Someone help me understand these new origin results

[SEdnaquest]

FTDNA recently updated their "MyOrigins" Results, and the results confuse me. I have 3 family members who've tested, a father, mother, and child. The results don't seem to add up. Before anyone suggests it, the DNA results do confirm the paternal relationship.
Below are the new results. In parenthesis is their ethnicity.

The Father (Sicilian)
Southeast Europe-85%
British Isles-8%
Ashkenazi-3%
East Middle East-<2%
Southeast Asia-<2%
Oceania-<2%
East Europe-<2%

The Mother (1/2 Hungarian, 1/2 Slavic)
East Europe-61%
Southeast Europe-37%
Asia Minor-<2%

The Daughter
Southeast Europe-42%
East Europe-41%
Asia Minor-16%
Ashkenazi-<2%
Siberia-<2%

This doesn't seem to make sense to me. Where did the Daughter's 16% Asia Minor come in, her parent's middle east DNA is only trace results? And how come neither of her parents have the Siberia DNA result? I suppose the daughter didn't inherit the father's 8% British Isles by pure chance alone. And just how does a Sicilian get Oceania DNA?
Also, is there a historical reason why Hungarian or Slavic has such a huge chunk of Southeast Europe in the DNA?

[Swennilsson]

There are probably lots of people who know this stuff better than I do, but I will take a crack at explaining it.
As I understand it, the company has collected samples of the DNA of various populations around the world. These are modern populations. They use these sample to get a general idea of the common DNA found in a particular population, then when one of us tests, they compare our sample to the population sample of the particular groups. This works pretty good for people who are relatively "pure", if there is such a thing. For example, A friend of mine has all Swedish ancestors as far back as he can tell. His profile fits the Scandinavian sample very well, and it shows him to be 99% Scandinavian.
However, the process works less well for those of us who are mongrels. I, for example, have a Scandinavian father and a German mother. The system sometimes shows me as British. I think this is because modern Brits, from whom the samples come from, are mixtures to some degree of Germans who invaded in the 5th century, and Vikings who came several hundred years later. I too am a Scandinavian-German mix, but my mix happened 58 years ago rather than 1000 years ago, but the testing can't tell WHEN the mix happened.
So, in your case, the daughter's profile may look different than either of her parents because the genetic mix might make her resemble a different population. Bear in mind that many of the European populations are very similar, so it doesn't take much to nudge one into another group. The system works much better if there is greater difference in the populations. For example, it is rather easy to see the difference between Native Americans and Germans, or Japanese and Norwegians.
As time goes on and samples get bigger, there will likely be greater accuracy. At present people should not take the ethnic estimates too literally, especially if the percentages are small.

[SEdnaquest]

The father's old results were:
Southern Europe-69%
Asia Minor-21%
North Africa-2%
Eastern Middle East-1%
Ashkenazi Diaspora-7%

I rationalized the Asia Minor (around Turkey) to be because he was Sicilian, and Sicily was invaded by the moors way back. But the new results show only a small percentage from the middle East, none from Asia Minor. Now how do I rationalize this.

[Biblioteque]

As Swennilsson says, "testing cannot tell WHEN the mix happened; and people should not take the ethnic estimates too literally."

Agonizing and obsessing over MyOrigins is a waste of time because as new reference populations, algorythms, chips, and testing tools are introduced, the results will change.

BTW, how do they factor in the changing of borders over thousands of years? Who are we really? Perhaps it depends on when in time. Ethnicities sometimes became a blur and a blend. Migrating from one country to another, assimilating and becoming them, assuming their language, and then in a few hundred or thousand years forgetting where they originated, and who they were at an earlier time.

Big border changes thoughout history. Poland was extinguished from the globe for hundreds of years. What was their ethnicity during that time.

They say French Normans were just "recycled Vikings."

Guess we are all "recycled" something.......

[georgian1950]

There are probably lots of people who know this stuff better than I do, but I will take a crack at explaining it.
As I understand it, the company has collected samples of the DNA of various populations around the world. These are modern populations. They use these sample to get a general idea of the common DNA found in a particular population, then when one of us tests, they compare our sample to the population sample of the particular groups. This works pretty good for people who are relatively "pure", if there is such a thing....

Thanks for one of the more perceptive posts on ethnicity estimation. The problem is there are very few "pure" groups. My research is showing a high level of common ancestry around the world within the past 300 years. That is why we have so many matches but only have a clue about a handful of them. Pretty much everyone is on the wrong track now with admixture analysis, not just Family Finder.

Jack Wyatt