View Full Version : DNA Circles is the right idea, but POORLY executed

08-11-2015, 02:06 AM
Ancestry is taking us in the right direction with their concept of DNA Circles to auDNA verify ancestors on family trees. But, they are executing so very poorly on this beta that it is absolutely maddening. Ancestry DNA users are simply not connecting their DNA kits to their trees. I looked at my top 100 auDNA matches out of over 6,000 on Ancestry. Over 35% of these top 100 had yet to connect their kits. When I emailed and asked why, the most common responses were that they didn't know that it was necessary, didn't understand the benefit, didn't know how to get it done. The computer modeling to execute on DNA circle verification is not meaningful if there is not enough participation. It's not that the users don't want to participate, it's that they don't know they need to participate in order to make the results meaningful for everyone. Ancestry hasn't made that clear. Ancestry is not being proactive in following up to ensure kits are connected to trees. Finally, Ancestry should come up with a method to make this connection automatically happen.

In addition, it is not clear whether those people who lock their trees are also being factored into the modeling. If not, why not? Surely there is a way to include, but not disclose information for anyone with locked trees.

As it stands now, between those who have not connected their kits, and those who lock trees .... there are not enough willing participants to make DNA Circles meaningful.

Ancestry is really blowing this opportunity to race out ahead of the pack. Not only do they have a miserable website, now they've teamed it with an ineffective effort at auDNA verification. Shameful.

Mac von Frankfurt
08-11-2015, 05:00 AM
To many family researchers a distant cousin DNA match falls into the category of "bright shiny object". It's already raining data on the internet the last thing we need is folks having us hop and jump all around our tree. As I said in response to your other comment, if somebody has taken the time to list all known surnames on 23andMe I can usually figure out pretty quickly where the connection is. And if I can't find it quickly (or they haven't taken the time to list the surnames) then we probably won't find it or it is irrelevant to the lines we are each working on.

08-11-2015, 11:00 AM
I don't even know what my Ancestry kit number is. And I don't spend much time looking at the few circles I have. I browse the trees/names of closer matches now and then. Mostly I spend my time in the old fashion genealogy section.

08-12-2015, 04:17 AM
You guys are nuts. Genealogy sections don't mean SQUAT any more. Unless your Ancestors are au or Y DNA verified, I don't care how much paperwork you have to back up your family tree ... your tree is a joke without DNA verification. Why? Have you bothered to check the bast*rds lists in county court records from about 1785 to 1850? ... if you did, you'd throw your *#%$ tree out the window. It's worthless.

The game of old in family tree building was spending months parked in a library. The game of new is to spend days on Ancestry leveraging off other people's trees. Then, using that info, put your best guess out there, submit DNA for yourself, a couple cousins, your parents, and let the computer do the rest. Then, you investigate anything not getting DNA verification. If you do anything otherwise, your tree isn't worth the paper it's written on. I can find the county court records bast*rds lists that proves it.

And to you Mac Von Franfurter ... you're doubly nuts. I used to manually look for matches using a Chromosome map and FTDNA data. I'd pound my chest if I found one. Now that I'm seeing Ancestry tools and data, I realize just how very stupid and naive the manual process is. It is HIGHLY unusual for any two people to have anything less than 8,9,10 surname overlaps out to 7 or 8 generations. I have a multitude of people where I have 20+ surname overlaps. Think about it. You get 8 generations out, you're talking at least 512 family surnames. It's not unusual at all to have a multitude of overlaps. When you manually find a match, you don't have a clue that the match is the result of that single common set of ancestors you spot on corresponding trees. It could come from any of a multitude of common ancestors. And, unless both you and the other party have fully developed trees out 7 - 8 generations, which very few people do ... there's no way for you to know for sure. These are the reasons that the matching process, done manually, is simply hogwash, absolutely meaningless. You must have a computer bounce the DNA against the trees and through a series of highly complex math and statistical models, determine if that match pertains to a specific set of ancestors on both trees ... DNA verification. And right now, the only service that offers this feature is Ancestry. And soon, it will put Genealogy research/libraries out of business, because these models will become significantly more accurate over time as the base of data grows. At some point in the future, you will be able to spit in a test tube and immediately receive a printout of all your ancestors with a probability of accuracy attached at each node on your tree.

You guys need to stop this silliness of doing tons of research and wasting massive amounts of time manually tracking down common ancestors. You're not accomplishing anything. Unless your ancestors on your trees get DNA verified by a computer model ... you don't have squat to hang your hat on. If you guys were smart, you'd start getting up to speed on what this DNA Circles verification at Ancestry is all about, and how to effectively use it.

Mac von Frankfurt
08-12-2015, 04:52 AM
You guys need to stop this silliness of doing tons of research and wasting massive amounts of time manually tracking down common ancestors. You're not accomplishing anything. Unless your ancestors on your trees get DNA verified by a computer model ... you don't have squat to hang your hat on. If you guys were smart, you'd start getting up to speed on what this DNA Circles verification at Ancestry is all about, and how to effectively use it.

We just have different goals is all.

I am more of a family social historian. I enjoy learning about the area the family lived in and what their life was like. I typically spend about a year studying one branch over about a century. Then we go to the area and visit the places that are part of the story as well as the libraries and genealogical societies that have information not available online or in microfiche repositories.

(Social history is a broad branch of history that studies the experiences of ordinary people in the past.)

08-12-2015, 02:23 PM

Having been a very early advocate of ANCESTRY atDNA and having tested at all 3 companies I think I have a bit to add to this discussion. There's lots that is maddening about ANCESTRY. Tha failure of testers to connect their DNA to their trees is frustrating however I do not think this can be blamed on ANCESTRY. There are many reasons people chose not to do so. Including that they are adopted or donor conceived (thus have no tree), they tested to get their ancestral makeup (and they do not care about genealogy), they are just starting out (having no tree), they have discovered some discrepancies between the DNA and their oaper tree (and have removed the connection for now while they sort it out) and finally they tested at the request if a family member (whi has agreed to keep them out if it).

AS FOR traditional genealogy being a thing if the past. That is simply ludicrous. Yes it is important to verify with DNa eveidence and I have done so with much of my tree (2nd great grandparents are all confirmed through triangulation) so in my tree people appear to be who I thought they were. You can see on this page all listed in red are confirmed https://sites.google.com/site/wheatonsurname/about-me/more-about-kelly. Traditional genealogy works hand in hand with genetic information. Both can be misinterpreted. Even if I have 7 children of an ancestral couple of 3rd great grandparents And they all agree that John JONES was the progenitor it may not be so if Mrs JONES had a love that noone knew about. You coukd oerhpas use YDNA to verify but there is always a certain amount of ambiguity in assigning DNA to specific ancestors. But We do the best we can.
As for the CIRCLES they work rather well as I have 28 of them--- I do find the two for aunt/uncles curious.

08-13-2015, 05:15 AM
Mac - Yep. I understand. To each, his own. My goal is a DNA verified tree. Yours is knowledge, history, hobby. Nothing wrong with either.

KWheaton - nice post. Nice work on your tree. Nicely documented. I am envious. Gives me something to strive for.

Also, reminded me of the frustrations of genealogy research. I had this Ancestor, Mary Titman. I had 2 abstracts with her name clearly indicated as Titman. But, for the life of me, I couldn't find her parents. I couldn't locate any other Titman's in the county, surrounding county, the state, or surrounding states. I spent a week looking for Titmans. I was swearing at Titmans under my breath and once screamed the name out loud in the library. Everyone thought I was cussing. It was driving me nuts. Then, out of frustration, I drove three hours to the county, pulled the underlying court case and land record. Head banging ensued. It wasn't Titman. It was Tillman. I concluded that the author of the abstracts simply wanted to play a cruel joke on Mary's ancestors. I was reminded of my Titman hunt when I saw the Tilghman and Tillman on your website. I'll check your GEDMatch file numbers later to see if there's any chance we're related.

08-13-2015, 05:48 AM
KWheaton - I have a question about Triangulation though. How can you effectively triangulate manually when you have 10,15,20 surname matches with just about everybody on your DNA match list? And, if you do it manually, how can you be 100% confident that you spotted all the possible surname overlaps, such that there is no other possible surname match accounting for the one you're investigating? And finally, how effective is triangulation when you're working with trees that aren't fully blown out 4 to 5 generations? Just about everyone has a couple, if not more, brick walls after a 3 or 4 generations.

See, I have a different view on triangulation. I don't think it's possible ... without the power of a computer doing the work. I have come to realize this after seeing the mind blowing surname overlaps I have with most people on my DNA match list. That, and just how underdeveloped the average tree is out there. Triangulation manually? Nope, it can't be done with any degree of reliability. It must be computerized.