Looking back over a career, I find that most of what went well traces to a small number of decisions with a particular shape: the downside was bounded and survivable, and the upside was not bounded at all. None of these cost much to attempt and none of them could have been predicted.

What strikes me about these in retrospect is how unremarkable they felt at the time. The large upside is invisible in advance, because if it were visible the bet would not be available, and so the thing shows up in your calendar as a minor distraction competing against work that is concrete, urgent and clearly valuable. Measured against that, the asymmetric bet loses every time, and it loses on a perfectly reasonable assessment of expected value conducted with the information available. The error is systematic rather than careless.

The only defence I have found is to stop evaluating these things by how compelling they look and start evaluating them by the shape of the payoff. Two questions: what is the worst realistic outcome, and can I absorb it? You are playing for the tail, and the tail only arrives for people who took enough shots at it. So, as a reflection for today: what small bet with a bounded downside have you been declining because it did not seem important enough?