(Finbar) wrote in message . com...
According to Van's Aircraft,

3,862 RV aircraft have been completed and flown. By far the most
popular model has been the RV-6 model. Here's a table, by model,
showing number ever completed and flown, with accident statistics from
NTSB since 1/1/99 (the search looked for RV in the model designator
and excluded "incidents"):

Fatal accident
Model Flown Accidents Fatal Killed rate per year

RV-6/6A 1,850 68 13 21 0.15%
RV-4 1,127 31 7 11 0.14%
RV-8/8A 441 12 4 5 0.20%
RV-3 217 5 0 0 0
RV-7 114 2 1 1 0.20%
RV-9/9A 112 5 1 2 0.20%
RV-10 2 0 0 0 n/a

The RV-7, 9 and 10 are recent models.

By comparison, about 36,000 Piper Cherokee (PA-28) variants were
built. In the 1-year period up to 9/1/2003 there were 14 fatal
accidents in this fleet with 26 fatalities. This translates to a
fatal accident rate per year of just 0.04%. However, this large fleet
includes many airplanes built decades ago. It seems reasonable to
divide the "ever built" size by 2 to reflect the relatively more
recent vintage of the RVs. This still makes the Cherokee rate about
0.08%, significantly below the RV rate.

Of course, this doesn't provide the required comparison, which is to
sailplanes. Anybody got a reasonable estimate for the number of
sailplanes in existence, say, completed and flown since 1970?

I have been involved in a reliability analysis at work recently (of
our super computer). There are about 2000 processors in the thing. The
numbers follow classical reliability theory fairly closely: on average
one of the processors is going fail about every 3 days. The mean time
to failure of an individual processor is 172 MONTHS. So, if I am a
processor I might be the one that fails (i.e. dies) within 3 days, or
I could last well past the 172 months. Should I be writing my will or
relaxing with a beer?

The issue becomes more complicated if you consider Bayesian
statistics. This would apply to situations such as: "What is the
probability of dieing if you fly drunk?". For "drunk" you might
substitute:
1. High performance glider
2. Low performance glider
3. Contests
4. First flight of the season
5. Fly less/more than 20 hours/year
6. Have less/more than 100 total flight hours
7. Fly in the mountains/ridges/prairies
8. Fly in wave/thermals/ridge

The moral is that statistics are meaningful for actuaries and
politicians, but aren't very useful for us pilots. OK, flying drunk is
a BAD idea! Don't do it! But what about the rest? While you can't
avoid the first flight of the season, some of the others are in your
control. Do you quit contests because of the higher risks? This must
be a personal decision.

The fatalities seem to be predominated by preventable mistakes, i.e.
stall/spin low to the ground. Don't do this, either!

Tom