Can we have downvotes and at the same time prevent 'flag wars'?

in #steem5 years ago (edited)

Hi Steemians, recently I already touched the delicate "DOWNVOTE" topic, but I think I have to dive a little bit deeper as it really concerns me. Also a discussion via discord with @hashkings was very interesting and convinced me to present my ideas in a new post.

In my opinion flags are necessary, but ...


In the post "Where do we take STEEM from here?" I already mentioned how some users are exploiting the rewards pool, and in a recent research I found out that the number of users who are writing posts or comments just for farming purposes is even higher than I expected.

Because of this problem, but also to be able to combat for instance spam, plagiarism or even worse, such as child pornography, I think the option to flag (now called downvoting) is necessary and helps to keep the 'STEEM organism' healthy.

Users who are against flags or at least against a limited number of 'free' flags ('free' until the downvote level has reached zero), couldn't convincingly explain me how to protect the rest of the community against the 'farmers'. For example, just consider the fact that @haejin is farming on STEEM, but he isn't farming on HIVE (even if he tried) ...

The crux, however, is that downvotes are often set for the sole reason of pursuing other users, solely because of their dissenting opinions or even completely independent of what they write(!), and denying them permanent visibility and any rewards. This is counterproductive to say the least and makes a devastating impression on newcomers who happen to observe such 'flag wars' or even get into them! We should be aware of this.

... if I am in favour this option I also have to offer ideas to solve the problem of personal 'flag wars'.


To contain 'flag wars' waged purely for personal motives, I think there should be a committee of respected users elected by the community and equipped with sufficient delegated STEEM power by Steemit, Inc.

This arbitration court could be called in cases when a user feels unjustifiably attacked, and then had to decide whether the flags were justified or not. In case the arbitrators agreed with the complaint they would counter the flag with an upvote (and possibly downvote the attacker).

My idea means that apart from curation accounts who reward quality content we would have arbiters to decide about flags. Then it would still be possible to fight abuse but as well curb unjustified personal attacks.

Of course details had still to be discussed. I imagine that the committee wouldn't have to discuss and get active because of every little flag, but in serious cases, if for example a whale hunts a small user and automatically flags every of his posts, then it would be time to step in.
If for example I flagged anybody who posted about dogs because I like cats, then the arbitors would counter my action, but when I downvoted him because of plagiarism they wouldn't interfere (or even add an additional downvote) ...
In addition, the arbitrators could also actively take part in the search and combat of spam and plagiarism.

So what do you think? "Great idea!" or "Typical jaki01 nonsense!"? :-)

I add an additional question as it fits the topic perfectly: a while ago @glory7 suggested that there should be the necessity to give an explanation for every flag. Would that feature be feasible and useful to curb flag abuse?

Sort:  

There is only one real solution. Remove all STEEM rewards and move the "Proof of Brain" completely to SMT's or other second layer solutions like the Steem Engine tribes. Only this will end all the Drama.

For example, nobody cares who is on trending in "Splinterlands" or "Palnet" or other tribes. It's just interesting inside the tribe.

I know you are in favour of that.
The problem which I see is where the value of all these (or at least some) SMTs would come from?
In case the SMT rewards had no real value, then I wonder what the motivation for the average user could be to change for example from facebook to STEEM?

I came here first because of "Censorship-Resistance". Maybe Hive will go full censorship-resistent and will skip all rewards because they will have no funds, and Steem goes to content rewarding for cat photo posts, but everybody knows they will censor non-cat content.

Yes, of course there can be several motives why to join a blockchain based social media platform.
Here I was only considering the topic from the perspective of what would contribute to or delay/prevent mass adoption of STEEM and the possibility to earn money there.
HIVE has its fascinating aspects, but I dislike the behavior of part of its leaders ... So I cannot do much more than waiting and see how things will develop on both sides of the big rift.
Maybe, as you said, both networks will occupy different niches of the crypto universe, aiming for different target audiences and thus will coexist in future.

Would be very cool if the could coexist peaceful.
Let's hope for this !

In the end for me is the question if downvoting content whom you don't like (I mean not something what is against law or other things like discussed in this post) is not similiar with censorship or at least close to them. And to let people not write something (even not in their own blog) cause they have not enough ressource credits is also near to censorship.

Very good point !

If he can write at facebook as much as he want and at steem he is strongly limited cause of resource credits and he has even no change (except to put money in the system) to change this when he cannot earn rewards to increase his steempower.

Not gonna work, Whaleshares did that exact same thing and well... No downvotes, No reward pool, some type of tribes called pods and other things that were added on.

Nobody cares about "Pods". Censorship-resistence is the USP for the chains, the rewarding is just overkill and leads to all the drama. Skip the rewards and focus on censorship-resistence. Make one thing excellent, instead of making 2 things only a little bit ...

No, I agree with you. Can you imagine a true censorship-resistence platform, governments would 💩 themselfs.

Now that would be a true cryptocurrency revolution!!

The question is, if STEEM would attract enough users from other platforms without a rewards pool?
Why should they join STEEM as long as all their relatives and friends are using for example facebook?
The ideal of censorship resistant is great, but alone not enough to attract the masses, because, unfortunately(!), many (most) poeple don't care about it.
OK, one option would be of course not to care about mass adoption and the monetary value of STEEM and just focus on non-material blockchain values ...

Free flags are like free ammo, you can't really expect something good from it.

I've read so many suggestions before about this issue (limit the reward per post to the average median, remove the reward pool, lower the reward pool, change the whole concepts to PoW, focus on Steem as a utility token, remove the reward pool and use SMT for the PoB, use the SPS as a decision making vector....)

People are currently too focused on what is happening now, but I believe we should look for a long term solution. Imagine if mass adoption comes to happen, how can any entity monitor millions or billions of genuine and fake users. It's impossible and no entity will have enough resources or manpower to constantly scan the whole chain to detected abuse. I've discussed before with some people the idea of how users could store underage pornography on the chain as text then use specific toolings to recompile the data into images. What should we do then? fork them out? Because hiding them from the UI is useless since pedos aim for secrecy in the first place. A delicate question that needs the consensus of the community, otherwise, we will be known as pedo heaven.

I believe there is a need to have a Steem Foundation, a structure that can lead the discussion to reach a clear consensus. There are a lot of things that need to be changed and discussed, but we need to make sure while moving forward to not ignore the bigger picture and push things based or sensationalism (aka changes that are based on emotions, anger....).

I believe we should look for a long term solution. Imagine is mass adoption comes to happen, how can any entity monitor millions or billions of genuine and fake users.

This is a good point, but I think the time between now and this hypothetical future need to be considered too. The norms and practices we have in the present shape what the future is like. And if the platform in the present is more attractive to spammers than it is to legitimate posters then we're unlikely to get to mass adoption (similarly, we also have a problem if whatever anti-spam/abuse ideas are implemented are more burdensome on new users than on spammers).

" how can any entity monitor millions or billions of genuine and fake users."

This is a very good point, I totally agree.
Their must be a automatic solution, it will be impossible to analyse all posts, at least if you have nobody who is ready to pay for this a very large amount.

For now there actually aren't that many posts. To spot abuse a tool like for example SteemWorld created by @steemchiller does a rather good job. Quite often a glance at "Voting CSI" tells you already enough ...

In case community grows, I think that the final decisions could still be met by humans, but the suggestions what to analyze deeper would be offered by algorithms.

OK sounds fine, I hope you have the people who are ready to do this job.
You think they should get some reward or something for this job ?

Yes, they should get rewarded!

Free flags are like free ammo, you can't really expect something good from it.

How do you want to prevent single users like @haejin and many others to extract hundreds of dollars (per person) every day?
I don't consider this, plagiarism and spam as minor problems.
It's so discouraging for newbies to observe these kinds of behaviour. Some told me that they don't like to take part in a community where pure greed gets rewarded ...
I think there has to be any solution, and if no flags then anything else.
Of course I respect your different point of view but cannot agree here.

It's impossible and no entity will have enough resources or manpower to constantly scan the whole chain to detected abuse.

Algorithms will do. They seek and preselect suspicious activities, so that humans only have to check whether these cases suggested by software are worth to investigate further or not. In addition users can always report abuse manually.
Anyway, there is still a long way to go until mass adoption, and I think it's worth solving problems now which in the worst case could prevent to reach mass adoption at all ...

Loading...

Don't count selfvoting for reward pool will solve this issue.
If you than also add the rule that all accounts who are created with ressource credits from another account (what you can analyse at the blockchain also for the past) count for reward pool to the account who created them you will also have solved the issue with multiple accounts.

First of all there are (many) other ways to create multiple accounts than using one's main account. I could for example ask you to create "jaki02" for me and I create "wer-verliert" for you. I can also create accounts via different front ends using different e-mail addresses and phone numbers.

Furthermore your suggestion would punish honest users who create new accounts for friends using their own resource credits, because then they couldn't support these friends anymore.

Your idea also doesn't prevent schemes in which for example every day whale A upvotes ten mini posts of whale B, who rewards ten mini posts of whale C, who upvotes ten mini posts of whale A.

"Furthermore your suggestion would punish honest users who create new accounts for friends using their own resource credits, because then they couldn't support these friends anymore."

Ok this is a very good point, to be honest I was not aware of this point.

Thank you for the post. I have thought about this, and it seems that this downvote issue is a lot more complicated then I had initially thought.

At first, I thought that making downvoter specify the reason would help alleviating personal retaliations or "emotional" attacks.

It seems that I was naive... After observing this Hive drama and what some of hive users are doing on steem... now I would not be surprised that they will automate downvote bot and spam the same downvote "reason" msg all over.

Moving forward, I believe that we should either remove the free downvoting mana and/or use the Steem inc stake delegation to counter spam/etc.

I'm for removing the free downvotes. The ability to decrease the payout of a post should come with a cost, as it was before implementing the EIP.

My full answer in @donekim's poll post:
https://steemit.com/hive-101145/@steemchiller/q8s0d9

During the time we had no 'free' downvotes @haejin (as my example here) exploited the rewards pool. When they got introduced he stopped. Now he is exploiting the rewards pool again.
For me it is obvious that farming like that won't be combatted if downvotes cost something.

I read your full answer. You wrote the same as I stated, too: free flags are used mainly by whales because all others fear retaliation. That's why I am in favour to implement arbitors/committees of users to protect people from retaliation and flgas given just for personal reasons.

Hm - If you recognise a crime and go to police to tell about this, what would you say if police want to have a fee from you cause you gave them this information ?

... now I would not be surprised that they will automate downvote bot and spam the same downvote "reason" msg all over.

What about adding the necessity to solve CAPTCHAs before giving a flag? That may help against automation ...

This right here.

Not sure about technical feasibility, but that certainly be a solution.

Any bot interacting with the API could get around this easily

Just don't allow users (bots) who are directly interacting with the api to flag? :)

For me best way would be to use the Steem inc stake delegation to counter spam/etc.

To remove the free downvotes means that many user don't want use their votepower for downvotes.

F.E. to announce child pornografy (as one example) should be honored and not punished (by loosing your votepower).

How many people would not go to police and announce their some crime if they would know that they had to pay a fee to the police for this ?

We need to act now -- or is it already too late ? this is gonna be my next post topic . To be honest and more relevant on why flags are needed -- or are they not needed stays where is steem moving towards ?

my personal opinion has always been the same . Steem is turning into the worlds most worst playground for capitalism , where richer is getting fuck load of more money and the poor peoples are getting nothing and leaving this platform (in this case the new comers / users )

Is still really a proof of bran now ?
where the fuck is brain ---

https://steemit.com/@raycoms/posts

https://steemit.com/@pharesim/comments

Above are few examples of why we need flags

Now is free flag worth it ? well to be honest -- I simply don't know , The idea of captcha means we will have to flag manually -- without automation you really cant detect all that users that are milking or spamming inside the chain .

My and my team are building a bot (Fully Automated) and Artificially Intelligent to detect the different patterns of Spamming , for example Bid Botting , Less Character Posts , Voting Patterns of Attached accounts , Reward count and many more . The logic behind this bot is very huge . If any one wants to look this bot in action you can look https://steemd.com/@steem-keeper .
we are currently upvoting everything with 1% for testing purpose only .

my Personal Opinion - Since Steem will only have value around the users in this great ecosystem , we can elect a group of committee , and take this automation in action where we can use the free flag option (Hard coded ) , not every people will have this feature , and this group can be elected with the process of how steem.dao works , but the voting pattern should be different (Not based on the stake you hold but based of votes your get from active users only ) . The members of this committee will be reviewed Time and again and if necessary can be changed using a permission grade system that is already in graphene , This way we can fight against spam , save the concept of flagging AKA Downvoting , and make steem great .

Hope you like my answer

Hope you like my answer

Interesting (and good) answer. I think there still would be a lot of details to be discussed, but most important is that people (and especially the 'leaders') get aware of the problems 'farming' and plagiarism are currently causing.
Anyway, removing flags completely can't be the solution in my opinion, and I think we agree.

Very interesting discussion, and some useful potential solutions put forward.

In terms of some arbitration committee, I guess the process of election would be most difficult...

I think there should be a committee of respected users elected by the community and equipped with sufficient delegated STEEM power by Steemit, Inc.

That would need some thought, and consensus building.

Yes, true, I didn't think too much about the details yet, because at first I wanted to see if anybody (respectively: enough users) are interested in my ideas and ready to discuss them further ...

I was reading this thread earlier this morning and then thinking about it while I took my morning walk. I know that some people think "let people handle it with SMTs" is a solution, but I'm thinking that just kicks the can down the road. How would an SMT handle these proof-of-brain abuse problems? Or are they going to be just as puzzled as we are about how to fix it today? Since any given SMT economy will be smaller, I think it would behoove us to figure out how people who want to launch an SMT could protect themselves from things like economic bullying. If some bright-eyed and bushy-tailed person launches an SMT for their hobby community and then an obnoxious Steem whale buys their way in to wreck the community for their own trolling amusement isn't that the kind of thing that would sour people on building things on top of Steem?

I think the distribution of SMTs could easier be to controlled within comparatively small communities with admins and experts concerning certain topics.
However, the problem which I see is where the value of all these (or at least some) SMTs would come from?
In case the SMT rewards had no real value, then I wonder what the motivation for the average user could be to change for example from facebook to STEEM?

A good idea with the arbitration court, at least for the time being, because in the long run it would probably have to be further elaborated if the number of users were to increase extremely and at some point there were simply not enough personnel left to be able to handle the flood of cases. Even I, as a user who is generally against downvoting, can't offer you a convincing alternative and, based on your arguments, (unfortunately) have to admit that they are necessary to a certain extent, even if not to the extent that they are done here.

To make a rough suggestion, how about some kind of committee that has to approve articles before they can be viewed by everyone on Steemit, this could make downvotes superfluous, which would otherwise only be necessary for posts (spam, child pornography, etc.) that, if sorted out, would not be published at all. Of course, sooner or later the problem with the mass of posts will come up, which cannot be handled by humans, but maybe it would be possible to limit this by a kind of "logarithm", so that some posts will be "sorted out" by the system in advance to relieve the mentioned employees.

As I said, these are only rough suggestions, which, if they bring something, will surely have to be thought over further, but who knows, maybe a professional can get some kind of food for thought in the right direction, so that his results will be more concrete than my suggestion. ;-)

To make a rough suggestion, how about some kind of committee that has to approve articles before they can be viewed by everyone on Steemit ...

I think that would be even harder than just concentrate on flags, and apart from that would be censorship ...

Ok the first question I have if steemit can handle all complaint about flags or who should do this ?

The idea from @glory7 is fine but the question is what if I save me one time a text and copy/paste them every time I flag something ? So in the end it would be also necessary in this case to analyse this reasons.

My idea would be that one flag is no flag, means if only one person (not account) add a flag this should not count. Their should be at least two persons who flag the content.

I am aware that this is not so easy cause many persons have multiple accounts and so steemit should calculate them together, but I think it should be possible.
As the blockchain saves all steemit should know when a new account is claimed at the blockchain what account has paied the ressource credits for this and add this new account for the purpose of flagging to this account. Means, that all accounts who was created with ressource credits belong for flag purpuse to the account who paied the ressource credits.

Ok the first question I have if steemit can handle all complaint about flags or who should do this ?

Not Steemit, Inc. had to handle these complaints but user-elected committees created for exactly that purpose. Please read my post! :)

The idea of @glory7 might prevent automated mass flagging. It could be discussed and improved further, for instance by adding the necessity to solve CAPTCHAs before giving a flag.

My idea would be that one flag is no flag, means if only one person (not account) add a flag this should not count. Their should be at least two persons who flag the content.

The idea is not bad but how could you know if two accounts represent two different persons? :)
I own quite some accounts, and most of them weren't created by using my own resource credits (actually most of them existed already before resource credits came into play at all).

"The idea is not bad but how could you know if two accounts represent two different persons? :)"

OK at the longterm you could analyse with KI, means if always (or often) accounts add together flags KI could bring them together and count them as one person.

But it would not work immediately and help not in the situation, just in longterm.

Another possible way would be to offer (free choice of every user) the possibilty to do ident check for accounts and add the rule that only accounts with checked identity can add flags. All other accounts can just exist and do all normally but would not be able to add flags. Or to change this a little bit and tell that they can also create flags, but it's necessary that two accounts with checked identity add a flag and only in this case let all flags count.

The problem with your last idea is that many blockchain users use crypto currencies and blockchain among others because of the possibility to stay anonymous/pseudonymous.
One of the main ideas in the crypto space is to guarantee that everybody can state his opinion without the fear to be suppressed by individuals (who for example hate homosexuals), groups of people (who for example don't tolerate different religions) or governments.
That's why many crypto users for instance dislike Voice.

Ok but I don't see any problem in this cause as you can have multiple accounts you can do identity check only for one account and can post critical content with another account and nobody must know that both accounts belong to you.

If you don't like this idea than I think it should be ok if you don't have the possibility to flag content.

In my opinion if you want flag content you must not be able to do this anonymous.

I think that the experience shows that (a lot of) people flag anonymous content cause of their possible interest and not cause this content is against law or against somebody. They even have a problem with one post of the user (maybe correct) but flag not only this one post but flag than every post of this user, although the other nine (from f.e. ten post) are quite ok - or even they write a bot who directly flag every content a specific user post new. Should it really be ok to do something like this anonymous ?

Then, where to put your SP?
To the identifed account? Then you cannot earn money with critical articles written with your small, unidentified accounts (people tend to follow big accounts).
To the anonymous accounts? Then the flags of your identified accounts wouldn't have much effect. However, if your anonymous accounts followed the flags of your identified accounts, then it would be obviuos that all accounts belong together.

To be honest, I consider my suggestion of elected committees as way superior compared to the implementation of identity checks on STEEM. :)

"To be honest, I consider my suggestion of elected committees as way superior compared to the implementation of identity checks on STEEM. :)"

OK there will be posts that are clearly objectionable and deserve the down vote. There will also be posts that are not objectionable and then the committee can remove the downvotes. As always in real life there will also be posts where it is a matter of judgment whether they already "qualify" for a down vote or not. In such a case, what if this committee may not agree and what options will I as a user have to defend myself against a decision by this committee? Is the committee almighty as the King of England used to be or will there be a legal process?

... what options will I as a user have to defend myself against a decision by this committee?

Don't plagiarise, don't spam with ten minimal posts per day, don't put child pornography online ... :)
If you read my article you know that I am against to flag dog posts just because I like cats. :)

Is the committee almighty as the King of England used to be or will there be a legal process?

I suggest the committee to be elected and reconfirmed regularly by the community. It would have quite an amount of STEEM power, but of course if many users upvoted a post they still could outvote the committee.

Anyway, I don't think it to be useful to discuss too many details as long as nobody is in favour of my idea ...

Great post!

Hi @jaki01

I just replied to one of your comments and then I've noticed your publication. Please allow me to repeat:

What would you say about an idea to build some blacklist and ensure that users who would end up on this blacklist wouldn't be able to upvote (their upvote wouldn't provide any rewards and it's value would be zero).

That would obviously have to be programmed by STINC, who would have to adjust their API.

Yours, Piotr

I think I would prefer to counter upvotes respectively flags of these accounts automatically to neutralize them.
Then no intervention on blockchain level would be necessary, and the decision could be easily reversed again.

Thanks for such a prompt reply @jaki01

I'm simply afraid, that abusers will outnumber those who are trying to fight with abuse.

And mostly that they will have more patience and will wear us out. Reality is that if I would have to keep tracking down and downvoting same people, over and over again - then I would give up sooner than later.

We need some solution, that would allow to block more "permanently" abusers from being able to place upvotes.

ps. do you use discord? perhaps you could join project.hope server and DM me. I would gladly get to know you closer, since we're on the same page now and we are both trying to figure out how to protect our investment (steem).

https://discord.gg/9wAJ4C

Yours, Piotr

I'm simply afraid, that abusers will outnumber those who are trying to fight with abuse.

I think the mentioned (or an additional?) 'committee' account with enough delegated SP from Steemit, Inc. would be able to successfully combat abuse.

I think your idea is very centralized (even more than mine) and conflicts with my ideal of censorship resistance.

Concerning Discord, I am using it but don't have much time for chatting. I will join your channel, but please don't expect too much ... :)

Dear @jaki01

I think your idea is very centralized (even more than mine) and conflicts with my ideal of censorship resistance.

i agree. it's very centralized. however downvotes already has been proved to create more tension, dramma, killing userbase and it's simply not working.

I've few good friends, who are right now "milking steem" like crazy. And when I asked them if they aren't worried to be downvoted - I received interesting answer: flaggers will get bored sooner or later.

That's the problem. Those who abuse system, they go "quiet" for several days after been downvoted and then they are coming back. Perhaps unnoticed. Or they simply carry on, knowing that they will win at the end of the day.
@ haejin is great example.

@haejin has no chance on HIVE ...
If persistent 'farmers' got flagged automatically, then they soon might stop having fun here. :)