Announcement

Collapse
No announcement yet.

No match big Y is useless

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

  • #16
    Originally posted by John McCoy View Post
    For me, the lack of close matches on FTDNA even before the "reprocessing" and also on YFull has been useful. It is now clear that my McCoy family is not related to the other McCoy's who have been tested, and that I can safely ignore them in my research! But the McCoy families from Brown County, Illinois that I would really like to compare myself with haven't been tested.
    I'm confused. Shouldn't the STR results already given you this information?

    Comment


    • #17
      I am not sure it is useful to argue, even assuming that one admin here reports to FTDNA 'customers wish this or that'.
      Do they care ?

      Quantity is a quality by itself.
      Such as for example, 10 matches at GD=40 SNP (that you won't get now) all spread around Poland are well worth 1 single match at GD=10 (that you won't get either) located in Warsaw, Poland.

      With GD=40 matches, you can add statistical weight to a geographic location, or a name, and then when the (Location X Name) pops up as a STR match (more likely than a BigY, let's get real) contact the person, ask to join a group and only then, this person may take a BigY.
      The idea that you can get close GD=10 BigY match straight, by just waiting, is like winning the lottery.

      In fact it is worse than that. Most people still test with an autosomal kit and when you get a favorable statistical (gepography X name) hit, you have to beg the person to take a STR test, because obviously nobody is going to take an expensive BigY test that brings potentially NO meaningful information (Yeah you are BY1843567 !) and No match. STR tests are still the best compromise to satisfy somebody's investment and keep the person interested, which it is all about.

      Comment


      • #18
        Originally posted by Armando View Post
        How in the world is anyone supposed to know if a match is going to show up without them getting tested first? How do you know if you will not get a significant match in the next few years? How do you feel about telling potential matches that they should not get a BigY test which could in turn cause you to not get a match for a lot longer time from now than would occur without the complaint?
        You can reasonably guess based on your haplogroup. Lots of Western Europeans test, so if you are R1b, I'd be surprised if you don't match. But how many native Americans, obscure African lineages, and other "interesting" results are going to have matches? Almost none, and these are the people who will advance science by testing.

        Comment


        • #19
          BTW - I have 39 unnamed variants. So it is unlikely I will see a match any time soon.

          Comment


          • #20
            Originally posted by Alhan View Post
            Great explanations thanks. I've taken the quick road from YDNA12 to BigY in August sale, I was pretty much clueless what's been going on.

            I now have 24 unnamed variants. However, I did have some matches before HG38 conversion. In fact my terminal SNP (R-FGC64132) formed after my results (my little contribution to Geneology I guess!)

            From your explanations, I assume my fellow who is sharing my terminal SNP had less than 16 unnamed variants, so we were matching on BigY. (considering old 40 unnamed variants threshold)
            The threshold wasn't added until the hg38 upgrade. They officially state that the cutoff is 40 non-matching variants but it is actually 30 non-matching variants because people that should have matches with 35 or 39 non-matching variants don't get those matches. So no, the fellow sharing your terminal SNP did not have less than 16 unnamed variants.
            Last edited by Armando; 8 January 2018, 08:56 AM.

            Comment


            • #21
              Originally posted by bvbellomo View Post
              You can reasonably guess based on your haplogroup. Lots of Western Europeans test, so if you are R1b, I'd be surprised if you don't match. But how many native Americans, obscure African lineages, and other "interesting" results are going to have matches? Almost none, and these are the people who will advance science by testing.
              Yes, a lot of R1b people can expect to find matches and should get BigY testing in case they do but, using your own example, not all R1b people will get matches so YMMV. If they don't test there is no way to know if they will get matches. The only way to know for sure is to test. Even if no matches show up immediately there is a possibility that a match will show up in the not too distant future.

              Comment


              • #22
                Originally posted by bvbellomo View Post
                But how many native Americans, obscure African lineages, and other "interesting" results are going to have matches? Almost none, and these are the people who will advance science by testing.
                Actually all people that get a BigY test can help create new branches that were previously unknown. That is how four SNPs previously thought to be phylogenetically equivalent to L21 were proven to be upstream of L21. http://ytree.net/

                Comment


                • #23
                  Originally posted by Armando View Post
                  Actually all people that get a BigY test can help create new branches that were previously unknown. That is how four SNPs previously thought to be phylogenetically equivalent to L21 were proven to be upstream of L21. http://ytree.net/
                  So we will see new branches popping up next to ours, or even below our own branch, but without names and locations associated to these branches, when the people are not matching.

                  In the example I mentioned above, a BigY kit was in a new branch below mine, but still too old for us to match with the new 30SNP stringency. It is lucky that another BigY kit matched this person so they could form a new branch, otherwise, if that person didn't match me with the new stringency, he would not have matched anybody else and stayed invisible for everybody, including the groups.
                  But that was 2 years ago and thankfully, we matched, and I am the one who take credit for contacting the person and asking him to join the relevant group.

                  So that now I am surrounded by very precious little flags, thanks to the Groups and the Big Tree, but none of them I will match now (I have 25 unnamed variants) and at least one would have been absent in the new system.

                  That raise the question of how many people take a BigY but don't join their Y-haplogroup group automatically. Probably a minority, but still, you don't want to loose anybody such is precious any information.

                  So now the new stringency is truly ~ 15 new unnamed SNPs, since we have to consider the other dude's unnamed SNPs.
                  That's what ? a 1500 to 2000 years old ancestor. The chance that one male descendant from this ancestor takes a BigY, are not great outside of the British isles and colonies.

                  Comment


                  • #24
                    You might have a different take on the matter if you were named Buchanan! A large number of Buchanans have taken the Big Y test, and it is now clear that they fall into distinct groups defined by multiple new SNP's that must reflect their ancestry and history. You can see this interesting result on Alex Williamson's "The Big Tree" website ytree dot net, look for the R-L1335 group.

                    Comment


                    • #25
                      Originally posted by dtvmcdonald View Post
                      But no matches at all is showstopper.
                      Just checked for my Big Y matches -- 0, none. I'd understand not matching on my terminal SNP (BY4064); it's bound to be a small group. But that's a couple of levels below Z253, a larger group. I just checked R-Z253 project stats; they have 822 members, ~380 with Big Y. Surely, at least some of the 380 are Z253+.
                      If the feature is non-functional now, there should be some notice.

                      Comment


                      • #26
                        Originally posted by rt-sails View Post
                        Just checked for my Big Y matches -- 0, none. I'd understand not matching on my terminal SNP (BY4064); it's bound to be a small group. But that's a couple of levels below Z253, a larger group. I just checked R-Z253 project stats; they have 822 members, ~380 with Big Y. Surely, at least some of the 380 are Z253+.
                        If the feature is non-functional now, there should be some notice.
                        You have to have 30 or fewer non-matching variants with people that are positive for R-Z253 in order for you to have a match. When I look at BY4064 at http://www.ytree.net/DisplayTree.php?blockID=2749 there are at least 20 variants downstream from BY4064 which means that you have about that many more variants and so does anyone else that is positive for BY4064 which amounts to about 40 non-matching variants. So there needs to be someone that matches you on about another 5 variants downstream from BY4064 at which point some of your unnamed SNPs variants will get an SNP name and will be added to the FTDNA haplotree and your terminal SNP will be one of those newly named SNPs.
                        Last edited by Armando; 25 January 2018, 08:13 AM.

                        Comment


                        • #27
                          It doesn't help when somebody says 'it works for me so who cares about the others ?'

                          15 novel variants is ridiculously small. It could refer to a known SNP/ancestor 1500 to 2000 years old and who said we are not interested in older ancestors ?
                          Anyways I have complained enough, a few here seem to agree that NO BigY matches is more damaging than a lot of irrelevant BigY matches (that can be easily filtered), but we don't seem to get any traction.

                          Comment


                          • #28
                            I think we are losing this battle, Admins in various groups stay silent about it and few people show the will to complain. Basically FTDNA can do whatever they want, once the money is spent, why should they care ?

                            Group admins should be more concerned. Since now we are entirely dependent on our groups to see who is in our subclade, every time a new members will pop up in our subclade we will ask the admin for name (since the ancestor's name is often not the same as the tester's name), background info and email. Multiply that by the number of members and the admins could be overflowed with requests.
                            Is that what you want, admins ?

                            Anyways, using the number of different novel variants SNPs to define who should be a BigY match doesn't make the slightest sense. When you are positive for a SNP that defines a branch/subclade, everybody under it, who is positive for the said SNP should be a BigY match, regardless of the number of different unnamed SNPs. It could be 30, it could be 40 or 50, it doesn't matter. The ancestor is the same. Some particular lineages could experience a slightly higher mutation rate and slightly shorter generations time in average, like 25 years versus 27 years and that could easily make up for the difference in SNPs.

                            Somebody explains to me why that total number of different SNPs should be 30.

                            Comment


                            • #29
                              Originally posted by Ric View Post
                              I think we are losing this battle, Admins in various groups stay silent about it and few people show the will to complain. Basically FTDNA can do whatever they want, once the money is spent, why should they care ?

                              Group admins should be more concerned. Since now we are entirely dependent on our groups to see who is in our subclade, every time a new members will pop up in our subclade we will ask the admin for name (since the ancestor's name is often not the same as the tester's name), background info and email. Multiply that by the number of members and the admins could be overflowed with requests.
                              Is that what you want, admins ?

                              Anyways, using the number of different novel variants SNPs to define who should be a BigY match doesn't make the slightest sense. When you are positive for a SNP that defines a branch/subclade, everybody under it, who is positive for the said SNP should be a BigY match, regardless of the number of different unnamed SNPs. It could be 30, it could be 40 or 50, it doesn't matter. The ancestor is the same. Some particular lineages could experience a slightly higher mutation rate and slightly shorter generations time in average, like 25 years versus 27 years and that could easily make up for the difference in SNPs.

                              Somebody explains to me why that total number of different SNPs should be 30.
                              As a project admin of several larger projects we are not silent as we realize there are squeaky wheels that need to vent. We would prefer that FTDNA invest their resources into finishing off the hg38 upgrade process and delivering the BAMs over being diverted making some enhancements which will make it harder for the bulk of users to properly interpret their results.

                              BigY SNP matching criteria are similar to the STR GD distance criteria. A MEANINGFUL cutoff needs to be in place to help less knowledgeable clients to understand the LIMITS of their results. With the BigY SNP limit that essentially means that an individual isn't going to share an ancestor with a matching result within the last 2000 years. If one wants deeper ancestral information associated with BigY matches that is where the haplogroup project grouping comes in to help provide that definition.

                              The correlation between the relevance of BigY matches and STR GD could vary some across the major haplogroup branches.

                              Why should the majority of the BigY test clients care to understand and work with matches where their shared ancestor was >2000 years ago?

                              Comment


                              • #30
                                Originally posted by wkauffman View Post
                                Why should the majority of the BigY test clients care to understand and work with matches where their shared ancestor was >2000 years ago?
                                It is totally inappropriate for FTDNA to decide for us what we should be interested in and what we should not.
                                I am interested in all my ancestry and I bet many others are. Why shouldn't we ? make a poll...

                                Also, who were our ancestors 3000 years ago(I mean their descendants as matches, you get that) provides information about where/who they could be 2000 years ago.
                                This is important precisely for those who don't get matches from a <2000years common ancestor, and who will never get any. Usually from countries who don't test much : France, Germany (all German matches are colonials), Belgium, Central Europe, Eastern Europe all the way to Russia. This is a big chunk. The chances of finding a match with a common ancestor younger than 2000 years from these areas is null.

                                With a bunch of 3000 years matches you can contact by email, make correlations and perhaps infer who or where are potentially interesting people who should test, who may already be STR match at a level insignificant, but good candidates for a nearby clade in BigY testing... and you get closer and closer. If you know nothing, nothing will happen.

                                But I concede that a 2000 years old clipping threshold is all good for the Anglo-sphere customers with a heavy testing demographic.

                                Comment

                                Working...
                                X