A new match (we’ll call him LHO) showed up on my DNA results recently. I suspected his connection based upon our shared matches and his username. The conversation we had confirmed how he and I are related. This post isn’t about that. When I reviewed the shared matches he and I had I was reminded of a few pitfalls of relying only on shared matches when doing analysis.

LHO and I are third cousins. Our great-grandfathers (Lewis Habben and Mimke Habben (1881-1969) were brothers). AncestryDNA indicated that we shared 28 cM of DNA across two segments. Their predicted relationship for us was in the 4th to 6th cousin range. That’s a reasonable prediction based upon the amount of DNA that we share. I won’t share the same amount of DNA with every third cousin I have and there are ranges of typical shared DNA for certain relationships that have been established (Blaine Bettinger’s shared cM project has ranges based upon submitted samples is probably the best, but is based upon users who knew their relationship and submitted the shared cM with that individual. AncestryDNA does have a page with predicted relationship information.) The ranges are broad and relationships at the range discussed in this post are far enough that there is variation. There’s nothing unusual with a predicted 4th-6th cousin actually being a 3rd cousin.

LHO and I share a common set of great-great-grandparents, Jann and Anke (Fecht) Habben, who married in Hancock County, Illinois, in 1881. There at least six of their descendants other than myself who have tested at AncestryDNA:

• MAT with whom I share 443 cM of DNA–she is a granddaughter of Mimke Habben (Jann’s son). I am Mimke’s great-grandson.
• PTG with whom I share 372 cM of DNA–she is a another granddaughter of Mimke Habben.
• PTGD with whom I share 163 cM of DNA–she is PTG’s daughter and a great-grandchild of Mimke.
• GHRD with whom I share 119 cM of DNA–she is a great-granddaughter of Mimke Habben.
• MHSG with whom I share 33 cM of DNA–she is a great-great-granddaughter of Jann Habben.
• LHO with whom I share 28 cM of DNA–he is a great-great-grandson of Jann Habben.

Looking at the histogram of 3rd cousin submissions to Bettinger’s “Shared Centimorgan Project,” (page 13) LHO and I are well within the typical amount of shared DNA for our genealogical relationship. We are lower than the average and that’s reasonable because the average, after all, is the average. Our shared amount of cM is typical for a more distant relationship which is why that’s what AncestryDNA predicted for us. But you can’t just rely on the predictions and they are reasonably correct at this genetic distance–not precise.

Note: of these six descendants of Jann and Anke (Fecht) Habben only two were identified with ThruLines being the initial tool to potentially determine the match. The rest were obtained after initially studying shared matches with MAT who was immediately known to me.