[DISCUSSION] TOP 20 Validators/Validators votes

[Conversation]

Positive this is a very popular, well rounded topic of discussion, however:

If we are all to simply accept the CP being equivalent to the [edited] 5th (currently) ranked Validator, in relation to its weight in ATOM (currently), curious:

1. Are we ready to remove TOP 20 validators influencing on-chain voting?
2. Are we ready to decouple Validators from their delegators votes?

If not, would sincerely appreciate your time in helping to understand why.

Thank you in advance.

I have made my peace with the idea after Atom One forked. Now we have two philosophies in parallel, and somehow it is good for the ecosystem.

What’s wrong with copying something that another chain is doing and it’s beneficial?

I also believe that, in the long run, a political reform will be necessary to strengthen the network’s resilience and its democratic structure. Imposing a cap on the voting power per validator (for instance, no validator should hold more than 5% of voting power over the chain) seems to me a strong way to improve resilience. Similarly, as you mentioned, separating the voting power of delegators from that of validators (or at least allowing delegators to freely choose the address that represents their voting influence) is also a valuable idea.

That being said, even though I’m strongly convinced of the need for such reform to ensure the network’s democratic health, I believe now is not the time to focus on resilience. The priority should be to establish a clear territory and a solid use case. But I can assure you that when the time comes, I’ll be the first to bring these ideas back.

Vote delegation is a very important topic indeed, one that we have covered years ago, even though it failed to catch the community’s interest so far. It may take time for many to fully appreciate where vote delegation is likely headed in the long run. Based on our research and past contributions on the topic, we believe the most robust governance architecture will naturally evolve into the following layered structure: 1. Direct Participation The default and most desirable outcome: The token holder votes directly. This ensures the most direct form of democratic expression. 2. Delegated Governance via Councils If the token holder chooses not to vote, the delegated representative — a member of a governance council — inherits the voting right . These representatives, or “governors,” would ideally be subject-matter specialists, selected by the token holder to reflect their values or interests. Importantly: • The token holder retains ultimate control and can override any vote cast on their behalf at any time before the voting period ends . • They can also revoke the delegation altogether if they feel it is being misused. 3. Validator as Final Fallback…

Excerpt (1194 of 2309 characters). Read the whole post on the forum ↗

I also think the direct votes should override validators’ vote.

Appreciate validators casting ‘abstain’ votes on topic influencing on their revenue or other topics where their votes can have a bias.

Firstly, thank you all for sharing your time/valuable input here. waqarmmirza: I have made my peace with the idea after Atom One forked. Understandable. StunZeed: What’s wrong with copying something that another chain is doing and it’s beneficial? Agreed, especially if the model chain specializes in governance issues, specifically. Quentin: The priority should be to establish a clear territory and a solid use case. But I can assure you that when the time comes, I’ll be the first to bring these ideas back. Respect the drive to be first. May the Cosmos Hub prosper. Govmos: it failed to catch the community’s interest so far. Your teams work and tonal posture is highly appreciated. Keep up the great work (+everyone here) Govmos: The default and most desirable outcome: The token holder votes directly. Govmos: • token holder retains ultimate control and can override any vote cast on their behalf at any time before the voting period ends . • They can also revoke the delegation altogether if they feel it is being misused. Agreed. Govmos: the validator may…

Excerpt (1197 of 1744 characters). Read the whole post on the forum ↗

[Correction] RE CP weight. It is roughly equivalent to rank #5 . Thank you, Govmos, for sharing this valuable thread “Ideation - Governance…” and the work you, and your team, provide. Some key points (that personally stood out): IDEATION : Governance Councils & Treasury Modules Validators start off with to much power. making their voting power flat self-stake would make everyone equal. Response: IDEATION : Governance Councils & Treasury Modules but our model doesn’t want to crush validator power, just rebalance. We think that this will mostly happend with the LSM (liquid staking module) if we improve it by adding a “Vote Transport” function that moves the vote along with the liquid staked representation. Response: IDEATION : Governance Councils & Treasury Modules that removes the necessity for those with the most power to have the most skin in the game. IDEATION : Governance Councils & Treasury Modules after separating security and gov, what stops the councils from voting to pay themselves the treasury, and minting 40 million more atom? Not sure how to feel about your ideation being based off…

Excerpt (1197 of 2391 characters). Read the whole post on the forum ↗

FHZ:

Validators voting power = self stake.

Considering the usual very low participation to online governance, don’t you think many proposals will not even reach Quorum with this kind of rule?

I like the idea and it totally makes sense but self stake voting power is no different from a delegated voting power where the staker doesn’t vote. That why I think :

• no one can overrule validators’ self stake voting power
• validators can cast ‘abstain’ votes to emphacise on unbiased position towards the subject; when they do, delegators’ vote overrule their ‘abstain’ vote
• validators can votes with full voting power (including delegated votes); however, if delegators decide to cast votes on a subject where their validator already has experessed their opinion, votes from delegators should overrule those from validators

This would allow delegators to express their opinion in a much mode direct democracy on any subject. For now, this voting power is being delegated and can’t be overruled unless validator decides to ‘abstain’ voting.

FHZ:

Validators voting power = self stake.

While relying exclusively on self-bonded stake to define voting power may be too rigid a solution, we believe you’ve touched on a compelling idea. At Govmos, we find it particularly elegant to consider incorporating self-delegation into the vote inheritance model.

Rather than applying it as an absolute rule, we would suggest a more nuanced, relative approach—potentially using self-bonded stake as a cap on the voting power a validator can inherit. This would mirror the logic used in the Liquid Staking Module (LSM), where validator-linked shares are bounded by self-bond levels.

We’d be especially interested to hear @Noam’s thoughts on the technical feasibility of implementing such a mechanism.

kov0x: don’t you think many proposals will not even reach Quorum with this kind of rule? Good question and unsure the answer. How is the quorum, in this proof-of-stake system, calculated exactly? If, within the current, low participation environment, Hub Validators were completely decoupled from delegators voting weight, would we still be reaching quorum? kov0x: this voting power is being delegated and can’t be overruled unless validator decides to ‘abstain’ voting. Was not aware of this. Unless a validator specifically votes ‘abstain,’ delegators casted vote are overruled, regardless? Govmos: consider incorporating self-delegation into the vote inheritance model . Absolutely. Govmos: potentially using self-bonded stake as a cap on the voting power a validator can inherit. Interesting and good suggestion, especially when personally against relying on LSM, completely. Reminder: Prop993. DROP now has “indirect” say RE Hub governance, through Neutron. Why? If DROP were to experience a bug, or their system goes down, those remaining silent on this matter, are not only saying they are fine with valueless…

Excerpt (1190 of 1661 characters). Read the whole post on the forum ↗

I think it’s a good idea to base the voting power on self bond, but you still need to set an asymptote (for example, a 5% cap on voting power), because otherwise you end up with voting power becoming centralized on centralized exchanges, which are mostly self bonded.
On the other hand, the issue you’ll face if you try to cap voting power in any way is that you might end up with the same validators appearing multiple times under different identities to bypass the cap. It’s tricky, and this would require regulating validators by prohibiting them from duplicating their nodes, with heavy slashing penalties enforced by governance, assuming good faith.

These are all absolutely valid points and we are 100% aligned with them.

Quentin:

if you try to cap voting power in any way is that you might end up with the same validators appearing multiple times under different identities to bypass the cap.

What are COI’s and do they matter? Like Govmos said, valid points made.

Curious to hear more thoughts on this topic.

Quentin:

5% cap on voting power

Great suggestion/start.

Quentin:

the same validators appearing multiple times under different identities to bypass the cap.

and ensuring proposals are approved(?)

kov0x:

For now, this voting power is being delegated and can’t be overruled unless validator decides to ‘abstain’ voting.

This is not an issue for the broader Cosmos Hub?

FHZ: and ensuring proposals are approved(?) A validator operating multiple nodes would already gain a significant advantage over others, regardless of how many millions the other validators have in staking, since their voting power would be capped. An actor that centralizes enough voting power could unilaterally influence governance in several ways: by approving or rejecting proposals, or even by abstaining from voting in order to prevent quorum and effectively block decision-making, this tactic is even more relevant today as proposal durations are being shortened. Another potential threat is the initiation of malicious proposals with significant support from fraudulent validators. Additionally, if the largest validators lose a significant portion of their voting power due to caps, smaller validators could gain relatively more influence. This shift could incentivize attempts to bribe or acquire lowertier validators, whose voting power would become more valuable in comparison. Of course, these are just hypotheses, but we should carefully weigh the pros and cons of such a mechanism. This is a complex and delicate reform. While it could lead to a more equitable…

Excerpt (1192 of 2184 characters). Read the whole post on the forum ↗

Govmos:

These representatives, or “governors,” would ideally be subject-matter specialists, selected by the token holder to reflect their values or interests.

May be worth implementing this inspired suggestion made by Govmos’ (thank you), with the caveat of removing, Validator’s as a “final fallback.”

Quentin:

this tactic is even more relevant today as proposal durations are being shortened.

We should be encouraging members (delegators/stakers) to stay active and vote, while completely separating voting weight from centralizing validator groups/committee’s.

Quentin:

we should carefully weigh the pros and cons of such a mechanism.

Are we surprised more Validator’s are not engaging with this topic? Is it due to COI’s?