Ancestry Matches 7cM and less

ThruLines does that with the "Potential Ancestors" that it creates for gaps that are in your set of ancestors, it is particularly interesting when you have a possible DNA link through one of these "Potential Ancestors".

 

It only shows people who have a DNA match directly with you, it does not show all the people who are descended from that ancestor (and have trees showing that potential) but do not match with you, despite matching with other descendants of the ancestor in question.

To illustrate, I have a 3xGGparent with 24 Thrulines. My fourth cousin has 36 Thrulines to the same ancestor, but 6 of mine are not on her list. Another 4th cousin from another line adds a few more matches. If we still had DNA circles, we could have 45 matches in the circle instead of 24 or 36. I only know this because I have been able to make contact with those two cousins and we have given each other reciprocal access to our DNA results. For this 3xGGparent the extra matches dont matter much as we have him well documented, but trying to confirm matches further back in the tree the extra matches can be more important. Equally so when a distant match to one of our research group might be 30cM but only single digits for another and not at all for the third researcher, who instead matches to a close cousin of the DNA match (completing the circle).

As it stands, Thrulines is of far more value than DNA circles was, but the later had great potential which was not fully realised.

 

Totally agree with Terry - I have now around 20 homemade DNA circles (I had only one when determined by Ancestry) that have all been created by sharing DNA results with 44 cousins. This included the connections from the official DNA circle.
As I have found all (but one) of my 3xg grandparents and most of my 4xg grandparents, most of my new finds are in the range of 4th to 6th cousins or more distant which are normally not found by Throlines (unless I build the link in my own tree),
ThroLines has no real use to me now whilst new Common Ancestors only happen once every few months if I am lucky. In the last three months I have found between 50-100 cousins via reciprocal access though most did not share DNA with me which would be expected for a 5th or 6th cousin (based on ISOGG statistics: 32% 5th cousin and 11% 6th cousin).

The main problem with the original DNA circles within ANCESTRY is that it had to rely on family trees that often contain errors that a human can intepret but a computer cannot.

 

If I am reading the notes on Ancestry correctly, the new updated algorithm has been in place since early August

No of segments discussion from Ancestry: Our updated matching algorithm may reduce the estimated number of segments you share with some of your DNA matches.
Length of longest shared segment discussion from Ancestry: Our updated matching algorithm may reduce the estimated number of segments you share with some of your DNA matches.
DNA matches must be greater than 8cM discussion from Ancestry: Our updated matching algorithm will increase the likelihood you’re actually related to very distant matches. As a result, you’ll no longer see matches or be matched to people who share 7.9 cM or less DNA

All that has changed is that the result of the longest match were added as an output to us earlier this week and in the next few weeks all matches less than 7.99999 (except those in groups, with notes or messaged) will disappear.
There is nothing in their literature to suggest anything NEW

 

I see that Ancestry are sticking to their August schedule of improvements with the aim of giving more accurate results. It puzzles me as to why then I have results with, say 18cM total match, with the longest segment being 20cM long? Strange mathematics. Not sure how confident I can be about these results. I should point out that I have not been looking for these anomalies, just happened on them when checking out potential matches. Am I missing something?

 

I think they have one algorithm working - it would not make sense to have two working (or two databases) as this would take up valuable computer memory for which they are lacking and partly the reason for the change.
Wrt to the 8cM and under, this would be simple to have not live within the algorithm until a date is reached. This is just a simple IF statement.

Coming back to the difference between Common Ancestors and DNACircles - these are out of sync as Common Ancestors will report relationships even from private trees - when they first came out, I found out that a person with 83cM was actually a 2nd cousin once removed and had been adopted (hence why it was private though searchable). I did make contact with her later but Ancestry seem to have bypassed their own privacy rules (the tree wassearchable but does that mean it can publish the relationship in ThroLines) which are always used for why there is no chromosome browser.

 

It looks the unmarked most distant matches have now gone. Mine have, anyway.

 

Mine are gone as well. I went from 35,356 the other day to 12,588 tonight.

 

Mine gone as well from 30556 to 10884 as expected. All those tagged appear to have survived so thanks to all messages telling us to attach to a group, message etc from Peter and others. Not sure I see any other changes so still have discrepancy between shared DNA and longest segment (my largest is probably 49cM v 60cM for my highest unknown cousin). Await Peter's reporting back from Ancestry.

 

I'm down from just over 39,000 to just under 15,000.
Looks like the shorter matches that I saved are still there.

 

I don't know if ThruLines has yet to be updated after the cull. I deliberately removed colour coding and notes from a couple of 6 cM Common Ancestor matches in the hope that a completely ludicrous pair of potential ancestors would be removed, but no such luck so far.

 

I've 'lost' more than half of my matches, having gone from approx. 50.25 K to approx. 23.25 K ... and over 7.5 K of those are shorter matches that I recently saved (no wonder I found August tiring)! So, I think I now have under 16 K in the 9 cM and over range (plus some saved matches between 8 and 8.9 cM - I have no ready means of separating the ones above and below 8 cM). Seems like a huge difference?

 

I have also lost ThruLines since cull - common ancestors still there but no ThruLines - I attached a new tree to my DNA but that was over a week ago
Wass this supposes to happen?
The reason I ask is that muy current subscription runs out this week and awaiting a new offer to "restart" it

 

I haven't lost ThruLines but the earlier information from Ancestry indicated that potential ancestors shown there might change or be lost during the changes. I had a couple of potential ancestors showing in my ThruLines that I was hoping I definitely would lose, but no, they are still there standing out like sore thumbs!

 

I think my matches are boosted a bit by having a marriage between cousins in one part of my tree and two sisters marrying two brothers on another line. I may be over-simplifying, but I think that means there are some matches that seem closer than they otherwise would?

I don't suppose very many of the short matches will be useful, but as I already had several that I could prove the links to, I thought I should hang on to as many with shared names or significant locations as I could. And yes, I am the sort of person that hates throwing things away.