I wanted to start a conversation about a topic thatās been on my mind lately: shielded voting.
For those unfamiliar with the concept, shielded voting involves encrypting votes during the voting period and decrypting them only after the vote closes. This approach can bring several interesting benefits to our governance processes:
Mitigates Bandwagon Effects: By hiding votes until the end, shielded voting prevents people from being influenced by early voting trends. This could encourage more independent and thoughtful decision-making.
Reduces Voter Apathy: When voters know their vote counts equally, regardless of when they cast it, they might feel more motivated to participate.
Prevents Last-Minute Strategic Voting: With shielded voting, thereās no way to game the system by casting a vote at the last minute to sway the results.
While shielded voting sounds promising, Iām curious to hear from you all:
Could shielded voting improve our decision-making processes in Optimism?
Are there potential downsides or challenges we should consider?
Should shielded voting be applied to all proposals or only to specific ones, like elections?
How could we implement this in a way that aligns with Optimismās values and goals?
Your thoughts, concerns, and ideas are all welcome. The more perspectives we have, the stronger our governance can become.
Some risks that I see: since votes are masked, malicious actors can congregate and vote in a certain directione (i.e. the recent Compound Gov Attack), and good actors are left without knowing any context on existing votes.
Personally not a fan of shielded voting, all the points you bring up are valid, but it also means that we might not even know if quorum has been reached on multichoice options where 1+ choice is valid (eg. electing 5 members with 20 nominees).
Additionally, with the current on-chain voting system someone could possibly miscast their own vote and not know until the vote has passed with no chance of knowing it. Itās already happened to me in other shielded voting systems, and feels like a tangible downside for an upside thatās not guaranteed
I see the value of not āfollowing the trendā; However, I find the point as a whole a bit conflicting. While it could prevent bias from popular opinions, it also eliminates the potential benefit of understanding the rationale behind othersā votes. Reading othersā reasoning before casting your vote can enrich your thought process by introducing new angles or considerations that you might not have initially considered (and even make you reconsider your initial stance).
Gm! Shielded voting has been discussed and tested amongst a few larger DAOs in the industry in the past.
Aave for example performed a shielded voting trial and found it to be a negative experience and thus did not carry-on using it.
Aavegotchi most recently voted down a proposal to enact shielded voting as well.
I presume you are only considering this being used on offchain signalling via Snapshot, as shielded voting at scale doesnāt exist onchain? Although itās something that could be built.
Interesting conversation! There are different trade-offs and attack vectors that come with public versus private versus shielded voting. All of these considerations need to be evaluated in the specific context of the governance questions to which they are applied. Shielded voting may be appropriate for a specific type of governance question or scenario but not another (and the same applies to the other types of voting.)
Creating a more robust framework to reason about the full set of tradeoffs for public versus private versus shielded voting is on our research roadmap for 2025. All of these options should be considered in relation to how they provide a solution to specific problem statement(s) with a holistic understanding of the impact implementation would have on the governance system as a whole.
