Curation problems (incentive, rewards and self-voting)

in #steemit7 years ago (edited)

Most of you will probably agree that there are problems with how rewards are currently distributed. Content quality doesn't seem all that important and it's apparent that the top 10% authors get the majority of the rewards. Also more and more people are becoming aware of the self-voting problem. The root of all these problems is curation. This article will cover ways to change curation incentives and algorithms to improve the user experience on SteemIt.

Current problems

  • Curation skill isn't monitored
  • Curation rewards are too small
  • Self-voting
  • Popular content creators are over-rewarded
  • Many curators are inactive (probably because rewards are so small)
  • Not enough curators are actively searching for under-rewarded content
  • There is no incentive to use flagging

Proposal:

Implementing the following systems might solve or improve some or all of the above problems.

1. Implement curation skill tracking

When you curate content, your curation will be rated based on the quality of the content and your action. When you upvote content and it ends up getting in the top 10 trending, your curation rating will increase dramatically. When you upvote content and it gets flagged heavily, you will lose significant curation rating.

The second factor that influences your curation change would be the curation rating of the curators that curated the content. If lots of whales with a bad curation rating voted on certain content and you vote on it as well, you will receive less curation rating or even lose curation rating compared to when you would upvote high quality content.

Implementing a curation rating will also add flagging incentive, because then curators actually get a reward, in the form of increased curation rating, when they flag bad or below average content when more curators flag that content as well.

Your curation rating will influence your curation rewards. The higher your curation rating, the higher the percentage of the curation reward pool you'll receive!

2. Increase curation rewards significantly for skilled curators

To solve 3 problems at once (low curation quality, curation inactivity and self-voting), the curation rewards for skilled curators should be increased significantly! Below is my proposal to achieve this:

Curation bonus pool

Next to the author pool and the currently existing curation pool, there could be a third pool implemented. This pool will make sure the curators that voted on content the earliest are very heavily rewarded. The way this would work is as follows:

The amount of shares in the bonus curation pool will be divided by starting at the first curator and ending at the last curator (if there are still shares left in the pool). When the bonus curation reward of a curator already exceeds the amount of shares that were assigned to the total reward pool by this curator, that curator won't get bonus curation rewards anymore. If there are still shares left after going through all curators, the remaining shares go back into the curation reward pool.

In reality this means that the first or first few curators will earn much, much more than the others when they vote on under-rewarded content that is older than 30 minutes. Voting on popular content creators in the first 30 minutes will earn less rewards, just like before (but still more than before). The whole curation meta would change so that you can't predict anymore which authors and which content will receive the highest rewards, so the content will be rated much more based on the quality of the content than the popularity of the author. Having more STEEM POWER also means having a bigger risk that you won't get paid enough curation rewards if you invest 100% voting power on certain content, so whales might want to spread out more unless it's exceptionally good content.

Example:

  • 10% of the reward pool goes to curation bonus
  • The curators all voted when the content was at least 30 minuted old (so no reversed-auction curation penalties)
  • All curators voted at 100% voting power at 100% strength
  • Curator 1 has 10,000 SP, which increases the total reward pool by $10
  • Curator 2 has 500 SP, which increases the total reward pool by $0.50
  • Curator 3 has 100,000 SP, which increases the total reward pool by $100
  • Curator 4 has 5,000 SP, which increases the total reward pool by $5
  • Curator 5 has 100,000 SP, which increases the total reward pool by $100

Total reward pool: $215.5
Curation bonus pool: $21.55

Curation bonus pool distribution:
  • Curator 1 will receive $10 bonus curation rewards (remaining curation bonus pool is now $11.55)
  • Curator 2 will receive $0.50 bonus curation rewards (remaining curation bonus pool is now $11.05)
  • Curator 3 will receive $11.05 bonus curation rewards (remaining curation bonus pool is now $0.00)
  • Curator 4 will receive $0.00 bonus curation rewards (remaining curation bonus pool is now $0.00)
  • Curator 5 will receive $0.00 bonus curation rewards (remaining curation bonus pool is now $0.00)

As you can see the first 2 curators receive a higher curation reward than when they would self-vote (because the bonus curation rewards are added to the default curation rewards). The other curators didn't earn more than a self-vote, this means that they didn't find the content fast enough and/or over-rewarded it!

Conclusion

When all these changes are combined, there is finally perfect incentive for curators to do a perfect job of curating content, because you'll get rewarded and penalized based on your curation behavior. I'm sure there are still flaws in this implementation, but it might inspire someone to improve upon it and/or come up with even better ideas.


Don't forget to follow, resteem and browse my channel for more information!

Sort:  

Congratulations! This post has been upvoted from the communal account, @minnowsupport, by Calamus056 from the Minnow Support Project. It's a witness project run by aggroed, ausbitbank, teamsteem, theprophet0, and someguy123. The goal is to help Steemit grow by supporting Minnows and creating a social network. Please find us in the Peace, Abundance, and Liberty Network (PALnet) Discord Channel. It's a completely public and open space to all members of the Steemit community who voluntarily choose to be there.

If you like what we're doing please upvote this comment so we can continue to build the community account that's supporting all members.

I don't like flagging, so I don't agree with you there, another thing I don't agree is content quality, this is subjective maybe you don't like what I like so how can you punish me by reducing my curation reward for upvoting something I like?

I think these are good points. I struggle with flagging. The only time I have used it is obvious plagiarism. I would be tempted if someone did a follow for follow on my blog because I really hate that. I just don't know where else I would use it, especially if it has many upvotes, as obviously those people find value in the post.

There's nothing wrong with flagging. Everyone is allowed to use their power however they want and if they want to flag to redistribute rewards, that's allowed.

Yes content quality is obviously subjective, but the average rating is pretty relevant. You will hardly ever get punished for upvoting something you'll like. You'll usually only get punished if you upvote scams or illegal content or something similar.

What would you consider to be illegal content? As for flagging, i don't think we can reach an agreement, i see it as detrimental and most likely to promote enmities and other problems, but everyone is free to think as he or she wishes.

I don't consider anything illegal, the government does. But fuck the government, right? :P

Well i'm sure some day you'll find some content harmful or some person harmful to the platform in some way and use the flag feature.

nice post! I completely agree with you.

Another problem to add to the list:

When flags 'are' used, they are sometimes used irresponsibly. Incidentally I feel that flags should always require the individual doing so to include an explanation. Its not like the community cannot see whodunit.

Incidentally, while we are talking about modifying curation rewards, might I humbly draw your attention to my post about data flowers? Its only semi-relevant but it is what it is (but might influence your thinking on this topic).

To bring this back on-topic however, I personally find the '30 minute rule' and the 'early curation rewards' to be rather conflicting. I have caught myself refraining from up-voting a post because its still minutes away from 30 minutes (I waited 2 minutes before up-voting yours...) - and thats not right.

I am not sure what I'd propose as replacement - but I am sure that such will come to me eventually. I am personally leaning toward suggesting that votes are anonymous up until 30 minutes or even an hour after publication.

As an aside - I really like any suggestion that encourages people to 'spread out more'.

However I find it tempting to suggest a slightly different curation rewards structure.

1st, = snipe
2nd-3rd = secondment
4th-7th = confirms
8th-15th = green light
16th-31st = etc.

These would have their pools mostly shared - so the 16th and the 31st are relatively equal, but there would be a small jump between the 16th and the 15th.

I find that I am rambling on, so I'd best post this before it gets longer. Thank you for seeking to tackle this topic @calamus056 :c)

I don't like it either that you can't vote whenever you want. Indeed spreading out might be the key to all this instant voting stuff without even reading the content whatsoever.

It's all really complex, but i'm pretty sure major improvements are possible without leaving it too vulnerable to attacks.

Absolutely. I also find that the minnows/dolphins/whales classifications seem to also affect who people hang out with. Small fish gravitate to big fish in the hopes of big crumbs.

And yet it seems to me that the 'smart' thing for whales to do is to engage in circular voting among themselves. I would 'really' appreciate it if you or anybody else could explain why I'm actually mistaken in this thinking. :c)

You're right, they're all asskissing each other for votes. STEEM POWER is way too influential when there aren't any systems in place that ensure quality curation.

This is regretable and a clear short-coming in the structure of the system.

It will be a challenge (for me) to think up any solution that would also serve the interests of whales. Still - it needs to be done.

Thank you for confirming my suspicions.

Changes to Steemit are certainly in the making, probably when they get out of Beta and clear up all the bugs as well.

Thanks for information keep in touch

Hey, thanks for the info and ideas. I'm still trying to figure out how all this stuff works. It's a whole new world, this Steemit. I appreciate the perspective!

I like the ideas. Not to name names, but some vlog entries that get $100+ and are full of nonsense do not deserve to be top content

It's sad to see so many under rewarded quality content. I've been actively looking to upvote new quality posts from minnows when checking on eSteem.

I wish Steemit will evolve and develop best solution for this issue.

It's sad indeed, and what's more sad is that the developers all have at least 2 million dollar accounts and are working on getting more users on SteemIt (HF20) instead of fixing the protocol and UI first.

Thank you for posting. You answered some questions and confirmed some suspicions for me. Every day is a learning experience on Steemit! Upvoted & resteemed.

"Content quality doesn't seem all that important" - my 6th day here and still same feeling ...

Some form of a Curation Rating may very well be a Great way to go... it seems to me, We need to All be equal at some level.. The Whales get Big Heads from all the Ass Kissn Sucka Fish . Same ShT different platform... know what i'm sayin' ... A curation rating > worked into the system may be "the Great Equalizer ... and make it feel less like a Pyramid Scheme .. HolySmoke! .. They also need a way to show a resteemed counter.... ReSteemed.. Following..

IKnowWhatYourSayin :)

Sounds complicated. Will the next 1 million users on Steemit have the faintest idea how this proposal is working or helping them?

KISSS - keep it short and simple and meant for the stupid

STEEM On!!