A non-ribbon chord diagram is a circular relationship graph: nodes are arranged as arc segments on the perimeter of a circle, and connections are drawn as curved lines (chords) spanning the interior. The perceptual mechanism is gestalt continuity — the eye follows each curve from one arc to another, reading the connection as a single unit even when it crosses other chords. The circular layout implies no hierarchy (all nodes are equidistant from the center) and no direction (chords are symmetric). Connection strength is encoded by stroke weight and opacity, both redundant encodings of the same variable — accessible in both color-normal and color-deficient viewing.
A full ribbon chord diagram fills the space between two arcs with a shaped "ribbon," encoding directional flow volume: the width of the ribbon at each end represents how much flows from that node. This is appropriate when data has meaningful direction (migration, trade, financial flows) and when per-direction volumes are precisely known. A non-ribbon chord diagram strips the filled shape away, keeping only the connection lines. This is appropriate when relationships are undirected, when the connection pattern matters more than volumes, or when the filled ribbons would create false visual depth that obscures the underlying topology — which is the case in this coordination network.
A force-directed network graph would show the same relationships but allows nodes to move — the viewer cannot directly compare node positions across views. The circular layout here gives every node a fixed, readable position with a labeled arc. A matrix view (heatmap of all pairs) would be more precise for exact value comparison but requires the viewer to trace rows and columns mentally — the chord diagram makes the relationship pattern immediately visual without any matrix traversal. For dense networks (all 8 nodes connected to all 7 others), the chord layout handles the 28 simultaneous relationships without the overlap chaos of a force-directed layout.
A hub node in this chart is one whose arc is the starting point for many thick, dark chords. OCHA and UNHCR are visually dense because their arcs originate the strongest chords — thick walnut and obsidian lines radiate across the circle. Government Donors and Local NGOs show thinner chords, indicating weaker or fewer coordination relationships. This hub pattern is impossible to see in a table and requires significant computation to extract from a matrix view. The chord diagram makes it perceptually immediate: dense = hub, sparse = peripheral. Hover any node to isolate its connections and read the strength values in the tooltip.
The FT Visual Vocabulary classifies this under Relationship: "Show connections and flows between entities." The chord diagram is specifically suited to many-to-many relationships within a bounded, enumerable set of nodes — it fails for open-ended networks where the node count is large or variable. Abela's framework routes here when the question is "who is connected to whom, and how strongly?" Tufte's data-ink principle is honored: every chord pixel encodes a relationship. The node arc segments are the only structural elements — they carry identity (label and color) and provide a consistent angular reference that the force-directed layout cannot offer.
Every chord uses a cubic Bézier with both control points at the
origin (0,0) — M x1,y1 C0,0 0,0 x2,y2.
This means all chords converge through or near the center of the circle,
regardless of which two nodes they connect. The geometric consequence:
chords between adjacent nodes (short arcs apart) curve steeply inward;
chords between near-opposite nodes are nearly straight through the center.
The visual consequence: the center of the diagram becomes a
focal density region — the viewer's eye is drawn there
precisely where the most cross-connections occur.
This is not an arbitrary choice. The alternative — Bézier control points
at fixed offsets — would produce chords that intersect at random interior
points, creating visual noise without the unifying center gravity.