No announcement yet.

I have a question about my matches...

  • Filter
  • Time
  • Show
Clear All
new posts

  • I have a question about my matches...

    When a match is estimated as a second-third cousin or a third-fifth cousin, how accurate is such a guess? And does that mean they're descended of your great grandparents ( or something like that ) or can they be descended of a sibling of a great grandparent?

  • #2
    The match estimates are only estimates, based on each testing companies' algorithms and so on. Some may be more on the nose than others. You can double check them using the tool at (there are links to several versions there - version 4 is the most recent). With FTDNA, check your match in the Chromosome Browser, and only add up segments which are 7 cM or over. Disregard any segments under 7 cM. Use the resulting total with the Shared cM tool. Version 4 gives a list of possible relationships.

    For your second question, yes, your match can be descended from a sibling of a grandparent, great-grandparent, or other. A relationship chart such as the one by Alice J. Ramsey, or this cousin chart at ISOGG will both allow you to see the possible common ancestral couple for the various levels of cousin. For example, per the ISOGG cousin chart, a third cousin would share 2nd great-grandparents (great-great) and a fifth cousin would share 4th great-grandparents. Those cousins at various levels of "removed" would be related at different levels of great-grandparents for each of them.

    The Ramsey chart is good for visualizing how cousins are related using a tree display, vs. a grid chart. You can see "removed" cousins who are related through siblings of great grandparents, etc. on that one.


    • #3
      Bottom line, the error that goes with the estimates gets very large by the time you get to 3rd cousins and beyond. For very close relationships, the estimates are much more reliable, but you MUST pay attention to all the different relationships (half relationships, "removed" cousins, aunts and uncles, etc.) that have EXACTLY THE SAME predicted shared cM as actual 1st, 2nd, or 3rd cousins. In addition to the shared cM tool mentioned in this thread, I like the very popular "DNA Detectives Autosomal Statistics Chart" that is now scattered all over the internet -- do a Google search for that title -- this is the one chart I keep by my computer. The relationships are grouped in such a way that it is easy to see what the possibilities are for any given total of shared centiMorgans.


      • #4
        Another thing that you have to consider when looking at the amount of DNA shared is the type of genetic community your family comes from. It is possible that many of my matches on my paternal side share DNA material with me through several of their lines. My dad's folks were from old settler families in the West Virginia area. Now, I have no cases of brothers marrying their own sisters, but I do have cases where early on brothers from one family married sisters from another family and then their children (1st cousins married) and then down the lines of those families, there were a number of more distant cousins marrying. I have found DNA match whose parents are both related to me, and all 4 of his grandparents are related to me and so on. I'm using a system on my tree of color coding the various lines back to help in tracing the various line connections. Fortunately my great grandparents were only 3rd cousins and related only on one line rather than multiples.


        • #5
          Keigh, a few times I have found my Phelps ancestral couple 4-5 times in the tree of a match.
          BTW, how many matches do you now have at that other company? You have always had more than anyone I know. I have 84,000.


          • #6
            Hi Biblioteque: Currently at the other company I have 105,327 matches. I've actually been able to send messages to some of those, that are actually doing trees, and tell them where they are located on my tree. Sometimes with the WV cousins, the tree will show only 4 or 5 names on their tree, and I can place them. It's a real kick to write a message and say, "Your mother was my 4rd cousin 1X removed down my Lynch/Gregory Lines and your father was down the same line too."

            Have you had any luck with the ThruLines program. That other company keeps wanting to say that a particular person is my 3rd great grandfather and that the woman they match with him for my 3rd greatgrandmother is married to a totally different man. The men have similar names, Henry Sutton. The way they stick people together in the ThruLines totally confuses me. I have over 30 DNA matches with the descendants of a Benjamin and Jemima Van Wart Sutton, starting at the 3rd cousin level (93 cM) going to distant. But I only have 1 DNA match who traces back to the Henry Sutton that Ancestry keeps showing me. That cousin match is 10 cM and shares no other DNA match with me. I don't know what the program runs on, but I don't think it's logic.
            Last edited by keigh; 11th July 2019, 05:24 PM.


            • #7
              keigh, I am just beginning to sort through and wrap my mind around the ThruLines. They give me lots of ancestors I do not believe fit within my tree. My head wonders if ThruLines is less scientific, and just more speculation. Some of these are surnames that I have never heard of. Sometimes a key/help is if a surname keeps popping up in trees of my matches. I then begin to wonder if that surname should be somewhere in my tree. This has proven to be true; and I have looked at every tree in my account which was very time-intensive, as is all of this! It is akin to a huge jigsaw puzzle.

              And, yes, logic should always be factored in. With your match total, you must be the leader of the pack!

              Genealogy is not a hobby I chose; it chose me!
              Last edited by Biblioteque; 12th July 2019, 09:54 AM.