Bayesian Updating
Updated 2026-08-17
INTRODUCTION
English translation pending.
CORE DEFINITION
Bayesian updating is the algorithm that follows from treating belief as a probability. The posterior probability equals the likelihood of the observed evidence under the hypothesis, multiplied by the prior probability, divided by a normalizing constant that keeps the total coherent. Its instruction for the mind is double-sided: do not hold rigidly to the old view, and do not let one piece of news flip it entirely. Instead, ask how much more likely the new evidence makes the old belief, and let the answer set the size of the move.
SCAFFOLDING EFFECT
Reduce cognitive load
- Calibrate the move: ask how much the evidence raises or lowers the belief, then move exactly that far. - Hold the middle: refuse both the frozen prior and the total conversion, since neither is what the rule prescribes. - Keep a running ledger: treat today's posterior as tomorrow's prior, so beliefs track accumulated evidence.
Anchor fast decisions
The formula works by weighing two sources of information: where the belief stood, encoded in the prior, and how diagnostic the new evidence is, encoded in the likelihood. Neither alone sets the answer, which is why a strong prior survives weak evidence and why a weak prior can be overturned by decisive evidence. Iterating the rule makes belief a path rather than a point, so credibility accumulates over many observations and is gradually eroded by contrary ones, rather than being reset with each new headline.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/Bayesian_inferenceverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS