Tools & thinking

How to think in frameworks

None of these are mine. Each is a shortcut past a mistake somebody else already made expensively, which is the best kind of thing to borrow.

1. Reversible vs. one-way doors.

   Jeff Bezos’ split: most decisions are doors you can walk back through, so make them fast. The few that are one-way deserve every hour you can give them. Most people invert this and agonise over the reversible ones.

2. The two-by-two product grid.

   Jobs, 1997: consumer and pro, desktop and portable, four boxes, everything else cancelled. Any time a list of priorities passes about six items, draw the grid and find out what it is.

3. Inversion.

   Instead of “how do I succeed here?”, ask “what would guarantee failure?” and avoid that. Carl Jacobi’s habit, Charlie Munger’s favourite. The failure list is always easier to write and more actionable.

4. Circle of competence.

   Know the edge of what you understand. The value isn’t in having a big circle -- it’s in knowing precisely where yours ends, and being honest at the boundary.

5. Expected value.

   Probability times outcome, for each option. It makes the case for asymmetric bets obvious: a 10% chance at something enormous beats a certainty of something small, and intuition never sees it.

6. Opportunity cost.

   The real price of anything is the best thing you gave up to do it. A good decision compared against nothing is not a good decision, it’s an unexamined one.

7. Second-order thinking.

   “And then what?” Asked twice. Most bad decisions are excellent first-order decisions, the cost shows up two moves later, which is where nobody looked.

8. Regret minimisation.

   Bezos again, on leaving a good job to sell books online: project yourself to eighty and ask which choice you’d regret. It strips out the short-term noise that dominates the decision in the moment.

9. The 10/10/10 rule.

   Suzy Welch’s question: how will I feel about this in ten minutes, ten months, ten years? Excellent for the decisions where emotion is loud and the stakes are actually small.

10. Compounding.

    The one underneath all the others. Small consistent inputs look identical to doing nothing for a long time, and then they don’t. Almost everyone quits during the part where it looks like nothing.