Suppose for X million dollars we can either: A) Launch one rocket with a 90% chance of success B) Launch two rockets each with a 70% chance of success We sho

Suppose for X million dollars we can either: A) Launch one rocket with a 90% chance of success B) Launch two rockets each with a 70% chance of success We should do (B) because it has a 91% chance that at least one of the rockets succeeds, and a 49% chance two rockets succeed--an outcome that is impossible under scenario (A). With the right telemetry, the 51% chance that at least one fails will give us valuable knowledge that helps us improve our rockets the next launch. Of course, the calculus changes if we let failure-averse marketing people run our rocket company.