Wednesday, September 28, 2016

Revisiting: Shape of a Complete Family Tree

I was recently thinking about this topic that I wrote about in a couple of my very first posts and began thinking that I could do some interesting calculations on my own 100,000+ database. Since I also have programmed my own genealogy software (Cumberland Family Tree), I wrote a routine to automatically calculate the number of actual and unique ancestors by generation in my own database for any given individual.

There are basically two numbers I wanted to calculate: (1) the actual number of "slots" filled in a standard ancestral pedigree chart which doubles every generation and (2) the actual "unique" ancestors which will always be equal to or less than the number of slots filled. This second happens whenever an intermarriage occurs, ie one of your ancestors happens to be a sibling of another of your ancestors. In such a case their parents and their ancestors get only counted once as a "unique" ancestor, although they may fill two (or even more) "slots" in a standard pedigree chart.

Take a look at Table #1.  There are two groups of columns, one for my father (Leonard Lund) and one for my mother (Frida Rage). The first row "Self" is always only "1", the person at the top, my father in the first group. Notice that 4 rows down my father is missing a known ancestor. That is his grandfather's father, whom we do not know since he was orphaned at age 5 when his mother died and we can locate no evidence whatsoever of who his natural father really was. However, on my mother's side we know 100% of her 32 third-great-grandparents (row 6).

Another interesting thing is that between the 11th and 20th rows (generations) we know only about 30 or so direct ancestors of my father, but 3-400 of my mother's.

But one of the most interesting things is 30 generations back - about 1000 years ago (using 30 years per generation), we know almost 20,000 ancestors on my father's side and over 30,000 on my mother's side. However, let me rephrase that - if we filled out a standard ancestral pedigree chart, 30 generations back my father's would have 20,000 "slots" filled, but my mother would have 30,000 "slots" filled.

But compare those numbers with the more than half-billion "slots" (536,870,912 to be exact), representing the actual number of ancestors a person should have 30 generations back. In both cases the actual slots I have filled in my own genealogy is a small fraction of even 1% of my true ancestry. And as I wrote in one of my first blogs, this is nobility and royalty only. The common people who certainly make up 99% or so of my real ancestry are not known.

But the 20 and 30 thousand values are misleading, since there was so much intermarriage, the actual "unique" ancestors are only 808 and 295 individuals. What is really eye opening is how few this is compared to the 10 and 30 thousand (about 4% and 1% respectively). This demonstrates how much intermarriage was occurring among the known nobility and royalty. My gut feeling - with no proof - is that for common people intermarriage would be even more common, due to less travel among the poorer and common folk. Even in the 1500's through 1800's in much of my own genealogical work in Norway, people tended to marry those from nearby farms who were already related to them often only a few generations back. The smaller the community, the more likelihood of close intermarriages.

Another thing that I find interesting is the difference between my father's and mother's unique ancestors in the 30th generation. On my father's side the only lines I can trace back to nobility are English which occurs within 10-12 generations. On my mother's side it is Norwegian nobility. But notice how the English side has less intermarriages based on the 4.1% unique individuals, while the Norwegian side has high intermarriages at only 0.9% unique - a four fold difference! Is this perhaps due to Norway being a smaller populated country than England was a thousand years ago? Although in both cases we are back a thousand years ago and looking only at nobility and royalty.

Anyway, just some interesting thoughts. By the way, I have not released a new version yet of CFT with this newly added "Ancestor Count" feature. When it comes out it will be part of Version 4.07.

Cumberland Family Software:

If you would like to have new blog posts emailed to you as they are posted, email me a note with your email address and I will add you


  1. Really useful CFT functionality. Very interesting calculations and explanation. Thank you, Ira!

    1. Thanks for the comment. I am glad you like it.

  2. Dear Ira,
    Great addition to CFTW
    Thank you