[Research] ENS-GR-01: A Ten-Proposal Governance Record for Delegation Incentives

[blockful]

Thanks for putting this together, and for running the Round 2 files through a different curve. The timing is good, because the three-month review we promised is due next month.

@alontweft

The point is not to claim that this is a finished anti-whale solution. It is to replace a hard per-address cutoff—which creates a visible incentive to split wallets—with a rule that rewards sustained governance behaviour and reduces marginal returns smoothly.

Agreed on the cap, and this is close to where we landed with nick.eth during the Temp Check (here). Caps are not a long-term sybil mechanism. We kept them as a guardrail for the pilot, and whether to remove or replace them is one of the questions for the three-month review. Your comparison of the seven capped holders with the cap, without it and under a continuous curve helps there, because it puts a number on the split incentive we had only described.

@alontweft

missed Delegate votes should reduce both the Delegate’s reward and the associated delegator reward

On grading delegate rewards by votes cast, our first model did exactly that and we dropped it before the RFC. With a small set of active delegates, a gap between 90% and 100% attendance pushes delegators toward the few delegates with perfect attendance, which works against the dispersion the program is meant to improve. The reasoning is in the RFC thread (here). The pilot also showed how sharp the binary rule can be. In August nick.eth, who held most of the active voting power at the time, voted on 6 of the last 10 proposals and stopped counting as active. The program’s growth metric for the month came out at −70.69% (details).

@alontweft

new participants should begin accruing from their first governance action, while fresh wallets should not immediately outperform mature participants.

The 180-day time-weighted balance does this job in the pilot. A wallet funded last week weighs close to nothing, whatever it delegates to. That rule replaced the original hold-days factor after nick.eth showed how many dust wallets older than 180 days were sitting around (here). Counting proposals instead of days aims at the same thing.

@alontweft

It does not clearly distinguish sustained personal voting, considered delegation, and delegation made only near the end of a round.

On the last one, the pilot deliberately does not distinguish them. An earlier draft used time delegated, and we moved to time held after feedback from AvsA and James (here). Whether that trade held up is a question for the round data, not for this thread.

Thanks again for the work. The three-month review, covering all three rounds, comes next month in its own thread, and that is the better place to test these ideas against the actual distribution data. We would like your read on it there.