Details on Proposed Comment Reward Curve
In preparation for the next hard fork our team has been working on an alternative to the N^2 reward curve that results in the top posts making significantly more than everyone else.
n log(n) curve and blue representing the n curve.
We found it was easier to understand these curves by taking the derivative, which is what we graphed.
Ideal Curve
In a world with honest people who don't vote on themselves to get "free money for nothing", a simple linear curve, aka n would produce a 1 share 1 vote proportional payout. This is the blue line and shows the ideal situation.
Realistic Curve
Unfortunately, we live in a world where people will attempt to game the system by voting for themselves. If everyone voted for themselves then the result would be simple interest payments and have no economic impact. We believe that groups are more honest in aggregate than individuals. We also believe that whales (accounts with over $500K Steem Power) have more to lose and are easier to police than the multitude of smaller accounts.
The green line, n^2 / (1 + n), shows a blend between n^2 and n where for small numbers of votes the reward schedule is closer to n^2 and for larger values it approaches n.
With these curves, any single whale in the top 20 will have the power to vote with n weight such that the reward is 75% linear. If two or more of the top 20 whales concur the the final result is over 90% linear. For all content on the top of steemit.com today the reward payout would be 99% linear under this curve.
n log(n) Alternative
In the past we suggested something closer to n log(n), but as you can see from the graph, it takes longer to reach a linear approximation and then it overshoots and remains biased toward concentration of votes.
Deployment Proposal
We propose to role out the new curve for the comments only because comments do not have curation rewards in the next release. If it is successful for comments, then we will have to derive a new curation reward equation based upon this curve before we can adopt this curve to top level posts.
All equations shown are normalized by removing constants we will scale by.
Please give your feedback and let us know what you think.
What about the idea that comments do not need their own reward pool in the first place? Is this simply not open for debate? The sentiment has been shifting about this topic over the last few weeks.
I vote no on a separate pool for a variety of reasons.
If the argument is that it will increase engagement, then I don't believe that's necessarily true. It'll be hard to increase engagement on posts via comments when there aren't many humans actually reading the posts in the first place.
These rewards will take a large percentage away from current posting and curation rewards (which is still the plan, as far as I know), to the tune of 38%. Curation rewards are only paying out about 12% of the pool right now. A further 38% cut will bring that down to around 7.5%. How exactly does this incentivize users to curate better and not simply lead to more automation and "sure bets" from such automation?
Having a separate rewards pool for comments adds another layer of complexity to an already complex system. New users already have a steep learning curve when using this platform. How exactly does this fit in with Steemit Inc.'s notion of KISS, which was stressed in the road map announcement?
Having a separate pool for comments - which are already not read much, due to the parent posts not being read much, overall - will likely invite more spam and potential abuse, since they're not even monitored by the user base as much as parent posts are.
If the plan is to increase engagement, perhaps the focus should be on growing the active user base beyond the ~800 daily authors and a couple of thousand active accounts, which include a large amount of alt accounts, automated voting accounts, and bots.
Basically, the problem isn't a lack of comment rewards. The problem is the lack of a large enough, active, human user base to make the platform function as intended. Instead of creating a separate pool for comments, why not treat blog posts and comments the same? Same curves, same curation rewards, same time frames for voting. Wouldn't that make more sense and actually adhere to the KISS principle?
I strongly disagree.
With a comment reward pool, steemit can become attractive to a much more broad community, extending out from "bloggers" to readers and consumers who can also make money just from adding commentary.
They can do that now, without a separate pool. And flattening the rewards curve on blog posts will invariably free up more rewards for comments, so that proposal alone could lead to better comment distribution. If comments and blog posts were treated the same way, that would also help the situation.
Ultimately though, a comment pool will not attract users if nobody knows about it...or the website itself. Unless there is going to be a marketing campaign to accompany these changes, they won't matter much. Steemit has a lack of active human users and an apparent lack of interested potential users. How does this get addressed and resolved by simply throwing a larger share of rewards at comments?
Never mind the fact that your "strong disagreement" didn't actually address any of the potential issues I raised. With what do you "strongly disagree?"
Basically, this is just the same wishful thinking that everyone seems to buy into with each new proposal. How, precisely, will a comment rewards pool grow the user base, given what we've seen so far with the user base, development/changes, and the marketing of the platform since it began? And how will the reduction in posting and curation rewards negatively impact everything else? Why do people only look at the potential good (which is largely unproven) and completely ignore the likely bad (which we can actually observe)?
Separate the rewards pool and see what numbers it yields & response that comes from the community. BETA is trial and error not assumption and deny correct? Create a Reward Pool and see what comes from it. If its good then so be it, if its bad, then so be it. Let the people decide, regardless of how small the number.
-V.O.T.U. Pirate King
We have already tested the current reward curve and the results were 1% rewards for commenters. With a result so low, yes I strongly disagree with your whole rejection of experimentation.
Yeah...that's why I'm not for keeping the current rewards curve. It would help if you actually read what I've been writing instead of just insisting that you "strongly disagree" without actually explaining what exactly you disagree with.
I made four specific points. Maybe you can address them one at a time, if it's easier?
I agree there is a need to attract more than just bloggers to the site. But, I disagree that a new reward pool for comments needs to be created to achieve this.
I have made probably about $100 from comments since joining. And that is without attempting to, and only engaging in conversation when I want to. One doesn't need a separate reward pool in order to be rewarded through comments.
I worry that with the creation of this second pool, people will not be engaging in conversation because they feel passionately about the topic and want to weigh in. Instead, there may be a wave of people attempting to game the comments reward pool as there currently are those attempting to game the post and comment pool. We may also get a lot of lazy, spammy comments, written purely with the intent of gaining a monetary reward, rather than inviting meaningful discussion.
It could result in less authentic engagement, which will make Steemit less inviting to newcomers.
This is my worry at least. I do hope that I am wrong.
Reddit was grown to be the top 23 most visited website without any monetary incentives, so I think the evidences shows the comment reward pool isn't necessary to attract people to Steem.
Also in the early days of Steem people could make a simple introduction post and make in the hundreds and even thousands of dollars rewards but a lot of people just wouldn't join Steem and make the no-brainer move of making those posts. (I can name many of my friends and I heard of many more)
There is no need for separate reward pool to attract people on Steem. Are there good reasons for separate pool? Maybe.
Trending comment would need a section on themselves or how are we suppose to find those great comments and conversations.
I don't know the best way to go. It's a tough call either way.
except that this site has been peddled as a money-making site
therefore thats what the expectation is
The same thing you are supporting here can happen with a single pool where comments are not absurdly penalized by the fact that the highest-rewarded posts get multiple whale votes resulting in potentially quadrillions of times more weight than comment posts with even a decent amount of non-whale voting. (Of course the same applies to posts as well.) Flattening the overall curve goes a long way toward alleviating a lot of these problems and should be tried first.
I think a separate reward pool for comments is much needed considering the current ^2 distribution rule. However, if you want to make these changes, I strongly urge you to do them one at a time so we can see what effect they're having on the system.
It's reasonable to think that separating the comment reward pool from the blogging reward pool will give people more incentives to vote on comments. However, at the same time, you're going to remove curation rewards from comments, giving people less incentive to vote on comments. It might be a good move to make both changes, but do them one at a time so you can know the results of each.
I don't think we'll need incentive to vote for comments. People already vote for them at a cost to curation rewards, so those comments that do get votes would get maybe 38 times what they get now while most people will still get 0 - which is ok because we don't want to encourage spam comments anyway.
I think this is true to an extent, but today people vote on blogs posts much more often than they vote on comments. Compare that to the trending reddit posts where it's very common to see comments that have almost as many or as many votes as the post itself.
While it will be significantly higher, I don't think it will be 38X for a few reasons:
And this cannot be achieved through the same reward pool? You have not explained/argued how or by what mechanism exactly more engagement happens because of a separate reward pool.
Because with separate reward pool comments compete against other comments, not posts. Allocating 38% for comment rewards instead of current nearly 10% is enough economic incentive to predict significant improvement in the quality of discussion.
Its audacious to say comments aren't worth as much as posts, and I think everyone agrees on that, at least I hope so, but splitting up the reward pool is treating comments as a different system and that will create more problems because of unnecessary complications. The quality of discussion won't improve simply because of a separate reward pool. You have argued not so much that it's necessary to have a different pool, you have argued mostly that it's necessary to treat comments with the same system/standards that posts get treated with to some extent, which I half agree with, because as long as posts are more valuable than comments the incentive for posts outweighs comments, even if it's marginal. Are we arguing that comments are not worth as much as posts or that they should not take from the rewards posts get, because I don't understand how a separate system is indeed the solution and not an unnecessary complication, wouldn't a better solution be to treat comments with the exact same standards the posts get, in turn removing a whole bunch of extra processing and speeding up the system, and in turn making the curves the same will not give any incentive to game one aspect over another.
Well I thought it was obvious. How about 38 times more rewards?
https://steemit.com/comments/@beanz/why-should-commenters-make-38x-more-rewards
I am all for removing curation rewards, but that doesn't mean that comments should be less valuable than posts, as many have pointed out they are not of the same opinion that devaluing comments is necessary. I believe that comments can and should be treated as posts (when curation rewards are removed of course) and should compete in the same reward pool then.
-nesting limit-
@beanz
I haven't read the whole whitepaper but I have just finished the section on payouts and rewards and I have a general understanding of how SP works now. I agree, curation rewards are necessary to give SP a use value like you said to curate content more effectively, therefore I believe that yes we need to curtail curation rewards to allocate the votes in the comments but I still do not understand how this needs it's own pool besides the argument that posts are going to suffer because people will allocate their votes in the comments instead of posts which in turn will devalue posts.
To counter that argument that a separate comment reward pool why not allocate more votes over the same period of time and in turn allowing people to compete in the same pool, of course with this new curve for voting power.
I think it would be a terrible idea to remove curation rewards completely to be honest. Curation rewards give SP a use value. Without it, this is not much better than the many many tipping platforms that have tried and failed. People will simply use their SP to upvotes themselves and their will be no point in curating content (organising it so that the better material is more visible)
I don't see people's behavior changing based on the rewards that curations pays (whales/hoarders excluded), at least to say that curation won't happen, curation rewards give meaning to SP, then shouldn't SP necessarily be scrapped because like you said it incentivizes votes based on payout/profit rather than on curation at that point.. or am I misunderstanding it. Curation won't stop simply because there's no paying incentive to do it. People will still type and post things because they have other incentives, they will curate in pretty much the same manner. Other platforms have a decider/admin/authority that supersedes users, steem does not, which is why I think steem cannot fail, but that is not to say that it won't stunt it's growth to have such discrepancy in power between users, it's very off putting to anybody that joins when they see not necessarily self voting, or even self voting whales, but collusion of power to gain more power. I understand the steem power is actually a great thing to invest into for arbitration, but as it is now it acts more like a savings account to use in a game of big wallet little wallet and less as a tool for arbitration.
Very good points here... I am inclined to agree.
Also its worth noting that just flattening the curve (and thereby removing a large portion of the penalty for fractional voting) will potentially have a significant effect of curating comments in and of itself.
The formula they used is great. Better than n log n. I just wish it didn't come with yet another side order of opening up additional vectors for abuse.
In fact, in terms of amount of reward distributed I think this curve might penalize large users mildly for not splitting their vote (and also for bandwagoning)
I think it's important to encourage curators to split their vote.
Not sure where you get these figures you toss around for the number of authors and active users. You maybe need to start following @elyaque who puts out those figures on a daily basis.
The numbers could be a lot better but you seem to prefer to them be a lot worse which to me places some of your other assumptions in question. I think we need to let the dev team try out what they have in mind knowing they will be willing to roll it back after looking at real figures on the effect the changes have.
Steempunks.com
Can you seriously not see how rewarding commenters would grow the userbase beyond bloggers to a wider userbase of commenters?
Can you seriously not see how comment rewards would attract a larger user base beyond "bloggers"?
Did you seriously miss the point where I said...
"Instead of creating a separate pool for comments, why not treat blog posts and comments the same? Same curves, same curation rewards, same time frames for voting. Wouldn't that make more sense and actually adhere to the KISS principle?"
Do you understand the implication there and the effect it could have on comment rewards and incentives for both commenters and comment voters?
Did you seriously miss the point I made when I said...
"Ultimately though, a comment pool will not attract users if nobody knows about it...or the website itself. Unless there is going to be a marketing campaign to accompany these changes, they won't matter much. Steemit has a lack of active human users and an apparent lack of interested potential users."
The problem with the tiny user base and engagement isn't a lack of a comment rewards pool. Surely, you should have been able to figure that out by now. It's not like the problems haven't been debated ad nauseam.
Did you miss my reply that said I disagree. I apologise for not phrasing it as completely disagree rather than just strongly, I hadn't realised you wanted me to waste my time breaking down your comment in a battle of sophistry.
I see no reason to treat "posts" and comments differently, personally. They're both words strung together to form sentences, which users may or may not find funny, entertaining, valuable, or whatever.
speaking about that...
there was an idea, that... pool for comments in the posts could be proportional to value of a post. That of course would cause that only top-trending post would be commented... so what if we would create a pool for a comments which would be inversely proportional to value of a post?
I am thinking about designing a system of incentives, which would encourage people to leave a valuable feedback and comments under undervalued posts.
That would cause, that people would start to focus on commenting new users :)
Or...and hear me out on this one...it may not be easy to follow...
We could make things simpler and treat all posts the same? :)
if you would like to make things simpler, then you could make a proposition to set equal vote for everyone - that is simple! But wait! That would not be steem any more.
Things can be and should be simple when they proved that they are working.
Whole my concept described above would be actually easy to explain: Help those, who are weaker. They have the greatest potential to grow!
Actually the whole purpose of curation rewards were to encourage exactly this behavior. The problem is... that vote is not the best way to give a feedback. There is a reason why when people are flagged, they are always looking for explanation from a flagger.
That makes no sense.
because...?
I don't see how people spamming "Good post :)" on every bad post on the platform will create any long lasting value.
only the pool for comments under low-value post will be greater. Someone still would need to vote for that.
If someone would leave a great comment under average post and only X would vote for that, this comment should receive more reward comparing to situation when exactly the same comment would be posted under post which already has great value and also would be voted be the same X.
That would actually cause that "great post" comments (which were posted only to be rewarded) would almost never appear under great-value post.. because those comment would need to receive tons of votes to get meaningful payout from smaller pot.
Though I am personally for removing curation reward, I totally agree your opinion. Removing curation reward and separating comment pool are not fully discussed yet, and there's no clear consensus among the community members. Let unfinished things opened.
I am opposed to this. Among other problems it will result in less honest (or at least not aggressively abusive) vote power being deployed into the comment pool and therefore facilitate abuse.
I think it is better to address the issue of revising curation rewards as needed, not have a separate comment pool at all, and just use the new curve (or one like it) for all content, with all content earning curation rewards.
Why treat posts as different from comments when comments can contain just as much information as posts, will engage users in a natural way by starting/continuing the conversation and exploring other discussions, and if we remove the incentives for commenting or experimenting with comments doesn't that affect a greater majority than posts themselves as there are more comments than posts?
I agree. I suggested above:
Yes I was agreeing in my own convoluted way, sorry for the confusion.
Currently what stands out to me is the lack of reading, nobody sees posts so nobody reads comments, and older posts are locked to comment so themes and conversations get diluted. I would say we need a natural way to incentives comments not by another reward pool, but rather by making them available to the readers, so far people can read only the title of the post and some arbitrary information, what if a substantial update was made, yeah I'm thinking comment rewards are a good thing, a few people have been offering them but those are people making good payouts. Tough question. I have a idea I'm not sure it's a good one. It will certainly not fit the Kiss strategy :D
Some here I agree. Dont change it.
Maybe N SQUARED is a feature and not a bug... The free market is currently finding the value of what STEEM is worth in spite of the highly visible selfish mining syndrome which plagues all communities. If one of the biggest whales like Smooth says he hates this proposed change, then I'd wonder how bad he hates this change as a shareholder. In other words, If I was a whale like Smooth, and someone took my cheese away, I just might shrug and leave.
https://steemit.com/steem/@steemjesus/1-good-thing-about-steem-is-that-the-weak-arguements-fall-away-so-fast#@gtg/re-steemjesus-re-
I think you misunderstand. I'm very much in favor of a flatter reward curve, but it should be for all posts and not just comments, there should not be a fixed percentage allocated to comments (which could very well be either too high or too low, or even both at different times) and posts and comments should be treated equally with respect to curation rewards.
Curation of comments is nowhere near the same importance as curation of posts. Rather than seek more ways to generate more steem, why not start thinking about what makes steem more valuable to people as a whole. 38% isn't really a cut if you consider what good it could do for the price.
It is very unclear to me that any of this has much effect on the price but if it does then flattening the entire reward pool (including comments) with one system that is less complicated and confusing would likely be more positive.
Agreed but who is deciding the road on what happens @smooth - who are the ones with the POWER within the company that drive the steem ecosystem? Not speaking about the people on top of the company to the public world here. Any hint would be cool - sorry for asking but I work in PR
I agree with you but I won' cast my vote on such a high payout post :D I agree with your point, but I would flag for an oversized payout :D
Three things:
n^2 + an, strictly superlinear. That is, its derivative is not 0 atn=0. I don't know if that has much effect on your conclusions, but I figured I might as well point it out.If adding a separate voting power pool for comments, they could do the "8X more rshares" thing they wanted to implement a while back.
Yeah, I think that would make a lot of sense.
I disagree that this is needed. I think that the opportunity cost to voting for comments may actually be a very encouraging thing for minnows who won't have any opportunity cost. (Remember most users start as a minnow)
https://steemit.com/comments/@beanz/why-should-commenters-make-38x-more-rewards
This will reduce curation reward a lot which is currently the primary incentive to buy steem power.
Steem should slow down money distribution to people who don't buy steem power. This is not going to end well . All effort should be done to rewards those who actually buy steem power.
https://steemit.com/steem/@snowflake/steem-inflation-a-tool-to-create-demand
That defeats the purpose in my opinion. The purpose was never to reward those who buy steem. The purpose is to reward participation on the platform because that distributes the currency to as many people as possible. That includes bloggers and commenters. Remember that steems value is speculative, so what 38% is worth now, will have completely changed over time and especially after recruiting a larger user base of commenters rather than just bloggers and curators.
https://steemit.com/comments/@beanz/why-should-commenters-make-38x-more-rewards
The reason people want steem is because it has value, if it was worthless people wouldn't give a fu* about it, so you have to make sure to align incentives with people pouring money into the system aka investors. If it was guaranteed that minnows would gain more influence I think it would probably be a good trade off and I'd go along with it, but whales will still be able to upvote and so I doubt minnows will gain much and be incentivized enough to buy steem power . This problem needs to be dealt with at the root. Btw I think that steem won't survive long term if 100% of the voting power is not given to the 99.8% so I am very into the idea of giving minnow influence but the way it's being done here bothers me.
The reason, why not say The only reason..
It's not just about investing capital but about investing attention and effort, and people do that for free everywhere else, so..
You can't judge of a curve over another until there has been enough users to the platform. The fact that we can't subscibe to tags, that the retsteem are at the same place than the followed posts. and that there is no more 30 days trending. (there should be way more time range) Also we should be able to select a bunch of tags and cutomize a homepage like reddit does.
I really wonder where the steemit energy is being put.
Mobile development.
esteem is very good and better than steemit.com, What's the point ?
Are you making steemit.com responsive or building iOS / Android native app ?
I was just saying what Steemit is doing ;) I don't get why they (sneak) are going full throttle into mobile, and ignoring the existing userbase of bloggers and requests for better features. They are going to focus on mobile users, and their attention spans, and what they can do (not blogging/writing obviously).
Because STEEM is a currency first, and a social media app second. The number 2 most used (#transactions) chain on earth.
steemit doesnt care about the past
it only cares about what it can make off your content TODAY
all posts should be on a constant 24-hr pay cycle but not all of them will get paid out after the first day because then they fall off everyones radar
thats why this place is not a source of information like it should be
no one comes to steemit to "look up that one blog post that was really awesome" because after it pays out its forgotten by everyone even the author
people come here to post things to make money
the whole site is so damn blinded by the money that no one cares what you post
Google remembers and generate a ton of organic traffic.
youve missed the point
the limited payout opportunity for a post results in people reposting because their original post got zero attention
not all good posts get the attention they deserve and its pretty crappy that after 24-30 hours its over
not once have i seen a 30-day post resurge into the trending section because people are conditioned to forget a post after it pays out
and you shouldnt be relying an ANOTHER site to do the job YOUR site should be doing
dont get me wrong - google is a necessity but its laziness to put some of the functional responsibility of your site on the shoulders of another
My personal opinion is we need to adjust
1part in the denominator. Currently the constant part is equivalent to about 400 MVESTS, so dolphins still have significant disadvantages compared to whales and this also allows bots to easily choose their strategies.I would suggest to decrease it below 100. With lowered bar, many more posts can reach the almost-linear section then people will have more leeway to choose posts to upvote, while bots will be confused by flood of candidates.
As I said in the chat room, we can also continue using the
n^2 + a*ncurve but increaseato a value equivalent to about 20000 MVESTS. With this curve the lottery effect will remain but only to posts with highly concentrated whales voting. Any single whale voting on a post alone will only have a near-linear effect.My head hurts. Apparently, it has been too long since my last math lesson.
I finished my draft mainly on technical/mathematical sides. Hopefully it helps your understanding (but I cannot guarantee LOL). It will be posted after polished by peers, maybe tonight or tomorrow.
So there is a lot of controversy around splitting the rewards pool. I honestly keep going back and forth on it, so I can't say I have an extremely strong position on it. I am still giving it a lot of thought. The one big argument that I will say against it though is it seems like it violates KISS.
So ignoring whether you are for or against splitting the rewards pool for a sec, I will say I am strongly in favor of using
n^2 / (1 + n)in place ofn^2. @clayop did an excellent job summarizing the views that came out of a very large community discussion here, and the consensus is that implementing this across the board (for comments and posts) is a really good idea.There has also been some talk about re-working the curation formula too, to incentivize better voting behavior. I think there is potential to make some significant improvements here, and I've heard some good ideas proposed by @smooth.
I do really like that the end goal includes having
n^2 / (1 + n)for both comments and rewards, and potential changes to the curation formula. Both of these are a big plus IMO.I don't see how this curve discourage self voting. All it does is that it reduces payout of every posts that is already earning very little. You basically reduce a few self voting posts at the expense of everyone else.
The only logical way to counter self voting is to downvote self voted content.
I have a practical suggestion. We could try to find a perfect curve with theoretical reasoning, and go back and forth for a long time...or, we could just decide to run an experiment. Suppose HF17 ships with
n^2 / (1 + n)which automatically expires, and reverts back ton^21 month from now. This way, there is limited downside risk, and we get to actually see how the new curve performs in practice.