wslyvh / tokenlog

This repo is no longer maintained. Check out the fork at General Magic
https://github.com/GeneralMagicio/tokenlog
MIT License
53 stars 22 forks source link

quadratic voting accuracy #30

Open liviade1 opened 3 years ago

liviade1 commented 3 years ago

We all have so many votes and it seems hard to adopt just one voting strategy. ( Maybe some people use all their votes, maybe others use just a little bit of it) So results could be easily manipulated by one person with high voting power, specially towards the end of the voting periods. Griff suggested adding percentages instead of single votes. Like using 30% of your voting power in one proposal and 70 in another for ex. Is this feasible to implement?

wslyvh commented 3 years ago

I had a more 'stepped' voting system previously, which basically allowed for people to vote in steps of 10 (basically 10-100%). Other projects suggested though that they wanted users to have full control and allow them to decide themselves. Which makes sense, imo.

I think high token holders, or whales always have the possibility to 'manipulate' or greatly influence the results. Quadratic voting prevents some of that already, by making it more expensive. I'm not sure if voting in steps, or any technical changes for that manner, would solve any of that. Open to more ideas though!

For Commons, there's a lot of coordination and support through different channels on the voting rounds. It might be more useful to set clear guidelines and strategies for each round.

E.g. set expected timelines (start/end dates) for each voting round Encourage everyone to vote with most/all of their voting power. Or maybe a limit? Since you're voting in rounds, there's basically no downside to using all. Tokens aren't locked up or anything. And votes are 'released' by the end of the round and available again for the next.

Would that make sense?