Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Need Accident Stats!

0 views
Skip to first unread message

jame...@news.delphi.com

unread,
Oct 5, 1993, 9:12:28 PM10/5/93
to

My daughter needs accident statistics for skydiving for a project she is
doing at school. She's an advertising major and is creating some sort of
ad campaign PR type of thing and needs to compare accident rates for
skydiving with common activities like driving a car.

Can anyone supply accident statistics?

TTFN--The Big Cheese

Patrick W. P. Dirks

unread,
Oct 6, 1993, 6:13:30 PM10/6/93
to
In article <28t61s$2...@news.delphi.com>, jame...@news.delphi.com
(JAMES...@DELPHI.COM) wrote:

Well, I recall a posting a while back which had statistics for a whole list
of sports, including skydiving. I only have some statistics I got few
years ago from the USPA for 1990:

115,500 jumpers [Don't know if this is 'active' or how this was
derived]
2.4 million jumps [This must be an estimate]
23 fatalities

Sorry, no statistics on the number of injuries or their severity (although
I imagine if you didn't fall in the USPA statistics category you probably
just sprained an ankle or broke a bone - that may have changed over the
recent years with the number of people hook turning their canopies for
landing).

Assuming both car accidents and skydiving deaths are completely randomly
distributed over the population [a BIG assumption], given the average
passenger automobile death rates in the 1987-1989 period (1.18 car deaths
per 100,000,000 passenger miles), the risk of dying on a given skydive are
about the same as the odds of dying over a 800 mile road trip: about 1 in
100,000.

Hope this helps,
-Pat Dirks.

Hector Mandel

unread,
Oct 6, 1993, 9:34:02 PM10/6/93
to
Patrick W. P. Dirks (p...@apple.com) wrote:

: Assuming both car accidents and skydiving deaths are completely randomly


: distributed over the population [a BIG assumption], given the average
: passenger automobile death rates in the 1987-1989 period (1.18 car deaths
: per 100,000,000 passenger miles), the risk of dying on a given skydive are
: about the same as the odds of dying over a 800 mile road trip: about 1 in
: 100,000.


Ugh ... here we go again :-) You say it is a "BIG" assumption. I say it is
a WRONG assumption. The more I think about these silly comparisons, the more I
am convinced that they are useless. We see these statistics repeatedly
misused, even when the very people who gather them are careful to warn against
misuse. There is no comparison between automobile and skydiving accident
rates. They are fundamentally different activities, with entirely
separate accident rates and causes. Car accidents and skydiving accidents
are, for the most part, *NOT RANDOM EVENTS*. They almost always have
preventable causes. Instead of wasting our time and credibility comparing
apples with oranges, we should concentrate on minimizing the risks in both
activities.

No matter how much some people rationalize or deny, the fact is that skydiving
is an inherently dangerous activity. People who have a frivolous attitude
towards it are kidding themselves.

===============================================================================
Hector Mandel, CHP (217) 244-7237 |E-mail: h-ma...@uiuc.edu
Head, Health Physics Section |Snail mail: 101 S. Gregory St.
Division of Environmental Health and Safety | Urbana, IL 61801
===============================================================================

Patrick W. P. Dirks

unread,
Oct 7, 1993, 3:32:17 PM10/7/93
to
In article <28vrma$5...@vixen.cso.uiuc.edu>, man...@romulus.ehs.uiuc.edu
(Hector Mandel) wrote:

> Patrick W. P. Dirks (p...@apple.com) wrote:
>
> : Assuming both car accidents and skydiving deaths are completely randomly
> : distributed over the population [a BIG assumption], given the average
> : passenger automobile death rates in the 1987-1989 period (1.18 car deaths
> : per 100,000,000 passenger miles), the risk of dying on a given skydive are
> : about the same as the odds of dying over a 800 mile road trip: about 1 in
> : 100,000.
>
>
> Ugh ... here we go again :-) You say it is a "BIG" assumption. I say it is
> a WRONG assumption. The more I think about these silly comparisons, the more I
> am convinced that they are useless. We see these statistics repeatedly
> misused, even when the very people who gather them are careful to warn against
> misuse. There is no comparison between automobile and skydiving accident
> rates. They are fundamentally different activities, with entirely
> separate accident rates and causes. Car accidents and skydiving accidents
> are, for the most part, *NOT RANDOM EVENTS*. They almost always have
> preventable causes. Instead of wasting our time and credibility comparing
> apples with oranges, we should concentrate on minimizing the risks in both
> activities.
>
> No matter how much some people rationalize or deny, the fact is that skydiving
> is an inherently dangerous activity. People who have a frivolous attitude
> towards it are kidding themselves.
>

I don't disagree: working from overall fatality statistics is a GROSS
generalization and is probably a poor assessment of your individual risk;
you can surely affect the odds greatly by your own behavior. However, in
both cases, there is a certain amount of uncontrollable risk: even though I
wear seat belts, I can be hit by a drunk driver, for instance, running a
red light as I'm crossing an intersection. Even though I have an AAD on my
rig, I could get hit by someone under canopy who happened to be looking the
wrong way at the wrong time. Just as there are people driving recklessly
or driving cars with bad brakes, inviting disasaster, there are people who
jump poorly maintained or unfamiliar equipment.

I agree people should do all they can to be careful and avoid what risks
they can. Nevertheless, despite the combined caution of most skydivers
around, some people DO get killed. I just thought SOME perspective on the
overall risks would be better than none at all; I think people mistakenly
picture a much higher risk level in the absense of any data.

Let's be careful out there,
-Pat Dirks.

0 new messages