About six weeks ago, Ancestry had a meeting with a few bloggers and educators in the genetic genealogy community and brought us up to speed on a new feature that was upcoming. Ancestry showed us their plans to expand the DNA Circles feature, although to be very clear, to the best of my knowledge, none of us were involved in any type of beta testing with Ancestry.
Today, Ancestry assigns you to DNA Circles based on a combination of your DNA results and your tree, based on common ancestors shown in trees of matching individuals. I wrote about Circles and how they are calculated in the article, “Ancestry’s Better Mousetrap – DNA Circles.”
As an enhancement to DNA Circles, today Ancestry rolled out their new feature which is called “New Ancestor Discoveries” where Ancestry assigns ancestors to you based on DNA matching alone, without matching ancestors in your trees.
And, in my case, they are wrong. Unquestionably wrong. What I hate the most about this situation is if you’re not a genetic genealogist, and you haven’t done your homework, you’ll be thrilled with your new wrong ancestors, “proven,” of course, by DNA.
We received a quick glimpse of the pre-beta product – and truthfully – if this was accurately done and appropriately portrayed as a DNA match with people who shared common DNA and maybe a common ancestor – I could be excited. In fact, I was excited.
I do believe this type of matching can be done accurately – but Ancestry has missed the mark – not just with me but from other early reports in the community as well – with lots of people. Portraying this match as a “new ancestor” is wrong and it’s terribly misleading.
Here’s what Ancestry has to say about the New Ancestors matching.
Ok, what does Ancestry have to say about Diedamia Lyon, my New Ancestor who is not my ancestor?
Clicking on the green “Learn About” button shows me the “facts” that ancestry has gleaned from their trees about Diedamia Lyon.
What this tells you that isn’t immediately evident is that Diedamia Lyon was married to John Curnutte, my second “New Ancestor.” There is a “Facts” tab that shows you the sources that Ancestry used to create Diedamia’s story. They have used compiled data from 215 trees. I cant’ speak for Diedamia, but I know several of my Circle Ancestor’s stories are wrong – based on the compiled trees – substantially wrong in fact. Because the trees are wrong.
So, in essence, Ancestry is saying that I descend from both Diedamia Lyon and John Curnutte, an ancestral couple. This would be invaluable, if it were accurate. Ok, how did Ancestry connect those dots to arrive at that conclusion?
Clicking on the “See Your Connection” button under the Circle icon shows you the members of the Diedamia Lyon Circle.
I have DNA matches with Don and Michael who are members of the Diedamia Lyon circle. Clicking on Don, I can see that he has DNA matches to Michael and three other individuals who I don’t have DNA matches with in the Diedama Lyon circle. However, all of those individuals also share a pedigree chart and Diedamia Lyon is their shared ancestor.
I can click on any of these people and see who they match in the circle, or I can see a list.
What I can’t see is how Ancestry drew those DNA conclusions. There are no tools, no chromosome browser, and obviously, “trust me” isn’t working.
I want to share with you how I know, beyond any doubt, that Diedamia Lyon and John Curnutte are NOT my ancestors. I am a long-time meticulous researcher. I would invite you to search for any of my ancestors’ names on this blog. I have been writing about one ancestor per week now for more than a year in the 52 Ancestors series and, if I have written about them, you can see the types of information we have on each one. I know which of my ancestors are proven and which are questionable. So, let’s see why Diedamia and John cannot be my ancestors.
First, we can eliminate my mother’s line. My mother’s ancestors are from Holland, Germany, Canada/Acadia and one line from Vermont/Connecticut. They are all accounted for and I know where they were, shown below.
The 6th generation shown above is the generation into which Diedamia Lyon and John Curnutte, both born about 1800, would fall. Mother’s generation 6 ancestors, at the far right, were all born between 1766 and 1805, many in Europe. You’ll note there are no blank spaces for missing ancestors and the geography is not southern – meaning no place near Wilkes County, NC where Diedamia was born in 1804. So, my mother’s side is immediately eliminated.
My father’s side, however, does have several lines that come through Wilkes County, NC and many other southern lines. So the connection would be through my father’s side of the family.
Again the 6th generation would be where Diedamia Lyon and John Curnutte would have to fit if they are my ancestors, and there are no blank spaces here either. All of these ancestors were born between 1759 and 1804.
Of the above generation 6 ancestors, the following have a Wilkes County connection:
- Elijah Vannoy born in Wilkes County about 1784
- Lois McNiel born in Wilkes County about 1786
- William Herrell born about 1789 in NC, possibly Wilkes County where he married in 1809
- Mary McDowell born 1785 NC, possibly Wilkes County where she married in 1809
Looking at the pedigree chart of William Herrell and Mary McDowell, you can see that indeed there are some unknown wives. John Herrell was born in about 1760, possibly in Frederick Co., VA and Michael McDowell in 1747 in Bedford Co., VA. While the connection may be through these lines, it’s clearly not from any two people born in 1800 and is at least in the 7th generation – IF the connection is through these lines. At this point, this is the most likely connection because it’s in the right location and there are two unknown wives. If I had triangulation tools, I could probably tell you immediately.
Now let’s look at the pedigree chart of Elijah Vannoy and Lois McNiel, also from Wilkes County.
As you can see, this pedigree is even more complete than the Herrell/McDowell pedigree. Not only is there no room for a couple born circa 1800, there are no unknown parents for another 3 generations prior, not until the 9th generation. The only individual here through the 8th generation not proven via both paper and genetics, meaning triangulation, is Sarah Coates.
So, not only are Diedamia Lyon and John Curnutte NOT my ancestors, it’s very unclear how they are related to me, IF they are related to me. It’s obvious that the only way we are related is that someplace upstream, I do share a common ancestor with both Don and Mike who share the Lyon/Curnutte tree with each other and several others as well, but that does NOT mean that I descend from Diedamia and John, nor that I share a common ancestor with them.
Now, if I share the SAME DNA segment with Don and Mike that could be triangulated to the Curnutte/Lyon descendants, then that would mean we do all share a common ancestor someplace along the line. But wait – Ancestry doesn’t use triangulation – nor do they give us the tools to do so. So we have NO idea if we actually share the same DNA segments or not.
So, let’s take a look at the trees of both Don and Mike to see if we share any common surnames that might be linked.
Fortunately, Ancestry does provide an easy way to do this. By clicking on your matches name to the right of the circle, and looking at their tree, Ancestry shows you the common surnames.
By clicking on the shared surname, you can see the people in both trees, theirs and yours, with that surname, side by side.
All three of us have a dead end Moore line. That is our only other surname in common, and Moore is very common.
So, it’s possible, given that we have no way to tell which segments are matching whom, that I match both Don and Mike through an entirely different ancestor, or ancestors, known or unknown. It’s also possible that someone upstream of Diedamia and John is a child of one of my unknown lines, and while Diedamia and John are not my ancestors, I do carry some of the same DNA as their descendants because we all share a common, unknown, ancestor. But I have no way of knowing.
What I can do is to contact my two matches and see if they will download their DNA to GedMatch where I can get at the truth via triangulation. It’s a shame we have to do that.
So, what is the net-net of this new tool?
- Ancestry missed, big time, especially by labeling the match as a “New Ancestor.”
- Ancestry can salvage the situation at least somewhat by renaming the “New Ancestor” something like “Common DNA Match.” This would alert people that there is some common ancestry someplace, but not mislead people into thinking that Ancestry really HAS discovered a new ancestor or ancestral couple. In some cases the named couple MAY be ancestors – but that’s certainly not always the case. And I don’t like the label “Potential Ancestor” either because I think it implies a much closer relationship than may be present. I remember how completely thrilled I was to see my “New Ancestors” names and without having enough experience to piece the puzzle together, both genealogically and genetically, I would never have known enough to be as disappointed as I am. I feel terribly sorry for the many people who will take this erroneous information as gospel – and the rest of us who will have to live with the incorrect fallout – forever. This amounts to a new way to create an incorrect ancestor and Heaven forbid, attach them to your tree.
- This would all be a moot point with a chromosome browser, but then again, Ancestry already knows that.
And I was so hopeful….
Fortunately, the New Ancestors feature is still in beta and changes can be made – and I hope they are. I know Ancestry has already incorporated at least one the suggestions made as a result of the meeting a few weeks ago.
As I looked back over the new features and the information I received from Ancestry, I am especially concerned about the verbiage accompanying this information.
Here’s what greets me on my DNA page.
Here’s the e-mail I received.
The problem is – it’s just not true. These matches may be valuable in some cases. But they are not as represented. This match is not my ancestor.
So yes, I do want Ancestry to “Show Me.” Show me the chromosomes. Show me how Diedamia Lyon and John Curnutte are my ancestors. Show me how you put 2 and 2 together and came up with this. Show me.