However, a piece of evidence can only be given its diagnostic value by relating the likelihood of finding that evidence under the hypothesis of a particular source to the likelihood of finding that evidence under a rival hypothesis of another source.