Elimination of Curation Rewards

in #curation7 years ago (edited)

"Eliminating curation rewards would be saying that curators should perform unpaid labor for the authors and the steem power holders." - @remlaps

There are a lot of reasons not to eliminate curation rewards:

  1. Curation rewards are currently one of the only reasons to power up / remain powered up.
  2. There are a huge amount of users that are actively involved in the platform through curation activities (developing bots, curation trails, guilds, manual curation, etc.).
  3. Curation rewards provides a financial incentive for users to spend a very significant amount of their time discovering good content.
  4. The goal of the platform is to reward users for their contributions for the platform, and curating is a form of contributing.
  5. Lots of users find earning curation rewards fun.

To a large extent, we are going to get what we curate. If the stakeholders are allocating a significant amount of rewards to photography, we will attract more photographers. As more photographers are attracted to the rewards, competition will drive higher quality photography. (That's the idea..)

Unfortunately the curation reward system that we have today has been shown to incentivize the wrong type of behavior:

  1. A large portion of voting is taking place without users even reading or evaluating the content. While bots could be used to perform a lot of automated evaluations of content that would not be easy for humans to do on a large scale, that is not what they are doing today. Most of them are designed to maximize earnings from the curation rewards game.
  2. Many voters calculate whether they will receive a good curation reward, over whether they 'like' the content.
  3. Lots of quality engaging content does not receive significant rewards, because it is not expected to receive a lot of votes.
  4. Established authors who are producing sub-par content still get lots of upvotes, because they are on auto-vote lists and are expected to receive lots of upvotes.
  5. Very few people vote on comments, because there is no expected curation reward.

Curation rewards are also not very appealing for new users:

  1. Voting before the first 30 minutes is bad. (Well, sometimes. It depends.) This adds confusion to what would otherwise be a natural/organic process.
  2. In order to maximize curation rewards, you have to vote on 40 posts per day. Every day. 365 days a year. If you don't, you will be losing a percentage of your stake to the users with auto-upvote bots. (That sounds like a job!)
  3. The amount that you see in the "Potential Payout" is not the amount that the author gets. It is split 75/25 between the author and curator. Although the author can get more if the curators vote within the first 30 minutes. (You lost me at The amount that you see in the "Potential Payout" is not the amount that the author gets.)
  4. To earn anything significant from curation rewards, you have to first have a lot of SP. (Wait, I have to give you money in order for me to make money..?)
  5. Voting on your favorite content is likely not in your best financial interest.

The benefits to removing curation rewards would be:

  1. It would remove a confusing aspect of the platform that is unappealing to typical new users.
  2. The decision to upvote would simply be based on what content you liked and wanted to reward.
  3. The upvotes would primarily be from users who were actively engaged in the platform.
  4. The motivation to have bots for purposes other than evaluating good content would be greatly diminished.
  5. Less voting stake would be used, making upvotes from users who do still vote more powerful.
  6. Regular users' perception of the platform would improve.

But wait, shouldn't curators get paid for their work?

If they do their job well - they will.

The value of SP will come if we can build a platform that attracts and retains billions of users, in a way that keeps them actively engaged in the site. With a large and engaged audience, that gives us the ability to build a revenue model (such as advertisements) on top of those users. That revenue can be turned into passive earnings for all SP holders.

It is up to all of us on Steemit to direct the rewards in a way that is beneficial to the platform, to give us all the best potential for return on our investment.

When you curate, you are participating in the community's decision on how to best allocate our limited rewards pool. By directing the resources towards the things that bring the most value to the platform, we are collectively deciding on what we feel is the most likely to give us the best return on our investment.

It is entirely up to each user how they want to use their votes, but paying them to do so is not necessarily going to incentivize better behavior.


Image CC0 public domain from pixabay.com.

Remember to vote for witnesses!
https://steemit.com/~witnesses
If you aren't sure who to vote for, check out this Witness Voting Guide.

Sort:  

One thing this debate has revealed to me... is that lots of people seem to think the best way to increase the value of their account... is by behaving in ways that earn them more STEEM rather than by behaving in ways that raise the price of STEEM.

It's like everyone's fighting for a slightly bigger piece of a really small pie, rather than trying to grow the pie. Maybe everyone's just assuming other people will grow it, so they're just jockeying for position. But IMO, this pie could be huge if everyone stopped trying to get more STEEM and started trying to make STEEM more valuable. Even the minnows on this platform right now will be whales over night if Steemit grows to be the size of Reddit or beyond.

Whales wanting more STEEM right now makes no sense to me. Even as a small dolphin... I'd way rather see the price of STEEM increase than see my stake increase. I don't vote for curation rewards. I vote for people who I believe are increasing the value of STEEM... call me crazy lol but that's how I think I'm going to make the most money.

One thing this debate has revealed to me... is that lots of people seem to think the best way to increase the value of their account... is by behaving in ways that earn them more STEEM rather than by behaving in ways that raise the price of STEEM.

BINGO!

One thing this debate has revealed to me... is that lots of people seem to think the best way to increase the value of their account... is by behaving in ways that earn them more STEEM rather than by behaving in ways that raise the price of STEEM.

Some of this is tragedy of the commons. Most of it is simply that most people who participate in any endeavor are short sighted because most people period are short sighted.

Yes or alternately (and I'd argue more correctly since there is no existing resource that is being overgrazed and therefore spoiled; instead this an attempt to build something) a public goods problem. That is addressed by setting things up so that incentives are at least somewhat aligned between short-run individual interests and longer-run social interests. In this case it is very questionable whether the existing system comes even close to that.

setting things up so that incentives are at least somewhat aligned between short-run individual interests and longer-run social interests

Word up !

Comments like this make me dream for an option to... resteem comments! How great would be such a future?

Well said!!

Amen. There a lot of truth in that comment which I think is worthy of its own post.

1000% !!!

Good point 😅

To the point. Great analogy!

well said. I only vote for things I like and don't really think about the curation rewards. I am concerned about the price of Steem continually falling though.

this pie could be huge if everyone stopped trying to get more STEEM and started trying to make STEEM more valuable.

How do we make Steem more valuable? That's the real question. Increasing the user base hasn't seemed to work. I joined when there were 60,000 users and it has doubled since then and still has not increased in value. It will take a lot more users and things to use Steem for to increase the price.

To a large extent, we are going to get what we curate. If the stakeholders are allocating a significant amount of rewards to photography, we will attract more photographers. As more photographers are attracted to the rewards, competition will drive higher quality photography. (That's the idea..)

I try. I have original work. One time, one got above $100, getting visibility, setting an example of quality to get rewarded. Then it got flagged for "too high payout" and "not being valuable to Steemit" or some such nonsense... That's while another post that same day by me, was not my original higher quality content, got close to $200, for cannabis.

Next, another time a post on quality important truth, morality, got above $100, that was too much again, knocked down with a flag. Other posts don't get this treatment. I have pulls from the blockchain to prove it.

Then again, another quality post on morality, this time it was at $70, so flag it down to $40. Hehe.

Visibility goes down, and attracting people to do the same is diminished.

I agree with how things should work based on how companies who produce things do succeed, and that is with a recognition of the importance of quality. But some people work against quality, because they don't think there is "quality", or that the content is "not valuable to the platform to attract new users", and other excuses.

So rather than spend time, effort, and creating higher quality original content, why would someone do that when it gets "punished" each time?Other posts, not original, less high quality that I make, can go to those payouts and they don't get flagged. Funny isn't it...


Curation can be free. If you care about the platform, you do it, for whatever content you want to promote. That's how I have done it. I don't look at curation rewards, although at some point I tried to do the 20minute mark thing. Vote for it early, later, whenever you want.

Early helps the author more, but so what if you're later? It shouldn't be about your curation rewards, but about valuing content for content. That's what needs to get into many people's understanding. Bots cant evaluate content for content. Consciousness is required. So what if you only do it at 9pm once a day for everything that day and don't get the curation rewards. Bots can be used to vote for trusted authors, then unvote manually on review once a day, etc. There are indeed better ways to use bots, but I'm still for consciousness being what drives actual social media.

I'm with you with everything you said about curation/voting. Flagging is a separate issue on it's own, which warrants it's own discussion.

The social contract is quite flexible round these parts you know?

The examples you've given are very much determined upon your lens or perception of the world. You say cannabis is less important but a 'quality post on morality' is...

Who's supposed to judge that and give it context?

Who's supposed to judge that

an expert on morality, ldo (though i do believe krnel is one of the best posters that lives on the front page_

Well that's what I mean. A post written about cannabis and one written about 'quality important truth, morality'.

The whales are mainly dudes that mine crypto currencies.

One likes smoking weed another thinks posts that are more like academic articles are less valuable to mass adoption.

In the same way they make their preferences known through their lens. Krnel makes his perceptions known in this response.

Hows a whale that enjoys smoking a bit of weed going to conform to such highly individualised and personal standards?

I'm not a fan of the 'weed' stuff, but my personal views are that we should find a way to encourage all types of content that are going to bring active and engaged users to the platform.

I don't think the issue is cannabis posts being good or bad.

I think it's that someone actively punished @KrNel by taking away some of his reward for a post that he highly valued. If it had happened to his lower quality post too, there would be less issue, but instead, people are deliberately trying to influence him to post things they want to see, and the form of influence isn't just carrots.

They include a stick, by removing rewards, and the biggest issue is that @KrNel is upset because his most favorite post he wrote got punished, while the lesser post he wrote got rewarded.

This truly irritates a writer, because instead of getting "commission" type influence, where he can choose to either write about his favorite topic, OR a topic he knows will make lots of money, he's instead forced to only write for the money, because people will actively ruin his profits by flagging what he just wants to share with people, despite any potential popularity payout.

I mean, we all know that some types of content won't make much money, but we publish it anyways, just to share the idea. It's fine if it doesn't make lots of money, but when it DOES make money, but someone else feels the need to remove that money, it's just a heartbreaking feeling.

Not of "oh no, I'm sad I don't get money", but a feeling of "This community is broken. Why am I punished for work that people liked and upvoted just because some rich person didn't like it? I deserve the money that people give me via votes."

It's the same as someone ordering food at a restaurant.

If it's good, people will pay for it and tip.

If it's bad, people want their money back. That's fair.

But what's unacceptable is for people to take more money than they spent, robbing the restaurant of money they collected from other patrons.

Same with flagging for these frivolous reasons.

Bots can be used to vote for trusted authors, then unvote manually on review once a day, etc.

With the current system, there's a big advantage to voting early so as an optimisation problem, this is not going to be the best way and so will not "naturally" attract people. And it's not just about curation rewards, voting early draws more attention to good posts, so helping it snowball. For posts you believe in this is desired.

There are indeed better ways to use bots, but I'm still for consciousness being what drives actual social media.

I'm coming round to this way of thinking and trying to think how bots can fit in better. The argument I make above is not necessarily the best way that Steemit can be, but it makes the most sense with the way it is. A change, probably a fairly core one, is needed to encourage more conscious engagement. That is if it's agreed there's an engagement problem. Some prominent witnesses I spoke to did not think it was 😕

I believe Curation Rewards could exist, but I do think the algorithm has some issues that do incentivize dog piling, finding people that are followed by bots and voting on them whether you read it or not.

In fact, the "shouldn't curators get paid for their work?" Payment can be viewed in several ways. If you are encouraging an author that you really like to produce more content, then their content could almost be viewed as a form of payment.

I do believe there can be a happy medium, but I believe the current algorithm places too much emphasis in such a way they results in dog piling more than quality. Quality still happens, but with bots you end up with people that always get up voted by the same bot. Some of their content likely should not be worth that much. What about the gems out there that are not an established author that are ignored, because your daily bot votes used up how many votes you are going to place for the day?

I look at steemit/busy.org a bit different than a lot of people.

I view my steem power more like the potential guaranteed income I have to purchase goods on steemit for that day. It doesn't actually cost me anything in reality, but it is there for me to spend like a virtual currency.

How do I spend it? I walk into a store full of content, and just like walking into a bookstore, or music store, or any other store I walk PAST the things I don't like, and I vote on the things I do. Me powering up increases my ability to reward content creators I like (posts, and comments) and being able to reward them better means I'll likely get more content I like.

I personally do not use a bot, though I am more than capable of writing one. I actually read the things I up vote. I will occasionally up vote things I may not agree with completely simply due to the sheer effort and thought I can tell the content creator put into the work.

I believe the current curation system has flaws. Yet so does many other things. We are working through those.

I do think that algorithm might be worthy of some tweaking...

Also as far as rewards from curation... Unless you have a substantial amount of steem power those rewards are very very low even with the current system.

https://steemit.com/curation/@sigmajin/an-opponent-of-the-exponent-making-the-case-for-vshare-linearity

Take a look at this, it goes back to some of the stuff we talked about in the downvote thread, and i think its a much more workable solution to curation problems

I am in favor of doing both. (Removing curation rewards, and switching to a more linear rewards formula.)

Obviously, im not. But i will say that even if I were strongly in favor of eliminating curation rewards, i would not be in favor of doing the two simultaneously.

I think that there is a very strong case to be made that a more linear reward distribution formula could fix many of the percieved problems with curation rewards. IMO, we should at least see what effect the changes have before we throw the baby out with the bath water.

I'd go in reverse, removal than adjustments, but I do agree that they shouldn't happen at the same time.

I think the bigger problem isn't how the rewards are distributed, it's the fact the motivation of each voter comes from the desire to earn a reward. It seems almost like a conflict of interest. I want popular/trending content on the site to be elevated because people find it interesting, not because they believe it will provide them with the greatest reward.

When considering the act of voting, you shouldn't be considering yourself, you should be considering only the content on which you're voting.

That's what I believe at least :)

Whatever you think which one is bigger, I think both of them are big problems ;)

Good comment! I just wanted to make the observation that people are going to think of themselves when voting anyways, even if what they are thinking is that they could later be asked to justify the vote because it can be traced back to them.

Well said. I totally agree.

One interesting thing though - technically the current algorithm discourages "dog piling". If you vote on a post that already has a high payout, you get less than if you vote on a post before it receives a high payout.

The problem is probably better described in that it encourages voting behavior based on the predictions of what posts will receive a high payout. In theory this was supposed to be one and the same as "good" content, but in practice these two things have turned out to be different.

Loading...

Craig still buying high and selling low. Nothing new under the sun :)

160% inflation rate

Thank you for leading this discussion. It is very important. We should consider some means of evening things out in terms of the voting power. I don't know if eliminating curation rewards is the best way, but it should be on the table along with any other options. Let's keep this discussion going and moving it forward.

You make lots of valid points here... it's the actual implementation of a fair and balanced curation system that is the hard part. I'm a newbie here, but have been part of dozens of "this kind of gig" since the very first revenue-sharing , user-generated-content sites back in the late 90's (Epinions, Themestream, WrittenByMe, etc) and they almost all fail, mostly for reasons related directly to curation.

So far, what seems to work best (again, hard to implement) is a system that awards the greatest curation influence to those who are the actual "community builders." In other words, those providing consistent high quality content also hold the most influence over curating to ensure that quality content rises to the top, and junk/spam/clickbait sinks into oblivion. Steemit already self-regulates because there are "sliding scale" controls in place... but maybe there also needs to be one that values an "actual eyeballs" upvote higher than a bot upvote.

The other source of many failures seems to be the broad underestimate of the impact of a simple piece of psychology: Human Greed. We worry about trolls, but there's a FAR larger, more pervasive and destructive element out there whom I've come to think of as "Money For Nothing Seekers." They are the ones who respond to someone (well meaning) saying "I am making some extra money social blogging on Steemit" for ALL the wrong reasons... at least "wrong" in terms of community building and creating lasting value and quality. Yes, they exist. Yes, they number in the millions. They can take an unprepared site down in a few months. I think most of us know this, but we tend to sweep the magnitude of their impact under the rug... but that would be a grave mistake, in my opinion.

Anyway, getting back to curation (and keeping in mine the desires for decentralization, support for alt viewpoints and no censorship), it seems important that there be system safeguards that (a) recognize when someone is trying to "game" the system-- most likely through automation-- both on the content and curation sides and (b) uses a sliding scale of sorts to render "useless/empty contributions" less and less valuable and visible.

One final thought, before I end this dissertation... IF the long term roadmap is to create a decentralized alt social network that potentially serves tens of millions, seems that would be hard to accomplish without promoting Steemit to a large audience of "outside eyeballs." That said, I think it would be wise to make sure the internal systems-- rewards, curation, social features-- are rock solid just with the existing user base as it grows slowly and organically, before any sort of large scale appeal goes out there. In other words, don't throw a big party till you're sure you know where to get enough food and beer and tables and tents in case it rains and volunteers to make sure nobody drives home drunk...

@denmarkguy, I could not agree more! Upped your comment for the emergence of commonsense, as rare as rocking horse poo, as I like to articulate.
There was recently a suggestion to increase the value of a captcha verified vote by 10 times and an eyeballed (human) with a comment added by 100 times. Bot votes to stay as they were. The realities of this were horrendous in absolute terms.
Perhaps the converse would work better - a bot vote whether private bot or service bot (doesn't that sound funny to you?) if it can be verified as such could have a value of, say, 2% or 5% or 10% or 20% - I am not qualified to rationalise the exactitudes - thereby giving the real content followers a far higher say in value of content. The bots really do little to enhance any attractiveness of the platform other than 'Games' players but the voting system is clearly a key factor.
The bots may cost money to run - but at the expense of the big picture - they are feeding off the good stuff and are, don't shoot me, acting as parasites.

The problem is that at a blockchain level (where the votes are calculated) there really isn't a way to distinguish between humans and bots.

A capcha system is the only such proposal, but I am not aware of a way to implement one within the blockchain. Plus, even if we could - it would add a mental burden to the act of voting that would be discouraging to a lot of users.

I prefer the elimination of curation rewards, as it essentially eliminates the incentive to use bots in the first place, unless they are the type that can evaluate aspects of post quality (like plagiarism) that are hard for humans to do.

@timcliff I was curious about your thoughts on a guaranteed payout of the curation reward pool (say 50%) for all whales above the arbitrary 250MV, on a prorata basis according to a weighted average of SP held in exchange for voting power eliminatin (like preferred shares). I proposed this to @snowflake earlier and he's not a fan.

My thoughts are:

  • (i) if they are gaming the system, they are probably getting 50% of the reward pool anyways , and
  • (ii) the proposal at least creates an incentive to keep increasing your SP as a whale if you are looking for a larger portion of the pool.
  • (iii) it keeps curation intact

What are your thoughts?

If you divide the existing curation rewards up across everybody (those curating and those not curating) then those who are actively curating will get less than they are today.

The 250 MV threshold proposal has been put on pause for now, but if we were to go down that path eventually - there is a very complicated equation to solve as far as making there enough incentive for whales to keep the SP above 250 MV powered up (without being able to cuate/vote), without paying them too much at the expense of the network.

If the 250+ users were only getting 50% of the curation rewards and had to give up their influence - they would be incentivized to just use a bunch of accounts < 250 - because they could make more that way and use all of their voting power.

there is a very complicated equation to solve as far as making there enough incentive for whales to keep the SP above 250 MV powered up

I went through the math on this... i think on the can opener post.

if you found a way to exclude the steemit account (which, i don't take it as a given that you could. Or that TPTB would be cool with it even if it were possible), and you limited it to just 1GV and above whales, you would be able to offer about 50K sp a week divided up among about 150 G vests, which amounts to around 260 SP a week per Gvest if divided porportionally. This amounts to around 2.6% per year non compounding.

That gets cut in half to 1.3% APR if you include the steemit account. I didnt go through and do an exact count, but my spitball guess based on some quick tallying on steemdb is that if you added in all the accounts between 250MV and 1GV as well, you would end up with something like 1.7% apr not including the steemit account and just under 1% including the steemit account.

You have got my vote, not that it counts! I really do appreciate the enormous amount of effort you are putting in to ensure the successful future of steemit.com - I hope that there is apt reward for you! Someone like me asks a question and they do not get a reply for a minnow, they get a reply for a human.
Respect aplenty!

There are a huge amount of users

No there isn't

Laugh. True. I was speaking relative to the size active user base. Good point :)

But wait, shouldn't curators get paid for their work?

If they do their job well - they will.

The value of SP will come if we can build a platform that attracts and retains billions of users, in a way that keeps them actively engaged in the site. With a large and engaged audience, that gives us the ability to build a revenue model (such as advertisements) on top of those users. That revenue can be turned into passive earnings for all SP holders.

Every word there also applies to author rewards. Should we eliminate those too? IMO, Paying someone for their labor means that they should be paid in proportion to the work they do or the value that they add. A generalized reward that's available to every steem holder does not qualify as payment. Lanier addressed this point, too:

The usual counterargument to that is that they are being paid in the sense that they too benefit from all the free stuff and reduced-cost stuff that comes out of the system. I don't buy that argument, because you need formal economic benefit to have a civilization, not just informal economic benefit. The difference between a slum and the city is whether everybody gets by on day-to-day informal benefits or real formal benefits.

Personally, I am an advocate of letting authors set their own curation percentage and use that as another lever in the competition for up-votes. New authors could set their curation percentage high to attract voters. Established authors could throttle it down and keep more for themselves. Although, I imagine that might be difficult to implement with the current blockchain design.

Every word there also applies to author rewards. Should we eliminate those too? IMO, Paying someone for their labor means that they should be paid in proportion to the work they do or the value that they add.

Probably the huge difference is that paying for content drives the quality of content up, as outlined in the photography example in the post. Paying for voting does not drive up the quality of voting, as described in the post.

Personally, I am an advocate of letting authors set their own curation percentage and use that as another lever in the competition for up-votes. New authors could set their curation percentage high to attract voters.

It's not a bad idea, and one that was brought up during the HF discussion. It would be interesting to see something like this played out, although it would add a lot of complexity to the system. And just stating the obviously (not that it is a horrible thing) but it would essentially be a form of vote buying / bribery.

Paying for voting does not drive up the quality of voting, as described in the post.

I'm not totally convinced of that. streemian, steemvoter, and autosteem have enabled hundreds of thousands or millions of votes that would not have been cast otherwise. That probably wouldn't have happened without the incentive that curation rewards provide, and we have no idea what the platform would have been like without them. Also, we have yet to see how things play out in the long term. Many people are still unfamiliar with the rewards rules.

I have thought that it would be interesting to see what happens if bots and guilds all shut down voluntarily at the same time for a couple of days. My guess is that a small number of authors would like it a lot and most would not. I doubt if we'd be able to coordinate that experiment, though.

And just stating the obviously (not that it is a horrible thing) but it would essentially be a form of vote buying / bribery.

Maybe. Or maybe it's just an inducement to evaluate a post that a curator might have otherwise ignored. I doubt if many people would intentionally vote on poor quality posts just for a tiny curation reward, especially knowing that it's likely to be a wasted vote because other voters won't do the same. Either way, as long as it's transparent, I don't see it as a problem. The complexity would be my bigger concern, and may well be a "show stopper."

Upvoted for visibility and to reward quality of presentation, not necessarily agreement/endorsement.

Thanks @smooth. As always, your level of caring for the platform and your ability to review the pros and cons of the various changes being proposed is an invaluable contribution!

Coin Marketplace

STEEM 0.30
TRX 0.11
JST 0.033
BTC 63968.82
ETH 3136.80
USDT 1.00
SBD 4.28