Are some 6-8 cM matches valid and valuable? If not, then are my 172 ThruLines that Ancestry created for me that include my 8 great-grandparents surnames at that level all wrong? Or the total of 552 ThruLines at 6 and 7 cMs all wrong?
We all know by now that about half of 6 and 7 cM matches will be identical by chance, meaning not valid, but that leaves about half that ARE valid. We need clues to be able to figure out IF these matches are valid, and the logical place to start is by utilizing three techniques.
- First, if both of our parents have tested, does the person also match our parent, and if a chromosome browser is available, on the same segment.
If the answer is no, no need to go any further, this match is not valid. If yes, then we know if phases through one generation and we need to keep looking for evidence.
- Second, the same litmus test, but with our closest known relatives that have tested. Does the match also match aunts, uncles, siblings, first cousins, or other known proven close relatives? Of course, if they match on the same segment, that’s family phasing and the beginning of triangulation and strongly, strongly suggests descent from the same common identified ancestor.
Note that Ancestry does NOT show you Shared Matches below 20 cM, so don’t assume those shared matches to family members don’t exist. Check your family members’ kits directly. Don’t rely only on Ancestry’s shared matches.
- Third, surnames and trees that suggest common ancestral lines of DNA matches. That’s what Ancestry does for us with ThruLines. Let’s take a look at what I’ve found sorting and grouping my 6-8 cM matches at Ancestry.
There’s way more information than I expected to find.
Focus on Grouping
With Ancestry’s upcoming purge of all DNA customers’ 6 and 7 cM matches, inclusive, I’ve been very focused on grouping and saving those matches for future use. Otherwise, they will be gone forever, along with my genetic connection and any useful genealogical information.
I’ve written about the upcoming Ancestry purge here, here and here – including preservation strategies and how to communicate with Ancestry to share your feelings about this topic if you so choose. Note that this disproportionately affects people seeking unknown ancestors a few generations back in time.
Raise your hand if you have no unknown ancestors before 1870 or so…
Ancestry’s 6-8 cM Matches
I’ve been recording statistics as I’ve been grouping and working with results, and thought I’d share what I’ve found with you.
I have a total of 92,931 matches at Ancestry. This includes endogamous Acadian, Mennonite and Brethren lines, which produce lots of matches, but also multiple German and Dutch lines of relatively recent immigrants with almost no testers. So it probably evens out.
You’ll note that of my matches, 3,757 are estimated by Ancestry to be 4th cousin or closer, and Ancestry categorizes the rest of them as Distant matches, from 6-20 cM, although some of those wind up being closer than 4th cousins.
I have 27,926 6-cM matches, 16,846 7-cM matches and 11,428 8-cM matches. I was initially saving 8-cM matches because Ancestry was initially rounding 7.6 up to 8 and the only way to save all 7-cM matches was to save all 8-cM matches. Last week, Ancestry added decimal points so you don’t have to save 8-cM matches anymore, just all 6 and 7.
Without additional tools, all of those matches are overwhelming – but that’s exactly WHY we need technologies such as clustering, triangulation, ThruLines which Ancestry provides, a chromosome browser, family phasing, shared matches below 20 cM, and more.
You can certainly look at known genealogy and make inferences about common ancestors when you match someone genetically, and that’s very useful in and of itself.
However, you need more than just the fact that you match someone to confirm that you share a specific common ancestor biologically, not just on paper. Having said that, just having the breadcrumb of a DNA match to lead you to your cousins isn’t a bad thing in and of itself.
Of my total matches:
- 18% are 7 cM
- 30% are 6 cM
- That’s a total of 48% of my matches that would have been lost later in August if I hadn’t grouped them.
Some people feel that matches at this level aren’t useful, but the line in the sand is very thin between a 7.99 cM deleted match and an 8.0 retained match where the former is lumped into the “not useful, so no big deal to lose” bucket and the other is just fine and potentially useful.
I get it, I really do, that everyone gets tired of explaining that NO, you can’t find one match and assume a valid connection, and yes, digging for evidence is work. There is no magic wand. Smaller or larger matches, they all need additional cumulative evidence to indicate that the match is valid, and how.
It’s time-consuming and frustrating educating people HOW to utilize all DNA matching appropriately. Those smaller matches take more effort to work with and require more evidence of legitimacy, but there are absolutely, assuredly many legitimate, useful, matches between 6-8 cM.
Furthermore, many of those matches reach back in time to those elusive ancestors we are seeking and can’t yet identify. We need more and better tools, not less data. Conversely, some 6-8 cM matches are as close as third or fourth cousins. I found 4 in one family and we’re sharing photos of our ancestors who were siblings, born in 1827 and 1829, respectively.
I’m not throwing half of my 6-8 cM coins away because some are gold and some are counterfeit.
If you are, I’ll take all of your coins and I’ll be happy to sort out the gold, thank you😊
Where’s the Gold?
You can search and sort in any number of ways at Ancestry. First, I checked to see how many of my 6 and 7 cM matches had common ancestors as identified by Ancestry via Thrulines.
|6 cM||7 cM||Total|
|Common Ancestors (ThruLines)||274||278||552|
If I had not grouped these, I would have lost all 552 matches that Ancestry connected to common ancestors through ThruLines. Of course, each connection needs to be individually verified using traditional genealogical record searches. Keep in mind that ThruLines can only find matches where people connect in trees.
Without these 6 and 7 cM matches, any connecting genetic path or breadcrumbs to these people is gone.
Since I can filter by segment match size and surname, combined, at Ancestry, I decided to take a look at my 6-7 cM matches that would be purged had I not grouped them, and see what I can discover by surname utilizing the surnames of my great-grandparents.
That’s just 3 generations for me, meaning I could expect to carry more of the DNA of these ancestors than of ancestors further back in time.
I started with the “Match name” of Estes, meaning that the person who took the test has that name. Of course, some women could use their married surname, so this doesn’t mean that my match to that person is via that surname. It’s just a starting point, but probably a good hint.
I had 12 Estes surname matches in the 6-7 cM range. Of those:
- 4 had no tree
- 1 had a private tree
- 1 had an unlinked tree
- None had common identified ancestors meaning ThruLines
- That leaves me with 7 candidates to work with directly, including the unlinked tree
- Of those, I knew how 5 of their trees connect to the Estes line
Of course, I have the benefit of having worked with the Estes genealogy for decades along with the benefit of trees and other resources not at Ancestry. Connecting these lines took me about 15 minutes. In essence, I’ve turned them into virtual “ThruLines” by identifying the common ancestor, even if Ancestry didn’t.
I have not yet worked with the rest of my surname matches in the same way, but by preserving them by grouping, I can in the future.
I searched for both the “Match Name” and the “Surname in the Matches’ Trees,” separately. Some who carry the surname aren’t going to have trees and conversely, finding the surname in your matches’ trees is by no means an indication that that particular surname or ancestor is why you’re matching. However, it’s a great hint and a place to begin your research, including shared matches.
Be sure to check alternate spellings of surnames too.
Note that a surname that can also be part of a name returns all possible connections. For example if I’m searching for the Lore surname and the name of my match is Loreal Jones, it will still appear in the Match name list. The same applies to the name of the managing person. However, scrolling through these is pretty easy.
So, what did I find?
I created this chart of what I discovered using the surnames of my great-grandparents along with common alternate spellings.
|Surname||Match Name||Surname in Matches’ Trees||Comments|
|Estes, Eastes||13 matches, no ThruLines||208 matches, 20 Thrulines|
|Bolton||6, no ThruLines||121, 14 Thrulines||All 6 surname matches have trees and I can place some immediately.|
|Vannoy, Van Noy||2, no ThruLines||49, 10 ThruLines||I can place 1 of the 2 surname matches and connect them to the Vannoy line. Their tree is unlinked and another is private. Checking the “include similar surnames box” resulted in 2355 results. Won’t do that again.|
|Ferverda, Fervida, Ferwerda||0||2, no ThruLines||Confirmed a common ancestor in the Netherlands with one tester. An 1860s immigrant line.|
|Miller||175, 1 ThruLine||2248, 95 ThruLines||Very common surname and Brethren. Shared matches, if over 20 cM which is Ancestry’s threshold would potentially be very helpful.|
|Clarkson, Claxton||2, no ThruLines||96, 22 ThruLines||I need to break down a brick wall in this line. Also, maybe someone has a photo of my great-grandmother. I was able to provide a photo of someone else’s ancestors discovered as a 6 and 7 cM match to 4 family members.|
|Lore, Lord||112, no ThruLines||209, 10 ThruLines||Acadian, endogamous. Lore is part of many other names.|
|Kirsch||0||18, 0 ThruLines||1850s German immigrant line. This was VERY helpful. I’ve already found previously unknown cousins and one line that I thought was defunct, isn’t.|
|Total||310, 1 ThruLine||2951, 171 ThruLines||Total 3261 matches and 172 ThruLines|
I’m not willing to throw these away.
Continue to Provide Feedback to Ancestry
I find the assertion that these smaller matches are neither accurate nor valuable simply mind-boggling. Clearly, as you can see above, these matches provide invaluable clues for us, as genealogists, to follow. Over time, I’ve proven many matches in this range (who have tested at or transferred to other vendors with a chromosome browser) to triangulate with several generations of family members using DNAPainter, so at least some matches are quite valid. And yes, we do have tools to accumulate evidence – the same exact tools we use for larger matches.
Imagine how much else is actually buried in those matches that could be distilled into useful information with technology tools.
I fully understand it’s in Ancestry’s best interest to delete these matches to free up processing resources, but I’m far from convinced that it’s in our best interest as avid genealogists.
I also realize that many if not most genealogists who aren’t as focused as many of you reading this article won’t notice or care, but that’s not the case for truly committed genealogists with years invested in this work. There’s valuable information there for those of us willing to commit our resources and invest our time to work on the matches.
The Proof is in the Pudding
The proof is in the results – those 3,261 surname matches that serve as immediate hints and 172 ThruLines that Ancestry themselves has assembled for us.
The more I work with these matches, the LESS convinced I am that they should be deleted. There is certainly chaff to be sifted and discarded, but Ancestry could take a more precise, surgical approach instead of a wholesale decapitation that will remove 48% of my matches and more for other people. I would certainly be more than happy to be part of a proactive discussion focusing on how to delete less useful matches or those we’ve determined to be invalid, but preserve the rest.
Of course, the easiest option would simply be for Ancestry to allow us to elect to retain current and elect to receive future 6-8 cM matches by checking a simple box and continue to provide those for those of us who care and are willing to work with them.
Yes, the remaining matches after the purge will indeed “be more accurate,” as Ancestry says, because fewer will be false, but many of the very matches you need to identify those elusive distant ancestors will almost assuredly be gone. The baby will have been thrown out with the bathwater.
It’s generally not any individual match itself, but groups or clusters of matches that point the way – shared matches and ThruLines. If half or more of the cluster we need is gone, with no way to connect the genetic dots, we may never discover the identity of those ancestors. That’s a shame, because it negates the very benefit of being in the largest autosomal database. In a way, both Ancestry and we as their clients are victims of their own success.
Perhaps Ancestry will yet reverse their decision and if not, perhaps Ancestry’s competitors will see an unfulfilled opportunity here. I’d be glad to be a part of those discussions as well.
Take a look. What valuable nuggets are hiding in your smaller matches? Be sure to group those matches to prevent their deletion.
Provide Feedback to Ancestry
There’s still time to provide your feedback to Ancestry if you don’t want to lose your 6-8 cM matches later this month. Ancestry needs to serve all of their genealogical customers who have taken DNA tests, not just the most convenient. I encourage Ancestry to develop useful tools as others have done instead of deleting the matches we need in order to unmask those unknown ancestors.
- Email Ancestry support at firstname.lastname@example.org although there have been reports from some that this email doesn’t work, so you may need to utilize another contact method.
- You can initiate an online “chat” here.
- Call ancestry support at 1-899-958-9124 although people have been reporting obtaining offshore call-centers and problems understanding representatives. You also may need to ask for a supervisor.
- Ancestry corporate headquarters phone number on the website is listed as 801-705-7000.
- You can’t post directly on Ancestry’s Facebook page, but you can comment on posts and you can message them.
- Ancestry’s Twitter feed is here.
Thank you so much.
DNA Purchases and Free Transfers
- FamilyTreeDNA – Y, mitochondrial and autosomal DNA testing
- MyHeritage DNA – ancestry autosomal DNA only, not health
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload – transfer your results from other vendors free
- AncestryDNA – autosomal DNA only
- 23andMe Ancestry – autosomal DNA only, no Health
- 23andMe Ancestry Plus Health
Genealogy Products and Services
- MyHeritage FREE Tree Builder – genealogy software for your computer
- MyHeritage Subscription with Free Trial
- Legacy Family Tree Webinars – genealogy and DNA classes, subscription-based, some free
- Legacy Family Tree Software – genealogy software for your computer
- Charting Companion – Charts and Reports to use with your genealogy software or FamilySearch
- Legacy Tree Genealogists – professional genealogy research