Tuesday, July 14, 2020

More About Those Odds

Battering down a genealogical brick wall can seem an enormous challenge, so I try to circle the problem and assess multiple approaches that might lead to my conquest. With DNA as one tool, we gain some possibilities open to us that a mere paper chase might not provide.

I've already mentioned one encouraging aspect of the sheer number of distant DNA matches when attempting to scale the brick wall surrounding the parentage of a distant ancestor: we have a lot more possibilities to examine. So it is with my husband's challenging ancestor, his second great-grandmother Johanna Falvey. Born in County Kerry—but to whom?—Johanna eventually married John Kelly, gave birth to at least four children, then left Ireland for, of all places, Fort Wayne, Indiana.

Now, four generations later, the DNA tests of my husband and his two sisters point to some promising DNA matches. Each of those matches include Johanna's surname Falvey in their family tree. However, the count of genetic material shared between my husband's immediate family and these others is rather low, hinting at the distance of the relationship—everything from eighty two centiMorgans down to a nice round number: zero.

I thought it might be useful to list all the possible relationships that can be represented by those centiMorgan measurements. After all, raw numbers sometimes tell us very little; it is when we translate them into concepts we can wrap our heads around that we get led to that aha! light bulb moment of clarity.

To try this exercise, I decided to begin with the Falvey match which contains the highest number of centiMorgans shared. That would be between one of my sisters-in-law and a gentleman in New Zealand, the very place where, post-emigration, our Johanna supposedly had a sibling living.

I took that number—82 centiMorgans—and entered it into Leah Larkin's "What Are The Odds" interactive chart at Jonny Perl's DNA Painter website. Here is what I got:

While I removed the relationships with the lowest percentage of probability (two categories, each ranked with three percent probability), there are still several connections for us to examine. If I remove all the half-relationships and focus on only the full relationships, we could be dealing with anything from a first cousin three times removed to a fourth cousin.

To help me avoid having my mind go cross-eyed, I then snagged a copy of my husband's pedigree chart to help keep the relationships straight—and to correctly count off all those x-times-removed connections.

To also help you avoid going cross-eyed, I've included the paternal side of his pedigree chart here. The little black rectangle in the lower left corner, labeled "Stevens," represents my husband. If you've been following along at A Family Tapestry for a long time, you've already met my husband's dad, Frank, and his paternal grandfather, Will. Johanna Falvey was Will's maternal grandmother.

If we take the simple relationships—third cousin and fourth cousin—as possible relationships, we are talking about finding a match for whom the Most Recent Common Ancestor would be Johanna herself, or her parents. Since I already have traced all the lines of descent from Johanna and her husband John Kelly, even though that thirty one percent probability looks the strongest on paper, it is not likely, given the DNA matches we are working with, thus I'll eliminate that third cousin possibility. On the other hand, the fourth cousin match would target Johanna's parents, the very ancestors I still need to identify—making this a key relationship to watch.

The more complicated relationships also pointed back to specific MRCAs. In the case of first cousin three times removed, that would represent a first cousin to Catherine Kelly, Johanna's daughter, and thus Johanna's parent as MRCA. A second cousin once removed would be a second cousin to Frank, yielding a MRCA as Johanna, herself, which I would again rule out as a possibility. Yet, the second cousin twice removed option would lead to Johanna's parent as MRCA, a possibility, despite the twenty four percent probability, which I could still accept. And the second cousin three times removed, at a twelve percent probability, would finger one of Johanna's own cousins, pointing to her grandparents as MRCA.

Thus, I can diagram the specific relationships which would be most likely, given a shared eighty two centiMorgans between my sister-in-law and this man from New Zealand. While I don't yet know the identity of Johanna Falvey's parents or grandparents, this gives me some specific connections which are most likely, based on the science of probabilities and what I already know about Johanna's own descendants.  


  1. A great exercise, Jacqi, I'm going to try this now. Unfortunately, I'm working with much smaller segments - wasting my time?!?

    1. Dara, it is the smaller segments which have been helping me push my way back a generation in some cases. Triangulating matches who share a known surname and geographic region, despite a smaller count of joint centiMorgans, can lead to answers you might not find any other way.

      I have seen that in the case of my Polish paternal grandfather's unknown origin, and I'm hoping the same will work out here. Of course, it all depends on what you mean by "smaller segments" but I've not eliminated using possible DNA matches until they go below 20 cMs...and even then, I've spotted cousins from a known pedigree with DNA match numbers approaching 10!


Related Posts Plugin for WordPress, Blogger...