After modeling this a bit more, here’s a few more questions/ideas regarding the distribution:
Time-weighted snapshot
Should we take historical yveCRV balances into account?
-
t_0→t_n - time-weighted snapshot since yvecrv creation
-
t_{n-1}→t_n - time-weighted one-week fixed time snapshot between snapshots
-
t_n - simple snapshot matching the time of the airdrop
Distribution schedule to maximize farming
I see two potential farming models with the second one being 2.25x more efficient in terms of active farming weeks.
-
After paying the initial 50% penalty and staking EPS, the principal can be unlocked on a rolling 3 month basis. EPS is delivered to holders with a 3 months lag. This way we always farm with 3mo worth of the airdrop, which translates to an average of 9.75 weeks of EPS airdrop in farming.
-
Increase the farming amount till the end of airdrop in 12 months, then airdrop a big chunk (9mo worth) and distribute the remaining portion as it unlocks over the remaining 3mo. This way we farm with 21.9375 weeks of airdrop on average.