Nate said:
I would like to see more solutions in the form of incentive structures.
This is my favorite thing about blockchain technology and that's what I was going for with my Flag Rewards article. We can talk about what is acceptable behavior but if we don't express and enforce the desired outcome with consensus rules, the "undesired behavior" (if there is any) will continue.
In my post, I say:
If you are the first person to flag a post that has a large payout, and the payout later becomes zero, there should be a reward for the first flag. The reward should be a 24 boost to voting power.
But if the post doesn't become zero, you should have a voting power reduction for 24 hours.
So, the idea here was to incentivize flagging of plagiarism and spam. Stakeholders who just want to reduce payout of "overpaid" posts would have less of an incentive to do so.
Setting aside whether or not this is a good idea to implement, this is an example of an incentive structure solution.
Absolutely inertia, that's the spirit of what I'm trying to say. I think you're right that undesirable behavior is going to continue if we don't inforce it. Asking the whole community to change is simply not an option.
If stakeholders want to protect their stake, and they think the payouts are too large, one way to fix this might be to get the witnesses directly involved using parameters. Witnesses could set the size of the reward pool. Then we can get some price discovery going.
If the stakeholders know that the reward pool is half the size today compared to yesterday, wouldn't they be half as likely to flag "overpaid" posts? Granted, any given author would also know this, and maybe he'll put off posting his big article until the reward pool goes back up. That's the whole point of price discovery: "Should I allocate this resource here, now?"
Last night on SteemSpeak, @noisy told us about an idea he had: one day a week, the max payout is set to something like $10. I thought it was a pretty neat idea. It would be interesting to see how posting habits changed if something like that was implemented. Would people just post twice as much because they have a good chance of getting $10 (each post), since all of the people who routinely get on trending would avoid that day?
I like @noisy's idea, but I don't think it should be once a week. I think it should be another witness parameter. If all of the witnesses decide on an exact day and an exact max reward, only then will it happen. Then there'll be a 1 week cool-down. Then they can try it again.
I wonder what bots would do in that situation? It adds a bit of uncertainty that would be difficult for bots to predict until a few of these "max-payout" days have gone by.
Fascinating ideas in this post. I love what @noisy proposed. If bots don't know the potential outcome then bots become useless. Seems this particular solution solves the trigger happy flagging and pushes bot to the brink of extinction.
I think the bots will know better than most human users.
It's true that bots would know with certainty that the witnesses have set these parameters, they just won't be able to predict the behaviour of the authors and stakeholders under those parameters.
I envision the creation of 'post rape gangs' designed for the specific purpose of reducing posts to zero. For those engaged in such behaviour outside of such gangs, it becomes a kind of gambling .. "Will people agree, and drive this thing down to nothing!? Let's seeeeee! "
These things aside, it is adding complexity .. this is certainly not easy to explain. I can't seem to find anyone to fully agree, but the problem is solved by making flags/downvotes of equal weight, so that 'numbers' of users are required to kick the shit out of bad actors.
BWAHA! 'Exactly!'
If you use numbers than you have bots flagging posts. You actually enable even worse bad-acting because it's practically free to make an account. It becomes very cheap to flag a post into oblivion.
Reputation would have to be a factor then, to limit the use of bots in this fashion .. I suggested 50+, or perhaps even more! There is no reason that people who have not invested any time in the platform be given the right to flag others who have.
Bots must be pretty easy to identify, given access to account / login information (location) and blockchain activity .. perhaps one downvote per post, per IP within a 24 hour period. Certainly some 7-proxy wizards can take on that challenge, but .. wow, what a waste of a life, just to sneak some literally-lousy bots in to do some entirely destructive flagging, that gains the bot-master nothing? Once identified, these bots should be completely destroyed for breach of terms, and whatever STEEMPower they have collected would go into a pool for later distribution to the entire LEGIT userbase! :)
Pretty interesting take 🙂 But that rejects the idea of reducing "overpaid" payouts, which is currently a feature of the system, not a bug. There'd need to be acceptance of that as a bug.