Strategy for finding common link

[cardinalcanary]

My AncestryDNA test shows that i share DNA matches with 3 people in the USA. I am in England and my family back for many generations are English or Welsh.

I have been in touch with 2 of the people in the US and they have absolutely no info on any English or Welsh ancestors. We are all keen to know how we are related.

WGolden is extremely likely to be my 4th to 6th cousin.
KB is a good match to be my 4th to 6th cousin.
CB is KB's paternal grandmother and is highly likely to be my 4th to 6th cousin.

WGolden is a good match to be KB's 4th to 6th cousin and highly likely to be CB's 4th to 6th cousin.

WGolden knows very little about his ancestry.

CB's grandmother was called Mary Anne Golden.

It seems to me that I must follow the Golden trail back to England if I can.

I do have an ancestor by the name of Ernest Golden Howlett in Utah but can't find where his Golden middle name originated.

Any suggestions what I should do next to pull everything together and join the dots?

Thanks

Stephen

[familydar]

Hi Stephen

The first rule of family history has to be "work backwards from what you know".  DNA testing doesn't change that.

So presumably you've worked your own tree back a few generations - needs to be all lines back to your 32 3xGGP to stand a chance of finding the common ancestor for 4th cousins, another 2 generations for 6th.

Your US cousins need to be encouraged to do the same.  They might not know a lot about their ancestry at the moment but by definition they must know more than you.

Upload your results and ancestor tree with dates and locations to GEDMatch.  Get your US cousins to do the same.  Also to FTDNA.  Both free, and may give you some more matches, perhaps closer than 4th cousin, which could help narrow down the field a bit.

The following two sentences from your post seem to be at odds with each other

I am in England and my family back for many generations are English or Welsh.

I do have an ancestor by the name of Ernest Golden Howlett in Utah but can't find where his Golden middle name originated.

Don't get fixated on GOLDEN.

best wishes
Jane :-)

[cardinalcanary]

Thanks Jane

I'm pretty confident that i have got my 32 great grandparents worked out but not so clear on 4xggp.

My 32ggp are all English or Welsh but some of their offspring ended up in various places around the World. One such person had the middle name Golden and ended up in the states. Golden is the name that seems to link all of us but it may be coincidental in my case.

I won't get fixated on Golden.

Stephen

 

[rsel]

Hi Stephen,
    The method I am using is a bit difficult to explain, so apologies if it becomes unclear, however I am slowly starting to get some success from it.
     The way I think about the problem is that you have two (or in this case three) trees, each of which somehow overlays, and each parent is a node that is a branch point, and somewhere those overlaying trees meet at one node. Now when you first look at the nodes each person has 510 possible relatives (if you go back to Great (x6) Grandparents), which could be the common one. So therefore you need to start trying to eliminate any branches you can. So for example if I can find that the person shares a match to another person who I do have a confirmed link to, the link must be above the node where that link occurs, which eliminates the other branch's.
    To give an example, I have had myself and both parents tested, so any match to me and either of them eliminates 50% of the possible connection nodes. I then have a know link to a 2nd Cousin of my father via our Great (x2) Grandparents, so any matches shared with her reduce the possible nodes from my side down to 62 (out of the 510), which is a big improvement. If you can then do similar on the other side it starts then t become a lot easier to find which areas of the tree you need to focus on as the possible matches. Now the more links you can do this for the better your triangulation can be.
     Now this bit is a lot of work, but I have been going through every single DNA match I have on ancestry, and looking at the 'shared' links and creating a spreadsheet of all the ones that share matches and creating what I have called pools, on the assumption that these pools of people must share a common ancestor at some point. So far I have identified approximately 80 different pools in my own matches and now its just a case of trying to link the people in them. This is where it gets real fun, as some people don't respond to messages or have tree's linked, so connecting them is not easy.

Richard

[familydar]

Richard, I follow your logic, it's a strategy I followed myself.  Until the following was pointed out to me.

Your father's second cousin has inherited from his GGGP a chunk of DNA that overlaps with a chunk that your father has also inherited, from the same GGGP.  But another second cousin may have inherited a completely different chunk, one that your father hasn't inherited.  Or your known cousin hasn't.  In simple terms, perhaps your father and your known second cousin got Granny's blue eyes, but the unknown second cousin got Granddad's brown eyes.  Yet all three of you are descended from the same couple.

It's a useful way of highlighting the more likely MRCAs but you can't use triangulation to definitely eliminate branches.

best wishes
Jane :-)

[rsel]

Richard, I follow your logic, it's a strategy I followed myself.  Until the following was pointed out to me.

Your father's second cousin has inherited from his GGGP a chunk of DNA that overlaps with a chunk that your father has also inherited, from the same GGGP.  But another second cousin may have inherited a completely different chunk, one that your father hasn't inherited.  Or your known cousin hasn't.  In simple terms, perhaps your father and your known second cousin got Granny's blue eyes, but the unknown second cousin got Granddad's brown eyes.  Yet all three of you are descended from the same couple.

It's a useful way of highlighting the more likely MRCAs but you can't use triangulation to definitely eliminate branches.

best wishes
Jane :-)

Hi Jane,   I don't think your logic is quite correct, if A & B are the cousins via my fathers fathers father, and both share a DNA match to an unknown C, then  I can eliminate any branches via my fathers mother....   I agree that you cant use triangulation if all 3 people don't share a match, but when all 3 do it does work.

Richard

[familydar]

Richard, agreed, I misunderstood what you'd written

Jane 

[cardinalcanary]

Thanks Richard and thanks Jane.

I am going to make an assumption and work through that to start with.

CB is a closer match than KB. CB is KBs paternal grandmother so that seems the most likely path to follow.

I am assuming that KB's father is my 5th cousin based on the known birth year of CB.  His mother (CB) is my then my 4th cousin once removed and KB is my 5th cousin once removed. I then know that KB's father shares a 4 x ggp with me. One of my 64 4 x ggp is the common link.

I then factor in that we are all related to Wesley Golden on another branch of the tree. Again, I am assuming Wesley is my 5th cousin once removed. And the Golden's in CB and KB tree are the same original Golden's. This reduces the number of people to look at significantly.

Mary Ann Golden is KB's 3xggm. I can jump to that node and work up that branch of the tree.

I have got the Golden line back to George C Golden who was born in England in 1770. He married someone called Mary. I am now a bit stuck.

However, George C Golden is in a tree called 1685 on the Francis. If George was born in 1770, he didn't come over on the Francis but presumably his relatives did. But who?

Now I am jumping to a name in my tree; Ernest Golden Howlett. He had a brother called Newel Howlett. Newel is a surname of passengers on the Francis.

There is also a surname Fendhall. This could easily have become Fendol as in Fendol Sylvester Golden as the family surnames got passed down to ancestors as first names.

I feel that I am getting tantalisingly close now but still lots of assumptions.