Can the Leeds Method Help Me Work with My DNA Matches?
The Leeds Method sorts your DNA matches into clusters. The members of each cluster are likely descended from a common ancestor or ancestral couple. Finding the common ancestor(s) for a cluster helps us to understand how these matches are related to each other and to us. And this method can help us to verify our ancestors or to break through brick walls.
While the Leeds Method is a powerful tool, unfortunately it cannot help everyone. Can the Leeds Method help you?
Works Well
The Leeds Method works well in the following situations:
If you have “quite a few” 2nd and 3rd cousins
The Leeds Method is based on, and needs, 2nd and 3rd cousins. As a generalization, those are matches between 400 and 90 cM. Although I cannot give an exact number, the Leeds Method often works well if you have at least 6 to 8 matches at this level. And some of them should be in the 200 to 400 cM range.
If you do not have endogamy or pedigree collapse
The Leeds Method works well if you do not have endogamy or pedigree collapse in your tree – or if it is on only one part of your tree. (Note: I will define both terms under the section for when the Leeds Method does not work well.)
If your grandparents are not related & came from different areas
If your grandparents are not related to each other, the Leeds Method works well. It also helps if they came from very different areas so there’s little chance that they might have a surprise connection or that a match might be related to two of your grandparents.
Doesn’t Work Well
The Leeds Method, unfortunately, does not work well in the following situations:
You have endogamy or pedigree collapse in your tree
Endogamy is the practice of a specific group of people marrying within their group for generations. Pedigree collapse is the occasional occurrence of cousins marrying cousins which used to be a fairly common practice. Both of these situations result in your matches being related to each other in multiple ways. And that causes the clusters to merge together instead of creating separate clusters.
However, if you have endogamy or pedigree collapse in just part of your family tree, you can still use the Leeds Method to separate out other parts of your tree.
You have few or no 2nd-3rd cousins
The Leeds Method is based on 2nd and 3rd cousins. If you don’t have any of these matches – or very few – the Leeds Method will not work as intended. However, you can still use it to create clusters with whatever matches you DO have!
Ms Leeds, I have several proven and unproven Revolutionary War Soldiers, meaning both Paternal and Maternal families have been in N America a very long time. It appears they started in PA and/or VA and migrated down the east coast to Georgia. Both sides appear to be in same economic strata, farmers and hard workers. Even though many do not like Ethnicity Estimates, I have found among my Matches that most of them fall within a few percentage points of each other: England, Wales & NW Europe; Ireland/Scotland; Germanic Europe; Norway and 1% Jewish. AND Yes, I am almost certain I have found the Jewish connection.
The more I sort my Shared Matches, the more Pedigree Collapse I get. I thought, because my parents lived in different counties, separated by a county that they would show little to no mixing. However, that is completely incorrect! That may have been true up until The Great Depression or even before…after The War Between the States. After Ancestry’s recent adjustment I am down to 107,282 Matches, 9,056 close. Tens of thousands appear to be kin to each other irregardless of Maternal or Paternal or where they lived.
I have been working on my Families since 1986. My most recent gold was to find more information about a husband and wife GGGrandparents. I am oldest living descendant (80 Yrs old) from my grandfather and maybe their oldest living descendant from all their children. I feel our DNA connection is aging out. I am getting nowhere with Bethena Bustre born about 1805 in Georgia. I can not even prove her sir name. Only clue we have is marriage registered in Henry co GA 1827.
I have looked at small amount of DNA matches, so far, finding no repeating names or patterns.
My question to you, with my DNA collapse, am I wasting my time? Most people lament about not having enough matches, I am overwhelmed with matches and do not know which ones to pursue.
I know you are busy; I have entered this question into several FaceBook groups with little response.
Sincerely, Judy McCleskey Webb Houston, Texas but born and raised, Atlanta, Georgia
Hi, Judy. (I’m in the Houston area, too!) So, first of all, if you still have over 100,000 matches it seems very likely you have some kind of pedigree collapse. Before Ancestry purged our matches under 8 cM, an “average” number of matches was about 60,000 on Ancestry. My dad was always right at average. He now has less than 30,000 matches. Do you have Colonial ancestors? We do not, but that could be your “problem.”
Have you found any matches that match with this particular line? If so, I would look at any of the shared matches of those people and see if you can make a connection.
You might also search your matches for the surname Bustre and work with those matches, if any. Although I have not worked with anyone from Georgia, the southern states had a lot of pedigree collapse – though it seems a bit odd since yours weren’t even in the same county. Maybe it was farther back in time.
Another possibility is to do targeted testing. This is where you research a specific couple and then trace the descendants down to the present day as best you can. Then, you ask if these descendants would be willing to test. You are hoping to find a match or matches of people you CAN prove are descended from a specific couple through their paper trail (or actual documents).
Those are a few thoughts! I hope this helps!
Dana
Thank you for devising the Leeds Method. I have a few over 80 matches in the 90 to 400 level, and they divide very well into 4 groups. My four grandparents were born in four different states with different backgrounds so it works so well. I also have lots of second cousins. I found 3 that were in two groups, but I knew why: one was a first cousin once removed and the other two were first cousins twice removed ( I am over 80).
Thanks again.
Joanne, You are so welcome! And, I’ve been meaning to make a post about first cousins once or twice removed and how that can affect your clusters. You did a great job working around those!
Thanks for this – what does it mean if you have 5 groups with no overlap? Does it just mean you need to include more cousins until you find that overlap? I’ve identified 2 of the lines as maternal, but am stuck with the other 3.
Hi, Des. Not everyone will get 4 clusters. The goal is to figure out what your clusters mean. You probably have 5 clusters because you don’t have any 2nd cousins on one of the family lines. So, I would suggest you now work with these 3 clusters and try to identify how they fit into your family. If this part of your family is unknown – or not making sense – try to figure out how the members of that cluster are related to each other. Hope this helps!
Most of my paternal ancestors, 100 years ago were born in the county of kent in England. They mainly had large families of about 10 children . I find the same surnames recurring – ie people seemed to marry second cousins or nieces married sons of aunts & uncles – is the Leeds Method unsuitable for this scenario?
Hi, April. It will really depend, but it is likely it will still work for you. I think you should give it a try and see what happens!
Dana
Hi,
Since you are using the 2-3 cousin matches, would you also use their 2nd and 3rd cousin matches to cluster?
Thanks,
Pat
Hi, Pat. Good question! When we do our initial chart, we are starting with 2nd grandparents since they share great grandparents with us. We want the DNA to sort our matches, but we are kind of hoping they sort into 4 great clusters. So, I would not add the 2nd & 3rd cousins of your matches to the initial chart. But, once it is created, you can definitely add them! Also, you can go ahead and try it and see what happens if you’d like. You will still get clusters, they just might form a little differently.
Hope this helps!
Dana
I am 100% Ashkenazi Jew, with endogamy and pedigree collapse. I ended up with 18 colors, and overlap all over the place. I know this method is not ideal for this situation, but any suggestions?
I started building trees for all sorts of people and no clue which side they related to me, or they are matching to both sides of my family.
Amy
Hi, Amy. Yes, unfortunately this method does not really work for endogamy and pedigree collapse. If you have a subscription to Legacy Family Tree Webinars, I highly suggest Paul Woodbury’s talk on DNA and endogamy called “Dealing with Endogamy.”
Dana
Would there be any way to use this to help identify the family of a 3Ggrandmother? This is only a record of her marriage to be found, and it’s ambiguous about her last name (long story).
I have identified a group of 5-6 matches that match the corresponding
2Ggrandfather that I can’t identify, and the entire family is fairly well documented, unless there is an NPE. Most of those match all the matches identified as descendants of the 2Ggrandfather. Unfortunately not a single one has a tree 🙄
Thanks in advance!
Hi, Dave. Clustering in general can help identify an unknown 3x great-grandparent. Since you’ve found a group of unknown matches that might be related to this unknown ancestor, I would concentrate on them. If they have shared matches who share less DNA, you can work with them, too. You also might try testing at additional companies to get (hopefully) more results. I suggest paying to be tested at both Ancestry and 23andMe and then uploading your results, for free, to MyHeritage and FTDNA.
Dana