GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2011-12 > 1322863398
From: Jim Bartlett <>
Subject: Re: [DNA] No matching segments
Date: Fri, 02 Dec 2011 16:03:18 -0600 (CST)
To piggy back on your info... you have a much better chance of having (but not necessarily confirming) a match, the more distant the cousin is. Here's why:
I have perhaps 20 first cousins - if any of them took the atDNA test, we'd sure match - none of my first cousins has tested.
I have perhaps 170 second cousins - if any of them took the atDNA test, they would match 99% of the time; alternatively the pool of potential matches would be 168 of them - none of my second cousins has tested.
I have perhaps 1,500 third cousins - if any of them took the atDNA test, they would match 90% of the time; alternatively the pool of potential matches would be 1,350 of them - one of my third cousins (that I know of) has tested - with 5 matching segments over 8cM.
I have about 15,000 fourth cousins - if any of them took the atDNA test, they would match 50% of the time; alternatively the pool of potential matches would be 7,500 of them - two of my fourth cousins and I have found a paper trail to a Common Ancestor (no guarantee that our one large shared segment each comes from the Common Ancestor we identified)
Similarly: 180,000 fifth cousins - 10% would match - or 18,000 in the pool - I have one 5th cousin match (with confirmed paper trails)
2,000,000 sixth cousins - 2% would match - or 40,000 in the pool - I have paper-trail confirmed three 6th cousin matches.
I have seven 7th cousin matches; seven 8th cousin matches; four 9th cousin matches; and then it trickles off, even though the number of probable cousins continues to rise - and the number of those who would be in the matching pool remains very high, although it probably won't increase forever .
I think the reason confirmed matches with paper trails trickle off is because I only know 236 of my potential 2048 ancestors (roughly 10%) that would lead to 10th cousins. I'm sure many of my atDNA matches reported by FTDNA and 23&Me are really my cousins, I just haven't been able to flesh out my Tree enough to identify them (and I suspect most of those matches have not either)
The odds and the pools above are the same if you rely on random cousins testing or if you target and/or fund many of your close relatives to test (however in this latter case you will get many more hits)
On 12/02/11, Paul Wright<> wrote:
I believe both 23andMe and FTDNA are claiming 99%+ chance of overlapping
segment share with 2nd cousin or closer, 90%+ for 3rd cousin. There have
also been studies shown that the 23andMe shared segment and total share
figures can predict actual relationship within 1 degree of relatedness out
to 2nd cousins.
The likelihood of a full sibling pair inheriting all of the exactly
opposite halves of DNA from each parent would seemingly be so microscopic
as to not warrant any reasonable consideration. While a statement like
you've made can be said about a single chromosome, the odds for all 22
autosomal chromosomes and the X would be almost infinitely small. Probably
better luck winning Poweball.
On Fri, Dec 2, 2011 at 6:59 AM, John Lerch <> wrote:
> Mary Alice,
> You totally miss the point. There is a chance (albeit a very small
> one) that you have ZERO matching segments with a TRUE BIOLOGICAL