How far back can AncestryDNA's "Common Ancestors" feature work?

[History Lives]

This is a question I've started wondering as all of the DNA matches I've managed to have a look at so far that have the predicted "Common Ancestors" leaf next to them and the "You may be related through __" have been for people related through 5x great grandparents or closer, which is the same as ThruLines.

Is this the limit or has anyone else found one that gives a predicted ancestor more distantly than that? If not, I assume this is a limitation of the algorithm and not a genetic limitation as I have a suspicion a few of my matches are from a more distant connection than that.

[davidft]

have a look at this diagram on DNApainter and you will see that it confirms that in most cases that is as far back as you can go

https://dnapainter.com/tools/sharedcmv4

[History Lives]

have a look at this diagram on DNApainter and you will see that it confirms that in most cases that is as far back as you can go

https://dnapainter.com/tools/sharedcmv4

I didn't mean in terms of centimorgans, but in terms of the actual algorithm that shows who your common ancestor is if both of you have this person in your tree. I know for a fact I have people who I'm related to further back showing up on my match list but the common ancestor feature isn't working for them.

[clayton bradley]

Ancestry only goes back to 5th ggparents. What is somewhat annoying is on the new feature where you can say which side a match belongs to, mother or father, I have several who are 5th cousin once removed so I add them as Distant Relationship and write the correct relationship in the note. In every case Ancestry has changed this back to their suggested 4th-6th.

[brigidmac]

I have a 6cm.match which goes back to 6x ggparents

I know this because they are my mothers 5 x ggparents and show up on her thru lines

Im just going to check dates

[Albufera32]

I think the answer to your question is that the label "common ancestor" as used by Ancestry means someone for whom they have established a (possible) Thru Line.

So Ancestry will only label a match as a common ancestor up to 5th great grandparents. You may be able to figure out more distant matches, but it won't label them common ancestors, even after you add them to your tree.

[brigidmac]

Thats interesting my mother has a 6th cousin showing up on her thru lines at 11cm
So if he doesnt have enough dna cm to match me he wont show up ??

Now a younger generation are testing we also get 4th cousins twice removed

[Albufera32]

If I understand it correctly (and it is perfectly possible that I  don't) the thru lines and common ancestor label go together. IF a match were further back than that yet still had enough shared dna to pass the minimum threshold for Ancestry, it would still appear as a match, but will never get the "common ancestor" label.

The same goes for shared matches - I have two matches, both of whom I have traced, and who, having chatted to one of them, I happen to know are in fact brother and sister. Yet although both appear as matches, neither appears as a shared match with the other, since the shared dna is below the threshold set by Ancestry.

In the case of your mother's 11cm match, the chances are that one more generation down he will be below the minimum to even show as a match, not to mention a common ancestor - but of course the problem with random things is they are, well, random, and it is possible you actually share that same 11cm segment (unlikely, but possible) so he still might show for you as a match at least.

But if he is a 6th cousin of your mother, he is your 7th cousin, so he won't show on your thrulines ever, and if I am correct, then he won't ever show a common ancestor label either. To be fair to Ancestry, that's one of the reasons they recommend people get any elder relatives they can to test. (Of course, the cynical view would be that it also generates more money for them.)

[phil57]

Thru Lines is only making a guess based on what it considers to be matching ancestors found in two or more users trees. Of course the assumptions made by Thru Lines rely on the information in those trees being correct. It may well be, but the not insignificant number of trees with errors in them that I have discovered on Ancestry makes me very wary of accepting any such indications as gospel without carrying out my own research on both connecting lines to prove or add support to the assumption. Too many trees on Ancestry seem to quote other Ancestry users trees as the sole source for corroboration, and there have been quite a few occasions where I have followed links on a particular individual through 20, 30 or more trees to find that no-one has a definitive source of proof - they seem to have just accepted an Ancestry hint or what another tree has told them without any thought or investigation