lying in the report, why wouldn’t they lie if releasing the data?

It's one thing to fake or massage a summary, it's an entirely different thing to fake the whole data set. This divergence increases based on the amount of data.

Your question is essentially equivalent to: "If you don't believe what the politician said about the event, why would you believe the video recording of it?"

It's also easier for someone to figure out they are faking it if they actually released the data. Or at the very least say, "Some of this data seems off"

If the politician is holding the camera and editing the footage, it's a valid question.

Anyway you wouldn't need to fake a whole data set; you just need to employ a little bias in what data you choose to collect, how you collect it, how you process it and how you present it. Those things happen all the time. Or on an only slightly more extreme level, you could also selectively censor it using automation. People are used to thinking data is truth, but even the best data is always filtered through a human source.

Were the first point accurate, it would instantly become not so upon recognition of the existence of the concept of delegation.

The point about selectively choosing data, how to process it, etc is important, and often overlooked. People are accustomed to working with what they're given, but objectivity may be a step further back.

Regardless, such things can only be better revealed by providing the data.

If the goal is greater illumination, there is simply no argument to be made against greater transparency.

Yep, transparency leads to illumination, so in that way those two analogies work together like sunlight and window panes. But you can lie with data, was my singular point. More data is harder to fake, seems to be yours, and I suppose I would agree. But once faked, albeit at whatever difficulty, more fake data is more dangerous and less illuminating than less fake data. (Edit: Not only because there's more of it, but because it ironically has that very property of being or seeming more truthy or trustworthy because there's more of it.)

Anyway I have no idea what you're saying in your first sentence I gotta say. I recognize the existence of delegation, and yet still trust any party's data (and the completeness, honesty and transparency thereof) in direct proportion to some estimation of that party's general trustworthiness and whatever I know or can surmise about their aims, agendas and interests in relation to the subject of the data. And when the subject of the data is the very party collecting it, you can surmise immediately some of the probable interests and aims. They probably want to look good and not bad for example, or make more money and not less, etc.

Convention, basically.

Releasing fraudulent data is different, culturally and legally, to releasing a powerpoint-style report summarizing curated "key points," selected, defined and quantified in a non transparent way.

Every year, public companies release annual reports. Hundreds of pages of largely BS. A few dozen pages of real financial data.

It's not uncommon for a page 1 chart of the company's market share, provided by a 3rd party to be total nonsense. The page 114 figure summarizing tax liabilities needs to be auditable. That's part way to transparent. Not everyone can see the data, but someone can.

There are a lot more techniques to see if a large data set has been altered than a summary.

Whenever a company releases a report like this it should be treated as marketing copy. I'm sure it's "true" in the sense that the data matches the report, but the relevant metrics and methodology was chosen such that the result reflects well on Facebook - otherwise this would not have been released.

The first sentence of the overview gives away the game:

> Transparency is an important part of everything we do at Facebook.

This is a marketing statement, not a statement of fact.

> This is a marketing statement, not a statement of fact.

Precisely.

It's also the reason that all those cookie banners titled "We care about your privacy" - followed by 20 different dark patterns, defaulting to ON, and (il)"Legitimate purposes" - make me irate, and make me groan quite audibly.

Facebook is smart (and devious) enough to be "transparent" in a way that will not be truthful, since being truthful would confirm many of the beliefs that FB indeed shares vast amounts of misinformation while profiting from that sharing.

Maybe they are scrubbing data like AirBnB did for their New York reports, or more likely they are just controlling the metrics and scopes carefully to provide no useful, or at least no damaging, results.

They have been caught repeatedly publicly proclaiming one thing while internally discussing the exact opposite. It's like the tobacco industry all over again.

I wonder why Facebook is so afraid of transparency.

I’d like to start by saying that I actually like using Facebook in 2021. I’m Danish and I’m on Facebook solely because I play Blood Bowl and Facebook is where the majority of the community organise things. From how it operates it’s local version of the NAF rules to events for tournaments and leagues and just general chat about collecting outrageously little plastic toys. Before I started playing Blood Bowl I had quit Facebook.

While Facebook is still the most popular social media in Denmark, seeing 65% of us login daily, it’s trustworthy ness is at an all time low post the COVID lockdowns. I’m 2018 more than 50% of us would use Facebook for news, by the end of 2020 that had fallen to 5%.

Leading me back to me having quit Facebook, and into completely anecdotal speculation, but I quit Facebook because it didn’t have any content for me. No one in my network posts anything of real interest on Facebook and after I stopped following news media’s on Facebook and went directly to their sites instead (as well as subscribing to a physical news paper) Facebook was just boring. Until I started playing Blood Bowl. If I’m not alone in this, and only 5% of us now follow news sources on Facebook then wouldn’t that be an issue for them in the long run?

Their key advantage now that people have abandoned news is hobby and group content, but Facebook only has an advantage here because all of us are on Facebook. Except young people aren’t joining Facebook. In 2018 1/4 14-18 year old Dane joined Facebook, in 2020 that has fallen to 1/10. Sure Facebook is sort of safe from disruption, but within the miniature painting communities we’re already beginning to see more and more of people not having accounts or people only having accounts for one specific purpose. Less anecdotal than that, research shows that 67% of us will use phrases like “I’m only on Facebook because x” and that it’s no longer “cool” to be on Facebook because of its bad PR.

So why is Facebook afraid of transparency? Trust for SoMe in general is an all time low, but Facebook is actually one of the few networks sharing at least some of their data publicly. So why not exploit that advantage and go all the way? Especially when their image is already so terrible that people come up with excuses die being there?

probably because if scientists could do rigorous analysis of the kind of content that's shared and what effect it has on public discourse the results wouldn't look great. I think it's that straightforward. Same reason cigarette companies were not big fans of independent research.

Hey! Academic research on social media.

The big thing that reddit has done, is by letting people see data (to whatever extent), studies can be conducted that tell us what is actually going on.

Right now, when I talk to anyone dealing with trust and safety, the discussions start around the point in a workflow where a case is filed in court.

But the amount of clarity we have on whether certain forms of forums result in happier communities, whether they reduce polarization and so on - is remarkably small.

And, from my experience, its mostly in English and mostly focused on the “global North”. W

I used the reddit pushshift data set for my work, and realized that conducting sentiment analysis for content relating to India was … beyond the tooling I could find available. Code mixed language analysis is its own kettle of fish. A problem Ive been thinking about every since.

Now multiply these limitations to under resourced communities around the world and its concerning.

As I see it, there’s tons of law related work being done in this space.

But it seems the data and code analysis of it is mostly held under an NDA. And to illuminate a future painful discussion - its held by companies in the US, a geo political fault line in the making. What happens to Puerto Rico? Kenya? Morocco?

This is the one thing that tech can still do, which is to build the tools that give us an objective view of how people actually behave online, and make that information public and a common good.

> without releasing any of the data a researcher would need to answer a question like “Is extreme right-wing content disproportionately popular on Facebook?”

I'm surprised to see The Epoch Times not mentioned in the article, right there as the #10 most popular link (between a fashion show and some local story about a 6yo missing). The only news site I'd call "traditional" media in that list is ABC News in #17.

https://www.wired.com/story/facebook-transparency-biggest-si...

"The answer: memes. From his personal account, which has more than 120,000 followers, Jacke posts a steady stream of low-rent viral memes that have nothing to do with the Packers, adding the URL of his business to the top of the post. We’re talking the likes of “Pick one cookie variety to live without,” or “Give yourself a point for each of these that you’ve done.” "

Perhaps a more interesting quote -

"It’s remarkable that the data Facebook chose to publish"

It’s just a more advanced version of those MySpace quizzes from 2007. Same thing, just in a shiny React app.

Those quizzes/surveys are making a comeback in some demographics on facebook, too.

https://www.facebook.com/chris.jacke.13/ for the curious.

I got fairly heavily downvoted last week for trying to point out that Facebook's "Suggested for you" stories were almost certainly paid content via some path (even if not "advertising"), given that they're uniformly awful clickbait garbage that Facebook knows I don't interact with. Folks really didn't want to hear that theory and insisted it was just random clustering or whatever.

Arguments like the linked article make it pretty clear that that's not the case. Facebook is either being scammed badly by players exploiting their algorithm to throw clickbait spam in front of their users...

... or they're on the take. I think it's pretty clear Facebook is too smart to be scammed like that.

It happens on instagram too of course. Start scrolling then its just suggested reels and posts are all major accounts with with thousands of followers which comes with some commercial angle.

Not sure if it is just me or if it gets harder and harder to curate my IG "recommendations".

I get presented an unreasonable amount of crap every day and always mark this stuff as uninteresting. Nonetheless I often see that same s*t again the next time I open the app.

I stopped FB and for IG I nearly only use my freelancer profile to keep in contact with some parts of the networks I am in. And to put out (hopefully interesting) material for my network of small organic food producers and farmers.

About half of my 'Activity Feed' is "unfollowing" this attention-grabbing crap. It's not even really clickbait, just a video to stare at for 45 seconds. I think Facebook just uses these to pump their time-spent-on-site metrics.

This week I've "unfollowed": Viral Hog, Crafty Panda, Crafty Panda How, Kids Crafts, 5 Minute Crafts, America's Funniest Home Videos, LAD Bible, and LAD Bible Australia.

I put unfollowed in quotes because that's what facebook calls it when I click "Stop seeing posts from this page" on suggested content.

It's amazing people keep using Facebook amidst such torrents of bullshit. Why do you?

tbh I Re-joined after a year because I needed to get rid of a bunch of furniture in a hurry and Facebook marketplace has a lot more traffic than Craigslist.

Now moving to a new city, the neighborhood classifieds and meet up groups are pretty compelling

So… they are selling ads, and correctly labeling them, and we’ve all seen that system, including the backend..

But then there’s another, shadowy market, where they sell far cheaper ads to really lowbrow content that people are more likely to click on, but that just doesn’t have an obvious way t monetize these eyeballs?

Getting less money, diminishing their platform with bad content, breaking laws by not disclosing it, and keeping it all secret?

What I’m saying is there may have been a reason why you were downvoted.

"Uniformly awful clickbait garbage" has its own kind of potency. When it's trying to pad your feed, Facebook might not know that "give yourself one point for each one of these you've done" has any specific relevance to you, but it does know that a lot of people seem to engage with it and you're a person. It's basically guessing when it's out of ideas. To infer any kind of malice, you'd have to define some sort of threshold for how much "uniformly awful clickbait garbage" you would expect to see just because of this guessing, and observe that the threshold is being exceeded. Have you done that?

US targeted ads on facebook cost a lot. Cannot imagine advertisers making a positive ROI , especially after accounting for fake engagement,but I guess some do

Jury is still out on whether advertising makes a difference at all, more so targeted ones.

It really does seem like a song and dance that doesn't do anything at all. Maybe the real correlation isn't ad spend to profit, but that companies that are able to spend a lot on ads just tend to be ones that can turn a profit on their products anyhow no matter if they spend on these ads or not.

Reply on news.ycombinator.com