Vote manipulation is getting more common. Some recent examples:
While the accounts were banned, the malicious voting activity stuck around.
Should admins have the ability to discard votes, and if so, which admins? Should community mods have that ability? Can you think of any ways that tools like this could be abused?
@rimu@piefed.social @wjs018@piefed.social a good topic to developp :)
When banning someone there is the option to remove their content too. It makes sense to include votes in that.
No one is here for the internet points. Why worry about imaginary karma?
Because it affects visibility of content.
Read OP’s post, they’re worrying about manipulation, not karma whoring or harassment.
Stuff like bots mass up or downvoting a post to promote or hide it.
Not if you sort be New.
If we expect to be remotely large, sorting by new only is infeasible.
As much as it pains me, I think the only solution to vote manipulation is to disable downvotes. Mind you, I don’t like it - I think downvotes are useful in a healthy self-governing community - but here’s my rationale as to why it’s the only solution:
1) As I’m writing this, it comes to mind that perhaps we can selectively disable downvotes? Just like some instances don’t allow fresh accounts to post, perhaps something similar can be done for downvoting. Maybe it can also be extended to accounts below a certain up- to downvote ratio, to avoid mass downvoters.
I suppose that would address only a part of the issue and there are other, less intrusive ways to mitigate the effects of malicious early down voting. For instance, early down votes could be weighed less.
It depends on the reason for banning, no? If the account was banned because it is a bot, it makes sense to remove all their activity including votes.
However, if the account was banned for misbehaviour, I think it makes more sense to remove only the offending posts and directly associated votes. E.g. all votes by the offending account in the thread in which the offence took place
Downvotes don’t seem to be much of a factor in post visibility, at least in scaled mode?
I think votes should honestly be a bit more like old school SlashDot voting, where you had several different types of votes you could leave on a comment like Insightful, Funny, Helpful, etc. Have a few negative ones like Bad Faith Argument, Spam, Advertisement, etc. And also like old school /., you’d have a limited amount of votes you can give. Make them replenish once per day, or have users earn additional votes for receiving positive votes on their comments, or something along those lines.
That would prevent bombing an entire comment thread with downvotes, and provides much-needed context for any given comment’s score.
I have to say, I’ve always admired the Stack exchange system. Yes, it’s a Karma-like system, and it’s obviously not perfect, but it means that accounts always start with very little abilities, most notably that they’re not able to downvote yet. And when those accounts do get the ability to downvote (which doesn’t come all too quickly), it costs a certain amount of their “reputation”, which makes them think twice about downvoting.
I would give you an Insightful vote but I don’t have any left. /s
Jokes aside, I like both limiting number of votes per day (or otherwise) and having different kinds of votes. The reason why something is up/down voted can make for a better discussion. But I am agnositc towards renewing votes bases on engagement. On one hand, it would increase engagement, and on the other hand, it could scare lurkers away from otherwise upvoting good content.
This! Lemmy/Piefed needs metamoderation.
Or disabled until a certain number of upvotes are reached. It could potentially be disabled again of upvotes falls down under the threshold again. Or just have them time gated.
Do away with voting altogether and force people to think for themselves instead of just following the herd.
It’s literally how what you see is regulated. If a company X wanted to hide products from company Y, they could make bots to auto-downvote Y products and upvotes X products.
Granted, I feel like more commonly vote manipulation is done for geopolitical reasons rather than astroturfing
Vote manipulation is done in both directions
I like this a lot.
Piefed has some comment emojis available. Not sure how they show up on other instances.
I used a “no smoking’ one on your comment. But did i use it properly or just to screw around?
Well, I’m here for the internet points. I’m a hoarder, so I like collecting stuff, internet points included.
Gog disabled down votes on its forum and now there’s a bot up voting every reply in derailed threads. Mass up voting can also be a problem in creative hands.
On Mbin, it shows as just a regular upvote. Emoji votes would also be a great change, too! I like the way Misskey-like instances use them.
I know? I didn’t say it didn’t happen, I said that positive vote manipulation can more easily be addressed with spam prevention measures.
Voting is an indicator of agreement/disagreement and will influence how people feel about a certain post.
Yeah, I haven’t seen that anywhere else. I also liked that each user had a limited amount of votes to cast.
Source: Excellent slashdot karma from when the site was good.
PieFed, at the discretion of community mods, offers restriction of voting to only subscribed community members. This limits drive-by downvoting from All, where people would not have read the community rules (which in PieFed are repeated in their entirety at the bottom of every post from that community).
It also offers restriction of voting to only “trusted” instances, thereby introducing a third category between the binary federation vs. defederation.
I have also seen communities on PieFed that disable downvoting entirely, even to subscribed members, even on the same instance.
Community mods can enable or disable these settings at will iirc.
I think the other place tried to solve this by weighing a certain number of votes up or down.
So if a post got 10 upvotes, the 11th would weigh less in the algorithm, meaning that it was harder to burry something that was already perceived as upvotable. If a post of comment got 5 downvotes, the 6th etc would “weigh” less in the algorithm making it harder to bury posts just by downvoting them. They also labeled posts as things like “controversial”, “popular” etc.
I don’t know that this is a solution, in part because our “algorithm” doesn’t really function on a karma system, and in part because I don’t have the kind of knowledge it takes to understand the finer details of how this arm of the fediverse works under the hood.
But I do like the idea of limiting the number of downvotes an account can make per day, and also perhaps automodding accounts that do upvotes or downvotes an a rate that a human user couldn’t.
I like the idea of a weighted rating or star system
This shows up regularly. It would definitely be an improvement over the current binary system.
Piefed already has the emoji reactions, so that’s a step in that direction
They are in /hot/, which is likely to be vastly more common than scaled.
This is just in fundamental contradiction to how the Forumverse works. No content regarded as highly relevant and interesting can be filtered via your system and sorting by /new/ on a wider scale just means that a new post you make can be completely missed if no-one notices it.
or have users earn additional votes for receiving positive votes on their comments
I found the slashdot system worse than the reddit/lemmy system, if you commented anything that offended the hive mind you got downvoted into oblivion and lost the ability to vote, which obviously ended up reinforcing the hive mind.
You must love cookie banners 😋
Because at least I’m piefed you get punished if downvoted too much
it worked. i also added a no smoking emoji to this comment.
Hard agree on the first part, hard disagree on the second part. Making the system into any sort of rewards system with counterbalancing not only makes the overall system tastier to exploit for Fake Internet Points, but also makes migrating less sellable to new users because their ability or value to interact is reduced or even nullified for a non-deterministic amount of time.
Yes piefed is known to exact CCP-style hidden moderation.
It’s a shame honestly, I feel like Piefed without up/downvotes at all would work better.
Eh, it still has some good things and in theory since this is FOSS someone could just, like, fork it and remove the whole shadow cabal moderation thingy.
I wish Mbin had even a fraction of the childlike whimsy that Misskey has.
For positive voting you could look at how quickly accounts upvote after a post has been made, combined with how new they are.
I suppose you mean the limitations per diem on voting is what encouraged the hive mind, but even without those limitations Reddit and Lemmy have developed hive minds of their own, with similarly SOHC behaviors.
oy vey, @goferking0@lemmy.sdf.org it’s your time to shine!
tell threadiverse your solution to this!
Uh, votes don’t matter here. That’s one of the improvements over Reddit.
But your (one of) solution is to kill half of the voting system to solve half of the vote manipulation. It’s like solving spam by turning off comments. I don’t think that is going to be a popular opinion
I agree this makes the most sense.
That’s not killing half the voting system to solve half of the vote manipulation. Downvotes do not even get used at the same ratio as upvotes. I’m sure someone can pull numbers, but I’d roughly estimate that in most communities no more than 10% of votes are downvotes. And even if they were, I’m not sure you quite parsed my full comment.
Blocking voting on fresh accounts is not a novel idea. As another commenter said, it’s the system used on Stack Overflow. Blocking all downvotes is not even the goal. The goal is to make brigading not worth the effort. The worst case scenario is that all downvotes get disabled (which still works, despite its unpopularity - it’s been implemented by instances like beehaw). But in the end, that’s just a baseline. It can be improved, and I like to believe that I was quite clear on that in my first comment.
But sadly do on piefed
What’s the virtue of everyone taking part in a conversation even when they have nothing substantive to add?
Who said you have to participate? Not interested? Move on.
Piefed doesn’t incorporate any of this into an algorithm.
IMO if you’re banned from a community, for good reasons or not, you shouldn’t be able to interact at all. If I kick someone out, I don’t want them peeing thru the mail slot.
Please elaborate.
Piefed ties your voting habits and how often downvoted so it can flag you if either downvoted too much or down vote up and down sort of equally.
Basically if someone goes against hive mind once or twice can cause getting removed or limited on their instance.
https://join.piefed.social/2024/06/22/piefed-features-for-growing-healthy-communities/
Find people who have low karma
When someone is consistently getting downvoted it’s likely they are a problem. PieFed provides a list of accounts with low karma, sorted by lowest first. Clicking on their user name takes you to their profile which shows all their posts and comments in one place. Every profile has “Ban” and “Ban + Purge” buttons that have instance-wide effects and are only visible to admins.
The ‘Rep’ column is their reputation. As you can see, some people have been downvoted thousands of times. They’re not going to change their ways, are they?
The ‘Reports’ column is how often they’ve been reported, IP shows their IP address and ‘Source’ shows which website linked to PieFed when they initially registered. If an unfriendly forum starts sending floods of toxic people to your instance, spotting them is easy. (In the image above all the accounts are from other instances so we don’t know their IP address or Source). Find people who downvote too much
Once an account has made a few votes, an “attitude” is calculated each time they vote which is the percentage of up votes vs. down votes.
People who downvote more than upvote tend to be the ones who get in fights a lot and say snarky, inflammatory and negative things. If you were at a dinner party, would you want them around? By reviewing the list of people with bad attitudes you can make decisions about who you want to be involved in our communities.
All these accounts have been downvoting a lot (Attitude column) and receiving some downvotes (Rep column). Their profiles are worth a look and then making a decision about whether they’re bringing down the vibe or not.
How does piefed treat them differently?
the childlike whimsy that Misskey has.
This, so much. I really wish I could read Japanese, because the really active Misskey instances look genuinely fun to be on. It reminds me a lot of the OMGPOP days, which I miss dearly.
Jesus Christ I hate that the internet is turning this space into your shitty fucking idea of a dinner party
Who said they weren’t interested? Your entire premise is based on them being interested.
So again, what’s the virtue and everybody having to comment even if they don’t have some substantive to say?
Who said they weren’t interested? Your entire premise is based on them being interested.
I don’t even know how you came to this conclusion. If you’re not interested in the discussion, and you’re not going to participate in it, why should you get to influence how others may see the content by voting on it and affecting its visibility? That’s the premise.
Buddy I don’t care how hard you’re trying I’m not going to let you move the goal posts. Answer the original question.
when did this change? I can’t vote if I’m even tempbanned let alone full banned.
Every single time I see someone with a low reputation warning, they are toxic users.
It’s an imperfect tool, but it helps identify trolls and sea lions.
https://piefed.blahaj.zone already has downvotes disabled, disabling the reputation warning features as well (you can’t get bad attitude if you can’t give downvotes)
This was about old votes from before the user was banned, specifically for vote manipulation bots
ahhhh, that makes sense.
they are visible, they do matter, just not that much as on reddit
as I understand it does not hide negatively voted comments. these are stats, for to moderators, for helping moderation decisions. It’s not automatic.
I’m sorry you feel that way, but since that isn’t what I said, implied, or alluded to, your question is fucking stupid and maybe you should learn how to read.
Basically if someone goes against hive mind once or twice can cause getting removed or limited on their instance.
Except it appears designed for that not actually detection of the bad users. And does nothing when the toxic users just say on their own instance or comm. (like goat, pugjeasus and with recent db0 votes the feddit.org admin)
That’s already the case, the question is what should happen to the votes before the ban
If “going against the hivemind” is insulting people (which is what I’ve seen most of the time with users with both warnings), then it works as intended.
Also, giving a lot of downvotes is usually a sign of toxicity, and that’s only based on the user’s actions, not the downvotes they receives.
And does nothing when the toxic users just say on their own instance or comm
As I said, it’s not a perfect tool. To solve toxic users creating their own communities where they reign alone would require admins stepping in. And in the case you mention, when the person is an admin themselves, there isn’t a lot you can imagine, no tool would be able to address that.
There has been many times on lemmy and reddit where that is not the case, just saying what people didn’t like was enough. From games to politics people love to dogpile. Making a system that helps do that is asinine.
Blaze you do so much for the fediverse but defending this type of system is very disappointing to me.
I said it above, but I’ll rephrase:
The point you are making with people going against the hivemind is related to people receiving a lot of downvotes, but doesn’t explain people giving a lot of downvotes.
Blaze you do so much for the fediverse but defending this type of system is very disappointing to me.
I indeed do a lot, and I’ve seen toxic users going rampant at a few moments. The lemm.ee shutdown due to trolling and toxicity is a sign that we needed a way to identify bad faith trolls and toxic users better to avoid mods and admins burnout.
The point you are making with people going against the hivemind is related to people receiving a lot of downvotes, but doesn’t explain people giving a lot of downvotes.
Because it’s treated differently in piefed but still an example of how the software is designed to punish those who do not act the way the creator wants.
It just doesn’t seem like a way to actually address bad faith trolls or bad actors just make it easier to purge. Especially with instances that love to keep those types around, see world and shit just works.
I have felt there needed to be a specific type of vote available only to the original poster and to the users individual reply.
An up/downvote from the OP or the user I responded to I think should be differentiated from another user who isn’t either.
If the OP or commentator votes that should be noted alongside the X number of random votes. It isn’t an anonymous vote, but those votes would be public acknowledgements tied to the user making the public post/comment.
giving a lot of downvotes is usually a sign of toxicity
Emphasis mine. When is it not a sign of toxicity? Rules are defined by their exceptions, so I am curious as to how this exception is navigated, if at all?
Essentially someone who posts with high frequency has a capacity to issue more downvotes without compromising this admittedly imperfect tool.
Now I was never really a reddit user, but the problematic karma farming of accounts associated with that place was directly linked to these kinds of tools and metrics, no?
Emphasis mine. When is it not a sign of toxicity? Rules are defined by their exceptions, so I am curious as to how this exception is navigated, if at all?
In my experience, it is almost always the case, but I said usually in case someone came up with a very unique situation.
Essentially someone who posts with high frequency has a capacity to issue more downvotes without compromising this admittedly imperfect tool.
Are you saying that because they would get more upvotes, they could offset the downvotes they receive? Potentially, but this is where the second metric comes in (giving a lot of downvotes), and as we said, the two are almost always present at the same time.
Now I was never really a reddit user, but the problematic karma farming of accounts associated with that place was directly linked to these kinds of tools and metrics, no?
Karma farming is an issue when users can see karma as an absolute value. It’s not possible on Piefed, which only shows a percentage of attitude (downvotes given, visible to everyone: https://piefed.zip/u/Blaze ) and reputation (downvotes received, visible only to admins)
Because it’s treated differently in piefed but
Then Piefed is fine?
Especially with instances that love to keep those types around, see world and shit just works.
Instances who tolerate bad faith trolls or bad actors are going to do so with or without tools such as Piefed’s.
Are you saying that because they would get more upvotes, they could offset the downvotes they receive? Potentially, but this is where the second metric comes in (giving a lot of downvotes), and as we said, the two are almost always present at the same time.
Right, though it’s a mitigating factor. I guess there’s something I don’t know about piefed: Lemmy comments all have a default upvote from the user that makes it. But it can be revoked by the user. Does Piefed work the same way? My thought only applies if that’s the case.
The upvote you give yourself is there, but IIRC it doesn’t count for your score.
While we are talking, this is the kind of users who gets the two warnings: https://piefed.zip/u/grimreaper@sopuli.xyz
While we are talking, example of a user with the two warnings: https://piefed.zip/u/grimreaper@sopuli.xyz
That’s an interesting example of a user this is designed for/around.
The general system of up/downvotes seems to be doing its job quite as intended: their views appear routinely unpopular and there’s a seemingly pretty strong community consensus around that.
It looks like their threads have comments that solidly and clearly refute the garbage manosphere stuff. For some people it’s the opportunity to express a refutation of it publicly and directly. The public viewer gets to read those responses too.
So with that example: what do the flags do that the content of their posts don’t already communicate?
I miss slashdot. My opinion is that if somebody was banned because of vote related chicanery, then their votes should disappear with them. If it didn’t have anything to do with votes, the votes should stay. Not sure if that’s feasible.
Because it’s treated differently in piefed but still an example of how the software is designed to punish those who do not act the way the creator wants.
If the part after the but is ignored sure. But that part is also the issue so 🤷.
The point is the tools of piefed will amply the bad actors by allowing them to do the worst parts of reddit in the fediverse.
And here’s an example of a bigger troll who it surprise surprise doesn’t get flagged https://feddit.online/u/FiniteBanjo
Would you rather discard a tool that identifies 90% of trolls just because it doesn’t reach 100% accuracy?
Piefed system doesn’t punish those who do not act “the way the creator wants”.
Downvotes given are only actions from the user, no interaction from the Piefed dev.
Downvotes received is feedback from other users, no interaction from the Piefed dev either.
allowing them to do the worst parts of reddit in the fediverse.
Trolls are already around, as you pointed out in another comment. I would rather have a tool that works 80% of the time to detect trolls than no tools at all.
So with that example: what do the flags do that the content of their posts don’t already communicate?
It warns other users that this commenter may be a bad faith user / troll.
Usually when I encounter a troll, I check their profile to see if they are indeed a troll. The warning saves some time on that, and is accurate the vast majority of the time.
I guess I approach it inversely. I encounter what looks like a troll post and I’ll only check profiles when either I am interacting with them, or there’s such deep downvoting already I’m just doing a morbid dive into someone’s history.
Most of the time though the user just has a deeply downvoted argument but otherwise normal and/or low engagement posts, so they wouldn’t be flagged by this.
So I understand that it can save some time with some niche cases.
But I can’t help but note that the system seems intentionally blind to targeted harassment, which can be a source, if not cause, of bad faith accounts. (And likely those need different approaches since those are also niche cases themselves.)
And maybe it’s all just because of my instance’s Local feed, so that’s what I see as a prominent problem on Lemmy.
But I can’t help but note that the system seems intentionally blind to targeted harassment, which can be a source, if not cause, of bad faith accounts. (And likely those need different approaches since those are also niche cases themselves.)
If you mean using puppet accounts to massively downvote someone, that’s also tracked, but with another tool
Not necessarily puppet accounts, just brigading in general.
It’s the rationale many instances used to defederate hexbear. (Even though iirc hexbear disables downvotes, so they’re defederated for users mass posting, usually that hogshit image, instead of mass voting.) It wasn’t puppets or bot accounts at any rate.
But then there’s repost communities where users share comments (especially in places they or their audience is banned from) or DMs for a group response.
Not to mention the whole ‘block and downvote all .ml on sight’ mentality. But hopefully that might be something this tool could catch.
Again it’s designed to empower downvoting to silence those against the hivemind
When you ignore all the downsides and concerns everything looks amazing.
I wish I had your confidence in it being positive
It’s the concept that that is what is a good thing by default
Like a good dinner party includes some heated debate and some reap feeling not some politeness theater to assure ourselves were betters
yes but these stats highlight those that are only participating or creating heated debates. users that are consistently downvoted, but also users who are giving lots of downvotes
We’re running in circles at this point. I still have to see an example of a false positive for this system, as a nontoxic user being flagged as such.
And when you only focus on the edge cases every tool seems insufficient.
Wouldn’t say running in circles as much as ignoring the issues
if a legit user gets flagged as toxic, there’s no other consequence besides the warning, that doesn’t prevent anything from being seen.
Yet