Minimum Effective Dose and Discrete Outcomes

Jul 31, 2023
The minimum effective dose (MED) is defined simply: the smallest dose that will produce a desired outcome. Anything beyond the MED is wasteful. To boil water, the MED is 212°F at standard air pressure. Boiled is boiled. Higher temperatures will not make it “more boiled.” Higher temperatures just consume more resources that could be used for something else more productive. — Tim Ferris

For tasks with discrete outcomes, the minimum effective dose (MED) is often the best strategy. All other things equal, accomplishing the same task with less energy is more efficient. It’s not so much about the calculus of putting in minimal effort but allocating your effort efficiently.

Ferris uses examples of fitness (e.g., 30 mins per day vs. 3.5 hours once a week), medication (taking more of certain drugs won’t help),

Some examples:

  • Building a minimal viable product to test a hypothesis. Given a hypothesis, adding more features or polishing won’t help validate the hypothesis past the necessary features.
  • Bin-packing applications — A program that fits entirely in memory won’t benefit from more memory. A file that fits entirely on a disk drive is just as “stored” as the same file on a larger disk (even with redundancy, there are diminishing returns). There are also diminishing returns to increasing the CPU cores an application runs on.
  • “Winning without fighting is preferable” — Sun Zhu

Cases where outcomes are not purely discrete, but there are severe enough diminishing returns to seem like it (or even negative returns to scale):

  • “Two pizza teams” / Mythical Man Month — Right-sized teams achieve the most. Past a point, adding more people slows projects down.
  • “C’s get degrees.” Courses are often graded on a discrete letter scale — your raw score isn’t recorded. A 95% vs. 100% score is usually registered as an “A.” (Of course, there are many other reasons to put in as much effort as possible in certain courses).
  • The costs of overengineering.
  • Ferris applies MED to fitness training, finding the minimum time needed to elicit the desired increase in some fitness metric.
Daily posts on startups, engineering, and AI