Ancestry added a greatly anticipated feature this week that promises to help genealogists – shared matches. This is similar to the “In Common With” feature at Family Tree DNA – at least in concept.
Previous to this announcement, when you match someone at Ancestry, the only way you can see who else they happen to match in common with you is if you are placed in an ancestor DNA Circle with them – and then you can only see the other people in that Circle.
For example, here is my Henry Bolton DNA Circle.
The people I match are shown with an orange line. Each of those people match me, and they may also match other people in the Circle that I don’t match.
Regardless of whether I match the individuals directly, or they match someone else that I match, the common factor is that we all share Henry Bolton identified as an ancestor in our tree.
What Ancestry introduced today is the ability to click on any of these people who match me, OR, the people in the circle who do NOT match me but who do share Henry Bolton in their tree and match others in the circle – and see who they match in common with me. This should allow people to group their matches, at least tentatively and is especially promising for those frustrating people with whom you match closely but have private trees and won’t reply to messages.
While this is interesting for circles, it’s not terribly useful in terms of breaking down walls, because I already know Henry Bolton is my ancestor. In other words, I wouldn’t be in the circle if I didn’t already know the identity of that ancestor.
What I’m particularly interested in, is applying this tool to my NADs, or New Ancestor Discoveries, because if I can figure out how these people truly are related to me, then I may be able to make a discovery of a new ancestor in my tree. Now THIS holds a lot of promise and intrigues me greatly. So, let’s take a look at my NADs and see how this new tool works and if it’s useful. I can hardly wait!!!
State of the NADs
If you’ve been following my blog, you’ll know that Ancestry and I have been having a bit of a friendly Bad NAD duel. Ancestry keeps giving me new ancestor discoveries (NADs) but in several cases, I have unquestionably proven that those NADs are not my ancestors – hence the term – Bad NADs. In one case, the new ancestor assigned to me is the husband of my ancestors sister. However, I currently have three NADS that are related to each other than may benefit greatly by this new shared matches tool.
Since my last NAD update, where Diedamia Lyon and John David Curnutte were given to me a second time, another NAD has been added – John David Curnutte’s mother, Deresa Chaffin.
Here’s the tree version of this relationship
NAD Circles and Matches
In the NAD Circle for Diedamia Lyon, John David Curnutte and Deresa Chaffin, we find both Don and Michael, whom I match.
First, keep in mind that I may match both Don and Michael on other lines – so the fact that I match both of them and they both descend from a common ancestor does NOT mean that is how I connect genetically to both of them. But for purposes of this discussion, let’s assume that it is and proceed.
The fact that we find these two individuals whose DNA I match in all three circles suggests that the relationship is through the Curnutte line, and not through Diedamia Lyon at all, except for the fact that these men also descend from her. Given that John David’s Curnutte’s mother is also a NAD suggests that the connection to Diedamia Lyon and John David Curnutte is through the Curnutte line. Although Deresa Chaffin’s husband is not listed, he is John Tolliver Curnutte and clearly, the connection might be through him as opposed to Deresa – just like the connection to the couple Diedamia Lyon and John David Curnutte was through the Curnutte husband.
The NAD Circle for Diedamia Lyon and John David Curnutte are identical, with two matches and 5 non-matching individuals.
For each one of these individuals in the Circle, if you click on their name on the right, you’ll be able to see a variety of information, including their pedigree and matching surnames, maps and locations, and the new shared matches tab.
The new shared matches tab is a great tool, and it’s particularly important, when unraveling NADs to use it in conjunction with the shared surnames, shown at left. These are the surnames found in both your tree and the person whose tree you’re comparing against.
Let’s take a look at one of these – Moore, as an example.
As you can see, these are either not the same line or at least can’t be identified as such. However, in some cases, you may recognize your matches’ end of line person as connecting with your tree further upstream. It’s times like this that having a robust tree where you’ve tracked downstream lineages of your ancestor’s siblings can be very beneficial.
By clicking on the shared matches option, you’ll see the following people who you match in common with the individual – in this case, Don, my DNA match. I could also compare to one of the people in the Circle whom I don’t DNA match.
What I’m particularly looking for are matches with that lovely shakey leaf by the View Match button on the far right. Ahem…there aren’t any, which means none of these matches match me with a known common ancestor. Rats!!!
While Diedamia Lyon and John David Curnutte have the same members as each other in their NAD circles, John’s mother, Deresa Chaffin, has more members in her NAD circle – which means more opportunities for me to find common line hints..
The DNA matches are to the same 2 people, but now there are additional people in the circle who also match Michael and Don.
The great news is that in addition to clicking on your matches to see who else they match, you can also click on any other circle member. I’m very, very hopeful that a distinct trend emerges so I can tell at least what line these NADs might be associated with.
I needed a mechanism to keep track of who all my matches match, that I match, and what lines they descend from – so I created a spreadsheet.
NAD Matches Spreadsheet
Column 1 – NAD – The ancestor’s name of the NAD Circle where these individuals are found as members.
Column 2 – Person in Circle – The “person in circle” is the individual whose name shows either as a DNA match or as a circle member who does not match my DNA, but does match the DNA of at least some of the other circle members.
Column 3 – DNA Match – Tells me if this person is a DNA match to me or not.
Column 4 – Common Family Line to Person in Circle – The common ancestral line (or lines) if I can determine whether or not we share a specific ancestral line. By the way, just because we share that line does NOT mean that is how we are DNA related – and no – there is no way to tell without a chromosome browser.
Column 5 – Common Surnames to Person in Circle – Common surnames between my tree and the person in the Circle, as identified by Ancestry.
Column 6 – Shared Matches with Person in Circle – Names of Shared Matches between me and the person in the Circle.
Column 7 – Common Line with Shared Match – Common ancestral lines with shared matches (column 6).
I combined the information from Diedamia Lyon, John David Curnutte and John’s mother, Deresa Chaffin. I sorted column 6, Shared Matches with Person in Circle, alphabetically, hoping that some of these matches would be the same, and they are, and would be identifiable to specific family lines.
So….Drum Roll….Who is the Common Ancestor???
I compared each person identified as a person in the NAD Circle (column 2), or any person that matches me and a person in the NAD Circle (column 6) with my other spreadsheet that I maintain listing all of my Ancestry matches and our common ancestors.
The group that includes the initials EVH are a family of siblings and their children, so they really only count once. The person by the name Mars has a private tree, but told me that our common ancestor was Joel Vannoy and Phebe Crumley, the same individuals as my cousin group through EVH.
It’s certainly possible that the common DNA that connects me with Michael and Don and possibly with John David Curnutte’s parents are through the Vannoy/Crumley line.
If indeed, our common ancestor is upstream of Joel Vannoy and Phebe Crumley, which is a VERY BIG if, but it’s the only lead I have – then they must fill a known pedigree void.
Deresa Chaffin, according to the Ancestry overview (which is all I have to go on at this moment and is compiled from 705 trees which makes me exceedingly nervous) was born in 1775 in Virginia to Simon Chaffin and Agatha Curnutte. She married John Tolliver Curnutte, so we have an intermarriage already (or incorrect surname information), which can mean a larger dose of the Curnutte DNA. Trying to follow these individuals up their trees at Ancestry was an exercise in frustration and futility with many of the wives surnames being the same as the husband and no sources or documentation of any kind. Suffice it to say, I can’t connect the dots through surnames or location, other than the state of Virginia.
However, looking at my tree, my vacancies for ancestors in that timeframe, in the Vannoy/Crumley branch of the tree are limited.
Phebe, Jotham Brown’s wife’s surname is unknown, but they were married about 1760.
William Crumley’s wife’s name is unknown, but they were married by about 1788. Clearly, Deresa being born in 1775 cannot be William Crumley’s wife (or Jotham Brown’s), and Deresa married a Curnutte, so she cannot be the ancestor in question for either vacancy.
John Tolliver Carunutte, Deresa’s husband was born about 1774, so clearly, he isn’t my ancestor either. One generation upstream, I have vacancies for six unknown parents, one of which would have been surnamed Brown. These people would have been born between 1720 and 1740, at the latest, and possibly earlier, so probably not John Tolliver or Deresa Chaffin’s parents either.
Unfortunately, we’re now back into the ether – and it’s very tenuous ether at that. Without a chromosome browser, I can’t confirm that the DNA of any of these matches triangulate with the Vannoy/Crumley DNA line – or any line for that matter.
However, in the spirit of running every lead down, right into the ground, and in this case, into the rathole – I view these new shared matches as my only hope of ever unraveling the mystery of the 3 related NADs. So far, I’ve proven they can’t be my ancestors, at least not in that line, but I still have absolutely no idea of how or if they are related to me – despite due diligence on my part- at least all the due diligence I can think of.
Suffice it to say I’m disappointed. It’s not my lucky day. No happy dance for me. I guess I probably don’t have to mention that if Ancestry provided a chromosome browser, I wouldn’t even have to be slogging around in the mud trying to piece these puzzle pieces together that might not even be from the same puzzle.
However, your mileage may vary and it may be your lucky day, so please give this new shared matches tool a try. If nothing else, it will help you group your matches by ancestral group and will give you clues as to the family groups of those people with private (or no) trees. And who knows, maybe you’ll unravel your NAD and actually discover a new ancestor!!! It could happen, especially if your matches are willing to download to GedMatch for verification!
Here’s Ancestry’s blog posting about the new shared match tool which includes a nice “how to” video.
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
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- Legacy Tree Genealogists for genealogy research
Always a joy to read you! I just wanted to note that you may have dismissed a marriage a bit too soon when you say Deresa born in 1775 cannot be married by 1788. Well, I have in my tree several of these “impossible marriages” when the girl was 12 or 13 and often the husband is 35 or more!
Hi Suzanne, You’re right about the age, but I dismissed her as a spouse to William Crumley because she could not be William Crumley’s unknown wife by 1788 because she married John David Curnutte.
You lost me!
I’m trying to figure out where, in my tree, that these three people could fit. There is only one place, IF, they do indeed connect in my line up the Vannoy/Crumley line. The only place Desara might fit is as the unknown wife of William Crumley – except she can’t be because she married a Curnutte male. That was my point. So I did not discount her because she was born in 1775 and could not have been married by 1788. I discounted her because we know she married someone else and she could not have been the wife of William Crumley who we know was married to someone by 1778. That’s what I was trying to say. The only other vacancy in that tree branch is Jotham Brown but his wife’s first name was Phebe and they were married by 1760, significantly before Desara was even born.
If you check shared DNA matches and get a list of six persons and then check the DNA matches for each of those six, you may find that four or five of them all match each other and maybe there are one or two who match only your initial match. Those who all match each other would be a triangulation group if what Ancestry is telling us is that we all share a DNA segment. We won’t know what segment, but if one or more of the matches has uploaded to Gedmatch, maybe we can find out.
I’ve played with this for one evening and solved a mystery of a third cousin match, whose only known genealogical connection was through a 9th gr-grandparent. Running his matches I found that he matched four of us who all matched each other through a 4th gr-grandparent. His match could be closer than that to me as he has several gaping holes in his tree. At least now I’ve identified his line and can put aside the green NPE spectre that I think we resort to more often than is justified.
I’m building a spreadsheet for these shared matches and will try to mate them with the TGs on my Gedmatch spreadsheet. Hopefully Ancestry will eventually allow us to specify at which level shared matches are reported. There is plenty of information to work with at the “very good” level in the meantime.
It means that the people match each other but it does NOT mean they match each other on the same segment, so they are not a triangulation group in the way we traditionally use that term – to mean three people that all match each other on the same segment and have a common ancestor.
But is we write to those 6 people and get them to upload to Gedmatch we may be able to get a bit further.
Don’t we get beyond the limits of probability when four people all match each other on some segment that the segment in question is the same segment? It would be nice if Ancestry at least gave us the chromosome we matched on. We might find that we are dealing with two or more groups of matches instead of one, however the four who inter-match would be in the same group.
I’m not in any circles and don’t have a lot of ancestors in my tree yet but I did find it interesting to play around with the “shared matches” feature that appears when you click on someone in your match list. I have found that a number of people that I match with also share matches with other people I match with. This is the sort of in-common comparison I frequently use in FTDNA to decide who do run a browser match on. Now we just need a Ancestry browser. Otherwise I’m not quite sure yet how to make further use of what I’ve found. Suggestions?
You’ve said it all. We need the browser.
Unfortunately AncestryDNA’s “Shared Matches” feature only works for “high confidence” 4th cousin and higher matches in the list – so this is a huge deficiency and missed opportunity. For the tests I manage, <2% of our matches are predicted 4th cousins or closer, so this feature is non-existent for 98% of our matches. Yes, AncestryDNA has made a tiny baby-step in the right direction, but it is still very disappointing and incredibly frustrating that they are so reluctant to provide more.
I couldn’t agree more.
That’s interesting; I didn’t realize there was a “4th cousin and higher” limit… that might explain why I clicked on one of my 5th-8th cousins (confidence = good) and found that her shared match w/ me is my dad. But when I click on my dad’s list of shared matches, she doesn’t show up on his list…
My comment wasn’t worded well, but CathyD’s comment is still correct. I should have said that the Shared Matches feature can be checked on everyone in your match list, irrespective of how high up (or low down) the match appears on your match list. However, the only individuals that are shown by the Shared Matches feature are those that are at least high confidence, at least predicted 4th cousins or higher. For the tests I manage on AncestryDNA, this means that 98% are filtered out by AncestryDNA), so we are only being shown a pitiful fraction of what could be useful.
Roberta, Thanks for all your postings, especially the recent ones about AncestryDNA.
After matching up the people linked to me on AncestryDNA, I made a list of the top 70 “A” people on my GEDMatch account. . Biggest problem is that the name given on Ancestry is usually different from their name on GEDmatch. I’ve only confirmed two people so far. Many of the matches are above 20 cM, some as much as 48 cM. I sure wish I could confirm the rest. Of course, I COULD email them. Fortunately Some of them did more than one DNA test. Can’t figure out how one of my proven “shaky leaf” cousins is listed as having no matches. And, how do I get rid of that really bad NAD, Milo Andrus who, they say, MIGHT be a relative (NOT!)
My 1st cousin ‘s son had his DNA done. He showed up on my matches. My Dad & his mother’s dad are brothers. Thing is there is No circle for us. Likewise, one of my 2nd cousins once removed (living) from my Dads line has no circle. No circles for my Danish line at all. My maternal line has 8 or so circles.
There are a lot of instances where it seems circles should be created and aren’t. Ancestry has never shared their exact methodology, but I know it has to do with the confidence score of the matches as well as the number of them. I also know that close relatives such as siblings and children are counted as 1 and displayed as a family group. So a direct family group will not create a circle even though logically it seems like they should.
Roberta, thank you for the details about the process you’ve described in this article. I very much appreciate all of your articles and was hoping that you would cover this new feature at Ancestry. At first glance, after some initial eagerness, it appears that this new feature might not be very helpful. A large percentage of my matches have no shared matches, particularly in more distant generations. Otherwise, the shared matches are very often people without trees or they have private trees. I have already e-mailed everyone in my circles, asking them to move their tests to Gedmatch, which seems to be effective for those matches who already had a “green leaf,” but I was hoping to get some clues to my “brick walls,” mostly between 1750-1820, with this new feature.
As June commented above, some of her matches match with other matches. Using Gedmatch and Genome Mate, I’ve noticed that there appear to be several “circles” of 5-6 matches who mostly all match each other and me through triangulation, but finding the common ancestor is elusive.
Right now they are only showing 4th cousins and high confidence scores, so it is limited. It didn’t help me, but some are reporting some success.
If would be very helpful if Ancestry would provide the chromosome and segment where we match other people. Then we could use triangulation to see if the match is paternal or maternal. Triangulation is necessary since each segment has two halves, and so we might match one person on one half and another on the other half. In that case those two individuals would not show a match. I like 23andMe for the very reason that they do provide that information.
I was in the middle of a major Ancestry project when this Shared Matches feature popped up. I decided to invite my top 200 Ancestry matches and all 68 shared ancestor (leaf hint) matches to GEDmatch. We all agree nobody should have to do this; Ancestry should listen to their customers and provide a chromosome browser already. But I’d be shocked if they did it retroactively.
In any case I now have a spreadsheet where I can compare everyone. It’s true that only 4th cousin+ matches are shown, but you can also click on a 5th-8th cousin and see these matches. I had the Snavely tool so some were known; many are new.
I’m asking my closer matches if they know how they match Shared Match XYZ. That may be the main value in the early going.
Scorecard after about 5 days:
Invitations – 150
People who joined GEDmatch as a result – 2
People who were on there already and I didn’t know – 4
People who gave me their password and said “you figure it out” – 1
People who said they knew how they match XYZ – 0
Oh Rich, those numbers are really depressing – about 3% total and a heck of a lot of work on your part. I have had very limited success with people at Ancestry joining GedMatch, if there weren’t already familiar with it. I had hoped that others might fare better. Thank you for sharing.
It’s early. But yes, the numbers are lower than I had hoped. I’m finding the leaf-hint matches are more likely to accept, because they’re the ones who are really into this and have built their trees out 8 generations. The closer cousins are less likely to know what GEDmatch is.
That makes sense.
” I had the Snavely tool so some were known; many are new.”
Rich, what is the Snavely tool?
June – The Snavely tool is a free app available at the Chrome store. If you google “AncestryDNA Helper” and Snavely you’ll see it. i can see the DNA results of about 10 of my cousins, and i use this tool to see our common matches. You can also use it to get a complete list of your match’s surnames in an excel spreadsheet.
Rich, Please contact me at firstname.lastname@example.org. I started seeing strange things when I tried playing with it. I’m hoping you know enough about it to explain what I’m seeing. Thanks,
Rich Capen, I got flagged as a spammer when I sent out that many invitations to my matches. As a result, people didn’t get my messages and I can’t use message center. I’ve been working on getting my account restored for 3 weeks now. So far, no luck.
One person said that happened to them when they sent more than 50 messages in one day. Another said 30.
Thank Jason. That issue had not occurred to me. The cutoff must be higher than 30; I sent maybe 40 on one day. I looked on the site for guidance and didn’t see anything. BTW I’m now up to 4 new GEDmatches and 5 that were already on. It was worth it : )
This NAD circle of mine, might have a connection with my mother’s Adam’s family. I am still confused how I am related to the Caudill Family. Could the connection might be with a Adams family member rather then with a Caudill? Here is what it keeps saying. “Your DNA matches the DNA of members of the John Adams Caudill DNA Circle. Because of this, there is a good chance (up to 70%) you could be either a descendant or relative of John Adams Caudill. John Adams Caudill was born on January 1, 1798 and died 1873”. My mother’s tree is pretty well researched, so I would think it be a slim chance we missed a Caudill relative. We have no Caudill’s on our tree. Could this be one of those cases where I just inherited dna from a possible relative way back in the day? Any thoughts anyone?
You may not be related to the Caudills. In one case, the connection was through the matches wife’s sister. In another, the spouses line, not that person’s line. And in one case, I match those same people in the Circle to a different known ancestor – so that NAD was only circumstantial. Take NADs with a very very large dose of salt.
Thank you for your NAD report. The Adams middle name part had caught my eye. So it made me wonder if it belong to my mother’s Adams family. None of the Adams names on our shared tree seemed to match my mother’s. I liked how Ancestry added the shared matches tab. Maybe this a step in the right direction for Ancestry’s updates.
Thanks Roberta for another excellent exposé.
I spent a couple of hours last night playing with Ancestry’s results and comparing known strings I share on GedMatch.with Ancestry’s results.
My results were mixed, to say the least. For example, I share a fairly long string (40+ cMs) with with 3 people also on GedMatch and a shorter string (17 cMs) , triangulated with the first three. Ancestry only picked up the first three in commons. At least one of their other in commons and I share a string but on a totally different chromosome.
I have another string with 4 triangulated Ancestry matches on GedMatch, but when I check Ancestry’s ‘in common’ list, it shows no matches. The matching strings range from 23 to 10 cMs but I would expect at least the longer strings (over 15 cMs) to display, but instead nothing. The high confidence is probably the reason. I have very few of those are most are locked or have no tree.
Then I tried my NADs. Fortunately a couple of of the Ancestry matches had also uploaded to GedMatch so I could evaluate the results. And, drum roll please. Three of the four all matched on the same chromosome sting on their maternal side to a common ancestor. That’s the good news. The bad news is that the NAD is on their paternal side. So almost certainly has nothing to do with the matching chromosome string.
This IS NOT the same as a chromosome browser – not even close. This new tool could lead you to jump to some very wrong conclusions. We still need a chromosome browser, NOW
I had previously done some work on one of my NADs. Using the Snavely tool to identify ICW matches for cousins I’m sharing DNA with and the GedMatch info for some known cousins, I was able to develop a hypothesis as to which line the NAD is probably connected to. Results using Ancestry’s Shared Matches tool seems to support my hypothesis, but as of yet no idea how the NAD is connected to my family line and certainly no MRCA identified. I can’t help but think how much easier it would be with a chrome browser.
Pingback: Autosomal DNA Matching Confidence Spectrum | DNAeXplained – Genetic Genealogy
Pingback: DNAeXplain Archives – Intermediate DNA Articles | DNAeXplained – Genetic Genealogy
Pingback: The Best and Worst of 2015 – Genetic Genealogy Year in Review | DNAeXplained – Genetic Genealogy
Roberta, would you compare/contrast AncestryDNA’s Shared Matches lists with the lists generated by FTDNA’s In Common With tab? Both seem to list people who match, at some level, me and another specified person that I also match. Except for FTDNA’s inclusion of ICWs with smaller amounts of matching DNA, do they mean essentially the same thing? (And thus can either tool be used on the pathway to triangulation?) Also, (I am about a year late to this party) is it your understanding that Ancestry still includes in its Shared Matches only 4th cousins or closer? My Shared Matches lists currently include people who Ancestry says are in a “possible range: 4th-6th cousins,” all of whom appear to share with me more than 20 cMs.
FTDNA only includes people who match over their threshold and it’s larger than the 5cM at Ancestry. You can select your own people to check against at FTDNA. Ancestry’s tool can’t be used on the pathway to triangulation because there is no way to see what segments anyone matches anyone on, including you to anyone. There is no way around that at Ancestry.