- Event and evidence
- 26.25 %
- Complement and evidence
- 9.75 %
- Total evidence probability
- 36 %
- Event probability given evidence
- 72.9167 %
On this page
Distinguish the three inputs
The prior is the event probability before the evidence. The other inputs describe how likely that evidence is when the event holds and when its complement holds. For this conditional model, the prior must be strictly between 0% and 100%, so both conditioning branches have positive probability.
Inspect the fictional default model
Suppose a control weakness has prior probability 35%. An exception is flagged with probability 75% given a weakness, and 15% given no weakness. Weakness and flag has probability 26.25%; no weakness and flag has probability 9.75%. A flag therefore has total probability 36%.
Condition on the observed evidence
Of the total 36% probability assigned to a flag, 26.25 percentage points belong to the weakness branch. Dividing 26.25% by 36% gives approximately 72.9167% probability of weakness given a flag. The likelihood of 75% answers the reverse conditional question.
Use the model under its stated assumptions
If the evidence has zero probability under both branches, conditioning on it is undefined and the calculator rejects the inputs. A computed posterior does not verify the prior or evidence model, and it does not prescribe an operational, investment or compliance action.