I have been a very strong advocate of "More Markers" for more than a decade. I particularly dislike to see men tested only at 12 or 25 markers and would like to be able to block those poor choices from being offered through projects that I administer. (while I acknowledge that there area few - limited - cases where 12 markers can make sense.)
If you want to see a dramatic situation where it took at least 67 markers to define two men's genetic family - take a look at the Eubanks Surname Project: http://www.worldfamilies.net/surnames/eubanks/results
There, you will see Eubanks Haplogroup R1b - Lineage I. This is a dramatically clear Lineage - clear because there are so few random mutations. The men included in this Lineage (and their paternal line families all clearly share a Common Paternal Ancestor - though he isn't so recent - based on reported EKAs (Earliest Known Ancestors) If the EKAs are approximately correct, I'd guess that the MRCA (most recent common ancestor) was born prior to the mid-1600s - and possibly much earlier.
Now - focus specifically on the point of this blog posting - kits N10706 & 220084 - two men, both reporting "Thomas Ubancke, Westmorland, bapt. 11.11.1552" as their EKA. These two men share 5 mutations that the rest of the men in the Lineage don't have (with two minor exceptions addressed in points 2 & 3 below). This keeps them from being reported as a match to the men in this Lineage. Once they upgraded to 67 markers, they found matches - though at a fairly large GD. (This is why it's so important to consider BOTH the GD AND the actual marker results!)
I wonder how many more matches these two men will have reported by FTDNA as more men test at 67 markers? There are 5 men reported as their 67 marker matches with different surnames who are not in the project. These men should be in the Eubanks project to see how their markers align in the Lineage - it could be a part of the story!
Here's my take on Eubanks R1b - Lineage I, based on shared mutations insights that I just shared with Kit 220084:
1. You and N10706 are clearly closely related - sharing 5 mutations that nearly all of the others don't have (see points 2 & 3 for the exceptions)
2. You and N10706 may be "distant" (some 100s of years) kin with kit 363756 - but closer to him than to the others - based on his 15 at marker 458. This is a probability - not a certainty, as his 15 could have come from a different ancestor (parallel mutations) instead of being passed to him from your ancestor (shared mutation)
3. Similarly, 62298, based on his 36 at CDYa, is in the same situation as 363756 with you and N10706. Probabilities say that either 363756 or 62298 - but almost certainly not both - could have a shared mutation with you. It's also possible that both are parallel mutations
4. Otherwise, I'd guess that all of the others now in Eubanks R1b Lineage I share a recent common ancestor with you and N10706 no later than the mid 1600s to early 1700s (based on reported EKAs and mutation differences)
5. The 4 guys with a 30 at 449 are probably more closely related to each other than to the rest of the Lineage - but it's also possible that one of the bunch has a parallel mutation. Paper trails suggest the 3 with an Albemarle Co EKA have a shared mutation - but no insight is to be had on the adoptee
Fine print - details you may want to skip: When comparing at 37 markers, the maximum difference to be reported as a "Match" is a Genetic Distance (GD) of 4. This can be one different on 4 different markers or some other combination of differences that add up to 4) [Note: don't get totally hung up on this counting - as there are special cases for counting GD at most "Multi-Markers - 385, 464, YCAII, CDY and others and another special case where multiple multi-markers are showing "doubles"]
So - several important learnings from this example:
A. Don't just give up if you don't have any matches at 37 markers (NOTE - I don't even consider 12 and 25 marker comparisons to be "matches" - so I am ignoring them here) Test at a minimum of 37 markers and consider an UPGRADE to at least 67 - particularly when you have no matches at 37!
B. Look at both the FTDNA-reported matches AND the Results table. It's VERY IMPORTANT to consider your Genetic Family (that I call "Lineage") There are often insights in the Lineage that are not available - such as the genetic kin that aren't reported as FTDNA-matches.
C. Look at mutations with the same value. These can indicate more recent shared ancestry between these men than with others in the Lineage.