Know Thy System

Have you ever fidgeted with a gadget in a way for which it hadn't been intended—and then became disappointed with the result? Those of us who are prone to improvising sometimes forget that tools are designed for specific purposes, and that using them otherwise does not come with the same guarantee. That's what engineering is all about: design with an end use in mind.

I forgot about that the other day when I received what I thought was an unusual observation about a tool I was using. Thankfully, it wasn't about use of a sharp, pointy object—I'm one of those people who are prone to improvising—but about a virtual tool.

The context of the comment had to do with that research project I've been wrestling with lately: finding the parents of my husband's second great-grandmother, Johanna Falvey. The tool—a virtual one, remember?—was the system at AncestryDNA, where my husband has several tiny matches which are about to become so small that they will be invisible. I was reaching out to all the DNA matches I could find whose tree included a direct-line Falvey ancestor—all of whose shared genetic connections are quite small, I assure you.

This particular DNA connection was a person with Falvey antecedents who, in leaving Ireland, had traveled from County Kerry westward to the New England region of the United States. In all, this DNA match and my husband shared a mere 18 centiMorgans—not much, I admit, but thankfully a smidgeon over that new cutoff point of eight cMs.

As I have been doing with all the other Falvey DNA matches, I sent this person a message asking if he would be interested in comparing notes and working with a team of fellow Falvey matches to untangle this puzzle. His response indicated he was quite open to collaboration, but he had one question.

Upon receiving my message, this researcher had clicked on the tool at Ancestry labeled "Shared Matches." Just as I had noticed in doing this, he received a message that went something like this:

This DNA match found it odd that even though he and my husband share Falvey ancestors to some degree, not a single other person appeared to share this genetic connection with them. He mentioned as much to me in his response.

Of course, I was surprised to learn that as well—until I learned that that wasn't the case at all. We could possibly share a common Falvey connection with others—it's just that it was too small a connection to register, based on the specific system we were using. In other words, that system is not designed, currently, to be useful in that way. If you and your match have other matches in common—but only at a shared amount of centiMorgans less than twenty—they will never be listed in response to your inquiry.

It wasn't until all the commentary on AncestryDNA's latest decision to revise their tools that I realized why these two matches didn't come up with any shared matches. Somewhere deep in all the articles about the latest Ancestry changes, I recalled seeing someone speaking to that issue. Of course, now I can't find the specific mentions of that design limitation, but I did locate a few indicators that that was how the system now handles "Shared Matches."

For one thing, Ancestry has already mentioned that "Shared matches are only available for fourth cousins and closer to you." Since I was starting out with an 18 cM match, it was likely that our connection represented at least a fifth cousin connection. But Roberta Estes, in her response to a comment on a recent post, explained that "Common matches are only shown if they are 20 cM or over."

No wonder we had "no" connections! The tool wasn't designed to provide that information. While of course it would make my research efforts much easier if such information could be provided—at least to the same level as the currently-bemoaned upcoming cut-off point—I still have what I need. Thanks to the data the system does provide, I was able to find a viable match, despite the small genetic size, and make a connection with another researcher sharing my goals to locate a specific Falvey family in County Kerry. While I can't just click a handy label like "Shared Matches," I can search the entire match listing for other DNA customers whose tree includes a Falvey ancestor from County Kerry.

While that lack in the system I'm using does cause me some additional work, it's certainly work that can be accomplished. In the meantime, it reminds me to know the limitations of the specific tools we choose to use. While an insightful customer service response would be to implement changes that are of use to one's subscribers, in the meantime, I can maximize what use I can gain from a tool, and look elsewhere for the services I'd prefer to gain.