I recently completed a autosomal DNA through ancestry. They reported that I had a possible match to another ancestry user with:

30.0 centimorgans shared across 1 DNA segment

However, it turns out that while we are related by marriage we don't appear to be related by "blood".

The only relationship we can detect is that my great grandfather is the other person's father-in-law of their great grandfather.

Assuming that our record genealogies are correct and that there was no "hanky-panky" it would seems that the suggestion of a DNA match is incorrect.

How often might this occur, and is it assignable to "experimental error" in testing? Other explanations?

  • 1
    Please provide more information about the match: the total amount of shared DNA and the longest shared segment. I ask because it is very important. If the numbers are low there is no confirmed relationship (but it may still exists and the relation is 5th or more cousins). Commented Jan 4, 2018 at 22:18
  • only info available: "30.0 centimorgans shared across 1 DNA segment" -
    – BobE
    Commented Jan 5, 2018 at 0:47
  • 1
    Are you only using AncestryDNA? Have you and your match downloaded your raw data and uploaded it anywhere you could use third-party tools?
    – Jan Murphy
    Commented Jan 5, 2018 at 1:05
  • 1
    @Jan Murphy Yes only Ancestry and No, I have not, and I suspect my "match has not.
    – BobE
    Commented Jan 5, 2018 at 3:31

1 Answer 1


You said that you share 30 cM over 1 segment.

A single segment of DNA that size could quite possibly have come from a common ancestor 10 generations back.

10 generations likely exceeds the limits of your genealogical research on many of your lines. So there could be a connection via that one segment that you are not aware of.

Matches of 15 cM or less may be by chance, but the likelihood of a by chance match for a 30cM segment is virtually zero.

If the connection is 10 generations back, you have up to 1,024 ancestors to check against up to 1,024 ancestors of the other person which is a million possible combinations. And what if it's 12 generations back, not 10?

So basically what I'm saying is not to expect that you'll be able to find the common ancestor.

Even so, you can narrow down the side that the person is connected on by triangulating that segment with overlapping segments belonging to other people.

  • 1
    Always possible could have come from a much earlier generation, but before I get wrapped up in trying to uncover common relations prior to 1750..... what are the possibilities that the "match" of that size is just random error?
    – BobE
    Commented Jan 5, 2018 at 4:58
  • The greater issue is that this segment could be compound or IBS (identical-by-state). To detect it one needs to make so-called phasing procedure. Which will separate matches from fathers's and mother's sides. Commented Jan 5, 2018 at 13:34
  • 1
    The chance of a random match of 30 cM is virtually zero. Unfortunately , no. I have a such one. And it was eliminated by phasing procedure Commented Jan 5, 2018 at 13:35
  • I've added to my answer to address the comment of BobE. And @George - I'd trust a 30 cM segment more than I would trust phasing. Phasing isn't perfect and there are no calls, misreads and mutations that can throw it off.
    – lkessler
    Commented Jan 5, 2018 at 13:51
  • @GeorgeGaál - I would be very interested in looking at the raw data behind your 30 cM match to understand why you eliminated it through phasing. Feel free to email me: lkessler at lkessler dot com
    – lkessler
    Commented Jan 5, 2018 at 16:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.