Jim,

I definitely agree with the scrawny (or as someone said "undernourished") trees being a big problem, and the value in attempting to look at all your matches.

As far as your original statement that I questioned about their legitimacy, you waited until the end to give the punch line: "I admit that when I say most Matches are legit, I'm not thinking about the thousands of Ancestry Matches at to very low end"

That's what I was saying. In my wife's sample, 70% of her matches are considered "very low confidence" and only 10% are considered "moderate" or higher, so I would say most of her matches are not cousins.

Now, people here have found real matches with a very low confidence rating, and as you said some segments can be pretty resilient through time. I don't think they were intentionally trying to inflate the number of matches. I think they knew there were situations like that and didn't want to miss anything. It's an age old problem - do you cast a wider net knowing you will get more false positives, or do you tighten things knowing that you will miss some actual matches? Plus, there is no need to inflate - we all have way more than humanly possible to go through without tools like Jeff's. (the exception being JBarry!)

David Mc