Calling all stats people (spinoff of 'Safety of Skydiving' Thread)

[skydivegirl 0]

Not to beat a dead horse but all these recent 'how unsafe is skydiving' threads have got me going in circles. Somebody please explain to me any error in my thinking.

I learned in two different graduate level risk assessment & analysis courses that to evaluate risk, I need to look at deaths per event/exposure... not deaths per participant (since the events/exposures per participant can either be a little or a lot). For this example the event will be 1 skydive.

According to the USPA website:
In 2002:
"In one year, current USPA members reported making a total of 2,151,228 jumps."

In 2002: there were 33 fatalities

So, 33/2,151,228 = X/1,000,000
33,000,000 = 2,151,228X
X= 15.34

So, in 2002, there are about 15 deaths per 1 million skydives made.

(I was going to subtract total jumps/fatalities of tandems but I couldn't find any reported information on that.)

In another thread someone mentioned there being approximately 30,000 licensed skydivers with approximately 30 deaths a year.... the conclusion being 1 in 1,000 skydivers will die per year.

This being the case... it isn't the same as having a 1 in 1,000 chance of death. Right??? 1 per 1,000 skydivers will die... but approximately 15 per 1,000,000 skydives will result in a fatality???????

Someone help me out here

Pink Mafia Sis #26

[nightjumps 1]

1 in 1,000 chance of death is not the same as 1 per 1,000 skydivers will die.

[councilman24 37]

Per exposure is more technically right. The media wants per participant (maybe read "average" participant). But we all know that neither are a true assessment of any one individuals risk. If you don't do high performance landings your risk probably just went down. If you don't maintain your equipment your risk just went up. I usually tell people that by in large skydiving is as safe as you make it and that most fatalities, with notable exceptions, are human error on the skydivers part. Many of the others are human error on someone elses part. The easy comparison is driving. Driving drunk increases your risk. Driving only when it's sunny decreases your risk.

As stated above 1 in a 1000 skydivers dying is not the same as you having 1 in a 1000 chance of dying. Your risk will vary with your actions or inactions.

I'm old for my age.
Terry Urban
D-8631
FAA DPRE

[newsstand 0]

Quote

So, in 2002, there are about 15 deaths per 1 million skydives made.

So doesn't this work out to 1 chance in 66,667 skydives of dieing? No guarantee that you are the one even if you have 66,666 skydives under your belt.

"Truth is tough. It will not break, like a bubble, at a touch; nay, you may kick it about all day like a football, and it will be round and full at evening."
-- Oliver Wendell Holmes

[skydivegirl 0]

Quote

1 in 1,000 chance of death is not the same as 1 per 1,000 skydivers will die.

OK, it was phrased poorly. How about:

1 chance of death per 1,000 skydivers AND
1 chance of death per 66,667 skydives.

Does this sound about right to everybody?
Where's benny & rehmwa when you need them???

edited to add: These are, of course, round-about numbers based on the only info that I could find.

Pink Mafia Sis #26

[quade 4]

Quote

These are, of course, round-about numbers based on the only info that I could find.

Which is more than likely the source of your problem.

The estimated number of skydives made each year is, I believe, way off.

quade -
The World's Most Boring Skydiver

[skydivegirl 0]

Yea, this is probably a complete waste of energy...

Pink Mafia Sis #26

[nightjumps 1]

Quote

No guarantee that you are the one even if you have 66,666 skydives under your belt.

That is correct. The odds on each skydive is 1:66,666. To think that each skydive increases your odds is referred to as, "Gambler's Fallacy."

Same as with the lottery. Those folks who play the same numbers each week in the hopes that "sooner or later" these numbers will hit since they haven't hit before. The odds for those numbers are the exact same as any other number for the week.

So, statistically speaking, your odds of dying on a skydive are the same as going to Vegas, laying down a dollar and winning $66,666.

From my experience in Vegas, Each dollar I lay down winds up with a greater probability that I walk away with less money than I do when I get there.

[nacmacfeegle 0]

"I need to look at deaths per event/exposure... not deaths per participant"

This could be described as Probable Loss of Life and Individual Risk. PLL is normally a global number, which if applied to say jumping in the US would be an average number of deaths per year. Basic individual risk per jump would be the PLL divided by the number of jumps, as you describe.
Pretty straightforward so far.
Things get complex pretty quickly though.
Factors need to be applied to account for visiting jumpers, seasonal variations, and oddball deaths (suicide, murder, aircraft crashes, etc).
Furthermore, if you base your risk meters on jump numbers alone you do not get an accurate picture of the risk, eg a jumper who makes 1 jump per year would, on the face of it, be substantially less 'at risk' than one who made 1000 jumps per year. This would be true if all jumpers were exposed to the same type of risks. The only 'event' that equalizes across this type of analysis is equipment or aircraft failure. This is clearly not the case with jumping (the majority of skydiver deaths are, I believe attributable to human errors), so a currency/skill factor could be applied.
Similarly other factors could be applied to cover things like equipment selection, wing loading, use of AAD, audible and analogue altimeters, camera jumping, jumping at a busy DZ, type of jumps preferred etc.
In order to develop an accrate risk model, we would also need records of all injuries and near misses. This aspect is traditionally very poorly recorded, compared to say aviation or nuclear power generation etc.

Pretty quickly, trying to define accurate quantified risk meters applicable to skydiving becomes quite farcical, therefore the only real way to risk assess this type of activity is to use qualified as opposed to quantified risk assessment.


If you adopt a qualified risk approach, your mitigations against 'events' that might harm you will be based on adopting a reasonably practical approach. This means investing in things like decent reserves, altitude management devices, AADs, and extended continuous education without any sort of cost benefit analysis. It also means having personal rules covering the type of jumps you get involved in, weather conditions, and even DZ selection.

At the end of the day nobody is arguing that skydiving is inherently hazardous. But 'how risky'...now thats another discussion altogether.

--------------------

He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me. Thomas Jefferson

[Fast 0]

wow, I haven't had time to read the site in over a month and the same discussion is going on.


lmao

~D
Where troubles melt like lemon drops Away above the chimney tops That's where you'll find me.
Swooping is taking one last poke at the bear before escaping it's cave - davelepka