This is the capstone. It pulls together everything the course taught: why one model turns every problem into a nail (the man with a hammer), why Munger’s cross-disciplinary latticework beats depth in a single field, the three ways models actually combine (they stack into a lollapalooza, they check each other’s blind spots, or they complete a picture by explaining different parts), how to run the whole lattice on a live case, how to choose which lenses fit a novel situation, and how a written checklist plus a pre-mortem keep you honest against the four pitfalls. Several questions hide a trap option — watch for the answer that says models merely add up, the one that crowns a single model as enough, the one that treats a stacked reading as PROOF the crowd is right, and the one that shouts “use the lattice!” without naming any specific models. Read every option before you lock.
How this exam works
One question at a time, and each answer locks the moment you submit — there is no Back, no retry, and no Restart. Your score stays hidden until the very end. The pass mark is 70%. Some questions are “select all that apply” and will say so — on those, you must tick every correct option and no wrong ones to get the mark. Take your time; a locked answer is final.
What is "the man with a hammer" syndrome?
Select an answer to continue.
Course Recap
Big picture
Combining Models: The Latticework — the five pillars
- Combining Models: The Latticework
- The man with a hammer
- One favourite model bends every problem to fit it — to a hammer, all is a nail.
- The model you know best is the one you over-apply, and it gives no warning at its edge.
- The latticework (Munger)
- Big models from DIFFERENT disciplines, held at once, with experience hung on the frame.
- Cross-disciplinary breadth beats depth in one field; a handful of big models pay across domains.
- Three ways models combine
- STACK: same direction, forces multiply into a lollapalooza — a danger flag, not proof the crowd is right.
- CHECK: one model vetoes another's blind spot (base rate vs vivid story, incentive vs stated reason, second- vs first-order).
- COMPLETE: each model explains a different part (price, players, dynamics).
- Running the lattice + choosing lenses
- Read a live case (bank run) through incentives, game theory, feedback loops, social proof, critical mass, base rates, second-order thinking at once.
- The meta-skill is CHOOSING which models fit a novel situation.
- The circle of competence tells you when you cannot even pick the lenses.
- Checklist + four pitfalls
- A WRITTEN checklist and a pre-mortem, run on purpose, keep you honest.
- Forcing a model where it does not fit; double-counting one effect under two names.
- Confirmation-shopping for the flattering model; paralysis-by-lattice past the point of deciding.
- The man with a hammer
Key takeaways
The whole course is one argument: no single model is enough, because the man with a hammer bends every problem to fit his one tool and gets no warning at its edge — so build Munger’s latticework of big models from many disciplines, held at once. Those models combine in three distinct ways: they STACK in the same direction and multiply into a lollapalooza (a powerful, self-reinforcing signal and a danger flag — never proof the crowd is right), they CHECK each other’s blind spots (a base rate disciplining a vivid story, an incentive auditing a stated reason, second-order effects catching what first-order missed), or they COMPLETE a picture by each explaining a different part. Running the lattice means reading a real case — a bank run, a mania, a red-hot fundraise — through many lenses at once, and the true meta-skill is CHOOSING which lenses fit while your circle of competence tells you when you cannot even choose. Externalise it all with a written checklist and a pre-mortem, and stay guarded against the four pitfalls: forcing, double-counting, confirmation-shopping, and paralysis-by-lattice. Breadth plus discipline plus honesty about your limits — that is the practitioner this capstone was built to make.