RootsChat.Com
General => Ancestral Family Tree DNA Testing => Topic started by: Norfolk Nan on Monday 29 September 25 16:31 BST (UK)
-
I've not taken much notice of ThruLines until recently and so far, so good. But this weekend I've had some suggestions with the prompt to evaluate that are totally not true - is there any way to say no or is the only option to ignore them? Call me picky, but they annoy me >:(
-
I'm assuming you're referring to a 'Potential Ancestor' on the main Thrulines page?
If so then unfortunately no, there's no option to reject the suggestion. If and when you identify the correct person and add them to your tree then the unwanted suggestion should disappear, but it won't be immediate.
-
Yes and no. I do have 'potential ancestors' that certainly aren't so I'll make an effort to place the right person - if you're right that should do the trick. What is bugging me today is a DNA connection with a tree that suggests we share a common 5xggrandfather because he married in Shropshire and had more children at the same time as mine was having kids in Bedfordshire, all late 1700s. The only thing in common is the name! Grrr!
-
Hi, is there a DNA match?
That will be real, what you need to do is build the tree from that match.
There may be a mistake somewhere.
Either that or you have got pedigree collapse somewhere and the real link in further back in time.
I had a couple like that eventually found the link, but it involved looking in places I would never have thought of looking, several counties away, but the documentary evidence was there as well.
-
Hi, there is a tiny match - 9cMs - so I'll do that just to be sure ;D
-
I would say that the combination of a 9 cM match and same surname suggests it may well be real.
Do you have Pro Tools?
Any shared matches, or just a signgleton?
-
No shared matches so far. I expect we do have a connection somewhere, it's just that Ancestry says we have a common ancestor but that person doesn't show up in her tree. Clearly Ancestry/ThruLines offers other member trees to evaluate each generation but the first two they've offered are ludicrously inaccurate and I can't see where they match with my match (if you're still with me). The AI is definitely off bonk!
-
Hi,
Yes, I know exactly where you are coming from. That is why trying to build a tree is important. It is much easier to build up than down, and sometimes that mad AI is telling you something significant. I have pushed one branch back an extra couple of generations this way.
-
Verify the facts shown in ThruLines and then try and contact your DNA match to fill in the gap to their most recent listed ancestor.
Za
-
Totally agree about building out a tree, and will try, but the match hasn't been on the site in over a year so I doubt she will reply if I message her.
I know I sound deranged but it infuriates me that it's claimed AI is brilliant when it's clearly shoddy! My match doesn't have our 'common ancestor' in her tree, and the later generations used by Ancestry to prove the connection don't feature in her tree either. It's the worst example of cutting and pasting I've seen since my kids were in school!
-
Cutting and Pasting is exactly what so called AI does, there is no intelligence involved, all the rest is the invention of Marketing Departments across the world.
A Computer Lecturer I know insists on calling it MR (Machine Regurgitation). I keep that in mind when I'm looking at Thrulines and Common Ancestors. Another computer term is apt - GIGO - Ancestry seemingly does no curation on the trees it uses as input.
-
It's the same with the links down the side of a page - they used to be pretty good, now they are diabolical!
-
The fundamental rule is trust no one, Artificial Intelligence or otherwise.
That said Ancestry’s hints and Thru lines are searching huge quantities of data and sometimes come up with gems.
I would much rather it came up with rubbish stuff than filtered out the gems.
What does frustrate me is hints that appear and then disappear.
Would you really be searching census information in the US to see if your ancestors went there before returning to the UK, that is what Ancestry is giving us, however much dross there is with it.
-
Good point about the occasional gems - and no, I wouldn't want to miss out on those rare treats. It's just a shame that things are getting less efficient, not more. Or maybe, my expectations are too high?
My first post was written in the vague hope there would be a way to say 'this is wrong' - we have the opportunity when rejecting a hint, Ancestry wants to know why - so why not in ThruLines?
Btw, I did eject one irritant as suggested, I filled the gap in my tree and ThruLines deleted the suggestion PDQ!
-
Yes, that can work well, though I have some persistent ones, other people have accepted the link.
That said, take a note of the match, they can be below what Ancestry will normally show you, and may be a crucial link to that net generation!
My greatest frustration is that Ancestry do not do Thrulines further back. I have several known an proved links further back and I have frustrating matches that Thrulines might just make an interesting connection.
They have the technology, they just don't seem to want to roll it out.
-
Further back would be good for me too. I hoped DNA would help with earlier brickwalls too but no such luck. Still, some help, even if it has to be dredged and treated very cautiously, is better than none. We shouldn't moan, it's much easier than when I began my tree back in the 1980s!
-
Building trees and grouping matches for early relatives is very time consuming, I have spurts of enthusiasm for it.
I am hoping that the posted Ancestry updates might help in that regard.
-
Thrulines.
Can be pretty accurate, can be wildly in error, & everything in between.
I spent a fair amount of time going through the Thrulines and for a while I had linked into my tree over 130 DNA Matches, both via the Common Ancestor route and my the Hard Slog Method.
Once I had marked up a 5 generation Pedigree Chart with the predecessors who were the MRCA with each of my DNA Matches I could then see that I had DNA Matches linking to and via each of my Great Great Grandparents.
Then I set about creating Groups based on each of the 8 G GP’s, so 8 Groups as per the text below.
I then went through each of the DNA Matches who are in my Family Trees and assigned them to the appropriate Group.
As I was assigning each of them I also looked at the shared matches and assigned them to the same Group(s).
Next it was back to the tree and assigning an image of DNA Icon to each person in the branch from the DNA Match to the MRCA, this now allows me to easily follow the DNA pathways in my trees.
Now when I look through my DNA Matches I can see those who have not been placed into the trees, the Unknown have their own Icon and Group.
Since I originally created the text below I have added another Group, this one is labelled with a ? icon to signify Thrulines that do not compute.
A lot of work, yes, but now dealing with new DNA Matches who are of interest is an easy weekly task.
.
.
.
DNA Grouping.
Firstly print out a 5 Generation Pedigree Chart.
For the 8 Great Grandparents label each of them on the Pedigree Chart with a number, starting at 1 and going down the chart adding 2, then 3 etc up to 8. Hence each Great Grandparent is given their own unique number in the 1 to 8 range.
There is method in this seemingly madness approach.
In Ancestry DNA there is now the possibility of creating up to 64 DNA Groups and you can assign DNA Matches to applicable Groups. You can then filter the Groups so that only those in a selected Group will be visible and this is the Batch of DNA Matches that are going to be the main focus of your research, at least until you have gone as far as possible.
In my case I have created Groups as follows but Do Note that the first character of the name is displayed inside the coloured marker and hence why I chose to use 1 through 8 for the GGP’s:-
1 = 1 GGP
2 = 2 GGP
3 = 3 GGP
4 = 4 GGP
5 = 5 GGP
6 = 6 GGP
7 = 7 GGP
8 = 8 GGP
U = Unknown, these are the mystery ones, the ones left over.
As you build DNA matches into your tree you can assign the appropriate Group marker(s) to each of them.
Initially I would suggest you look at the highest DNA Match and then work down them to say those that share 75cM. Assign each DNA Match to the applicable GGP Group(s) and assign the Shared Matches, irrespective of the cM that they share with the DNA Match to the same Groups.
If you are short of DNA Matches move down to include DNA Matches who share 60 cM then down to 50 cM if necessary. Again assign Shared Matches to the same Groups.
With the DNA Matches who have not been assigned to any of the numbered 1 to 8 Groups, assign each and their Shared Matches to the UNKNOWN Group.
As an example say the Maternal Grandfather of the Home person is not known, then there should be no DNA match showing as included in Groups 5 and 6. These ungrouped DNA Matches will be the centre of research and all of them and their Shared DNA matches will be assigned to the UNKNOWN Group.
Filter to view the UNKNOWN Group and then work through them one by one.
Whilst this process may seem a long and arduous process it will provide a strong foundation to develop your Biological Family Tree not just in finding the unknowns but in expanding the Family Tree.
-
That is a painstaking and methodical way of getting matches into order and under control. I'm sure it helps if, like most of my tree, the people concerned do everything in the right order and keep things open and above board. But what about the others? :o
My biggest Ancestry match is a total mystery, her tree (her version nor mine) doesn't match mine in any way and she doesn't answer my dms so I'm reliant on our shared matches. The next biggest shared match has no tree at all, provides no clues and doesn't respond either. In all there are around 27 shared matches and those with a tree, however small, generally give a hint at how they share with each other - the same surname in the same towns - but not with me!
Obviously I'm looking at illegitimate births now but if one parent isn’t ever acknowledged there will always be some matches that won't be explained. And if this is back in the dusty past I'll never prove a thing. And they all stay in that box marked 'annoying mystery' ;D ;D
-
Nan, you are going to need a bigger box.
Zaph
-
;D ;D ;D ;D ;D ;D
-
Thrulines [....]
DNA Grouping [....]
Thank you so much for this, Biggles50. I'm going to print your instructions and give it a go.
-
That is a painstaking and methodical way of getting matches into order and under control. I'm sure it helps if, like most of my tree, the people concerned do everything in the right order and keep things open and above board. But what about the others? :o
My biggest Ancestry match is a total mystery, her tree (her version nor mine) doesn't match mine in any way and she doesn't answer my dms so I'm reliant on our shared matches. The next biggest shared match has no tree at all, provides no clues and doesn't respond either. In all there are around 27 shared matches and those with a tree, however small, generally give a hint at how they share with each other - the same surname in the same towns - but not with me!
Obviously I'm looking at illegitimate births now but if one parent isn’t ever acknowledged there will always be some matches that won't be explained. And if this is back in the dusty past I'll never prove a thing. And they all stay in that box marked 'annoying mystery' ;D ;D
This is one reason why I now always suggest that you try to get siblings and First Cousins on both parental sides to take a DNA test.
With DNA unearthing skeletons we need to make sure close Biological Family is actually Biological Family.
As I found out to my cost, First Cousins were not Biological First Cousins.
-
Yes and no. I do have 'potential ancestors' that certainly aren't so I'll make an effort to place the right person - if you're right that should do the trick. What is bugging me today is a DNA connection with a tree that suggests we share a common 5xggrandfather because he married in Shropshire and had more children at the same time as mine was having kids in Bedfordshire, all late 1700s. The only thing in common is the name! Grrr!
You can get rid of potential ancestors that you know are completely wrong, not only by inserting what you know is right (i.e. the correct ancestor) but also by inserting 'unknown' in the add father/mother box. Cos this doesn't match the potential that Ancestry are showing you it no longer displays their option.
By the way, potential ancestors update immediately if you just refresh your tree once you have filled in a box.
Pheno
-
As one particular fake offering annoyed me so much I did insert my own suggestion but it wasn't real. That worked but a name's a name and no doubt suggestions will appear eventually - I much prefer the 'unknown' idea, it's to the point. Thank you ;D