What is this? Most Epstein network projects show what's known — who's connected, what documents mention whom.
This one measures what's missing. It uses CUE's type system as an investigative tool:
schema violations become leads. Dangling references = unknown persons. Missing evidence = unverified claims.
Type inconsistencies = logical gaps. All analysis is pre-computed at export time — the browser renders static JSON.
Current state: 132 entities, 398 connections. Evidence coverage is intentionally low (11.4%) — most entities
were added from structural relationships before evidence was linked. The gap is the work queue, not a surprise.
The structural analysis is more revealing: most cross-cluster relationships depend on a single bridge entity.
Discredit that person and the connection between those worlds disappears from the public record.
Reading the graph: Node size = degree. Opacity = evidence (opaque = documented, ghost = unverified).
Red/orange rings = gap severity. Solid lines = bidirectional. Dashed = one-way claims.
Toggle buttons switch between manual Clusters (19 curated),
algorithmic Communities (12 from NetworkX — stroke shows cluster for comparison),
and Bottleneck heatmap (composite risk score).