Reading the Trust Graph

The trust graph turns your completed help exchanges into a picture. Each person you've built trust with is a node; each relationship is a link. Stronger, more recent relationships draw heavier lines. It's the platform's memory of who has actually shown up for whom.

One number: the decayed trust score

A node's trust score is the sum of the current (decayed) strength of all its trust relationships. Trust that isn't maintained fades, so the score reflects where a relationship stands today — not its all-time peak. The same person therefore shows the same score in every view.

One visual language

Across all views:

• Every node is the same size. Node size doesn't encode a variable, so it can't mislead. What you read is structure — who clusters together, who bridges groups.
• You are enlarged and white-ringed — a "you are here" anchor.
• Your connections are amber. Lines touching you stand out from everyone else's.
• Clusters share a color. Within-cluster ties are indigo; cross-cluster ties are slate. A cluster is a tightly-knit pocket of people, detected from the strongest ties.

The views

• Community — every member of a community, grouped into clusters by how closely they connect.
• My Network — your first-degree network within one community, clustered.
• Your Network (dashboard → People) — your trust network aggregated across all your communities.
• Communities (dashboard → Communities) — the inter-community depth view: each community is a node sized by membership. Organic ties (solid, slate) accrue as members exchange help across communities; fission lineage (dashed, violet) traces parent→child links left behind when a community splits.

How relationships form and fade

Completing a help exchange strengthens a tie; endorsements, karma, and shared events also contribute. Without continued interaction, a tie's weight decays over time — which is why the trust score is the decayed current weight. Active trust is what counts.