Simulation Decomposition — Making Monte Carlo Simulations Explain Themselves

Monte Carlo simulation is powerful—but blind.
It can show how uncertain a model’s outcomes are, yet it doesn’t explain why those outcomes occur.

That’s where Simulation Decomposition (SimDec) comes in.

In “Monte Carlo Enhancement via Simulation Decomposition: A ‘Must-Have’ Inclusion for Many Disciplines” (Mariia Kozlova & Julian Scott Yeomans, INFORMS Transactions on Education, 22(3): 147–159), the authors present a remarkably simple but transformative idea:
instead of producing a single undifferentiated histogram of outcomes, color-code it by the states of key input variables.
That one move makes the simulation results interpretable—and useful for decision-making.

🔗 Read the full article: https://doi.org/10.1287/ited.2019.0240

How It Works

As illustrated in the schematic on page 4 of the paper, SimDec divides input variables into meaningful states (for example, low–medium–high or optimistic–pessimistic).
During the Monte Carlo run, it records which state combination produces each outcome and builds a stacked histogram, where each color segment represents one scenario.

The result?
You can immediately see how different combinations of inputs drive the overall distribution—without rerunning the simulation dozens of times.

Examples Across Disciplines

The authors showcase three compelling examples:

• Geology: estimating oil-field reserves under uncertainty; SimDec reveals that thickness drives most uncertainty and that 3-D seismic analysis best reduces it.
• Business: evaluating investment profitability; decomposition exposes which mix of cost and production levels can secure profitability even under volatile prices.
• Environmental Science: comparing the carbon footprint of wooden vs. plastic pallets; SimDec uncovers how end-of-life treatment (burning vs. landfilling) reverses which material performs better.

Each case uses the same logic—and in each, decomposition exposes relationships invisible in a standard Monte Carlo plot.

From Research to Education

Because SimDec is easy to implement in Excel or any modeling tool, Kozlova and Yeomans advocate its inclusion in simulation courses.
It trains students not only to model uncertainty but also to understand it—making simulations a real learning instrument rather than a black box.

Why It Matters

Simulation Decomposition transforms Monte Carlo from a passive observer into an interactive explainer.
It keeps all the rigor of stochastic modeling but adds clarity, making uncertainty analysis transparent, visual, and actionable—exactly what both researchers and practitioners need.

Reference:
Kozlova, M., & Yeomans, J. S. (2022). Monte Carlo Enhancement via Simulation Decomposition: A “Must-Have” Inclusion for Many Disciplines. INFORMS Transactions on Education, 22(3), 147–159. https://doi.org/10.1287/ited.2019.0240

‍