Blockchain Governance Models | Should Steem Governance Model Be Changed?

in #hive-1679228 months ago (edited)

With the recent events on our beloved blockchain Steem and the softfork that freeze the Steemit Inc stake it is worthy to have a look at the governance model of the blockchain.

Aberrated Path.png

Before we dip into this topic first a glance of something called The Blockchain Trilemma.

How can a blockchain simultaneously solve for scalability, security, and decentralization? The consensus is it cannot.

This means that the level of decentralization and scalability/speed/efficiency are opposite. In order to achieve higher scalability and speed one need to sacrifice some levels of decentralization.

What Type of Blockchain Governance Model Exists?

Blockchain governance models are a wide topic and it can be debated a lot. Also, I won’t pretend I’m a great expert on the topic. Here I will just mention a few.

The governance models in most cases are divided in these two categories:

  • Off Chain Governance
  • On Chain Governance

In a lot of the cases something like direct and representative governance is also a metric.

For the proof of work coins (PoW), where miners are rewarded with the inflation, like Bitcoin and Ethereum the governance model is usually Off Chain. In these models there is usually some foundation or key players that make all the decisions. Everyone is welcomed to make suggestions but at the end the decisions are made by few. Some of the key players are core developers, node operators and major token holders. There is also something called Benevolent Dictator for Life, like Vitalik for Ethereum. The final decisions usually are approved by him.

The On Chain Governance usually means there is a voting mechanism for proposals, build in the blockchain. Some of this mechanisms can be trough proof od stake, delegated proof of stake, liquid proof of stake and sometimes a hybrid between proof of work and proof of stake.

Delegated Proof of Stake

As most of the Steemians know we have a Delegated Proof of Stake governance model for the blockchain. Witnesses are being appointed from the community to run the nodes and secure the blockchain. The more stake you have the heavier your vote is.

It is also a representative governance model, since the witnesses represent their voters indirectly.

Some of the pros for DPoS:

  • Speed
  • Scalability
  • Energy efficiency
  • Flexibility
  • Real time voting
  • Better distribution of rewards?

That last bullet is with a question mark, cause the theory here is that people will elect only those delegates who give them the most rewards. This may not always be the case and sometimes something called cartels forming can occur. In the Steem case witnesses gets only 10% of the inflation so this is not as hot topic. Imagen if witnesses got 100% of the inflation.

Actually, Steem makes much better job at reward distribution with its Prof of Brain mechanism than most of the other coins.

Some of the cons for DPoS:

  • it’s easier to organize a 51% attack
  • The rich may get richer
  • Cartels forming

Fewer people are in charge of the network (20 in our case), that leads to more efficiency but also it can be easier to organize 51% attack. Since the voting comes from the stake you have the more you have the more you control the network and the rewards distribution. Forming cartels is obvious, since I support you, you support me, is the easy thing to do to stay on the top.

Steem DPoS

In the Steem case we have a DPoS. By definition this is a representative governance and some level of decentralization are already sacrificed for speed and scalability. But not all DPoS are the same.
Steem goes further in the centralization with its 30 votes rule. One stake can vote 30 times. There are only 20 witnesses. If someone has 51% stake it gives him immediate control over the blockchain.

This is doubling down on centralization!

If the rule of one stake, one vote is applied, if someone has 51% stake in theory, he can vote half of the witnesses, just the top 10. The other top 10 wouldn’t be affected.
How many times can one stake vote can be a topic of discussion and the theory is that it should be between 1 and 20. The more times one stake can vote the higher the centralization of the blockchain is.

The one stake, one vote rule can be even further decentralized if weighted stake based voting is applied. This means that in order to increase the influence of the small stake holders, different weight of the different stake size can be applied. This means reducing the large stake holders influence and increasing the small stake holders influence. The stake limits and weights are up to discussion.

Incentive for voting. We have seen a lot of posts where everyone is calling for STEEM stakeholders to vote for witnesses to protect the blockchain. In a lot of the DPoS coins block producers/superrepresenitves/witnesses are sharing their rewards with their voters. This is the case for TRON. Now the rewards for those blockprodusers are much higher than our own here on Steem and usually a larger share of inflation goes to them. (10% in the case of Steem). I can guess if there was a reward for voting for witnesses the amount of stake that is voting would have been much higher.

Maybe we can allocate larger share from the inflation to witness rewards. The proof of stake (15% inflation share) would be no.1 candidate, and then allocate that to the witness voters.

Make Steem More Decentralized

As a final thought on the possibilities to make Steem governance model more decentralized (more secure) would be these things:

  • One Stake, One Vote
  • Weighted Stake Voting
  • Introduce Witness Voting Incentives

All the above are just an ideas and topics for discussions.

Maybe if we had a better governance model, the softfork wouldn’t been needed at all. The one stake votes 30 times rule makes Steem DPoS more centralized than the general DPoS models.

All the best
@dalz

Posted via Steemleo

Sort:  

I’d always side with having more witnesses and nodes but I guess theirs a cost involved that may put people off unless say the price was worth it

Maybe reduce to 10 votes and increase the top 20 to 25 or 30 and spread the stake backing around a bit! We still tiny and most people don’t even know about witness voting especially the new peeps here and well so many have voted and left so their stake is on witnesses not even here

I think the overall cost for a node is not as big, as the time, skills and effort needed.

I can't agree more. We could also look at increasing the number of witnesses to a larger number then 20.

When i 1st joined this blockchain I saw the TOP 20 as waaayyyy too centralized. I do agree with what they did as reading the history of the blockchain it should have been addressed long ago, but we need more decentralization for sure.

but we need more decentralization for sure.

Agreed on that

There are way more than 20 witnesses total (over 600 I believe last I checked?) but it is the top 20 that basically decide the direction of the blockchain. I agree increasing that number makes sense though.

Tweet:

I believe if we dropped the votes to 5 from 30 it would make a big difference in terms of security and decentralization. The more I think about it the more I like that number. The sooner we make that happen the better.

I still think one stake one vote is the right thing. But everyone can scale their vote. Meaning if I have 10M, I can vote 5 times by 1M, and 10 times by 500k ... so basically one can vote even a 100 times, but scaled until they dont drain their witness voting power.

Yes that would work as well. It would prevent the top 20 witnesses basically being selected by only a handful of accounts like we have now. I also would be on board with expanding the Top Witnesses classification from 20 to something larger, like 30 or 50.

There are some scenarios where only having 5 votes can lead to a minority stake taking control of consensus. Imagine a situation where everyone votes on the 5 more popular block producers...in that case a minority stakeholder can vote on 15 other witnesses and effectively own 2/3 plus one of the top positions.

It is not very likely but possible none the less.

What is more likely is that a minority stake can block a necessary HF by controling 1/3 of the top witness positions...we need a mathematician that is good with game theory to go over the possible solutions.

There isn't going to be a solution that covers all possible scenarios, only better ones than we have right now, and this likely is one.

In DPOS 51% of the stake does not give a stakeholder control of the blockchain. In order to take control you need 2/3 plus one of the top witness positions. The 51% figure only applies to POW or traditional POS.

This is all in theory since in practice voter apathy creates a situation where not all of the stake votes on block producers and you need much less stake to take control. If it wheren't for that soft fork 0.22.2 would not have been necessary since steemit only has ~30% of the staked steem.

Yes true the 51% is a figurative speech for taking control of the blockchain.

That is way its marked with italic ... but maybe quotes should been better :)

The overall staked STEEM is around 210M. The share of known SP to Steemit is around 60M SP + 5M liquid. This is on chain stake.

This post definitely deserves resteem by many.

Thank you!