"Jose" wrote in message
m...
He may have been just a bad pilot that flight...

Is he a good pilot if he makes lots of mistakes that never result in an
accident?


No he is a lucky Bad Pilot.


To me, the usefulness of the categorization "bad pilot" is predictive.
Prediction is based on a propensity to do something. Statistics of one do
not show a propensity. Although it calls attention to a pilot which may
belong to the class, it does not =put= that pilot in that class.


If you show me a pilot that regulary breaks the rules, ignores safety
concerns and does things that most of us in this forum would catagorize as
"Bad Pilot" tricks. Then that pilot is more likly to have an accident than
someone that most of us would catagorize as a "Good Pilot"

Your constant harping that pretty much can be summed up as "****
happens"...

That's not the point of my harping. The point is that, using statistics
of one to label somebody with a moniker that is presumed to have
predictive value is erronious, and wrong thinking leads to wrong acting.


Entire industries are based on doing just that. I deal with workers'
compensation insurance on a daily basis so I will give you an example from
that arena.

For a given type of work let's say masonry there is a given "manual rate"
for each state. Let's say that rate is $10/$100 of payroll or 10%. If you
have two companies both open for business on the same day. A little while
after the two companies have been in business for over a year and have a
claims history and organization called the NCCI is going to assign to each
company a Experience MOD rate. This number for a company that has performed
equal to the average company in that business will get a 1.0 mod rate a
company that has done worst than the average will get a MOD of say 1.1, a
company that has done better will get a mod of say 0.9. The total premium
the company will pay for the next term is then the manual rate times the MOD
rate.

Let's say are two make believe companies have a history now and company A
had 10 injuries that cost the insurance company a total of $100,000. Company
B only had one injury but it was a big one and cost $100,000. One might
thing that when the MOD rate was calculated for these two companies that it
would be the same. Well guess what? One would be wrong. Company A with a
bunch of injuries would be considerably higher because in comparison they
are a more dangerous place to work and statistics show that there will
sooner or latter be a large accident that costs more than the little
injuries combined plus the little injuries will still be there.

Company B on the other hand doesn't have little injuries and statistics show
that it might be years if ever that they will have another big accident.