… even when they’re frustrating
The Legal Genealogist owes readers an apology.
Between the June week of the Genealogical Research Institute of Pittsburgh (GRIP) and a hard-to-diagnose technical issue with the website, the blog has been really quiet this week.
And it may not get a whole lot better for a while as the July GRIP is coming up fast, and the Midwest African American Genealogy Institute the same week, and then the Institute of Genealogy and Historical Research, and…
You get the picture.
But I did take a break for a short time and play a little with my own genealogical DNA results yesterday, just to see how the numbers stack up with even the limited support we can get from Ancestry’s ThruLines.
We all have 64 fourth great grandparents and 128 fifth great grandparents. If cousins married cousins, those numbers can change a bit in what’s called pedigree collapse.1 But I don’t have any known pedigree collapse that close (I have a set of fifth great grandparents who are also my sixth great grandparents so I start to lose unique individuals from my tree at that level), and that makes the fifth great grandparents a good level to look at.
So… remember how ThruLines works. It takes our trees and the trees of our known DNA matches and suggests the likely common ancestors. If the trees are wrong, the suggested common ancestors will be wrong. And even if the trees are right, ThruLines won’t account for DNA that may come down from another as-yet-unknown common ancestor who isn’t in one or both of the trees.2
All that being said, it still highlights the one reality of my ancestry: I have some lines that have a lot of genealogists (or at least a lot of DNA testers), some lines with darned few, and — sigh — the Germans still aren’t testing in high enough numbers to help out at all. Oh, and an awful lot of people who — like me — have no clue who the MRCA is in a particular line.
The lines with an abundance of testers? All from the U.S. southern states:
• In my Baker-Davenport line (Virginia to North Carolina), more than 100 matches even excluding those who are my close kin (sister, cousins, aunt and uncle).
• In my Wiseman-Davenport line (Virginia to North Carolina), more than 130 excluding close kin.
• In my Buchanan-Boswell line (Maryland to Virginia to North Carolina), more than 180 excluding close kin.
• And in my Jones-Pettypool line (yeah, you got it: Virginia to North Carolina), we’re pushing close to 300 matches.
And yes — sigh — the Bakers married the Davenports married the Buchanans married the Joneses, all in western North Carolina.
The lines with the fewest testers? The Germans, by a long shot. I have at least some documentation for 28 German fifth great grandparents. For 17 of them, I have exactly one match: my sister. For one more, my matches are my sister and my niece (who should match me on the others but hasn’t added them to her tree). For another eight, I have two matches: my sister and a fourth cousin, whose second great grandparents emigrated to the United States in the 1850s. We have, of course, no matches in common since you don’t get shared matches on Ancestry if you don’t share at least 20cM of DNA in common, and I only share one 8cM segment with this fourth cousin.
And for the remaining two German fifth great grandparents, I have three matches besides my sister: a possible fourth cousin once removed whose almost entirely undocumented tree shows a shared ancestor married to two different men with the same surname both born in the same year, and a possible fifth cousin once removed and a possible sixth cousin — whose trees show a second German surname that could just as easily explain the match in Germany. Oh, and yeah… and who have a Baker ancestor in colonial Virginia.
Sigh…
Of course that’s still better than the ThruLines showing that Matthew Johnson, my third great grandfather in another out-of-Virginia line, could be the son of at least two different guys. Like John Joseph Daniel Johnson based on (for example) a tree where (a) the sole source is (yep!) other undocumented Ancestry Family Trees and (b) the Matthew Johnson shown and all his siblings were born in North or South Carolina rather than Virginia and (c) that Matthew died in 1887 in Georgia — whereas mine died about 1863 in Kentucky.
But hey… it’s a hint, right?
Now… despite my frustrations, let me make it clear: these can be valuable hints. They can still do a lot to help put families back together. I noted today, for the first time, a tester who descends from a second great granduncle who died in Iowa in the 1850s where we share a whopping big 42cM segment.
So even when — in moments of frustration — I think I may be through with ThruLines, in fact I know I never will be. Not when they can put me in touch with cousins like that.
How are your ThruLines helping… or not?
Cite/link to this post: Judy G. Russell, “Not through with ThruLines,” The Legal Genealogist (https://www.legalgenealogist.com/blog : posted 26 June 2022).
SOURCES
- See e.g. DiAnn Iamarino, “3 Ways to Find Double Ancestors in Your Family Tree,” Family Tree Advice, posted 16 Aug 2019 (https://family-tree-advice.blogspot.com/ : accessed 26 June 2022). ↩
- See “AncestryDNA® ThruLines®,” AncestrySupport, Ancestry.com (https://support.ancestry.com/ : accessed 26 June 2022). ↩
I was most excited when I discovered the paper trail of a sister of a German immigrant second great grandfather. I can track two lines of her descendants thanks to DNA and Thrulines (unfortunately none of these folks seem interested in talking to me. Oh well.)
But… then… I discovered a DNA match on My Heritage, 20cM, one segment, that is related through the unique surname of a husband of a descendant (a granddaughter to this second great aunt). On Ancestry I have 12 DNA matches to this couple, ranging from 11-43cM, six in Thrulines. They are all shared matches to each other, no one that appears to be an “outlier” to this German group. While my connection to the My Heritage match is unknown, it looks like this husband is either his uncle or great uncle. *Many* of the My Heritage shared matches appear to triangulate with the two of us, and there are a lot of ’em. My gut feeling is that the MH match is a Maryland connection to my paternal grandmother’s mother’s side, where the German connection is through my paternal grandmother’s father’s side.
Oh there is a Theory of Family Relativity for this DNA match on My Heritage, but it’s totally wrong, but that’s another story!
I keep hoping a few more of my Germans will end up with MyHeritage… c’mon, cousins…
I wish some of these Ancestry DNA matches would test elsewhere!
As to another point you made, because of a lack of descendants and people who won’t test, many of my ThruLines are only there because of my two young adult children who come up as shared matches. While I’m glad ThruLines recognizes them, I feel I’m cheating the system. The “legs” of this “triangle” have a very narrow base and could topple over quite easily! LOL
It really would help if we could review some of these in a chromosome browser, but… ain’t gonna happen on Ancestry.
Judy, I share your love/hate relationship with ThruLines as well as the imbalance of family groups who test. I’m very fortunate that my paternal uncle agreed to test; I’ve used his kit to break through a long-standing brick wall in my Dad’s Irish line. Thank goodness a previously unknown 2nd cousin tested, which opened up an NPE line and provided her family with answers to questions about their father’s ancestry.
At the same time, I’ve been able to wade through my uncle’s 149 matches at 4th great grandparent line (Leatherman/Schwalley) to be able to focus on the less tested lines. I’ve also come to the conclusion that, despite lacking a single piece of reliable documented evidence, DNA points to a confirmation that my Beightler/Bigler line married a Luckinbill as the ThruLine indicates.
I believe the key is utilizing the tool that’s available and leveraging it with traditional genealogical research to locate evidence to support or refute the “suggested” connections. Which can be fun…or frustrating…depending on how much time you have to spend!
Or fun AND frustrating both at the same time! 🙂
I have 860 matches in Thrulines with descendants of Gutridge Garland and Bridget Hampton, who are my 4th and 5th great-grandparents, depending. Others of interest:
Wiseman-Davenport = 276
Baker-Webb = 309
Buchanan-Jones = 391
Of course, there’s a huge amount of overlap among these groups so it’s not quite as bad as it looks. I wish there was an easy way to identify unique matches.
That whole side of the family is ridiculously prolific — and soooooo many of them do DNA test. I feel for you, cousin!
ThruLines for me is worse than useless.
It suggests 12 4th-greats me, but…
Six of those are easily proved to be wrong… “ancestors” that folks related to me have latched onto without evidence or in the face of easily found evidence against, and have copied from each other hundreds of times, citing either nothing but each other or else actual records that do not provide evidence for the claim.
Four of them are doubtful… evidence hard to find, but circumstances suggest the construction is fairly likely to be incorrect (none of the other trees providing any evidence).
The last two are unproven – no evidence in the other trees, and none that I can find.
Because of how it is organized, all that noise in 4th-greats makes 5th-greats unworkable… there might be some nuggets there, but they are swamped by the continuation of the prior generation.
I say worse than useless, because it seems that other users keep piling on to those “false consensus” and “undocumented” lines and copying them, and it seems that pretty much nobody is doing any actual research. Or if they are, they don’t seem to be putting documentation in their public trees.
In my view, Ancestry ThruLines is part of the problem, not part of the solution.
I know it doesn’t help when people keep “adopting” ancestors from undocumented trees. But every so often… there’s a hint buried, and this is one tool to find the hints.
ThruLines helped me break down a brick wall. (Also from a U.S. southern state, haha.)
My 3rd great-grandfather, William, had been almost a complete mystery to even his own son — all we had was a name and date of birth, the initials of a probable family member, a probable general location in the late 1850’s, and a dramatic rumor about his death.
My grandparents did some initial research and found a few families with young men by his name of approx. the correct age in that area in the 1850 census. More recently I’d found a birth record for his son and a possible death record (albeit so damaged it’s almost impossible to read). But there was no “connecting” piece to any of the potential families from the census.
Several descendants of William have done AncestryDNA — and according to ThruLines we all matched with branches from 1 of the potential families from the 1850 census. Some of William’s siblings lived much longer lives and their descendants had more detailed genealogical info to go on.
ThruLines hasn’t been 100% accurate for me, of course. There are a few branches where it seems like all the matches just have the same “guess” on their family tree, and it’s clear that there’s no real evidence for the connection. But in this particular case I was really grateful for the help from ThruLines.
Glad to hear about your breakthrough!
ThruLines got me some Mayflower ancestors and 1 Jamestown ancestor. I had a couple of brick walls about 4 generations back. Thanks to ThruLines, it shows a possible ancestor with several DNA matches (admittedly at really low levels – many below 20cM), but it gave me the clue I needed to find the correct documentation. It’s also starting to break a hole in the wall that is my 3xgreat-grandfather who was born in Quebec in 1824 but married and had family in Michigan in the 1860’s (no idea of his life before 1861). On the flipside, it keeps suggesting parents of my Nathaniel Foster as being Lemuel Foster and Dolly Davis when all the documentation that I can find has NO connection. ThruLines is basing it on the hundreds (yes, really hundreds) of trees that give the relationship with literally not one piece of documentation. So, I may never break that brick wall.
Gotta love those clues… and never give up hope for more.
File this under “preaching to the choir” but of potential interest to some of your followers:
ThruLines are potentially helpful when identifying more descendants of known ancestors but—with respect to finding unknown origins of “dead-end” ancestors— ThruLines can only provide false leads!
False leads cannot be detected at Ancestry, because Ancestry has no Chromosome Browser or segment triangulation tools. Also Ancestry, limits Shared Matches to 20 cM.
For these reasons (and more), everyone is encouraged to download DNA results from Ancestry and upload them to GEDmatch.com for free. GEDmatch use a threshold of 7 cM for matching and has excellent tools.
GEDmatch is powerful but intimidating to some. Once you’ve created a free GEDmatch Kit, you can simply take a “set-it-and-forget-it” approach and let others reach out to with “new” cousins and—if you’re lucky—“new” ancestors.
I do not — repeat do not — recommend GedMatch. It is owned by and operated under the auspices of a company providing forensic services and does not follow good ethical practices for genealogy. If you want a chromosome browser, upload to MyHeritage. DO NOT use GedMatch.
My experience is the same as yours. Some lines have many (no where near as many as your lines do) and some have so few. One of my second great grandparent marriages that produced 12 adult children has only 4 matches! None have helped me with brick walls. I have been doing fairly good research for 30 years and could contribute to others’ but so few ever respond. So many who take DNA tests aren’t interested in genealogy just did it at the request from a relative. My wish for Ancestry is that it has a way to separate the active genealogists from others and does not allow someone to attach your tree to theirs. Do your own research!
The lines I am most interested in have few to zero ThruLines. My mom and I are the only people showing up for one set of my 3rd great grandparents. I agree about differentiating between the active vs inactive genealogists. I’ve received no reply from several DNA matches. Very sad as one match was my great grandfather’s (who I never had the chance to meet) niece and my mom wanted to reconnect.
I have mixed feelings towards ThruLines. Some helped break through research questions I had on my maternal side. Then there’s ones that are not as helpful that link to the wrong people, based on sparsely documented or undocumented trees, the ones where users just keep copying other people’s trees. This created confusion for people re: one of my 4th great grandfathers. You have some descendants listing the correct info and a lot going with another guy with an unusual middle name. They lived in the same place, same time period. For my father’s line nothing earth shattering just a lot of DNA matches on branches I am not researching.
I think ThruLines is great, although I wish the List view showed the Star and/or the dots. There’s plenty of room in their current format.
I also find its value diminishes at the end.
At 3rd great gp level, I find the suggested ancestors to be 87% reliable.
At 4th, 75%
At 5th, 54%, so its starting to get rather low.
Also at 5th, many of the ThruLine ancestors point to matches under 10 cM, so the matches themselves are suspect.
Nearly all of the Potential/Green ancestors are based upon a whim. Not reliable at all.
I generally tell people that my Thrulines are reliable through 4th gr gp. And Shaky at 5th, although even some of those do turn out to be valuable.
Cheryl Chasin: One of my 5th gr gf is Samuel Garland, brother of your Gutridge. So we might be 7th cousins 🙂 But I have “only” 157 suggested matches. 🙁
TruLines has been, on average, a good thing. Of course sometimes it’s garbage in, garbage out but like MyHeritage’s Theories, one can figure it out fairly quickly. I have a lot of County Mayo Ireland endogamy and feel teh pain when a dna cousin to both my mom and dad only has tree data for one side or the other. No idea who MRCA is. But the fact there is a tree is a plus and if it provides new clues re: a previously unknown child or marriage or the maiden name of a spouse – that has often helped me out as much as it otherwise being accurate. It has found enough to get me to look at this person’s tree when I might otherwise have not done so.
Once in a while it strikes gold, more often it’s a head slap on the quality of the tree and often enough to keep me checking, it helps me out in some way.