About that “closest” match…
It was one of The Legal Genealogist‘s goals for 2022: to identify the nearest as-yet-unidentified autosomal DNA match at each of the major testing companies.
Solving the first one, at MyHeritage DNA, was easy, thanks to some very big hints (like a full real name for the tester). He turned out, as I expected, to be a third cousin, descended from Gustavus Boone and Isabella Rankin (Gentry) Robertson — not surprising, since those Robertson second great grandparents had 11 kids, and they have left a truly prodigious number of descendants.1
The next one — at Ancestry DNA — is “match EL, a predicted 2nd – 3rd cousin with whom I share 125 cM … (and clues that) suggest this could be another Robertson cousin…”2
And, oh boy, is this one not going to be the charm…
There are some tantalizing clues here:
• The match size of 125cM gives me odds of roughly 75% that this is a third cousin or closer, according to DNA Painter’s Shared cM Project tool,3 and 96% that this is a third cousin or closer using Ancestry’s own data.4
• The longest segment is 36cM — a healthy size for a third cousin or closer.5
• The top shared matches between me and EL are all Robertson matches and represent most, if not all, of the children of Gustavus and Isabella.
• The family locations of folks in EL’s trees include locations where the Robertsons lived in Texas and at times overlapping the Robertsons’ time frames.
But then there are some complications:
• EL uses a username that not only isn’t related to any real name, it’s also not one that appears to have been used elsewhere online.
• None of EL’s Ancestry trees is linked to EL’s DNA test, so there’s no way to know for sure if the tester actually represents any of the lines shown in the trees.
• None of the online trees shows a link to any of the Robertson side of my family.
• EL hasn’t responded to messaging on the Ancestry system.
So… I enter my data into the WATO tool at DNA Painter to see what relationship is most likely for various levels of descendants from Gustavus and Isabella, and it tells me my most likely relationship is second cousin once removed (2C1R), with EL being one of Gustavus and Isabella’s great grandchildren.
But then it occurs me that I’m looking at much less of the data than I should be. I have access to other tests and can see how much shared DNA those other kits have with EL. And that’s where the big glitch shows up. Because the results are all over the map. In my generation, the shared DNA ranges from my high of 125cM to a low of 41cM. And a generation earlier? My aunt shares only 5cM more than I do with EL: 130cM. And my maternal uncle shares less than I do: 108cM.
Now what does the WATO tool say? It tells me the strongest case is for EL to be the second great grandchild of Gustavus and Isabella — so a 3d cousin to me and a 2C1R to my aunt and uncle. Or — with exactly the same likelihood, of course — EL could be a 3d cousin to me and a 2C1R to my aunt and uncle if EL is the grandchild of any sibling of either Gustavus or Isabella. And though it may not be terribly likely that a living tester today would be a sibling’s grandchild because of the time frames involved, the possibility of long-tailed generations means I can’t rule it out completely.
And, of course, it also tells me that it’s still entirely possible for this to be a second cousin to my aunt and uncle and a 2C1R to me and those in my generation, as a great grandchild of any sibling of either Gustavus or Isabella. I do in fact have a known 2C1R in exactly that line of descent — he’s younger than I am, born when his father was 51 years old. That father, in turn, was born when his own father was 58 years old.
Sigh…
This one’s gonna take some work.
Two is definitely not the charm.
Cite/link to this post: Judy G. Russell, “Two’s not the charm…,” The Legal Genealogist (https://www.legalgenealogist.com/blog : posted 15 May 2022).
SOURCES
- See Judy G. Russell, “One down…,” The Legal Genealogist, posted 10 Apr 2022 (https://www.legalgenealogist.com/blog : accessed 15 May 2022). ↩
- Judy G. Russell, “The 2022 DNA goals,” The Legal Genealogist, posted 9 Jan 2022 (https://www.legalgenealogist.com/blog : accessed 15 May 2022). ↩
- “The Shared cM Project 4.0 tool v4,” filter for 125cM, DNA Painter (https://dnapainter.com/ : accessed 15 May 2022). ↩
- “Possible DNA relationships,” JR and EL, Ancestry.com (https://www.ancestry.com/ : accessed 15 May 2022). ↩
- See Blaine T. Bettinger, “The Shared cM Project – Longest Shared Segment,” The Genetic Genealogist, posted 1 June 2015 (https://thegeneticgenealogist.com/ : accessed 15 May 2022). ↩
Interesting to read about your mysteries and your methods, and the WATO tool you are using. I enjoy these puzzles, although they can get confounding pretty quickly. Feeling a bit inspired, I think I will go have a look at my closest unidentified matches. I haven’t looked at matches recently and there are probably a lot of new ones! I tend to find my European matches on MyHeritage and my American ones on Ancestry.
Sometimes it’s good to know you are not alone in this type of struggle, so thanks for that. I am currently working on something fairly similar. If the match’s tree is correct, then the match is much further back, which would probably be outside the experienced range in the Shared cM project – but I need to do a recount to check.
Which brings up the possibility of an NPE in the match’s tree.
And I currently have only a vague idea of where it might be, because neither geography nor DNA provides any clues so far. Maybe segment analysis might help.
That is going to be a challenging conversation when it comes!