Eric, If you've got surnames, give them surnames--lots of them! The 3rd-5th cousin estimated range may be inaccurate. Also, the last ancestor they know of may be a ways down in a female descent from the common male ancestor. I focus on geographic areas rather than surnames, because half of my ancestors were Scandinavian and had no fixed surnames before reaching the US. Geography is also of use for my Scots because if a match's Todds (for instance) are all from southern England, chances are it's some other line of Todds (not that I rule them out, but I don't focus on them). Karla On Thu, Oct 10, 2013 at 7:17 AM, Eric S Johnson <crates@oneotaslopes.org>wrote: > Am a gen-genealogy newbie. Am contacting matches on 23andMe and ftDNA. I > don’t have much exome triangulation to go on yet, so if I correctly > understand the process, ideally one compares patriarchs’ surnames lists > hoping to see an intersection. > > > > Is there an optimal way to share this list with matches depending on how > related we seem to be? I.e. if someone is offered as a “3rd to 5th > cousin,” it would seem to me un-useful to share with that person my > surnames more than X generations old, since it’s less likely our common > ancestry goes that far back. Does one provide maybe two alphabetical > lists-first, that of all patriarchs up to (e.g.) five generations ago > (since > that’s likely where we’re most likely to find the common ancestry), and > then another list of the next three generations (since that’s a > second-most-likely range)? I realize recombination is unpredictable and > also > that a given DNA overlap is indicative of sum of our “vertical” and > “horizontal” relationship distance, but it still seems to me there ought > to be one or two statistically-sweet-spot groups of surnames in which to > search. Or do we order our surnames from “most likely” to “less likely”? > > And for a “5th cousin or more distant,” add another list of > the next three generations or whatever? > > > > AIA if any of my terminology is incorrect. > > > > Best, > > Eric >