APA-Style Statistical Reporting
No results yet. Click Run Simulation (or Step), then pause, to generate an APA 7th edition results paragraph for the current run.
Market Emergence: Agent-Based Simulator
Simulation & Causal Dynamics
Simulate how micro-level customer motives, social proof, and word-of-mouth cascades produce unexpected macro-level market emergence, segregation, and churn.
DATA SOURCE
📚
Use a Case Study
Select a preset scenario above to view the business context, or upload your own agent lattice CSV.
📤
Upload Your Data
Drag & Drop custom agent lattice CSV (.csv, .tsv, .txt)
Required columns: grid_x, grid_y (0-based or 1-based whole numbers), and initial_segment (or brand_choice).
Optional: agent_id, reference_price (discount case). One row per cell; the lattice must be 10×10 to 100×100.
🧭 New to Agent-Based Modeling (ABM)? How 2,500 individual rules create market-wide patterns in 3 steps Click to collapse / expand
1. Local Circle Only
Each cell on the grid is one household or customer. It cannot see the whole town—only its immediate neighbors (r = 1 means the 8 surrounding cells in a 3×3 square).
2. One Simple Micro-Rule (τ)
Each year, every agent counts its neighbors and checks its threshold τ (e.g., "Are at least 33% of my neighbors in my group?" or "Is my block ≤75% homogeneous so I have diversity?").
3. The Emergence Chain Reaction
When one unhappy agent moves or switches, it changes the neighbor ratio for 8 old neighbors and 8 new neighbors next year—sparking a cascade no survey could predict. Click any cell on the map to inspect its 3×3 math!
Model Parameters (Micro-Decision Rules)
At Year 0 changes apply immediately; mid-run they wait for "Apply & restart".
33%
Minimum tolerance for dissenting neighbors before acting.
1 (8 Neighbors)
Social circle width: 1=immediate Moore (8), 2=local suburb (24), 3=broad (48).
15%
Brand loyalty, switching costs, or relocation friction resisting change.
15%
Percentage of uninhabited houses available for relocation.
🔬 Advanced ABM Dynamics (Heterogeneity, Topology & Competitive Game)
Compare passive one-sided tolerance (≥ τ) against active "Goldilocks" diversity preferences (τmin ≤ same ≤ τmax).
Adds non-local social shortcuts; click agents to inspect graph connections.
Aggressive rival launches conquest campaigns when defection/churn crosses the critical threshold.
🗺️ Spatial Market Lattice
Agent Spotlight
Coordinate:
(25, 25)
Current State:
Active Customer
Local reading:
—
Rule met:
—
Relocations:
0
Status:
—
Rogers Cohort:
Early Majority
Individual τ:
33%
Topology:
Moore (8 local)
3×3 Local Neighborhood & Rule Math
Click any cell on the lattice to inspect its 8 neighbors and micro-decision rule.
📈 Market Share / State Distribution
📊 Spatial Clustering & Realized Outcome
📋 State-by-State Breakdown (Year 0 vs. Current Year)
| State / Group | Year 0 | Now | Δ Shift | Avg Local Reading | Rule Met |
|---|
Spatial Clustering
0.00
Moran's I: Random mixture
Primary Metric
--
Baseline state
Transitions This Year
0
Agents changed state last year
Emergence Gap
--
Macro pattern vs. micro rule
Cascade Momentum
—
Transitions this year ÷ last year
Scenario Metric
--
Managerial Interpretation & Causal Dynamics
Select a scenario and click Run Simulation to see how simple individual decision rules add up to a market-level pattern. The interpretation appears whenever the run is paused or settled.
Key Marketing Takeaway: A survey tells you what each customer would do alone. A simulation shows what happens when customers react to each other, and replications show how much of any single run was luck.
Replications: Was This Run Typical?
One run is one draw of luck. Replications rerun the same settings with many different seeds. Each replication runs a baseline world and a shocked twin from the same starting grid, so the difference between them isolates the shock's effect.
No replications yet. Choose settings and click Run Replications.