Open AccessUNDERESTIMATING THE PROBABILITY OF COINCIDENCE
Author(s) -
M. A. Muller
Publication year - 2014
Publication title -
obiter (port elizabeth. online)/obiter (port elizabeth)
Language(s) - English
Resource type - Journals
eISSN - 2709-555X
pISSN - 1682-5853
DOI - 10.17159/obiter.v35i2.11895
Subject(s) - coincidence , bayes' theorem , point (geometry) , context (archaeology) , epistemology , psychology , computer science , mathematics , mathematical economics , statistics , bayesian probability , philosophy , medicine , history , alternative medicine , geometry , archaeology , pathology
Coincidences are more common than most people might expect. It is quite possible that different pieces of evidence that seem to point in the same direction do so coincidentally. We come to the best possible conclusion about (say) the probability of guilt only after careful analysis of the combination of probabilities of the respective pieces of evidence has been performed in conformance with the principles of probability theory. Several methods are available for the evaluation and handling of such contingencies. Depending on the way a particular situation presents itself, Bayes’s theorem in one of its equivalent guises is often used. The danger in avoiding this type of reasoning is that incorrect conclusions may be drawn, believing that events are somehow beyond coincidence. When it happens in a court of law it may be extremely prejudicial to the accused. Coincidences are best understood within the context of probability theory.