I have worked on this during the last month, and here goes my current thoughts:
It can either be between 0 to 1 (i.e. related 0.5 or 0.0001) but since most of us is related 6 cousins etc. that would lead to a very small number which would be difficult to understand. (While the same parents = 1, and only one parent = 0.5). So that would not work.
So I propose the index must be a gross number of the sum of common ancestors, the higher, the more related.
I propose one must go back 6 generations for this number not to be too large, and also meaningful. But lets say the number is N generations.
For each level x (working from your parents), a score of 2^(N-x) is assigned for each ancestor that is the same with the profile compared against. (For x levels, i.e. level 2 for me and level 3 for you, it would depend on the highest level, i.e. here 3)
For example, if there were only two levels (N=2), and the ancestors for Person A and B are:
a b c d a g h i
e f j k
A B
The score would be a = equals on level 2, so 2^(2-2) = 1
Duplicates should be added of course, as it shows an even stronger link.
Follow Us
Like Us
