TLDR
Let’s build a Prop House-like site and run a quarterly retroactive rewards round where a defined number of ETH rewards are given out to a set of the most impactful on-chain props ranked by nouner voting.
Suggested implementation:
-
Passed on-chain props each have their own page where they can post updates of how their prop went (delivered outcomes) leading up to the voting. Addresses approved by the proposer of the original prop have access to post these updates.
-
On the first week of every quarter, we take on-chain props that delivered their props in a defined lookback period (let’s say 1 year) and include them as options for nouner voting.
-
Like Prop House, each noun has 10 votes to distribute across props as their nouner or delegate sees fit (ideally weighing the success of individual props in consideration). Like on-chain props, voters have the option to supply reasons for voting X number of votes on Y prop.
-
At the end of the voting week, ETH prizes are sent to the top winning props. ETH prizes could be defined in various ways, but a potential set could look something like 200 ETH to the most voted prop, 100 ETH to the next two props, 50 ETH to the next four props, and 25 ETH to the next eight props, for a total of 800 ETH given out to 15 props for that given quarter.
Past efforts:
At various points in time the DAO and the community has felt that there should be a mechanism that gives retroactive rewards to the prop builders that delivered outsized outcomes for the Nouns ecosystem (see this forum post).
The most direct effort we passed to move towards a mechanism that supports this need was Prop 63 (Pilot Collectivized Nounish Allocation), but it had a slightly different aim. The aim there was to experiment with a smaller budget (100 ETH) to see if empowering nounish builders to allocate some portion of the DAO’s resources could lead to better outcomes. This prop seems to have stalled due to wag going MIA (we miss you wag!).
Of course, there’s also the small grants committee that delivers retroactive rewards on a more real-time discretionary basis but those are small amounts (usually less than 1 ETH).
Thoughts on the nitty-gritty:
-
Why limit the universe of candidates to on-chain props?
- I’m a firm believer that trying to build a mechanism that addresses the entire spectrum of retroactive rewards (from hundreds of ETH for the best on-chain prop builders to sub 1 ETH for being active on discord) is likely a futile task. It’s better to build multiple retroactive reward mechanisms for different segments of the reward amount spectrum in a similar way to how we approach funding with different mechanisms depending on the funding amount (on-chain prop, prop house, and small grants).
- This vastly reduces the complexity of this prop. The exercise of defining the universe of candidates (“who nominates”, “should the nominees be projects or individuals”, “can individuals be included multiple times”, “did the nominee agree to be nominated”, etc) is in itself a whole project. This is the rock that Prop 63 died on.
-
Why quarterly?
- Weakly held view, but gut feeling that monthly would be too often (can lead to performance evaluation fatigue) and something beyond quarterly would lead to too much recency bias (ppl tend to forget events rapidly, including the impact of a delivered prop) in favor of the props that delivered outcomes in the most recent months.
-
Why have a lookback period and not just look at the props that were delivered in the given quarter?
- Sometimes it takes a bit of time for props to show their full impact. I feel like we didn’t quite know how valuable FOMO Nouns was until more aesthetic nouns started getting higher bids consistently, we had a need to aim for a specific head (beer noun for Bud Light during the super bowl), and we understood the cultural/community value of Noun O’clock. So it would have been a shame if FOMO Nouns only had one chance to win the retro prize and missed it in its first chance and never had a second chance.
- The reason to not have an infinite lookback period is to not have the number of candidate props become ever growing into an irrelevantly large size.
-
Why not expand the voting to community (non-nouner) voting as well?
- We could select a group of community members to have additional votes, but we effectively have a decent amount of this happening through delegation. I don’t know the latest number but the Nouncil has a decent number of nouns delegated so they will represent the nounish builder perspective. Also a fair amount of the most engaged community members have nouns individually delegated to them as well. I prefer this decentralized approach, rather than the Retro House running some separate centralized process that somewhat arbitrarily gives out more votes.
- There is some potential value in the voter group of the Retro House being the same as the on-chain prop voting. Prop builders can see what props were deemed most successful by the results of past Retro Houses and use that to gauge what props will be more likely to be funded in the future. This isn’t the main reason to do the Retro House imo but is an added benefit that we would lose if voter groups were widely different.
-
Why allow updates on delivered outcomes for each prop?
- Better for voters to be more informed. Engaged nouners will have existing intuition of outcomes of different props, but good to have it laid out so that on the week of voting voters can go through them and have a bird’s eye view of all props.
- Prop builders can have insights as to the effectiveness of their props that are not visible to the public.
- Proposers (Nouner sponsors) can also have additional (non-public) perspectives that could be additive as well (for example, the Tech Pod could have a view on the performance of the Verbs team in their exploration of dynamic quorum voting).
-
Why allow voting reasons to be optionally expressed for each prop?
- Mostly just intuition from on-chain prop voting allowing reasons. Different nouners that are engaged in different focus areas often write a reason that informs my view on how I should vote. I expect similar positive dynamics to play out in the Retro House as well.
-
Why do fixed ETH prizes rather than dynamically sized prizes based on quadratic voting or some other reward function?
- The above bullet point in general is a weakly held view, but I tend to think there’s much value in simplicity. The fixed ETH prizes makes this retroactive rewards mechanism something you can explain to any non-crypto native prop builder in a sentence or two (“these are the reward prize amounts and winners are ranked by nouner voting”).
- I also tend to think there’s less risk in the reward distribution outcome being something the nouners, community, or the prop builders didn’t expect. Essentially, we have the distribution curve fixed and agreed on. We’re just deciding which props end up filling each spot.