09 June 2020

How to Diagram a Mystery DNA Match

This is what I'll do when a DNA match is too abstract to make sense.

One of Mom's DNA matches is driving me crazy. Mary and Mom share 250 centiMorgans. That's a lot for someone I can't identify. Ancestry DNA says there's a 67% chance they are:
  • 2nd cousins
  • 1st cousins twice removed
  • half 1st cousins once removed, or
  • half 2nd great aunt/niece.
Mary learned from her DNA test that her father was not her biological father. A mutual DNA match led to the name of her father. I recognized the last name from my ancestors' hometown.

It looks like Mary's biological father's grandmother is my blood relative. Her name was Maria Grazia Sarracino. Sarracino is my maternal grandmother's maiden name. Everyone named Sarracino in their little hamlet in Italy is family. But the town has very limited vital records available.

I looked at Mary and Mom's shared matches. Two people have the last name of Mary's birth father, 1 has the name Saracino, and 1 has a Sarracino grandmother.

I decided to research Maria Grazia Sarracino—the grandmother of Mary's birth father. Her daughter was Angela Coviello—Mary's biological grandmother. She was born in 1895 to Angelo Coviello and Maria Grazia Sarracino. Angela's birth record does not say Maria Grazia Sarracino's age. I searched for more of her children, hoping for more details.

It was obvious Maria Grazia was Angelo's 2nd wife. They kept having babies until he was at least 62 years old. And there were no children born before 1893. Maria Grazia probably married Angelo in 1892. The last child I found was born in 1904. The 1905–1909 records are missing, but Angelo was getting quite old. None of their children's birth records include Maria Grazia's age or the name of her father.

Maria Grazia Sarracino was most likely born between 1860 and 1875. But without the records, I can't look for her marriage to Angelo Coviello. I can't look for her death. And I didn't find her birth.

Would a diagram of all Mary and Mom's possible relationships help solve this DNA mystery?

This turned out to be incredibly helpful. I started with my Relationship Calculator spreadsheet. (Download the file for free. Can't open an Excel file? Here's the Google Sheets version.) The calculator tells you your exact relationship to a cousin.

You can also use this relationship calculator to narrow down your connection to a DNA match. Here's how I did it:
  • Duplicate the worksheet onto a new tab, leaving the original untouched.
  • Highlight the cells that match your estimated relationship to your DNA match. (Note: The estimated relationships come from AncestryDNA. If you're not a subscriber, enter the shared amount of centiMorgans in this tool: https://dnapainter.com/tools/sharedcmv4.)
    • I highlighted full relationships in yellow: 2nd cousin, and in 2 locations, 1st cousin 2 times removed.
    • I highlighted half relationships in orange: 1st cousin, 1st cousin 1 time removed (in 2 locations), and 2nd Great Grand niece/nephew (in 2 locations).
  • Delete or empty all the rows, columns, and cells you haven't highlighted.
  • Add names to make things clearer. If Mom is the child of the common ancestor, then the common ancestor is Grandma. Put the appropriate name in each highlighted cell.
  • Think through each relationship. You'll know that some are impossible, so you can cross them out.
  • The remaining cells are your strongest possibilities.

This simple diagram showed me the most likely relationship between Mom and Mary.
This simple diagram showed me the most likely relationship between Mom and Mary.

I started with the easiest one: both locations of 2nd Great Grand niece/nephew. For one to be true, Mary would have to be the 3rd great grandchild of my grandmother. Mary's older than me, and a stranger, so that's impossible.

For the other to be true, Mary would have to be the child of my mother's 3rd great grandfather who died in the early 1800s. Also impossible. I crossed out both cells.

Next I looked at Mom's Grandchild column. These cells all have my great grandfather Giovanni's name in them. I know all Giovanni's children, grandchildren, and great grandchildren. So I know Mary can't be his grandchild, great grandchild, or 2nd great grandchild. I crossed out all 3 cells.

Then I skipped a column and looked at Mom's 2nd great grandfather, Antonio. He couldn't be Mary's grandfather because he lived too long ago. I crossed out his cell.

That left 2 possible relationships for Mom's great grandfather, Giuseppe. Could Mary be his grandchild or great grandchild? For Mary to be Mom's full 2nd cousin, she'd have to be Giuseppe's great grandchild. I can't rule this out, so I'll leave the cell alone.

That leaves me with 1 more cell: half 1st cousin 1 time removed, with a common ancestor of Giuseppe Sarracino. But Giuseppe can't be Mary's grandfather because of the enormous age difference. I crossed out this cell.

I've eliminated all but the full 2nd cousin relationship. How would this relationship work? I took a look at my 2nd great grandfather, Giuseppe Sarracino. Giuseppe Sarracino married my 2nd great grandmother in December 1864. I can name 5 of their sons, born between 1865 and 1879. But I can't guarantee there were no other children.

Giuseppe had a habit of not reporting his childrens' births. By law, Italian citizens had to report births and deaths promptly.

But my great grandfather's 1876 birth went unreported until 1898. I found it completely by chance. They didn't report my 2nd great uncle Angelo's 1865 birth until 1894. And I've never found my 2nd great uncle Domenico's approximately 1866 birth record.

There's a gap from 1869–1875 during which Giuseppe could have had a daughter named Maria Grazia.

I wish I could add Maria Grazia Sarracino as a child of my 3rd great grandparents with a dotted line. I may never be able to prove she belongs there. But thanks to DNA probabilities, at least I have good reason to believe she is my 2nd great aunt.

I'm eager to try this method on more mystery DNA matches. I'll add a new spreadsheet tab for each person I want to put to the test. If figuring out potential relationships leaves your head spinning, give this a try.


  1. Great post! I love how you used the spreadsheet to rule out relationships.

    1. Thanks! It worked very well for this example. The next one I tried was a bit fuzzier, but I did figure her out.