Author Topic: The perils of Ancestry trees  (Read 868 times)

Offline Petros

  • RootsChat Senior
  • ****
  • Posts: 261
    • View Profile
The perils of Ancestry trees
« on: Sunday 15 November 20 12:04 GMT (UK) »
Some time ago I was contacted by a 37 cM match who has a tree of 729 people on Ancestry. However, we couldn't spot any obvious link with nearly all his tree being from parts of England where I have no ancestry.

The only possible overlap in birth locations was his  3 x great grandmother but the surname given was no match. When I looked again the German surname given for her seemed implausible for rural Bedfordshire in the 19th century. And checking the GRO index for the maiden name of the mother of his GGF confirmed this, and also provided the key clue as to the link. The tree he, and others, had posted had used the wrong marriage (3 years later than reality). The real marriage having been in London it was possible to confirm that this was of my relative to his 3 x great GF.

His tree also indicated that his 3 x great grandmother had marrried again and this provided the clue to our link to a shared match. Her true also had an erroneous assignment of a parentage which again hid the true connection.

Offline Flemming

  • RootsChat Veteran
  • *****
  • Posts: 624
  • Census information Crown Copyright, from
    • View Profile
Re: The perils of Ancestry trees
« Reply #1 on: Wednesday 18 November 20 13:07 GMT (UK) »
I think this is one of the biggest challenges for Ancestry in finding common ancestors and not sure how it can be addressed. For example, I have one match who's gone wrong with their grandmother (just say she was) called Ethel Bloggs born in Worcs, but my match has picked up another one born in Sheffield. The former isn't in my line, the latter is, and so Ancestry picks up a common ancestor when there isn't one. Our real common ancestors are on another line but Ancestry can't see this. Their new format, apparently still in Beta test, might get round this as it shows multiple connection options.

There are so many simple mistakes in trees (we're all fallable) that make it difficult to find the connection, although perhaps that's part of the challenge. It would be nice if people were more open to being challenged on info in their trees but there are many posts on here about rude reactions to trying to help - how very dare you and so on. And it cuts the other way. I had a stiff lecture from one match about my own line and, when I provided the evidence to support my tree, they refused to engage any further.

At least your own match was willing to review their work, which is nice.

Offline GingerVicky

  • RootsChat Senior
  • ****
  • Posts: 366
  • Census information Crown Copyright, from
    • View Profile
Re: The perils of Ancestry trees
« Reply #2 on: Wednesday 18 November 20 21:21 GMT (UK) »
That is amazing that you managed to figure it out. I think there should be a send message to all affected message as the amount of trees that have just copied others people's mistakes is amazing. I know mine is not perfect but I'm definitely open to being challenged about it.


  • Guest
Re: The perils of Ancestry trees
« Reply #3 on: Wednesday 18 November 20 23:17 GMT (UK) »
When we started our Ancestry Tree building we started on my Wife’s family.

Dorset OPC records are brilliant and we built the Dorset side of her tree fairly quickly using the OPC records to fill in and verify Ancestry blanks.

We got back to a xGGP with the surname Parsons and in the hints was a batch of family trees.

The first one looked promising, plenty of attached records, so we opened the tree and started to follow the line back.

Then we started to get Sirs, then a King.

Was this too good to be true, Royalty from a family of farm labourers.

Then we went back to the starting person and examined the tree more closely and found one was born sixty miles away and then from them the generations back were all in Wiltshire on this tree.

So we then search again for the same Parsons and found a baptism for them in the same village as all their descendants.  We could then follow the Parsons line in the same village back a few more generations.  Much happier then to have an accurate tree with each person supported by as much records as possible.

I did send a message to the owner of the tree in error suggesting a different baptism, but never received a response.  We did not bother sending any more messages pointing out the error to any other tree builder, not worth wasting our energy.

So we learned very early not to trust Ancestry trees, to paraphrase Captain Barbosa from Pirates of the Caribbean ... they are loose guidelines.

Offline brigidmac

  • RootsChat Marquessate
  • *******
  • Posts: 3,548
  • Computer incompetent but stiil trying
    • View Profile
Re: The perils of Ancestry trees
« Reply #4 on: Sunday 22 November 20 16:51 GMT (UK) »
100cm sounds like a big match but you still have to go back to great grandparents to find link

I missed a link because I hadnt realised someone had remarried and had a son with second wife who was lady matching me

There are all kinds of pitfalls.
Id be gateful to any one pointing out mistakes or omisions
Roberts,Fellman.Macdermid smith jones,Bloch,Irvine,Hallis Stevenson

Offline midlandslass

  • RootsChat Extra
  • **
  • Posts: 43
  • Census information Crown Copyright, from
    • View Profile
Re: The perils of Ancestry trees
« Reply #5 on: Sunday 22 November 20 18:32 GMT (UK) »
I keep my tree private so that any mistakes I make aren't perpetuated but am happy to share my tree with matches.
One tree that shares common ancestors had unfortunately picked the wrong lady, our common ancestors were from Cheltenham but he'd chosen a lady from Essex with the same name,
I have pointed out to them that they have the wrong lady (they reside in the USA) but the tree still has the wrong person.
Unfortunately it seems others have copied his tree and copied the wrong lady, and yes I'm 100% positive my tree is correct. :)

Hollingsworth - Lincs, Notts       Webster - Notts
Wakefield - Oxfordshire             Moor - Lincs
Green - Lincs                              Thorpe - Lincs
Duckmanton - Derbys, Notts      Clutterbuck - Gloucs 
Willetts - Worcestershire
Shaw - Warwickshire                  Dyer - Gloucs
Priest - Worcs                            Haden - Worcs
Hewitt - Worcs                           Bissell - Worcs
Scarfe - Lincs                            Tooley - Lincs
Turnell/Turnill Notts, Nhants, Camb