OK, here's a synopsis of results so far. What these show is the probability that Test 6.1 is not part of the analytics and can be ignored.
These are the numbers exactly as reported, so I think there are a couple of errors which I've commented on. I also have a feeling that some of these data suggests there are some worn floats out there.
Can everyone interested in this thread send us some numbers? We need a lot more to get a clear picture of how it works.
Cheers,
TC