Announcement

Collapse
No announcement yet.

Question about differing match results between my Mother & I; why the huge disparity?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by NYMark View Post
    That's very interesting, and I wasn't aware that number of blocks were counted, and that number does seem to be very significant. It would make sense if it were done differently, but the total includes numerous blocks that are surely noise.

    I think there's something else in the mix too. Looking at three of my closest AJ matches, one is projected as 3rd and the other as 4th.

    "3rd Cousin" - Longest Block: 21.46
    Total Shared: 110.13
    Total Segments: 27
    Second Longest: 10.27

    "3rd Cousin" - Longest Block: 18.11
    Total Shared: 94.81
    Total Segments: 22
    Second Longest 9.74


    "4th Cousin" - Longest Block: 21.91
    Total Shared: 91.86
    Total Segments: 22
    Second Longest: 7.89

    Second longest seems to have been the deciding factor in designating one as 3rd and the other as 4th. I suspect the "4th Cousin" is actually closer.

    In looking at some of my other "4th cousins", it appears that the algorithm takes a number of factors into account: longest block, total shared, number of blocks, and second longest. Thus, I have someone who shares 11 blocks total but is ranked a page higher than someone who has 16 total.

    FWIW, my lowest ranked "4th Cousin" has a longest of 7.27, a total shared of 125.03, and 34 blocks. Only 8 of the 34 are above 5 cM.

    I just noticed Tomcat posted the same as me that they count blocks before cm. He posted a way to change it so I have to figure that out. The way they have their matching system set up our eyes are trained to look at cm first before block since cm is listed first.

    Im seeing a pattern in our different kits. The closer matches have shorter blocks and are not listed first because of the blocks

    Here is that match I was talking about that shows up in the middle of my distant relatives. They have not emailed me back so I cant say if the prediction of closeness is accurate without knowing who is on their tree. This is my closest match to date outside known family. Same with my sister she has this match too listed at the middle of her pages

    2nd Cousin - 4th Cousin 3rd Cousin CM 87.34 longest block 19.55

    __________________________________________________ _____________

    This is my uncle moms side. This is his closest match to date and also listed in the middle of the page because of shorter block. They were our neighbors and we have an NPE so not sure what ancestor we share.

    2nd Cousin - 4th Cousin 3rd Cousin cm 82.36 block 29.54


    This is my uncles 2nd closest match and they are listed as top distant match This person and us share two 3rd great grandparents and two 4th great grandparents and anyone behind them

    2nd Cousin - 4th Cousin 3rd Cousin cm 66.20 block 30.33

    Comment


    • #32
      OOPs

      I forgot about this match. This is my uncle's closest match outside our family. listed again in the middle of the page. We share same surnames but cant connect on paper. They did come from our area. I dont know how match 2 connects to match 1 since no one is answering emails.


      2nd Cousin - 4th Cousin 3rd Cousin cm 90.89 block 19.92

      and this match too who also matches the match above but on different chrome in different locations

      2nd Cousin - 4th Cousin 3rd Cousin cm 64.66 block 19.91

      Comment


      • #33
        I think you're misunderstanding. You can't change the calculations, but you can sort your matches in different ways by using the menu bar at the top of your Family Finder matches page. For the examples I provided, I sorted by "Suggested Relationship", so the order goes from closest to most distant in FTDNA's estimation.

        You can also sort your matches using other criteria: match date, relationship range, longest block, and total cM, and you can go from most distant to closest or most recent to oldest, etc.

        To find the total number of blocks, you have to go to Chromosome Browser.

        They're not counting total number of blocks first, else my most distant "4th cousin" would be much higher up on the list. I haven't looked at them all, but I think 34 blocks is the most for any of my matches. They seem to be using a mix of criteria, the specifics of which are proprietary, but it's possible to tease out some hints.

        Originally posted by Yaffa View Post
        I just noticed Tomcat posted the same as me that they count blocks before cm. He posted a way to change it so I have to figure that out. The way they have their matching system set up our eyes are trained to look at cm first before block since cm is listed first.

        Im seeing a pattern in our different kits. The closer matches have shorter blocks and are not listed first because of the blocks

        Here is that match I was talking about that shows up in the middle of my distant relatives. They have not emailed me back so I cant say if the prediction of closeness is accurate without knowing who is on their tree. This is my closest match to date outside known family. Same with my sister she has this match too listed at the middle of her pages

        2nd Cousin - 4th Cousin 3rd Cousin CM 87.34 longest block 19.55

        __________________________________________________ _____________

        This is my uncle moms side. This is his closest match to date and also listed in the middle of the page because of shorter block. They were our neighbors and we have an NPE so not sure what ancestor we share.

        2nd Cousin - 4th Cousin 3rd Cousin cm 82.36 block 29.54


        This is my uncles 2nd closest match and they are listed as top distant match This person and us share two 3rd great grandparents and two 4th great grandparents and anyone behind them

        2nd Cousin - 4th Cousin 3rd Cousin cm 66.20 block 30.33
        Last edited by NYMark; 26 March 2013, 09:36 AM.

        Comment


        • #34
          I do understand you. I have not changed the sorting on my matches and Im guessing neither has the op of this thread and why the op's and my matches are showing up in the middle of the pages when they should read closer. FTDNA should automatically have the matches sorted by cm first rather than block. Thats what the confusion is. Im sure others might be missing their closer matches if they dont change how their matches are sorted. Most people will go to look at cm before block
          Last edited by Yaffa; 26 March 2013, 09:41 AM.

          Comment


          • #35
            I also had Ann Turner read our raw data on my uncle's closest match that we cant connect paper to. The 90 cm match ( it was lower before upgrade) She explained to me that your closer matches will match on more than one chrome but could appear closer if you happen to share more than one ancestor back there. Example our cousin and my uncle who share two 3rd great grandparents and two 4th great grandparents. the share a total of 60 cm. To date we don't know if this DNA could be coming from just one ancestor or all 4 ancestors they share.

            Comment


            • #36
              Then I misunderstood you. I apologize.

              I'd prefer the default to be "Suggested Relationship." The problem with sorting by total cM first is reflected in that projected 4th cousin who shares a total of 125 cM and has a longest of 7.27. I've been at this for a while, so I know my way around, but if I were a newbie and that came up at the top of my list, it would be very confusing. Things might be different if they stopped counting the under 5 cM segments.

              Originally posted by Yaffa View Post
              I do understand you. I have not changed the sorting on my matches and Im guessing neither has the op of this thread and why the op's and my matches are showing up in the middle of the pages when they should read closer. FTDNA should automatically have the matches sorted by cm first rather than block. Thats what the confusion is. Im sure others might be missing their closer matches if they dont change how their matches are sorted. Most people will go to look at cm before block

              Comment


              • #37
                Originally posted by NYMark View Post
                Then I misunderstood you. I apologize.

                I'd prefer the default to be "Suggested Relationship." The problem with sorting by total cM first is reflected in that projected 4th cousin who shares a total of 125 cM and has a longest of 7.27. I've been at this for a while, so I know my way around, but if I were a newbie and that came up at the top of my list, it would be very confusing. Things might be different if they stopped counting the under 5 cM segments.
                Thats ok no need to apologize. For the newbee I just think they should do automatic cm match first before block until people learn their way around. Im not a newbee but I did not realize they were matching by block first so I almost missed closer matches but caught on to what was going on.

                I also wish they would lower the threshold for matches. My sister had one match at FTDNA that did not match me. On gedmatch she not only matched me on autosomal but also X for both my sister and myself. This person was an adoptee and figuring out that match was on my father's mother's side of the family was helpful to all of us.


                All my uncle's closer matches all match me accept the one confirmed cousin who shares 4 + ancestors with us. I suspect if they lowered the threshold she might match me. I need this cousin to upload to gedmatch to see if we could match or I just did not inherit any of that DNA

                Comment


                • #38
                  Originally posted by Yaffa View Post
                  2nd Cousin - 4th Cousin 3rd Cousin CM 87.34 longest block 19.55

                  2nd Cousin - 4th Cousin 3rd Cousin cm 82.36 block 29.54

                  2nd Cousin - 4th Cousin 3rd Cousin cm 66.20 block 30.33

                  2nd Cousin - 4th Cousin 3rd Cousin cm 90.89 block 19.92

                  2nd Cousin - 4th Cousin 3rd Cousin cm 64.66 block 19.91

                  90.89cm = 4th Cousin
                  87.34cm = 4th Cousin
                  82.36cm = 4th Cousin
                  66.20cm = 5th Cousin
                  64.66cm = 5th Cousin

                  Comment


                  • #39
                    Originally posted by NYMark View Post
                    Then I misunderstood you. I apologize.

                    I'd prefer the default to be "Suggested Relationship." The problem with sorting by total cM first is reflected in that projected 4th cousin who shares a total of 125 cM and has a longest of 7.27. I've been at this for a while, so I know my way around, but if I were a newbie and that came up at the top of my list, it would be very confusing. Things might be different if they stopped counting the under 5 cM segments.
                    125cM = 4th Cousin

                    Comment


                    • #40
                      I don't understand the purpose of posting these numbers without comment. They're nonsense when they're presented in isolation, and as I've explained repeatedly, counting segments of 1-5 cM frequently leads to inflated estimates of proximity, especially in endogamous populations. 23andMe doesn't count anything under 5; Gedmatch goes down to 3.

                      If I look at my non-AJ projected 4th cousins, the range of total shared is 20.18-40.52. Longest blocks range from 18.81 to 21.96. The one who is projected as being the closest is the one with 21.96 and 40.52. Only one of her other shared segments is over 3 cM, 3.29. She is probably in the 6th-8th range. 23andMe has her sharing one segment and projects her as 3rd-6th. There's no paper trail for me, but triangulation with another match suggests that the common ancestor was born no later than the early 18th-century and probably before that.

                      The range for AJ or partial AJ is 40.68 to 125.03. I suspect that some of the matches in the low end of the range may not be getting adjusted. The longest blocks range from 7.23 to 21.91.

                      It's possible, but unlikely, that the 125 cM individual is a 4th cousin. He does have a common ancestral surname, and his ancestry is German Jewish unlike the majority of my so-called 4th cousins, but over 75 cM of the total shared are segments of under 5 cM.

                      Originally posted by madman View Post
                      125cM = 4th Cousin

                      Comment


                      • #41
                        Originally posted by madman View Post
                        90.89cm = 4th Cousin
                        87.34cm = 4th Cousin
                        82.36cm = 4th Cousin
                        66.20cm = 5th Cousin
                        64.66cm = 5th Cousin
                        It does not always work that way. the 66 cm is confirmed cousin who shares 4 plus direct line ancestors. making them like 3rd cousins genetically because of double relations even though the ancestors they share are further back than 3rd cousin. FTDNA does have them predicted that way too as 3rd cousin.

                        Comment


                        • #42
                          Originally posted by Yaffa View Post
                          It does not always work that way. the 66 cm is confirmed cousin who shares 4 plus direct line ancestors. making them like 3rd cousins genetically because of double relations even though the ancestors they share are further back than 3rd cousin. FTDNA does have them predicted that way too as 3rd cousin.
                          You stated yourself the connection is further back then 3rd cousin.


                          Originally posted by Yaffa View Post
                          making them like 3rd cousins genetically because of double relations even though the ancestors they share are further back than 3rd cousin.
                          It doesn't matter how many double relations you have if it not a 3rd cousin or closer you cant get a 3rd cousin.

                          You can have 4th cousin and 4th cousin marry there kids will be a 4th cousin instead of a 5th cousin.

                          now if its a 4th cousin and 2nd cousin then there kids will be a 3rd cousin.

                          but if there all further back then a 3rd cousin its impossible to be a 3rd cousin.

                          Comment


                          • #43
                            The key point here is "like 3rd cousins genetically". No, they are not 3rd cousins, but the amount of shared DNA they have is compounded by (somewhat) close cousins marrying.

                            Also, you're off on your cousin #s. Two fourth cousins marrying and having a child will give you a 4th cousin once removed, or fifth cousin to your child. Genetically, they are more likely to have more shared DNA with you than if one fourth cousin of yours married someone unrelated.

                            Comment


                            • #44
                              Originally posted by sjadelson View Post
                              The key point here is "like 3rd cousins genetically". No, they are not 3rd cousins, but the amount of shared DNA they have is compounded by (somewhat) close cousins marrying.

                              Also, you're off on your cousin #s. Two fourth cousins marrying and having a child will give you a 4th cousin once removed, or fifth cousin to your child. Genetically, they are more likely to have more shared DNA with you than if one fourth cousin of yours married someone unrelated.

                              Actually I had it right two 4th cousins who had kids would have same amount of dna as a 4th cousin.

                              Originally posted by Yaffa View Post
                              my uncle who share two 3rd great grandparents and two 4th great grandparents
                              now Yaffa has a 4th cousin and a 5th cousin who had a kid and calling it a 3rd cousin thats impossible it would be a 4.5 cousin and still register at the high end of a 5th cousin with dna.

                              Comment


                              • #45
                                Originally posted by sjadelson View Post
                                The key point here is "like 3rd cousins genetically". No, they are not 3rd cousins, but the amount of shared DNA they have is compounded by (somewhat) close cousins marrying.

                                Also, you're off on your cousin #s. Two fourth cousins marrying and having a child will give you a 4th cousin once removed, or fifth cousin to your child. Genetically, they are more likely to have more shared DNA with you than if one fourth cousin of yours married someone unrelated.
                                yes like 3rd cousins genetically. Though it's not an issue of close cousins marrying.We happen to share 4 direct line ancestors and their direct line ancestors behind them

                                we also have a match at ancestry to a separate line that has not tested here. 2 brothers of one family married two sisters of another family. Their children double 1st cousins. I dont have their raw data but Im sure this match also reads closer but I dont have cm amount.

                                in general if there is incest in a family or closer cousins marry, they will be genetically closer related. If you share several direct line ancestors you will be genetically closer related.

                                Comment

                                Working...
                                X