Human Web CollectiveWhat PieFed's "social credit system" actually does
Snoopyjoin.piefed.social/2026/08/06/what-piefeds-soci…
Just like Reddit, PieFed keeps track of how many upvotes and downvotes your posts receive and the net value of all the votes received is called your “reputation” (Reddit calls it “karma”). This is not shown in the UI anywhere (except to administrators), to dissuade the karma-farming behavior common on Reddit. Reputation is not a game or status symbol, it is a mechanic for enhancing trust and safety on PieFed.
Some people have criticized PieFed for this, calling it a “social credit system” similar to the one in China so I thought it’d be good to explain exactly what it does so people don’t need to guess or rely on hearsay.
36 Comments
Comments from other communities
Appreciate you taking the time to articulate your thoughts on different features over the past week or so.
I’m open to hearing other people’s critiques on the other aspects (if they have any), but everything makes sense, to me, until the “Low quality communities”. I recall this “feature” being a point of contention in the past.
You said its not as effective as you thought it would be and you don’t use it much, is it an opt-in feature? If not, can it be?
Alternatively, can admins make, alter, or add their own list?
As far as I know it’s a custom list. Admins will manually go into the admin settings, communities, pick one and edit… And set the checkbox for low quality to flag it on their particular instance.
Appreciate the explanation. That seems like a reasonable approach to me. Might be worthwhile having that info be transparent to the users.
Yeah. I really appreciate how Rimu is doing the blog posts to explain what PieFed does. And why. I think traditionally, we’ve been more one of those projects with a minimum of documentation. Half of it is self-explanatory, and you’d need to figure out the rest of the settings and intricacies by skimming the source code and old chats. We got some people volunteer to add a bit to the documentation in the past. Other times it’s like it always is with Free Software developers. They rather add 5 new features than write all the paragraphs to explain what the already existing features do. ¯_(ツ)_/¯
…Just my personal opinion…
(In case someone is interested in contributing to more transparency/documentation: https://join.piefed.social/get-involved/ )
HarkMahlbergMaybe the issue is the name? “Low Quality” makes it sound like you’d rather not have the community at all, but reading the description it sounds like the appropriate targets are Meme communities. “OkBuddy***” “***Circlejerk,” etc.
It’s not likely you’d often see productive or interesting conversations there, where rep would matter. They’re mostly ha ha funny image posts. It doesn’t make them useless, it just means they’re disposable.
I don’t particularly care for memes and as a general rule I would agree with you, but, for example, history memes with pugjesus had some good and informative posts, comments, discussions.
My potential issue is related to dictating what is considered “worthwhile” as an absolute, from a single unalterable source. If every admin can make/alter/delete list(s) and the list(s) is transparent, I don’t see an issue with it.
Oh yeah fair point, PugJesus is pretty chill. It’s a blurry line to draw where memes end and srs bsns begins.
nymnympseudonym
Kierunkowy74@kbin.social(…)history memes with pugjesus had some good and informative posts, comments, discussions.
At least PieFed has thought already about it: (from OP’s blogpost)
In some communities, which the instance admin designates as ‘low quality’, upvotes on posts do not increase reputation (upvotes on comments on those posts still do, though).
had some good and informative posts, comments, discussions
It was also heavily curated and controlled. You likely didn’t see any of the comments pointing out omissions and inaccuracies since they’d be hidden and the users shadowbanned.
Which is part of the reason keywords on an instance level can be a poor way to control and curate content. It’s just going to miss those who know how to game and bypass these systems and use it as a weapon as part of some kind of moderation arms race.
Call them “Shitpost” communities and call it a day.
I was about to type basically the same thing. There has to be a more professional word than “shitpost” though, surely.
Well, it is certainly a social credit system, but we need to learn to not just think China when we think about it. the US credit rating system, most countries individual court files, all those are social credit systems, but some rules make some of them useful and some of them are tools of oppression.
Slashdot had an interesting system to dissuade karma farming. They would be capping karma at 50.
And mostly the only impact it had would be that you would have your publication start at minus one when you had negative karma. And above 10, you would get some moderation point, if I recall correctly (in slashdot, you don’t get inifinite votes automatically)
The devs considered that most users who were contributing honestly would end up between 40 and 50 pretty quickly, and that it was okay to be downvoted a bit from time to time.
I’ve mentioned this before in other context, but I feel like there’s value in potentially looking at who upvotes who, and following that graph.
If you have like a closed circle all upvoting each other, but everyone outside the circle downvotes, that might be useful information.
… as part of an anti-bot add-on which would presumably also consider things like IP address and posting times
I think it would be helpful for more than just bots. Right now every upvote is equal, but I probably personally value the upvotes of people I like more. You might see a 5 star review online, but if four of your friends all say it sucks that counts for more.
Maybe activity would also be an authenticity indicator. For example, I’ve noticed that some threads will get consistent votes from accounts that have been inactive for months or even years.
CombatWombatIt’s interesting because I encounter the red triangles fairly often, but generally have good interactions with those accounts. Everyone in the wnba community has them, but my interactions there have been fine and reasonable so far, presumably because the poor reputation is from interactions in other communities — someone could be very confrontational about politics, but amiable about basketball, for instance. I suspect there’s a reasonable chance I have the red triangles as well, because I do a lot of posting in communities that I suspect have been flagged as “low quality” because of their low subscriber count relative to the frequency of posts. I get what you’re going for, and I think it’s a good direction to be heading, but it feels like there’s a little more tinkering to be done before it’s working as intended.
RimuI just had a look at that community.
Those people are all Donald J Musk’s alts and that guy is universally hated.
Working as intended in that case.
I suspect I can infer from this I should create a separate wnba community on a separate instance, then, rather than continuing to post there?
Yes all his communities will never go anywhere, he’s banned from most instances.
I understand your meaning more broadly, but thinking narrowly about the mechanics of blocking and federation, does a community’s owner or moderator being banned from an instance effect how posts to that community from other actors are disseminated across the fediverse?
No, it doesn’t stop posts in their communities (except their posts in their communities). It might stop their moderation actions from being accepted though, which effectively makes the community unmoderated and a liability.
This seems like a bug. Probably when someone is banned all the communities they’re the sole active moderator of should be locked. Or a warning shown to the admins, or something.
Thanks for explaining the mechanics. I’m kinda glad you’re thinking about federating moderation actions for blocked accounts instead of me cuz that sounds complicated and difficult to get right.
Btw papasqat is another one of his alts and plays games like this https://piefed.social/c/modabuse@lemmy.sdf.org/p/2265140/i-got-banned-from-seven-pointless-slop-subreddits-that-i-d-never-heard-of-and-want-noth
Lemmy.today is silenced now, which will stop people from stumbling into their Comms.
🇰 🌀 🇱 🇦 🇳 🇦 🇰 🇮I knew that guy’s behaviour seemed familiar, somehow. Just couldn’t place why.
similar to the one in China
Please don’t spread propaganda. This isn’t a real thing.
https://en.wikipedia.org/wiki/Social_credit_system#For_individuals
Educate yourself, comrade.
I’ve lived in China, done field work in China, and I’m a PhD candidate on China. You educate yourself beyond a one second search. Yet another reason to avoid online discourse with bozos like yourself.
Wikipedia says it, so it must be true.
Edit: Actually, read your own link
https://en.wikipedia.org/wiki/Social_credit_system#Misconceptions
This is the AI brain that is making profs retire left and right.
I’ve lived in France most of my life and I am still discovering some of the fucked up things that the French government does.
Do not assume you know China because you live there.
Actually, it is usually easier to know the evil side of the country you live in from the outside.
Did you not read past the first sentence fragment? Fieldwork and PhD candidate on China. I didn’t assume. You did.
Yes I should have said “similar to what people in the west think China has” but that’s not the focus of the blog post. Instead I just linked to the wikipedia article which is pretty clear about that.
What is a “low-quality community”,.something arbitrarily decided by the instance admin?
JohnnyEnzymeI imagine those to be ones with some combination of low-effort posts, low-effort comments, low-effort community design, and in general, absentee moderation & leadership, and for reasons like that, unappealing to almost all audiences. A good example might be some place created by someone that only makes sense to themselves, and is generally useless.
All that’s just my own spitballing of course, but no– marking it as such would be anything but arbitrary. Just look at the thought and care behind the implementation and purposes of the social credit system, as detailed in the blog, to get a sense of how that might work.
katy ✨the best thing to come from the whole tesseract is that everyone else is more open about the processes love this :)
Ek-Hou-Van-BraaiMaybe a bit off topic but:
The Truth About China’s Social Credit System
https://www.reddit.com/r/socialism/comments/1tlx7n0/the_truth_about_chinas_social_credit_system/
tl;dw
China doesn’t really have a Social Credit System, but the US does.
Are you talking about the attitude rating on your profile? If so I can see that and others too.
No that’s how much you upvote vs downvote. Your “positive attitude”
As opposed to how people vote on your posts
And just a note on that– if I understand correctly, a 0% attitude means that you’re upvoting as much as you’re downvoting, a 50% attitude means a 2:1 ratio, and 100% means you’re only upvoting. Correct me if I’m wrong on that, but I think that’s in the ballpark.
Also, the sampled votes to produce that score don’t go back too far. Maybe just the last 100 votes? Because you can pretty quickly alter that score significantly within only a day or two, is my point.
Thanks for reminding me, I’ve added a comment to the bottom of the blog post about that.
I should keep track of these blogs more. My opinion is a bit limited because I don’t care for public voting and don’t peruse by top or hot. I was peripherally aware of the vote limit thing from peoples posts in communities but then comically did not connect it to issues I had with voting which I only did to clear my feed. Once I realized I asked you if a neutral option could be made and you graciously got that going like way faster than I expected. Thank you again. So any issue I had with it is gone now. Again I just don’t get why people want to see things based on what other people feel are important things. Granted much of that is due to the limited size of the fediverse as it is now. If this became the size of reddit then I would likely have to rely on public voting although I would prefer the rank everything as a user approach.
tuckermto dissuade the karma-farming behavior common on Reddit. Reputation is not a game or status symbol, it is a mechanic for enhancing trust and safety on PieFed.
Thank you! This is the kind of thinking that has made me realize that I can never go back to big, profit-driven social media websites.
The profitable decision would be to promote karma-farming, since karma-farming boosts engagement. Thanks for choosing the “respect your users’ sanity” decision instead.
Looked back for this to bring up some things I have noticed with the new read. Im like 50/50 on reverting back to upvoting due to muscle memory but I think also a bit because up and down voting is so convenient on the left side and my brain sees it and goes into pattern. It takes a bit more to find and use the read button. Also the delete icon does not scream read. just the word read might be better. What I really wanted to mention though that if poor quality voting is an issue then why have it available in the feed. I mean shouldn’t people vote when they have clicked in and read the post. Granted I get media posts show in feed fully so it makes a little senese there but if they really like or hate it so much is it to much to ask to click into it? After all you have to do that to comment no.
I think the attitude system is insane when the fediverse is filled with so many people who deserve every downvote they get.
I think that there is a lot more content worthy of upvotes than there is worthy of downvotes in the fediverse.
Both things are true.
Shit. What do I do now.
One day altruism and hope will fail you and you will become a jaded curmudgeon like me.
i read your comment with the voice in w40k mechanicus trailer
On the flip side, the fediverse tends to have a pretty strong monoculture. You can collect down votes just by disagreeing with someone respectfully or stating your view point here.
Maybe it it’s numbers grow, it will become a bit less of an echo chamber.
I found that out the fun way by having the unpopular opinion of “we shouldn’t encourage 4chan-style doxxing threats over server drama”. The admin of lemmy.zip banned me from the server, presumably because it was somebody from that server pushing for that behavior.
Yes although if you’re only here to downvote bad content then perhaps your focus is in the wrong place.
I introduced some friends to piefed, and told them about the attitude system. They said they only downvote because remembering to upvote is too hard for their ADHD. So they said on piefed, they just won’t vote.
Yeah if someone can’t tell the difference it’s probably better if they don’t vote?
They can tell the difference. What they’re saying is, good posts don’t elicit a strong enough emotional reaction to get them to upvote. It’s not part of their workflow.
Reading insightful conversations and giving them a little ⬆️ gives me a positive dopamine rush.
Not naming any names coughfrongtcough
FWIW I looked at your Attitude rating and can say that unless you were trying to juice the numbers somehow, you fall well within what I’d consider normal patterns of upvote/downvote. The algorithm considers you to be far less negative than you may think you are.
I mean it’s all relative right? Sure we all get downvoted sometimes but if you’re getting downvoted way more than the average person you might be doing something the community doesn’t want you to.
The attitude system is how many times you downvote other people.
Oh thanks I felt like I was missing something.
The reputation is the hidden score that if you’re below whatever threshold it displays the red triangle(s) by your name.
Attitude system?
The attitude system is about the downvotes somebody gives, not the ones they receive.
And I think it’s a good thing. If downvoting is disincentivized, then downvotes themselves are more significant.
I… Know. I read the post. And thatsl’s what I’m saying.
Guess I’m confused about why you brought up the Attitude rating then.
A profile’s attitude (votes given ) has no correlation to a “low reputation” status (votes received ).
Saying people have a bad attitude because other people suck is rude.
Attitude only affects people who only downvote. Let’s say there’s a community with 100 people. 99 of them participate in good faith and make good content and the final one participates in bad faith and makes bad content. One day, all of them make a post and vote on everybody else’s posts.
The result: 99 good faith community members have posts with score of 98 and an Attitude of ~98%. 1 bad faith community member has a post with a score of -98 and an Attitude of -100% (or 0% because I forget how this works exactly)
Those sound like fightin’ words!
I think it might be fine? I like systems that reinforce trust and dissuade bad behavior as long as they aren’t draconian.
We use it that way, we want to disuade bad behaviour. As it is explained, there is very few users that reach veny low score.
Yeah I saw people freaking out over it like it was the end of the world, or even a bad thing really. Thanks for the clear explanation.
That said, I wish I could at least see my own, of nobody else’s.
I always thought the two red triangles were a UX bug. Looks weird.
If you move you mouse over, or finger, you can obtain the reason :)
They mean the fact there’s two instead of one
Oh i get it, thank :)
Deleted by author
I think I get Grail’s point. One red warning makes sense, but most of the time I see one, there will be a second with a nearly identical description.
I wish “votes” were multi-dimensional. “Yes/No”, “Good/Bad”, etc. is just too simple for the amount of nuance that exists. For example I could agree with somebody’s statement, but disagree with their tone.
That’s why i think we can find something else. The tone should be managed by mod