Week 9: Deciding in advance

Commitment, gates, and how a decision that is usually fine becomes the one that eventually is not

19 April 2027

The last lecture before you fly, and the one with the least aerodynamics in it.

An approach compresses time. The decisions it demands arrive in a fixed order at increasing speed, each with less margin than the one before — the opposite of the condition under which people decide well. The standard answer is to move the decision earlier: make it on the ground, in a form that requires only a check at the moment of use.

The lecture is mostly about why that answer, which everyone agrees with, fails so reliably in practice.

The mechanism is not recklessness. Continuing a marginal approach almost always works, and each time it works, the margin that was skipped gets retrospectively reclassified as margin that was not needed. The standard moves. Nobody decides to lower it — it lowers through a series of individually reasonable calls, until the accepted practice is one that would have been rejected outright on day one.

This pattern has a literature and a name, drawn from the study of organisational failures well outside aviation. We look at one case from a different industry, because the shape is easier to see when the technical detail is unfamiliar.

Then the direct part: each of you states, individually and out loud, which gate you are going to rationalise past in week 10.

Reading#

Your restriction list from week 3, and the published circuit gates. Arrive having written your own in the form if not X by Y, then go around.