Over the years of working with my DNA matches I had figured out how a number of them were related to me, but there are still a lot that I have no clue how we are related. So, I decided I wanted to systematically work through some of my Ancestry matches. To do that I need to figure out a plan of attack.
While looking for techniques I might want to try I came across some YouTube videos about doing Genetic Genealogy research. One of them was by an adoptee who talked about forming groups of matches to sort things out. She took sticky notes and wrote names on them and then looked at their shared matches and put them up on her wall based on groups of shared matches. This is the concept I am going to write about in this blog.
Since I don’t have the wall space to put a bunch of sticky notes up, I decided to use 4x6” index cards for this project. Easy choice, I already had a bunch of that size index cards I had bought for a different project. They are large enough to hold a good amount of info without being so large handling them becomes difficult. I also decided index cards would be easier to rearrange than sticky notes.
Since I wanted to be able to label each card as to what group I was assigning each person to the first thing I had to do is decide how I was going to group these matches and what name I was going to call each group. Since my first chore was dividing between Maternal and Paternal side, I decided that would be the basis for my group names. This is what I came up with:
Because right now I will primarily be working with my first 50 matches I figured this would give me enough groups to be able to sort out who the common ancestors are but not so many that sorting becomes unmanageable. I decided if I come across a 4th cousin match I would just put them into the group that was their descendant's group. For example if I found a 4th cousin who was descended from Antrim/Mason (parents of the Antrim part of Paternal 2B) I would just label them as Paternal 2B.
The next decision was the layout of the individual cards I would be using. It would need to be something I could quickly look at and be able to see the basics about that match. I decided there were some basics I wanted to be able to see quickly: Name, Group, Relationship, amount of shared DNA over how many segments, and if they have a tree. This is the layout I decided on:
First line on front of card: match’s name, included the managed by name if applicable, on left and on right space to put a group name.
On the 2nd line I will put my relationship to the match. If I know the relationship I will use pen to write it on the card. But, if I don’t know the relationship I will use pencil to write predicted range. I will then write amount of shared cMs and over how many segments.
On the 3rd line I will write "Tree" then note if tree is public/private/none.
I skipped a line then wrote “Contact Information:” and left 4 blank lines before writing “Notes:” these are areas I will fill in as I work the matches.
After I have put the basics on the front of the card I then will flip it over and make a list of all the shared matches between me and that person.
The first sort for most of my matches is really easy. I have a maternal aunt who has tested so if someone matches both me and her the match is on my maternal side. In addition I have a paternal 1st Cousin Once Removed and a paternal 2nd Cousin who have tested, which means if they have one of them as a shared match they will fall on my paternal side.
Like I said earlier, I decided to start with my first page of matches on Ancestry, which has 50 people on it. This will give me time to refine what I am doing without being too big of an issue if I decide to start fresh with something else. I decided to also include anyone I had a "Shared Ancestor Hint" (SAH) with. This gave me about 65 cards I am going to be working with to start.
After writing out the front side of the cards with the basics but not all the shared matches lists, I took a quick look at the groups that I had found. I discovered I had lots that were maternal, some that were paternal. However, there were 7 out of the 65 that I couldn’t tell which side they fell on. They didn’t match my maternal aunt nor did they match my paternal 1C1R. So I decided to take a closer look at those.
I started by writing the shared matches list on the backs of those 7 cards. 5 of the 7 had a lot of the same names on them. This is the basis for what I did next.
I opened excel and wrote down my name as #1 and then every name of my matches and all of the shared matches that showed up for any of those matches, numbering them as I added them so I could number the columns and know who that column was referring to.
I then went through each name and marked the intersecting square with an X if the two people were a shared match and a 0 if they were not a shared match. I then filled in the squares that had an X in them in yellow to make them easier to see. This is what I came up with:
First thing that popped out is that the first group of 10 not only matched me, but each one of them matched everyone else in that group. I am thinking with that level of matching with each other there is probably a shared ancestor there.
Then, at the suggestion of someone else, I shifted rows around a little (and I was very careful that I didn’t accidentally change who was sharing with who). That led to discovering a 2nd grouping where 58 of the 64 possibilities also showed up as shared matches. Not quite as good as first group, but probably still worth working to see if there is a shared ancestor there.
Here is that chart:
If you are still with me here, I posted this just to share what I am working on, with the hopes that someone can verify whether or not my hypothesis is even a possibility or if despite the pretty chart I still have no new real information to work with. I am really hoping that I do because many of the 25 matches have no tree or a tree with only a couple of names. And those that do have a few names, some of those have no dates or locations attached to the names. I am guessing my next step is to work on building some mirror trees but with such limited information to work with it will make building mirror trees difficult.