#dusk $DUSK @Dusk I was checking Dusk documentation again when one detail caught me: a larger validator set doesn’t automatically mean broader committee representation.
Dusk uses deterministic sortition to select provisioners, with stake influencing participation. So the intuitive assumption is that 50 validators should mean roughly 50 voices.
But I’m not convinced that’s enough.
I’d want to track the top 1%, 5%, and 10% of stake across 1,000 rounds, comparing their stake share with committee appearances and voting credits. Then repeat the analysis at 100, 500, and 1,000 rounds to measure unique validators, repeat selection, and concentration.
The real question isn’t how many validators exist; it’s how much voting influence repeatedly reaches the committee.
A validator set can look diverse while effective influence remains concentrated.
Maybe Dusk shows strong rotation. Maybe stake concentration creates a different picture.
That’s exactly why I’d rather measure committee representation than assume it from validator count alone.
Dusk uses deterministic sortition to select provisioners, with stake influencing participation. So the intuitive assumption is that 50 validators should mean roughly 50 voices.
But I’m not convinced that’s enough.
I’d want to track the top 1%, 5%, and 10% of stake across 1,000 rounds, comparing their stake share with committee appearances and voting credits. Then repeat the analysis at 100, 500, and 1,000 rounds to measure unique validators, repeat selection, and concentration.
The real question isn’t how many validators exist; it’s how much voting influence repeatedly reaches the committee.
A validator set can look diverse while effective influence remains concentrated.
Maybe Dusk shows strong rotation. Maybe stake concentration creates a different picture.
That’s exactly why I’d rather measure committee representation than assume it from validator count alone.