No announcement yet.

People with shared DNA don't show up "in common with"?

  • Filter
  • Time
  • Show
Clear All
new posts

  • People with shared DNA don't show up "in common with"?

    There are two people I found on Gedmatch who match me (and each other) on a segment of a chromosome. Both are also on FTDNA but when I use the "show matches in common with" for either of them, the other isn't listed. This makes me wonder how many other common matches I'm missing when using this feature, matches which could potentially help me figure out our common ancestry.

    Additionally, I share X-DNA with these two people according to Gedmatch but it doesn't register on FTDNA. I get that many people dismiss a match if there isn't a significant amount of shared DNA but knowing that we share X-DNA, even a small amount, can really help narrow down which branches of our trees we have a shared ancestry from.

    I feel like if it weren't for Gedmatch, a FREE service, I'd be missing a lot of information that FTDNA doesn't provide.

  • #2
    I have seen the same thing. I believe it because FTDNA uses more restrictive filters.

    Off the top of my head i think you have to have at least 20cM of shared and at least one segment that is at least 7cM in length. So if you had a match with only 1 segment that's 19.9cm long, they won't show up in FTDNA. Or if you had many segments and they were all below 7cM then they wouldn't show up either on FTDNA.

    Again I'm just writing this from memory.


    • #3
      Thanks, that's what I suspected. Like I say, I wish they would allow us to choose our own criteria for filtering. Thank God for Gedmatch or I'd be missing a LOT of important data. It's just sad that a company I had to pay for provides less than a free service.


      • #4
        I have often wondered just why IT imposes limits on what a customer can do by means of filtering, etc.

        I must assume that it can only be some sort of money-saving ploy or they just think customers are incompetent or - very likely - they do it because they can.

        I, for instance, hate most PC automatic functions because they are seldom exactly what I want.

        Naturally, nobody in IT will see this unless by accident and then I'll be scoffed at no doubt.


        • #5
          I wish FTDNA would lower the 20 cM threshold for matches by removing the sum total for matching, but maybe use sum total to help determine closeness of said match. I imagine there would be happier customers, and better word of mouth.


          • #6
            Originally posted by MitchellSince1893 View Post
            Off the top of my head i think you have to have at least 20cM of shared and at least one segment that is at least 7cM in length. So if you had a match with only 1 segment that's 19.9cm long, they won't show up in FTDNA.
            What's even worse is when that 19.9 cM segment shows up for Person A thanks to a little extra noise, but it doesn't show up for Person A's mother, even though that's where the match is (as shown at GEDmatch or 23andMe, etc.). That can lead to some very dangerous assumptions about connections to matches.

            I complained about this on FTDNA's Facebook page a few months ago. I got a reply from FTDNA that said the 20 cM minimum didn't exist and was outdated/wrong info from ISOGG. (I hadn't even mentioned ISOGG.)

            All someone has to do is sort their matches by smallest amount shared to see that the 20 cM limit exists, so I don't know why they're pretending it doesn't.

            They did ask me to send in examples, which I appreciate, but so far I've had other uses for my time rather than try to convince FTDNA through multiple ironclad examples with screenshots that they can't dismiss as one-offs, faults of other systems, or whatever BS a person who claims the 20 cM minimum doesn't exist might say. Soon, hopefully, I'll get grumpy enough to put it all together.

            I urge others to complain about this with examples via email or Facebook. (Facebook is nice because it's fully public, but beware a certain fan-customer whose mission in life seems to be to give any constructive critics a completely unrelated lecture on how autosomal DNA works... bleh.)


            • #7
              Need to Adjust Thresholds for Declaring a Match

              I also see quite a need to allow the customer to adjust their own filter settings for declaring a match.
              I have been analyzing a number of matches to family members and then comparing the matches in detail to their parents match list. I find that the actual chromosome locations of the child's match have contributions of DNA from both parents. I suspect that this is true even in the case where only one parent is a match to the same "cousin" as the child. But to confirm this, I need to know if the other parent has contributed DNA in the missing locations even though they may be below the threshold for them to be considered a match to this same "cousin". I cannot get at this data, if that parent is not declared a match.


              • #8
                Probably the ideal would be for FTDNA to use some sort of cutoff for its main match display (thus limiting false positives, etc) BUT to provide Gedmatch-like options so that we can check for smaller matches using kit numbers, etc.

                There have been plenty of times when Gedmatch has revealed information about FTDNA matches that I would never have learned using just FTDNA.

                Gedmatch is wonderful, but it's often down, and more importantly, there's no way to get every match persuaded to use it.


                • #9
                  Considering that all Projects have to have Public Websites -- and that the Public Websites list folks by their Kit Numbers -- ie all the folks in the Project have Kit numbers everyone in the world can see. Then you would have FTDNA give anyone with FTDNA results who asks the info on those Kits, just cause they copied the Kit Number from the Project's public website?

                  I sure hope not!!!