How to Keep AncestryDNA's Small-Segment Matches

From late August 2020, AncestryDNA will delete matches who share less than 8 cM with you - unless you have added a note about them, added them to a custom group, or messaged them. Some of my known relatives only share 6 cM or 7 cM with me, and I'm sure there are others whom I haven't identified yet. Here is my strategy to prevent those people from disappearing.

#1. In my AncestryDNA account, I created a custom group called '6-8cM Keep' (but you could skip this step and just add a Note to matches you want to keep).

#2. In the 'Shared DNA' filter, I select 'Custom centimorgan range' and set it to min=6 and max=8; then I do a series of searches, and on the results screen I either add a Note or assign people to my group '6-8cM Keep'. Examples of searches (with 'Custom centimorgan range' set to min=6 & max=8):

#3. Filter for 'Common ancestor'. (Some of the 'potential ancestors' will be wrong, because Ancestry's Thrulines derives those ancestor predictions from other people's trees, many of which are wrong - but I need to prevent those matches from disappearing so that I can later check the Thrulines theory.)



#4. Using 'Surnames in matches' trees', I do a series of searches for end-of-line ancestors' surnames, 'brick wall' surnames, unusual surnames, etc.



#5. Using 'Birth location in matches' trees', I do a series of searches for the main places where my families lived. I've also done some combined surname + birthplace searches, eg: "Surname in matches' trees" = CAMPBELL, combined with "Birth location" = Argyll, Scotland. (CAMPBELL is so common that I don't want to keep all the matches, but some small-segment matches with CAMPBELL ancestors in Argyllshire will be my relatives.)

How close is an 8 cM match? In some cases it will be closer than a 4th cousin. To see the full list of possible relationships, with probabilities for each, enter '8' in the interactive Shared Centimorgan Project tool.

* Roberta ESTES explains a more thorough strategy for preserving small-segment matches that may be useful.

* Blaine BETTINGER explains why many small segments are *not* valid shared DNA.

* My personal thoughts on using DNA as a tool for family history.

(This post first appeared on https://genie-leftovers.blogspot.com/2020/07/how-to-keep-ancestrydnas-small-segment.html.)