Monday, October 21, 2019
Make That Seven Ways to Czarnylas
Since I've been puzzling over those six matches to my paternal grandfather which popped up over the summer, I thought it was about time I drew up a massive tree to connect them all on one chart. I pulled up the names of those six matches, noted their likely relationship estimates given by the respective testing companies, and also marked down each match's total centiMorgans of connection to me, as well as the count of how many segments that measurement covered.
While I was in the process of doing that, I remembered that, since I had long ago tested my brother, I should look back at his results, as well. That's when I recalled one match I hadn't been able to place—someone who matched my brother, but who didn't match me. Over the years, we had written to each other, trying to figure out our connection but not quite reaching any conclusion. Until now, of course. Knowing what I know now, I can see exactly where this match fit in: in the same set of family members from that Michalski line in Milwaukee.
So now, I can say I have seven connections to my mystery grandfather. And I've now completed a very rugged, hand-drawn, tree drawing just where almost each of these matches falls in the big genealogical picture.
With that detail—and if my guess about Anna Zegar being Anastasia Zegarska is correct—I figured many of these matches relate to me at the level of third cousin once removed. The remainder—with one obstinate exception—can be listed as third cousins twice removed.
However, just to check whether this is within the range of possibilities, I headed over to a particular website to get some help. The website, DNA Painter, contains two freely accessible tools which I consider invaluable: an interactive version of Blaine Bettinger's Shared cM Project, and Leah Larkin's "What are the Odds?" chart.
What turned out to be most interesting was finding that some of those matches listed as fourth to sixth cousins also calculated out to be the same third cousin once removed level (by the pedigree chart) as others estimated by the DNA companies to be within the range of third to fourth cousin.
The corollary to this was likely seeing that the 100 cM cousin turned out to match the 49 cM cousin at the same level of relationship, despite the difference in centiMorgans: they were both my third cousins once removed. That was upheld by the percentages of likelihood of a specific centiMorgan measurement for the given relationship: it was only 19% as likely (as other relationships) to see a match with 100 cM in common at the level of 3C1R.
The encouraging news was that each of the DNA matches which I have been able to plug into this hypothesized family tree seem to be reasonable assumptions, based on these tools at DNA Painter. Of course, they could also fit a given number of other relationships, as well. But for now, I know that this proposed pedigree chart scenario is within the realm of reason.
The only drawback is that there is one match with one ancestral surname—Krzewinski—which I cannot confidently place within that chart, even though I can find Polish records showing a marriage between one of the Zegarska sisters and a man by this name. I have DNA matches with that surname in their family trees, including this one key match, but I have no way yet to verify through the paper trail that this is the right couple. I can't find the right way to attach this couple to the tree.
Other than that, it looks like these DNA matches do confirm my hypothesis of how they fit into my paternal grandfather's tree.
But am I happy yet? Not quite. I'm not sure why, but I don't yet feel confident about accepting the genetic genealogy indicators without more support from the paper trail. After all, I could have the wrong Zegarska women confused for my own relatives. And I'm not sure how to remedy that. Yet.
Before we attack ways to resolve that issue, though, let's take a look tomorrow at what I concluded from the hand-sketched pedigree chart which bundled all these DNA matches into one enormous family tree.