Academic Opinion Clustering — 3D Simulator

Bounded-Confidence Debate (3D)

Confidence threshold ε 0.15

Convergence rate μ 0.30

Academics 120

Opinion clusters

–

Polarization (var)

–

Interactions / s

–

Each sphere is one academic, scattered around a shared skeptical → convinced axis. Drag to orbit. Spheres drift along the axis and bunch into distinct camps as bounded-confidence interactions play out.

How it works

Each point's position along the axis is its opinion o ∈ [0, 1]; its spread off the axis is fixed per-academic noise, purely so overlapping opinions stay visible as a cloud instead of collapsing into one dot. Every tick two academics are picked at random and compared under the Deffuant–Weisbuch bounded-confidence rule (Deffuant, Neau, Amblard & Weisbuch, 2000): they move toward each other only if they are already close enough.

if |oᵢ − oⱼ| < ε:
    oᵢ ← oᵢ + μ·(oⱼ − oᵢ)
    oⱼ ← oⱼ + μ·(oᵢ − oⱼ)
else:
    no change  (too far apart to interact at all)

Confidence threshold ε — small values keep spheres locked into several separate clumps along the axis (real, stable polarization); above ≈0.5 almost every pair interacts and the whole cloud drifts into a single clump at consensus.
Convergence rate μ — controls how fast clumps tighten, not how many clumps form.
Why clumps instead of one spreading front — unlike a contagion model, an academic at one extreme and one at the other never influence each other no matter how many times they meet, so multiple camps can persist side by side along the axis indefinitely.
Sphere color tracks each academic's live opinion (cool blue = skeptical, warm orange = convinced), so a forming cluster shows up as a tight, uniformly-colored knot of spheres.

Skeptical end Convinced end Opinion axis

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.