DIFFUSIONCORE◉ LIVE LAB
Bass Diffusion (S-Curve)
Innovators plus imitators make the S-curve. Spend early on seeds, later on social proof.
How it works
- Innovators buy on their own; imitators buy because others did.
- Early curve is flat, middle explodes, late saturates — the S.
- Two knobs (p, q) fit most launches eerily well.
Use → Launches, rollouts, feature adoption.
Watch out → Reading a slow start as failure instead of the flat foot of the S.
Try it → Drag p and q in the lab until the curve matches a launch you know.
REDCAPE jobs → Predict · Design
Free sources → NetLogo Diffusion models • Wikipedia: Bass diffusion model (CC BY-SA)
INTERACTIVE LABoriginal recreation — poke it
PEAK SALES AT
t = 11
HALF ADOPTED
t = 12
Set q to 0 and you get the r-shaped broadcast curve; raise it and the S appears.
Attach to a lattice — combines with
Different families, shared jobs. Disagreement between them is where insight lives.
LEARNING
Diversity Prediction Theorem
Crowd error equals average error minus diversity. Diverse models beat the lone genius.
SHARED JOBS: PREDICT · DESIGN
DISTRIBUTIONSPoisson Arrivals
Counts of rare independent events per window: calls, mutations, clicks.
SHARED JOBS: PREDICT · DESIGN
LINEAR & NONLINEARCategorical & Threshold Models
Buckets and cutoffs turn messy attributes into testable if-then rules.
SHARED JOB: PREDICT