Mistakes bad enough you learn, but not bad enough they kill you

When we suggest to learn by doing, what we tend to mean is learn by making mistakes.  Or teach your system 1 to be better calibrated by making mistakes and adjusting.  Which is fine for the little mistakes: you learn which brand of cereal you like by buying the wrong one, trying it and then writing a note on your shopping list for yourself to buy the right one next time.

When you get to big mistakes it’s a bit harder to learn by doing because you don’t always get a second chance.  Some examples include:

  • Losing a million dollars on a sure bet.
  • Breaking your neck in a recklessly fun car accident
  • Committing suicide instead of asking for help from your peers when you just went through a relationship ending
  • Injuring yourself with industrial machinery

These all carry a pattern in them.  It doesn’t take a genius to say; well if I die – I can’t sit around and say – well next time I won’t be so careless or reckless or risk taking or next time I will be more careful or conservative…


So how is it?  Why is is that we care about learning by trial and error yet that looks like a terrible idea, leading to death and bad decisions over and over.  I propose a maxim:

You want to make mistakes that are bad enough that you learn from them, but not bad enough that they kill you.

To look back at the early examples while applying this maxim.  Examples that might teach you to be careful without killing you:

  • Losing a few hundred dollars on a sure bet.
  • Breaking your leg in a recklessly fun car accident
  • Getting piss-blind drunk and making a fool of yourself when you just went through a relationship ending
  • injuring yourself with hand tools (that don’t nearly spin as fast as industrial ones)

So how do we have the right kind of mistakes and not the wrong ones?  I don’t really know, but for starters I would say doing less dangerous behaviours and more safe-risk behaviours


It’s very hard to know what it feels like from the inside of a bad risk taking behaviour VS an okay risk taking behaviour.  It might be hard but I suggest classifying risks into multiple categories, and permitting yourself to go through risks that are safe, while restricting risks that are dangerous.  Risks that are safe:

  • low bets, in the few dollars, less than a handful of times a week
  • running late
  • trying new foods
  • trying new hobbies
  • doing activities slightly outside of your comfort zone
  • asking difficult questions
  • trying tasks that are a little difficult or new
  • working hard
  • sleeping a bit less

Dangerous risks might include:

  • big bets
  • signing up for dangerous activities
  • extreme sports
  • changing your entire diet, or fasting without planning or warning or knowing what you are doing.
  • having unprotected sex
  • taking drugs without experience or knowledge about what they will do
  • not sleeping
  • getting into fights with strangers
  • using powerful tools or equipment (industrial anything, or anything that spins very fast)
  • getting very very drunk
  • doing anything at high speeds – i.e. speeding in a car, boat or plane.
  • Using sharp objects while not concentrating

Near miss

Also in this category of risky things is near-misses.  We don’t always have accidents bad enough to register as “dangerous” or “deadly”, but we still want to learn from them.  This might involve being extra attentive to anything that might be noticed as a near miss.  I can’t tell you how to do that (not today in this post, but maybe in the future).


As a reminder again:

You want to make mistakes that are bad enough that you learn from them, but not bad enough that they kill you.


Meta: this took less than an hour to write.

This entry was posted in models of thinking and tagged , , , , , , , , . Bookmark the permalink.

Leave a Reply