Creating Twins and Scenarios

Creating Twins and Scenarios

A twin is a slice of your model brought to life over time. A scenario is a specific projection within a twin — its timeline, its assumptions, its data. You'll usually create the twin once and add several scenarios to compare alternatives.

Create a twin

1. Open the Tree View and find the Simulation section
2. Use the + button (or right-click context menu) to create a new Twin
3. Give it a meaningful name (e.g., "HR Operations", "Quarterly KPIs", "Widget-A Supply Chain")

Set the scope

In the Properties Panel (right side), pick which node types should become agents in this twin. Anything not in scope stays in the regular model and is ignored by the simulation.

• Narrow scope — pick only Employee and Team if you're modeling HR operations
• Wider scope — pick more types if your simulation needs them

Optionally, anchor the twin to a single M1 node — for example, a specific company or supply chain. The twin then represents the live state of that one element.

Create a scenario

1. With the twin selected, find its Scenarios branch in the Tree View
2. Use the + button to add a new Scenario
3. Name it (e.g., "Baseline 2026", "Aggressive Hiring", "Shanghai Disruption")
4. In the Properties Panel, set the timeline:
   • Start time — when the simulation begins (e.g., 0 or 2026)
   • End time — when it ends (e.g., 24 or 2030)
   • dt — the step size (e.g., 1, 0.5, 1/12)

Pick a unit (months, weeks, days, hours) and stay consistent across all three values. The scenario timeline determines how many cells appear in the time series grid for every agent.

Many scenarios per twin

This is where what-if comparison shines. You typically build the twin once (the structure stays the same) and add several scenarios where the assumptions vary:

Twin: Sales Pipeline
Scenario: Baseline	conservative growth
Scenario: Optimistic	new product line ships in Q3
Scenario: Pessimistic	top customer churns

Each scenario keeps its own timeline, agents, and links — they don't interfere with each other. Start with one scenario; add more as the questions emerge.

Next steps

• Adding Agents to a Scenario — populate the scenario with live elements
• Entering Time Series Data — record values over time