Loops HouseVOID

NEAR HACKATHON
Team
1 member- AOOwner
Aoyama
Overview
VOID is a re-runnable investment thesis for NEAR built on builder retention, not flows. Everyone counts what launches on NEAR; VOID counts what survives.
Chapter 1 reads every DAO from sputnik-dao.near and every proposal timestamp via plain RPC (no indexer, no key). Chapter 2 uses Google's public NEAR dataset on BigQuery to track every contract launched since 2025. Both use Kaplan-Meier survival curves per birth cohort, censored 90 days before the data ends so recent deaths are not missed.
The data is cleaned before it is judged: 223 DAOs touched only by sponsor bots are excluded, contracts that predate the window or were merely upgraded are dropped, and memecoin launchpad tokens are reported apart from apps.
Four claims were written with kill conditions before the data was pulled. Bottom line across both chapters: a growth bet, not yet a durability bet. DAOs are fragile (24% survive year one) with an early sign of improvement that rests on a small cohort. Apps launched since 2025 are durable: 65% survive year one vs 34% for launchpad tokens. Against Snapshot DAOs with 5+ proposals, scored with the same code, 36% of NEAR DAOs survive year one vs 27%.
A sensitivity table shows which verdicts flip under 60- or 180-day death rules. One command re-runs everything; a GitHub Action rebuilds daily. Live dashboard: https://aoyamari.github.io/void/