RE: Time To Wake Up and Fix Steem's Voting Problem
Before implementing linear reward Steemit Inc was against it, saying it would equal to giving away free money, leading to undesirable results. The exact quote:
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.
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. @steemitblog (source)
The current design, linear reward, leaking away free money forever, explicitly to those who don't engage in proof-of-brain is less desirable than superlinear reward which prevented such leak.
To put it more succinctly, investors set the price of Steem in fiat/btc/eth. Investors want to increase their wealth.
If investing in Steem and voting for themselves all the time is the best investment, they will do it.
The selfless investors want to incite cooperation and will avoid participating in a system unfairly favoring the selfish.
Obviously, the rationale behind selfish behavior will push Steem votes ever closer to simple interest payments on SP.
Last month @abh12345 asked: "Are the selfless delaying the inevitable?" (link)
Some of the conversations I had with prominent members of Steem in regard to Steem's economics incentives and more precisely the reward curve.
- @ned (6 months ago, 5 months ago, 3 months ago)
- @kevinwong and @trafalgar (1 month ago)
- @timcliff (1 month ago)
- @aggroed (1 month ago)
- @lukestokes and @smooth (this week)
Linear could be proven to be a better system than n2 but so far, no such proof exist. Proof that n2 is preferable to linear exist and have never been disproven.
My intentions are humble and just like every other Steemians I'm looking to better Steem.
This is false. It was demonstrated to work terribly. The premise that whales would behave better because they have 'more to lose' was and is entirely false.
I would love to have a better voting system and better align the incentive to adding value. I don't see it in n^2 (nor in any curve really, without other changes). I'm open to other alternatives.
Agreed, n^2 was a disaster then, and will arguably be even worse now. In a world where voting bots command much of the SP, you wouldn't want heavy super-linear. Linear or slight super-linear is better, but isn't an acceptable solution either.
Stake-weighted voting has failed. This is no surprise, given weighted democracies have always failed throughout history, sinking into greed and corruption. There needs to be a radical rethinking of how the reward pool is allocated. Changing curves are no longer the solution.
I don't have an alternative, and neither have I seen one. The only suggestion I can make, that'll make Steem appealing to me again, is canceling the common Steem reward pool. Instead, let SMTs compete for the best solution. I'm willing to bet the SMTs that are moderated with account-based style voting will attract much of the actual community engagement. But the totally free market SMT competition (as Steem is today) has every chance too, if that's what the market leans toward. Steem benefits by being the core asset for all SMTs. I know this will come as a ludicrous suggestion, and I don't expect anyone to agree to canceling the Steem reward pool. But that's all I have to say, I've given up on trying to reconcile stake-weighted voting at this stage.
PS: The idea that the largest stakeholders will "act in the best interests of the network" has proven to be false, time and time again. This is no surprise, either, as throughout history the powerful have always acted in their own self interests. There will be always exceptions, of course, but such a system is unsustainable until every single stakeholder of the network is not selfish.
They may well capture community engagement, which is fine, and may add a lot of value (censorship resistance, accountability, etc.) But I doubt they will have large reward pools. I don't really see investors voluntarily signing up to have their capital redistributed via a one-account-one-vote system.
I'm perfectly okay with this.
I also agree with you that some SMTs may come up with their own hybrid or different (including semi- or completely-centralized) solutions. Steem itself doesn't seem to have one.
I don't think so. I've proposed the same.
Interesting idea, and not all that ludicrous to be honest. Not sure exactly how you would implement it because you would have to change the whole premise of the system but sometimes a radical approach is needed.
Implementation - Minting of SBD and Steem to curators and authors will stop, and the reward pool will cease to exist. It can be gradual. Steemit.com, Dtube, eSteem can have their own SMTs. (Or they can collaborate to create one.)
Some SMTs can be stake-weighted, some SMTs can be completely centralized. (i.e. influence over reward pool is allocated to vetted accounts by owners or admins of that SMT. Miscreants and abusers are promptly banned from allocating the reward pool.) Or the best case scenario, an SMT can have a combination of multiple approaches. One possible approach would be people vote for curators instead of witnesses, and the top voted curators get to allocate the reward pool relatively. This will have to be weighted by community engagement metrics or some kind of reputation system to avoid Sybil attacks. It'll remain decentralized, but also try to achieve a more sensible way to allocate coins. Just thinking out loud here.
The point is - this will lead to competition, and who knows - maybe a couple of SMTs will hit upon an actually sustainable solution to making it work. It's clear that the Steem reward pool has failed, as far as content creation and valuation goes. Open it up to the free market, let it come up with solutions and the market will decide which one "wins".
Steem will benefit being the core asset to all SMTs.
Of course, SMTs are many months away at best, and I don't have an interim solution.
Yeah the last sentence is what I would worry about most. SMT's won't be launching till years end and by removing the reward pool I think you would be removing the biggest incentive for new people to join the platform. For all its downsides, steem has created something never before seen and there might be more bandaid solutions until we can reach an agreement on the proper direction to go.
Why not something as simple as a self-vote drains your voting power exponentially rather than linearly like a regular vote, this would do a lot to curb behaviour. Weighting content creators vs curators at 60/40 would also help with this. I think these little things could be enough to right the ship, but I agree with you that all options should be looked at and let the free market decide
Of course, none of this can happen before SMTs are ready for prime time.
There's no way to separate a self-vote from a regular vote. People can make multiple accounts and end up with the same result. Agreed on the creator:curator split, but it's another bandaid at best.
Sometimes though a bandaid is all you need to fix a cut. I'm not sure we are at the point where surgery is needed to fix the problem, but there is definitely a cut that needs immediate attention. I think maybe focusing on good rather trying to stop the bad is one way to look at it. If those who curate are receiving sufficient rewards for their effort I think the snowball effect will take over just like the slef-voting epidemic has done ths far.
There are the obvious self-votes, and there is covert self-voting - like delegating or giving away steem power to a bot that will always upvote your own content, having agreements like "I always vote for you, you always vote for me" with other account holders of the same size, etc. I think it's not good to promote the latter.
Yeah the last sentence is what I would worry about most. SMT's won't be launching till years end and by removing the reward pool I think you would be removing the biggest incentive for new people to join the platform. For all its downsides, steem has created something never before seen and there might be more bandaid solutions until we can reach an agreement on the proper direction to go.
Why not something as simple as a self-vote drains your voting power exponentially rather than linearly like a regular vote, this would do a lot to curb behaviour. Weighting content creators vs curators at 60/40 would also help with this. I think these little things could be enough to right the ship, but I agree with you that all options should be looked at and let the free market decide
Would behave within the limits of what the system allows.
Rational people act according to what they have to lose or gain. The more they have to lose according to their pyramid of values, the more cautious they will be.
My premise is that, under superlinear reward, if the top of the pyramid sends too much of the reward back to themselves, this should decrease the demand for Steem, making the price to go down, decreasing the wealth of Steemians, making the system self-regulating.
I'm also very much open to other alternatives.
Yes and it was well-demonstrated during the n^2 era that this often involves/involved self-enriching behavior. Having more to lose was actually synonomous with having more to gain by behaving selfishly, so there is an unfortunate symmetry which doesn't really help anything.
This ignores tragedy of the commons dynamics. Perhaps it might be true of stake became so concentrated that the top of the pyramid consists of stakeholders with enormous stakes (say 90%). With top stakeholders having 1-3% each this is not demonstrated at all. It is a hard case to make to someone with 1% that how they vote with their 1% has a huge difference to the overall price of STEEM and very clear that how they vote has a huge difference to their own personal wealth.
I do not think we want a system with top stakeholders having 90%. We need a better solution (or alternately conclude that proof-of-brain has no solution and rationally pivot accordingly).
Self-enriching can only occur if the demand for Steem goes up which is what is at stake here, the primary goal.
How I understand it, I came to the conclusion the current system should be modified for reasons I've stated.
I don't know what is the best solution and I respect and understand the changes can only come from consensus.
I'm grateful to be able to voice my opinion and have this discussion.
No that is not true. Most of the time when considering a specific action, the demand and price of Steem doesn't change, or change due to a variety of market factors which aren't connected to any individual stakeholder's action. If I vote myself a $400 reward, there is no mechanism by which that will reduce my personal Steem holdings by $400 or more. I just get about $400 richer. Maybe my actions reduce the value of my Steem holdings by a dollar or two, maybe not even that.
The self-enrichment is quite obvious.
Even if everybody does this 100% of the time (which isn't the case), it is far from clear that it would destroy the value of Steem. The value of Steem doesn't derive solely and directly from rewards being allocated in a particular way, or at all. The value of Steem comes from many different factors.
The model of destroying the value of one's investment as an incentive to behave in a particualr way is profoundly broken at multiple levels.
In case it isn't clear, I'm not advocating for everyone to take the most short-sighted and selfish possible action (though many will often do so anyway). I'm examining the logic supporting the model and finding it lacking.
If the order book doesn't change then, no matter how many more Steem someone receive, if there's no one to buy them, then in btc/fiat/etc, you haven't enriched oneself. That's my point or at least a gross simplification.
When everyone votes themselves exclusively, then votes become akin to interest on SP. As more investors join, we'll get ever closer to such a situation.
As this situation progress, SP use would be reduced to witness voting exclusively.
I have no idea what you are talking about in terms of the order book. At any given time it isn't hard to sell $400 worth of crypto at close to the market price (almost any crypto). If I earn $400 worth of rewards that I can sell it is very unclear what mechanism exists which would cause my stake to devalue by $400, and there probably isn't one. If it is just general inflation, then it doesn't much matter whether whether I earn the $400 or someone else does, but my personal wealth would prefer that I do, obviously.
No. In addition add: bandwidth, store of value, plus STEEM and SBD (both interchangable with SP to some degree) can be used a means of exchange. As bandwidth, SP may be valuable for other applications such as SMTs, Steem Monsters, censorship resistant accountable social platform, etc. Speculation about future value of these or other applications may impute a significant current value.
If someone increased their Steem stake by 1% every day but the price of Steem goes down by 2% every day then this person didn't make any money in fiat/btc/eth term. Their wealth has decreased.
How much money someone makes is dependant on the demand.
The demand can't be taken for granted.
True. I agree.
I agree with teamsteem. Why would people , stakeholders, want to shoot themselves in the foot?
This algorithm is the basis for an emerging competitor to steemit.. It's called Primas and is moving out of open beta and coming online this week. It's like Medium quality content with a community incentive system just like steemit but more fair and without the centralisation.
As I try to spread the ideas below I add my comment not only under the main article but also as response to your comment, because otherwise I guess you wouldn't get aware of it:
Some time ago I formulated some ideas to make self-voting, circle-voting and spamming less attractive by ...
@clayop had a similar idea.
When I wrote these articles some time ago I really had hoped that more witnesses/big account holders would have joined the discussion!
I can see some merit to that. It would have to be thought of more carefully.
The other 3 points should also be considered.
I think the "selfless/selfish" behaviour terms are not at all times representative and can be misleading so I'd prefer to avoid using them. With regards to the notion of free money, I'm having some reservations using the term as well, especially considering consistently high rewards some users are getting whether or not if PoB is effectively involved. But that'd be an argument of semantics in relation to behavioural expectations, not what reality actually is since we aren't 1-dimensional agents. I'd agree the leak is more flatly distributed across the board, unlike n2 (or any superlinear curve) which congregates such leaks into something more manageable.
Proof to be a better system in what terms? I guess there are different expectations all around. To me, linear has the shape of a fair consumer product, in that an acquisition and exercise of SP at any level isn't shortchanged by the system in favour of those with higher SP. I'd be pretty pissed off under n2 if I purchase or earn 100 SP, but someone with 10x more SP has 100x influence when it comes to stake-weighted voting, instead of a linear 10x. However, n2 does have its qualities, and on balance, I'd prefer taking only the best parts of it which are useful and hence, settling for something between linear and n2, even if only composed of multiple linear approximations to form something like, maybe, n1.5.
To repeat what I've posted previously about "slight superlinear":-
I think @trafalgar has commented with some better explanation and insights. So far, that's my take on it..
update: edited a little to clear up some wording..
A sinking boat vs a floating one.
The current system favors those who do engage in absolutely no work, aka proof-of-no-brain-needed. How could it be worst?
That would be irrational. Linear devalue the whole platform and thus Steem.
Regarding slightly superlinear, I think it would only slow down the desirable effects of superlinear.
I think a good way to fix steem's voting problem is to give big votes to all the guys from Portugal who have an idiot photo as a profile photo and whose nickname starts with a "d" and ends with "nunes"
see how wise I am?
Indeed..
Not sure if I'm willing to concede here with that statement, but I get what you mean. It'll be a hardsell for anyone who can put their fingers together in figuring they're being shortchanged immensely for the system.
Yup, our proposal here is an in-between solution for the individual and the collective. I think there's good evidence that n2 did not work as well as expected. Sure it amplifies and contrasts contributions to the greatest degree, but are users generally happy? Maybe in a purely elite community. I'd give it to Dan for figuring out that linear is highly undesirable in first pass though. That's certainly not something that's obvious to most developers.
Let's make it n^99999 so no one has any influence at all, good idea dude, you must be great at economics and by your suit I can see that you know what people on modern social media are engaged by.
Or we could just decide that 2 is clearly the ideal power to use in terms of economics and incentives. What a convenient coincidence that n^2 is easy to calculate in integer code using
rshares*rshares!I'm sure everyone in the >99% at least secretly likes >1.
Steemit users circumstances vary. Some post "full-time" and need an income from this and also to cover expenses involved in making good posts. We like to believe blockchain social media has a long term future. So with a modest account myself I would rather give the votes away at least in the forseeable future to encourage Steemit user number growth. Hopefully the Steem price will rise sufficiently one day to make super capital gains.
Why did not they block the opportunity to vote for themselves?
Because it's not a problem to make a second account.
Well, you create it and what's next? Vote for yourself is already prohibited. What's the meaning?
You vote for this 2nd account, or the 3rd account.........
Is the big profit from the plankton account?
I'm not sure what you mean, but I a plankton account is unlikely to make 'the big profit'.
I mean that the new account it's a plankton and what profit from plankton?
It's not that hard to send some steem to a second or third account. So in the end there are for example 3 accounts and each is holding one third of the amount of steem. In the end they are also voting for themself
Posted using Partiko Android
nope. I've got two; this is my personal one, and @stormlighttech is my project-based one.
You got a 17.27% upvote from @emperorofnaps courtesy of @teamsteem!
Want to promote your posts too? Send 0.05+ SBD or STEEM to @emperorofnaps to receive a share of a full upvote every 2.4 hours...Then go relax and take a nap!
This looks a tad shady :P
I bought the vote to make my comment appear at the very top. If I could go back, I would avoid buying it.
I BOUGHT THIS VOTE BECAUSE I FEEL THAT MY OPINION IS SUPERIOR TO EVERYONE ELSE'S AND I SHALL PROVE THAT SUPERLINEAR IS SUPERIOR BY SHOWING AN EXAMPLE WHERE SOMETHING WITH MORE MONEY BACKING IT GETS MORE VISIBILITY THAN EVERYTHING ELSE IN THE THREAD.
I tend to think Steem would probably be better without bid-bots but I could be wrong about it. Steem rules get decided by consensus.
I wouldn't want to be the one who has to decide them. I'm glad I can voice my opinions and participate in DPOS.
Now that bots exist, I've encouraged people to use them and I've done so many times and publically.
My intentions are also inline with the universe's intention to make a positive impacts in many people's life.
Good steemains over Evil steemains.. I know which side I'm in :)
“You can fool some people some times but you cant fool all the people all the time” - Bob Marley
N2 can be exploited equally as linear, just need to divide steem among many accounts. The solution is account based voting imo
Gaming it the way you're mentioning it isn't so straightforward.
https://steemit.com/non-linear/@felixxx/more-n-2-47
And as Dan pointed out:
There are very good arguments that n^2 has too much evil and prevents (or at least inhibits) Steem from growing.
I'm sorry, but you are elevating the Steem white paper to some sort of holy text or carefully researched and peer-reviewed science, and it really isn't that. It is a marketing pitch that got written and implemented in a few months to launch a blockchain product.
I've never implied the Steem white paper couldn't be wrong. I even said I could be wrong about linear rewards many times and I also said I would prefer to be wrong about linear rewards.
It's not necessarily the most fun feeling to disagree with people you care about. It's important to stand up for what we think is true yet remain open about the fact we could be wrong, which is something I've been upfront about.
Fair enough. And disagreeing does not mean that you do not respect or like the person.
For that matter, often when I 'disagree' I'm not even necessarily claiming the statement is incorrect as much as not well supported. (Unsure without review how much this applies to our recent spread-out conversation.)
N2 has it's own problems, like barrier to entry. Whales like Curie can't give massive upvotes to minnows and keep them motivated.
I think a hybrid way of keeping stake linear rewards on steem, and N2 on SMTs with account based voting is ideal. Keep everyone happy. You need to keep investors happy too or they'll just power down and look for different opportunities.
I agree other solutions are possible.
Dividing among accounts makes no difference at all. Votes are added together before n^2 is applied. So (vote_A+vote_B+vote_C)^2 is the same as one account D=A+B+C with all the stake voting once.
Nice huge self upvote :-) attempt at irony perhaps ?