joellewis/mental-model-library

Overview

A "Mental Model" is a representation of how something works—a simplified map of reality. This skill provides a "Latticework of Mental Models" from major disciplines (Physics, Biology, Psychology, Math) to prevent "Man with a Hammer" syndrome—the tendency to treat every problem as a nail because you only have one tool.

Iron Law

NO COMPLEX ANALYSIS WITHOUT CHECKING FOR APPLICABLE MENTAL MODELS

Relying on intuition alone ensures that you will fall victim to predictable cognitive biases. Systematic model application is the only defense against human misjudgment.

State Machine

digraph mental_model_flow {
    "Problem Identified" [shape=doublecircle];
    "Invert: Define Failure" [shape=box];
    "Filter: Identify Core Discipline" [shape=box];
    "Apply: Select 2-3 Models" [shape=box];
    "Gate: Bias Check" [shape=diamond];
    "Synthesize: Find Intersections" [shape=box];
    "Decision/Insight" [shape=doublecircle];

    "Problem Identified" -> "Invert: Define Failure";
    "Invert: Define Failure" -> "Filter: Identify Core Discipline";
    "Filter: Identify Core Discipline" -> "Apply: Select 2-3 Models";
    "Apply: Select 2-3 Models" -> "Gate: Bias Check";
    "Gate: Bias Check" -> "Synthesize: Find Intersections" [label="clean"];
    "Gate: Bias Check" -> "Filter: Identify Core Discipline" [label="bias found"];
    "Synthesize: Find Intersections" -> "Decision/Insight";
}

When to Use This Skill

• When a problem seems "stuck" or repetitive.
• Before making an irreversible decision (e.g., product pivot, major hiring, market entry).
• When analyzing why a competitor is succeeding or why a project failed.
• During a "Pre-Mortem" to identify hidden risks.

When NOT to Use This Skill

• For routine, low-stakes operational tasks where speed is more valuable than depth.
• When the goal is creative expression rather than objective analysis.

Core Process

Step 1: Invert, Always Invert

Instead of asking "How do I succeed?", ask "How could this fail spectacularly?" Define the conditions for failure and work backward to avoid them (Source: Munger).

Step 2: Traverse the Latticework

Apply models from at least three different disciplines to the problem:

• Physics: Is there Entropy (disorder) increasing? Is there a Leverage point?
• Biology: Are Incentives driving behavior? Is there a Red Queen Effect (running to stay in place)?
• Chemistry: Is there enough Activation Energy to start this change? (Source: Parrish).

Step 3: Identify the Lollapalooza

Look for points where multiple biases or models act in the same direction. When 3+ models align, the outcome will be non-linear and massive (Source: Munger).

Step 4: Run the Misjudgment Checklist

Audit your conclusion against the 25 Psychological Tendencies (e.g., Social Proof, Reciprocation, Availability Bias) to ensure you aren't fooling yourself (Source: Munger).

The Model Registry (High-Leverage Only)

| Discipline | Model | Principle (Source: Parrish/Munger/Bevelin) | |---|---|---| | Thinking | Inversion | Turn the problem upside down to see what to avoid. | | Thinking | First Principles | Strip a system down to its fundamental truths and build up. | | Physics | Entropy | Disorder increases naturally; order requires energy. | | Physics | Leverage | Small inputs at the right point produce massive outputs. | | Biology | Evolution | Individuals and organizations that don't adapt go extinct. | | Biology | Incentives | "Show me the incentive, and I'll show you the outcome." | | Systems | Feedback Loops | Outputs that circle back to inputs (vicious or virtuous). | | Psychology | Social Proof | We look to others to determine correct behavior. | | Psychology | Loss Aversion | The pain of losing is 2x stronger than the joy of winning. |

Cross-Skill Invocations

• REQUIRED SUB-SKILL: getting-started — To route the problem to the correct family before applying models.
• RECOMMENDED SUB-SKILL: decision-frameworks — For formalizing the output of the mental model analysis.
• RECOMMENDED SUB-SKILL: devils-advocate — To use Inversion as a tool for stress-testing.

Rationalization Table

| Thought | Reality | |---------|---------| | "I don't have time to go through 100 models." | You only need the 5-10 "big ideas" to catch 90% of errors. | | "This is a business problem, I don't need Physics." | Economic systems are subject to the same laws of scale and limits. | | "I already know why this is happening." | Your "knowledge" is often a story masking a cognitive bias. | | "Mental models are just fancy names for common sense." | Common sense is not common; models make it systematic and repeatable. |

Red Flags

• "Man with a Hammer" → You are forcing the problem to fit your favorite model (e.g., only seeing incentives).
• "Confirmation Bias" → You are only looking for models that support your current plan.
• "Ignoring Scale" → You assume a model that works at small scale (Start-up) works at large scale (Enterprise).

Diagnostic Checklist

• [ ] Have I Inverted the problem to identify "how to fail"?
• [ ] Have I applied models from at least three different disciplines?
• [ ] Have I checked for Lollapalooza effects (confluence of forces)?
• [ ] Have I audited for the Social Proof or Incentive biases?
• [ ] Am I working within my Circle of Competence?

Sources

• Munger, Charles. Poor Charlie's Almanack, Talk Eleven (Psychology of Human Misjudgment).
• Parrish, Shane. The Great Mental Models, Vol 1 (General Thinking), Vol 2 (Science).
• Bevelin, Peter. Seeking Wisdom, Ch. 2 (Psychology), Ch. 3 (Physics/Math).