Doug, A few years ago I caught Sorenson with 20% errors in the ten samples I sent them, so I hope that testing HAS improved in the interim. FTDNA has a good reputation, so I think it's below one percent there. I, too have higher confidence in the testing results, but you still have to consider mutation errors, even though they are buried in the correlation. That's the issue of over-interpretation I have been writing about. - Bye from Bill On Jul 23, 2011, at 5:49 PM, tuulen wrote: > Bill, > > Lab error could occur, but apparently today's testing standards are quite > high and so the rate of lab error is likely quite low. I have higher > confidence in the test results than in the DNA itself, and a comparison > between two people's DNA could today be based on very accurate test results. > > > Your theory sounds convincingly good, and I look forward to learning more > about it, please. > > Doug >