100+ users under Ancestry's "Shared Matches" tab

[melba_schmelba]

Hi all,

I'm interested to hear how many results you find with matches under the Shared Matches tab of Ancestry DNA. [It's in the centre of a user results page with Tree, Ethnicity and then Shared Matches]

I find this a strong and accurate feature in showing results who share the same ancestors. When I click this tab on my user matches I get 0 to 12 user names appear under shared matches - normally there aren't that many.

However there is one area that brings 100+ shared user matches, which seems unusually large in comparison with everything else. Do you have any shares with such a number of results?

With such a big number of matches is it possible this is a strong hit of DNA that sits further back in the tree (6-8GG). ie: with it bring further back it is pulling in more shared matches, whereas something at 3GG is unlikely to have that amount of shared users.

I also can't identify where this share is coming from.

These shared match DNAs start at 36cM and downwards. (For what it's worth an identified match elsewhere on my tree with 35cM is a 3GG.)

I can account for all 5GG of my tree (and nearly all 6GG) as I'm Scotland and all my ancestors are born/died in Scotland, can find the paperwork trail and DNA matches. There's nothing odd about my tree - it's just this one area.

I've tried to build a tree based on these 100+ user trees but it's very patchy, no clear pattern, and that may indicate something further back. To draw all these trees into a single 3GG point isn't happening. Sure with a 7-8GG then creates many more intersection possibility.

Incidentally all these users trees are USA with no obvious Scottish connection.

I have similar to this, with over 100 shared matches for one match in the 40 cMs. I have identified which branch of my family several tie to, and most of the others are located around the American Mid West - I suspect an unknown branch of this family settled, perhaps several brothers or siblings and had large families generation after generation, possibly then intermarrying which would magnify the shared matches. Also I suspect that area is possibly one of those that is more highly represented in the Ancestry DNA database. The problem is linking American trees back to the UK accurately - many have been traced wrongly and you are led on a wild goose chase if you follow American family trees to the letter.

[Ayashi]

The highest shared match group that I've noticed I think is up to nearly 30. The vast majority of them descend from the same couple but I have no idea whatsoever, surname or geographically, how they are related to me. Maybe one day...

[ikas]

The vast majority of them descend from the same couple but I have no idea whatsoever, surname or geographically, how they are related to me. Maybe one day...

That's the limitation of the 20cM cut off. If you could look at shared matches between 10cM and 19cM there may be clues in that list of their relationship to you. I have two or three similar situations. I don't see a way round it.

[Ayashi]

Having someone I can identify on my side of the tree would be nice... The last time this situation happened, where all of our shared matches were on my side of the tree to my confusion, we eventually realised that there was a person in her tree descended from the common couple... married to her grandmother's sister 

In theory, if I match to them, someone related to me should eventually come up as a match... maybe? 

[melba_schmelba]

The vast majority of them descend from the same couple but I have no idea whatsoever, surname or geographically, how they are related to me. Maybe one day...

That's the limitation of the 20cM cut off. If you could look at shared matches between 10cM and 19cM there may be clues in that list of their relationship to you. I have two or three similar situations. I don't see a way round it.

There are tools you can use to analyse multiple accounts at once to get lower than 20cM matches such as Genetic Affairs. I also just found this blog where an American lady talks about the perils of clustering (and shared matches on a smaller scale) for southern families in the US where there was a lot of cousin marriage

https://www.theoccasionalgenealogist.com/2018/12/dna-auto-clustering-genealogy.html

[gizmo303]

Gizmo303 - When you say there are 30+ from 19cM to 6cM do you mean in your shared matches list you have some below 20cM? I understood there was a 20cM cut-off.

Yes only if I go looking at users below 20cM. As explained some of these came from matches above 20cM that control more than one set of DNA results. And I also look at new DNA matches in general and yes even some of the ones below 20cM do pull in these usual users under Shared Matches.

[DavidG02]

I clicked a random 26cm match  and found 2 x 3-4th cousins at 90cm+

There were also 120 20cm+ matches

I dont know who they are or where they fit as the info is private and profile is near dead

[Galium]

Not as many as a hundred, but I do have one bunch of, as yet unexplained, matches of around 20-25cM, who mostly link to each other.  The common ancestor that I can find for many of them is a man who emigrated to the USA in the early 19thC after being converted by the Mormons.  He had three wives on the go, and lots of children. Still no idea how he might be related to me, though.

[Petros]

My wife and I both have groups of shared matches similar in numbers to those described by Galium. Each group also appears to come from a common ancestor who emigrated, but with no discernible connection. Possibilities may be hinted at by geographic origin in the UK where their trees go back far enough