kept re-reading dusk’s tokenomics docs because the slashing section didn’t say what i expected.
most pos chains i’ve looked into burn a misbehaving validator’s stake — gone, network-wide loss, nobody gains.
dusk doesn’t do that.
per dusk’s own tokenomics docs: they use “soft slashing.” a validator with repeated faults or long downtime doesn’t get their stake destroyed. instead, a portion of it moves directly into the claimable rewards pool — the same pool that pays every other honest staker.
read that twice. the penalty for one validator’s bad behavior becomes a direct payout for everyone else still doing their job right.
that’s a genuinely different incentive shape than burning. burning makes misbehavior a pure loss for the whole network — nobody profits, the bad actor just loses. redistribution turns it into a transfer. the bad actor loses, and somebody else specifically gains.
which raises a question the docs don’t really address: does that quietly change how an honest validator feels about a neighbor cutting corners? burning gives you zero reason to want anyone else to fail. redistribution means somebody else’s downtime is, indirectly, your upside.
probably too small an effect to actually change real behavior — the amounts are likely modest relative to normal rewards. but i can’t rule it out, and it’s a strange thing to leave unaddressed in the docs for a chain built this specifically around institutional-grade design.
#dusk @Dusk $DUSK
most pos chains i’ve looked into burn a misbehaving validator’s stake — gone, network-wide loss, nobody gains.
dusk doesn’t do that.
per dusk’s own tokenomics docs: they use “soft slashing.” a validator with repeated faults or long downtime doesn’t get their stake destroyed. instead, a portion of it moves directly into the claimable rewards pool — the same pool that pays every other honest staker.
read that twice. the penalty for one validator’s bad behavior becomes a direct payout for everyone else still doing their job right.
that’s a genuinely different incentive shape than burning. burning makes misbehavior a pure loss for the whole network — nobody profits, the bad actor just loses. redistribution turns it into a transfer. the bad actor loses, and somebody else specifically gains.
which raises a question the docs don’t really address: does that quietly change how an honest validator feels about a neighbor cutting corners? burning gives you zero reason to want anyone else to fail. redistribution means somebody else’s downtime is, indirectly, your upside.
probably too small an effect to actually change real behavior — the amounts are likely modest relative to normal rewards. but i can’t rule it out, and it’s a strange thing to leave unaddressed in the docs for a chain built this specifically around institutional-grade design.
#dusk @Dusk $DUSK