Why our way of distributing Steem by voting sucks, why it does not scale and how we can improve this situation

in #steem6 years ago (edited)

Have you heard anything like that in the past?

Reward distribution sucks, because of circlejerking and greedy whales!

When you see some crappy content which receives huge upvote, it is easy to notice that something is wrong. But I believe this is not a result of people being greedy... but rather this is a natural consequence of asking them to do something impossible:

Simply find the best undervalued content among thousands of posts in hundreds of languages about on topics which you know nothing about and vote for it :smile:

And if you are a whale...

...just make sure, that your vote is not too big for some posts. Simply divide it among hundreds of posts, to make sure that distribution is fair!

Simple, right? What can go wrong?

I do not consider myself as a whale right now. For sure around 50k SP is a lot, but still, it's almost nothing in comparison to stake of people which has over 1M SP. But there was a time when I felt like I was I whale...

Few months ago, when STEEM price spiked, my single upvote was worth almost $25.

And because back then a Steem Dollar was worth much more, my single vote was actually worth over $50. That was crazy! 10 votes per day, 30 days per month... I realized that now I am responsible for distribiution of over $15000 per month. This is 4-6 times more than a sallary of a senior progammer in Poland.

I realized, that I am overwhalmed by this amount.

It was then, when I realized... that being a whale is not only a priviligae, but also a huge responsibility.

$15000 per month could for example help me reward 1000 new users. Did I reward 1000 users? Not even close. I failed because of many reasons. But if I failed, do you think a whale which has over 500k SP have succeded in rewarding 10000 new users per month?

Why voting is so damn hard!

When my vote was worth up to $1, voting for me very easy. It was trivial for me, to find 10 posts a day, which deserved to be rewarded by $1 upvote. Things become more difficult while my vote become stronger and stronger.

When my vote was worth $10, I was able to find a few undervalued post good enought to be rewarded by $10 upvote, but rest of the posts I tried to reward with partial upvote (50% upvote, 25% upvote, etc). The problem with that approach was... that now I was forced to vote not for 10, but for 30-40 posts if I wanted to use all of my voting power. And because not every post which I read deserved to be rewarded, I was forced to read over 100 posts a day, if I would like to vote and support only good content.

Is this really possible? And how many posts would need to read a whale, if he or she would like to vote only for great and undervalued content?

Do you still think that whales are responsible for poor distribiution? Or maybe we have a more general problem - a problem with system, which simply do not scale when it comes to voting, as a fundamental mechanism of token distribiution?

We are jammed, more than a Bitcoin

When I first said that at loud to few of my friends on Steem, they ware really surprised. They was sure, that Steem has a capacity to handle 10 times more transactions without any problems. That is also a conclusion, when you are looking at blockactivity.info:

But exactly like number of megapixels in your camera is not only one thing which matters, similar big number of processed transaction also is not a recepy for an efficient blockchain. Blockchain is responsible not only for processing transaction, but also for distribiution of rewards.

Bitcoin use Proof of Work alghorithm to reward one miner at a time, which calculated a proper hash. It is important to notice, that while producing a content (a hash) for bitcoin is extremly difficult, evaluating this content is rather very very simple. That's why Bitcoin can reward objectivly "best hash" every block.

basically, whole situation we could illustrate like this:

Increasing an efficiency of content evaluation

Let's first ask silly question:

Could we implement an alghorithm for Steem to reward objectivly best content on the platform every single block?

Nope. Why? Because evaluating a real content is extremly hard, and it is even more difficult if you would like to be objective about that. If we would like to improve voting on Steem, and increase an efficiency of content evaluation, we have to understand why evaluation of content is hard in the first place:

Why evaluation of content hard?

  • lack of knowledge about a topic of a post
  • lack of interest about a topic of a post
  • plagiarism
  • different language
  • lack of time

As a community, we are very fortune that from early days, we have bots like cheatah (build by @anyx), which detects plagarism on Steem. Without cheatah situation would be much, much worse (thank you @anyx for being a hero of Steem!).

But still, we have a problem with 4 other causes. If only we could figure out, how to design a mechanism, to overcome those difficulties and to distribiute in a right way milions of dollars... if only we could learn how manage huge budgets efficiently and not spend it for crap...

Delegation of responsibilities.


Progress of humanity, would not be possible if people would not specilized. We need to understand one fact: sadly, you will never be an expert about everything.

If you have a problem with a law, you do not want to spend a hundreds of hours to learn all the law, you simply goes to a lawyer. If you want to know why your car isn't working, you are not studing mechanics, you simply goes to car mechanic.

So, when you are seeing a nice post about medicine but you are not an expert in that... why you suddenly want to evaluate it on your own seeing it has a few nice pictures and a lot of well formated text, which you don't even want to read?

If you want to achieve a lot, you should focus on things which you are good at, and you should delegate the rest of things to professionals which you trust.

This is how each and every empire was build. This is how Steem can be improved or maybe even fixed.

Allow others people to help you curate a great content

Did you had a situation like that, when your friend asked you to vote for some content, because it is awesome acording to this person? If your trusted friend would be a medicine-expert and he would say to you, that some post about medicine is world-class, I bet you would not use to much your brain power, to decide to support this content. After all, you got a recomendation from an expert in this field. Probably you will just vote for that, because you trust your friend and his expertise.

  • Suddenly, your lack of knowladge about a topic of a post was not a problem
  • Suddenly, your lack of interest about a topic of a post was not a problem
  • Suddenly, your lack of time to read this post was also not a problem

and even if that article about medicine would be written in language which you do not know, but an expert which recommended this to you would know this language, probably this would also not be a problem for you. You would just vote for that.

Don't trust, verifyTrust, but verify what you don't trust

People in blockchain space are saying, that you should not trust anyone. From a technical point of view I fully understand that. But we are building a social network here... which is about being social!

You cannot build a lasting-long relationships without trust, so you will not be able to build fully functionall social media network without leveraging a trust between individual members of this community.

Trust is subjective. Trust is also not a binary function. You trusts your friend, but this doesn't mean, that you would like to give him an access to your private keys. You can trust your friend that he is a good guy, but maybe you do not trust his ability to keep passwords safe.

You can trust your friend about everything related to medicine, but this doesn't mean, that you would like to give him an access to your SteemPower, so he could not only vote for posts about medicine... but also to vote for posts about politics which you dont like.

This is why SteemPower delegation do not solve a problem. SteemPower delegation is a binary thing while trust is not. With SP delegation you need to trust someone 100% or at all. Yes, you can delegate only a 50% of your SP, but you still need to trust this person 100%, to not use this SP against your will. And if this person will do something with your Steem Power what you don't like... you can do nothing about that. You can cancel a delegation (and wait a week, to get it back - LOL), but you cannot cancel upvotes which you don't like even despite the fact, that you funded them.

What a wise person like you, can do in this situation?

You can use a help of your wise friends.

What if you could use a some kind of system, which you could configure in such a way, so your friend could use your single votes on his own, but...

  • without compromising privacy of your password or private keys
  • only for post with tags specified by you, like #medicine
  • no more than 3 times a week
  • only if you have more than 90% of voting power
  • and with many other rules which you could customize...

What if you could stay in total control of your votes and always be able to cancel those votes which you don't like? What if you would not have to worry anymore, that your voting power is wasted, because you want to go for a 1-week vacations?

What if we could truly leverage a value of human connections, trust between people and their huge expertise?

And what if we would not have to ask what if, and we could simply check all of that? :)


Sort:  

@noisy Darn this is well thought out and written. I don't always read these types of posts. You have taken a really hard hit. The fact you didn't jump ship says something about you.

Also, you have given me a better understanding of the Whales voting situation. The ones that don't abuse their power.

I agree and feel you are right on when explaining why evaluation of content is hard. Some of your solutions I see whales already implementing.

The only point I disagree on is

What if you could stay in total control of your votes and always be able to cancel those votes which you don't like?

I am in observation mode at this time so you post could not have come at a better time for me.

What if you could stay in total control of your votes and always be able to cancel those votes which you don't like?

What I mean is a such possible situation: Imagine that with current system you delegated your SteemPower to your friend, hoping that he will vote only for good posts. But for any reason he decided to vote (with your SteemPower) for something what you don't like. Currently with SteemPower delegation, you can only ask your friend to revoke his vote, he he do not have to listen. Or maybe he will listen, but he is not available to cancel a vote.

New single-vote delegation system, which we've actually build on top of Steem, will allow you to stay in a control. Truth is... that this fact alone, that you can cancel a voted casted by a decision of your trusted person... makes this person even more reasonable.

You don not need to cancel any votes, but we believe that you should have a right to do so :)

The only point I disagree on is

Could you elaborate what drawbacks of such approach do you foresee?

Well, this, of course, is from a narrow perspective, mine. A little fish on Steemit. The hardest part about getting a whale vote is not getting one the next time. Especially in the beginning, later you learn if you write something good and the right person sees it it happens.

Then you know of other little guys who are in the Whales circle. They continue to get the vote even for some little post of really little value to many people. (again my limited view)

Your question:

Could you elaborate what drawbacks of such approach do you foresee?

Here goes:

It would piss me off and probably other people too. The thought would be why do you let someone curate that upvotes things you do not agree with?

Suffering withdrawal from not getting another whale vote is an actual vote is an actual thing but there are plenty of whales in the sea.

The hardest part about getting a whale vote is not getting one the next time.
Exactly! Because even if you encounter a good post of one author once, it is really difficult to come back to everyone and reads all those posts. Simply... there is too many of them!

But if a voting power will be spread among 20 trusted experts... there is much higher chance that one of them will come back next time.

Then you know of other little guys who are in the Whales circle. They continue to get the vote even for some little post of really little value to many people. (again my limited view)

From my perspective... this looks exactly the same. But I do not believe whales are doing that on purpose... in my opinion many of them are doing that mostly because finding new good authors is really hard.. so they prefer to avoid the pain of looking for... and they decide to reward an author which they know, which is above average.

But this is a problem, because this author have less and less reasons to produce better and better content... because he gets rewards anyway. That's why I do not like an idea of auto-upvotes.

It would piss me off and probably other people too. The thought would be why do you let someone curate that upvotes things you do not agree with?

I agree with that. But once again... this is an option, which nobody have to use knowing what a reaction of people can be. But in my opinion, this is nice to have feature.

I am sure you will do what is best. You have earned my respect by talking about solutions instead of leaving. :) Wishing you much more success.

I like your responses they are reasonable. I do think people make friends and there is nothing wrong with supporting your friends. :)

OMG,previously you can upvote for $25 now only $2...

Ups and downs....we're used to it. By the end of the next bear market that $25 will be standard

I suppose all of these things could/should be implemented to aprove the Steem blockchain!

Very good reflection, a truly important post ! Looking forward to meet and get to know you at #Steemfest3 in Krakow and take this discussion further. Steem belongs to us, those with SP, whether minnows or whales, we need to reflect - as you are doing here - and then get together to take it in the right direction.

Very well thought of ideas. I actually have 3 things in my mind how I can reverse it. One is making kind of a list, Second, is using a beneficiaries mechanism and third is to be hands-on and make sure you are not voting people and communities that are full of shit, if you know what I mean. I am far from being a whale but soon, I will do a run of these things and see how it goes... Great post @noisy.

One is making kind of a list, Second, is using a beneficiaries mechanism and third is to be hands-on and make sure you are not voting people and communities that are full of shit, if you know what I mean.

could you elaborate this little bit more? I think I didn't fully grasp your idea.

It is a really tough thing to solve and I have thought a lot about the game theory behind all of this and where other social networks have in a way succeeded and became what we were used to and why STEEM feels weird and unnatural to most people.

Pretty much anywhere else curation of content isn't being paid for. It is more naturally currated by how entertained or engaged a person is.

Also there is no other social media site that is this open to where so much can be accomplished openly with automation.

Since curration could be passively automated with upvote bots this encouraged the content creators to make "content mills". Producing cookie cutter content and cranking it out consistently so they could be a "safe bet" for those with the ninja mine stake.

We see content mills / content farms on YouTube that somewhat exploit what they perceive to be the current YouTube algorithm but for the most part most of the channels that have grown a big audience made sense and was entertaining to a large set of REAL viewers

The next nail in the coffin was the ability to earn a bigger profit selling votes instead of doing manual voting or even the automated voting.

Now we have these people who have been rewarded for having a cookie cutter "content mill" where none of the content they produce could go viral anywhere on the Internet able to then turn around and say

Thanks for the votes....... now you can buy my vote!!!!

Solutions

Not easy to solve at all. First of all the distribution was messed up on this platform from the start. That made the system feel way more unnatural and weird. This could be resolved on here but won't be. Strategic airdrops from the Steemit account would be my solution but that won't happen.

Trending and Hot pages have to be changed to reflect actual content that is viewed a lot. Doesn't have over a certain percentage of the vote weight from bots, takes into consideration voters reputation and at least 200 SP so that the algo can't be gamed by large bot trails...etc. Nothing will be perfect but if those pages were changed it would take the incentive for people to be playing king of the mountain with the vote bots.

There is no content produced for the STEEM blockchain that could go viral anywhere else on the Internet basically the way things are now.

Also I think that ads being implemented on the platform and then doing a revenue sharing with content creators could help make there be an incentive to actually create content that drives traffic. Instead of just having content mills.

It is a tough problem to solve but ultimately I don't feel that being able to take away a vote after delegating power will do too much.

Surely your answer can be found in a curation trail but this isn't going to save Steem.

Steem needs real-world adoption before it'll be taken seriously.

Imagine people using Steem to buy things in the real world posting a payment proof and you upvoting 10 of them a day so they can recoup some of their costs. Would this not be another way curate positive actions on Steemit?

Now imagine if we targeted a particular business like this for a number of months, then moved onto other companies as the movement grew. Wouldn't this real-world adoption be newsworthy?

Here in Australia Travelbit opened their platform to Steem and nobody used it, and now they are trying another shitcoin.

Another missed opportunity. IMHO

People seem to think the way you make money here, is to accumulate more Steem, but the real way to make money is to ensure the price of the Steem you already have goes up.

Real-world adoption is the tipping point we need.

And a little creative thinking could move that needle.

Bonus Points: Choose a 3-World Country and give the unbanked a reason to adopt Steem. It'll only take one merchant to get started and when a poor community realize they can get stuff for free by posting receipts, the idea will spread quickly.

"Steem feeding the hungry" - Now this is news-worthy.

Good Luck with your burden, but it's a probablem I'd like to have.

Great write up. For Steem succeed, the successful delegation of whales (and or orcas + dolphins too) to responsible curators is key.

It's not easy to find great individual curators who consistently vote on content however, esp in a bear market.

great individual curators who consistently vote on content

with such single-vote/voting-power delegation system, you do not need to care so much about consistent voting.

You can delegate you voting power to 30 people you trust, restrict, that they can use your voting power only if you have more than 95% VP... and those 30 people will utilize it according your rules. If 10 people among those 30 will not vote - this is not a problem, because other 20 still will be able to utilize your voting power :)

I think Steemit works pretty well considering the complexity and originality of what is being done here. But it sounds like your wise person system would be a help. Also it seems like it should be pretty easy to program something like that. I've actually thought of something very similar and it's on my list of things to do. Are you planning to create some sort of application that would implement this wise person's system?

Posted using Partiko Android

i see this post correcting a lot of mistakes done by several steemians...i feel it is very important to note that most original contest are also written in different languages but you cant vote it because you cant understand what the blogger is writing...
i dont mind serving a purpose of assisting any whales of searching for good content or assisting in upvoting original content if delegated to.
as a minnow i find it very difficult to attain a 1$ payout on any of my post so i prefer commenting on different post and expanding my knowledge for my own good.
good post @noisy, i can assist you spread the word.

Coin Marketplace

STEEM 0.23
TRX 0.12
JST 0.029
BTC 66687.47
ETH 3486.23
USDT 1.00
SBD 3.19