RootsChat.Com
General => Ancestral Family Tree DNA Testing => Topic started by: GailB on Friday 21 November 25 09:34 GMT (UK)
-
There is a new look for Thrulines on Ancestry. Not sure I like it, but at a glance it is much easier to see how many matches for each ancestor.
-
Interesting.
It has not reached our system yet.
-
Must be a gradual rollout.
-
You seem to have a choice between old style and new style and you can switch between them. A minor improvement if you prefer the new style. I do wonder who/how the priority for developments is set. Certainly not near the top of my wish list.
-
I find it a bit easier on the eye but the fundamental flaw of relying on inaccurate trees is still there.
-
I still have the old style, cannot see anything to switch between 2 styles.
Possibly rolling out gradually...
I agree Glen & Ikas and also would rather they concentrate on the DNA side and improvements rather than how things look
Cas
-
All very nice and pretty but they don't include all the dna matches that we identify but they don't, and worse yet they include all the nonsense ones that they identify and we can't reject.
-
Having used this new look Thrulines over the weekend I have changed my mind. The old style evaluation is not available. I have a minor issue with the new display as you now see much less of the relationship line on the screen which means you have to scroll up and down to see the entire relationship.
The major issue is that you cannot evaluate each step as you could do before. It seems to be AI led and it forces you to add individuals to your tree before you can evaluate steps down your matches line. I don't want to add individuals to my tree until I have evaluated all the steps in my matches line. Unless I am mistaken there is no way now to do that.
Also in the course of one hour it changed the relationship line of one match (admittedly low level) by replacing the CA at the top to a completely different ancestral pair. Thrulines always needed to be checked, of course, but I never saw a change in CA in such a short time before.
It looks like another miserable AI failure. Definitely a backwards step for me.
-
I got the new system on Saturday and don’t like it at all. The tree has very tiny print and shows only great grandparents to the present. Too much scrolling is involved to see older ancestors and those are the ones I am most interested in. The system was very balky this weekend, although it is better this morning. Initially, they did not offer the option of going back to the previous system. That is now coming and going. It first appeared on my desktop before my tablet. And of course, as someone else said, the problem of inaccurate trees remains.
-
There can be real issues with half relatives too, manually editing the relationship on the dna match list has no impact on the relationship displayed in tree view and the tree view version can't be edited.
Thrulines is trying to use two conflicting sets of data (dna based and tree based), to suggest who my grandfather is and gets it wrong every time. I can only see two ways to stop it, either accept the suggestion knowing it to be wrong (and mislead others), or enter 'unknown' or a fictitious name to occupy that spot and block Thrulines trying to fill it.
-
Having used this new look Thrulines over the weekend I have changed my mind. The old style evaluation is not available. I have a minor issue with the new display as you now see much less of the relationship line on the screen which means you have to scroll up and down to see the entire relationship.
The major issue is that you cannot evaluate each step as you could do before. It seems to be AI led and it forces you to add individuals to your tree before you can evaluate steps down your matches line. I don't want to add individuals to my tree until I have evaluated all the steps in my matches line. Unless I am mistaken there is no way now to do that.
You can still evaluate matches without adding them to your tree. Select one of the green people on the potential line, and then select "review and add to tree" from the right hand side pop up. This will not add the entry to your tree, but as you will see from the screenshot you can evaluate.
-
Thanks for that Gail. I will have another shot at evaluating and report back
-
Just checked, and I only have the old style Thrulines, with no option to change.
Presumably it’ll turn up at some stage.
Romilly.
-
You can still evaluate matches without adding them to your tree. Select one of the green people on the potential line, and then select "review and add to tree" from the right hand side pop up. This will not add the entry to your tree, but as you will see from the screenshot you can evaluate.
As far as I can see you can only evaluate if you start at or one below the top of your matches line. Then it works as you describe. If however you start further down the matches line you are faced with a different screen (attached). This adds the "missing" people and therefore forces you to evaluate from top down. Previously, from memory, you could evaluate anyone in the line.
-
As far as I can see you can only evaluate if you start at or one below the top of your matches line. Then it works as you describe. If however you start further down the matches line you are faced with a different screen (attached). This adds the "missing" people and therefore forces you to evaluate from top down. Previously, from memory, you could evaluate anyone in the line.
If you click on the blue "Review" on Catherine Cameron rather than "Continue" to the next step, it will take you to the evaluation screen like the one I posted.
-
In most cases you really should evaluate from the top down. If you spend hours evaluating five or six generations from the bottom up and then the last one doesn't connect to your tree it's very frustrating. Been there, done that etc !
-
If you click on the blue "Review" on Catherine Cameron rather than "Continue" to the next step, it will take you to the evaluation screen like the one I posted.
Yes, I realise that but I didn't want to start with Catherine Cameron. I wanted to start with a person two places below. I found out this morning it must have been a bug which they have fixed. Now I can start with anyone and just hit evaluate as I did previously. The screen I had yesterday has disappeared. Happier now but still agree the new display means much more scrolling.
-
In most cases you really should evaluate from the top down. If you spend hours evaluating five or six generations from the bottom up and then the last one doesn't connect to your tree it's very frustrating. Been there, done that etc !
I prefer to scan down quickly to identify any red flags and evaluate those first. I find it is often quicker.
-
All very nice and pretty but they don't include all the dna matches that we identify but they don't, and worse yet they include all the nonsense ones that they identify and we can't reject.
I have yet to receive the new version and quite disappointed. I would prefer they fix the evaluation to allow me to reject peoples dubious tree matches which are obviously wrong. Personally I don’t think AI should be part of family history when there are so many people believing in Ancestry’s claim “just enter a name….”.
-
Personally I don’t think AI should be part of family history when there are so many people believing in Ancestry’s claim “just enter a name….”.
To be fair I don't know for sure if the new display involves AI. I just think it has a feel of AI.
-
If it has the feel of AI then it's not very good. You shouldn't be able to tell the difference. Sadly you almost always can. I strongly believe that this has been unleashed on the world two years too soon, before it has been adequately developed and fully tested.
My list of bizarre AI experiences gets longer by the week.
Barclays: Say briefly what you are calling about.
Me: Change PIN.
Barclays: Did you say pin?
Zaph
-
There must be problems, I have been reverted to the original Thrulines view.
-
Thanks for that. I just checked and I’ve been reverted also.
-
I have been reverted too. Personally I hope it never comes back. Too much scrolling.
-
I have just looked and I have the new version ::)
-
I am sorry I don't understand what people mean here about 'accepting' and 'rejecting' information from Thru lines. Is there an option to add something to your tree from Thru lines? Personally, I don't add anything to my tree unless I have found it for myself using primary sources. Why would I even think about believing something in Thrulines when it is so obviously wrong most of the time. The best way to improve thrulines is to be meticulous with your own tree, extending it beyond your direct line. Surprisingly, Thru lines can then straighten itself out.
-
I am sorry I don't understand what people mean here about 'accepting' and 'rejecting' information from Thru lines. Is there an option to add something to your tree from Thru lines? Personally, I don't add anything to my tree unless I have found it for myself using primary sources. Why would I even think about believing something in Thrulines when it is so obviously wrong most of the time. The best way to improve thrulines is to be meticulous with your own tree, extending it beyond your direct line. Surprisingly, Thru lines can then straighten itself out.
Yes, there is an option to add directly to your tree. It gets worst, in that you do not even need to accurately research what you are adding.
That said I have had a lot of success being pedantic in working through Thrulines one person at a time since I use it as a rough unproven guideline.
If only people would be more careful it would help in them having accurate trees.
-
valk - the reason I would like to be able to reject Thrulines is so that it can present me with the next most likely solution.
I seems to me that although there is only one 'correct' solution, there are many 'possible' solutions, we are presented with the most likely of them. I want to reject the most likely if I don't like it and look at the next, etc etc.
MyHeritage's version of Thruline's shows more of their working out and I am assuming Ancestry's uses similar methodology.
I know that by adding dummy ancestors I can stop Thrulines but that's not exactly what is needed here.
-
I like it but i wish there was a reject function. On my mum's line it suggests links which match my fathers DNA. Its frustrating I cant reject this. Distinct lines are blurring between folks.
-
Thanks both for your replies. I generally reject the ludicrous matches by ignoring them. Ancestry does show the trees that might be the source of their calculations and adjusting those might provide a different thruline but contacting the owners of these does fall on stoney ground. People fall in love with their trees and don't like to change them even though the error of their ways is clear to see.
I was led to believe by a friend that DNA analysis would open up my tree but I now realise that I was doing remarkably well with paper records and can probably live with not knowing the father of the odd illegitimate child. Although I have spent many absorbing hours puzzling over mysterious matches whose family haven't left the US for 400 years, I would generally say that for my objectives DNA has been a waste of time.
-
DNA results are the icing on the cake for a good accurate paper based tree though that accuracy can also make Thrulines almost worthless. On the other hand DNA can uncover NPE's & be huge for adoptees & those trying to identify an unknown ancestor though some rely too heavily on Thrulines and other trees.
I'm an adoptee with NPE parents & an NPE grandparent and had to ditch twenty years of traditional research though the biggest issue for someone such as myself is low cM matches. Thrulines is no benefit until the link is established by which point the work is done so I don't really need it.
As someone who builds deep and wide floating branches the tree checker has been a big help, I might build two branches months apart and the tree checker has picked up duplicates that merge branches together. It's all about using the right tool for the job even if some of them are cheap and cheerful.
-
I am in a very similar position to Glen.
I have always been sceptical about Thrulines, in that in any suggested line from your Thruline match to your MRCA pick a person in the middle of the line and click on Evaluate. Look at multiple trees and note the parent, who frequently does not stack up as being valid as the parent shown in the predicted Thruline can be different.
Thrulines IMO is best treated as Rough Guidelines, work each person in the traditional Paper Chase manner. When you get the match linked you have the best end result, a documented line validated by DNA as being as accurate as it currently can be.
-
I can see the new view for ThruLines now. I like it.
It would be more useful if I could download it to my computer as a .pdf and print it. The same goes for the clusters.
-
I'll note that ThruLines still shows my full brother, whose DNA is attached to the same tree as mine, as my 3rd cousin.
It correctly shows me as a descendant of our great-grandfather (Mr. X). Then it creates a duplicate Mr. X as his brother, and shows my brother as the descendant of our Mr. X's brother.
He's my "3rd cousin," with 2680 shared cMs.
-
I'll note that ThruLines still shows my full brother, whose DNA is attached to the same tree as mine, as my 3rd cousin.
It correctly shows me as a descendant of our great-grandfather (Mr. X). Then it creates a duplicate Mr. X as his brother, and shows my brother as the descendant of our Mr. X's brother.
He's my "3rd cousin," with 2680 shared cMs.
I had a similar problem with my mother and my sister on both the old style view and the new. It deleted the tile for the mother and Thrulines showed her as my father. Depending on which ancestor's Thrulines I was looking at my mother showed up as my aunt, as a half sister to herself and other such rubbish.
To fix this I unlinked both myself, parents and sister from my tree and from each other and then I relinked us and it fixed the problem.
-
Thanks for the tip, GailB. What a pain in the butt, LOL.
-
I just had to post this supposed 'relative' suggested to me by ancestry on the basis of mutual DNA. I am 60% welsh 38% English with no ancestors in the USA and certainly not Hawaii. I suspect the names suggested may not have the same lineage as me...............
-
Hi,
Well, that does rather depend on the size of the match.
They seem quite recent dates.
Thomas Blake Glover was an Aberdonian in Japan in 1859.
It only takes one relative to tramp the world, that is the kind of thing that takes me down a rabbit hole.
I have a lot of DNA “relatives” who seem to be in America for over 200 years, but the matches seem very consistent so they must be real in some way.
Have you actually looked at their Origins, one may be 25% Welsh!
-
I just had to post this supposed 'relative' suggested to me by ancestry on the basis of mutual DNA. I am 60% welsh 38% English with no ancestors in the USA and certainly not Hawaii. I suspect the names suggested may not have the same lineage as me...............
If you share DNA with someone then you are Biologically related to them.
OK, maybe if the cM is low then they can be false, but they can also be right. My lowest DNA Match shares only 6cM with me and yet there is a document trail to support the relationship.
My Italian ancestors came to England in 1850, another batch of Italian relatives went to the USA where I now have hundreds of very distant DNA Cousins.
Do keep an open mind and all options on the table.
-
I remain open minded but the sheer number of supposed 3rd cousins spread across the world based on minute amounts of DNA seems overly optimistic to me.
-
I remain open minded but the sheer number of supposed 3rd cousins spread across the world based on minute amounts of DNA seems overly optimistic to me.
Yes, there is quite a spread, for me a lot of those in the 8cM up to 100+cM range are given as potential 3C’s.
Most of those of mine with below 15cM end up being tagged as Distant Cousins if I get then linked.
-
I spend a lot of time verifying the various ThruLines suggestions. The aspect that I find hardest is looking at the two previous generations before my suggested match.
I've done a lot of work following my tree back 200 years but it is very difficult following another branch down from an ancestor before 1850, to finally arrive at a suggested DNA match.
The next two or three generations down from a shared ancestor can usually be readily confirmed.
Does anybody else have this problem or a solution? Quite often the matches' parents and even grandparents are redacted or referred to by meaningless nicknames. And the lack of recent birth records add to my problem.
Zaph
-
I spend a lot of time verifying the various ThruLines suggestions. The aspect that I find hardest is looking at the two previous generations before my suggested match.
I've done a lot of work following my tree back 200 years but it is very difficult following another branch down from an ancestor before 1850, to finally arrive at a suggested DNA match.
The next two or three generations down from a shared ancestor can usually be readily confirmed.
Does anybody else have this problem or a solution? Quite often the matches' parents and even grandparents are redacted or referred to by meaningless nicknames. And the lack of recent birth records add to my problem.
Zaph
I've encountered the same thing. If one or more grandparent is visible, I'll look for obits or news articles that might list children and grandchildren. It's time-consuming.
If the person is a DNA match with me, I'll check all our shared matches, because sometimes he or she will have a sibling, parent or child who has also done their DNA and has a more complete tree with enough names for me to figure it out.
I'd be interested to hear what other people do.
-
Yes, I do both of those ideas, but I would also welcome hearing what other people do. Those last two generations are so tricky because of privacy. I wonder if family historians have a different view on data privacy from the rest of the world.
Now there's a topic for another conversation.
Zaph
-
Yes, I do both of those ideas, but I would also welcome hearing what other people do. Those last two generations are so tricky because of privacy. I wonder if family historians have a different view on data privacy from the rest of the world.
Now there's a topic for another conversation.
Zaph
Having a DNA Match with a useable User Name is the first thing to check, if they are female trying to determine if their surname is their married or birth surname?
An odd User Name can be traced but that depends upon shared matches and if they have a tree all be it a rudimentary one.
Fingers crossed that there is a tree to use as a guide.
Somewhere in the tree hopefully there will then be the Paternal and Maternal surnames and I use them, so where Ancestry shows them as Private I initially use the surnames with no forenames.
Then add the DNA Match and search for their birth, especially their location which I then use as the likely Marriage location. I then look for Siblings to see if the location remains consistent and add them to the tree.
With luck we then have the final branch of the tree accurate.
We start with the MRCA in the Thruline and work forward for each of the people.
-
Alan o,
One thing to bear in mind is that the maximum number of relatives if you go back 5 generations an each generation has 6 children, with no pedigree collapse is of the order of 130,000. If you allow for a bit of pedigree collapse which gives you visibility of the next generation, you are up to about 1.9 million, given the various European diasporas over the last 200 or so years, it it eminently possible to have relatives all over the world.
From a point of identifying people I tend to only check down to two generations before present, I am most interested in trying to push back the brick wall.
I will reach out if I think I may assist in someone’s search, but usually just build to a dangling branch from my tree.
-
My own tree is deep and wide, which means that most matches where ancestry suggest a thruline are easily "placed" as I already have them in my offline tree. Even if they're using an alias, so it's not clear which particular sibling they are, I know they are the son or dau of a particular couple.
Sometimes the thrulines are wrong, most commonly because of a skipped generation, but I don't think I have any matches where anc has suggested a thruline and I've not been able to either agree it or come up with a close alternative (I'm not including the private individuals who are the root people of entirely private 7 person trees).
Matches with no suggested thruline are a different matter, particularly small matches using aliases, but although it's been a slog, building my tree deep and wide is definitely paying off.
Jane :-)
-
My own tree is deep and wide, which means that most matches where ancestry suggest a thruline are easily "placed" as I already have them in my offline tree. Even if they're using an alias, so it's not clear which particular sibling they are, I know they are the son or dau of a particular couple.
Sometimes the thrulines are wrong, most commonly because of a skipped generation, but I don't think I have any matches where anc has suggested a thruline and I've not been able to either agree it or come up with a close alternative (I'm not including the private individuals who are the root people of entirely private 7 person trees).
Matches with no suggested thruline are a different matter, particularly small matches using aliases, but although it's been a slog, building my tree deep and wide is definitely paying off.
Jane :-)
I guess it depends on where the most recent generations were born and are living. If I didn't know my first cousins, it would be extremely difficult to impossible for me to track them. They haven't done DNA, they don't have any online trees, and we're all in Canada with strict privacy laws regarding official documentation.
And now I've got the song running through my head...
-
Didn't think of that Josephine, now I have the same earworm! I suppose old Hickory might be responsible for some of my NPEs ;)
Jane :-)
-
I just had to post this supposed 'relative' suggested to me by ancestry on the basis of mutual DNA. I am 60% welsh 38% English with no ancestors in the USA and certainly not Hawaii. I suspect the names suggested may not have the same lineage as me...............
I don't think it would show up on Thrulines but some UK people joined the navy (official or unofficial) and travelled to Pacific islands and left descendants there so that you both have common ancestors.
I would love to track the common ancestors for my matches in the Pacific Islands and New Zealand that must be DNA matches by this route.
-
Hi,
Well, that does rather depend on the size of the match.
They seem quite recent dates.
Thomas Blake Glover was an Aberdonian in Japan in 1859.
It only takes one relative to tramp the world, that is the kind of thing that takes me down a rabbit hole.
I have a lot of DNA “relatives” who seem to be in America for over 200 years, but the matches seem very consistent so they must be real in some way.
Have you actually looked at their Origins, one may be 25% Welsh!
I have matches in the US (I am UK) who have pedigree collapse trees and the same individuals turn up four or five times in their trees, I consider these legit matches and just mark them with the name of the common surname in case one day I can work out the common ancestor who had one child stay and one emigrate.
-
I have one line with naval associations. I have one maternal branch in the Plymouth area in the 1870s with 2 great great uncles in the RN. One made CPO and served 14 years: I have a list of all of the ships served on but he was on short engagements in UK waters or the Mediterranean fleet and certainly no long journeys to the Pacific.
The second one made Lt Comd and had a much longer career with service in Eqypt in 1882, West Indies in the 1880s and Venezaula in 1902.
His obituary lists a number of ships he served on. It rather helpfully lists the name of those he served on between enlistment in 1874 and 1890 when he was made a warrent officer. They seem to be Armoured Frigates in the Mediterranean fleet. So again no sign of any Far East service.
I fear that any 'relationship' that has resulted in a DNA mix with a Japanese family may have been not a love match but transactional so I doubt I will ever find him in their family tree!
Thank you all for your input.
-
Alan, "transactional"! I love it! And I usually hate euphemisms.
Zaph
-
alan o
Don’t forget that this is Ancestry’s best guess at the relationship, it could quite easily be a 4th or 5th cousin, so going back to the 1820’s in your timeline.
-
Thanks. That gets beyond my scope to identify uncles who might be candidates. Rather frustrating but them my cousin had his ancestry done and came back 99% Welsh with not many overseas matches so that was not a lot of variation for him to investigate!