# Ideas
Mental models worth internalising. The source count shows how often an idea appears across books—higher counts signal ideas worth understanding deeply.
| Idea | Sources | What it does |
|------|---------|--------------|
| [[Activity-Level Analysis]] | 4 | Profitability belongs to activities, not industries |
| [[Base Rates]] | 4 | Start with how often things happen in general |
| [[Bayesian Probability]] | 4 | Update beliefs proportionally to evidence strength |
| [[Black Box Thinking]] | 4 | Treat failure as data; build systems that capture learning |
| [[Calibrated Questions]] | 2 | "How" and "What" questions that guide without commanding |
| [[Chesterton's Fence]] | 2 | Understand why something exists before removing it |
| [[Complex Adaptive Systems]] | 3 | Systems where agents adapt to each other—unpredictable but patterned |
| [[Complements]] | 2 | Products that make yours more valuable |
| [[Counter-Positioning]] | 2 | New model incumbents can't copy without damaging themselves |
| [[Crystals of Imagination]] | 3 | Products embody knowledge; the economy amplifies access to others' knowhow |
| [[Cynefin]] | 2 | Different domains need different approaches (simple/complicated/complex/chaotic) |
| [[End-to-End Process]] | 3 | Own the whole process, not fragments; eliminate handoffs |
| [[Ergodic Hypothesis]] | 2 | Time average ≠ ensemble average; what matters is your path |
| [[Explore vs Exploit]] | 4 | Balance trying new things against using what works |
| [[Goodhart's Law]] | 2 | When a measure becomes a target, it ceases to be a good measure |
| [[Growth Mindset]] | 3 | Abilities develop through effort; failure is feedback |
| [[Hanlon's Razor]] | 2 | Never attribute to malice what's explained by incompetence |
| [[Inversion]] | 4 | Solve backward: avoid stupidity rather than seeking brilliance |
| [[Kelly Criterion]] | 2 | Size bets proportionally to edge; never bet everything |
| [[Law of Requisite Variety]] | 3 | Only variety can absorb variety—match complexity to environment |
| [[Little's Law]] | 1 | Cycle time = work in progress / throughput |
| [[Loss Aversion]] | 7 | Losses hurt ~2x more than equivalent gains feel good |
| [[Marginal Gains]] | 3 | Break problems into parts; improve each by small amounts |
| [[Mimetic Desire]] | 3 | We want what others want; desire is imitative, not autonomous |
| [[Multi-Armed Bandit Problem]] | 1 | Balance exploration and exploitation dynamically |
| [[Network Effects]] | 4 | Value increases as more people use it—winner-take-most dynamics |
| [[OODA Loop]] | 2 | Observe → Orient → Decide → Act; speed of iteration wins |
| [[Optimal Stopping]] | 2 | When to stop searching and commit (37% rule) |
| [[Optionality]] | 7 | Prefer choices that keep options open; asymmetric upside |
| [[POSIWID]] | 4 | The purpose of a system is what it does, not what it claims |
| [[Process Power]] | 2 | Operational excellence that takes years to replicate |
| [[Process vs Outcome]] | 6 | Judge decisions by quality of thinking, not results |
| [[Recognition Primed Decision Making]] | 2 | Experts pattern-match, then mentally simulate |
| [[Scale Economies]] | 3 | Unit costs decline as volume increases |
| [[Scale Economies Shared]] | 4 | Pass scale savings to customers; create compounding loyalty flywheel |
| [[Second-Level Thinking]] | 3 | Ask "and then what?"—first-level thinking is consensus |
| [[Secretary Problem]] | 1 | Explore first 37%, then pick first option better than all explored |
| [[Self-Image]] | 2 | Performance equals self-image; change identity to change results |
| [[Skin in the Game]] | 3 | Only trust those who bear consequences of their advice |
| [[Switching Costs]] | 2 | Friction that keeps customers from leaving |
| [[Systems Thinking]] | 7 | See interconnections, feedback loops, and delays—not just events |
| [[Tacit Knowledge]] | 5 | Knowledge you can only get from doing, not reading |
| [[Tactical Empathy]] | 2 | Understand their world to influence their decisions |
| [[The Barbell Strategy]] | 3 | Combine extreme safety with small high-risk bets; avoid the middle |
| [[The Five Focusing Steps]] | 2 | Identify constraint → exploit → subordinate → elevate → repeat |
| [[The Value Stick]] | 3 | Strategy = raise WTP or lower WTS; capture the spread |
| [[Theory of Constraints]] | 4 | System output is limited by one bottleneck; fix that first |
| [[Via Negativa]] | 4 | Improve by removing, not adding; less is often more |
| [[Viable System Model]] | 2 | Five functions every organisation needs to survive |