Pre-defined patterns that map common problem types to recommended mental models
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.
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
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
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
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)
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
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
{
"tool": "search_problem_patterns",
"arguments": {
"query": "innovation"
}
}Returns: Pattern 2 (Stuck in Conventional Thinking) with top 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.
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.
Common pattern sequences for complex challenges:
- Pattern 1 (Unclear) → Define the problem
- Pattern 4 (Complex) → Break it down
- Pattern 2 (Conventional) → Generate novel approaches
- Pattern 3 (Assemble) → Combine into solution
- Pattern 4 (Complex) → Understand current state
- Pattern 6 (Strategic) → Align incentives
- Pattern 5 (Feedback) → Create learning loops
- Pattern 1 (Unclear) → Clarify vision repeatedly
- 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
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.
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.