# How many generations back is suggested by 6 mutations on a yDNA-67 match?

I have a yDNA-111 test that finds a match with a yDNA-67 test with 6 mutations and a name variation, `Thorp` vs `Tharp`. Statistically, how many generations away would our common ancestor be to give a mean of 6 mutations? How many generations plus or minus would you need to add to the mean to include one standard deviation? It is suggested that the common ancestor might be 9 generations away, so I'm wondering if that is realistic? I would imagine that mutations are rarer and thus 6 mutations out of 67 markers would take more than 9 generations to get an average of 3 down our line and another 3 down their line from the common ancestor. Note that the haplogroup is predicted to be: R-M269 and the Ysearch ID is `SC8P7`.

## 1 Answer

If you use Moses Walker's MRCA Calculator and plug in their suggested mutation rate of .0043, you can get an approximate answer to your question. The tool only goes down to 62 of 67 matches, but it is reasonable to extrapolate an extra column to the right of the table to represent 61 of 67.

Doing so, it looks like you need 12 generations to have a 50% chance of a match (i.e. have a mean of 6 mutations).

For 105 of 111 markers, you need just 6 generations to have a 50% chance of a match. (6 mutations).

So mixing the two tests, depending on which side the mutations occur, there could be between 6 and 12 generations to the MRCA and that averages to about 9 generations.

One standard deviation on each side of the mean would be at the 16% and 84% levels. In an extrapolated column for 61 of 67, these would be about 8 generations and 15 generations. For 105 of 111 it would be 3 generations and 10 generations. Averaging these would give you standard deviations that range from 5.5 to 12.5 generations.

• A problem with this estimation is that if 111 (an additional 44) markers are tested, it is assuming no further marker mismatches are discovered among those 44 markers. Without further knowledge, a better estimate would be that 4 more mismatches would be detected, so the rest of the estimate would probably be better based on an estimated result of 101 of 111 markers instead of 105 of 111. This would increase the estimated generations. – mgkrebbs Dec 14 '19 at 19:14