This is Usenet! Just 'cuz the horse is lyin' down and hasn't moved in a
week
don't mean we can't kick it again


You know that's the truth!

There's another rule in statistics that the smallest sample about which
you
can make a statistically sound statement is 30. It's statistically dubious
to draw assumptions from very small samples but there's also a point of
diminishing returns to having ever-larger sample sizes.

Interesting. What is the sample size here really? Is a sample a single
hour, an average flight of some number of hours, an accident, a plane, or
what? It would seem if the best performer runs at .28 failures per 100,000
hours (I think this was the rate for the DA20 at last look) then you would
need about a million to ensure a good number because there is no such thing
as a .28 dead person.



Well, let's be fair and say that there are a significant body of "OH S#*T"
experiences that are more survivable in a Cirrus than anything else. Of
course, from the insurance company's perspective, a total hull loss is a
total hull loss, whether it's due to CAPS or an engine-out night landing
in
the mountains.


I am not yet ready to agree. I await more hours free of fatalities. Sure
the chute is good, but what would the designer have done without the BRS
option. Would the plane be less safe? I would rather have the Lancair I
think, but its more money.