FWIW, there’s a fairly rich research literature that argues abductive reasoning is the primary form of reasoning entrepreneurs engage in.
I can’t find a use for it (though b-schools are obviously interested, because pedagogical reasons), but I thought it was really interesting.
Conversation
The basic form of abduction is "I observe X, I abduce Y as a reason for X." The nuance comes from the fact that Y is testable. It allows you to verify via experimentation/practice. (Screenshot from en.wikipedia.org/wiki/Abductive)
1
You can see why this might be the sort of thinking entrepreneurs do. When faced with uncertainty, you could a) deduce from some theory that Y is going on; b) perform the sort of calibrated probabilistic judgment to induce a possible explanation (i.e. Bayesian updating).
1
OR c) you could just pick the most likely explanation, and then move quickly with a course of action that assumes it is true and protects your downside, and re-eval later when your actions give you new information.
1
1
A simpler way of putting this:
Quote Tweet
I keep coming back to this model of decision making that @berkun articulated in The Year Without Pants.
I didn't have the words for it before. The time you spend doing expected utility calculations is better spent picking one path and running hard and getting new information.
Replying to
The next time you are faced with uncertainty (e.g. you observe X in the market and X may be explained by any number of reasons) remember that generating a calibrated judgment is only half the challenge. The other is being decisive.
1
1
The way I try to remind myself of this is to say to myself 'if there is an action A that has low downside risk and will generate more valuable information than is currently available, the right move is nearly always to stop judging and to do A instead'.
1
6
This is abductive reasoning because very often action A will assume a reason P for observation X. And P may be shitty.
The trick is to not get too caught up over analysis. You don't want to spend too much time arguing P vs Q vs R as reasons. Just go do A and re-eval later.
2