No announcement yet.

Half-cousins and calculating years to most common ancestor with FF

  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by mkdexter View Post
    This goes back to the difference between sum and blocks in how FF makes predictions.

    1st cousins with both grandparents in common will share about 700cM in sum. It will be a sum of grandfather's and grandmother's shared DNA between the two cousins. If there is only one grandparent there is less DNA shared, 300cM instead of 700cM for example (I am using FF numbers here not 23andMe which is different sums). FF will see the 300cM and predict the relation to be 2nd cousins even though they would be half first cousins.

    Now that is not the same if FF sees a longest block of 12cM and a sum of 25cM. FF will predict that small amount of DNA to be a 4th cousin based on the longest block shared by two people that originated from one ggggrandparent.

    Sums of cM DNA include paternal and maternal blocks combined (hence the need to fudge the sums for half relationships) and longest blocks just include one parent in most cases.

    If the relation is 1st or 2nd you do need to look at the sums expected but not the case for most of the 3rd cousins and beyond since at that level FF does not use sums to make a prediction as much as it uses the longest block.
    This also makes sense: so a half cousin has the status of 2nd cousin (since calculation is based on total amount of dna), but a 4th half cousin would show up the same as a 4th cousin in the FF calculator, since you are looking only for lengths of shared material, and whether you share 4 grandparents (doh!) or 1, after 4 generations, the dna from each of them is broken up, and you are looking for sonsecutive lengths in common, not the total amount of dna.