DECISIONSCORE
Decision Trees + Value of Information
Fold back probabilities; pay for information only if it changes the action.
How it works
- Branch on chance, fold back expected values from the leaves.
- Information is worth exactly what it changes: EV(with) − EV(without).
- Decide first, then buy info only if the order flips.
Use → Product, medical and policy bets.
Watch out → Paying for precision that changes nothing.
Try it → Price one real decision's 'would I act differently?' test.
REDCAPE jobs → Act · Design · Communicate
Free sources → Coursera M2 (Decision Models)
Attach to a lattice — combines with
Different families, shared jobs. Disagreement between them is where insight lives.
GAMES
Collective Action & Commons (Ostrom)
Shared pools survive with boundaries, monitors and graduated sanctions — no panacea.
SHARED JOBS: DESIGN · ACT
LINEAR & NONLINEARConcavity & Diminishing Returns
Curved payoffs change optimal bets: diversify when concave, concentrate when convex.
SHARED JOBS: DESIGN · ACT
LEARNINGDiversity Prediction Theorem
Crowd error equals average error minus diversity. Diverse models beat the lone genius.
SHARED JOBS: DESIGN · COMMUNICATE