Proposal for Flagging System (No drama here, please. Only discussions of agreeable solutions.)
This idea could probably use the community's input, as I'm sure it'll need some brainstorming if we want to find any holes, or perhaps aspects that might not be agreeable.
I've been thinking on this for a couple of days now, given the recent disagreements going on in the community, and I've tried to come up with my own offering of a solution, so please take it as you will.
Disclaimer: I'm not a developer. I'm not an economist. This is just me trying to approach a problem as logically as I can while trying to find as even a middle ground as I can think of that most people can agree on.
If this idea is stupid - great, that's one less stupid idea. But please tell me why if you think so, so that I can learn.
I'm also not creating this post for bashing any members of Steemit, so I'd prefer if we can keep those comments elsewhere.
First off, keep upvotes the way they are. As far as I can tell, I think most people can agree that that part is pretty satisfying.
Secondly, don't get rid of the flag, but change how it functions.
Let me give a quick scenario:
Let's say Mr. Plagiarist arrives and posts an obvious news article copy/paste straight from cnn.com. Or let's say he's an abusive troll, whatever.
As it currently stands, the community can downvote this post, but usually it really doesn't mean much in terms of impacting any potential payout or hit to their reputation until a sensible whale comes along and really gives that flag a punch, knocking down payout as well as reputation.
Here's the problem with that scenario: It relies on the whales to be the police.
I may be wrong here, but I'm willing to bet that some of the biggest stakeholders in Steemit are probably annoyed with the amount of policing they've been needing to do in order to make sure the payout is as evenly distributed as possible.
Here's the meat & potatoes of my proposal: What if we turned flagging into a community poll so it can police itself?
Hear me out...
That post by Mr. Plagiarst/Abusive Troll gets its first flag. Anyone can initiate it, even rep level 25 users (negotiable).
At that moment, for all other users, the flag is now a poll with "yes" and "no". It hasn't applied a flag to the post yet.
Majority rules here, so if the flag wins at payout time then BOTH the amount of pro-flag voters as well as the percentage leaning pro-flag will determine the damage to rep & payout. So if there are a lot of votes and it's a landslide, the damage is heaviest. If there are only, say, 5 total votes and 3 wins, the flag still gets applied, but with only 3 votes and 60% of the total number of votes, the damage will be light.
- The number of pro-flag votes would determine the payout drop.
- The percentage of the pro-flag vote would determine the reputation drop.
In other words, if you've got a lot of people voting for a flag on one of your posts, you'd better hope you've got as many people to come and defend you and vote "no" to the flag, because if you can get the vote as close to 50/50 as you can, you'll only get a drop in payout if the flag wins, not a drop in reputation.
But, if you've got a total of 20 votes, and 19 are voting for the flag, it's obvious at that point how the community feels about you and your rep takes a hit along with your payout.
EDIT: @neoxian brought up a great point that I neglected to address:
Combating Bot & Sock Puppet Abuse
My proposal for that is to not grant flag voting abilities to accounts that have less than a pre-determined reputation level, such as maybe 63-65. Anyone can initiate a flag, because that would be an alert system for the rest of the community, rather than having to wait for a 63+ user to see it and initiate it. But the moment the poll is started, only 63+ or whatever rep level accounts can participate in the poll. Since accounts have to be nurtured to get to these levels, it discourages that kind of abuse. And while the reputation level requirement may be viewed as unfair, it could be compared to having to wait until you're a certain age to vote, such as 18 years old here in the US. So not only are we combating abuse, we'd also asking new members to be patient and work on getting more experience on the platform, thus resulting in more informed voters. It's a little like earning your vote slider.
If the people ain't happy, ain't nobody happy.
I would like to submit that Steemit.com is the engine that powers Steem, and as such, the success of the whole system relies on the satisfaction of its community. The happier a community is, the more likely it is to draw in new members and retain them, not to mention attract new investors and keep them invested. By giving the people a voice in times when decisions of change are being made (maybe...hopefully), or when something negative - such as a flag on a user's post - must be dealt with, the more the community can be its own immune system, to figure it out and even the playing field on its own, the more likely the losing side will say "this sucks, but the people have spoken", rather than posting complaints about the system because "a couple of whales of spoken".
Let me know what you think, whether or not you think this sounds plausible, or maybe any holes you can think of.
I'd say it's time to start making the community happier, and I hope you agree that giving the people a common voice in polling can be a good first step in that direction.
@dan / @dantheman - I hope you see this. I know in the past you you were looking for a solution to the flagging/downvoting conundrum, and if by some grace of the universe you find this approach interesting, I'd love to hear your thoughts. :)
Voting that is not based on Steem Power is open to bot abuse.
Edit:
I think we should wait for the new power curve before trying other ideas.
Oh you bring up a great point that I forgot to include, and I'll be sure to edit it to reflect this.
I was going to mention that voters must have a certain rep level, maybe 63? This discourages using multiple accounts or bots because one must nurture their account enough to even get to that rep.
I had that idea before also, and I'd like to see if there's a problem with it, mayeb @neoian you can see it? What I said yesterday was:
I think we should wait to see how the new power curve pans out before even more ideas..
For sure! One step at a time.
I really wish there was an ideas backlog / approval system the devs used. I guess steem kind of does that but it's frustratingly dispersed.
Oh yes, ok added a rep requirement would improve the idea for sure.
You mean build the rep thing into consensus? It's hard, at least current rep scoring algorithm is flawed.
I believe his comment was in response to my suggestion that we add a minimum rep scoring for any given account, giving that person ability to take part in the community poll to decide if a flag gets applied to a post. Perhaps 63-65. This discourages bots and sock puppets from abusing the system since accounts really have to be used heavily to get to those rep levels.
Yes I know the idea, but IMO it's not a good solution: it's a big change, so takes time to implement, and not easy to do it right, but likely easy to be gamed.
Nesting:
The gaming part is what I was hoping to get input on. Whatever idea we might decide on (mine or not) has a possibility to be gamed, perhaps there should be a think-tank to put efforts into predicting those exploits and cut them off at the pass. Otherwise, my only question would be why you believe it would be hard to do it right? I'm not asking to put you on the spot, I just want to learn. :)
Simple answer: if it's easy we should have done already. :)
Actually whenever there is a new idea, at the first place I would think about how to game it, so comes this conversation. :)
In my mind, the easy solution is the easiest one to game. And that's what I mean by a think-tank - a group assigned to think of ways to game whatever system that gets decided on, thus coming up with ways of patching those exploits. Ultimately, I would LOVE to see a flagging system that is community based rather than based on the opinion of a few individuals that the community doesn't necessarily agree with. To me, the community is an organism that needs to build up its own immune system and decide what goes and what stays. That may not be easy, but just because it's hard doesn't mean it might not be worth the effort. :)
I hear you.
In reality it is open to bot abuse with steem power too, but it does take a lot more effort. I was going to mention how 1:1 polling could easily be gamed by bots though.
If there were a way to do a verified account and allow those to be the ones doing a 1:1 poll that might work, but it would add additional barriers for new people joining. It becomes a decision on what we are willing to do...
At this point since we are still in beta I believe we should be open to trying most things without assuming they will fail before trying them. We need the negative/failure results and success results to help us come up with more information. So while it is in beta TRY things. If they don't work, remove them. If they do work keep them. If some parts don't work, and others do then use that to come up with new results.
We all have our opinions on probabilities and what might happen and that can lead to paralysis in trying anything.
So if you guys want to try things. I'm open for pretty much anything as long as we learn from our mistakes, and our successes.
Please review the post - it has been edited to address potential bot abuse.
I didn't write about such abuse. But I was inspired.
Here is my post I just made. I hope you enjoy it...
Decision paralysis... solutions don't come from only taking the known safe path
I believe you'll find that it clearly indicates I'm willing to TRY any idea. ;)
It's a feature not a bug 😉 Needed to keep sock puppets from having an influence, with the sad effect of excluding newcomers. However there is precedent for unlocking features which have larger community impact. It's like this on StackExchange for example.
The "verified account" is just one which has above a minimum rep, say 55 or 60.
Totally agree.
Resteemed. I am going to write a companion piece and reference this as my inspiration. It is not to take away from what you are asking, or to ask the same thing. I wanted to talk about something even broader than steemit/steem/busy.org that I believe is directly relevant.
Good post. My comment to your post concerning what you are asking is my reply to @neoxian. That also is kind of foreshadowing what I want to write about.
EDIT: Here is what you inspired.
I love coming back to a great post with a good discussion going on. It's like sucking on a tootsie roll pop that lasts for a long, long time. I have to resteem these kinds of posts or miss out!
At Steemit we no longer have to be just pieces in their games.
I'm not sure but wouldn't this be susceptible to sybil attack?
Edit: Nvm forgot about your edit :P
This idea sounds better than what's going on now, to me! ;)
I've mentioned this in response to several posts today - flagging should only be used for plagiarism, abuse and SPAM- not to be the "great equalizer" of the rewards pool.
I've never downvoted a post because I choose not to. If I don't like a post, I click away with no comment and no vote. It's that easy. And if I see abuse, plagiarism and SPAM I report it to Steemcleaners because that's exactly what they were designed to do - isn't it?
But if a person wants to flag someone then there should be a pop up with 4 choices:
This notification could then go to the appropriate authorities to investigate and see if the flag is appropriate or not.
And WHY NOT have an unlike or thumbs down button that doesn't affect the value of the rewards. A downvote hurts, expresses the dislike from the reader but doesn't hurt the proverbial wallet and it's distinctly different from a flag.
AND why not have recaptchas for comments, upvotes, and flags for bot prevention?
This right here could even be something worth polling, to see if we can apply rules to the platform as many people are saying we should. So many people are saying different things about what should be done about any particular thing. Okay, that's great - if that's how the majority feels, then let's do that. If we're all just shouting from our own individual rooftops what should be done, nothing will get accomplished. We need a consensus.
As my father used to say: Opinions are like assholes - everyone's got one. LOL Sorry for the language but I think it's appropriate. I worry about an overall consensus because mob mentality is a real thing. I would love to see a roundtable discussion between the owners, witnesses, and community members because we may have differing opinions or pathways but the end goal is really to make Steem(it) flourish. We all want what's best.
But this is why I'm all for testing all limits and tolerances. We're still in beta - now's the time to push the buttons and see what explodes and what works. As someone said quite a few months ago "Tach it out!"
The overall community needs to take a chill pill - but that's just my opinion - for what it's worth.
Mob mentality is a reality, I feel. The question is not how to eliminate it, because I feel like that's a losing battle. You're fighting against human nature. The question then becomes, how do we minimize it? Do we try to talk to everyone individually? Or do we narrow things down to the best, say, 4 or 5 options and poll them? I prefer the latter, personally, because the former sounds too complicated drawn out - ultimately we need to face the reality that we will never agree on everything, especially if we're not voicing our opinions out as a community. Individually is just too chaotic, and everyone will have different directions they'll say we need to go. It's like having 10,000 hands on 1 steering wheel.
And yes, we're still in beta - people keep saying this. And since that's the case, what better time to say, "Maybe we should consider this for the real deal?"
If there are ideas that would work better than others, we need to hear them now, not when it's too late.
Exactly. No one likes change so people will grumble no matter what happens. I say test those boundaries. The diehard Steemians will stay the course.
It might also be worth a thought to consider a system where bad flags can be discouraged by a community vote on whether a flag was in good faith or just in spite from disagreement. I figure fear for losing reputation for bad votes is just as good a deterrent than gaining a reputation for being a good voter is an incentive.
Exactly, and I think the inherent structure of the idea would include that. If it's a simple yes/no vote on a flag, I think people are more likely to look into what they're voting on - especially if the poll is popular no matter which way the vote is going.
Interesting idea! 🙂
Couple of questions:
Good questions.
I would say to make it as democratic as we can, one-user-one-vote sounds like the best approach. Maybe we can vote on it. ;-)
I would think anyone can start a flag, but the weight of it will depend on how many people vote on actually applying it to the post if it wins.
Please review for an update to the post which addresses potential abuse.
Thread depth reached, come on HF 17!
No problem 🙂 I think the poll idea is actually much better than my idea of revoking a flag if "reported" enough by users which high enough rep. I prefer this, well done on the idea.
What about point 2? This is not addressed.
I addressed it. The inherent design of the idea would mean that anyone can start a flag, but its weight is ultimately determined by how popular that flag is. Let's say only 5 people vote and the flag wins by getting 3 votes. In that case, 3 people have won, but their flag will be weak because A) there are only 3 of them, and B) their win percentage isn't that great (only 60%, just over half). So while that person does get flagged, it's damage is light.
I apologize, I was looking at the wrong #2 point.
I do foresee it getting a bit cluttery, but perhaps a new section can be created which shows only flagged posts. In fact, it would be nice to be able to filter them by date posted, most popular, category, etc - not unlike how we do with all posts now, but just for flagged ones. I can think of a good handful of people who might spend time browsing that side of Steemit. LOL
I have a great idea. It's called freedom of speech. Create a block "button" or create an option that enables users to "approve all comments" before they show up publicly. I'm tired of these useless eaters who can just say whatever they want to sway perception about one's posts, and then my feed being a giant complaint thread...
It's not fucking rocket science, Steemit. Thank you for providing me a platform to vent :)
There is a "disable-reply" feature in the back end, just not yet on the UI.
I don't feel the need for a big change to the actual system structure. What we've been seeing occur is almost all psychological in my opinion. Simple User Interface changes may be enough.
Modify the Trending page algorithm to something that factors number of votes and rep of the voters, comment activity, and number of views as primary factors over payout alone. 3 whale votes shouldn't put something on trending over 50 votes from active, respected minnows and dolphins.
Don't display pending post payout in the list summary/thumbnails. If someone wants to find "over rewarded" content to down vote , they'll at least have to click on every article to see what the payouts are. The rewards system is great, and one of the selling points of Steemit, but making it the primary motivation isn't working. Imagine if everyone wore nametags with their annual income printed on it as they went about their day. I think we'd see some interesting new social dynamics and an increase in feelings of envy/resentment/hostility in our daily lives.
Change the flag icon to a down vote like the UI at Busy.org.
The one possible system change I could possibly see as good would be adding a "block" function. At the very least it would block another account from seeing your posts, at most block them from interacting with them altogether.
If blocking merely limits visibility, it'll just be another roadblock and deterrent to trolls or accounts acting as "value police" although they still could by interacting with the blockchain in other ways if they are savvy enough.
If interactions between accounts can actually be disallowed, I'd find that interesting. It would require people to accept that they can't drive out viewpoints that they don't agree with. For good or bad, there would be terrible communities on Steem that would block anyone who disliked them and could ensure they survived as a group. Of course the act of blocking means they are cutting themselves off from that rewards potential and the visibility of spreading whatever information or message. As community features are rolled out this could also manifest as private communities. Again, the pros and cons are that content within limited social circles is free from flag attacks and comment from the "outside" world, but it's also impossible to garner more rewards than that which can be generated by the combined steempower of your group.
If people want a shot at the whale up votes, they've got to be willing to accept whale down votes as well.
And if that is the system that ends up working the best, then so be it. Otherwise, I have to file that one in the pile with the other things which don't necessarily cater to a happy, satisfied community. The more a community finds things here to be unfair - especially in a place where it's 100% voluntary to be in the first place - then the more likely that community will shrink rather than grow. Catering to community happiness is just good business.