Divvi uses a proportional prize contest model, meaning rewards are split based on how much a builder contributes relative to others.
If someone tries to spam low-quality activity (e.g. fake transactions), their share of the rewards shrinks as builders gaining authentic user activity continue contributing; each additional tx from authentic builders’ users decreases the ROI for spammers at no additional cost to said builders (since their users pay gas costs for each transaction).
So, you can't just brute-force your way to a big payout— you’re competing for a slice, not a fixed dollar prize.
Extractive tactics become a net-loss for spammers as real usage grows.
The whole incentive program is intentionally experimental and modular:
Like a live immune system: Divvi evolves as attackers probe it.
Divvi is hard to game because it’s designed around relative contribution, costly-to-fake metrics, with built-in flexibility to course-correct.
Mercenary players can’t just show up and extract; they’d have to actually contribute in meaningful, verifiable ways, which defeats their business model and turns them into contributors.