I have been working on a reputation system for DAO delegates (eventually other types of contributors) and am requesting feedback from the community to make the algorithm better to reflect the true contribution of delegates.
You can see the current ENS DAO delegates leaderboard here
Karma Score
The delegates are listed in order of their Karma score. Karma score is a quantified score representing the contribution of a DAO member. Given a Signal S and weightage W, karma score is the sum of the product of each signal and its weightage.
Signal (S) can be on-chain or off-chain voting percentage, total votes delegated, forum activity, badges accrued and so on.
Weightage (W) is the multiplier you apply to each signal to indicate the importance of that signal and normalize it.
Karma Score = Sum(S1*W1, …, Sn*Wn)
ENS Karma Score Calculation
- Delegate does not have an ENS domain: (-20 points)
- Has link to delegate pitch: 10 points
- Total proposals initiated in the forum: (10 * total) points
- Total proposals discussed in the forum: (2 * total) points
- Total posts created: (1*total) points
- Total likes received on posts in the forum: (0.5*total) points
- Total posts read: (0.25*total) points
- Snapshot voting percentage: (1*pct)
- POAPs collected: (0.1*total) points
- More than median votes delegated: 20 points
- Experience for being in other DAOs
- <3 DAOs: (count*15) points
- 3 - 6 DAOs: (count * 10) points
- 7 - 10 DAOs: (count * 5) points
- User Type
- Individual: 10 points
- Company: 0 points
- Non-profit: 50 points
Benefits of this system
- Tokens holders can easily decide who to delegate their tokens to
- Make changes in working groups based on activity
- Get proactive in retaining active contributors
- Discover new and upcoming contributors and recruit them into working groups
- DAO contributors can show their reputation and find opportunities in other DAOs
Request for feedback
I believe for this reputation score to be accepted by everyone, it should be community driven and so I am requesting feedback from the ENS community to improve this algorithm. Do all the signals and the weights being assigned make sense, should we not consider some of those, should more signals be added (ex: discord)