RE: Voting Abuse and Ineffective Curation: A proposal for blockchain-level change
I think delegations can potentially help new users have a curation impact quicker than the traditional method of buying SP directly, because there's a leverage factor (it's cheaper to rent than to buy). But this only works well if curators can get enough curation rewards from the delegated SP.
My motivation for delegating my steem power was to lower the impact of my own votes and to increase the number of curators that could give meaningful rewards since I don't have time to read all the posts that are made. With my manually negotiated deals, I generally gave up 50% or more of my potential curation rewards, but with curation rewards as low as they are right now, this wasn't a major loss.
Originally, I negotiated deals with about 6 or 7 people for good sized chunks of SP (160-500K SP if I recall correctly), some for a split of the curation rewards and some for no profit (free delegation). These are all people that were all pretty well known on Steem (popular authors/curators) and that I had personal chats with over time. Despite this, I had to deal several times with conflicts between just those few delegates, which really surprised me.
But more importantly, personally negotiating delegations just doesn't scale when you have as much SP as I have. I can't personally negotiate enough such deals to get a good spread of influence. The other possibility for personally negotiated delegations becomes to hand over the negotiation to trusted proxies who could negotiate and manage such arrangements. Several people seem willing to step forward for such a role, but I don't really like delegating that much power to people anyways and while most of them are probably well intentioned, I don't always agree with their rationales for where they would award the SP.
Ultimately, I decided that a capitalistic system for awarding delegations would allow for the most scale and least cronyism, provided I put in a couple of safeguards. The primary safeguard was to make sure that the cost of purchasing a delegation was more than it's possible to make back in income just by self-voting alone (so you might profit if your self-vote gets you more attention that leads to upvotes from third parties, but you won't profit if just end up with the money from the self-vote).
As a backup safeguard, against someone who just can't do economic math and doesn't realize they are losing money by self voting bad posts, or against someone who doesn't mind losing money if they can multiple their ability to troll others more than they could by directly buying SP, I added terms to our delegation agreement that allows us to drop a delegation without refund if we feel it's being abused. The terms include what we consider abuse and instead of monitoring for abuse directly, we investigate reports of abuse by the community. The penalizing aspect of providing no refund for abusers seems to be an effective deterrent so far to the purchase of delegations by abusers. I think we've had around 4 or 5 reports of abuse, 2 of which resulted in the delegations being dropped.
I very much appreciate this candid discussion of your struggle to design appropriate delegation arrangements.
Given that such contracts can be standardized, can you not offer contracts that depend on curation split, designate self votes abuse, monitor those users, and pull the delegation if they self vote?
I realize this would be but marginally improved over pro bono delegations, and would diminish your ROI. However, if such delegations to manual curators resulted in upwards pressure on Steem, then the losses on the front end might be more than made up on the back - particularly if this became a model other delegators followed.
Thanks!
Have you seen this?
yes, I tend to at least scan-read most any post that makes it on trending or hot page related to a suggestion or attempt for improving steem in some fashion. Seems like a commendable way to try to delegate a moderate amount of Steem. But I don't see any method like that impacting curation results of the blockchain without changes to the code itself.
Not even if there were an order of magnitude more delegates?
The median payout is dismal. These moderate delegations deliver from ten to fifty times the median payout in one vote.
1000 such delegations would potentially reach 10k authors. With only ~20k posts a day, that's significant reach.
Thanks!