Ancestry Matches 7cM and less

[Winksetter]

For the record ( I have no idea whether Gedmatch has been hacked or not) I never load any sensitive or personal info onto any website unless vital, there is no need to with Gedmatch, so no data that cannot be freely accessed on Ancestry family trees is held other than dna data. So really I couldnt care less whether this is true or not. Same goes for FTDNA.

 

[Winksetter]

Ancestry backend services now overtaxed all the time when trying to mark lower matches. Proof of the fundamental issue here or Ancestry stopping users keeping matches???

 

[Pauline]

Ancestry backend services now overtaxed all the time when trying to mark lower matches. Proof of the fundamental issue here or Ancestry stopping users keeping matches???

I suspect that the main problem at present is lots of users trying to get their distant matches into groups etc before the changes are made.

But as I mentioned before, Ancestry acknowledge in the White Paper that reducing the demands on data storage is a consideration in the upcoming changes.

 

[Bob Spiers]

So really I couldnt care less whether this is true or not. Same goes for FTDNA.

Nothing for you to worry about unless you are wanted by Interpol, the US secret service or MI5/6 who are not averse to sampling DNA profiles (often quasi legally) found at both Gedmatch and & FTDNA. And of course one must not overlook IT savvy hackers with their own criminal motivations.

 

[Pauline]

Ancestry backend services now overtaxed all the time when trying to mark lower matches.

Seems to be working OK at present so I guess it's variable and you need to keep trying.

 

[Chris from Stansted]

Ancestry match losses. I have 128 common ancestors flagged by AncestryDNA indicating we have common ancestors, this number excludes three that are incorrect. 26 of these 128 have match values in the 7/6 cM range. 11 of these 26 are also in my brothers matches and flagged by AncestryDNA showing our common ancestors, the match value varies with one match being 26cM for my brother and only 7cM for me. Most match values with my brother vary by just a few cM. For 6 of my sub 8cM matches I also have one of their parents or siblings in my list with a higher then 8cM match value, this indicates that these are genuine matches.

I have triangulated, using GEDmatch, 3 of my matches of 7cM, these matches are on the same segment and portion.

It is probable that for some of these 26 matches the cM valve may be by chance but the relationships are correct. I have not been able to break any brick-walls using DNA but it has certainly confirmed my paperwork.

My feeling is that with this AncestryDNA matching change potential new matches will not appear.

 

[Bryman]

I think the White Paper actually acknowledges that one reason for the change is the need to reduce demand on data storage. For example, on page 14 they say:

"The cutoff of 8 cM was chosen after considering several factors. The first factor is data storage. Since the number of matching segments grows exponentially with decreasing length, we dramatically reduce the storage requirements of our matching database by increasing the cutoff."

Well spotted. I missed those sentences 'buried' in the discussion about "the phenomenon of decreasing accuracy with decreasing IBD length". I should have been more careful to read every sentence and understand the significance. I was beginning to lose the plot at that point.

My eyes were drawn too quickly to the following scatterplot shown in Figure 3.2 and I was wondering if there was any significance in the apparent secondary 'lines' formed either side of the central orange dotted line. Those concentrations of dots are displayed between about 11 cM on each axis and 30 cM on the orange line. Perhaps I was seeing shadows as there did not appear to be any acknowledgement or explanation of what I thought I saw. Too many similar charts examined many years ago when trying to make sense of particle physics experiments!

 

[Chris from Stansted]

Ancestry’s White Paper of 2016 used the same wording to change the cut-off point to 6cM. You can read it here. https://www.ancestry.com/cs/dna-help/matches/whitepaper

 

[Pauline]

All the relevant evidence should be in my previous post; I included it as a guide to others who might be faced with a similar situation.

Returning to this point, I am still bothered that there may be a danger that this could lead others in similar circumstances to draw conclusions from distant matches that cannot safely be reached. I do in fact have a very similar situation in my family where one of my 7G grandparents is alleged to have founded a dynasty in America following an emigration in the early 1700s.

I say 'alleged' because there is considerable uncertainty as to who actually went to America and if they had any descendants there. I have no distant matches at all in this line, but my sister has distant matches with descendants in both the documented English branch and the alleged American branch, as do some closer documented cousins from the English branch. So is this proof that the American family is descended from my 7G grandfather?

I would say not. Two things stand out in reaching this conclusion - firstly the highly dubious nature of many of the trees alleging the American connection, and secondly the sheer number of people claiming a descent from my ancestor. Rather than there being 'safety in numbers', the more Ancestry members claiming a descent from a particular person or couple, the higher the chance of one of those descendants appearing by chance among our distant DNA matches.

In fact, when I search the matches of my husband and his brother, I find they have a greater number of distant matches to supposed American descendants of my 7G grandfather than my sister does.

Weighing all of this up, I have to conclude that despite the apparent DNA evidence, it cannot confirm either my research or the alleged American link.

 

[peter]

Then if this was my family I would not view these DNA matches as confirmation, only that it seems consistent with other evidence.

But this is precisely the kind of situation where we each have to make our own judgement and much will depend on the strength of the other evidence, and how much weight this DNA evidence is being expected to provide.

As you say, we have to make our own judgement. I've been working with probability and statistics for most of my life, so I know the value of information from an independent source - and in this case I have DNA information from two independent sources. No matter how many historical records I might find, I couldn't possibly achieve the same level of certainty without DNA, because records alone cannot prove anything.

 

[Pauline]

I know the value of information from an independent source - and in this case I have DNA information from two independent sources.

Yes I understand that but it isn’t what I was getting at. It’s about the reliability of the DNA information - what is the statistical likelihood of a small shared segment being identical by descent, and even if it is, what are the chances of determining which ancestor it has been inherited from?

 

[peter]

Yes I understand that but it isn’t what I was getting at. It’s about the reliability of the DNA information - what is the statistical likelihood of a small shared segment being identical by descent, and even if it is, what are the chances of determining which ancestor it has been inherited from?

As I explained in my earlier post, in this case the chances are very high. Of course, this wouldn't be the situation had my cousins' ancestor not emigrated to the US over 300 years ago, virtually eliminating the possibility of a more recent match involving ancestors yet to be identified.

The main point I was making originally is that Blaine Bettinger was writing from the perspective of an American matched with other Americans - his evaluation was correct in those circumstances, but for LostCousins members, most of whom aren't American but do have thousands of matches with Americans, the situation could be different (depending when their cousins migrated).

 

[Pauline]

As I explained in my earlier post, in this case the chances are very high. Of course, this wouldn't be the situation had my cousins' ancestor not emigrated to the US over 300 years ago, virtually eliminating the possibility of a more recent match involving ancestors yet to be identified.

And I'm afraid this is where you lose me, and I don't understand the logic you are using here - it's coming across to me as a bit of a circular argument.

But if for you the chances are high that the shared segments are IBD, and inherited from your 7G grandparents Francis Medley and Mary Linnett, then what are the implications for others of us who have distant matches with descendants of this same couple?

I don't have any matches with descendants of this couple (at least, none with trees) but my husband has one, my brother in law 4 and my sister 3 - all sharing 6 or 7 cM. My husband's match (which he has in common with his brother) has a private tree so I can't see where the connection is, but his brother's other 3 matches are with American descendants of the emigrant son, diverging into 3 different lines of descent by the mid 18th century. Two of my sister's matches are in the American branch, again diverging by 1750, while the other is with a descendant of the son Francis in Essex.

What conclusions, if any, should I be drawing from these matches - bearing in mind that I haven't personally verified any of the trees concerned?

 

[peter]

What conclusions, if any, should I be drawing from these matches - bearing in mind that I haven't personally verified any of the trees concerned?

It sounds to me as if the two of you are in much the same position as me, assuming that you had already documented your ancestry back to Francis Medley & Mary Linnett. We should probably compare matches since it's unlikely we match to the same cousins.

 

[Pauline]

assuming that you had already documented your ancestry back to Francis Medley & Mary Linnett.

No, neither of us has any known connection to this couple nor to any other Medleys.

 

[peter]

No, neither of us has any known connection to this couple nor to any other Medleys.

I imagine that like most of us you have quite a few 'brick walls' in the last 9 generations? It's possible that on the other side of one of those 'brick walls' are ancestors you share with Francis Medley or Mary Linnett.

As a control I searched the matches of three of my cousins whose tests I manage, but who are on the other side of my tree. None of the public trees of their matches include Francis Medley and Mary Linnett.

 

[Pauline]

It's possible that on the other side of one of those 'brick walls' are ancestors you share with Francis Medley or Mary Linnett.

Yes, it's possible, but on the whole I'm inclined to think it's somewhat more likely that these are not valid matches.

But I still don't understand your reasoning with your own matches in this line (as I wrote in #73 above).

 

[peter]

Yes, it's possible, but on the whole I'm inclined to think it's somewhat more likely that these are not valid matches.

But I still don't understand your reasoning with your own matches in this line (as I wrote in #73 above).

I had a documented connection to Francis Medley and Mary Linnett before I found the DNA matches. When two completely independent processes lead to the same conclusion, it's very likely that it's correct.

Without DNA we rarely have this luxury - the so-called Genealogical Proof Standard is largely smoke and mirrors because records are rarely independent of one another.

 

[Pauline]

I had a documented connection to Francis Medley and Mary Linnett before I found the DNA matches. When two completely independent processes lead to the same conclusion, it's very likely that it's correct.

I'm afraid this still seems to me ominously like a circular argument, and that the two strands of evidence (DNA & documentary) are not being used independently of each other here.

When we already have a documented connection with a distant DNA match, there is a considerable risk of this influencing our assessment of the likely validity of small shared segments, at which point the DNA evidence ceases to be independent.

Without a documented connection we are likely to assess small shared segments more objectively, and to recognise that these are not only more likely to be invalid (ie not IBD) than valid, but that pinning down any valid segments to a particular ancestor is verging on the impossible.

But I guess we will never agree on this so should probably draw this discussion to a close now (interesting though I have found it). I continue to believe that small shared segments should be treated with extreme caution, and that we all need to take care not to fall into the circular argument trap.

That said, I think you quite right in saying that documentary evidence is far from infallible, although it is in my view more likely to be reliable than anything gleaned from distant DNA matches.

 

[Bryman]

I'm afraid this still seems to me ominously like a circular argument, and that the two strands of evidence (DNA & documentary) are not being used independently of each other here.

Sorry if I am continuing this discussion after you wish to close it down but I do not understand why you believe this to be a circular argument. I know nothing about this particular instance other than what I have read in this discussion but I believe that the DNA and documentary evidence should be considered independently. They do not offer proof that they must both be correct but the two together do make that outcome more likely than if only one was known.

I agree that without a documented connection, small shared segments are more likely to be invalid than valid, but that does not mean that they must be invalid. The addition of the documented connection would suggest that the unlikely connection is in fact valid, however great the odds are that it should not be so.