Friday, March 31, 2017

Before Starting a Mirror Tree: Finding Common Ancestors

If you are adopted and you are trying to build a Mirror Tree, finding a common surname or set of grandparents among all the names in the trees on AncestryDNA can seem a daunting task. This is what I do to find and sort common ancestors (CA) and knock down brickwalls. This saves having to print out numerous trees and looking at all the different pages. (We will bounce back and forth from spreadsheet to View all Matches to do this.)

The view match page displays the person tested and five more generations of grandparents. If your match shares little as 14 cM avg for 4th cousin 2x removed the info from this page should often be enough to provide the information you need.
Click on the name of your match to go to the contact, notes and information page. This page also often shows a hidden tree you can click on at lower left. Click on the blue circle with and i to find amount of shared DNA.

You are going to be making a spreadsheet of your closest matches closest matches. Got that?

Don’t worry about how to work with spreadsheets at this point. Just understand: cells are all the little rectangles. Rows are numbered and go horizontally across the page. Columns are identified by letters and go vertically down the page. Got that? You’re good to go. (Youtube search spreadsheet basics if you don’t “get it”.)

You can use any spreadsheet . (Excel, Open Source…)

1.     Open a spreadsheet.

2.     Go to AncestryDNA and open View all matches.

3.     Enter the name of your closest match as a heading in row 1, column A. (You can adjust the column width as needed by hovering your mouse on the lines between the alphabet letters until you see a cross with an arrow on each end. Click and slide to the right.)

4.     If they don’t have a tree, skip to step 9.) If they do have a tree, proceed to the next step. (Note: Sometimes you cannot see that a tree is available until you click on the View Matches page. There may be a link to a tree at the bottom of the info screen that allows you to see one or more trees.)

5.     On the left side of the View Match page is a list of all of their surnames. Click on the greater than symbol (>) and open all the surnames. Do this all the way down to the bottom of the page. Depending on the amount of DNA you share with your match

6.     Starting from the bottom of the page, left click your mouse and hold down as you scroll up the list and highlight all the names in the surname list. Right click to copy.

7.     In the first empty cell under your closest match’s name, row 2 Col A, right click and paste.

8.     Next, you will repeat what you have just done, with one change; you will not be working with your next closest match, but with the Shared Matches of your closest match.

9.     View All Matches page your closet (top) match click View Match, find Shared Matches tab and click. You are now looking at your match’s matches.

10.  Find the first person in their shared matches list that has a tree. Use this person’s name as the heading for row 1 column B in the spread sheet. Repeat steps 5, 6,7,8. Copy and paste at least four shared matches trees in the spreadsheet of your closest match. Now you can compare all of the surnames in alphabetical order side by side.

11.  I start with the first surname and do a search. (click in any empty cell in your spreadsheet and press control+f). Type a surname into the pop-up box. Click find.  If there are no matches go to the next surname on the list. Continue until you find a common surname shared in different columns.

Update: On the spreadsheet you can also highlight each column by clicking on alphabet letter at the top of the column, then go to the A-Z sort and click. This will make it easier to see common ancestors. 

12.   Highlight any repeated surnames that show up in different columns. If you find more than one repeated given+ surname, you may want to highlight it in a different color. Go back to Ancestry and check trees to find the name of the spouse for the most recent (by dates) common ancestor.

13.  Create spreadsheets for each of your top twenty matches shared matches. You will find different sets of common surnames. One may be a surname you have already identified with a different spouse. These are most likely linked together in one side of your tree. A different set of surnames may (not always) indicate the opposite side of your tree. Mark these different sets of linking common surnames into Group A and Group B.

14.  Use the closest matching couple (amount of dna and segment size) and begin building your Mirror Tree.

If you have a tree (gedcom) program. (many free online) You can build different trees on your computer based on a different set of surnames without uploading them on Ancestry.

For adoptees seeking assistance you will want to join the private FB group DNA Detectives. When you request to join, tell them "Hi" from Barbara.

Use Caution with this note below. Some people have huge trees.

Note: I was just informed by a very savvy Ancestry user at DNA Detectives; if you have a full Ancestry membership (not limited to AncestryDNA) you can go to the tree of your match and click on Find Person then select "list all people." Thank you Sharon! 


  1. Will this work for my Donegal ancestors where there's endogamy? I'm trying to find my grandmother's father. I have worked out from DNA matches that he's from Donegal.

    1. Pye, there are some free videos on the Legacy software site that have a lot of info on finding people Ireland. Look at their webinars. I thought Ireland would be hard to search, but evidently one company has loads of records from other sources. Watched a video just last night.

  2. If you are on Ancestry it should work fine. Keep looking for the common surnames, pay attention to the dates, someone may be a sibling. IF you keep coming up with a different surname that fits location and time, you might want to consider the possibility of a non parental event. Has anyone in that surname line done Ydna testing?

  3. Could a mirror tree work for a person with a brick wall in their tree? I've got two great-great grandmothers with NO past. One, ugh, has the surname of "Miller."

    1. GizaCat, I do this to breakdown brickwalls. I would do shared matches with anyone who has your greatgrandmothers in their trees on Ancestry. See if you come up with a common surname in their tree and use it (if they appear to link to your grandmothers) use that surname and study it carefully. Look at other people's tree with surnames for your grandmothers. Many people used their middle name or nickname which can throw your genealogical research off. Pay close attention to dates and locations. See if you can find siblings for your grandmothers. Look for people born within one or two years of their birthdates in the same geographical area. Sometimes parents died and grandparents raised the children. That might have happened with your grandmothers. Depending on the time, some parents sent their children to live with relatives before the Civil War broke out. That may be a reason you are having challenges finding their parents. I would also recommend doing Gedmatch Triangulation and segment comparison. Perhaps they will show up there.

    2. GizCat, you could also do triangulation and see who you link to and capture their trees. Most triangulations hit in 4th cousin category, so it does take some work.

  4. What do I put in the B Column?

    Also...most (all?) of my closest matches are on my mother's side. Would doing the mirror thing with a more distant match of someone I suspect to be on my father's side help me figure out more on his side?

    1. I would make sure I have run "are your parents related" tool on Gedmatch.

      You start with your closest match in Col A. Col. B you begin with SHARED matches of your closest match. I would do several entries for their closest SHARED matches and keep adding to this page until you get repeats of the same surnames. Eventually you will come up with the same given and surname with correct locations and birthdates (I allow a 2 year play one birthdates because so many people estimate from census records.

      After you find common ancestor for your closest matches matches. Do the same thing with your second closest matches shared matches. By the time you have done this for your top 15 closest matches shared matches. you should have at least a few for your different parents. Much depends on how close all your closest matches are.

      When you are certain (you have made notes on your maternal line) then go ahead and do the closest matches shared matches. Don't be surprised if you discover they are from your maternal line, only further back. I spent 3 weeks sure I was working with an adoptees bio father's line; turned out to be her mother's. Bummer. But I was able to build her maternal line much further back in the process.

  5. This is genius, I wish I'd known about this method earlier. I attempted my first matrix, however I only had 2 other shared matches from a 4th-6th cousin to work with. One of the shared matches has NO surnames common to all/any 3 of them - I was wondering what this might indicates? I was surprised as they have over 500 names in their tree.

    1. I am so sorry I didn't see this earlier; I am not exactly a computer genius and thought I had this set up to notify me when I received a message. April, it means you probably only match on one name and they don't share DNA with the others. Or they don't have that cousin in their tree. Remember you have two genealogy trees one is paper, the other is a genetic tree. Not all cousins share DNA. I would highly recommend you donate $10 one time to use Gedmatch Tier 1 and download from their triangulation program into a spreadsheet. I will be covering that info in the coming weeks to explain it to everyone. If you have the surname Brady anywhere in your tree, I am doing a triangulation and segment study on that surname. PM me on FB if you do.

    2. Thanks for your reply; I'd be interested in your post about triangulation on GEDmatch, I'm keen to learn how to utilise that tool

    3. Susan, don't know if you have been following. I have posted on the tree graphic and on the segment bar graph re: triangulation. I am a visual learner as are many people. That is what I start with so people can get used to the concept of triangulated groups and working with them. My focus is on what you see when you look at the tool. To "read" what it is telling you and how to respond to matches and get responses back. For the science behind the tools I always recommend reading Jim Bartlett's Segment-ology blog. Blaine T. Bettinger's books are very good. I am doing the posts as I can. Seriously, I DO blunder my way through. If I can succeed with all my blundering, anyone can. I just love sharing my successes and my failures. Let me know what you want to learn and I will do my best to blog what has worked for me.

  6. Am I correct in assuming I should only compare my closest match with their 4 closest matches. And not compare the 4 closest matches with each other.

    1. So sorry I just realized my program isn't alerting me to comments. bmac200t, What you are going to do is make a spread sheet for each of your closest matches closest matches. Eventually you will find several of your closest matches showing up as their closest matches. That will assist you in sorting your lines, your maternal and your paternal. (unless they are related) If you do a bunch you will actually find at least four lines. A's maternal and paternal and B's maternal and paternal. It pretty much depends on how close your matches are. You should end up with at least four distinct groups.