Ancestry’s New ThruLines Utility Needs More Work

Last week, AncestryDNA® unveiled a new utility called ThruLines.™ You can read more about getting and using ThruLines™ from Ancestry’s article, here. Like many of you, I was anxious to play with it and see what, if anything, it did for me. I must say, I’m underwhelmed. Granted, the tool is still in Beta testing, so hopefully improvements will be made to the accuracy of the matching algorithm as time goes by. But as it is now, my concern is that ThruLines™ will only add to the existing confusion and misunderstanding of fledgling genealogists. Let’s walk through this utility to see what it offers and where the problems lie.

This shows my new Ancestry DNA home screen. I can access ThruLines™ on the right, and there’s a link at the bottom to click if I choose to continue using DNA Circles.

ThruLines first screen

The second screen gives me a portal to each of my ancestors to explore.

ThruLines second screen

When I first scrolled down on this screen, before I began to write this article, Ancestry highlighted a Potential Ancestor named Marianna Kozłowska, and informed me that she was my great-great-grandmother. Intrigued, I clicked on this person to examine the evidence for this assertion. On the next screen, Ancestry informed me that Marianna Kozłowska was the mother of my great-grandfather, Joseph Zielinski, and that she was mentioned in the family tree of a particular Ancestry member. If we take the information in the family tree at face value, Marianna Kozłowska was the daughter of Antoni Kozłowski and Tekla Stępkowska, born 1863 in Nowy Garwarz, Mazowieckie, Poland, near Glinojeck. (Antoni and Tekla were also reported to be my potential ancestors.) Marianna was married to Stanisław Zieliński, who was born 1863 in Wkra (also near Glinojeck). That fact is apparently the basis on which Ancestry’s algorithm determined that Marianna Kozłowska was my ancestor. I, too, have a great-great-grandfather named Stanisław Zieliński, who was born in 1863 and was married to a woman named Marianna.

The problem is, I have good documentary and DNA evidence that proves that my great-great-grandfather Stanisław Zieliński was born in Mistrzewice, Mazowieckie, Poland, not Wkra, and was married to Marianna Kalota, not Marianna Kozłowska.1 Moreover, my Marianna Kalota was the daughter of Roch Kalota and Agata Kurowska of Budy Stare, Mazowieckie, she was not the daughter of Antoni Kozłowski and Tekla Stępkowska. Marianna Kalota’s parents’ names and grandparents’ names are stated in my online tree, so it’s not as if there’s anything to suggest to Ancestry’s algorithms that I’m uncertain about the identifies of those ancestors. Closer examination of the tree which mentioned “my” ancestor, Marianna Kozłowska Zielińska revealed that the tree owner really has no good evidence for her claims about Marianna Kozłowska’s place of birth. For example, the passenger manifest that was supposed to document Marianna Kozłowska’s emigration to the U.S. was for a woman whose husband’s and children’s names did not match the data in the tree. Furthermore, the Marianna in the manifest was from Eckardtsfelde, Prussia, which is some 230 km west of Glinojeck.

There’s no shame in being confused about the origins of one’s ancestor, and everyone makes mistakes when they’re starting out in genealogy, so I’m not using this example merely to criticize the research of the woman who posted this tree. But I thought that surely there must have been some other basis for Ancestry’s conclusion that Marianna Kozłowska was my ancestor. Since this tree owner was clearly confused about where her Marianna Kozłowska was born, was it possible that she’s nonetheless a distant cousin of mine who simply made a few wrong turns while tracing her tree?

I checked out the profile of the woman who posted the family tree in question. If you’ve never done this before, you can access the profile of any Ancestry member with an online tree by clicking on the username found at the top left corner of the screen showing their tree. That will bring you to the screen shown below.

Ancestry Member Profile page

If that person is a match to you, or if any of the kits that (s)he manages are a match to you, it will be noted here. Additionally, if you manage other DNA results besides your own, you can use the drop-down menu, “Select DNA Test,” circled here in red, to compare this particular Ancestry user with any of the kits you manage.

To my surprise, she was not a DNA match at all. Not only did she not match me, she did not match my mother, the great-granddaughter of Stanisław and Marianna Zieliński. At this point it was pretty clear that the only basis for the assignment of Marianna Kozłowska Zielińska as my ancestor was her marriage to a man with the same name and year of birth as my great-great-grandfather. Never mind that Zieliński is the 8th-most popular surname in Poland, so there were undoubtedly quite a few Polish men named Stanisław Zieliński who were born in 1863. Sigh.

There is a bright side to this story, however. Ancestry requested feedback on my experience with ThruLines,™ via a little popup window, so I gratefully obliged them and expressed my concerns about their algorithm. Ancestry responded with lightning speed, such that when I returned to the site a few hours later to grab some screen shots for this blog post, there was no longer any mention of Marianna Kozłowska or her parents among my Potential Ancestors. Whew! Kudos to Ancestry for taking such prompt action in response to critical feedback. If nothing else, it underscores their desire to do the right thing by their customers.

Let’s examine another Potential Ancestor and see how that one shapes up. To quickly find these, I sorted my results according to this “Potential Ancestors” option using the drop-down “Filter by” menu at the top left. Once filtered, the results are shown below.

Potential Ancestors

Mary Cebulska intrigued me because there is a Maria Cebulska in my family tree, although she’s on my husband’s side. I also have Cybulskis in my tree since they married into the Zieliński family in Poland. However, examination of the family tree from whence this data came reveals that this is a reference to a fictitious Mary Cebulska who was purportedly married to my great-great-grandfather, Józef Grzesiak. This case was a bit trickier, since there was actual documentary evidence from a U.S. marriage record which stated that Józef’s wife’s name was Mary Cebulski. It turns out to be incorrect, and I wrote about this evidence previously. However, it was at least an honest mistake that any researcher might make if they were to base their case only on U.S. records instead of examining the evidence from Polish records. I won’t fault Ancestry for that one.

Next up, we have Walburga Meinzinger. I was a little surprised to find her in the list of “potential” ancestors because she’s an actual ancestor identified in my family tree on Ancestry. When I click on her name in this list, I arrive at a screen that tells me a little more about the connection.

Thrulines Walburga Meinzinge3r

Ancestry’s proposal of Walburga Meinzinger as my 4x-great-grandmother is based on her appearance in a tree posted by my paternal aunt, with whom I collaborate. Clicking on the “10 DNA Matches” brings me to a screen which may be the best part of the ThruLines™ utility, thanks to the clear graphic depiction of the relationships between me and my DNA matches who are also descendants of my great-great-grandparents, Wenzeslaus Meier and Anna Goetz.

Meinzinger tree

Is this new information for me? No, I had already discovered my connection to these folks by clicking on “shared matches” and either examining their online trees (where available) or writing to them. And the information about number of shared DNA segments and centimorgans of shared DNA is no more useful now than it was previously, in the absence of a chromosome browser which would allow me to paint these shared segments onto my chromosome map. Moreover, it’s misleading for Ancestry to highlight Walburga Meinzinger as the common link between me and all of these matches, since the most recent common ancestral couple isn’t Walburga and her husband, Christoph Meier, but rather Walburga’s grandson, Wenzeslaus Meier and his wife, Anna Goetz. At this point we have no evidence that Walburga is necessarily the ancestor “thru” whom I’m related to these 10 DNA matches, since it’s entirely possible that none of the DNA that we share came from her, but instead came from (for example) the Goetz side.

Finally, let’s take a look at Ancestry’s suggestion of Elisabeth “Lizette” Christina Gross as another one of my potential ancestors. This time, Ancestry informs me that Elisabeth was the mother of my 3x-great-grandfather, Carl Goetz. According to the tree which was supposed to be the source of the information, Elisabeth was born 2 February 1833 in Heilbronn, Baden-Wuerttemberg, Germany. She married Carl Wilhelm Friedrich Goetz, and they were the parents of one Carl Wilhelm Christian Goetz who was born 5 October 1853 in Bavaria (Bayern), Germany and died 19 March 1933 in Buffalo, New York. 

My Carl (or Charles, in English) “coincidentally” also died on 19 March 1933 in Buffalo, New York,2 and equally “coincidentally,” was born on 5 October 1853 in Leuchtenberg, Neustadt an der Waldnaab, Bayern, Germany.3 However, he was the son of Ulrich Goetz and Josephine Zenger, as evidenced by his death certificate (Figure 2).

Figure 2: Death certificate of Charles Goetz (Carl Götz), 19 March 1933, with parents’ names underlined in red.2Charles Goetz death 1933 marked

Moreover, there’s no evidence that my Carl ever used the middle names Wilhelm and Christian. While the birth dates quoted by this tree owner for her Carl Wilhelm Christian Goetz are a match for the documented birth and death dates of my Carl Goetz, the parents’ names and place of birth are clearly not a match. So this tree owner is erroneously conflating my Carl Goetz with her Carl, who may or may not have been the son of parents named Elisabeth “Lizette” Christina Gross and Carl Wilhelm Friedrich Goetz. It’s not clear to me precisely how she came to the conclusion that my Carl belonged in her family tree, beyond indiscriminate borrowing from online trees, but it’s very clear that he does not. Once again, I thought perhaps there was DNA evidence linking me to this tree owner through some other line, that might have been the basis for Ancestry’s identification of Elisabeth “Lizette” Christina Gross as my Potential Ancestor. Once again, I was disappointed. This tree owner isn’t a match to me, or to my father (Carl Goetz’s great-great-grandson).

The point here isn’t that there are inaccurate family trees online; we all know that already. But I think Ancestry’s ThruLines™ tool exacerbates the problem. Since ThruLines™ are accessed through the “DNA” tab and not the “Search” tab, it suggests that the highlighted “Potential Ancestors” are proposed on the basis of DNA matching rather than being based solely on the existence of trees containing individuals with the same names as one’s own ancestors. Unfortunately, in all the cases I examined, the DNA matches were too far “downstream” for them to be useful in drawing any conclusions about my potential relationship to more distant ancestors. The fact that I share DNA segments with my mother, my sister, and my four children cannot be used as evidence of our common descent from someone purported to be my great-great-grandmother. So if these “Potential Ancestors” are being identified solely on the basis of online family trees, then it would be more honest to have them suggested under the “Search” tab rather than the “DNA” tab.

If beginning genealogists are going to use these ThruLines,™ they need to understand that the the “Potential Ancestor” designations are no more reliable than the record hints or “shaky leaf” hints which Ancestry provides. While I love Ancestry for the convenience it offers in allowing me to locate and download documents pertaining to my family online, in the comfort of my home, at 2 am, I do wish they would leave well enough alone. I think it would be much better to put the records online, put the family trees online, and put the DNA data online, and then leave it to genealogists to connect the dots between those data sets themselves.

Perhaps the best thing to do is to hope that, in time, the usefulness of tools like ThruLines™ will increase. There seem to be plenty of people who are raving about this tool in the various Facebook genealogy forums, but so far, my personal experience with it has not been positive. As Blaine Bettinger wrote in the “Genetic Genealogy Tips and Techniques” Facebook group, “As a community, we need to decide whether we want automated tools that will unavoidably perpetuate mistakes, or whether we want NO automation. Those are the only two options.” Call me a Luddite, but I don’t think automation like this is doing us any favors. I look forward to the day when Ancestry proves me wrong.

Sources:

Roman Catholic Church, Młodzieszyn parish (Młodzieszyn, Sochaczew, Mazowieckie, Poland), “Akta stanu cywilnego parafii rzymskokatolickiej w Młodzieszynie,” unknown dates, 1885, marriages, #21, record for Stanisław Zieliński and Marianna Kalota, 15 November 1885.

2 New York, Department of Health, Division of Vital Statistics, County of Erie, City of Buffalo, Death Certificates, 1933, no. 1688, certificate for Charles Goetz, died 19 March 1933.

3 Roman Catholic Church, St. Margaret’s parish (Leuchtenberg, Neustadt an der Waldnaab, Bayern, Germany), Band 6, Taufen 1848 – 1869, p. 26, no. 38, birth record for Karl Götz, Bischöfliches Zentralarchiv Regensburg, St.Petersweg 11-13, D-93047 Regensburg, Germany.

4 Blaine Bettinger, “Genetic Genealogy Tips and Techniques” Facebook group, post on 27 February 2019, (https://www.facebook.com/groups/geneticgenealogytipsandtechniques/permalink/594183891045315/ : 3 March 2019).

© Julie Roberts Szczepankiewicz  2019

11 thoughts on “Ancestry’s New ThruLines Utility Needs More Work

  1. Thank you so much for your detailed explanation of the dangers of accepting any data without critical thought. I understand that mistakes happen, I discovered one of mine just yesterday, but I don’t need help making more.

    Liked by 1 person

  2. I am finding similar assistance and problems. When the trees and DNA acctually match, the pictures are illuminating. It takes discernment to determine when they really do not match. I hesitated working online because I had seen how errors in DAR genealogies were perpetuated when published. Automation accelerates the process. Thanks for explaining it so cogently!

    Liked by 1 person

  3. I had an initial bad experience. In a nutshell they were using someone’s incorrect tree to connect me to a 6cm match. It made a lot of sense and I worked on it for a while thinking a great find. All wrong. They had connected my relation and her parents to their tree. My relation never married, theirs had children. Wrong dob too.
    So be careful using-just like the shaky leaf hints they are suggestions only.
    Good explanation!

    Liked by 1 person

  4. i do not  know how to get around in ancestry’s DNA stuff and this is all a mystery right now. But I do know of mistakes. In several lines or something on Ancestry ‘whateve’r I went on earlier they had me as my father’s daughter but they had me dead!!! I can assure me that I am not. Hope I continue to get better and have more energy and whatever else it takes to get me working on genealogy again and I also want to write my life story. Started it and never finished. Does not tell my  story and no one knows it better than me.Thanks again for all your work on genealogy. BONNIE

    Like

  5. I am relieved that you voiced your experience with the “beta” Thrulines. You pointed out exactly what needs to be done; Thrulines needs to be removed until it’s improved. It could be a search tool but not in the category of DNA speculation. I find it a very sad development because it’s catering to the instant gratification of the commercial non-genealogy DNA testers community. (Magda Maria)

    Like

    1. Yes, I wish I could be more enthusiastic about this. I keep hearing reports that some people have made discoveries this way, but I keep checking back on the new “Potential Ancestors” that have been identified since I wrote this, and none of them are helpful hints that utilize DNA evidence appropriately.

      Like

  6. Good post! It would be nice if Ancestry would create a way to accept or ignore the ThruLines matches much like we already can with the shaky leaf hints.

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.