mcp-server

HUMMBL Problem Patterns Guide

Pre-defined patterns that map common problem types to recommended mental models

Overview

Problem Patterns are curated mappings between frequently encountered challenges and the most effective mental models to address them. Each pattern identifies:

Pattern Name: The type of challenge
Recommended Transformations: Which of the 6 core transformations apply
Top Models: The 3 most effective mental models for this pattern
Example Scenarios: Real-world situations where this pattern applies

These patterns provide fast onboarding for new users and proven starting points for experienced practitioners.

The 6 Problem Patterns

Pattern 1: Unclear Problem Definition

Description: The problem itself is ambiguous, poorly scoped, or symptoms are confused with root causes.

Recommended Transformation: P (Perspective) Frame and name what is. Anchor or shift point of view.

Top Models:

P1 - First Principles Framing: Reduce to foundational truths
P2 - Stakeholder Mapping: Identify all affected parties
P4 - Lens Shifting: Adopt different interpretive frameworks

Example Scenarios:

“Our product isn’t selling” (but unclear if it’s pricing, positioning, features, or market)
Stakeholders describe the same situation in completely different terms
Team debates solutions before agreeing on the actual problem
Symptoms (low engagement) confused with causes (poor onboarding)

How to Apply:

Start with P1 to strip away assumptions and find foundational truths
Use P2 to map who’s involved and their different perspectives
Apply P4 to view the problem through customer, competitor, and future lenses
Converge on a clear, shared problem statement before moving to solutions

Success Indicators:

✅ Problem statement is specific and testable
✅ All stakeholders agree on what’s actually wrong
✅ Root causes distinguished from symptoms
✅ Constraints and scope are explicit

Pattern 2: Stuck in Conventional Thinking

Description: Team, organization, or individual trapped in incremental thinking. Same ideas recycled. Innovation feels impossible.

Recommended Transformation: IN (Inversion) Reverse assumptions. Examine opposites, edges, negations.

Top Models:

IN1 - Subtractive Thinking: Remove elements instead of adding
IN2 - Premortem Analysis: Assume failure, work backward
IN3 - Problem Reversal: Solve the inverse problem

Example Scenarios:

Product roadmap is just “more features” with no breakthrough thinking
Strategy sessions produce predictable, safe ideas
Team says “we’ve always done it this way” or “that’s impossible here”
Competitors are disrupting but internal thinking remains incremental
Innovation workshops generate only minor tweaks to existing approaches

How to Apply:

Use IN1 to identify what to remove or stop doing (subtraction often > addition)
Run IN2 premortem: “It’s 12 months from now and we failed spectacularly—what happened?”
Apply IN3: If the problem is “increase revenue,” solve “how to minimize revenue” to reveal blind spots
Challenge every assumption: flip constraints, invert success criteria

Success Indicators:

✅ Team generates genuinely novel ideas
✅ Sacred cows are questioned
✅ “Impossible” ideas are explored seriously
✅ Simplification competes with addition

Pattern 3: Need to Assemble Solution

Description: Solution requires combining multiple elements, disciplines, or approaches that don’t naturally fit together.

Recommended Transformation: CO (Composition) Combine elements to create emergent properties.

Top Models:

CO1 - Building Blocks & Lego Principle: Combine reusable components
CO2 - Cross-Pollination: Borrow ideas from other domains
CO4 - Synthesis & Integration: Merge opposing or complementary elements

Example Scenarios:

Building a product that needs AI + hardware + service components
Creating a new business model from existing capabilities
Bridging multiple departments (engineering + sales + support) for a solution
Innovating by combining ideas from unrelated industries
Assembling a team with diverse, non-overlapping skills

How to Apply:

Start with CO1 to identify modular components that can be combined
Use CO2 to look at how other industries solve similar problems
Apply CO4 to synthesize seemingly incompatible elements (e.g., premium + accessible)
Iterate: combine, test for emergent properties, refine

Success Indicators:

✅ Solution is greater than sum of parts (emergent properties)
✅ Components remain modular and reusable
✅ Innovation comes from novel combinations, not just new invention
✅ Cross-domain insights are integrated

Pattern 4: Complex System to Understand

Description: System is too complex to grasp as a whole. Overwhelmed by interactions, dependencies, and moving parts.

Recommended Transformation: DE (Decomposition) Break down complexity into manageable components.

Top Models:

DE1 - Modular Decomposition: Break into independent components
DE2 - Layered Architecture: Organize into hierarchical strata
DE7 - Abstraction Hierarchies: Hide complexity behind interfaces

Example Scenarios:

Legacy codebase with no clear structure
Enterprise organization with unclear responsibilities and overlapping roles
Market with too many variables to model simultaneously
Product with feature bloat that nobody understands end-to-end
Process with so many steps that quality suffers

How to Apply:

Use DE1 to identify natural boundaries and create modules
Apply DE2 to organize into layers (infrastructure → services → interface)
Use DE7 to create abstractions that hide lower-level complexity
Document interfaces between components clearly
Test each component in isolation before integration

Success Indicators:

✅ Components can be understood independently
✅ Changes in one area don’t cascade unexpectedly
✅ New team members can contribute to specific areas quickly
✅ Complexity is managed, not eliminated (it’s still there, but organized)

Pattern 5: Feedback or Iteration Issues

Description: System doesn’t respond well to changes, learning cycles are broken, or iterations aren’t improving outcomes.

Recommended Transformation: RE (Recursion) Apply patterns at multiple scales and iterations.

Top Models:

RE1 - Iterative Refinement: Successive approximation toward target
RE2 - Recursive Application: Apply same pattern at multiple scales
RE3 - Self-Similarity & Fractals: Recognize repeating patterns

Example Scenarios:

Product launches with no feedback mechanism to improve next version
Organization makes decisions without learning from past outcomes
Process improvement efforts don’t compound over time
Team keeps making the same mistakes (no learning loop)
Scaling challenges: what worked at 10 people fails at 100

How to Apply:

Establish RE1 tight feedback loops: measure → learn → adjust → repeat
Use RE2 to apply successful patterns across scales (team → department → organization)
Identify RE3 self-similar patterns to predict behavior at different scales
Create deliberate learning cycles with retrospectives
Build compounding improvements (each iteration informed by previous)

Success Indicators:

✅ Feedback cycles are fast and actionable
✅ Learning compounds over time
✅ Same patterns successfully applied at different scales
✅ Team explicitly reflects and adapts

Pattern 6: Strategic or Coordination Challenge

Description: Multiple actors, incentives, or systems need to align. Game theory, ecosystem thinking, or meta-level analysis required.

Recommended Transformation: SY (Meta-Systems) Understand rules, patterns, and systems governing systems.

Top Models:

SY1 - Feedback Loops & Causality: Map reinforcing and balancing dynamics
SY2 - Incentive Design & Mechanism Design: Align incentives with goals
SY19 - Ecosystem Thinking: Model interdependent actors

Example Scenarios:

Market with conflicting incentives (short-term vs long-term)
Multi-sided platform struggling to balance different user types
Partnership where incentives aren’t aligned
Organization where departments optimize locally but harm global outcomes
Industry transformation requiring coordination across competitors

How to Apply:

Map SY1 feedback loops: where do reinforcing/balancing cycles exist?
Use SY2 to design incentives that align individual and system goals
Apply SY19 to model the entire ecosystem and interdependencies
Identify leverage points where small changes cascade
Design coordination mechanisms (standards, protocols, shared metrics)

Success Indicators:

✅ Incentives are aligned across actors
✅ System-level outcomes improve (not just local optimization)
✅ Coordination mechanisms work at scale
✅ Unintended consequences are anticipated and mitigated

Using Problem Patterns with the MCP Server

Quick Pattern Lookup

{
  "tool": "search_problem_patterns",
  "arguments": {
    "query": "innovation"
  }
}

Returns: Pattern 2 (Stuck in Conventional Thinking) with top models.

Get Recommended Models

{
  "tool": "recommend_models",
  "arguments": {
    "problem": "Our team keeps repeating the same mistakes with no improvement"
  }
}

Returns: Pattern 5 (Feedback or Iteration Issues) with RE transformation models.

Pattern Selection Guide

Not sure which pattern applies? Ask these questions:

Question	If YES → Pattern
Do we even agree on what the problem is?	Pattern 1: Unclear Problem Definition
Are we stuck with the same old ideas?	Pattern 2: Stuck in Conventional Thinking
Do we need to combine multiple things?	Pattern 3: Need to Assemble Solution
Is the system too complex to understand?	Pattern 4: Complex System to Understand
Are we not learning/improving?	Pattern 5: Feedback or Iteration Issues
Do we need to coordinate multiple actors?	Pattern 6: Strategic or Coordination Challenge

Pro tip: Most real problems touch multiple patterns. Start with the most acute one, then layer in others.

Pattern Combinations

Common pattern sequences for complex challenges:

Wicked Problems

Pattern 1 (Unclear) → Define the problem
Pattern 4 (Complex) → Break it down
Pattern 2 (Conventional) → Generate novel approaches
Pattern 3 (Assemble) → Combine into solution

Organizational Transformation

Pattern 4 (Complex) → Understand current state
Pattern 6 (Strategic) → Align incentives
Pattern 5 (Feedback) → Create learning loops
Pattern 1 (Unclear) → Clarify vision repeatedly

Product Innovation

Pattern 2 (Conventional) → Break out of incrementalism
Pattern 3 (Assemble) → Combine novel elements
Pattern 5 (Feedback) → Iterate based on learning
Pattern 6 (Strategic) → Design ecosystem fit

Beyond the Patterns

These 6 patterns are starting points, not exhaustive. As you work with the HUMMBL Base120 framework:

You’ll discover your own common patterns
Patterns will blur and overlap
New patterns will emerge from experience

The goal is not rigid categorization—it’s faster pattern recognition so you can apply the right mental models quickly.

Contributing

Found a new pattern or want to improve these descriptions? See CONTRIBUTING.md for how to submit feedback.

Problem patterns are validated against 22 months of framework development and 140 chaos tests. They represent common, repeatable starting points—not universal rules.

This site is open source. Improve this page.