¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. R1b-U106
  2. Topics

Haplogroup creation date uncertainty

 

My haplogroup is R-BY71870. My closest matches, MF and PHVH, are in daughter haplogroup R-BY81828. FTDNA estimates that R-BY71870 was formed in 1650, which is off in my opinion by at least 30 years. (The MRCA of PHVH and myself was first recorded as receiving supplies in 1638 in Rensselaerswyck, New Netherlands.)?

R-BY81828 is estimated to have been formed in 1750.

MF was an adoptee whose birth father has been discovered from his original birth certificate and verified to be within a generation of said birth father by DNA matches. I traced the ancestry of that birth father, and the paper trail peters out with a Henry Van Houten, aged 77 (so born ca. 1773) in the 1850 census of Salem, Washtenaw County, Michigan, USA, born in New York state, living with presumed wife Susan, aged 76, born in New Jersey. I believe but can not be certain that they are the Henry Van Houten and Susan Eckerson who married in the Kakiat Reformed Church in what is now New Hempstead, Rockland County, New York in 1799.

The most recent ancestor of PHVH who could have been Henry's father is Roelof Van Houten, born 1741. He and his first wife baptized 5 children in the New York (City) Reformed Church between 1766 and 1775, with a 2 1/2 year gap between children 4 (baptized 1772) & 5 (baptized 1775). None of his recorded sons are named Henry.

Roelof's father was Cornelis Van Houten, born 1717. He is recorded as having two other sons that lived to adulthood and had children. Of those, one had a son Hendrick, but he was born in 1783, rather too late to be the Henry who married Susan Eckerson, or be the Henry in the 1850 census. The other brother of Roelof is recorded as having had one child by his first wife in 1766, then there is a 19 year gap before he started baptizing children with his second wife in 1785.

How much uncertainty should I assume is in the 1750 estimated date of formation of R-BY81828? Roelof's birth in 1741 is likely well within the uncertainty, but would Cornelius' birth in 1717 fall outside of it?
 

There is a LOT of uncertainty in all dates associated with haplogroups. The rough estimate is the broad observation that the average time between SNP mutations is 83 years. Say if your line has accumulated three new mutations since the last shared mutation with another lineage (both descendants of your most recent common ancestor MRCA) then, very roughly, about 250 years has passed since the split. Additional information from additional testers, documented relationships, and associated haplogroups can be worked into the estimate to refine it but it is still VERY rough. For example, I manage a group who have a documented common ancestor born in 1646. Thirteen descendants have tested who represent lineages from four of his sons. Nine generations have passed and in that time from 2 to 7 mutations have occurred in the descendants. If you simply applied the 83 years per mutation to those, you come up with wildly different time estimates since their MRCA. But since we have 13 testers instead of 2, we can get a better estimate. The average was 3.6 mutations which gives us an estimate that is quite close to the real time [1950-3.6x83=1651]. You'll notice that FTDNA gives ranges as well as dates for haplogroup formations. In the case of R-BY81828, the range is 1416 to 1810.
?
Mike Tryon
 

Thank for that. Where do you find the ranges for haplogroup formation? I was looking at Discover, but that didn't give a range.

I can definitely constrain it tighter than 1416 to 1810. Henry Van Houten was born about 1773, and he wasn't the ancestor of PHVH, so formation of the haplogroup is at least one generation earlier. The most recent patrinear ancestor of PHVH that could have been Henry's father assuming my tree is right) was baptized in 1741, so that puts an upper bound on the formation date. PHVH and myself are descended from different brothers of the second generation in America, baptized in the New Amsterdam Reformed Church in 1648 (my ancestor) and 1651 (his), so that puts a lower bound on the formation, but that encompasses four generations... again, if my tree is right.

To constrain it further, I need to find a tester who is descended from Cornelis Roelofs (baptized 1651) who is not a descendant of Roelof Van Houten (baptized 1741.)


On Wed, Jul 31, 2024 at 3:31 PM, Mike Tryon
<michaeldtryon@...> wrote:
There is a LOT of uncertainty in all dates associated with haplogroups. The rough estimate is the broad observation that the average time between SNP mutations is 83 years. Say if your line has accumulated three new mutations since the last shared mutation with another lineage (both descendants of your most recent common ancestor MRCA) then, very roughly, about 250 years has passed since the split. Additional information from additional testers, documented relationships, and associated haplogroups can be worked into the estimate to refine it but it is still VERY rough. For example, I manage a group who have a documented common ancestor born in 1646. Thirteen descendants have tested who represent lineages from four of his sons. Nine generations have passed and in that time from 2 to 7 mutations have occurred in the descendants. If you simply applied the 83 years per mutation to those, you come up with wildly different time estimates since their MRCA. But since we have 13 testers instead of 2, we can get a better estimate. The average was 3.6 mutations which gives us an estimate that is quite close to the real time [1950-3.6x83=1651]. You'll notice that FTDNA gives ranges as well as dates for haplogroup formations. In the case of R-BY81828, the range is 1416 to 1810.
?
Mike Tryon

1 - 3 of 3