Priors
A prior is the belief already in place before current evidence arrives; sound judgment depends on noticing that starting point and letting evidence move it by an appropriate amount.
You do not begin deciding when the evidence appears. By then, part of the decision is already present: an expectation about what you are likely to find.
E1The starting point inside the verdict
That pre-evidence expectation is your prior. New information does not produce a conclusion on its own; it acts on the confidence you brought into the situation. The prior supplies the starting belief, and the evidence determines how far that belief should move. This is the core logic extended by Bayesian Inference.
E1A starting belief is not a protected belief
A prior is useful only as a provisional baseline. If contrary evidence is automatically dismissed or bent to preserve the original belief, the prior has become a Trapped Prior. But discarding every strong prior after one weak signal is also poor updating: the point is not to have no expectations, but to change confidence in proportion to the evidence.
Write the odds before the result
Before checking the outcome of one decision tomorrow, record what you expect and how confident you are—for example, “70% likely.” After seeing the evidence, record the new percentage and the specific observation that changed it. This makes the prior visible instead of letting it quietly masquerade as the conclusion.
Episodes that teach this
-
From Instinct to Insight - How Do We Really Decide? Bayesian Thinking - Future IQ
· explained at 2:28
7,456 views
"These are your beliefs before you saw any evidence... these are called priors."