Ancestry’s new tool, ThruLines has some good features and a lot of potential, but right now, there are a crop of ‘gators in the swimmin’ hole – just waiting for the unwary. Here’s help to safely navigate the waters and not get bit!

Let’s start with first things first.

Maybe it’s wishful thinking or a slip of the keyboard (or phone) but I see people referring to ThruLines as TrueLines, and that’s absolutely NOT true. ThruLines must be verified and proven, just like connections with trees must always be proven – especially when new ancestors are suggested.

ThruLines are not necessarily true lines.

Having said that, there’s lots to unpack here, so let’s get started looking at Ancestry’s latest discovery tool, ThruLines.

It took me a few days to wrap my head around this – meaning what ThruLines is attempting to do, with what information, and why. In other words:

• How is ThruLines supposed to help us?
• How can ThruLines actually help us?
• What are the limitations and dangers?
• How can we avoid the problems?

The difference in the answer depends on your goals. Let’s dissect ThruLines into pieces to see how it can help with genealogy, when and where – along with what to avoid. There are some useful features here alongside some very large neon-flashing danger signs right beside the ‘gator pond.

What is ThruLines?

Ancestry’s blog announcing ThruLines can be seen here.

ThruLines does two things:

• ThruLines groups DNA matching descendants by ancestor. ThruLines shows you, by ancestor, which people match your DNA and claim to be descended from that same ancestor. Notice the words “claim to be.” As always, when other people’s trees are involved, there is a danger of conflated trees and other concerns. We’ll discuss that shortly.

Especially for more closely related relatives, this grouping of DNA matches by ancestor is a great feature because their trees and who they believe they are descended from are more likely to be accurate in the past 3 or 4 generations when the relationship with the ancestor is a result of direct knowledge as opposed to further back where the relationship to an ancestor is a result of genealogical research. This also means the relationship to your match is easier for you to confirm, if you don’t already have your match in your tree. These gathered matches allow you to add family members and cousins to your tree. You never know who might have photos or other information, so matching and discovering that you are connected makes it easy to reach out.

These match groupings use ancestors that you already have in your tree.

• ThruLines provides “suggestions” for potential ancestors by extrapolating from other trees. These “suggestions” include replacing your researched ancestors with other ancestors from other people’s (often inaccurate) trees, many of whom are not DNA matches to you. This is Gatorland!

Ancestry has provided this product announcement but has not yet released a white paper about how ThruLines performs the following:

• Utilizes DNA matches.
• Utilizes trees, including how decisions are made about which trees to use and how to suggest potential ancestors.
• How Ancestry determines when to ignore your existing ancestor in your tree in favor of a suggested replacement.
• The weighting between DNA and tree evidence. In some cases, nonsensical ancestors are being suggested through people’s step-parent’s lines, or ancestors’ “other spouses” lines that the tester is not biologically connected to, so the Ancestry selection process cannot be solely based on DNA matching and in some cases, is clearly not based on DNA matching at all.

Hopefully, a white paper will be coming shortly.

Caveats

If you’re an experienced genealogist, I’m not worried about you. You already understand about ‘gators, meaning the caveats and concerns about the massive number of incorrect trees.

My primary concern is that ThruLines encourages people to believe that ancestors are being suggested because DNA has confirmed that a specific ancestor is theirs. In many cases, erroneous trees have propagated for years, and now all of those people are “wrong together” so their incorrect ancestor is being suggested as an ancestor for many more people. Worse yet, multiple wrong trees are being stitched together by Ancestry in ThruLines.

I wish Ancestry provided a visible warning someplace where users CAN’T MISS IT. MyHeritage does exactly that, even in the name of their similar feature, Theory of Family Relativity.

Buried in the Ancestry support document for ThruLines, I found this:

This should be a required clickthrough before anyone can use ThruLines.

Accessing ThruLines

Not everyone at Ancestry has ThruLines yet. Ancestry has been struggling the past few days and ThruLines have been coming and going. ThruLines is in beta and will be rolled out during the month of March.

For ThruLines to work, you must be sure:

• Your tree is connected to your DNA.
• Your tree is either public or a private searchable tree. Unsearchable trees won’t have ThruLines.
• Your tree is at least 3 or 4 generations deep.
• You only have one kit for any individual person connected to that person in the same tree. If you have multiple kits for the same person connected to one tree, only one kit will have ThruLines. If this is your situation, you can create a “twin” to yourself in your tree and attach the second kit to that person and both kits should get ThruLines. There aren’t many people like me who have tested twice with AncestryDNA, so this shouldn’t be a problem for most people.

You can have multiple kits attached to the same tree, but each kit must be connected to a different person in the tree

If you want to know if ThruLines is available on your account or you are having problems with ThruLines, I wrote about that in the article, “Ancestry’s Disappearing ThruLines – Now You See Them, Now You Don’t.”

Myths and Misconceptions

• ThruLines is NOT telling you or confirming that a specific ancestor IS your ancestor.
• ThruLines is NOT modifying or automatically doing anything to your tree. The ThruLines “trees” you are seeing are constructed for ThruLines.
• You are NOT necessarily related to, nor have DNA matches with the people whose trees are used to suggest potential ancestors.

Warning

For individuals seeking unknown parents, if you see your name beneath a placard that shows a “private” individual in a hashed box above your name, this does NOT mean your parent has been discovered. This only means that Ancestry has “paired” you with a potential ancestor that happens, in your case, to be a parent based on some combination of name similarity (yours) and a person with a similar surname in someone else’s tree.

If you have a parent/child match, it will be the first match on your match list. Look there.

It’s heartbreaking to get excited only to learn that the ThruLines “potential parent” shown for you has nothing to do with genetics – so be prepared for this possibility and don’t get excited. Check your DNA match list.

Facts

• According to Crista Cowan in her RootsTech video, ThruLines is a replacement for Circles. After working with both my ThruLines and Circles, it became obvious very quickly that Circles have not been updated for some time, probably months. If you have Circles or New Ancestor Discoveries (which have been dormant for the past couple years), please archive them so you don’t lose any information you might have. I wrote about how to do that here.
• New Ancestor Discoveries were discontinued roughly two years ago, so you may not have any.
• Shared Ancestor Hints (green leaves) is no longer working. Neither is Filter By Common ancestors. I sure hope they fix both of these (probably related) bugs.
• Any ThruLine card in a dotted edge box is a POTENTIAL ANCESTOR and is very likely incorrect.
• Any ThruLine card in a solid line box when you mouse over the card is an ancestor currently found in your tree. It’s as correct as your tree.

• Just because you have an ancestor in your tree does NOT mean that Ancestry will use that ancestor. Ancestry may provide another potential ancestor from someone else’s tree. Watch for the hashed lines and be cognizant of who is already in your tree, and why!
• On the ThruLine trees, any person, meaning ancestor OR other person in a grey hashed line box is only a suggestion based on someone else’s tree or multiple trees.
• ThruLines shows you which tree that ancestor was “suggested” from, allowing you to click through to that tree and view their documentation.
• On the ThruLine trees, any person in a solid edge box is from your own tree, shown with red arrows below, while any suggested individual is shown with hashed edges, shown with green arrows.

• The summary below the ancestor’s name may indicate that you’re related to XYZ ancestor in <some number> of ways, but review the people you match very closely because you may be related to them, but not through the ancestor or in the way shown.

• If there was a second marriage, ThruLines may be attributing your relationship to the un-related spouse. The descendants shown may be from the “other” marriage and that person’s ancestors. If you’re thinking the unrelated spouse’s ancestors can’t be genetic, you’re right – at least not through that line. Be very careful. You’re in ‘gator territory.
• Furthermore, Ancestry may be suggesting ancestors of the “step-spouse” or other tangential line as well. More ‘gators.
• Just because you match 5 descendants of XYZ ancestor, that does not mean any of these people match each other. In fact, you may match some of these people through another line entirely.
• You can still click through to view the DNA comparison feature at Ancestry. However, since the Shared Ancestor Hints (green leaves) is not working at all, you will not be able to see the side-by-side tree comparison feature☹

• Using Shared Matches on the comparison page, you may be able to determine if some of these individuals do in fact match each other which helps to increase the likelihood of common ancestry in a specific line.

• ThruLines does not replace Shared Ancestor Hints (green leaves) although ThruLines organizes the Shared Ancestor Hints by ancestor. Currently Shared Ancestor Hints is not working and says you have no matches with shared ancestors which is clearly incorrect if you previously had any Shared Ancestor Hints.
• ThruLines may “add” projected ancestors to matches whose trees don’t reach far enough back in time, but who connect with another tree who connects with your tree. This occurs in ThruLines, not in your own tree unless you specifically add the information there. This additive “tree extension” effort is very similar to the WeRelate application which was infamously wrong – more like WeDontRelate.
• You will still receive potential parent hints for ancestors on your actual tree, some of which will (or may) also be reflected in ThruLines. In some cases, the hint on your tree and the ThruLine suggested “potential ancestor” are different, so check both places.

• ThruLines only reaches back 7 generations, so if you’re looking for a breakthrough or descendants from ancestors before that time, you won’t see them in ThruLines. Previously in this regard in relationship to Circles, Ancestry had said that there were too many unknown family lines and multiple relationship paths that far back in time.
• Many of the ancestor ThruLines share exactly the same descendants. For example, my Dutch line only has a limited number of testers, so the same 10 people are listed for generation after generation going up in the pedigree (back in time.)
• Once ThruLines offers a potential ancestor, they continue offering parents and grandparents of that potential ancestor until they run out of ancestors, reach the 8^th generation or some other criteria for stopping. If this is a legitimate line, great – and if not, it’s a royal pain without an “off” button to reject known erroneous “potential ancestors.” “Gators having baby ‘gators!
• There is no way to “ignore,” “reject” or tell Ancestry to “disconnect” or remove a potential ancestor. I hope they will add this feature soon. This could be useful if they suggest another ancestor, especially in lines where you are truly at a brick wall. New “potential ancestors” would provide you with ideas for who/where to search.
• There is also no way to mark a ThruLines card as “seen” so you don’t review it again.
• Having two kits connected to the same person in your tree will (at the time of this writing) prevent one of those kits from receiving ThruLines. In my case, I took a V1 and a V2 test and had them both connected to my own record. This can be solved by adding yourself as a sibling and connecting one kit to the sibling.
• ThruLines is free for everyone for now but may require a subscription in the future. (Don’t forget about the Insight subscription to access DNA features only, assuming it still exists, but you must call support to obtain that limited subscription.)

Suspicions

• That trees with “more” sources are weighted more heavily than trees without “more” sources. Case in point is my own tree for an ancestor who little is known about, so I only had an estimated birth and death year. However, Ancestry suggested a replacement with a very robust but nonsensical tree that incorporates “lots” of documentation. In fact, this amazing woman has birth records from Washington DC 1830-1955 (which didn’t exist in the 1700s when this woman lived), New Hampshire 1714-1904, PA 1669-2013 and who died in TN but is buried in PA. Any modicum of logic would have immediately down-weighted this tree’s veracity.
• That not all relationships are genetically based. For example, I have 15 ancestors for whom I’m the only DNA tester listed, legitimately, so why are those ancestors shown as a ThruLine for me? No DNA is involved.
• That Ancestry created, updated or still has a version of that “One World Tree” someplace, because some of this information is drawn from old trees with information removed long (as in years) ago. Does Ancestry know where this information was obtained from customer trees, and how they selected the specific tree to use? Do they update it? How often and what logic decides what is incorporated into that tree? I notice that in some cases, suggested ancestors’ spouses came from different users’ trees, even when the same spouse was in the same tree.
• That projected ancestors are entirely tree based only, not genetically based.
• Although Ancestry has not yet told us how they weight tree matches, it stands to reason that the most complete records are the most likely to be matched successfully – so be sure your ancestor’s records are as complete as possible.
• Having said that, some of the suggested potential ancestors replacing my existing ancestors have much LESS documentation than my own tree, including some with the only “source” being Ancestry trees. I have no idea what Ancestry is actually doing, when, or why.

Accessing ThruLines

Sign on to your account and click on DNA and then “Your DNA Results Summary.”

Click on “Extras” and then “Ancestry Lab.”

Once there, enable the beta functions. I’m not positive you need to do this for ThruLines, but there have been so many issues that I’d recommend doing this, just in case.

If you have ThruLines available on your account, you’ll see this on your DNA Summary page.

Click on the green box to access ThruLines.

Please note that as of this writing, ThruLines is not stable, meaning that ThruLines and ancestors, as well as matches are tending to come and go. Some features are working sporadically and some not at all. The Shared Ancestor Hints and Common Ancestors filter is not working at all, even when ThruLines is functioning.

Sorting ThruLines

After working with ThruLines, I discovered, for me, working the different types of records together was easiest, because what I do with those records differs.

Records fall into the following categories:

• Existing ancestors in your tree
• Potential (suggested) ancestors

In fact, Ancestry provides the ability to filter in exactly that fashion, at the top of the ThruLines page on the left side.

Let’s look at these two types of records individually, because I use them differently.

In case you’re wondering how I track my ThruLines, I created a spreadsheet that includes columns for:

• Number – numbered so that I know I’ve accounted for all 254 ancestors through 7 generations
• Generation in which that ancestor is found – for example my 4 grandparents are generation 2
• Surname of Ancestor
• First name of Ancestor with birth year if multiple people by that same surname
• “Should Be” column for when Ancestry suggests a replacement ancestor for a correct Ancestor I have in my tree. “Should be” is the correct ancestor’s name.
• Existing – meaning does the ancestor exist in my tree already and is the ThruLines card provided by Ancestry for this existing ancestor from my tree
• Potential – meaning is this a “potential ancestor” as indicated by a hashed line. Believe it or not, I have several cases where I have an ancestor by a specific name in my tree and Ancestry has suggested a different ancestor by the same name that is a conflated version of my ancestor and another person by that same name, so the answer can be yes for both “existing” and “potential.”
• Members – how many people match in this ancestor’s group. The number of matching people is easy to see at the top of the ThruLine card when you click on the Ancestor card to open.

• Comments – anything that comes to mind such as why the suggested ancestor is wrong, something to look at, if they are a Y or mtDNA candidate, etc.

Here’s an example of my spreadsheet.

I also color coded the ThruLines ancestors according to the groups identified later in this article so that I could filter by color. In the example above, the grey entry is an example of a correctly gathered ancestor and the red entry is a hypothetical example of an incorrect entry. If an incorrect person was listed, I would enter the correct ancestor in the “should be” column.

I had to create this spreadsheet to wrap my head around what Ancestry was doing with ThruLines, and to some extent, perhaps determine why.

I would suggest that you read through this entire article before deciding how to handle your ThruLines, then come back and create this spreadsheet if you want to. I had to create a spreadsheet to wrap my head around what was going on with ThruLines so I could write this article.

Existing Ancestors

ThruLines only “sees” ancestors on your linked tree. That means the tree linked to your DNA results.

If you switch trees, you’ll have to give Ancestry some amount of time to switch your results to the new tree. No, I don’t know how long that is in actuality. Hours to days. Ancestry suggests two days. Many people are reporting much longer waits.

What you’ll see when the process is complete is a very nicely organized set of “ancestor cards” that begins with the closest ancestors you have in your linked tree.

Linking Your Tree

If your tree is not linked, PLEASE LINK IT. You will not have ThruLines if your tree is not linked correctly.

To link your tree, click on “DNA” at the top of the page, and then on DNA Summary.

Click on the Settings gear in the upper right corner of the page.

Be sure you are participating in matching and then link your tree in this section:

Exploring ThruLines

I have 161 individual ancestors listed on cards at Ancestry, along with 49 potential ancestors, although this number varies from hour to hour and day to day.

Existing ancestors in your tree have a solid line around their card when your cursor is above the ancestor.

Potential ancestors have a hashed line around their card.

This might be a good time to mention that Ancestry includes information from searchable but private trees. This means that information from many of those private trees that so frustrate genealogists is included. It may also mean that trees people are using as “quick and dirty trees” and they forgot to make unsearchable are included too.

However, if you have a private searchable tree, this now means that you too will have ThruLines.

As frustrating as these “private” cards appear at first glance, they actually aren’t useless. I clicked on this private placard and look what I see.

This potential ancestor happens to be inaccurate, but at least I can see something.

Sometimes you’ll see this instead (even if the person lived so long ago that they can’t possibly be living), or if you’re lucky, the following which at least provides the name of the suggested ancestor so you can search elsewhere.

I’m very grateful for this change to provide the ability to at least identify the ancestor being referenced.

Looking at my Ancestry tree, to the 8^th generation (meaning 7 generations, inclusive, counting from my mother, means that if you expand your tree once, every ancestor other than the last column should be shown on a ThruLines card as illustrated below.

In 7 generations, there are a total of 254 ancestors, counting our parents as generation 1.

Let’s break my 254 ancestors down into categories based on the Ancestry ThruLines.

Group 1 – Ancestors with No DNA Matches

Based on the fact that I’m the only child that has tested as a descendant of my mother, and she has a card, Ancestry appears to have taken every one of those 254 individuals and processed them in some fashion. I say this because I have a total of 20 ancestors in my tree with whom there are no DNA matches attributable to that ancestor.

In fact, it’s this line of relatively recent German immigrants, the parents having arrived in the mid-1850s. Jacob Kirsch and Barbara Drechsel (Drexler) didn’t have a lot of children and many of those children didn’t marry and have children, which leaves a small descendant pool to test.

Clearly, based on this, the ThruLines, meaning the cards shown, aren’t generated based on a DNA match. That’s fine, except that I understood that a ThruLines card meant that you HAD a DNA match first, then secondarily a matching tree as well.

Obviously, that’s a misconception.

I’ll be keeping a running scorecard of my 254 ancestors and how they break down in ThruLines.

Ancestors	Number	Comments
Total ancestors in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Remainder	234

Group 2 – Missing Ancestors Altogether

This next group is probably the easiest to account for, because they are missing in the Ancestry ThruLines cards altogether. They are clearly in my tree, but they have no ThruLines card showing that they exist. If they were only in the 7^th generation, I could understand that they are missing AND don’t have hints about ancestors in earlier generations – because Ancestry (unfortunately) doesn’t provide anything in the 8^th generation, but that’s not the case here. Two full generations are missing entirely.

This entire branch of my mother’s tree is missing altogether – both parents and all 4 grandparentts of Angenietje Houtsma.

It’s clearly NOT because there aren’t any DNA matches, because the Kirsch branch in the last example has no matches and still has ThruLines cards for ancestors.

It’s not because there aren’t parents, because Angenietje Houtsma has grandparents who should have cards as well, AND, those grandparents have record hints. So, it’s not like these people are unknown to the system, because they aren’t. In this one line alone, 6 ancestors are missing.

In the 6^th generation, I have a total of 4 missing ancestors who are in my tree but have no cards, and in the 7^th generation, 10, for a total of 14 missing ancestors. Where are these ancestors and why don’t they have a ThruLines card?

I have no idea.

Ancestors	Number	Comments
Total ancestors in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14
Remainder	220

Group 3 – Ancestors in My Tree with Gathered Descendants

This next group is the largest group of matches after eliminating the missing ancestors and those with no DNA matches.

This group consists of ancestors who have cards from my tree shown by Ancestry AND with whom I have DNA matches attributed to that line.

Keep in mind that many more people may have DNA tested and are descended from these ancestors, but their DNA doesn’t match my DNA. The only resource available to see that those people match others descended from that ancestor is if you have a Circle for that ancestor, you can check for people NOT shown in this ThruLine grouping.

Ancestry has stated that they are not going to continue to add to the Circles, so if you want that information, archive it now. I wrote about how to do that here.

I will be doing that for every ancestor with a Circle.

Let’s look at Lazarus Estes. He’s my great-grandfather and I know of most of his descendants, or at least I think I do. I have 6 DNA matches that descend from Lazarus.

Ok, maybe I don’t know most of his descendants. I know most of his descendants a generation ago. One of these names I’ve never heard of. The good news is that they might have information that I don’t. Pictures, stories, something.

If your goal is to connect with LIVING people, you’ll love this ThruLines feature.

In recent or relatively close generations, people are likely to know their genealogy which means their parents and grandparents. For example, I don’t question for a minute that the three descendants of Gracie Estes Long know that she’s their grandmother. I would hope that Tyler knows that my half-sister is his great-grandmother, but I suspect he has no idea who I am. His mother and grandfather are still living, which is why they are marked as private and have hashed lines, so he could ask them and I’m sure they know both who Edna was and who I am.

As you move further back in time where people are depending on historical research, that’s when the trees become more problematic, entering ‘gator territory, because they adopt and incorporate other people’s trees, believing them to be accurate.

One point that this graphic illustrates quite well is the difference in inherited DNA in the green boxes. Note that with my three 2C1R (second cousins once removed), I share 170cM, 161 cM and a paltry 25cM of DNA with them. That’s a very large difference. Then note that I share LESS with my half 2^nd great-nephew, with whom I’m more closely related than with two of my 2C1R. Roll of the genetic dice.

You might notice that I can’t drop down the middle box because there’s not enough space for all 6 matches to show simultaneously. Sometimes you have to scroll back and forth to see the entire graphic, including all the siblings, so you can click at the top on the “List” link to see the people you match who descend from this ancestor in a list format.

There are three additional pieces of information available from this “List” screen.

If you click on “View Relationship,” it takes you back to the tree where you will see only your relationship with that person.

Notice that the solid lines mean these people are in my tree, but there’s another hint too. You can see that Becky’s father was taken from her own tree, but her grandmother, Lucy was taken from someone else’s tree. Is that accurate information? Don’t ever assume that it is. The trees are all hotlinks. Verify, verify, verify!!!

If you click on the person’s initial box or name, you’ll be taken to the DNA comparison screen that we’re all familiar with. Be sure to note how you’re related so you can check easily.

This confirms that Becky didn’t provide any more information than her parents in her tree.

If you click on the segment information in the middle of the “List” screen, you will see the following:

Please note that these percentages do not correlate with the DNAPainter tool here which I use extensively. Ancestry does remove segments that they feel are “too matchy.”

There’s a pretty large difference between 40% and almost 52%. I wonder if Ancestry is a victim of their own incorrect trees where relationships are reported inaccurately. If that’s how they are calculating these statistics, it could well explain the discrepancy.

I would think that genealogists who care enough to make the effort to enter their DNA information into Blaine Bettinger’s Shared cM Project, from which the DNAPainter tool is derived would care enough to make sure the relationships reported are accurate. You can contribute to this crowd-sourced effort here.

I have a total of 148 “Ancestor Gathered Descendant Trees.” I know for a fact that not all of them are accurate for any number of reasons, but what I do know is:

• That my path to the ancestor is accurate because it’s my tree and I’ve spent 40 years performing original research and documenting those ancestors.
• That I’m somehow related to these people, assuming that the segment is not identical by chance.
• The identical by chance scenario can be lessened for each match by looking at the Shared Matches for hints based on other people that also descend from the same ancestor.

Checking my match with cousin Beverly to help eliminate the identical by chance scenario, I discover that I do have shared matches with her, and that two of the closest common matches are people I recognize. Becky from my example above and another cousin I know well – both who descend from the same lines and help confirm the legitimacy of Beverly’s match.

Ancestors	Number	Comments
Total ancestors in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14	No ancestor cards at all
Ancestors from my Tree with Gathered Descendants	149	My ancestor is accurate. Ancestor of matches may or may not be accurate.
Remainder	71

Group 4 – Ancestors with Unknown Parents But No ThruLine

These are the individuals I was truly hoping would have a potential ancestor.

With one exception, all of these 9 ancestors are females with no surnames. In the one case where the ancestor is a male, the potential father is incorrect and no mother is offered. Based on the other mothers offered connected to incorrect fathers, the mother would be the wife of the incorrect father.

Ancestors	Number	Comments
Total ancestors in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14	No ancestor cards at all
Ancestors from my Tree with Gathered Descendants	149	My ancestor is accurate. Ancestor of matches may or may not be accurate.
Ancestors with Unknown Parents	9	Generally, missing parents of females with no surnames and no potential parents offered.
Remainder	62

Group 5 – Ancestors Shown as Potential Ancestors are Already in Tree

In 3 cases, I have Potential Ancestor cards for the same exact person that is listed in my tree already, with much the same information, making me wonder why mine was ignored and the other offered as a replacement.

The good news is that the other person’s tree from where these suggestions arose looks to be quite well documented, so I look forward to contacting them and researching what they have attached.

Ancestors	Number	Comments
Total in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14	No ancestor cards at all
Ancestors from my Tree with Gathered Descendants	149	My ancestor is accurate. Ancestor of matches may or may not be accurate
Ancestors with Unknown Parents	9	Generally, missing parents of females with no surnames and no potential parents offered.
Potential Ancestors Already in Tree	5
Remainder	57

Group 6 – Possibly Accurate Potential Ancestors

Only two Potential Ancestors are possibly accurate. Both of these individuals are the parents of a known and proven ancestor. A cousin has done some research on this line and eliminated a number of candidates, but I need to work with her to research further to determine if the suggested couple has been researched or eliminated.

Ancestors	Number	Comments
Total in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14	No ancestor cards at all
Ancestors from my Tree with Gathered Descendants	149	My ancestor is accurate. Ancestor of matches may or may not be accurate
Ancestors with Unknown Parents	9	Generally, missing parents of females with no surnames and no potential parents offered.
Potential Ancestors Already in Tree	5
Possibly Accurate Potential Ancestors	2
Remainder	55

Group 7 – Inaccurate Potential Ancestors (‘Gator City)

I saved this group for last because it’s just painful. As a genealogist, I have to say that truthfully, the fact that Ancestry has suggested 55 ancestors that I know positively are inaccurate terrifies me for the sheer fact that less experienced genealogists will grab gleefully onto these “new ancestors” and perpetuate the Ancestry-provided incorrect trees like kudzu vines. The perception is that these trees are now “proven” by DNA – a statement I’ve seen repeatedly the past several days.

THESE ANCESTORS AND TREES ARE **NOT** PROVEN BY DNA!!!

These trees are predicated upon other people’s inaccurate trees with suggestions being made to replace your ancestors, currently in your trees, with other ancestors from other people’s trees. There seems to be no consistent logic that applies in ‘Gatorland.

The end result will be that even more people will receive inaccurate “Potential Ancestors” because there are yet more of those incorrect trees skewing the algorithm. I don’t know if the criteria for ancestor suggestion is “most numerous” tree or something else. This scenario is the very definition of a vicious ‘gator circle.

The incredibly frustrating aspect of ThruLines for me is when Ancestry ignores the ancestor in my tree that I’ve spent years (if not decades) researching and documenting, and instead suggests a “Potential Ancestor” that defies logic. LHM!

In some cases, such as with Lydia Brown, wife of William Crumley III, the widely disseminated Elizabeth Johnson is shown as the mother of my ancestor, Phoebe Crumley, instead of Lydia Brown. Not only is Elizabeth Johnson incorrect, it’s been proven incorrect for several years now via mitochondrial DNA, AND, I’ve written and published about this case.

Imagine my frustration, to put it mildly, to discover that Ancestry is now ignoring my carefully proven ancestor and suggesting that I replace her with the unproven, erroneous ‘gator. Not only that, but I fully suspect that my tree is NEVER going to be suggested to others, because it’s a (nearly) lone voice lost in the forest of incorrect ancestors.

Truthfully, this makes my blood boil – 55 separate painful times. Once for each incorrect ‘gator masquerading as an ancestor. Why would Ancestry think that replacing my ancestors in my tree with ones from other people’s trees is even remotely a good idea?

To suggest that I “consider” a different ancestor or information in another tree is vastly different than simply ignoring the ancestor I have in my tree and providing a “Potential Ancestor” replacement, like the one in my tree doesn’t even exist. (Not to mention that this attitude in and of itself is both arrogant and condescending.)

If I should consider a different ancestor, I’d at least like for that ancestor to have lived in remotely the right place and time. People did not have children at age 5 nor after they died (except occasionally for men within 9 months), nor did they have 30 children, nor were they married, having children and living in multiple places at the same time. Well, at least not most ancestors😊

The Quality of a potential tree should be part of the recommendation factor, especially if Ancestry is blithely ignoring my existing ancestor in favor of another potential ancestor from someone else’s tree.

Simply telling you how wrong these suggested Potential Ancestors are would not do the situation justice. I’ve documented the circumstances, briefly, with the hope that you will use my experience trying to sort through this ‘gator swamp as a warning for what to look for and consider in your own ThruLines, and how.

What’s worse, when Ancestry ignores your existing ancestor and suggests others, they don’t stop with that one ancestor, but then continue to suggest and propagate ancestors on up your tree for the erroneously suggested ancestors. These recommendations are not based on DNA or your existing ancestor in your tree but on “those other” trees alone.

Let’s look at an example of what Ancestry “suggested” for my Crumley branch. The red Xs document where Ancestry replaced a known ancestor with suggested incorrect ancestors – including on up the tree. (I should have used little ‘gators instead of Xs.)

Unknown H2a1 is an unknown wife of William Crumley II with the H2a1 mitochondrial haplogroup. Ancestry did not assign a potential ancestor for her, but obviously Ancestry “believes” that she was a Johnson, because her father is suggested as Andrew Johnson. Of course, this means that H2a1’s mother is incorrectly “suggested” as well as Andrew Johnson’s wife.

I know this is wrong, because Elizabeth Johnson was a second wife who married William Crumley in 1817, long after his son, my ancestor, William Crumley III was born in 1785. Therefore, it’s impossible that Elizabeth was William III’s mother. Not only that, she was 12 years YOUNGER than William Crumley III. Twelve years younger than her step-son.

Furthermore, Lydia Brown, the proven mother of Phebe Crumley through William Crumley III in the next generation, was ignored as well, and his wife was also given as Elizabeth Johnson through a different tree. This Elizabeth Johnson’s parents were assigned as different parents than the Elizabeth Johnson who married William Crumley II in 1817. Are you confused yet? Believe me, so was I and obviously, so are other people as well as Ancestry.

The end result of this is that not only were my existing ancestors ignored and replaced, but the erroneous trees that are themselves full of impossibilities for the person they are documenting are then suggested to replace mine. Those trees are then cobbled together by Ancestry in a Frankenstein mosaic of patched together ancestors that are blatantly wrong and very difficult to unravel.

And this in only one branch of my extended tree. This scenario happened on multiple branches. If you’re thinking to yourself, “How bad can this really be?”, here’s the graphic of every branch affected, and how.

That old “picture is worth 1000 words” thing.

If you think I’m overreacting, take a look at these graphics which do NOT include missing ancestors or the two that that might potentially be accurate – only the “Potential Ancestors” provided by Ancestry that I know positively are inaccurate.

The red Xs show where my ancestors have been ignored and alternative incorrect ancestors suggested as “potential ancestors.”

My tree is literally bleeding red Xs. And I just realized while proofing that there are now more than there were initially, and I missed one X. Sigh. The ‘gators are “propagatoring.”

If your jaw just dropped looking at those red Xs, let this serve as a warning for your own tree.

Below, brief descriptions of what is wrong, and how. Think of this as the ‘gator trap.

Ancestor	Suggestion to Replace Accurate with Inaccurate Ancestor
Joseph Bolton	My ancestor ignored and suggested similar Joseph Bolton from tree with significantly less information than mine.
Lydia Brown	Proven incorrect ancestor based on widely spread speculative misinformation.
Elisabeth Mehlheimer’s mother	I already have her mother, Elisabetha Mehlheimer, in my tree. Why suggest a “private” person instead? (This has since disappeared.)
William Moore	Replaced my William with another William Moore proven via Y DNA not to be from the same line. The William they suggest has 30 children from 3 “wives” who are obviously the same woman by different variations of a common surname, with many “children” by the same name. This tree has obviously been constructed by indiscriminately “gathering” from other trees. Yet, that tree, according to Ancestry, trumped my own carefully curated tree.
Lucy, wife of William Moore	Suggested wife of yet a different wrong William Moore, above, ignoring Lucy in my tree. Birth shown in Giles County, VA but also with an attached England birth document. Shows marriage to two different William Moores, at the same time, neither one where mine lived.
Daniel Vannoy	Suggested his brother, Francis, ignoring Daniel in my tree. The Francis tree has many spurious references to IGI records, Family Data Collection and Ancestry trees, but does NOT include my Elijah as his child, so how Ancestry decided to make this connection is baffling.
Sarah Hickerson	Suggested Daniel’s wife from a different tree than above where Elijah is included as a child. Which Vannoy brother fathered Elijah Vannoy was proven through DNA matches to the Hickerson family, not Millicent Henderson, wife of Francis Vannoy.
Jotham Brown	Because the right wife, Lydia Brown, was ignored, the wrong line continued to be suggested upwards in the tree instead of Jotham Brown, Lydia’s proven father. This “private” tree is for Zopher Johnson, as shown by his connected children even though “Zopher” himself is private. Another Potential Ancestor shows Zopher as a card a generation earlier, along with wife Elizabeth Williamson Cooper, perpetuating this wrong information up the tree another generation.
Phoebe Cole	Because the right wife, Lydia Brown, ignored, the wrong line continued to be suggested upwards in the tree instead of Phebe Cole, the wife of Jotham Brown, Lydia’s proven father in my tree. This “private” ancestor is for Catherine Harrison, wife of Zopher Johnson.
James Mann	Substituted James Robert Mann, the wrong person, ignoring the accurate person. This continues upstream for 2 more generations.
Mary Cantrell	Substituted Mary Jane Wilson, wife of James Robert Mann. The wrong line continued up the tree for 2 more generations.
Michael McDowell	Suggested replacing my Michael with a different, less correct, possibly conflated, Michael McDowell.
Samuel Muncy	Suggested Samuel Munsey-Muncy, ignoring my Samuel, from a tree that shows a Civil War service record for a man who died in the early 1800s and would have been about 100 during the Civil War. Miraculous! Lots of Family Data Collection and Ancestry Trees sources.
Andrew McKee	Suggested replacing my Andrew with a different, incorrect “private” Andrew McKee. The original tree for Andrew has now been made completely private.
Elizabeth, wife of Andrew McKee	Ignored my Elizabeth and suggested replacing with another Elizabeth. The matches look to be correct, so the other tree has the two Elizabeths conflated. The only source for the replacement tree is “Ancestry Family Trees.” Sheesh, Ancestry.
James Moore	Ignored my proven ancestor and suggested replacing him with a William Moore proven via Y DNA not to be of the same line. This William in the person’s tree was said to be born in Henrico Co., VA, but has an attached birth record from England. Can’t be both.
Mary Rice	Ignored Mary Rice and suggested Margaret Hudspeth, wife of the incorrect William Moore, above. Again, shown to be born in Henrico County, with English birth record attached along with IGI record as only sources.
Charles Hickerson	Since Sarah Hickerson was ignored, the incorrect family line was perpetuated up the tree with a wrong ancestor for the second generation instead of Charles Hickerson.
Mary Lytle	Since Sarah Hickerson was ignored, the incorrect family line was perpetuated up the tree with a wrong ancestor instead of Mary Lytle.
Sarah Rash	Ignored mine and suggested replacing with a Sarah Rash that appears to be more fleshed out, but the dates are all based on records not belonging to the correct Sarah, including a birth in England despite a shown birth date in 1732 in Spotsylvania Co., VA. We have our Sarah’s birth from the family Bible.
William Moore’s wife, Lucy’s father	Nancy Moore’s erroneously given mother’s supposed father, Samuel Little Harwell. This erroneous tree now perpetuated to the second generation.
William Moore’s wife, Lucy’s mother	Nancy Moore’s erroneously given mother’s supposed mother, Anne Jackson. Woman who died in 1765 has a SSDA claim record and an 1800s immigration record, even though she was supposedly born in Brunswick Co, VA in 1712. Married Samuel Harwell 5 different times – clearly has simply collected and attached data to their tree without evaluation, but Ancestry thought it was “better” than my tree.
Jotham Brown’s father	Because Lydia Brown was misidentified, so was Jotham Brown, and now his father as well, perpetuating garbage up the tree for two more generations
Jotham Brown’s mother	Because Lydia Brown was misidentified, so was Jotham Brown, and now his mother as well, perpetuating garbage up the tree for two more generations
John Cole	Since Phebe Cole was misidentified, so was her father.
Mary Mercy Kent	Since Phebe Cole was misidentified, so was her mother
Michael McDowell Sr.	Since the wrong Michael was identified earlier, a wrong father, John McDowell, was also identified, proven by Y DNA not to be related. The son of this John McDowell is yet a different John McDowell than the one Ancestry substituted for my John McDowell.
Wife of Michael McDowell Sr.	We don’t know who she was, but we know she wasn’t Magdalena Woods, married to John McDowell. The Magdalena Woods tree they recommended includes 13 Ancestry Family Tree, Family Data Collection and IGI records as sources, plus a German birth record for a person born (supposedly) in 1705 in Ireland, according to the tree itself. Pretty tricky!
Isabel, wife of Michael McDowell Jr.	Suggested private father, Ebenezer Hall, and erroneous mother, Dorcas Abbott who lived her entire life in New Hampshire, not Virginia.
James Claxton’s parents	Suggested erroneous private father and mother, widely distributed but proven via Y DNA testing not to be of the same line. Ironically, there are other trees for this person that are not private. How and why Ancestry selected the private one is a mystery. Ancestry also suggested his mother, from another tree as Catherine Kathryn Caty Middleton which is incorrect as well.
Joel Cook	Ignored Joel, proven ancestor via military records to suggest Henry Cook. Henry’s only daughter, Sarah, in the suggested tree would have been 5 years old when she married and 6 when her first child was born. Another miracle!
Alice, wife of Joel Cook	Ignored Alice to suggest wife of Henry Cook from a different tree with the only source being Ancestry Family Trees.
Samuel Muncy	Ignored Samuel to suggest Obediah Muncy.
Agnes Craven	Ignored, Agnes, wife of Samuel Muncy to suggest “private” person who was the wife of Obediah listed as “Mrs. Obediah.”
Andrew McKee	Ignored Andrew McKee who is proven to suggest Hugh McKee who did not live in the correct state to have his child. The next generation up the tree is also incorrect, suggesting a George McKee who had Elizabeth Barnes, mother of Ann McKee, according to Ancestry. Ann McKee’s mother was actually Martha (probably) McCamm. Another potential ancestor card suggests George McKee’s wife is Anna Elizabeth Carney with no sources at all.
Martha possibly McCamm	Ignored Martha and suggested wife of Hugh McKee, Mary Nesbit, perpetuating erroneous information for another generation up the tree. Some very convoluted trees in this mess with only source being “Ancestry trees.” This has now been made entirely private.
Elizabeth, wife of Steven Ulrich.	Ignored my Elizabeth and suggested Elizabeth Cripe, a surname/person that has been disproven but rampant in trees.
Marie LePrince	Ignored proven genealogy that Marie LePrince is the mother of Marguerite de Forest and replaced with Marie Claire Rivet. Shown marrying both in France and in Nova Scotia.
Hannah Drew	Ignored my Hannah Drew and replaced with wrong Hannah Drew from another tree with less information, showing only an Ancestry Family Tree and a record showing the birth of her son in England, while the rest of the tree shows his birth in the colonies.
Samuel Mitchell	Ignored my Samuel Mitchell and replaced with an erroneous Samuel Mitchell who was supposed to have died in 1756 in Maine, but has a Michigan death record from after 1867. I had no idea people lived to be almost 200 years old. Wow!
Elizabeth, wife of Samuel Mitchell	Ignored my Elizabeth and suggested Elizabeth Penglase as the mother of Catherine Mitchell. Elizabeth Penglase was born in 1698 in Kittery Maine, but with at London England birth/baptism attached to the record. Also shows her father as Christopher Mitchell and two marriages to her husband. Tree is very confused and conflated.
Susanna Koob	Ignored Susanna who is proven and suggested replacement with Anna Margaretha Kirsch with only source being “Ancestry Family Trees.” No DNA matches, so how was this done, exactly?
John Herrell	Ignored my John and substituted John Isaac Herrell, incorrectly from a different location, with no documentation.
John Herrell’s father	Because John Herrell was incorrectly substituted, so is his father, shown as Davie Harrell.
Francois LaFaille’s father	Suggested Jean Francois LaBelle who did not live in the right location to be the parent of Francoise LaFaille, nor is the surname correct.
Francois LaFaille’s mother	Suggested Marie Genevieve Auger Baron who as the mother of Francois LaFaille, who was the wife of Jean Francois LaBelle.
Jacques De Foret	Ignored proven ancestor, Jacques de Foret and replaced with Bonaventure Foret who did not live where my ancestor lived.

Evaluation

If you are looking for close cousins and know your tree well, you may well find some individuals with whom to collaborate based on the grouping of DNA matching descendants by ancestor. Perhaps you’ll be fortunate and discover previously unknown family photos, history or stories.

The further back in time ThruLines reaches, the more problematic the frankentrees that are unrelated to DNA become.

The suggested “replacements” of known, proven, ancestors with incorrect ancestors are found in my tree as follows:

• 1 individual at generation 4
• 2 individuals at generation 5
• 12 individuals at generation 6
• 40 individuals at generation 7

Here’s the final scorekeeping chart.

Ancestors	Number	Comments
Total ancestors in 7 generations	254
Ancestors with no DNA matches	20	German immigrant line
Missing Ancestors	14	No ancestor cards at all
Ancestors from my Tree with Gathered Descendants	149	My ancestor is accurate. Ancestor of matches may or may not be accurate
Ancestors with Unknown Parents	9	Generally, missing parents of females with no surnames and no potential parents offered.
Potential Ancestors Already in Tree	5
Possibly Accurate Potential Ancestors	2
Correct Ancestors Replaced by Incorrect Potential Ancestors	55

By any measure, this is an abysmal report card relative to “Potential Ancestors,” with only 2 potential new ancestors that could be accurate and 55 wildfires that can never be extinguished – with gasoline thrown on, encouraged and propagated by Ancestry themselves. What a terrible example of stewardship. This is not just a disservice to me, but to the entire genealogy community. We should be striving for accuracy, not feeding the ‘gators and fertilizing Kudzu vines frankentrees.

Goals and Benefits

My goal with genetic genealogy, and genealogy as a whole, is fourfold, as shown in the following chart. The questions is, how does ThruLines help me achieve these goals?

Goal	ThruLines
To confirm known ancestors through DNA	This is best achieved by segment matching which Ancestry does not provide, but less conclusive evidence can certainly be obtained through close matches and shared matches that match both me and close family members. Unfortunately, Circles which is a form of genetic networks would provide additional confirmation but is being discontinued.
To document the lives of my ancestors accurately for future generations	ThruLines encourages the propagation of erroneous trees by suggesting them, by linking to them, and by failing to use any discernible quality measure. A quality tree is NOT a tree with conflicting sources about the same event, the same timeframe or unreliable sources such as “Ancestry Family Trees.” We, as individuals, can’t put these fires out as fast as Ancestry flames them, especially if quality trees are discounted for larger “scavenged” trees. Size does not = accuracy.
To break through brick walls	The two individuals that I have yet to research, as well as perhaps viewing trees for others whose DNA I match and share a common ancestor may be illuminating. It would be difficult to sift through the chaff for a newer genealogist.
To reconstruct and paint the DNA of my ancestors	For this, I need at least segment data, if not a chromosome browser. I hope that Ancestry customers will transfer their DNA files to FamilyTreeDNA, MyHeritage and GedMatch to reap the added advantages of the tools at those sites – including the availability of segment information and the resulting confirmation ability.

ThruLines Recommendations

The “Potential Ancestor” feature could have been, and still can be presented in an entirely different way, facilitating responsible genealogy, including the following:

1. Extremely visible and repeated warnings cautioning that ThruLines are not definitive, only hints and MUST BE VERIFIED with research.
2. Do not ignore the ancestor in the customer’s tree.
3. Providing “suggestions” to look at alternate ancestors or trees for additional information for ancestors, not doling out “potential ancestors” to replace your existing ancestors.
4. Implementing artificial intelligence (machine learning) for accuracy including factors such as looking for multiple births in various locations (a person can only be born once), cobbled together frankentrees, multiple marriages at the same time in different places, births too late or early in the lives of potential parents, and more red flag factors that should down-weight trees as being “recommendation worthy.”
5. Sharing with the customer why these trees were considered as recommendation worthy, similar to the MyHeritage confidence factor and side-by-side comparisons.
6. Eliminating “Ancestry Trees,” IGI records and other similar “word of mouth” types of sources as being “recommendation worthy.”
7. The ability to “dismiss” a “Potential Ancestor” suggestion and for that dismissal to be part of the AI learning process relative to future recommendations.
8. Ability to group ThruLines, such as by categories: Dismiss (inaccurate), Accurate and processed, Reviewed but unknown accuracy, In Process, New and not yet reviewed, etc.
9. Restoration of shared ancestor hints.
10. Fix bug in common ancestors causing no matches with common ancestors to be found, which I would presume is supposed to replace Shared Ancestor Hints.
11. Permanently archiving Circles and NADs.
12. If the Circles must be replaced, find another way to provide a genetic network that includes people who descend from the same ancestor, have DNA tested and match some of the people in the Circle or NAD but not everyone.
13. The ability to know when looking at a tree of a descendant of an ancestor if they have tested, and if so, if they match you.
14. Much more adequate product testing before release. By any measure, this release has been miserable and was not adequately tested in advance. No one expects new code to be bug-free, but this is unacceptable.

Ancestry, I hope you’re listening, cause the ‘gators are circling and you need to help us escape from this mess you created.

______________________________________________________________

Disclosure

I receive a small contribution when you click on some (but not all) of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

• Family Tree DNA
• MyHeritage DNA only
• MyHeritage DNA plus Health
• MyHeritage FREE DNA file upload
• AncestryDNA
• 23andMe Ancestry
• 23andMe Ancestry Plus Health
• LivingDNA

Genealogy Services

• MyHeritage FREE Tree Builder
• MyHeritage Subscription with Free Trial

Genealogy Research

• Legacy Tree Genealogists for genealogy research