Bayes' Theorem Calculator

Bayes' Theorem mathematically formalizes the process of belief revision. The theorem states that the posterior probability of a hypothesis A given observed evidence B is equal to the likelihood of observing B if A were true, multiplied by the prior probability of A, divided by the marginal probability of B. The prior P(A) encodes your initial belief about A before seeing any evidence — this might come from historical base rates, expert knowledge, or previous experiments. The likelihood P(B|A) quantifies how probable the observed evidence is under the assumption that A is true. The marginal probability P(B), sometimes called the 'evidence' or 'normalizing constant', is the total probability of observing B across all possible hypotheses, ensuring that the posterior probabilities sum to one.

The marginal probability P(B) can be expanded using the law of total probability: P(B) = P(B|A)·P(A) + P(B|¬A)·P(¬A). This means if you know the prior P(A), the likelihood P(B|A), and the false positive rate P(B|¬A), you can compute P(B) without needing it directly. This decomposition is critical in practice because P(B) is rarely measured directly — it must be derived from the constituent parts. The result, P(A|B), is the posterior probability: your updated rational belief about hypothesis A after accounting for the evidence B.

The theorem underpins Bayesian inference, a complete statistical paradigm that treats probability as a degree of belief rather than a long-run frequency. In machine learning, Bayesian classifiers use this theorem to classify text, filter spam, and diagnose medical conditions. In epidemiology, it is used to determine the probability that a patient truly has a disease given a positive diagnostic test result — often revealing startlingly counterintuitive results when base rates are low. In court, it has been applied (and sometimes misapplied) to assess the weight of forensic evidence. Understanding and correctly applying Bayes' Theorem is a cornerstone of probabilistic reasoning in science and engineering.

Bayes' Theorem is mathematically exact, but its practical application depends entirely on the quality of its inputs. Choosing an appropriate prior P(A) is often subjective and contested — different researchers may assign different priors to the same hypothesis, leading to different posteriors. In low-data regimes, the choice of prior can dominate the result, making conclusions sensitive to assumptions that may be difficult to justify. The marginal probability P(B) must be computed carefully using the law of total probability; omitting relevant alternative hypotheses will produce an underestimated P(B) and an overconfident posterior. Additionally, Bayes' Theorem in its simple form applies to binary hypotheses (A and ¬A); extending it to multiple competing hypotheses requires the full sum over all alternatives in the denominator. It also assumes that probabilities are well-calibrated and that events are defined precisely — in real-world applications, ambiguous event definitions or poorly measured base rates can severely distort results. Finally, the theorem does not account for model uncertainty or the possibility that the hypotheses themselves are incorrectly specified.