Unnamed fathers .outside the box thinking

[brigidmac]

No they don't need to have fathered another child .

My grandmothers birth  father was identified through DNA matches to  his sisters descendants .

It helped that there were ethnicity differences and location clues as well as. An affiliation order Document to pay the birth mother .

[coombs]

My 3xgreat gran Clara Auber was born in 1835 and her documented mother was 51 at the time, she was born 1784. The previous child was born in 1827, and Clara had a few older sisters of childbearing age, and Clara's middle name was that of an older sister Emma Auber born 1813. Emma herself went on to have more illegitimate children 1840-1855 and fudged details on their birth certificates. Also Clara and Emma and her child born 1840 are untraceable in the 1841 census, and definitely not at the rest of the main family residence in Shoreditch, but it seems Clara later went to live with another sister.

Not impossible to have a baby at 51 but I am smelling something fishy, that Clara was actually the granddaughter of the documented parents.

[Ayashi]

No they don't need to have fathered another child .

Agreed, I've found the right family and considered that to be the closest I'm likely to get.

Technically if the bio-father had more children one should in theory have higher cM matches to them rather than the bio-uncles (if any), but DNA is so finicky that I've had cousins of the same relation to a target individual have vastly different cM amounts to each other. The closer to present day and the older gen the people matching, the more reliable, but in my cases as long as I have the father's parents the tree still works. Narrowing it down that far does often require a lot of matches though, to join parts of the tree or eliminate others, unless the father is suspected in the first place.