Conviction Voting in ENS Small Grants

Conviction Voting in the ENS DAO: Why One-Token-One-Vote is a Better Alternative

I am concerned that our current conviction voting mechanism may not be the best approach for the DAO. While conviction voting has its merits, I believe that there are significant drawbacks that warrant a reconsideration of our voting system. In this post, I will outline the main arguments against conviction voting in the ENS DAO, focusing on the potential issues faced by small grant participants, and propose the one-token-one-vote alternative to better serve our community.

Arguments against conviction voting in the ENS DAO:

  1. Concentration of power: One of the primary concerns with conviction voting is the potential for some large token holders to exert disproportionate influence over the decision-making process. Since some members have magnitudes more tokens than the majority of token holders, this system may create an unfair advantage for proposals supported by these whales, leaving others with little to no chance of approval.

  2. Unfairness for small grants participants: Conviction voting allows whales to vote for an unlimited number of proposals with their full voting power, potentially ignoring smaller proposals that might not have the backing of large token holders. This can create an uneven playing field for small grant participants who may struggle to gain traction for their proposals. This is seen with proposals on winning streaks.

  3. Complexity and confusion: Conviction voting can be challenging to understand and implement, especially for newcomers. The complexity of the voting mechanism may deter participation, create confusion around voting outcomes, and lead to decisions that don’t fully reflect the community’s wishes.

Alternative voting mechanism for the ENS DAO:

One-token-one-vote: This straightforward voting system assigns one vote per token, ensuring that each token holder’s influence is directly proportional to their holdings. While this method does not completely address the concentration of power, it offers simplicity and transparency that could encourage broader participation. Most importantly, it prevents whales from voting on unlimited proposals with their full power, creating a more level playing field for small grant participants.

I welcome further discussion on this topic and am eager to hear the thoughts of my fellow community members. Let’s work together to create a more inclusive and equitable decision-making process for the ENS DAO that gives every proposal a fair chance.

I appreciate you reading, first time posting a fresh topic in here.

2 Likes

I don’t think I understand - the DAO uses one-token-one-vote, not conviction voting. Are you referring to approval voting, which is used for anything with multiple choices?

3 Likes

Ty for the response Nick. Bringing this up from Observations in the small grant rounds.

Please correct me if i’m wrong. A person holding 10k tokens can cast that same vote weight on multiple proposals.

e.g 10k votes on 5 proposals. Each proposal gets 10k votes. The delegate voting doesn’t have to split up into smaller votes. 2k x 5 votes for one example.

2 Likes

The Small Grants platform voting is conducted differently from the on and off-chain proposal voting.

If there is an available snapshot strategy, we’re open to trying it out on the Small Grants. The small grants, by design, is meant to be iterative and experiment with different strategies if requested.

Are you available next Thursday at 1pm EST to join our call?

3 Likes

Appreciate the breakdown Colton, Thank you.

Next Thursday sounds great mate :ok_hand:

1 Like

To clarify, in Snapshot’s terminology, the strategy determines how many votes each person gets, and then the voting system determines how those votes decide the actual outcome.

ENS Small Grants currently uses the Approval Voting System. I believe in this case, @workfromhome.eth, you’re talking about a different voting system like Weighted Voting:

3 Likes

Small grants utilizes snapshot which offers:

Single Choice Voting
Approval Voting
Quadratic Voting
Ranked Choice Voting
Weighted Voting
Basic Voting

I like the willingness to consider experimenting with the small grants platform.

Some additional ideas:

  1. Mirror the round using one or more alternative voting styles just to see if the outcome of the winners would be any different than the approval vote. (This is additional work/voting for the delegate, but probably not more than an extra minute per alternative voting style being tested).

  2. Bifurcate 1ETH from the current Approval Voting to one of the alternative voting styles which would go to the winner of the alternative vote such as single choice vote. (if the winner of the alternative single choice vote already won in the main Approval Vote, then go to the next in line so as to not duplicate a winner from the Approval Vote).

2 Likes

Thanks for this!

Great!

Oh, and we chatted about this a few weeks ago on the call. I believe @avsa and @184.eth both ran simulations of quadratic voting. I know this is different, but no significant changes would have occurred in the previous rounds.

Repeat Winners

There has been a skewed distribution of PG winners who have more than one win

To address the repeat winners, we’re hoping that our forthcoming larger grant streams will reduce these occurrences. I want winners of our forthcoming Growth Grants to be ineligible for small grant during the term in which they win.

Some Considerations

Even with a different strategy and graduating repeat winners we will have to accept some of the limitations/concessions that any erc-20 voting system brings.

6 Likes

There is another recently deployed module for Snapshot that allows for shuttered voting, or private voting. During the voting period the votes are kept private but then revealed at the end of the voting period.

Not necessarily good for all voting things, but might have value in some multiple choice, social votes.

It’s worth just adding to the overall vote strategy conversation. The module is from Shutter Project. Here’s a tweet with some links.

https://twitter.com/jnptzl/status/1640900035423395841

6 Likes

@Coltron.eth I loved the breakdown for this in the ENS Town Hall Q1 2023 stream. Super insightful, and excited to see the different/new strategies being thought about for the future.

@workfromhome.eth I see it’s your first post, welcome!

4 Likes

good work

1 Like

I hope early enjoy ENS Email service, :slightly_smiling_face:

1 Like

Personally, I’m a fan of quadratic voting - interesting to hear that @AvsA and @184.eth ran the simulation. Would love to check out the data and see how the data changes upon the implementation of a quadratic vote.

I think that this is the most fair approach for the way token distribution sits.

If a delegate ‘VOTER’ has 10,000 ENS tokens to use as voting power; I believe the following is fair and reasonable.

In my opinion a vote should not look like this by ‘VOTER’

Project A: 10,000 votes

Project B: 10,000 votes

Project C: 10,000 votes

If ‘VOTER’ allocates 10,000 votes to one project, then all their voting power is consumed per se.

If ‘VOTER’ wants to vote for more than one project then if:

Project A: (receives) 5,000 votes, then that would mean that ‘VOTER’ only has 5,000 more votes available to cast.

Project B: 4,000 votes

Project C: 1,000 votes

etc.

I’ve been closely observing the voting mechanisms in various DAOs and have had experiences with conviction voting in the past. One of the primary concerns I’ve noticed with conviction voting, especially in the context of ENS small grants, is the potential inefficiency in fund allocation.

I’d love to introduce Pairwise, a voting mechanism designed to simplify choices and amplify voices. For voters, instead of viewing all the proposals at once, which can be overwhelming, they are presented with a prompt featuring two projects. This approach prevents voters from merely selecting familiar proposals and overlooking unfamiliar ones. Voters choose one of the two projects, and an ELO system ranks the proposals. For a deeper understanding of this system, please refer to the Colony docs.

f0b35e2b99a4f05aca04c727637a5143

The reason we present two projects for comparison, akin to the Tinder style, is that it simplifies decision-making. For instance, deciding whether to spend 10k on design might be challenging. However, choosing between spending on design or development becomes much clearer.

After the ranking is established, there’s an option to edit votes before finalizing them, adding a human touch to the decision-making process. We’re currently developing this feature for Optimism.

For ENS Small Grants, Pairwise can be utilized in various ways. One approach is rank choice voting, where, for example, the top five projects each receive 1 ETH, as done previously. All projects that win a vote against another will have a certain percentage, and we could distribute the entire 5 ETH based on the percentage each project holds.


To bring this vision to life, further development is required. I’m keen to gauge community interest in this approach. The ultimate goal is to foster digital democracy, simplifying decisions to garner more community signaling. Moreover, we can integrate any Snapshot strategy we desire, including quadratic voting.

2 Likes

I’ve long been a fan of pairwise voting. I’d love to see Small Grants experiment with it.

3 Likes