Concepts – Mirror Trees

What are mirror trees, and why would I ever want to use one?

Great question.

You’ll hear genealogists, especially adoptees or persons trying to find a missing parent mention using mirror trees.

Mirror trees are a technique that genealogists use to help identify a missing common ancestor by recreating the tree of a match and strategically attaching your DNA to their tree to see who you match that descends from which line in their tree.

I have used mirror trees to attempt to determine the common line of a close cousin whose common ancestor (with me) I simply CANNOT discover. Notice the words “attempt to.”  Mirror trees are not a sure-fire answer, and they can sometimes lead you astray.

Foundation Concept

The foundation concept of a mirror tree is very straightforward.

Let’s say you match Susie as a second cousin. This means that you should share a great-grandparent with Susie. A relationship this close OUGHT to be relatively simple to figure out – except sometimes it isn’t.

Note that vendor relationship estimates are just that, estimates of relatedness based on total and longest cM, and they can be off in either direction.

In the case of third cousins or closer, vendor estimates are generally pretty accurate.

You can view the ranges of cMs and relationships in this chart.

Of course, when you match someone, you don’t know who the common ancestor is, nor do you necessarily have access to their pedigree chart or tree. If you do, and you can easily see the identity of the common ancestral couple, that’s great – but life isn’t always that simple.

In Practice

In my case, I match Susie, and no place in our trees, at ALL, is a common ancestor, let alone three generations back in time. Furthermore, her entire line and my father’s line were all from Appalachia, so common geography doesn’t help.

We matched at Ancestry, so we both uploaded to GedMatch, where we match almost exactly the same, and the relationship prediction is the same as well. Someplace, in one of our trees, is an NPE, a misattributed parentage – because both of our trees are complete back beyond those generations.

Uh oh.

So, I created a tree in my Ancestry account, duplicating Susie’s tree, and making it private – at least one generation beyond great-grandparents – just in case the estimate is wrong. Then, I connected my DNA to her tree, as her.

In my case, I have two DNA tests at Ancestry, my V1 results and my V2 results. I never really thought about this as a way to keep one set of results working for me, connected to my own tree, and to have a second set of results to connect to mirror trees – but that’s exactly what I’ve done. I utilize the second set of results as my “working on a problem” results while the first set of results just stays connected to my own tree.

After connecting my DNA results to the mirror tree and giving Ancestry a couple of days to cycle through, creating connections and green leaf “shared ancestor” hints, I checked to see who my DNA attached to her tree says I match, and which line in her tree “lights up” with match hints. If I can’t tell by connecting my DNA as her, I can also connect my DNA to her parents and grandparents, one at a time – again – looking for green leaf shared ancestor hints in those lines. No hints = wrong line.

This process shows me in which of her lines our common lineage is found – even if I can’t exactly pinpoint the common ancestors just yet.


I had planned to provide step by step directions for how to create a mirror tree and then how to utilize the results, but then I discovered that someone else has done an absolutely wonderful job of writing mirror tree instructions. There is absolutely no reason to recreate the wheel, so I’m linking to two articles from the blog, Resurrecting Roots, as follows:

After building a mirror tree, their next article explains what to do next.

Now, if I could just figure out that common ancestor with my second cousin match. You may encounter the same type of challenge.

If the right people haven’t tested yet, you may not be able to achieve your goal on the first try. Or, in my case, it appears that we may have more than one common ancestor – complicating matters a bit. If this happens to you, wait a few weeks/months and connect the tree again, or build it out another generation to increase your changes of a green leaf hint.

The great thing about genetic genealogy is that more people are testing every single day. Give mirror trees a try if you’re an adoptee, trying to find an unidentified family member in a relatively close generation, or are being driven absolutely batty with a relatively close match that you can’t solve!

If you need help solving these types of problems, I suggest contacting dnaadoption and taking one of their classes.  They aren’t just for adoptees.


Standard Disclosure

This standard disclosure will now appear at the bottom of every article in compliance with the FTC Guidelines.

Hot links are provided to Family Tree DNA, where appropriate. If you wish to purchase one of their products, and you click through one of the links in an article to Family Tree DNA, or on the sidebar of this blog, I receive a small contribution if you make a purchase. Clicking through the link does not affect the price you pay. This affiliate relationship helps to keep this publication, with more than 850 articles about all aspects of genetic genealogy, free for everyone.

I do not accept sponsorship for this blog, nor do I write paid articles, nor do I accept contributions of any type from any vendor in order to review any product, etc. In fact, I pay a premium price to prevent ads from appearing on this blog.

When reviewing products, in most cases, I pay the same price and order in the same way as any other consumer. If not, I state very clearly in the article any special consideration received. In other words, you are reading my opinions as a long-time consumer and consultant in the genetic genealogy field.

I will never link to a product about which I have reservations or qualms, either about the product or about the company offering the product. I only recommend products that I use myself and bring value to the genetic genealogy community. If you wonder why there aren’t more links, that’s why and that’s my commitment to you.

Thank you for your readership, your ongoing support and for purchasing through the affiliate link if you are interested in making a purchase at Family Tree DNA.

32 thoughts on “Concepts – Mirror Trees

  1. Until i read this article, i did not know what a mirror tree was. I am an adoptee. In 1998 i found out who my mother was (she died in 1991). Only last year (2016) did i find out who my father was (he died in 1998). I found out his identity with 100 percent certainty through DNA matching and the analysis of DNA matches’ Family Trees. I did not make a Mirror Tree, which would have linked my DNA to one matche’s line. I don’t see how this could work in searching for a father or mother, (or both.) What i did was to create at Ancestry a private PaternalDNAMatchesTree which i did not link to my DNA test results. I added as many paternal matches’ trees that i could find (or invent myself) at Ancestry, 23andMe, FTDNA. I entered about 10000 names altogether, facilitated by the fact that i have an Ancestry subscription, so that i could find and add names and records more easily. My matches at first were 3rd and 4th cousin matches. At a certain point i noticed that the same ancestor Konrad Jung of Siefersheim, Germany appeared in three trees, so i connected those trees. Then i found several trees linked by another common ancestor John Whittaker of Manchester, England. By then I knew that i was descended from the Jung and Whittaker families. I added as many siblings and descenants as i could to all the branches of those two families. Eventually i found a marriage between Frank Jung and Nellie Whittaker (my paternal grandparents). I connected myself to them through ”Unknown Jung” (my father). Then i found out that they had two daughters and two sons. The younger son was the only one who could have been my father (right age, right time and place). Later i found two 2nd Cousins once removed, and eventually a living PATERNAL FIRST COUSIN 1R, whose name was given me by one of my first 3rd Cousin Matches. He agreed to test at Ancestry and we indeed proved to be First Cousins 1R. His mother is my First cousin, the daughter of my father’s brother. He has been wonderful to me, sending me much information and many photos of my father, his parents, siblings and forefathers. I am still in absolute wonderment at the power of the new science of genetic genealogy. And it only took a few months intense work, once i got started. I cannot describe who complete and happy i feel now that i know who my father was, as well as my mother.

    • First of all, thank you for this….When you say above that you “connected those trees,” what does that mean? Do you mean layered in info from those trees or did you connect to their trees somehow? HELP! (and thank you!)

      • In my own PaternalDNAMatchesTree I connected the trees of my DNA matches which had the same Most Recent Common Ancestors to one another by simply merging the varous Konrad Jung and John Whittaker profiles, or, by supplying the missing Mpst Recent Common Ancestor to the tree where i knew that the oldest shown ancestor was without a doubt a son or daugther of one of the above-named Most Recent Common Ancestors. That way the trees were joined at the respective MRCA. Then i connected my own profile to those trees by a series of three or four profiles called ”UNKOWN JUNG” or ”PRIVATE JUNG” back to the MRCA: Konrad Jung and his wife, and John Whittaker and his wife. Eventually i figured out that i was descended from Konrad’s son Casimir Jung, and thus filled his name in the profile labled ”UNKOWN JUNG GR-GR-GRANDFATHER” because i realised that i was a third cousin to three Jung descendants of Konrad Jung through his son Casimir, and a 4th cousin to two descendants of Konrad Jung through another son Conrad Jung, Jr. The same with the Whittaker side. Eventually though, when i found out that my grandparents were Frank Jung and Nellie Whittaker, i filled in the Gr-Grandparents’ names as well, and had to rearrange the profiles connecting them to my own profile, for i had originally randomly determined (incorrectly) that my direct male surname line might be WHITTAKER, and my paternal grandmother’s line JUNG, but it turned out to be the opposite: JUNG is my direct male surname line, and WHITTAKER is my paternal grandmother’s line. It is really simple! It just takes lot of work inputting the trees and then analysing them. Good memory for details helps too.

  2. When that happens I find using the Gedmatch tool two people who match on one or two of both kits to be especially helpful. Often the results will lead you to one or more people who link to one of you but not the other. Finding who those people link to may result in your answer.

  3. Thanks, Roberta … good article and the links were great. I always hate to disconnect my DNA from my tree, but having a second test is a great idea.

  4. I did this, and found it very useful. At some point I may share some more details, if you’d like (of course you already know some of them), but for now let me simply say that my Mirror tree at Ancestry eventually led to dozens and dozens of Shared Ancestor Hints and inclusion in six DNA Circles related to my hypothetical ancestors.

    I have to say that there’s an element of luck involved, but it’s really luck AND persistence. You need both, as well as a strong sprinkling of patience. But let me just say that DNA testing makes possible many discoveries even that were consciously concealed.

    Some may regard this as an invasion of ancestors’ privacy rights, but those who are now deceased no longer care. I also believe that it is a child’s right to know who his biological parents are.

    In my case, the issue didn’t involve my parents, but my mother’s father’s biological parents. And thanks to my Mirror tree and some other factors — including, as I say, quite a bit of luck — I believe I’ve found both who they were not, and who they were.

  5. This article popped up just as I was having a “what is this moment” in my genealogy program. I have color coded my four grandparent lines and and was suddenly faced with this situation: My paternal grandfather’s line intersects with my maternal grandmother’s line. What I discovered is a set of 6th greats with two daughters. One married an ancestor of my paternal grandfather and one married an ancestor of my maternal grandmother. It made it really easy to understand your hunt for common great-grandparents and to have your situation, with a 2nd cousin, so close you feel like you should know the great-grandparents by now! At least your 2nd cousin is communicating. I have a “second cousin” who I can’t even identify. One of those, “I tested but I won’t share people”. Thanks for another great article and good luck!  

  6. I read your post and her two posts,and I don’t understand the advantage of doing this. For my missing great-grandmother, I’ve already researched the trees of my third cousin who won’t reply and worked on seeing how his line connects to the lines of others we both match. How is this different?

      • Roberta…I have a mirror tree (actually two since I don’t know EITHER of my grandmother’s parents). In the one that I am working in currently, I KNOW that I am swimming in the right pool of ancestors as each person I connect to (have connected to five or six possible ancestors) shows Shared Ancestors. Some people I connect the tree to comes back with five shared ancestor hints and another shows 40– all going back to the same gr gr gr gr but it’s like mercury that keeps slipping out of my hands. I can’t seem to definitively nail down that critical definitive ancestor or set of ancestors at the gr grandparent level. Any suggestions?

      • I would track all descendants forward from that couple, to some extent looking for common geography. Use the dnagedcom tool to download the tree info so you can sort in a spreadsheet.

  7. Thanks for the article on Mirror Trees, Roberta. I asked you awhile back to do an article on the subject – you said you would – and you delivered!

  8. Why build another tree? Why not just create the newly found cousin in my tree, even if I can’t figure out our common ancestor? Then I could link my DNA kit to my newly found cousin.

  9. Hi Roberta,

    Question: If we make a mirror tree and attach our DNA to a person in that tree, are we also causing the new found matches to our DNA to receive false (I.e. Misleading) matches pop up on their searches?

    I know what I’m thinking, but sorry if I’m not doing a good job communicating it to you.

  10. In 2004 my husband and a presumed 1C on his paternal side both tested their yDNA. My husband did not match his cousin, but the cousin had a perfect match with a known distant cousin. The paternal paper trail proved that hubby’s presumed paternal grandfather was married to someone other than hubby’s grandmother when his father was conceived. The Y test quickly determined the biological surname and we joined the appropriate project in 2005 on FTDNA. He did his first atDNA test three years ago and convinced several close maternal relatives to test as well. Earlier this year I finally convinced him to test with AncestryDNA. One week before his results posted I discovered his first 3C match with someone with his bio surname on FTDNA. I contacted his match and received a short list of his direct line ancestors back to his gg-grandparents. I added them to my husband’s tree and built it out adding siblings, their spouses, children, grandchildren complete with as many sources on each that I could find. It has taken many weeks, but my husband now has 28 ancestor circles on Ancestry. Most are on his maternal side, but he has quite a few from the paternal line I built out. I have connected the most likely candidate as his father’s dad. Through him there are ancestor circles for one pair of g-grandparents, two sets of gg-grandparents, and three ggg-grandparents. The locations match where hubby’ grandmother lived, and armed with the new surnames from all the g and gg grandparents, I’ve been able to triangulate some additional paternal matches on FTDNA, GEDmatch and by downloading data to DNAMatch. Even with all of this I’m hesitant to say that I’ve “discovered grandpa”. His atDNA matches with everyone except those with his biological surname are strong. Everyone tells me that his biological surname ancestors intermarried, so there is supposedly a lot of endogamy. Would that lead to smaller cM with those matches?

  11. As Roberta has pointed out, even if one is not an adoptee, mirror trees can be very useful. I’ve modified the mirror tree concept to try to make some sense of the complexities of my many Irish ancestors.

    We all have 16 2gr-grandparents. Of my 16, 11 are Irish – 7 on the paternal side, and 4 on the maternal side. So, I’ve got lots of Irish blood. Trying to construct accurate family trees for Irish lines that go back into the “old country” is a challenge, to say the least. Many of you will agree with me, I’m sure. The plethora of common Irish surnames… Murphy, O’Brien, Sullivan, Kelly, Walsh, O’Connor, etc., doesn’t help. Anyone who is Irish has plenty of them, and usually in multiple lines. And complicating the common surnames picture is the fact that first names are common also – almost every family had a John, Mary, Patrick, Bridget, Catherine, etc.. Even with the important recent influx of online Irish records, it is still difficult to know if you have (for example) the correct baptism record of John Sullivan, parents Patrick Sullivan & Bridget O’Brien, of Cork as there sometimes so many other similarly named families.

    For awhile now, I’ve been grouping my autosomal DNA matches who also have Irish ancestors into clusters based on a number of factors such as chromosome position, relatedness to other common matches, common surnames, and geographic places in Ireland. Of course, this is greatly helped if the match has a family tree I can view.

    Where I differ from the usual mirror tree model is that I put all “promising” matches into the same overall tree. Call it my Irish tree. I’ve created it to be private and unsearchable so I can play “what if” with it as much as I want. I started the tree with one promising match, then filled out only his Irish ancestor lines. Then I added another promising match, and did the same thing. To physically get them inserted into the tree, I start by temporarily adding the new match to someone in the previous “island” tree as a “relative” and then I immediately go to the relationships page for that new match and disconnect the relationship. Now I have two separate “islands.” Two separate little trees within one tree. Then I add more matches, and “lather, rinse, and repeat” as the saying goes. As I build, I keep a separate list of each match added to the tree in a spreadsheet. This is important, as what you’ll end up with is a large tree with many separate “island” trees. There is no real “home person.” None are obviously related… at first. But that is the beauty of it… as you add more matches and their Irish surname lines… eventually some will.

    A way to visualize it is to imagine looking out over an apple orchard. Each apple tree is similar to one of the “islands” in your overall mirror tree. Each tree is planted a certain distance from each of its neighbors. But over the years, as each tree’s root system grows, expands, and spreads out; roots from different trees will begin to make contact. The same with your “island” trees in your mirror tree. As you add more matches, and fill out their applicable surname lines, some will start to connect. Usually, after filling out a surname line for a new match, I’ll then look over my list of all people in the overall tree for that surname to see if I’ve added someone that looks like a duplicate. That usually lets me know I’ve reached a connection in those lines.

    As of today, I’ve got approximately 3,350 people in my Irish tree based on approximately 70 “island” trees. I’ll admit, it’s a lot of work and there are no connections in the beginning. But it is worth the effort, as the more matches and their trees you add, the more connections you will eventually get.

    As an added boost I also use the DNA attachment method in the tree to let Ancestry’s family tree green shaky leaves do their magic. I’ve found some connections that way too. What I do is connect my DNA to each “island” match for about a week. Then I disconnect, and connect to another one. I track the dates I’ve connected to each one in the list of matches in the separate spreadsheet.

    I realize this is sort of a bastardization of the original mirror tree concept, but it works for me and has yielded fruit. I mention it here only in hopes it might give others ideas to “think outside the box” and come up with other ideas they will share when it comes to what we can do with our genetic genealogy data and tools.

    • Very good explanatoin! This is more or less exactly what i did to find my father, except that i added my DNA matches in capital letters, with amount of cM shared under ”DNA markers”, in order to recognise them as matches and as heads of their trees in my Paternal DNA Tree’s List of Persons. I also added an asterisk in the suffix box to a profile for each DNA Match whom i found to be descended from that person, as well as an asterisk for each person linking that ancestor the DNA match, making a visible trail. You have explained this ”bastardised mirror” concept very clearly. Last year when i was busy with it, i thought that this was the MIrror Tree method! Anyway it did work for me. I found my father’s identity and close living paternal relatives.

      • Thanks for the reply, Albertus.

        I’m so glad you were able to find your father doing essentially the same method I described. I’m only trying to connect trees of possible distant Irish cousins. But you were seeking a parent – that’s a lot more important.

        The only drawback I can see to most folks using our method is that it takes lots of time. You were DEFINITELY going to invest the time. But many folks won’t take that kind of time to just find cousins. Honestly, if I wasn’t retired, it would be difficult to devote the time required to do it.

        But we’ve proved it does work. Maybe some other brave searchers will be willing to give it a try. Or… maybe… Roberta will be able to figure out an easier, less time-consuming way. She’s pretty darn good at coming up with innovative solutions to genetic genealogy problems, ya’ know. 😉

        Good luck to you, Albertus!

      • Hi Tom, for some reason I have a hard time wrapping my head around mirror trees. I’m not an adoptee and my tree is filled back to 3rd g grandparents and the occasional 4th’s. I have a Mary McNeill for example and she is a brick wall. Would your method of creating islands of matches who have different McNeills work for me? For example: some of my matches have Barra McNeills, some have Argyll McNeills, yet others have some from Nova Scotia Would your method at least point me into the direction of which McNeill family I descend from? Many thanks.

  12. Roberta – I appreciate all that you do! Your articles have been so helpful to me, many, many times. Keep up the good work!!!

  13. For some reason, I’m having trouble posting with my regular e-mail address so I’ll try a secondary address. Hope this works.

    Thank you for this article. I wonder if you have any information from others who have attempted to link their DNA results to a mirror tree and have messed up their true results. When you link your test to another tree, then at some point later, relink it to your tree, will you regain all of your original leaf hints and DNA circles, and will you still have all the information entered in the notes field?

    I have a third cousin match, shown as “160 centimorgans shared across 7 DNA segments”. Up until now, with a match this close, I have always been able to determine which of my lines we are connected to by looking at our shared matches. But in this case, I was stunned to find we have 36 shared matches, and except for the match to my brother, I didn’t recognize any of the others. This match only has a small tree back to her grandparents, but using that information, I started building a private mirror tree for her, researching her lines back further than she had. I have now determined her tree back to all but 2 of her great great grandparents. Still, none of the surnames are familiar. I have messaged her but no response yet. Based on the ages of her parents (dad born in 1886 and mom in 1899), she must be in her eighties or possibly nineties.

    The only thing left to do now is to connect my DNA test to her tree and see what happens. But I’m really skittish. Right now I have over 700 leaf hint matches and I’m in 48 DNA Circles. I also use the notes field for lots of information. I don’t want to mess up any of that. If I connect my DNA test to her tree, then reconnect it to mine, what are the chances I won’t get back everything I’ve gained so far? Do you know of anyone who has lost any information doing this? Thanks.

    • That’s one reason why I’m glad I did the second Ancestry test. I have all the same Circles, but I have no NADs on the second test. I have 14 NADs on the original test. For a long time, I had no Circles on that test either. I understand your nervousness and I would be very hesitant without the second test to play with as well.

      • I’ve done it several times without losing anything. You need to disconnect your DNA from your tree first … that will cause you to lose leaves and circles, but they will come back (it will take a few days) when you re-attach it after you finish with the mirror tree. While your DNA is attached to the mirror tree, you won’t have those features on your own tree. Your notes will stay, though.

        I dislike disconnecting my DNA from my tree, as it will appear to your matches that you don’t have a tree. I like Roberta’s idea of a second test. Until I get around to test again, I have been using my brother’s DNA for the mirror tree.

  14. I get the concept of mirror trees and how this could be useful for testing theories. But I have a problem at Ancestry with a person who is not the testee, who has made the mirror tree public and arbitrarily made one of my grandfather’s siblings the mother of the testee! And, no, the sibling cannot be the mother because she never left the small valley that was home and that was filled with all her relatives. I have explained that this sibling cannot be the testee’s mother, and have asked this person who set up the mirror tree to make this tree private, with no result. So I appreciate your emphasizing that this kind of tree should be private!

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s