RE: PROPOSAL - How to DEFINE and FIGHT the BAD - Create a COMMITTEE with STRONG ARM to Support Creation of a GREAT STEEMIT CULTURE

[joshua-pace]

Thank you for articulating your thoughts and concerns so well. I love your thoughts on flagging and I have a couple of thoughts to run by you in that regard.

1. Make Reputation matter in flagging. This might already be the case, but someone with a high reputation should carry more flagging weight than someone with a low reputation. In addition, perhaps there should be hard rule that you can only flag content from people who are at (or lower) than your reputation level.

2. If you flag someone with a higher reputation score than yourself, it doesn't have any immediate impact, but it goes to a randomized pool of higher-reputation individuals who would then confirm or reject the flag. For instance, I'm at (38). If I flag someone who is a (55), someone else with a (55) or higher must sign off on it. An alert of my flag would be sent to a random sample of (55) or higher individuals and would be queued or passed around until someone confirms or rejects the flag.

I have a lot more thoughts on this matter... but since I'm not deeply knowledgeable with how the system currently works, i'll leave it at that for now. Cheers

Comments

[edje]

Both items you mentioned are no issues as far as I know the rules of flagging. Only higher Rep's have effect on Rep of the user they flag.

Wrt to Flagging: I think we need more than 1 type of Flag: 1) effecting Rep 2) effecting rewards. I'm not the best person to determine what the most effective tools or processes will be. I think that requires a larger community debate involving at least those who have more experience in this matter than I, also those who know what can be coded and what not, as well as the wider community voting for detailed proposals including everything from what the proposal will solve, how it will solve it, what the negative effects will be and so on.

A processes in which other members have to agree to something is not a bad idea at all. That is something that we shall not forget. It makes things more complex, but maybe it serves a purpose.

[joshua-pace]

Thank you for your response and again I appreciate your thoughts on this. The more I think about it, the more a distributed, two-level flagging system makes sense. Even though it adds a layer of complexity, I think you could get around it by adding a reward to the "confirmation" level of flagging. And since the confirmation layer is randomized, it would be very difficult to game the system.

Here's a theoretical walkthrough:

1. I flag someone and specify the offense (maybe using different flags as you mentioned). In this case, let's say it's plagiarism.
2. My action of flagging for plagiarism is then randomly assigned to a handful of people with a higher reputation score than myself.
3. Motivated by some reward (whether financial or reputational), the higher-reputation individuals see if the plagiarism flag is deserved.
4. Final judgment is rendered by quorum majority--multiple people have to confirm if the flag has merit (this eliminates bias or lazy judgements)
5. Perhaps as an added incentive for the quorum to make qualitative and just judgements: if a judge/confirmer is routinely out of the majority with their judgement, it could affect their reputation score--or simply disqualify them from participating in confirmations in the future--and thus forfeiting the opportunity for reward.

[edje]

Your proposal could be a very interesting one to further look into. For sure, however is the 2nd, 3rd, 4rd decision maker, shall follow some rules on which we agree within the community and these individuals shall be 100% transparent in what they do and reviewed by the community from time to time.

[joshua-pace]

That's a great point and something to consider. Also, I agree that 100% transparency is definitely a good thing :)