Forensic judgment of paternity depends on genetic and nongenetic evidence. Sometimes genetic marker tests can provide clear evidence to exclude a falsely accused man, but they do not always succeed. On the other hand, if the accusation is true, the alleged father will not ever be excluded by genetic marker tests. When a nonexclusion case occurs after one or more marker tests, a report of paternity probability is required in the court. The current methods for calculating the paternity probability are the "paternity index method" and the "nonexclusion method." A number of recent articles have openly debated the fallacies, validity and utility of both methods. This paper briefly reviews the two methods and proposes a non-excluded set method along with examples for illustrating the various patterns of paternity probabilities of the three methods.