Voting Test - Do You Think Changing to a Flattened Curve Is Urgent?
This is a test post for stake-based governance.
This post and comments for vote all decline payout. Please choose one of the options in comments and upvote it as much as you want. Please don't downvote comment that opposites your opinion, but if do I will remove it from calculation.
Question: Do you think changing to a flattened curve (more linear compared to n^2) is urgent?
Options: 1. Yes / 2. No (Upvote comment)
Add: By this way, we can efficiently figure out how much stakes are for or against on a certain issue. This can be a governance model of Steem and if it will have neat UI and be integrated to Steemit seamlessly, it would be awesome.
Excellent we have stake weighted polls!! We should do this for all the proposed changes!
Exactly!
YES!
As much as it might be urgent, I would not want to delay the bug fixes in HF17 because of contentious changes. I would implement ONLY ONE of these 3 in HF17
Flatter reward (but not flat)
7 day payout
separate reward pool for comments.
Unless I missed something everything else is good with me to be bundled in HF17.
Work on steemit.com UI is much more needed, and a way to finance work by turning pool reward into bounties.
A better text editor and a way to subscribe to #tags are some of the things that come to mind in terms when I think of urgent.
amen!
Ideally the user interface should present yes/no and other options in random order to the user.
Yes, it is clearly the communities overwhelming consensus and I think it increases fairness.
Abstained because I think the issues are complex. I dislike n^2 as much if not more than you do but replacing with something radically flatter changes the nature of the system a lot. I would probably support it but I don't think it should be rushed "urgently" nor do I necessarily believe it solves all related problems. Stake weighting, especially with wildly unequal stake (which is likely in any case) in a social context carries its own baggage independent of the curve. Possibly acceptable but again this is a more complex question and not one to be solved with a quick fix.
1. YES
Upvote this
I think the expirement is what really got people excited again because it flattened the rewards curve. Changing the formula is the best way to sustainably change the reward curve
I think everyone is for a flattened curve. That's even what Steemit inc has stated. They didn't revealed much but from what I get, I think they don't want to make it easier to rigged the vote in a selfish manner.
So I think the proper question should be, what would be the best curve and with it provide the most detailed math possible to try to prove or disprove what should the curve look like.
How about ^1.2
It might be it. Someone should graph a couple of curves and begin to argue for the best.
In my opinion, what will be most effective is a limit of posting rewards on a single post ($2,500 ??). This amount is still enough to get posters very excited about the opportunity, but also allows for spreading the wealth further field so more posters benefit. Additional, voting power should increase based on steempower, but only up to a limit (50k /100K ??) this will have the effect of curbing the influence of the very large steempower holders and make the posting platform a slightly more level playing field. Therefore as enticement to the very large steempower holders, any steempower over XX amount should not be rewarded with voting power on the blog platform, but with a higher interest rate on their steempower and also with the exclusive rights to vote on structural and organizational decisions of the organization. The current method of funding software development, marketing, etc through blog posts rewards is not a effective method. Structural and organizational funding to be distributed using a process more tailored to project budgeting. See Dash-coin masternode project proposal page for a general idea: https://www.dashcentral.org/p/DASHPAYCARD-DEBIT-CARD
I don't think there should be a max. Payout in fact I think voting and payouts should continue as long as the content is up. Just like you get paid with Youtube or most other places. Here there is only 2 pay periods and right now it's very hard to make money if not noticed by or in favor of the whales.
What Steemit needs is more equal voting or totally equal voting, continual payouts, a block feature and a rewards/affiliate program that pays people to bring new people on board and an ads program to bring in more money to the rewards pool.
Thanks for your reply. You have some decent points. For one, continual payouts would be great,( though don't know how that would be implemented), however I'm looking at this from the standpoint of a viable business and the investors who give value to Steem. What we saw last summer with some single posts making up to $30k and $40k was rediculous and not sustainable. It as simply a wast of organizational resources.
The rewards pool I don't feel is as important. Steemit was a very attractive and unique proposition for bloggers. One Steemit gets the incentive structure right and starts advertising/marketing it should not be a difficult sell.
If all votes were counted equal that would never happen without a massive following, and then if it did it would be deserved as that's how Youtube stars make millions.
It will be a difficult sale if you only get paid or two pay period unless you are getting insane money. Why would a Blogger with a following making continual Ad Rev and Youtube Payout come here to only get paid twice and a cap on payouts. This is also the problem with not having equal votes because if a person were to bring over 1,000 of their following, what that worth? $10-$50 (maybe it's more) if they are all new people that have no steem power? And there is no affiliate reward for bringing over those people.
If all votes were counted equally, you 1,000 people you brought over would make you money immediately.
Point taken. I still strongly think that there should be a graduated increase in voting power (up to a limit) based on the about of Steempower held, but to your point it would be good if a way could be found to give people holding zero (or very little) steempower a small bit of voting power (this would be for rewarding bloggers, but not for getting curation rewards). The problem is this typically opens the platform up to spam.
I would be great if a post could continually received payment, with no time limit, but don't know how they can work that would technically. I think for now Steem could concentrate on attracting, not the major utube stars, but the regular folks who are currently not really earning on other media platforms. This will start getting the buzz going while they workout the economics for the people who need to use the platform to generate major income.
I've read a lot about the experiment and the flattened curve. I think the experiment is working even though the implementation is imperfect. If flattening the curve will codify the experiment, which I think it might, then I am for it.
I think that changing to a flattening curve would be good for everybody, including the whales, as it would contribute to raising the price of STEEM.
I cannot comment on the urgency.
Obviously, Yes.
I also think that downvoting, in its current form, is the most annoying "feature" of Steemit, driving people away in droves. Something must be done about it. I personally declined my witness and seed operator position specifically because of this. When and if this will be addressed I'd be happy to rejoin the witnesses.
First you should rejoin the witness list. That's how I feel about you dropping of the witness list.
What the majority of people act like isn't an indicator of truth. Most people don't have a healthy lifestyle. In truth everyone gain from an healthy lifestyle yet most don't live by it.
I read downvoting is a necessary part of stake weighted decision making which Steem is based on because of some part of game theory which would make Steem prone to be gamed or something. I cannot explain it in detail even thought I read a couple of article on it here on Steemit. I never used downvoting once but I would be against removing down votes at this point.
Thanks for you words about my activity as a witness. Won't rejoin the list, because supporting something that I consider abusive would make me hold conflicting views, I'm not into this anymore.
As for downvoting, the current implementation allows for much more abuse than in any game theory. There is no stake in votes, just influence. It would be stake if every vote (up or down) will put some money on the table. I already made a proposal for this, back when I was a witness. https://steemit.com/steemit/@dragosroua/yavap-yet-another-voting-algorithm-proposal-or-what-i-actually-understand-by-proof-o-stake
Downvoting belongs to the Steemit logic such as upvoting does. The only thing that hopefully will change over time is how downvotes are perceived by users. Ideally as something quite normal to occur and nothing to be upset about.
I'm not for downvoting or flaging, but if has to exist the the downvoter/flagger must comment and state the reasons why. This way he/she is risking to be downvoted for the comment.
I think the downvotes are confusing. Is it a dislike or is it flag for breaking the laws/rules? Why isn't it separated into a downvote = dislike and a flag = bad actions?
I disagree. Even @ned wrote almost a year ago that downvoting should be reserved for blatantly negative posts or spam. It's hilarious to wait for people to accept negativity, instead of listening to them and eliminate it. What happened with "the client is right" thing? Here's what @ned said about downvoting a year ago.
Everyone accepts that downvotes cause negativity and problems. But nobody has been able to effectively argue against the cases where they are genuinely necessary, including where reward allocation is concerned.
Those who are campaigning against downvotes are getting nowhere because they are failing to convince people that they're not necessary for reducing over-allocation of rewards. Right now the downvote has been applied so liberally that it has effectively enforced a flatter rewards curve and demonstrated exactly that problem, resulting in greater engagement over recent days (posts and comments have spiked, even accounting for bots).
If you oppose downvotes, the best case you can make is to minimize their use. Harm reduction is generally better than prohibition and arguing to deny stakeholders their right to have their say. Here is my proposal which I think can virtually eliminate the impetus to use a 'rewards allocation' downvote:
https://steemit.com/steemit/@demotruk/up-and-down-votes-are-a-big-part-of-the-problem-on-steemit
Coincidentally, I wrote Down-votes: Steemit's Achilles' Heel? within a day or so before "the experiment" was announced.
While I remain neutral on "the experiment" itself, I do think that it proved the larger point that the downvote represents a major threat to the platform, because it can be used by a hostile entity, such as a well-funded competitor or state actor to severely disrupt and erode the user base. Imagine the negative impact that the experiment would have if abit and smooth were whales with harmful intentions, and weren't "pulling their punches."
If downvotes are necessary, then I think steemit's growth potential is perpetually limited to the footprint that its competitors will tolerate.
I would support something along the lines of your median proposal, although I think it may need to be modified from its current form to be compatible with the blockchain.
While downvoting is perceived as a flag and rightly so as it is refereed to by the system as such, and constantly refereed by the community as downvote, it will garner all the negative attention associated to flagging, and act as a confusion and misunderstanding primer. It's kinda a flag but it acts as a downvote with the power to also hide content, not just decrease rewards based on VP and SP. Downvote and flagging should be two different systems, flagging is for abuse and flag abuse, and that should be equal in power across the board, and the current flagging should just share a place next to the upvote as a downvote and it should be much more draining on voting power, as a means to balance it out and de-incentivize abuse and then remove it's ability to hide posts and give the new flagging function that ability and also to negatively affect the reputation, as well as allow for flags themselves to be flagged, and if there is a consensus of 2 or more affect the reputation of the flagger, as to stop sybil attacks with flags peg the flagging ability to a good reputation score that can be achieved by new users, maybe 50ish.
As for downvoting being used for purely negative posts or spam it's clear that those parameters have changed considerably as it is explicit in the very first reason it lists when you hit the flag/downvote button and I quote:
A lot has been learned in a year. Too much reluctance to downvote is also bad for system as a whole, at least under the current ruleset.
The downvote started to be used in an abusive way just 2-3 months ago. You being one of the most prominent downvoters here. I don't see how this equals as "a lot being learned". People are leaving because of this. People are the flesh and blood of this product. No people = no product. Simple.
how did you decline payout from comment? I gues from cli.
There should be special checkbox for declining payouts for comments - then everyone could create such survey :)
Yes from cli. I hope we have more fancy frontend for survey, not like this coarse thing :)
+1 this 🙂
You should be able to ReSteem comments. In fact, spin comments into a post :)
In general, "yes."
In spite of what I consider some dubious execution, the experiment showed that most people were more enthusiastic about their participation with the "no whale votes" in effect. I could see my votes have more effect (even if it was just 3 cents instead of NO cents!), so it felt like my reading and commenting actually meant something; had a value.
On a parallel topic, I still have reservations on the matter of urgency. I would like to see some further consideration of the relative weight of "bot votes" vs. actual human curation. As I am not a developer, I don't have a good sense of the implementation issues... but I struggle with the idea that a bot-- which, after all, does little more than confirm that the "content EXISTS" should carry the same relative influence as a human being actually reading a post, upvoting it, and perhaps leaving an engaging comment.
Whereas I have no argument with the existence of bots-- per se-- and believe they have a definite function (flagging spam, copying, abuse, trolling and so on), maybe it should be considered to have a measure where a bot can never vote with more than 5% or 10% voting power, to give more weight to a human voting at 100%. Just a thought... based on the premise that we are trying to build a COMMUNITY... and that's about people, not about automation.
this sounds interestingly...
(scratches head)
Most important thing, because it creates a feeling of unfairness , which is the most toxic thing a community can have.
More power for minnows is needed. This is social platform after all - we need users! If reward distribution is only in the hands of few whales this project will have high change of failure. When the distribution of rewards is in the hands of many, this will drive up demand for all kinds of different categories.
Giving users more power to impact posts cleary had positive results. I'm not sure how well that works at scale though. I think it's a good thing, but the returns may be smaller with a larger user base.
Absolutely yes!
It's the single most important change
Let's make social democration economic for steemit community !
Dang.. too much weight? The post should be trending now.
lol, funny (brings and shares popcorn)
Sorry late to the party, but I found it trending...
I removed my vote after seen not many high stake holders participating in this poll.
I'd even like a logarithmic curve, but I guess it wouldn't make sense because of account splitting.
As many have already said, the experiment showed the HUGE benefits to the larger population of minnows and dolphins having a greater effect on the reward distributions. It's hugely empowering and can you imagine that feeling of the first time you were able to regularly give a single penny multiplied by a thousand (arbitrary number here) people.
The other side of the coin here is that when the refinements to the curve (I highly doubt it'll be chosen perfectly on the first try) such that the whale counter-flagging can lessen (because of increasing the defining whale SP threshold) and ideally over time cease. Beyond this I would love to see the ability of flagging going back to focusing on abuse versus disagreements. It's somewhat harder to disagree with huge payout when it was upvoted by hundreds of users (and yes I acknowledge this could be skewed or abused too with trails.)
2. NO
Upvote this
I voted no because "flattened" could be too flat for my taste.
But should have voted yes, because I'd want "Slightly" flatter ASAP.
No brave guys? (or a cat can click it...)
Add: I will play devil's advocate!
Edit: Changed my vote
Well, this test will show only what a tiny minority of the whole community - those who actually have seen this post - is thinking about - for now I can see only 52 views. Mostly that would be your followers and therefore it might be expected they would agree with you.
I wanted whats best for all of us. I dont know which way to vote because I dont understand the issue but like I said, I want whats best for everybody.
CarbonVote voting model seems good
Whatever that will give more voting power to users with non-trival amount of SP (total circulating supply / number of users (accounts)) ratio, which is currently about 2400 SP, is right direction. I would give best power ratio to users on 10% SP behind this ratio to create additional demand for STEEM. Logarithmic function with center on this ratio or slightly beyond it would be best for platform and users but whatever that goes close to this solution is good. (I know you're also aware of splitting accounts and bots, but with bots we need to deal another way). If I understand it correctly, currently reward power is exponential which is totally wrong. Linear is better but still not ok, we need logarithmic.... totally reversed to current state IMHO.