Archiver > GENEALOGY-DNA > 2012-07 > 1342112805

From: Jim Bartlett <>
Subject: Re: [DNA] IBS vs. IBD - Family Finder Results
Date: Thu, 12 Jul 2012 12:06:45 -0500 (CDT)


The test is designed to find matching segments of atDNA. The algorithm that decides which matching segments to report to you as a probable cousin includes the length and number of matching segments and the number of SNPS in the segments. Each company uses a different algorithm that is based on their experience and is disigned to insure that most of your matches will indeed be cousins in a genealogy timeframe. A lot of judgement factors are involved. No system will be perfect. Some of your real cousins will not make this cut; a few who are not related will be included.

I don't really understand IBS very well. I think it's marketing talk to explain that some segments of atDNA that pass the algorithm don't really come from a Common Ancestor. It happens.

I think the answer to your last question is yes.

The big problem is how much of our ancestry is documented. It's not uncommon to match 6th cousins (despite what FTDNA says). So how many of your 128 5Great-grandparents have you documented? The matching algorithms, computers and atDNA don't care - they report matching segments from Ancestors you know and from those you don't know. So you can assume that the percentage of your Tree that you have documented is the maximum percentage of matches with whom you'll be able to find a Common Ancestor. How much ancestry your match knows, may drop this percentage even lower.

Hope this helps,

Jim Bartlett

On 07/12/12, BARTON LEWIS<> wrote:

I've had both parents take the Family Finder test through FTDNA, and
have the following question. FTDNA's "primer" describes the difference
between IBS and IBD. IBS includes matches which may occur "by
coincidence" (as opposed to sharing a common ancestor), according to the
primer. What is the likelihood that an IBS match is "by coincidence?"
Is the test designed to pick up matches which almost certainly are
through shared ancestry?

Thank you.

Barton Lewis

This thread: