> What I was looking for from Jeff, that he has so far refused to provide,
> is an approximate accident/100k vehicle mile figure for the general
> public.
Why spend effort researching the wrong metric?
> Using raw time
> would imply that a small region that has a Solo II incident every ten
> years is statistically as safe as a much larger region - i.e., one with
> more autox entries and more vehicle miles/year - that also has an incident
> every ten years. It should be apparent to anyone that that's VERY faulty
> logic.
Yes, that's not a good metric, but that wasn't what I was asking for.
I said miles is wrong, time is better. I didn't say the "per vehicle" part
was wrong.
And just so another obvious thing doesn't get misinterpretted, by time I
don't mean just the time of the actual run - that's not the way the site
owner or insurer views it. They insure events and taking into account
some expected number of participants (there is some implied factor
for the length of a run, but it won't be used to determine whether your
insurance is cancelled). Similarly for regular car insurance - you are
insured for a period of time, not a number of miles (though of course,
there is some factor there too - but again you don't get cancelled for
having so many accidents per mile, its a time factor).
Perhaps there are some insurance companies that explicitly look at
number of vehicle miles run in an autocross and base the rates (and
cancellation decisions) primarily on that statistic, but they would be an
exception. Its just not practical/cost effective for insurance companies
to be measuring/verifying those kinds of things. Same for regular car
insurance.
Time is the key.
|