Ann >From Larry's explanation, it seems our matches from long segments would rarely match a haplotype of one of our two parents. It would not even match the haplotype of one of the parents of our match. (I hope I'm using haplotype properly here) The matching process accepts any combination from either person (from either parent of either person), and then moves to the next location, looking for a match. I come back to what I thought I understood, which was that this could happen in short segments, but the 7.7cM cutoff was chosen because "usually" such a long segment of matches would only occur between folks who got that long segment from their Common Ancestor. This means to me that the entire long segment was from an Ancestor on the father's side or the mother's side. And therefor all the ACTGs in the matching segment were from one side. Otherwise I don't see the FF test as being much help to genealogists. Or we need to shift the cutoff? I believe some large segments survive somewhat longer than advertised by FTDNA, and our inability to find Common Ancestors stems from the fact that we don't know enough of our Ancestry. Jim - Sent from my iPhone - FaceTime! On Jan 6, 2012, at 6:00 PM, Ann Turner <[email protected]> wrote: > It could be a sign that some of those segments are Identical by State. As > noted by others in this thread, we're looking at genotypes, where we don't > know which allele came from each parent. If you could split the genotypes > into their two haplotypes, this would clarify the picture. That generally > takes father/mother/child trio data, though. For the initial query in this > thread, pair-wise comparisons of everyone with everyone else is a > reasonable first pass at deciding whether the segment comes from a single > common ancestor. > > Ann Turner