Author Topic: How far back can AncestryDNA's "Common Ancestors" feature work?  (Read 3054 times)

Offline Albufera32

  • RootsChat Member
  • ***
  • Posts: 204
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #9 on: Monday 07 June 21 10:16 BST (UK) »
Thru Lines are a form of hint - and are as reliable (or not) as every other hint.

Like every other hint, they can be both enormously helpful or frustratingly misleading.

Howie (Riccarton Ayrshire)
McNeil/ McNeill (Argyll)
Main (Airdrie Lanarkshire)
Grant (Lanarkshire and Bo'ness)
More (Lanarkshire)
Ure (Polmont)
Colligan (Lanarkshire)
Drinnan (New Zealand)

Offline brigidmac

  • RootsChat Marquessate
  • *******
  • Posts: 5,986
  • Computer incompetent but stiil trying
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #10 on: Monday 07 June 21 10:39 BST (UK) »
I agree you always have to check thru lines with documents

I have some relatives where different thru lines are possible because 2nd cousins in small villages married so we are related to both sides

I think ancestry prioritises thru lines from the ones that have the most documents and numbers of connections ..whereas I'd like to see the more obscure links

In the same way looking for a person on family trees the trees with most records are computer generated first
So if the information is wrong it's perpetuated by being copied by more people

Having said that thru lines have been able to confirm matches for distant relatives . One lady named her son after one man but named another as father on birth record ...DNA showed matches with thru line to the family whose surname he had as a middle name !
Roberts,Fellman.Macdermid smith jones,Bloch,Irvine,Hallis Stevenson

Offline Albufera32

  • RootsChat Member
  • ***
  • Posts: 204
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #11 on: Monday 07 June 21 12:26 BST (UK) »
I have three matches who appear (all three of them) on not just two but three thru lines through three different common ancestors.

The slightly horrifying thing is I rather suspect all three thru lines are correct for all three matches.
Howie (Riccarton Ayrshire)
McNeil/ McNeill (Argyll)
Main (Airdrie Lanarkshire)
Grant (Lanarkshire and Bo'ness)
More (Lanarkshire)
Ure (Polmont)
Colligan (Lanarkshire)
Drinnan (New Zealand)

Offline History Lives

  • RootsChat Senior
  • ****
  • Posts: 305
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #12 on: Monday 07 June 21 12:57 BST (UK) »

I've been in a similar situation. It essentially makes grouping matches via colour coding impossible for those lines.
North West, North Wales and Isle of Man:

Ward, Campbell, Cowin, Cowell, Parry, Fryer, Davies, Hughes, Briscoe, Jones, Spencer, Brownbill, Crowfoot, Nield, Randles, Youde

Gozo, Malta:

Apap / De Apapis, Camilleri, Grech, Muscat, Micallef, Custo, De Nasi

Angus, Scotland:

Ormond, Salmond, Cook, MacDonald


Offline LizzieL

  • RootsChat Marquessate
  • *******
  • Posts: 7,946
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #13 on: Monday 07 June 21 14:07 BST (UK) »
I have a 10cM match with a person who Ancestry flags up as us having a common Ancestor and correctly predicts he is my 5th cousin once removed. Shared ancestors are my 4 x great grandparents and his 5 x ggp.
I also have a 10 cM match with his son, but Ancestry is not flagging up that we have common ancestors, because the shared Ancestors are his 6 x great grandparents - too many generations for Ancestry's algorithm.
In both cases the individuals are correctly linked to an extensive tree
Berks / Oxon: Eltham, Annetts, Wiltshire (surname not county), Hawkins, Pembroke, Partridge
Dorset / Hants: Derham, Stride, Purkiss, Sibley
Yorkshire: Pottage, Carr, Blackburn, Depledge
Sussex: Goodyer, Christopher, Trevatt
Lanark: Scott (soldier went to Jersey CI)
Jersey: Fowler, Huelin, Scott

Offline History Lives

  • RootsChat Senior
  • ****
  • Posts: 305
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #14 on: Monday 07 June 21 15:06 BST (UK) »
I have a 10cM match with a person who Ancestry flags up as us having a common Ancestor and correctly predicts he is my 5th cousin once removed. Shared ancestors are my 4 x great grandparents and his 5 x ggp.
I also have a 10 cM match with his son, but Ancestry is not flagging up that we have common ancestors, because the shared Ancestors are his 6 x great grandparents - too many generations for Ancestry's algorithm.
In both cases the individuals are correctly linked to an extensive tree

Would be great if they expanded this out at a later date; thankfully, most of my brick walls are now further back than this. Interestingly though, the shared surnames feature seems to work back to the 1600s (possibly 1500s?).
North West, North Wales and Isle of Man:

Ward, Campbell, Cowin, Cowell, Parry, Fryer, Davies, Hughes, Briscoe, Jones, Spencer, Brownbill, Crowfoot, Nield, Randles, Youde

Gozo, Malta:

Apap / De Apapis, Camilleri, Grech, Muscat, Micallef, Custo, De Nasi

Angus, Scotland:

Ormond, Salmond, Cook, MacDonald

Offline clayton bradley

  • RootsChat Aristocrat
  • ******
  • Posts: 1,060
  • Census information Crown Copyright, from www.natio
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #15 on: Tuesday 08 June 21 19:50 BST (UK) »
Using the shared surnames I found a cluster of 9 Americans, from 23 cMs down to 10 cMs, descended from Cornelis Pieter Jongewaard born in the Zaanstad (Netherlands) 1641. The ancestor of the cluster left for America, mine remained in the Zaanstad and my father came to England in WWII. So it does seem you can go a long way back, in certain circumstances.
Broadley (Lancs all dates and Halifax bef 1654)

Offline Petros

  • RootsChat Senior
  • ****
  • Posts: 349
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #16 on: Tuesday 08 June 21 19:56 BST (UK) »
I actually have 9 6th cousins show up by common ancestors, most (all?) of whom appear to be genuine, some up to 14 cM

Offline LizzieL

  • RootsChat Marquessate
  • *******
  • Posts: 7,946
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: How far back can AncestryDNA's "Common Ancestors" feature work?
« Reply #17 on: Tuesday 08 June 21 20:26 BST (UK) »
Using the shared surnames I found a cluster of 9 Americans, from 23 cMs down to 10 cMs, descended from Cornelis Pieter Jongewaard born in the Zaanstad (Netherlands) 1641. The ancestor of the cluster left for America, mine remained in the Zaanstad and my father came to England in WWII. So it does seem you can go a long way back, in certain circumstances.

Did Ancestry flag this group up as having common ancestors with you? Or did you find them by your own research?
Berks / Oxon: Eltham, Annetts, Wiltshire (surname not county), Hawkins, Pembroke, Partridge
Dorset / Hants: Derham, Stride, Purkiss, Sibley
Yorkshire: Pottage, Carr, Blackburn, Depledge
Sussex: Goodyer, Christopher, Trevatt
Lanark: Scott (soldier went to Jersey CI)
Jersey: Fowler, Huelin, Scott