Master the 5‑Step Method to Reveal Hidden Models Behind Everyday Phenomena

What does “seeing through a model” mean?

A model is not just a mathematical formula; it is a simplified representation of complex phenomena, ranging from functional relationships to behavioral rules, evolution mechanisms, or feedback loops. Its purpose is to extract core mechanisms, giving you a key to understand causality and structure.

How to quickly identify the model behind a phenomenon?

Based on years of strategy analysis and mathematical modeling, the author proposes five practical steps.

Step 1: Quantify the phenomenon into variables

Any phenomenon can be broken into a set of describable variables. For example, the difficulty of career promotion can be expressed by Ability (A), Resources (R), Emotional intelligence (E) and Luck (U).

Step 2: Determine the relationship types between variables

Linear relationship : A increases by 1, B increases by k.

Non‑linear relationship : A changes cause B to follow exponential, logarithmic, S‑shaped, or inverted‑U patterns.

Feedback mechanism : A → B → A (positive or negative feedback).

Delay effect : A changes, B responds after a lag.

Competitive relationship : Increase in A reduces B (resource competition).

Step 3: Identify the structural form of the model

The core is Input‑Process‑Output:

Input variables (external inputs or internal resources)

Processing mechanisms (functions, rules, feedback)

Output variables (behaviors, phenomena, decisions)

Example – “excess shared bikes” in a city:

Input : platform expansion decisions, user registration growth, bike deployment volume.

Process : operation rules, usage rate, maintenance cost, market saturation feedback.

Output : street clutter, rising idle rate, user complaints, increased municipal management pressure.

This is a typical system dynamics model where delayed feedback creates a “bullwhip effect”, amplifying local demand changes and eventually collapsing the system.

Step 4: Compare known models with the current phenomenon

Many real‑world situations share structural patterns. By maintaining a “model library”, you can quickly match observations to existing models.

Phenomenon: user churn & retention → Possible models: Funnel model, lifecycle model

Phenomenon: public opinion swings → Possible models: Information diffusion model, SIR model

Phenomenon: social‑media photo sharing incentive → Possible models: Reinforcement learning model, attention‑mechanism model

Phenomenon: educational resource stratification → Possible models: Multi‑stage game model, Matthew‑effect structure

Phenomenon: new product pricing strategy → Possible models: Nash equilibrium, price discrimination model

Phenomenon: workplace “involution” → Possible models: Prisoner’s dilemma game model

Step 5: Validate the model’s explanatory and predictive power

An effective model should:

Explain the majority of observed behaviors.

Account for key outlier cases.

Provide reasonable predictions or decision recommendations.

Offer sensitivity insight into core variables.

Maintain an appropriate fit with real‑world feedback.

For instance, analyzing “students’ smartphone addiction harming learning” with only “phone time” as a variable yields poor predictions; adding course attractiveness, task challenge, and instant feedback creates a richer incentive model.

This iterative process—hypothesize, observe, revise—produces a structure that explains the past and guides the future.

Common everyday model perspectives

Social behavior – Small‑world network + emotional contagion mechanism

Consumer decisions – Prospect theory + temporal discounting

Work performance – Game theory + signaling game

Policy response – System dynamics + framing effect

Family relationships – Emotional accounting model + feedback loop

These perspectives let you think in terms of “structure‑relationship‑mechanism” rather than simple good‑bad judgments.

How to cultivate intuition for “model thinking”?

Daily modeling practice: when you notice a phenomenon, sketch a “variable‑relationship‑structure” diagram.

Build a model library: systematically study classic economics, behavioral science, and systems science models.

Read model‑centric works such as “Rationality in Human Activity”, “Thinking, Fast and Slow”, “Modeling as Mathematical Thinking”, “Structured Analysis”.

Structured dialogue: train by constructing problems, discussing derivations with other modelers.

Record and reflect: keep a “model notebook” documenting “phenomenon → model” cases.

Experts often see the underlying mechanisms and variable relationships before anyone else, giving them structural foresight in a complex world.