Finished 24/01/2021
Reckoning with Risk
The book tackles the issues involved in understanding what risk actually is in our decisions, especially in medicine and finance. It starts well, but toward the end it drags into a long critique of the U.S. system. Still, it has some worthwhile key points.
Notes
- Relative risk (+100%) vs. absolute risk (from 1 to 2 per thousand): big difference.
- Example: 9/11, fear of flying but cars are much more deadly. The risk we perceive is often shaped, and those who market medications know it well.
- Risk and uncertainty apply to different situations; we often try to treat things whose probabilities we do not know as “risk” (example: the turkey that does not know about Thanksgiving).
- Error culture: huge waste because people do not trust intuition (which they use a lot and is often more convenient than complicated rules), but they must “explain” their choices.
- Defensive choices are common: a doctor might recommend an unnecessary test to avoid risk.
- Screening and misleading advertising: when not needed, screening often harms.
- In finance, simple thumb rules are often much more effective than complex methods, which still fail (remember the euro-USD forecast charts, they often miss).
- Fear of risk is social. We usually fear what our peers fear (e.g., guns in the U.S.).
- In the end, after a long analysis of errors in medicine and finance, the gist is that you often do not need complicated models or convoluted decisions: trust the rules of thumb you adopt. Here are Gerd’s:
- Always ask for the reference class: Percent of what?
- Always ask: What is the absolute risk increase?
- Find the most important reason and ignore the rest.
- Ask Your Doctors What They Would Do, Not What They Recommend.
- Choose the alternative that avoids the worst outcome.