Voting based on media influence score – lessons learned
One of the projects that really interests me is Ethereum Signals Ring, where we’re looking for solutions to identify, filter and highlight important signals from the ethereum community. For me this work started from the Phil Lucsok’s initial presentation on the Magicians meetup in Berlin. The proposed comparison of the hashpower, full-node, stake-weighted, identity and gas voting methods started the conversation on finding legit and practical way for finding the appropriate signaling model.
While some of those methods are already used (like carbonvote) or are being tested (like gas voting by slock.it), my attention was caught by the cryptoinfluencers.io project. They are basically ranking the influences, based on twitter performance (currently), the amount of followers, importance of those follower and some other metrics.
The idea that came to me was around using this stats on top 100 or 500 influencers give those people the representation power for non-binding votes, connected to their influence score. As those people have some approved identity and are mostly known in the community it may have seamed to be good idea for them to indicate the whole community sentiment. Inspired by this idea i’ve put it to magicians forum for criticism and… well I got enough =) While some of that felt a bit harsh, I really appreciate the time and passion of the commentators. Here’re my outtakes from this activity, maybe used for further development.
1. Qualitative and Subjective metrics are controversial
There’s no general readiness to use a list generated by the 3rd party, especially if there’s somebodies manual moderation. As Phil have mentioned:
(those) reputation metrics have not been clearly defined, agreed upon, or legitimized
I wonder if some public TCR will make a better work here. Though even deciding on using a TCR (without the algorithm behind) may be a hard task. Quantitative statistics are to be more trusted by engineers.
2. The scope of governance is not fully clear
While the ring purpose is now mostly focused on sourcing the info to core dev on EIPs, there’re other aspects of governance in my opinion, besides the protocol itself. It can be the process of protocol management (e.g. who and how decides on merges), work of rings, priorities and policies.
I perceive the core devs as some kind of policy makers and representatives of the network. Some of them (like Vitalik, of course) are known for their statements on ICO mechanics, how industry works, what are the right values and so on. Maybe it’s only my opinion, and it shouldn’t be mixed with the core dev role. Though somebody will need to take that role anyway, and they would benefit from measuring the sentiment.
3. Loose governance is working fine for now (at least as it seems)
Core developers groups seems like a tight circle of people who mostly know ant trust each other. There are rare cases like the DAO or EIP-999, where community splits, most of the tech questions are solved quietly on the regular calls (pls correct me if I’m wrong). So getting a group of people with unclear interests even for non-binding voting may seem like a risk, with not much gain.
I doubt it will stop some projects to do some voting any way, though it may be better if a bunch of different groups with difference methods are available to be compared at once. Thus gaming the algorithm or bribing will become much more difficult.
4. Calcification threat
Not sure if it’s a general fear, but it seemed pretty interesting. The idea was in constantly refreshing the list of people we treat as influencers with new faces, so it won’t be the same giving the rising power.
Also can’t leave this twitter thread on reputation without the mention: https://twitter.com/sinahab/status/1027639769910525952 Funnily it happened at the same time as forum discussion.
Thanks again to all the magicians who provided their thoughts, it helps a lot!
In dgov foundation we’re forming the community around the distributed governance from practitioners, providers, academics and policymakers.
Support the Movement!
If you’re enthusiastic about the distributed governance — go give us a hand. Here’re some hot links on #dgov network:
