What is bayesian reasoning?

Example: You believe there is a 70% chance your team will ship the product on time. Then the lead engineer quits. A naive thinker either ignores this (conservatism) or panics and drops to 5% (overreaction). A Bayesian updater asks: how likely is a lead engineer quitting if the project were on track versus off track? The answer shifts the estimate — maybe to 40% — proportional to the diagnostic value of the evidence, not to its emotional intensity. The number is not the point. The discipline of proportional adjustment is the point.

This concept is part of Perceptual Calibration in the How to Think curriculum.