First of all, let me say that I like AncestryDNA Circles. Yes there are problems and still things to work out, but all said, I think they have the potential to be beneficial – so long as they are interpreted correctly.
Much of the grumbling about Circles comes from the fact that Ancestry promised the community a “better mousetrap,” and instead, released DNA Circles. DNA Circles is just fine for what it is, but it’s a far cry from a better mousetrap – meaning it’s not a replacement for, nor better than, a chromosome browser. It’s like being promised a “better car” and getting a bicycle while the “gifter” is trying to convince you it’s a Mercedes bicycle and you should love it, plus you don’t need a car anyway – you only think you do.
The biggest problem with DNA Circles is that people perceive Circles to be proof, when they aren’t – partly because of how the DNA is paired with no matching segment proof (chromosome browser) available – and partly because of inaccurate trees. That said, twice now I’ve found a really good hint through Circles, even if one did subsequently disappear. For that very reason, I check every single day.
But hear that word – HINT.
What is a DNA Circle and How Does It Work?
Just to be sure we’re all on the same page, let’s take a quick look at what a DNA Circle is.
Circles are created around a specific ancestor. To be included in a DNA Circle, you must match at least one other person’s DNA who shows that same ancestor in their tree, plus there must be at least another person who matches your DNA or that of the person’s DNA you match, and also has the same ancestor in their tree. So, a minimum of three people.
Now, we don’t know if these people match on the same segment, or if they also share other ancestors that might cause DNA matching, but the more people who match each other (and you) in the Circle, the better chance there is that there is a connection through that ancestor.
Let’s take a quick look at segment matching and how it’s done from Ancestry’s document titled, “Do all members of a DNA Circle have the same matching segment?” This paper along with others is available on your DNA results page by clicking on the question marks on the right hand side beside your DNA Circles header. So, no DNA Circles, no help information about DNA Circles.
In the segment matching example, below, you can see that the DNA of these three people shown do share DNA segments, but they are not triangulated. A matches with B and C on different segments, and B matches with C and A – but again, not on the same segments. A triangulated match that proves a connection through a common ancestor requires 3 individuals to share the same DNA segment.
If these three segments below were triangulated, you would see the same segment colored in blue (for example) on the chromosomes of all 3 individuals, not just two of the three.
So, what’s wrong with matching? Not a thing. Not one single thing. But a match to someone is not proof of this specific common ancestor. You could match these two people on different common ancestors, because none of the DNA is shared by all 3 individuals and matches on the same segment.
The bottom line to this is that if Ancestry were showing you only matches that were triangulated to this ancestor, you might not have any at all. Using Circles rather than triangulation gives you many more matches and many more individuals in your Circle – and Circles where you wouldn’t have Circles otherwise – and it’s much less accurate. It’s a group of cumulative hints that together are really compelling – but they are not proof. Proof does not exist in the Ancestry system with the current set of tools.
Proof is triangulation of at least three people who share the same ancestor that also share exactly the same DNA on the same segment.
This however, does not mean Circles are bad – it means they need to be used with full understanding and care.
Now, this does not HAVE to be the case, because every Circle could have an option to “see only triangulated matches.” That would be the best of both worlds.
Henry Bolton is one of my most robust Circles, so let’s look at that. You can see that I match the DNA of 5 individuals in the Circle who also show Henry Bolton in their tree. These are not triangulated matches, because we don’t know if any of us match on the same segments (at least not through Ancestry’s system) – but we do know THAT we match and THAT we share common paper genealogy with the same ancestor.
You can see the greyed out individuals and the greyed out connection network. These greyed out people don’t match my DNA, but they do match the DNA of at least one person that I match, and also share Henry Bolton as a common ancestor, so they too are in the Henry Bolton DNA Circle.
By clicking on one of the other Circles, you can see who they match. So, for example, by clicking on my first match clockwise, below, you can see that they match 2 of the people that I do, plus four people that I don’t. These matching networks, while not proof that this is how you are genetically connected to these people, do serve as great research connections. Hopefully, at least some of them will be willing to download to GedMatch or transfer to Family Tree DNA, or both, where you can utilize a chromosome browser.
Regarding reliability, I’m much more likely to place confidence in a robust Circle like the Henry Bolton Circle with multiple DNA matches between me and other members than the Diedamia Lyon Circle shown below. Diedamia is one of the “ancestors” identified and assigned to me as my “new ancestor” that isn’t.
Keep in mind that the Diedemia new ancestor match is WITHOUT the common tree match required to create a Circle – so just utilizing the DNA. So what this “new ancestor” Circle is telling you is that I match both of these individuals and THEY have a common ancestor between them, so I have been assigned to that same common ancestor – which is incorrect. Diedamia Lyon is not my ancestor. We have no idea if I share any common segments between these two matches, which would at least increase the chances that I share a common ancestor with both of these people, because Ancestry does not do segment matching, and they don’t give us any tools to do it either.
Circles are not predictable and often come like the tooth fairy, at night, so every morning I check to see if I have a new DNA Circle, a new “New Ancestor” or any new shakey leaf DNA matches.
How Did I Get My DNA Circles and Matches?
I decided I wanted to see if I can make sense of how Ancestry’s Circles and New Ancestors are actually assigned or created, based on my matches.
As you know, I did the little experiment where I recreated myself as a newbie, so I can compare my results with my regular robust tree and a mini-tree with only me and my parents. Through that experiment, we discovered that of my 16 DNA Circles, 2 got “assigned” as new ancestors when using a bare bones tree. In addition, the son of that ancestor was also assigned, correctly as an ancestor, even though he didn’t have a prior Circle. Of course, there are still the two incorrect ancestors assigned in both circumstances, the robust and the mini tree scenarios.
Let’s see if we can figure out some logic behind how this actually works.
I have been keeping a spreadsheet of my shakey leaf AncestryDNA matches with whom I share an identifiable ancestor on paper. Why? Because both matches and Circles tend to come and go.
I’ve assembled the chart below based on my DNA matches with shakey leaves, meaning we have both a DNA match and a tree match. The column titled “Number of DNA+Tree Matches” is the number of DNA matches to someone who also shares that ancestor in a tree. Of course, the DNA match could be from another line entirely, but this is based on what Ancestry has provided.
“Gen from Me” means the number of generations removed from me, according to Ancestry’s calculations shown with the match.
“Circle Member Robust Tree” means that I either do or do not have a Circle for that ancestor using the robust tree.
“Assigned as New Ancestor” indicates whether this ancestor was assigned as a “New Ancestor” using only the mini-tree.
Please note that some shakey leaf matches were lost when phasing was introduced.
Let’s see how this works.
For example, in the first row, Henry Bolton and Nancy Mann had 8 DNA matches total, but lost 3 with phasing. Currently, only 5 are shown as matches. Henry is also shown twice as a match to just him and not Nancy Mann. This makes sense, as he was married twice and I can clearly match Henry through a child of his first marriage. Henry and Nancy are 5 generations removed from me and both Henry and Nancy are shown as Circles with my robust tree. Both Henry and Nancy are also shown as “New Ancestors” in the mini-tree “recreate myself as a newbie” version. My only other accurate “New Ancestor” is Henry’s son, Joseph Preston Bolton, who is not a Circle.
I sorted this first table by “Number of DNA+Tree Matches.” Let’s see how that looks. To begin with, my second highest match is George Dodson and Margaret Dagord with 7 matches, but they don’t form either a Circle or get assigned as a new ancestor. However, I have lots of Circles with fewer matches. Go figure.
| Ancestor(s) | Number of DNA+Tree Matches | Gen from Me | Circle Member (Robust Tree) | Assigned as New Ancestor (Mini Tree) | Comment |
| Henry Bolton, Nancy Mann | 8 | 5 | Yes, both | Yes, both | 3 matches gone with phasing, also shown by himself, 2 marriages |
| George Dodson, Margaret Dagord | 7 | 8 | No | No | Margaret also listed separately with 1 match |
| Johann Michael Miller, Suzanne Berchtol | 7 | 8 | No | No | |
| Jotham Brown | 7 | 7 | No | No | 3 matches gone with phasing |
| Joel Vannoy, Phoebe Crumley | 6 | 4 | Yes, both | No | |
| Abraham Estes and Barbara | 5 | 9 | No | No | |
| George McNiel, Sarah Coates | 5 | 7 | No | No | |
| John R. Estes, Ann Moore | 4 | 6 | Yes, both | No | 1 match gone with phasing |
| Elizabeth Shepherd | 4 | 6 | No | No | Wife of William McNiel, not shown |
| John Francis Vannoy, Susannah Anderson | 4 | 7 | No | No | 2 matches gone with phasing |
| Philip Jacob Miller, Magdalena | 4 | 7 | No | No | |
| Robert Shepherd, Sarah Rash | 4 | 7 | No | No | |
| John Campbell, Jenny Dobkins | 3 | 6 | Yes, both | No | |
| Joseph Preston Bolton | 3 | 4 | No, but his parents have Circle, Henry Bolton, Nancy Mann | Yes | Two wives, my wife Margaret Herrell has 1 match, but she is not listed |
| Daniel Miller, Elizabeth Ulrich | 3 | 6 | No | No | |
| Stephen Ulrich | 3 | 7 | No | No | Married to Elizabeth Greib, Cripe, shown separately |
| Thomas Dodson, Dorothy Durham | 3 | 8 | No | No | |
| Andrew McKee | 3 | 7 | Circle disappeared | No | Had Circle, then gone |
| Fairwick Claxton, Agnes Muncy | 2 | 5 | Yes, both | No | |
| Jacob Lentz, Fredericka Moselman | 2 | 5 | Yes, both | No | |
| Nicholas Speak, Sarah Faires | 2 | 6 | Yes, both | No | 1 match gone with phasing |
| Henry Bolton | 2 | 5 | Yes | Yes | He was twice married |
| Charles Dugas, Francoise Bourgeois | 2 | 9 | No | No | |
| Cornelius Anderson, Annetje Opdyke | 2 | 8 | No | No | |
| Francois Broussard, Catherine Richard | 2 | 9 | No | No | |
| Gershom Hall | 2 | 7 | No | No | Son of below |
| James Lee Claxton, Sarah Cook | 2 | 6 | No | No | Gone with phasing |
| Joseph Rash, Mary Warren | 2 | 9 | No | No | |
| Joseph Workman, Phoebe McMahon | 2 | 7 | No | No | |
| Thomas Dodson | 2 | 9 | No | No | |
| Francois Lafaille | 2 | 2 | Matches both gone with phasing | ||
| John Hill, Catherine Mitchell | 1 | 6 | Yes, John Hill only | No | |
| Charles Speak, Ann McKee | 1 | 5 | No | No | |
| Edward Mercer | 1 | 7 | No | No | |
| Elisha Eldredge, Dorcas Mulford | 1 | 8 | No | No | |
| Elizabeth Greib (also Cripe) | 1 | 7 | No | No | Wife of Stephen Ulrich, shown separately |
| Elizabeth Mary Angelica Daye | 1 | 8 | No | No | |
| Francois Dugas | 1 | 8 | No | No | |
| George Shepherd, Elizabeth Mary Angelique Daye | 1 | 8 | No | No | |
| Gershom Hall, Dorcas Richardson | 1 | 8 | No | No | Father of above |
| Gideon Faires, Sarah McSpadden | 1 | 6 | No | No | |
| Honore Lore, Marie Lafaille | 1 | 5 | No | No | |
| Jacob Dobkins | 1 | 7 | No | No | |
| Jacque Bonnevie, Francoise Mius | 1 | 8 | No | No | |
| James Hall, Mehitable | 1 | 7 | No | No | |
| Jan Derik Woertman, Anna Marie Andries | 1 | 9 | No | No | |
| Johann Nicholas Schaeffer, Mary Catherine Suder | 1 | 8 | No | No | |
| Lazarus Estes, Elizabeth Vannoy | 1 | 3 | No | No | |
| Margaret Dagord | 1 | 8 | No | No | Wife of George Dodson, also listed with him |
| Michael de Foret, Marie Hebert | 1 | 9 | No | No | |
| Moses Estes Sr. | 1 | 8 | No | No | Wife Elizabeth, LNU |
| Pierre Doucet, Henriette Pelletret | 1 | 9 | No | No | |
| Rachel Levina Hill | 1 | 4 | No | No | Wife of Antoine Hill |
| Raleigh Dodson, Elizabeth | 1 | 7 | No | No | |
| Suzanna Berchtol | 1 | 8 | No | No | |
| William Herrell, Mary McDowell | 1 | 5 | No | No | |
| Charles Hickerson, Mary Lytle | 1 | 7 | Circle disappeared for both | No | Had Circle, then gone |
| Francis Vannoy, Catherine Anderson | 1 | 8 | Match gone with phasing | ||
| John Vannoy | 1 | Match gone with phasing | |||
| Lois McNiel | 1 | 6 | Match gone with phasing |
This comparison of matches to the Circles created is actually very surprising, because Circle creation seems to have very little correlation to number of DNA matches. Circles require 2 people to match each other’s DNA, plus a third person that matches at least one of the other two – minimally. There are very obviously behind the scenes criteria too, or I would have at least 36 Circles based on my matches to 3 or more people and 26 additional Circles if you could matches to only 2 people. That’s a total of 62 Circles, not 16.
Of the 7 ancestral couples with 5 or more matches, which give us the potential for 14 individual Circles, only two couples, or 4 Circles exist.
The chart below is sorted by “Circle Member (Robust Tree),” so only the ancestors who are in Circles are shown. The number of DNA matches range from 1 to 8.
| Ancestor | Number of DNA+Tree Matches | Gen from Me | Circle Member (Robust Tree) | Assigned as New Ancestor (Mini Tree) | Comment |
| Henry Bolton, Nancy Mann | 8 | 5 | Yes, both | Yes, both | 3 matches gone with phasing, also shown by himself, 2 marriages |
| Joel Vannoy, Phoebe Crumley | 6 | 4 | Yes, both | No | |
| John R. Estes, Ann Moore | 4 | 6 | Yes, both | No | 1 match gone with phasing |
| John Campbell, Jenny Dobkins | 3 | 6 | Yes, both | No | |
| Joseph Preston Bolton | 3 | 4 | No, his parents have Circle, Henry Bolton, Nancy Mann | Yes | Two wives, my wife Margaret Herrell has 1 match, but she is not listed |
| Andrew McKee | 3 | 7 | Circle disappeared | No | Had Circle, then gone |
| Fairwick Claxton, Agnes Muncy | 2 | 5 | Yes, both | No | |
| Jacob Lentz, Fredericka Moselman | 2 | 5 | Yes, both | No | |
| Nicholas Speak, Sarah Faires | 2 | 6 | Yes, both | No | 1 match gone with phasing |
| Henry Bolton | 2 | 5 | Yes | Yes | He was twice married |
| John Hill, Catherine Mitchell | 1 | 6 | Yes, John Hill only | No | |
| Charles Hickerson, Mary Lytle | 1 | 7 | Circle disappeared for both | No | Had Circles, then gone |
Next, let’s take a look at the ancestors who have Circles created for them.
Surprisingly, I do have a DNA Circle based on only 1 DNA match. The entire Circle is made up of three people.
I’m confused as to why this ancestor, John Hill, with one DNA match and one additional person in the Circle would have a Circle, but an ancestor like George Dodson and Margaret Dagord with 7 and 8 matches, respectively, wouldn’t. Both of the two Circle matches also show the same wife as I do for John Hill, Catherine Mitchell, but there is no Circle for her. Why not?
There is a line in the sand at which point Ancestry no longer creates Circles because they feel they are too far back in time to be reliable, but George Dodson/Margaret Dagord are at 8 generations, as are Henry Bolton/Nancy Mann who both have Circles, so that can’t be the problem.
Now, let’s look at who, from the matches and Circles was assigned as a “New Ancestor” when switching out my robust tree for the mini-tree.
Other than the two incorrect ancestors assigned, there were only three ancestors assigned from the more than 60 possibilities from ancestors who have 2 or more matches. Henry Bolton and Nancy Mann both have a large number of matches, so I can clearly see why they were assigned as an ancestor.
What I’m unclearly about is how their son, Joseph Preston Bolton, was assigned as an ancestor. He is not assigned as a Circle, or maybe that’s intentional because he’s downstream of Henry Bolton. In any case, this a correct ancestor assignment. I do have to wonder how Joseph Preston Bolton was assigned as a “New Ancestor” with only 3 matches and other ancestors with far more weren’t.
I thought perhaps it was because Joseph isn’t really that far upstream from me, at 4 generations, but then Joel Vannoy and Phoebe Crumley are also at 4 generations and have 6 matches and a Circle, but weren’t assigned as a “New Ancestor” using the mini-tree. I can find no consistent theme here.
The following chart is sorted by “Assigned as New Ancestor (Mini Tree).”
| Ancestor | Number of DNA+Tree Matches | Gen from Me | Circle Member (Robust Tree) | Assigned as New Ancestor (Mini Tree) | Comment |
| Henry Bolton, Nancy Mann | 8 | 5 | Yes, both | Yes, both | 3 matches gone with phasing, also shown by himself, 2 marriages |
| Joseph Preston Bolton | 3 | 4 | No, his parents have Circle, Henry Bolton, Nancy Mann | Yes | Two wives, my wife Margaret Herrell has 1 match, but she is not listed |
| Henry Bolton | 2 | 5 | Yes | Yes | He was twice married |
| George Dodson, Margaret Dagord | 7 | 8 | No | No | Margaret also listed separately |
| Johann Michael Miller, Suzanne Berchtol | 7 | 8 | No | No | |
| Jotham Brown | 7 | 7 | No | No | 3 matches gone with phasing |
| Joel Vannoy, Phoebe Crumley | 6 | 4 | Yes, both | No | |
| Abraham Estes and Barbara | 5 | 9 | No | No | |
| George McNiel, Sarah Coates | 5 | 7 | No | No | |
| John R. Estes, Ann Moore | 4 | 6 | Yes, both | No | 1 match gone with phasing |
| Elizabeth Shepherd | 4 | 6 | No | No | Wife of William McNiel, not shown |
| John Francis Vannoy, Susannah Anderson | 4 | 7 | No | No | 2 matches gone with phasing |
| Philip Jacob Miller, Magdalena | 4 | 7 | No | No | |
| Robert Shepherd, Sarah Rash | 4 | 7 | No | No | |
| John Campbell, Jenny Dobkins | 3 | 6 | Yes, both | No | |
| Daniel Miller, Elizabeth Ulrich | 3 | 6 | No | No | |
| Stephen Ulrich | 3 | 7 | No | No | Married to Elizabeth Greib, Cripe, shown separately |
| Thomas Dodson, Dorothy Durham | 3 | 8 | No | No | |
| Andrew McKee | 3 | 7 | Circle disappeared | No | Had Circle, then gone |
| Fairwick Claxton, Agnes Muncy | 2 | 5 | Yes, both | No | |
| Jacob Lentz, Fredericka Moselman | 2 | 5 | Yes, both | No | |
| Nicholas Speak, Sarah Faires | 2 | 6 | Yes, both | No | 1 match gone with phasing |
| Charles Dugas, Francoise Bourgeois | 2 | 9 | No | No | |
| Cornelius Anderson, Annetje Opdyke | 2 | 8 | No | No | |
| Francois Broussard, Catherine Richard | 2 | 9 | No | No | |
| Gershom Hall | 2 | 7 | No | No | Son of below |
| James Lee Claxton, Sarah Cook | 2 | 6 | No | No | Gone with phasing |
| Joseph Rash, Mary Warren | 2 | 9 | No | No | |
| Joseph Workman, Phoebe McMahon | 2 | 7 | No | No | |
| Thomas Dodson | 2 | 9 | No | No | |
| John Hill, Catherine Mitchell | 1 | 6 | Yes, John Hill only | No | |
| Charles Speak, Ann McKee | 1 | 5 | No | No | |
| Edward Mercer | 1 | 7 | No | No | |
| Elisha Eldredge, Dorcas Mulford | 1 | 8 | No | No | |
| Elizabeth Greib (also Cripe) | 1 | 7 | No | No | Wife of Stephen Ulrich, shown separately |
| Elizabeth Mary Angelica Daye | 1 | 8 | No | No | |
| Francois Dugas | 1 | 8 | No | No | |
| George Shepherd, Elizabeth Mary Angelique Daye | 1 | 8 | No | No | |
| Gershom Hall, Dorcas Richardson | 1 | 8 | No | No | Father of above |
| Gideon Faires, Sarah McSpadden | 1 | 6 | No | No | |
| Honore Lore, Marie Lafaille | 1 | 5 | No | No | |
| Jacob Dobkins | 1 | 7 | No | No | |
| Jacque Bonnevie, Francoise Mius | 1 | 8 | No | No | |
| James Hall, Mehitable | 1 | 7 | No | No | |
| Jan Derik Woertman, Anna Marie Andries | 1 | 9 | No | No | |
| Johann Nicholas Schaeffer, Mary Catherine Suder | 1 | 8 | No | No | |
| Lazarus Estes, Elizabeth Vannoy | 1 | 3 | No | No | |
| Margaret Dagord | 1 | 8 | No | No | Wife of George Dodson, also listed with him |
| Michael de Foret, Marie Hebert | 1 | 9 | No | No | |
| Moses Estes Sr. | 1 | 8 | No | No | Wife Elizabeth, LNU |
| Pierre Doucet, Henriette Pelletret | 1 | 9 | No | No | |
| Rachel Levina Hill | 1 | 4 | No | No | Wife of Antoine Hill |
| Raleigh Dodson, Elizabeth | 1 | 7 | No | No | |
| Suzanna Berchtol | 1 | 8 | No | No | |
| William Herrell, Mary McDowell | 1 | 5 | No | No | |
| Charles Hickerson, Mary Lytle | 1 | 7 | Circle disappeared for both | No | Had Circle, then gone |
| Francois Lafaille | 2 | 2 | Matches both gone with phasing | ||
| Francis Vannoy, Catherine Anderson | 1 | 8 | Match gone with phasing | ||
| John Vannoy | 1 | Match gone with phasing | |||
| Lois McNiel | 1 | 6 | Match gone with phasing |
If you’re looking for answers to this mystery, you won’t find them here. I don’t know. All things considered, here is my collective wisdom on this subject.
- Enjoy your DNA Circles. Communicate with your matches. Ask them to download to Family Tree DNA and/or GedMatch where you have tools to work with. Watch for secondary lines through which you might match. I have found several where the DNA match is not to the ancestor in the Circle, but to a different, common line entirely. Of course, we still share the ancestor whose Circle we are in, assuming the trees are correct – it’s just that the DNA match is not from that ancestor.
- Understand that DNA Circles do not prove descent from that ancestor. The more people you match, the more strongly it implies a connection, but keep in mind that the DNA connection and the tree may not be connected either. Circles provide a “wider net” but also increases the potential for inaccuracy.
- Enjoy your “New Ancestors” but be extremely skeptical. Some of them will be ancestors. Some may be related but not ancestors. Some not only won’t be ancestors, you may not be able to figure out if or how they are related, no matter how large and robust your tree.
- Use all of this as a shakey leaf hint – which we all knows means that there’s something to check out – not gospel being dispensed.
- Make a spreadsheet to keep track of shakey leaf DNA matches, Circles and other people in the circle whose DNA you don’t match. Just because your Circle or match is present today doesn’t mean it will be tomorrow.
- New Ancestors and Circles are both beta software. There is a feedback button at the top of every DNA Circle, at the far right. Please submit courteous suggestions and comments. Oh yes, and don’t forget to mention that we need a chromosome browser:)
______________________________________________________________
Disclosure
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- AncestryDNA
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- LivingDNA
Genealogy Services
Genealogy Research
- Legacy Tree Genealogists for genealogy research