No announcement yet.

Gedmatch - full siblings, utterly different results? Am I overthinking this?

  • Filter
  • Time
  • Show
Clear All
new posts

  • Gedmatch - full siblings, utterly different results? Am I overthinking this?

    Okay. 1st - please move if this needs to be elsewhere.

    2nd I know gedmatch is autosomal - hence no sex [x/y] linkage.

    However, I am confused with a few things. I have done all the worlds [except for those aimed to those of African descendent as it doesn't apply] and in reviewing it noticed a few things.

    The most predominant is the fact that with particular reference populations I can have a certain number of "matches" [as in if I do a CTRL + F under Y population it shows up X times] have a considerable size difference.

    Welsh for example - shows up 44 times for myself, 16 times for my sister.

    French - 104 times for myself, 72 for my sister.

    I understand that the genes are inherited randomly at best but such things - the Welsh for example - seems to be a heck of a gap. Particularly when viewing the single population % per world the Welsh is closest at 4.12 but can range as far back as 19.56 [in comparison sister's single Welsh % is 5.45 to 21.09].

    And Welsh for neither of us shows up in any of the 2 to 4 population mixes which if I remember correctly goes from self to parents to grandparents on connection.

    On the reverse French is close & shows up in the 2 to 4 population mixes.

    My question is this, how would I calculate the closeness of the sample populations with such size gaps?

    I mean it's simple enough with close populations - German, 29 / 28 respectfully.

    In the Welsh example would I do an average - 44 + 16 = 60 / 2 = 30.

    The reason why I ask, is I want to draw a comparable to the reference populations - to whom are we the most related to - to narrow down possible origin further than we have already.
    FTDNA Customer
    Last edited by James78; 1 August 2015, 03:38 PM.

  • #2
    Good questions. I don't have the answer, but I do recommend getting parents tested if possible. My kids had totally different myOrigins and when you compare them to my and my husband, they begin to make more sense. Then, when I added my parents, that furthered the effort to made sense of the "weird" stuff in the myOrigins. I just added my Grandpa too... it all helps.

    Personally, I don't feel confident on making any comparisons to other people based on the MyOrigins/other ethnicity calculators. To me, they are just predictions/estimations, but I don't feel one calculator can accurately compare people because the reference populations are not equally represented. I made a spreadsheet of everyone in the family and their myOrigins, but also of many of the calculators from gedmatch. They all have similarities and differences in the results..