Shared Matches and Sorting Your DNA List into Family Lines

Thousands of DNA matches become manageable once you sort them by grandparent. How shared matches, the Leeds Method and clustering tools work, and why close relatives sometimes land on the "wrong" side.

3 min read

A typical DNA match list has thousands of people, and most of them are distant cousins you'll never place. The key to using DNA for genealogy is sorting: figuring out which side of your family, and ideally which grandparent's line, each important match belongs to. Shared matches are how you do it.

What "shared matches" means

When you look at one match, the testing company can show you other people who match both you and that person. These shared matches (or "in common with" matches) are probably related to you through the same ancestral line, though not always.

Each company sets its own limits. Ancestry, for example, shows shared matches only when each person shares at least 20 cM with both of you, so two people who really do share a line may not show up as shared matches if one of the connections is small.

Start with known relatives

If a known cousin has tested, your shared matches with them are on that family line. A known first cousin on your father's side sorts out a big chunk of your paternal matches at once. That's why testing older relatives and cousins on different lines is so valuable. See Which DNA Test Should You Take?

If one of your parents has tested, even better: any match who also matches that parent is on that side.

Company tools

• Parental sides: Ancestry and some other companies try to assign matches to Parent 1 and Parent 2 automatically. It's often helpful, but not perfect, and the labels "Parent 1" and "Parent 2" don't say which is your mother or father.
• Groups, colored dots and labels: most companies let you tag matches. Use them to record which line you think each match belongs to.
• Suggested connections: tools like Ancestry's ThruLines and MyHeritage's Theory of Family Relativity suggest how you might be related, based on family trees. They're hints, built from other people's trees, which can be wrong. Verify every suggestion.
• Clustering: tools such as AutoClusters (offered by MyHeritage and independent services) automatically group matches who match each other, showing you clusters that usually represent separate ancestral lines.

The Leeds Method

Developed by genealogist Dana Leeds, the Leeds Method sorts your matches into your four grandparents' lines using a simple spreadsheet:

1. List your matches in the second-to-third-cousin range, roughly 90 to 400 cM. (Skip closer relatives, who may share more than one line.)
2. Take the first match and give them a color. Color everyone on your list who is a shared match with them.
3. Move to the next uncolored match and give them a new color. Color their shared matches.
4. Repeat. You'll usually end up with about four main colors, one for each grandparent.

Then look at the family trees of the matches in each color for common surnames and places to identify which grandparent each color represents.

When close relatives land on the "wrong" side

A common surprise: a match you know is related through your mother shows up with paternal shared matches, or an automatic tool puts a close relative on the wrong side. Possible explanations:

• You're related in more than one way. In small communities, your parents may be distant cousins, so some matches connect to both sides.
• The tool made a mistake. Automatic parent assignments are estimates.
• A family relationship is different than you thought. Occasionally, sorting reveals an unexpected parent or adoption. See Unexpected DNA Results.

Check the shared cM against the expected range, look at the match's tree, and ask whether a double relationship is possible before concluding something is wrong.

Working a cluster

Once you have a cluster of matches who share a line, look for the common ancestor:

1. Build out each match's tree as far back as you can, using records.
2. Look for the same couple, or the same surname and place, appearing in several trees.
3. Check whether the shared cM amounts fit the relationships those trees imply.

This is the core technique for using DNA to solve brick walls. See Using DNA Matches to Break Through Brick Walls.