Narrative Shapes

# Narrative Shapes

Shapes classify the **macro-structure** of delivery curves — how intensity rises and falls across the full story.

| Shape | Description | Curve Pattern |
|-------|-------------|---------------|
| **Climactic** | Build, climax, release — one dominant peak defines the arc | Steady rise → sharp peak (mid/late) → decline |
| **Episodic** | Multiple peaks of similar weight — no single climax dominates | Repeating rises and falls, no clear maximum |
| **Rebounding** | A meaningful dip followed by strong recovery | Start high → collapse → strong recovery |
| **Peaking** | Dominant peak early or mid-arc, followed by decline | Early high → sustained fall |
| **Escalating** | Momentum rises overall — intensity concentrated toward the end | Gradual, sustained rise to finish |
| **Flat** | Too little structural variation — no meaningful peaks or valleys | Near-constant delivery values |

**Detection Metrics:**
- **Overall Slope** — Macro trend (rising, falling, stable)
- **Peak Count** — Number of detected local maxima
- **Peak Dominance** — Largest prominence / total prominence
- **Peak Position** — Where the dominant peak falls (0..1)
- **Trough Depth** — Magnitude of central valley (V-shape detector)
- **Flatness** — Standard deviation of smoothed curve

---

Raw Markdown source