Governance Decentralization Through Imposing a Hardcap of up to 5M OSMO Per Validator

Proposal Discussion
1

We recognize the pivotal role that decentralization plays in fostering a truly inclusive and democratic governance model within our Osmosis community. Currently, we observe that a substantial portion of the Osmosis validator sets is concentrated within the Top 50, comprising over 70% of the total 352M Osmo staked. This concentration results in a governance structure heavily influenced by the top 50 validators.

To address this governance concentration and promote a more evenly distributed staking landscape, we propose the implementation of a hard cap on maximum staking and voting power per validator, set at 5M OSMO. By enforcing this limit, we aim to facilitate the redistribution of voting power across the entire spectrum of the 150 validator set. Even this hardcap might not be perfect to achieve the decentralization among validaotr set but atleast it will make some balance compared to current state.

This measure is designed to enhance the decentralization of governance, ensuring a more diverse and inclusive decision-making process. We believe that by encouraging broader participation and stake distribution, Osmosis can strengthen its resilience, security, and overall community engagement.

Your support for this proposal is crucial in shaping the future of Osmosis, fostering a governance system that truly represents the collective voice of our diverse community. Let us work together to fortify the principles of decentralization and ensure the long-term success of Osmosis.
Looking forward to get your feedback on this proposal

2 Likes
2

Glad this came up again! I am very much for decentralization initiatives, and they haven’t really come up since we moved the discussion forum!
(I got passionate enough that I needed subheadings :zipper_mouth_face: )

Validator interests

One issue is that these votes require either convincing larger validators to vote against their own interests (or potential growth) in order to make the chain itself healthier or having a situation where the bottom 51% are able to outvote the top 49%, without worrying that they will be unfairly penalized for entering the top half of the set.

Since a 5 million cap would directly impact the top 10 validators (34% of stake) as well as be something that the middle 33% may also see as impacting them so I think this is likely too low.

Fixed vs Percentage cap

A percentage cap may be better, as it would scale with the amount of stake bonded as well as inflation.

Sybiling

The other, larger issue that we have run into before with decentralization initiatives is that validators may just spin up multiple entries into the set in order to obtain a greater share than the cap.

So far, we have not found a way to combat this since it would rely on a social slashing mechanism. While keeping validation open to all and anonymous, this is the only available penalty - where stakeholders make the active choice to avoid sybiling validators. However, the fact that we need to have these discussions at all shows that stakers often delegate to high stake validators!

Why is there any issue?

Some of this is due to the displays, until recently Keplr didn’t show warning for the top validators, and some other wallets actively promote their own validators. Hopefully, the new Osmosis staking interface will begin to provide a better staking interface to encourage the selection of multiple lower-ranked validators.

Some of this is also that validators stake their own, or custodied funds. These can often be spotted by the high commission rates compared to standard.

Statistics

Finally, I think this is a great conversation to have, but Osmosis is actually doing really well for decentralization compared to most Cosmos chains with the 5th highest Nakamoto coefficient of 10!

https://decentralization.smartstake.io/?ref=osmo

This seems to be trending upwards over time too!

image
image888×409 18.2 KB

Extra pushes will help for sure, but this definitely isn’t a crisis compared to when the original discussions were being had. Dropped these here if anyone wants to read the archives.

3 Likes
3

Osmosis is becoming better and better over time.

I personaly don’t believe in a delegation cap indeed due to the risk of sybilling.

One thing where I would believe in (but is a bit more sophisticated) is to find a way where non-voting validators lose their voting power over time. So not touching their delegations or staking rewards (because there is where it becomes painful for delegators as well), but losing rights to vote in governance.

If it would be for example possible to include a factor which involves the voting participation of the last 10 proposals.
A validator participating in all proposals gets 100% voting rights.
A validator participating in 8 / 10 proposals gets 80% voting rights.
And so on.

That might trigger lesser active validators to step up (which is a win) OR make the share for governance-active validators in governance bigger (also a win). That way some of the really bad governance validators will fall out, making it easier to bring changing proposals to the chain in due time (if needed).

2 Likes