The account-based voting revolution
Steem has evolved dramatically from the initial implementation in 2016. However, one of the key features has persisted - stake-weighted voting.
The idea was based on the assumption that the highest staked holders of Steem (i.e. whales) would act in the best interest of the platform and delegate their stake to the best curators of the platform. Not just this - they were assigned exponentially more influence with an n^2 reward distribution curve. I.e. someone with 1 million SP will have allocate 1 million times more reward than the average person with 1000 SP. Not 1000 more as you'd expect.
Of course, such an assumption is utterly ludicrous and goes against human psychology 101. As expected, whales started acting selfishly in their best short-term self interests. In June 2017, the n^2 curve was changed to n^1, vastly democratizing voting. Of course, the 1 million SP still had 1000 times more influence. It still wasn't enough.
At the GOPAX meetup, Ned had a lot to say about this.
(Skip to 1:04:11)
Firstly, Ned clarifies that Steem's algorithm was designed to reward the best quality content the most, and that is still the intention. So it's reassuring to see that he still cares about quality of content. Granted, that's a pretty vague goal, and not realistic, but there are magnitudes to how far one is from that goal.
Given they still have this goal, Ned was fairly contrite in acknowledging that the current algorithms have failed entirely. We can see this clearly - there's zero relation between quality and rewards. Indeed, over time, as Steemians have learned to exploit and abuse the system, there appears to be somewhat of an inverse relation at this time.
A few weeks ago, I made a comment that I don't see this system as sustainable, and it will need a radical solution people may not like. Ned seems to agree.
He identifies stake-weighted voting as the single culprit for this vast mismatch between quality and rewards. Instead, in SMTs, we'll have the option for a different type of reward distribution based on account-based voting. I.e. one vote means one vote, irrespective of how much SP one holds. There'll still be a superlinear curation rewards curve, so curators who find the post early will be rewarded much more. I'm assuming magnitude of curation rewards are still tied to stakeholding, but there's no mention of that. The first issue is - this totally kills demand for SP. But then, one can argue there are shitcoins worth far more than Steem that have no such concept similar to SP - it's all liquid.
Important caveat - this is an optional feature of SMT and may not be something Steem adopts. But let's assume it does.
The obvious effect is, it'll shut down voting bots, abusive whales, scammers, curation projects overnight. Each to his own. Of course, people can cooperate and collude, but there's no longer a way for one person to dictate thousands of times more rewards than everyone else. Of course, there's a massive downside to this - you are effectively empowering random spammers to allocate the same rewards as bonafide members of the community. Whales can make thousands of accounts to effectively Sybil attack the system too.
Ned does not address this issue, but I must assume Steemit Inc are working on a solution. My suggestion would be along the lines of active members from each community votes for curators, and the highly ranked curators then have higher influence over the reward pool. No such ranking system is beyond abuse, but there'll at least be some method to the madness. Furthermore, I'd also suggest a decentralized judiciary system to deal with cases of abuse. Yeah, I know Steemians will hate this. Of course, SMTs can take a more centralized approach and appoint curators themselves. This centralized approach will always be far more efficient than any decentralized method of curation. Yeah, Steemians will hate this implication even more. Sorry folks, but this is how human society works.
This is a complete revolution. There's a lot to digest here, so give it a watch, think about it, I'd love to hear your thoughts. Equally, there's precious little information, and it raises many questions. This was just a tease, I hope for a detailed post and whitepaper on the account-based voting concept from Steemit Inc soon.
What I would love to see is a mixed rewards distribution algorithm that takes into account:
Of course, this would require a new reputation algorithm whose result would actually resemble the reality.
Just a quick shot, but I really think this could serve...
Some time ago I formulated some partly similar (diminishing returns) ideas to make self-voting, circle-voting and spamming less attractive by ...
@clayop had a similar idea.
It seems obvious to me that it has to move in this direction sooner or later as the blockchain (hopefully) begins to go more mainstream. Obtaining and retaining great users and curators will be increasingly difficult if the price of tokens increase and lower inflation + more users fighting for the scarce distribution makes it harder for new motivated users/curators to establish themselves. If so, why should they choose STEEM and not an inevitable competitor?
The main challenge will not be to make it better in terms of how it elevates the best content (any idiot can achieve that as it can hardly get much worse). The difficulty will be to maintain - ideally strengthen - the utility and thus demand for the token at the same time while also making it seem fair and robust to large and small stakeholders both.
I'll see if I can get a long piece on how I believe this can best be accomplished within the frameworks that already exist (or is not too much of stretch from what we have). But this is where we need more of the brainpower we have on the platform to work together for sure. I imagine the solution will be a new form of collaboration between Dapps and Community projects that we haven't yet seen.
Hi friend...how are you .. ??
I like this your post....
Please Stop - @mdsujonislam
You just said "your post" and in your your last 100 comments you used 66 phrases considered to be spam and you made this exact same comment 1 times. You've received 0 flags and you may see more on comments like these. These comments are the reason why your Steem Sincerity API classification scores are Spam: 76.60% and Bot: 2.40%
Please stop making comments like this and read the ways to avoid @pleasestop and earn the support of the community.
My two cents.
First of all, I agree the current reward system is far from being optimal (it is even wrong). This being said, one sentence got my particular attention:
I like this. It goes very well with the development of the communities. But how is this different from delegations? I would say not much, somehow.
However, one may need a pool of people (manually) checking the behavior of the curators in order to prevent from abuses at this level.
So in short, thus may go along with the community deployment and kill the entire voting-bot mess. If I am not wrong. This is nice :)
Crucial difference. Currently, whales dominates the reward pool. After something like that, reputed curators and community leaders such as yourself will dominate the reward pool. Doesn't matter if you have a million SP or 1000 SP, at the end of the day your voting influence is based upon how good a curator the Steem community thinks you are.
Of course, this is prone to abuse and circlejerk too, but at least it makes some sense and reflects the public, rather than just some nonsense algorithm currently that assumes "the richer you are, the better curator you are".
I agree, and I suggested this in the next sentence - a judiciary system to stamp out abuse. Steemians can vote for people to be elected as judges etc.
Besides that, the community can choose to unvote their curators, just like they do with witnesses now.
Note - all of the above are simply my vague suggestions, and I don't expect Steem to actually adopt any of it. I'll be interested to see their solution though.
Agreed. Something (anything) should be done, in one way or the other. The current system is not sustainable on the long run.
Or downvotes could be used as a measure of curator quality. If after the vote there is flags, the curator reputation Go down.
Not necessarily in the way reputation is currently implemented (if I am not wrong).
i like the idea of bringing the best content at the top, but it depends of who are curators and who are the readers who decide what is a good content, because everybody has an opinion, so this could influence the future of steemit , we need to choose who are the right curators and the readers because this could bring to another issue like censorship.
As you yourself state, this is a revolution but i found the same issues on this as you did. My fear is that by making it all an equal playing field those that invested the most into the platform, ie. whales, will not be rewarded sufficiently, thus pulling out their SP and crashing the price.
The stake based reward system, i really dont believe was ever created with the intention of creating a platform that rewards the highest quality content, but rather it was used as a way to attract investments. They would have gone with this route in the first place if that wasnt the case.
As much as i love 1account/1vote, equality and all that, that just wont work by itself. We need a somewhat of a hybrid system. I really believe if he flips the board this hard Steemit and all the other front-ends will crumble.
One suggestion that they should be doing, is coming up with an internal condenser that no one outside of Steemit sees... that uses the new algorithms to see how the new voting system would change the vote rewards over a period of time, and have reports ready to show that data when the time comes.
(During the alpha roll out of analyzing the effectiveness, so they can tweak things)
Sort of "snapshot the chain for a week" so we could go to a website like this after we move that algorithm to BETA for public review:
http://votingsnapshot.steemit.com and everyone can then view steemit for an entire week frozen in time with the only content available during the 7 day payout period of that snapshot.
They could also include data from the current chain rewards next to what the algorithm payouts would have been under the proposed system.
So you'd see someone with $1.42 payout (old) and $8.25 payout (new)
Conversely, you'd see someone with a $242.95 payout (old) and a $80.52 payout (new)
Of course though, when you change the tide this way, you should expect the tide to cause a huge reorg of value on the chain.
For example..
Make no mistake, the massive whales would simply start transferring wealth to their individual user botnets, and adapt to the new system within a week.
I'm quite nervous how this would play out. The one caveat we have is there are smart people working in development, and every issue we can think of has probably already been talked about and addressed in their meetings.
So I suppose we have no choice but to just wait and see this play out.
Not sure I follow how this would mean the end of curation projects - actually it seems this would boost them up. Curie for example has the second most followed curation trail on steemauto, currently with 995 users trailing their vote behind. Under current system, Curie's influence comes mostly from the large @hendrikdegrote vote trailing and to a lesser extent the ~100k SP that Curie itself votes with. So yes, if it goes to one account, one vote, those two large votes would not have as much influence. But currently, the 995 trailing votes from steemauto mean next to nothing in terms of the payout that a post receives on a @curie upvote because the vast majority of the trailing voters are low stake users. But it is still nearly 1000 users who think that Curie is doing a good job curating. Wouldn't those trailing votes mean that Curie actually would have more influence if this is the way things went?
My own curation project, @r-bot, has ~100 voters who have joined its curation trail on steemauto. People have joined the trail because @r-bot consistently upvotes good content. I haven't promoted it much, @r-bot does not regularly post, nor leave a comment. The team of manual curators just quietly goes out and upvotes good content. Wouldn't this change mean that the @r-bot vote would suddenly be worth a hell of a lot more than the ~ $.50 the manual curators who vote with @r-bot can hand out now?
Yes, that depends on the type of curation project. The projects which have 2 million SP delegated to them by one or two people, but don't have much of a trail will lose out; while something like Curie which has a large popular following but few delegations may benefit. I'm fairly sure Ned didn't absolutely mean "one vote is one vote", there has to be some systems to keep spammers and dust accounts from exerting so much influence; while also offering greater influence to active members. Let's wait and see what they come up with.
Thanks for sharing this @liberosist - Even interesting thoughts and messages, nothing new! And maybe too late to implement this into an existing ecosystem ( mean the community internal ecosystem which was from the start a platform with 1) Creators, 2) Engagers, 3) Investors (or call them Steeem / SP stake holders).
My main struggle(s) with implementing such a fair / equal vote model is indeed three fold:
We all know the current system is not fair and enable abuse - no doubt, this is not new indeed, we all talk about even before the last Hard Forks
Changing this now will probably mainly harm the guys working most here to increase their SP being not park of circle jerks or benefiting from early invests
Spammers and Scammer will kill the blockchain if no ID check is in place - even ID checks might not help as people are creative when it goes towards gaming a system
Resteemed.
I think you're right, if an SMT proves itself in the wild to be better in configuration than the main Steem coin, we will probably see it's config adopted. I'm still not sure how that would look from a business perspective at the time, but this is what Stinc have implied I think if you read between the lines.
I don't think they are. Really. All bets are on SMTs period, I would be very surprised if we saw any significant change to the Steem coin config until after SMTs are released.
If you're not just talking idling here, draw out a short spec for an SMT based around this.
But this is a feature of SMT. There's no way they can implement an account-based voting feature without effective mitigations for the obvious problems with it.
That was idle talk, of course. I'm not really interested in SMT, neither do I have the skill to draw actual specs or implement them.
Oh, well I just re-read what I wrote and it could have sounded pouty, I meant it sincerely! I've been sketching similar ideas for the last 6 months so am genuinely interested, but sure, idle talk is good too 😅🤟
And perhaps the main reason why this 'optional feature' of SMT won't come over to Steem?
I have thought for a long time that SP should be held as tightly as possible as the Steem ecosystem grows. A change to remove demand for SP is a huge step backwards isn't it?
SP, an investment with a 13 week withdrawal period will bring more stability as time progresses, while SMT will grow the value of STEEM?
Ah well, need to rip my blueprints up!
I don't think crypto traders like SP. It used to be a 2 year power down before. When it was changed to a 3 month schedule, it was met with a price increase from $0.10 to $1. (After the first 3 month period ended)
Not to mention, random shitcoins with zero usage are worth billions, far more than Steem, despite having nothing like SP. It's all liquid, and it works for Bitcoin, Ethereum, and hundreds of other shitcoins.
Thank you for this weeks golden snippet of information. It's things like this that you don't find in the whitepaper :)
While i cant current get my head around how account-based voting would work here with regards to SP becoming less useful, taking a route that negates the use of bid-bots and circle-jerks is certainly a good thing.
Boosting incentives for curation as much as possible has to be the way to go. Those that don't have the time can delegate to communities or people that do - they will get an honest share (if they so wish - many delegate for nothing) and the race to find the best content will be back. Great :D
I note the point has often been made regarding locking up investments is generally time-limited in other investment vehicles which become completely liquid after some point, and isn't a feature of many cryptos, like BTC, that don't seem to lack investors.
Perhaps @ned has another trick up his sleeve to maintain some limitation on liquidity. Perhaps he reckons it isn't necessary, nor beneficial, any longer.
I will eagerly await more details, as I'm sure you will also.
They can increase the inflation/interest rate you get for locking your coins in Steempower. Currently you get 7% and it diminishes each year. But they can always increase it to provide a "return" for holding steempower.
I hadn't realized that incentive.
Thanks!
Food for thought isn't it.
Just when you thought they'd blown it, or at least read 500 times that they had :)
I confess that eating my words of doubt will taste delicious!