Ancestry Releases Updated Ethnicity Estimates – Hope You Still Have Your Kilt!

Ancestry has been rolling out their new DNA ethnicity results over the past couple of weeks. By now, pretty much all customers have updated results.

When you sign on and click on your DNA tab, you’ll see a message at the top that tells you whether you have new results or they are coming soon.

I wrote about how ethnicity results are calculated in the article, Ethnicity Testing – A Conundrum. You might want to take a minute and read the article because it applies to methods generally and is not specific to any one vendor.

Ethnicity analysis is quite accurate at the continental level, plus Jewish, but less so within continents like Europe. Your results will vary from vendor to vendor and from update to update with the same vendor over time.

To be very clear, your DNA doesn’t change – and neither does your genealogy, obviously – but the evaluation methods used by various vendors change as more people test, reference populations grow, and the vendors improve their algorithms.

Of course, “improve” is subjective. Changes that “improve” one person’s results have the exact opposite effect on other people.

The Eye of the Beholder

Every time vendors release new population or ethnicity results, everyone runs to check. Then – queue up either “they finally got it right” or teeth gnashing! 😊

Everyone hopes for “better” results – but expectations vary widely and how people determine what “better” means to them is quite subjective.

So yes, the accuracy of the results is truly in the eye of the beholder and often related to how much genealogy they’ve actually done. Surprises in your genealogy can equal surprises in your ethnicity too.

Quantitative Analysis

First, let’s be very clear – you do NOT inherit exactly half of the DNA of each of your distant ancestors in each generation. So you might have NO DNA of an ancestor several generations back in time and multiple segments contributed by another ancestor in the same generation. I wrote about how inheritance actually works in the article, Concepts: Inheritance.

Obviously, if you don’t carry a specific ancestor’s DNA, you also don’t carry any genetic markers for any portion of their ethnic heritage either.

Measuring

The best you can do in terms of ancestral ethnicity percentage expectations is to methodically analyze your tree for the geographic and ethnic heritage of your ancestors.

I explained how I calculated realistic ethnicity estimate percentages in the article, Concepts – Calculating Ethnicity Percentages.

In summary, I made a spreadsheet of my 64 great-great-great-great-great-grandparents, each of which, if the DNA was divided in exactly half and passed to the next generation, would contribute 1.56% of my DNA.

Vendors can typically measure geographically-associated DNA less than 1%. At some point, however, the segments are simply too small to reliably identify and associate with a geographic location or population.

Over time, how different vendors refer to and label different parts of the world both vary and change.

Region Names and Ancestral Assignment

I created a spreadsheet where I track both my “expected” DNA based on my genealogy and the amount of reported DNA from that region by each vendor. As I added vendor results, I sometimes had to add categories since their categories aren’t exactly the same as mine. You’ll observe this in the following sections.

You might notice the “inferred” category. I wrote about this in the Calculating Ethnicity Percentages article, but the inferred locations stem from situations like an unknown wife of a man who is living in England or Germany. We can probably infer that they are from that same country.

In the US, an earlier era spouse’s ethnicity might be inferred from marrying a Scot’s-Irish person, living in a Scots-Irish community or being a member of a Scots-Irish church, for example. Chances are very high that a Scots-Irish man’s wife is also from the “British Isles” someplace.

When creating my spreadsheet, I was intentionally conservative in my genealogical estimates.

Ancestry Update in General

Are there any trends or themes in this most recent Ancestry update? As a matter of fact, yes.

Everybody’s Scottish it seems. I hope you didn’t trade your kilt in for that liederhosen a few years ago, because it looks like you just might need that kilt again.

In fact, Ancestry wrote a blog article about why so many people now have Scotland as an ethnicity location, or have a higher percentage if they already showed Scotland before. I had to laugh, because let me summarize the net-net of the Ancestry article for you, the British Isles is “all mixed up,” meaning highly admixed of course. That’s pretty much the definition of my genealogy!

Another theme is that many testers have Scandinavian origins again.

Back in 2012, Ancestry had a “Scandinavian problem,” and pretty much everyone was Scandinavian in that release, even if they had nary a drop of Scandinavian ancestry. And no, not every person has an unknown paternity event and if they did, the Scandinavians cannot possibly be responsible for all of them. The Viking prowess was remarkable, but not THAT remarkable.

Eight years later, Scandinavian is back.

So, how did Ancestry do on my percentages?

Well, I’m Not Scottish…

In the greatest of ironies, I now show no Scottish at all. My calculations show 5.46%, and it’s probably higher because I descend from Scots-Irish that I can’t place in a location.

I guess I need to turn in my Campbell tartan along with a few others.

I do, however, have Norway back again, but no Scandinavian genealogy.

This chart shows all of the Ancestry updates over time, including this latest, plus a range column for this update.

In addition to the 2020 percentage numbers, I’ve included the ranges shown by Ancestry in the far right column for the 2020 update.

Ranges

When viewing your own results, be sure to click on the right arrow for a population to view the range.

You’ll be able to view the range and additional information.

In this case, Ancestry is confident that I have at least 35% DNA from England & Northwest Europe, and perhaps as much as 41%.

You’ll note that my range for the questionable Scandinavia is 0-5. The only two ethnicities that have ranges that do not include zero are England & Northwestern Europe and Germanic Europe.

My Opinion

I know that I have Native American heritage and that it’s reflected in my ethnicity – or should be.

23andMe results, below, shows me the chromosome locations of Native American segments, and when I track those segments back in time, they track to the ancestors in the Acadian population known to have married Native American partners as reflected in church records. Those ancestors were proven as Native through Y and mitochondrial DNA of their descendants which you can view in the Acadian AmerIndian DNA Project, here.

I wrote about using ethnicity segments identified at 23andMe with DNAPainter to triangulate ancestors in the article, Native American and Minority Ancestors Identified Using DNAPainter Plus Ethnicity Segments.

For me personally, including my Native heritage in my ethnicity results is important. I can’t “do” anything much with that at Ancestry, other than view my match’s shared ethnicity. Since my Native heritage doesn’t show at Ancestry, I can’t use it at all genetically.

Why is this important? Looking at a match on my Acadian line and seeing that we share at least some Native heritage MIGHT, just MIGHT be a hint about a common ancestor. Of course, that’s just a clue, because we might both be native from different sources. If my Native ethnicity is missing at Ancestry, I can’t do that. It’s worth noting that in 2017, Ancestry did report my Native heritage and other vendors do as well.

23andMe provides detailed, downloadable, segment information that translates into useful genealogical information. FamilyTreeDNA has announced that they will be providing ethnicity segment information as well after their new myOrigins release.

The Big 4

How do the Big 4 vendors stack up relative to my genealogy and ethnicity?

And for Native American heritage?

I took the liberty of highlighting which vendor is the closest to my estimated genealogy percentages, but want to remind you that these percentages will only be exactly accurate if the DNA is passed exactly in half in each generation, which doesn’t happen. Therefore, my genealogy is an educated estimate as well. Still, the results shouldn’t be WAY off.

An appropriate sanity check would be that my genealogy analysis and the DNA ethnicity results are relatively close. Many people think they are a lot more of something because those are the family stories they heard – but when they do the analysis, they realize that they might expect a different mixture. For example, my aunt told me that my paternal grandmother’s Appalachian family line was German and Jewish – and they are neither. However, German and Jewish lived in my head for a long time and that was what I initially expected to find.

What’s Next?

Both MyHeritage and Family Tree DNA are slated to release new versions of their population genetics tools – so you’ll be seeing new estimates from both vendors “soon.” Both announced at RootsTech they would deliver new results later in the year, and while I don’t have a release date for either vendor – keep in mind that both FamilyTreeDNA and MyHeritage have brought new labs online from scratch in record time in a humanitarian effort to fight Covid. This critically important work has assuredly interrupted their development schedules. You can read about that here and here.

Kudos to both vendors. Ethnicity can wait.

_____________________________________________________________

Disclosure

I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

Genealogy Products and Services

Genealogy Research

36 thoughts on “Ancestry Releases Updated Ethnicity Estimates – Hope You Still Have Your Kilt!

  1. Last week’s update was interesting for me. Everyone was actually making sure I was okay with it. I was like “I’m fine. Really!”

    What happened was that along with everyone becoming a little more or less Scottish, Italy was split into north and south. So, this meant that my dad, my great-aunt and myself now have a split between north and south Italy when our trees go right to southern Italy. As an added bonus, we got Greek and Arabian. So, as a result my 46% Southern Italian became 34% Southern Italy and 11% Northern Italy. You add them together and it’s still almost half and half of my tree is still Italian.

    I have no issue with it as it still gels with the tree and I understand the history of the region. Everything adds up to “Romans”. That’s my go-to answer. =)

    The Mediterranean has been a hotbed of invasion, colonization and people intermixing for centuries. Even my great-aunt, who was born in Italy, has shown some interesting changes that reflect the region’s history.

    The estimate is still an estimate and it’s cool to see how things changed. This does synch with Myheritage a bit. But, since you said that’s going to be changing, soon, I wonder how it’ll look. I’m excited and I can’t wait! Still….ya gotta love those Romans, right?

    34% Southern Italy and 11% Northern Italy. You add them together and it’s still almost half and half of my tree is still Italian.

    I have no issue with it as it still gels with the tree and I understand the history of the region. Everything adds up to “Romans”. That’s my go-to answer. =)

    The Mediterranean has been a hotbed of invasion, colonization and people intermixing for centuries. Even my great-aunt, who was born in Italy, has shown some interesting changes that reflect the region’s history.

    The estimate is still an estimate and it’s cool to see how things changed. This does synch with Myheritage a bit. But, since you said that’s going to be changing, soon, I wonder how it’ll look. I’m excited and I can’t wait! Still….ya gotta love those Romans, right?

  2. Roberta,
    just to interject my two cents…

    My confirmed grandfather is the son of a full Greek man, and a half Greek woman, which should make him around 75percent greek… which his test results on ancestry are showing that, but its dividing it by 40percent Greek and 40 percent Italian… (Which is what its doing to many Greeks)
    Now before the update it was showing me and my 2 sisters as No Greek or Italian at all… Which would of been impossible… but after the update it shows us now as having about 5 percent of each….

    On 23andme for my mother(his daughter) it pinpoints to North Agean which is exactly where the Greek is from..
    But it still shows me as having about both 5percent Greek and Italian…

    But just from observation, 23andme appears to be more accurate as far as ethnicity is concerned, and way more informative with the Dna painter and so forth…

    As for the Native American, my moms .2percent is still their at 90percent confidence levels, which i assume 23andme is the far better company for this type of Ethnicity!

    • I would like to point out that after doing a spread sheet as you suggest at the 4th great grandparent level… my Greek is only 12 percent… which is more in line with the roughly 10 percent ancestry is quoting…

      As opposed to me thinking I should be 18 percent if my grandfather should be 75 percent…

      So at this view point, ancestry appears to be pretty spot on!

  3. Thank you for an informative article. I am so happy to learn Family Tree DNA is planning to have segment ethnicity matching!

  4. My sister and I share 2877 cM. Looking at these new numbers I guess we came from different parents. It was better before and since I have done our tree back past 9 generations I have an idea where we came from. Russia and Scotland are not in the mix. Ancestry gets us right every other time. And in the off times it’s REALLY off!

  5. I was previously near enough 50% England from my mother and 50% Scotland/Norway from my father. Now I have 22% Irish out of my mother’s half despite her tree having nothing outside Yorkshire and Durham in 300 years.

  6. Hi Roberta,

    With my update on ancestry, things simply went back closer to the way they were before the last update, and for me, perceptually, the last update seemed to be a slight improvement. I wonder if my cousins got their native ancestry listed again, this go around, I will have to check. At least mine still shows up on 23andme.

    Like you, I also have Campbells on both sides of my tree. One group settled in KY and another in NY. When you add in Bruce, Henderson, Sutherland, McKaughan, etc, one does expect to see a lot more of Scotland represented. I kept some in the update but not what would be expected. I will have to try your estimating method above. Maybe my perceptions are off.

    I have a reasonably close living relative, a 3rd cousin who is listed in a Clan book, as well. I read that many people moved to Ulster Ireland, from Scotland, for a generation, before coming here as they sought employment, and many may have set sail from Ireland as well. Perhaps this all causes the mix up.

    Like you, I also have Norway, but my ancestors who settled in NY seem to have left good records before and after coming over in the 1600’s from Amsterdam and the original immigrant’s wife’s father’s family, Bradtt, seems to have been from Norway, so my Norway puzzle is solved.

    At least ancestry kept my African ancestry – thank goodness!

  7. I don’t believe Ancestry is getting any closer to reality, no matter how many times they attempt to “improve” their results. Well, maybe a little closer. As you know, I’ve been testing at all the companies from when they first started. To see how all but 23andMe have changed the results over time is, to to say the least, disconcerting. I appreciate how imprecise the results can be considering all you’ve pointed out over the years, but still, this is how these companies market their product. The results shouldn’t be so far off the mark that it creates so much skepticism amongst the general public regarding genetic testing in general. Genetics is much more of a science than these tests imply.

    My paternal line is mostly French Canadian with proven genealogy back to the arrival of the immigrants in the 1600s, with three non-French arriving after the Conquest. They amount to 6.25% of my total heritage, leaving 43.75% French. The closest is 23andMe with 32%, Ancestry went from 0, yes 0, to 15%.

    And like you, I’ve found genealogically proven Native American ancestry of an 8th great-grandmother on my paternal line. This shows on 23andMe as 1/2 of 1%, but not on Ancestry nor FTDNA as they do not consider percentages less than 1.

    For these reasons, I recommend 23andMe to friends and family who seek ethnicity results that are at least close to what the genealogy proves. Ancestry is great for cousin matches and their databases, and FTDNA for Y-DNA and mtDNA, but neither compare favorably on ethnicity percentages. Thanks as always for highlighting how your own tests compare.

  8. Various companies have trouble with Sephardic Jewish ancestry and I think Ancestry “removes” some DNA so those who are 100% Ashkenazi Jewish (or almost) don’t have so many misleadingly close matches. Of course, if you have distant AJ only, your estimates might not appear for that reason, at Ancestry. — I have more Swedish now (13%, probably too high) and a bit more Eastern European (4%) … overall a tad less homogenous. I like 23andMe better for ethnicity estimates but rely the most on matches (including some, even at Ancestry, who are 100% AJ, or mostly “Iberian”, or a combo of Middle Eastern and Iran/Cyprus/Turkey… which may or may not be truly IBD). I really like your take on ethnicity, shortcomings, not getting too fixated, etc. Thanks again for a great blog post.

  9. I’m probably noted for criticizing AncestryDNA here, particularly with regard to them not offering a chromosome browser, but I will have to give them credit for one thing, they have my ethnicity as 100% French, which corresponds with all of my genetic genealogy research. I just looked again yesterday at MyHeritage, and they still have me at 51% Iberian (Spain and Portugal). While there was a Portuguese settler in New France named Rodrigue, I doubt I’ve got that much of his DNA in me!

  10. A very interesting article – thank you for keeping such good records and doing your analysis. My family comes from England midlands all the way down to Kent and across to Devon and Cornwall and southern Catholic Munster in Ireland – and absolutely no Scottish ancestry in our tree as far back as I can reasonably go (unless we accept the family story that since we are by folk lore descended from Robin Hood and he was clearly descended also from Scottish Kings if you follow another story) then that must be our Ancestry Scottish side coming out – except for the huge amounts of Scottish now given to us all some 900 years later – ha ha! I felt the previous Ancestry ethnicities were fairly “right” and that this is all just going to confuse the heck out of newbies starting out with Ancestry. It was hard enough previously to get people to accept what NorthWest Europe meant in terms of English heritage – but that was more understandable than the concept that most of England (seemingly) was conquered by Scots invaders.

  11. For myself and my close relatives that have tested with Ancestry, this update is MUCH closer to our known genealogy than the last update. The only new question that it has brought up is the Welsh. I currently have 5% listed, but the Irish is now down to 2%? I have no known Welsh ancestors, but known Irish ancestors. They finally got the Scottish more correct at least!

  12. I’m glad I’m not the only one. Thought the new results were crazy. The Scotland portion through me off,big time and I even received some Norway.The only ethnic results that I really consider is from 23andme. They seem to be the most reliable.

  13. Love your article Roberta–I did not really pay attention to the ranges! Clicking on the arrows shows me my 12% Irish could be any amount from 0% to 13%, and my 30% Scotland could be anywhere between 0% to 30%, and my 5% Wales might be from 0% to 13%. English, well, that’s solid: the range for my 47% England could be between 44% and 48%. It seems like they’re hedging their bets for the British Isles. From reading your article, maybe because it is admixed. I’m going to try to figure out what tartan to get! Fun!

  14. Excellent article! I really appreciate your scientific approach and explanations. I have long felt the the ethnicity estimates are somewhat of a gimmick (I use my DNA results much more for the matches and trying to validate connections using matches’ trees.) Like you have done, I have my genealogy expectation of ethnicity based on my carefully researched tree – in my case, I am ~40% Irish and ~3% English (balance 25% Polish, 22% German, 5% French, 3% Dutch, 2% Swiss). Yet, the new Ancestry update has me at 35% Irish and 23% Scottish (!) I would say the update is much further to my expectation than the last. (also interesting is that my 100% French/Belgian father-in-law now shows up at 29% Scottish! lol!)

    I wonder if one thing not really considered is that perhaps some of these Scottish populations had moved to Ireland or France further back in time. The fact that my McCarty ancestors came to North America in 1737 from County Cork Ireland, or my McGill from County Donegal in 1848 does not preclude the possibility that perhaps some Scottish groups had come on a boat to Ireland in 1500? or 1100?

    Regardless, it is alot more interesting to me to know my immigrant ancestors came from a specific place (for some, I know the village) and their historical context, rather than some other possible migration several hundred (or thousand?) years before that…

  15. Thanks for writing all these wonderful articles, Roberta. Really enjoy your blog!

    I’ve tested with Ftdna, MyHeritage, and the Genographic Project. I’m 100% German, in fact my oldest paternal and maternal ancestors lived roughly fifty miles apart, some 400 years ago.

    Yet Ftdna has me at 59% East European, 27% British Isles, 9% Scandinavia, 2% Iberia, 2% Sephardic Jewish, and some East Middle East trace results. No West/Central Europe at all, though my dad and my sister have 32% and 53% respectively. Weird.

    MyHeritage tells me I’m 48% Scandinavian, 25%Northwest European, 19% Italian, plus some small amounts of Greek, Southern Italian, Balkan and 1%(!) East European.

    The Genographic Project, RIP, came closest. It compared me (35%Northwestern Europe, 24% Eastern Europe, 17% Southwestern Europe, 16% Italy and Southern Europe, and 8% Northeastern Europe) to their reference population of Germans (53% Northwestern Europe, 29% Eastern Europe, 16% Jewish Diaspora, and 2% Northeastern Europe) and correctly identified me as… German. My second closest reference population was French, which makes total sense, since I’m from Baden-Württemberg.

    So much admixture in European populations, no wonder ethnicity estimates are so complicated.

    By the way, your ancestors and mine may have crossed paths. My most recent research led me to an 8th great-grandmother from Beutelsbach/Schnait. 🙂

    • Your backgound sounds like mine… growing up with the idea I’m pretty much German (with a Huguenot from way back when). But there was a rumour about Jewish ancestry and of course Karow is a Sorb/Wendish name. With DNA testing and comparing, I get results like you – NOT 100% German or even close really. Are you not ALSO getting matches to people who are Jewish? My highest matches share segments 20-30 cM… most a bit less or involving 2 or so small segments. I think I can figure this out better, with time and effort. And I definitely embrace it all, knowing I wouldn’t be here without each and every one of those ancestors.

      • Heidi,
        When I first decided to test, it was for fun. Ever since reading about NatGeo’s Phoenician project, I was fascinated with the idea of tracking migration through thousands of years with the help of DNA. I was half convinced my results would be boring, and never expected that much interesting admixture! The various Gedmatch tools and projects are also pretty entertaining.

        About the matches – I really couldn’t say, I don’t do much in the way of checking them out. Germans don’t seem to be into DNA testing as much as Americans, so I have comparatively few matches to begin with (about 700 on Ftdna). The closest are in the four to five generations removed range; most don’t have trees or surnames listed (on MyHeritage or Ftdna). I’ve tracked down a few that I was able to fit into my tree, but that’s it.

        • WIth 2 parents born in Germany (1929 and 1931) and so far all my ancestors born in areas considered German (West Pomerania included), I get about 1,400 matches. 700 is quite low. I wonder why. (Note: my maternal grandmother had a half sibling (maybe) and my maternal grandfather definitey had one brother. That reduces the number of matches in the 2nd cousin range considerably. So I have mostly distant matches always. But still I have twice as many matches, after the USA, it’s Germany, Netherlands and Sweden. — You can sort for Ashkenazi Jewish among your matches. If you get many people who are 100% AJ… well you can speculate a bit more. BTW, until today, I was 100% NW European at FTDNA, no extra detail for the past 3 years approx. I can’t get on just now. Apparently there’s some kind of update there today.

  16. Your Dutch ancestry is mostly Frisian right? This region often gets a lot Scandinavian (in contrast to the south ot the country), perhaps that is part of the reason why you get Norwegian.

    • That’s possible. I’d feel better about it being legitimate if it didn’t come and go and was consistently found.

  17. At the time I first tested at Ancestry DNA, early 2015, they were using an older chip. I tested again in late 2017 after they switched to a newer DNA chip. So, I have two samples that I labeled V1 (version 1) and V2 (version 2).

    With the new updated ethnicity estimates, I am both amazed and perplexed. There is substantial difference in the estimates between V1 and V2. Between the two, there is a 3% difference in England & Northwestern Europe and Ireland, 4% difference in Sweden (perhaps because Norway disappears in V2), and now 2% and 1% for Spain and Wales, which are not present at all in V1. Why such a difference between two samplings from one body, mine?

  18. My grandfather was half French and half Northern Italian. That’s well confirmed, and I have lots of matches in northern Italy, especially. My latest results from AncestryDNA give me ZERO Italian. I looked at their map, and the city where my g grandfather came from (Biella) was almost on the French border. So I figured that Ancestry dumped my Italian ancestor into France. Nope. It says I’m only 2% French. My g grandmother was all French, from Lyon for many generations. So they’ve not identified my Italian or my French ancestors right. Ancestry will end up discrediting themselves with this latest ethnicity revision. What they need to do is to be brave and say that European ethnicities are cultural, not biological. This would be a good time to address the issue of tribal thinking, which borders on racism.

  19. My changes were pretty minor this time (as opposed to losing my Iberian peninsula and later my “southern European” in the last 2 updates!). But my “Sweden” moved completely over to Norway. I have no records pointing to either one, but I figure it’s tied to my north-central Germany ancestors. I’m not sure why there was that shift, but it wasn’t too upsetting to me. I like your spreadsheets–I may have to go back and cobble mine together!

  20. Mine is actually worse than before — but then, the 2019 update was in many ways worse than 2018, which was certainly worse than 2012. It just goes to show that increasing the size of reference panels and other forms of “tweaking” doesn’t always give you better results.

    Actually, from my standpoint even FTDNA is now doing a better job than Ancestry. (And they were *really* bad before MyOrigins 3.0). At least FTDNA is able to find my Spanish ancestry, which Ancestry is unable to do at all.

    Prior to the 2020 update, Ancestry was able to find a mere 1% Spanish, even though both of my maternal grandmother’s grandfather’s were actually immigrants from Spain. But at least the 1% meant that I could see a match’s Spanish even if the match elected to only show in-common ancestries. Now, I won’t be able to do that.

    Ancestry also did a flip-flop of my “Indigenous Americas” ancestry. In 2019, this was 1% “Indigenous Americas – North”. Now, I’m 1% “Indigenous Americas – Mexico” and <1% "Indigenous Americas – North".

    What's rather unfortunate on Ancestry's part is that the same "Indigenous Americas" ancestor would be likely to test in more than one "Indigenous Americas" category *if* Ancestry could test that person. So their descendants might also appear in different categories. But once again, the only way they'll be able to see a match's other IA categories is if that match has chosen to show *all* "ethnicities", and not just the ones that are in common.

    So previously, if a match had 1% "Indigenous Americas – Mexico" I would not be able to see that if the match did not choose to show all ethnicities. I've said it before, but I'll say it again: Ancestry knows that they should treat at least the "Indigenous Americas" categories as just one category for comparison purposes. So they should show these percentages to matches who have a percentage in *any* of them, whatever the match's election.

  21. Here’s another example of how Timber’s “discounting” works. My daughter has a match with reported sharing of 63 cM. Ancestry predicts this person as my daughter’s 4th-6th cousin.

    The thing is, my own sharing with this person is 160 cM and she’s actually my half 2nd cousin — we share a great grandfather but have different great grandmothers. So that means the relationship between my daughter and this person is actually (half) 2nd cousin once removed.

    Now, the unweighted cM is not that much more than Timber’s “adjustment” — just 74 cM — but it looks to me like a completely unnecessary adjustment. The larger number is easily believable, and is about half of my own sharing with the same person. Of course, the 74 *could* be correct, but I wouldn’t even know there was a difference without Ancestry’s decision to report unweighted sharing.

    With another half 2nd cousin — this one and I share a great grandmother and have different great grandfather’s — Ancestry reports my sharing as 68 cM. The unweighed cM is 78 cM. But guess what? Ancestry reports my daughter’s sharing with the same person as 70 cM with unweighted sharing of 79 cM! This means Ancestry thinks my daughter shares 2 cM more with her half 2nd cousin once removed than I do with the same person. Her “unweighted sharing” is still a bit higher than mine, by 1 cM. Still, it doesn’t look as if any adjustment by Timber was either necessary or helpful here.

    Or there’s this match: Ancestry says she and my daughter share 48 cM. Unweighted sharing, however, is 77 cM. My sharing is 136 cM, so the unweighted sharing for me is the same. Once again, this is one of my half 2nd cousins, and therefore one of my daughter’s (half) 2nd cousins once removed.

    I could go on and on with this, not only on my side of my daughter’s ancestry, but on her mother’s side as well. Example after example of where Timber adjusted sharing downward in a way that actually made much less sense than the unweighted sharing would have.

    Even if Timber is *sometimes* right, how much harm does it do that it is *often* wrong. I’ll give Ancestry credit for reporting unweighted cM — finally! — but you don’t even see it until you take a closer look at the match. The match can be pretty far down you list of “4th cousins”, or potentially even beyond, when it turns out they might have been a 3rd cousin or even closer.

  22. My etnicity estimate on ancestryDNA is 65% Swdish 30% Noregian and 5 % Finnish. I and all my ancestors the last 5-7 generations back are born in Sweden. So it clearly shows it is very difficult to separate swedish and Norwegian DNA. Ancestry clearly separates out DNA from the province of Värmland and the city of Arvika!! As well as from the province of Ångermanland. My maternal grandmother was from Ångermanland and my paternal grandfather from Värmland(which borders to Norway).Some of his ancesors came from the Arvika area but it is 200 years ago. Some of my American DNA matches know of Norwigian background(but not swedish).

  23. I have no complaints. I knew my ethnicity from my parents and what they told me about their ancestry. I am not of recent mixed origins like my fellow Australians but of the one origin and from a small part of the world where immigrants did not make much impact. The new estimate is close to my origin which makes sense. I am 99% of my ethnic group, with minor, questionable Middle Eastern ancestry. The point is that if you are not mixed as many Australians are being of English, Scottish, Welsh and Irish ancestry with minor German or other European ancestry, any estimate of your ancestry will be iffy, but if you come from a small village in a quiet part of a country with little immigration, you will get a better estimate.

    I have no Scottish ancestry, and did not receive any. I also do not understand why any dna company would break down ancestry beyond a couple of hundred years because beyond that time, many of our ancestors whether they were Scottish or otherwise would not pass down any dna to you, as their contribution would not make the cut when you were conceived.

  24. Pingback: Genetic Genealogy at 20 Years: Where Have We Been, Where Are We Going and What’s Important? | DNAeXplained – Genetic Genealogy

Leave a Reply to Roberta EstesCancel reply