You are viewing a single comment's thread from:

RE: Been a while since I've posted about curation

in #curatemethis8 days ago

It's hard to tell who's manually curating or not, it'd have to be decided on in some different ways if you mean the DHF would reward them additionally compared to autovoters, votetraders/selfvoters through delegations, etc, which this place is swarming off.

I haven't been a big fan of the DHF, there are a few things I could imagine asking for help for there to help hive out but have attempted to scrape some rewards off of the posting reward pool for now in a smaller scale to see how things develop - for instance POSH, onboarding and curation&moderation.

Sort:  

This is exactly what I mean: reward them more than automatic voters.

The higher the manual healing activity, the higher the reward should be. Of course, as you rightly point out, it is difficult to establish indicators or metrics that make it easy to measure that kind of activity. I don't know if in your observations you created your metrics (of which this place is rife) that show you numerical results. I don't know, maybe it's 70% or 80% of all daily posts on Hive😅.

There are ways to at least notice the automatic votes. By consulting Hivestats, you can log in and perform a search for an account you suspect of receiving mass votes. If a recent post receives votes in less than five minutes, this may be an indication of automation. You can see that in the table called “pending curation rewards” in the “voted after” column. This column shows the difference between the time the post was published and the time the vote was cast. Maybe 10 or 15 minutes, it is a matter of analyzing it carefully and drawing conclusions.

But be careful, this does not mean that everything is automated. Let's take the case of a publication that has an estimated reading time of 6 minutes. If I, as a manual curator with a voting power of 1HBD at 100% of the hand, notice that the publication is very good and it takes me 2 minutes to read it completely, I would undoubtedly vote for it immediately. That also needs to be taken into consideration when performing curation. This can also be analyzed with HiveSQL or with the API.

The point is that you could create clusters and thus sort publications and curation activities into different levels, for example: from least to most automation. But the goal is always to make those distinctions to have good filtering and to dig for those publications that focus on manual curations. For that reason, I state to you, “I don't know if in your observations you did your metrics.” This example I've provided you is one way to do that.

But let's focus on getting activity in publications more related to manual healing. Something called “Proof of curator (PoC)” could be created. PoC, by definition, would be a system designed for the purpose of incentivizing and rewarding manual curators, taking into account their degree of participation in content selection. One could also include curators' comments (the **commentrewarder **project does this, but not under the name PoC and in a different way) or, also, reasoned arguments supporting the criteria by which they cast their vote for a publication, but I think the latter would add more complexity.

And it also occurs to me, apart from designing a system like PoC, to make a change in the curation pool of publications. I don't mean “Rewards Distribution”, I mean reducing the amount of curation rewards for auto-vote and automated votes, including delegations. If a user self-votes with a voting power of 1 HBD his own publication, let's say it is halved and he will receive 0.50 HBD. This, in practice, would be equivalent to wasting your voting power, when you could have used it in another way, by curing manually. If you vote other publications and make sure you do it right, your voting power will remain at 1 HBD without suffering any reduction for self-vote, except what deducts the mana for each vote cast.

However, this is also compatible with the delegations. If I delegate voting power to a specific project and that project autovotes my publications, for example, say the first time, it reduces the amount of rewards to half proportional to the voting power I assign to a delegation.

And the other aspect would be automated votes. This could also be included in the reward pool. Automated votes could “continue to exist”, but over time their voting power would be penalized for not betting on manual cures. Of course, all of this I think needs to be thought through in more detail.

Total, the idea that occurs to me is to add to the “rewards pool” a system of decreasing rewards that considers the aspects I mentioned or what you have observed for a long time, or else, consider the Proof of curator(PoC) as a proposal in the DHF.I am curious to know if you have come up with any other ideas after reading my comment. I'd be happy to read you. And I apologize for the great length of the text.

I don't think we can directly reduce self-vote/curation returns of autovoters at the blockchain level, it would become kind of messy. I prefer the other route by increasing the rewards of active authors and active curators and I'd be open for additional tools and methods to help with that. I think most of those rewards would have to come from the author reward pool or DHF since curation is hard to adjust aside from downvoting where one would deem votes to be mainly auto and author undeserving as well.

I don't think we can directly reduce self-vote/curation returns of autovoters at the blockchain level, it would become kind of messy.

I also thought about it, it would be messy, it could generate complications that would affect us all.

But I think the second way to increase the rewards could come from the DHF or author (you have to determine the one that generates better results) is easier to implement and manage in the short term to incentivize and reward manual cures and the creation of quality content.

You have touched precisely a nervous and worrying point: “a lot of autovotes are just bloat and we the manually active ones are carrying the weight of them more often then not getting the same returns as us”.

One of the ideas of @commentrewarder is to in a way also reward manual voters more indirectly, because manual voters may also often leave a comment. This is something I wanted to post about next, so if authors were using the project more to forfeit part of their author rewards to the comments, then also only upvote the comments that seem genuine, they would in a way reward the curators as well.

I'm all for rewarding manual activity more, a lot of autovotes are just bloat and we the manually active ones are carrying the weight of them more often then not getting the same returns as us, so author rewards is one way to beat them at that at the inflation level.