A question: Persons A, B, and C are supposed second cousins to each other based on traditional genealogy, with a combination of well sourced and unsourced but detailed family trees. On GEDMatch/FTDNA, persons A and B match each other but neither one matches C. (Also add to this person D, a third cousin to the whole group, who is a DNA match for A and B, but not C.) How confident can I be that there was an NPE somewhere between person C and the MRCAs of A and B (great grandparents)? Seems like it, but is it even remotely possible that the relationship is correct but there's a lack of match just due to the quirks of probability? I looked at the Genetic Genealogist's shared cM project tables and it really looks like I should expect at least some matching at this level of relatedness.