Skydiving is safe – but not as safe as you might think

After repeatedly hearing the myth that the most dangerous part of a skydive is the ride in your car to the dropzone and my inner bullshit detector constantly being alerted as soon as I heard this or anything similar, I tried to find some statistical proof for or against that statement.

On my research trip I discovered the concept of measuring the risk of a certain event or activity in the unit micromorts (https://en.wikipedia.org/wiki/Micromort). To summarize – if an event has a risk of one micromort, your chance of dying while participating in the event is exactly one to one million. From the Wikipedia page about micromorts I gathered the following values:

• Travelling 370km (230mi) by car: 1 micromort
• Travelling 9,7km (6mi) by motorcycle: 1 micromort
• Doing a skydive in the US: 9 micromorts
• Doing a skydive in the UK: 8 micromorts
• Doing a skydive in Germany: 18,55 micromorts (Source: http://www.dfv.aero/downloads/Jahresvergleich.pdf - why this number is so much higher in Germany is something that I too don’t really understand)

So to compare travelling and skydiving, depending on where you live and jump, you would have to travel about

• 2960 – 6863,5km (1840 – 4265mi) by car
• 77,6 – 179,5km (48 – 111mi) by motorcycle

to match the risk of dying in an accident during one single skydive to dying in an accident while travelling to the dropzone. Even if these are average numbers that might differ from person to person and country to country the trend should be clear. Skydiving is so much more dangerous than your trip to the dropzone. I think this myth should be considered as busted.

Going deeper into the whole statistical thing I tried to find ways to “measure” which impact skydiving actually has on your probability of dying. The following data is based on numbers from Germany, as I live there and therefore had the least difficulty of finding reliable numbers.

In the end of the year 2010 there have been a total of 81.752.000 people living in Germany (1). A total of 852.328 people have died during the following year 2011 (2). The year 2011 has had 365 days which means that a total of 29.839,48 million days have been lived in that year (number of people multiplied with 365). The chance of dying by “living one day in Germany” therefore equals a risk of 28,56 micromorts (number of deaths divided by number of totally lived days multiplied by one million). I now do about 500 jumps per year which gives me an additional risk of dying in one year of skydiving equal to 9275 micromorts (number of jumps per year multiplied with micromorts per jump) – that again increases my risk of dying in one year by a factor of 1,887 (risk of dying in one year with jumping divided by risk of dying in one year without jumping). The risk that I die during the next twelve months is about 88,7% higher than the risk of the average German person dying in the same time. This might sound like not too much of a risk, especially at my jump numbers per year – but if you look into the total number of deaths, you will discover that only 32.988 of the 852.328 deaths are caused by unnatural causes – accidents, crimes, suicides, etc. (2). Most of the deaths in this statistics are older people dying for natural causes and health problems. So, let’s calculate again – the risk of dying in one day by an unnatural cause is about 1,106 micromorts (number of unnatural deaths per year divided by number of totally lived days multiplied by one million). Adding the additional risk of my 500 jumps per year, my total risk of dying in one year has increased by a factor of 23,98 over the risk of someone that is not jumping (risk of dying in one year with jumping divided by risk of dying in one year without jumping). I am almost 24 times more likely to die by accident than the average German person. This number seemed extremely high – almost unrealistic, but numbers don’t lie.

What bothered me was that I did not consider the chance of dying by a natural cause (cancer, infection, etc. – I simply did not know how high the probability for something like this is in my age, so I did not know how far my thoughts and numbers were off).

Let’s step back from taking only unnatural deaths and consider the total amount of deaths again, including natural causes. To get rid of the completely normal deaths of older people, I decided to look at deaths per year per age group. I myself am 23 years old and am therefore part of the 20-29 year olds, which comprises 9.947.000 people in Germany, living a total of 3.360,655 million days per year (number of people multiplied with 365). In the year 2011 a total of 2857 people from this age group has died (3) – all causes of death combined. So if you are a 20-29 year old and live for one day in Germany you are exposed to a total risk of dying equal to 0,85 micromorts (number of deaths divided by number of totally lived days multiplied by one million). Comparing myself to a non-jumper of the same age group I have a 30,9 times higher probability of dying – not considering the exact reason. While I originally expected the number to go down in comparison to the previous calculation, the opposite happened. Of course I know, that not everbody is doing 500 jumps per year and not everybody is from the same age group as I am, so following are few calculations based on an estimation of 100 jumps per year in Germany:

• 20-29 year olds: 7,0 times higher probability of dying within one year
• 30-39 year olds: 5,4 times higher probability of dying within one year
• 40-49 year olds: 2,7 times higher probability of dying within one year
• 50-59 year olds: 1,6 times higher probability of dying within one year

Of course as you could already see in the comparison right at the beginning, skydiving in Germany seems to be much more dangerous than skydiving in the US or the UK (I still don’t know why) and I am pretty sure that “basic” living in Germany is less dangerous than in the US or UK. Alltogether the increase of risk by skydiving might come down a bit more and statistics will always stay statistics. Also, I am fully aware, that the statistical calculations are not 100% correct, but the approximations done only cause falsifications in the 2nd or 3rd digit after the comma. There is no average person and there will never be – so nobody will be able to predict their personal future based on these numbers, but the trend should be clear.

While we consider skydiving to be super safe and completely controllable, it is not. It is still f***ing dangerous (even if not as dangerous as 30 years ago) – and I believe that is something we have banned from our mindsets and tend to forget to tell. Don’t be so casual about the risks you are taking - I know them and I accept them, but I guess not everybody is like that! Be aware of what you are doing – and please, don’t give rookies the impression, that driving their car to the dropzone is more dangerous than the skydives they are doing!

Sources:

(1) http://www.bpb.de/nachschlagen/zahlen-und-fakten/soziale-situation-in-deutschland/61538/altersgruppen
(2) https://www.destatis.de/DE/Publikationen/Thematisch/Gesundheit/Todesursachen/Todesursachen2120400117004?__blob=publicationFile
(3) https://www-genesis.destatis.de/genesis/online/data;jsessionid=EB6EC773C665904C219D19973DEE5D54.tomcat_GO_2_2?operation=ergebnistabelleUmfang&levelindex=2&levelid=1437993011892&downloadname=12613-0003


EDIT: Typos...

Thanks for crunching all those numbers to give us something interesting to read.

although statistics are always flawed in this case it may paint a general picture. Unfortunately I think it may be ever higher than 30 times for you since dangerousness depends on many factors and by doing 500 jumps a year you are probably engaging in more difficult and dangerous forms of skydiving (faster canopies, wing suits, large group jumps -> increased risk canopy collisions and the list goes on) furthermore by doing so many jumps you are more likely to find yourself getting lax with a gear check from fatigue or complacency. Perhaps your numbers account for all of it, perhaps it evens out when you factor in your experience that adds to your safety as a jumper.

At the end of the day calling skydiving safe is really just a defense mechanism against wuffos in our workplaces or outside life who pass judgement on us and think they are better and more mentally sound than us for not jumping out of airplanes. And to those I say it's safer than crossing the street to get a hamburger!! Could get hit by a bus

chemist

although statistics are always flawed in this case it may paint a general picture. Unfortunately I think it may be ever higher than 30 times for you since dangerousness depends on many factors and by doing 500 jumps a year you are probably engaging in more difficult and dangerous forms of skydiving (faster canopies, wing suits, large group jumps -> increased risk canopy collisions and the list goes on) furthermore by doing so many jumps you are more likely to find yourself getting lax with a gear check from fatigue or complacency. Perhaps your numbers account for all of it, perhaps it evens out when you factor in your experience that adds to your safety as a jumper.

Those were exactly my thoughts but I think it is very difficult to get personal experience and behaviour into a trend analysis without getting away from cold facts and mixing too much subjective impressions into it.

As already stated - statistics are only valid for the average. And as nobody is average in all perspectives, statistics have no influence or predictive power on ones personal future. There will always be people that are above the calculated risk and there will always be people that are below the calculated risk - but no one can tell for sure for a single person. Everybody can just estimate for himself and I have the feeling that a lot of skydivers overestimate. Maybe this can be regarded as identical or similar to the known issue, that 90% of car drivers consider themselves as above average drivers. Of course this is simply not possible - but still, people believe it. Maybe a lot of us are not as safe as the think they are (myself not excluded).

My comment from an earlier thread on skydive safety:

Quote

I also wondered about the micromort of a single tandem skydive. If that was only 1 - 2 micromorts, I think you could call a tandem skydive safe, because it close to the baseline of 1 micromort of all non-natural deaths we face every day.

Seth

You are making a bit of a mistake by considering all "skydiving" the same.

Choices that jumpers make can drastically change the risk level of a jump.

For example, jumping with an AAD & RSL, jumping a larger, lighter loaded canopy and not making aggressive landing approaches will greatly reduce the risk of a jump.
At least the past fatality reports seem to indicate that.

I'm old enough to remember when sex was safe and skydiving was dangerous ........

..... Somebody had to be the first to say it .....

wolfriverjoe

Choices that jumpers make can drastically change the risk level of a jump.

For example, jumping with an AAD & RSL, jumping a larger, lighter loaded canopy and not making aggressive landing approaches will greatly reduce the risk of a jump.
At least the past fatality reports seem to indicate that.

I don't see it. If all would jump large only or small only - proportions of numbers would be the same.
When data is mixed, the results are misleading.

format

***Choices that jumpers make can drastically change the risk level of a jump.

For example, jumping with an AAD & RSL, jumping a larger, lighter loaded canopy and not making aggressive landing approaches will greatly reduce the risk of a jump.
At least the past fatality reports seem to indicate that.

I don't see it. If all would jump large only or small only - proportions of numbers would be the same.
When data is mixed, the results are misleading.

I don't understand what you mean. Can you elaborate please?

wolfriverjoe

For example, jumping with an AAD & RSL, jumping a larger, lighter loaded canopy and not making aggressive landing approaches will greatly reduce the risk of a jump.
At least the past fatality reports seem to indicate that.

This is true - of course it would be something that is nice to be taken into account. But you now what? The skydiving world is to small to get reliable data to predict to risk of any of these special activities. Just look at myself. There are many factors that indicate higher risk, but also many factors that indicate lower risk - but in total? No idea...

• I do high performance landings
• I fly a highly elliptical canopy
• My wingload is at 1.6
• I do not use an AAD
• I do not use an RSL

But

• I am highly current (during the summer about 100 jumps per month)
• I am under constant supervision of my “teacher” with 15000+ jumps
• I only jump on my home dropzone (usually)
• I only jump from the same small airplane (usually)
• There will never be more than 5 other people around me in the air
• I do not do big way trackdives or freefly jumps (4+ people)
• I do not do wingsuit jumps
• About 20% of my jumps are tandems (as an instructor of course)

So... Is my risk higher or lower than the risk of the average skydiver? Can you give any evidence for that (other than the feeling in your stomach)? I can't - and therefore the average skydiver is the best estimation I can give.

Congratulations, like 99% of the population, you don't know how to do statistics and probability.

I looked at the DFV numbers and did the math and I got answers varying from 21.something to 15.something, depending on whether I used the whole set or just part of it. Where did you get 18.55?

Numbers don't lie, but your math is way way off.

First off, you used Wikipedia for your information, and their method of cataloging it is suspect at best. Their link to the BPA's stats (found here) don't even have the same numbers. ( 4,864,268 jumps and 41 fatalities over 20 years vs. 5,012,215 jumps and 40 fatalities over 20 years ). The BPA dataset breaks it down by type of jumper. Does the DFV? What kind of trend do we see in the numbers? Is it steady over the last 20 years, increasing, decreasing, anomalous? Are there other factors involved? Did the Tandem Instructor who died on the ride up count?

Second, the USPA and BPA stats are based on 365 day jumping years, German skydive season is barely half the year if the weather cooperates, and many (most) DZs are still weekend only even in the summer time. Do you consider only actual open days in your total annual risk calculation? Do you do a similar consideration for the average German your age? (This is not even to mention that in comparison to almost every other country on earth, Germany is incredibly safe. Congratulations on living in the place with the best average drivers in the world) Of all German Skydivers, are 100% of them (and the fatalities) only in the age 20-29 group?

Third, The vast majority of DFV members are inactive because they got their license, did a few jumps and then quit. How many are actually considered experienced? How many of the fatalities are current vs. novice vs. noncurrent? Wikipedia simplified this to the average for all jumpers, but the BPA Statistics clearly show that the injury/fatality rate drops by half or more for experienced jumpers vs. students, (1.5 per 100,000 for SL students, 2.5 per 100,000 for AFF students, vs 0.8 per 100,000 for experienced jumpers, not counting pro/demo jumpers) so Experience clearly is a factor in the actual statistics, but not the Wikipedia statistics or yours.

Put another way, if you go out and do 500 skydives in a year, you're actually reducing your risk, all other factors being equal.

That is to say, Skydiving is dangerous, but not as dangerous as you seem to be suggesting.

The BPA takes the total amount of fatalities in the UK over the total amount of jumps in the UK.
The USPA takes the total amount of fatalities in the US over the total amount of jumps in the US.
I took the DFV numbers for total amount of fatalities in Germany, excluding fatalities of German jumpers in other countries - the numbers in brackets, (85) over the total amount of jumps in Germany (4.580.865).
This is the same calculation method for all three numbers.

Dividing the number of fatalities (85) by the number of jumps (4.580.865) gives a probability of 0,000018555 for dying by doing one jump. Multiply this number by 1.000.000 and you get your risk in micromorts (18.55 or actually 18.56 - sorry for that mistake).

Using the same calculation method for the BPA you get a risk somewhere between 7,98 and 8,42 micromorts per jump (depending on which of the two numbersets you use).
For the USPA you get a risk of 8,85 micromorts per jump (41.900.000 jump, 371 fatalities).

The number of jumpable days per year does statistically viewed not affect this value - this should be clear by already seeing that the number of jumpable days is not part of any calculation. I did never calculate the risk of one day of skydiving, only the risk per jump. There might be currency issues causing the higher fatality rate in Germany but this is again something you will not be able to measure.
What is of course estimated is that the fatality rate did not really change in the slightly different timespans (BPA - 1995-2014; USPA 2000-2014; DFV 1996-2011) - and I think I am not completely off with this estimation or at least the differences are so minimal that they do not really have any significant influence on the result.

Experience is a factor. Like there are many others. And with my high currency I am definitely reducing my risk per jump, but still I do more jumps than other people. So - where's the break even? As long as you can not give me any clear numbers with clear calculations and clear sources, the average is the best and only estimation one can make.

maggyrider

The number of jumpable days per year does statistically viewed not affect this value - this should be clear by already seeing that the number of jumpable days is not part of any calculation..

It's not part of any calculation for micromorts for skydiving, it certainly is if you're comparing your chance of dying compared to the average German your age.

When I do not jump on a certain day, I can not have an accident jumping on that day.
When I jump 15 times on a certain day, I can have an accident 15 times on that day.

Of course the risk is higher during the summer, when I jump, and much lower if not equal to the standard persons risk in the winter, when I jump less or not at all. I took the average of a whole year because it is not really possible without a lot of effort to take jump frequency over season into the equation. I said ones risk of dying within the next 12 months is x times higher than the risk of an average non skydiving person in the same timespan - and that is correct.

Quote

the BPA Statistics clearly show that the injury/fatality rate drops by half or more for experienced jumpers vs. students, (1.5 per 100,000 for SL students, 2.5 per 100,000 for AFF students, vs 0.8 per 100,000 for experienced jumpers, not counting pro/demo jumpers)

Then again, you don't see many students cranking out 500 jumps per year. To simplify: If you make 10 times as many jumps as the average student but one single jump you make is "half as risky" due to your experience and being current etcetera you end up with five times the risk involved in participating in the activity i.e. not being around next season.

This of course is only looking at the risk of dying. All the other carnage (broken legs, pelvis, spine etc.) I have seen clearly shows me that Skydiving is not a safe activity.

You just have to ask yourself is the risk worth the fun you get out of it? As time goes on I find myself coming ever closer to the answer of "no".

maggyrider

This is true - of course it would be something that is nice to be taken into account. But you know what? The skydiving world is to small to get reliable data to predict to risk of any of these special activities...

...Can you give any evidence for that (other than the feeling in your stomach)? I can't - and therefore the average skydiver is the best estimation I can give.

I understand that the statistical pool is small, and getting good data is difficult.

But if you look at the USPA fatalities, lack of RSL, lack of AAD & small canopies are major factors in a large percentage of deaths.

Cutaways at adequate altitude, but no reserve pull. Greatly reduced by having an RSL (not eliminated).

No pull fatalities. Greatly reduced by having an AAD (again, not eliminated).

Bad landings under small canopies, either intentional swoops or just bad landings (often off-field or in marginal conditions). Again, reduuced but not eliminated by larger canopies.

I just dug out my Parachutist from April, and with a quick look, 12 of 24 deaths could have been (not would have been) prevented by these three things.

IIRC, previous years have shown similar trends.

So, while it is a lot of "gut feeling" (stomach), and injuries are not in the mix; I can show at least some evidence to back up my claim.

And as a sidenote: I am not advocating any sort of requirement for AADs or RSLs. I am in favor of minimum experience requirements for advanced canopies.
I'm mainly pointing out that the "average" risk in the stats can be reduced by a significant amount by making certain choices.

wolfriverjoe

I don't understand what you mean.

I meant that risk factor may be "useful" if it is given for SAME control group (age/experience/weather/canopy/wing load etc.) But it is not as far as know. These micromorts means nothing to me, they don't make me feel more or less safe at all.

I woudn't ever bet on predicting skydiving accident based on statistics (or based on anything else).
There is enough jumpers with constantly high risk factor and nothing happens. On the other side, ultra safe jumpers die and vice versa. What ever happens doesn't surprise me and numbers do not help.
What I think is that statistics gives a false idea of having knowledge, therefore is more risky to know it than not.

Ahhh...

Thank you. That makes perfect sense.

I fully agree that statistics only tell us about the past.

They can be helpful in predicting the future, but only to a certain degree, and that degree drops greatly as the sample size does.

I also find it interesting that risky activities are viewed almost as zero sum. Which is riskier? Doing five jumps on a Saturday or riding the motorcycle an hour each way to get there? Turns out that the risks don't cancel each other out. They just stack up.

It's an interesting philosophical question, but the more risky activities one does the greater the opportunity of encountering an event.

MikeBIBOM

Congratulations, like 99% of the population, you don't know how to do statistics and probability.

Put another way, if you go out and do 500 skydives in a year, you're actually reducing your risk, all other factors being equal.

You pointed out some very valid items in your post, however I think your last statement here is very incorrect. I included your first statement as well because irony ftw! Yes, if you do 500 skydives per year you will be very current and your per-jump injury/fatality probability will be less then someone who only makes 50 jumps a year. 500 jumps is a lot. All factors are not equal, one of these people is exposing themselves to the event by a factor of 10x more and has the higher over-all risk of something bad happening. Logical fallacy imo.

It's pretty simple, really. In my entire lifetime, I have known three people that died in a car accident.

Since 1980, I have been friends with 46 people that have gone in skydiving. Not included in this statistic are friends that died in skydiving aircraft accidents.

Carry on...

I did a similar analysis a while back using total miles driven per year versus the number of traffic fatalities per year here in the United States, compared to the total number of skydives made versus fatalities (in the US) and came to a similar conclusion. The number I came to was 1,000 miles. Skydiving is about as dangerous as driving 1,000 miles in your car.

Now, let's see interesting statistics
by The School of Life

..............and ? So now what ?