Hacker News Reader

Comment by clueless123 | Hacker News Reader

Here we go again.. (I think we discussed this topic previously)

Skydiving, hang gliding,Paragliding and many other "adventure sports" contain a significan amount of risk.. the difference is that skydiving (not base jumping) is tightly regulated by the FAA, there are strict training requirements, equipment maintenance etc etc .. it is almost like flying a plane. so.. it is super dangerous compared to knitting.. but it's quite safe compared to many other risk activities.. (Paragliding, hang gliding, motorcycle racing on a street, experimenting with drugs, proximity flying etc)

Most of the people I've met during 20 years skydiving are very much alive and still jumping, several hundred people..

View on HN

Oh hey, you're right, we did discuss this a while ago. Nice to see you again!

Yes, you're right that skydiving is safe compared to the sports that are more dangerous than skydiving. I agree and admit there are sports and activities that are more dangerous than skydiving. Wingsuit flying is one of them.

Last time we talked, you said the risk of dying skydiving is 1 in 100,000 and pointed at the comparative table here: http://www.besthealthdegrees.com/health-risks/

https://news.ycombinator.com/item?id=12419507

That number is per jump, not per year. The other 11 pictures in the top 12 of that page are per year figures. The units are different, the comparison is invalid. Juxtaposing these numbers is wrong, and quoting 1/100,000 per jump is super misleading when you compare it to the annual fatality rates of other sports.

This is what I was talking about above. The annual fatality rate of skydiving is 1 in 1,580. Skydiving is a bit safer than hang-gliding, roughly 3x. It's definitely not 180x safer, which is the very first thing the infographic implies.

The infographic depicts skydiving as safer than boxing, football, canoeing, and SCUBA. That is flat false, skydiving is much more dangerous than all of those combined.

BTW, with respect to regulation, about one in about 1600 skydivers die each year after having gone through training, after safety inspections, after exposure to the risks and liability waivers, after subscribing to USPA magazine and reading about each and every death in the U.S., etc. etc. Since nobody needs a license to get in a canoe, you may expect a higher percentage of people who are unprepared for the risks of canoeing to die doing it. Skydiving has a relatively high fatality rate despite the fact that it's regulated and the skydiving population has more training and safety equipment than other sports.

  Since nobody needs a license to get in a canoe,
  you may expect a higher percentage of people who
  are unprepared for the risks of canoeing to die doing it

Breaking down the numbers, the skydiving fatality rate is a reported 1 in 101,083 jumps. To normalise that to a per-year rate, we can take the USPA rate of 350,000 participants conducting 3,000,000 skydives per year [1]. That gives a mean of 8.6 jumps per year for each participant.

Given the 1 in 101,083 risk, that amounts to roughly a 1 in 11,750 risk of fatality per year.

The risk of canoeing in the linked article [2] is 1 in 10,000, so according to those reported statistics, skydiving is slightly safer than canoeing, significantly safer than boxing, is roughly 3 times as risky as scuba diving, and 5 times as risky as football.

[1] http://adventure.howstuffworks.com/skydiving8.htm

[2] http://www.besthealthdegrees.com/health-risks/

Wrong. USPA has 35k participants, not 350k. http://test.uspa.org/facts-faqs/safety

  Decade                           2010-2015
  Average fatalities per year      22.3
  Average membership per year      35,189
  Fatalities per thousand members  0.63

You didn't bother reading my comment? You just shared the same link I shared and discussed above. Those two links you shared are both examples of failures of statistics. They are comparing the risk of a single jump to the risk of a year of participation in other activities. That's a math fail.

Skydiving is most definitely not safer than canoeing or boxing or SCUBA or football. It doesn't even pass the smell test that football could be more dangerous than skydiving. You've been misled by a cute poster that someone who's bad at stats screwed up.

I did read the comment, and reposted the same link to identify it as the source of the canoeing stat.

My reference to 350,000 participants is this text [1]:

  The United States Parachuting Association has nearly 35,000 members.

  It estimates that about 350,000 people complete more than
  3 million jumps in a typical year.

3,000,000 jumps / 350,000 people = 8.57 jumps per person per year (approx).

  It doesn't even pass the smell test that football could
  be more dangerous than skydiving.

The perception of football as "safe" can be attributed to its popularity: many people play football, therefore it's viewed as normal, just like driving.

[1] http://adventure.howstuffworks.com/skydiving8.htm

You've used an unsourced figure from a website that used faulty logic to evaluate the risks of skydiving relative to driving, in an intentionally misleading way.

You've done some calculation based on this unsourced number, while ignoring the mortality rate that the USPA publishes.

The USPA themselves says the annual risk of mortality is 0.63 per thousand members. That's about 1 in 1600.

You are off by an order of magnitude. It's no wonder you think skydiving is safer than canoeing, but it is not true.

The stats we have are for members and excludes non-members. But one thing you might be missing is that non-members jump an average of once. At least 90% of the 3.2 million jumps are repeat jumps by members, members comprise the bulk of the statistics anyway.

I agree that the reference to 350,000 is unsourced - the article claims that number comes from the USPA, but I can't find any relevant data on the USPA site.

  non-members jump an average of once.
  At least 90% of the 3.2 million jumps are repeat jumps by members

That was the figure that I was struggling to source, but 90% feels high to me. At 90%, that gives a mean of 82 jumps per member per year, which feels high to me (although my experience as UK skydiver may be skewed by the amount of time we spend on the ground waiting for the skies to clear!)

It would be useful to see a more complete breakdown of the jump stats and fatalities, because it might indicate the riskier elements of skydiving (my gut feel says swooping or high wing-loading), but I haven't found a data breakdown that detailed.

You're right; it would be useful to see a more complete breakdown, and get a better sense of what specific aspects are riskier. Hook turns are definitely a significant culprit, I agree with you. The BPA website (link below) has some claims about what the risk factors are.

> 90% feels high to me.

Well, I'm guessing on top of the unsourced 350k number you shared, so maybe it would be best if we stop guessing. But most non-members that jump are doing a tandem jump. A tandem jump in the US involves a USPA member doing a repeat jump, and most of the time a non-member making their first jump.

All we can say firmly is that 1 in around 1,600 trained and qualified USPA members die each year. That is the actual mortality rate for people who continue to participate in skydiving. Since you're in the UK, the BPA also publishes safety info separately http://www.bpa.org.uk/staysafe/how-safe/

The one thing you can't say is that canoeing is more dangerous than skydiving. According to the sources we have, canoeing mortality rate is 1/10k while skydiving is 1/1.6k. Skydivers die ~6x more per year than canoers.

Drivers in the US die at the rate of 1/6k per year according to the sources you shared. So skydiving is ~4x riskier than driving. And canoeing is less risky than driving.

We can ignore the question of how many jumpers are USPA members if we look at the per-jump fatality statistics. The 1 death per 101k jumps lets you evaluate the risk of doing a single jump vs doing 100 jumps in a year. Let's use the current USPA numbers instead, which is now averaging around 1 death per 140k jumps (shown here http://www.uspa.org/facts-faqs/safety).

Your chance of dying doing a single jump is 1 in 140k. That seems low, but it's a single jump. Your chance of dying in a given year, if you make 100 jumps, is 100 in 140k, or 1 in 1,400. Some people do more than a thousand jumps per year, so if you jump 1000 times, your odds are 1 in 140. Looking at it this way, it's easier to gauge the risks because jumping more means risking more.

Please look at my comment above on Accidents vs Recklessness. As an example of a group that is well trained, follows procedures and plans properly, take the Golden Knights skydiving team.

They had a handful of deaths since they started in 1959. And their worst accident was a fatal aircraft failure, which is technically an aircraft accident, not a skydiving accident.

Read & replied. Valid point, IMO. I would agree completely that people who make safer choices will live longer statistically speaking than people who push the limits, cross more lines, fail to plan, etc.

The numbers we have are averages, there are definitely people beating the average through greater attention to safety. And, unfortunately, some groups that are on the other side of average.

Equipment is getting better too, and I have no doubt is helping lower the death rates. Pretty sure that loads more people were dying under old rounds that today's controllable squares. Lots of people died in WWII in Jeeps under parachutes. I remember the museum in New Orleans saying that Jeep jumps had a fatality rate of something like 30%. Can't tell you how thankful I am that I've never been asked to do that.

Would be interesting (but small sample size) to do the math on the Golden Knights and find the mortality rate. I totally expect it to be lower than for the general skydiving populating.

Some drivers take more risks than others, and insurance companies gather a bit more data on that kind of stuff. Rates might be higher if you get a speeding ticket, precisely because that correlates with accidents and claims more than people who have no record.

There is a convenient unit for assessing risk : the micromort. 1 micromort = 1 in a million chance of dying.

Micromort values from wikipedia :

- skydiving : 8 per jump

- hang gliding : 8 per trip

- motorcycle : 1 per 10km

- Ecstasy (MDMA, drug) : 0.5-13 per pill

- BASE-jumping : 430 per jump

So as you can see; skydiving, hang-gliding, motorcycling and drugs are in the same ballpark. BASE-jumping is way worse.

When you are young, you "spend" around 1 micromort per day just being alive, it goes up steeply as you get older.

I hadn't heard about micromorts before, thanks! That's definitely an interesting and valuable unit for comparison. It adds perspective to the fatalities per year.

I'd add that looking at participation-hours is another interesting way to look at it.

For example, my average day skydiving was 3-10 jumps. So I might have spent up to 80 micromorts in a day. But a freefall is at most 60 seconds, so I could also say I spent 80 micromorts in 10 minutes. Hang-gliding is something you only do maybe once or twice in a day, with flight times being 30-60 minutes.

It might not be fair to use participation hours for various reasons, but if you do slice it that way, skydiving looks a lot more dangerous.

There is a hidden factor here that I don't think is being taken into account. Accidents vs Recklessness.

Let me explain: When you do a risk activity, the way to mitigate risk is to bound the situation to keep all variables within safe margins.. For example, on aircrafts we look at the mean time of failure on a part, then you replace the part before it gets there, so you reduce the chances of it failing..

On adventure sports we do the same. When Skydiving, Rock climbing, Paragliding etc.. you keep looking at accident causes and develop a plan to avoid those situations.

So, If you do follow all the existing safety procedures,and still get in trouble, it is probably an accident, something you could not prepare or account for. (procedures are then created to mitigate its repeat occurrence)

On the other side.. if you disregard the safety procedures (push de envelope, as we say) and get in trouble, then you were reckless, and you paid heavily for it. That is NO accident. (if you get away with it.. you were just dumb lucky)

My point is.. I bet the micromorts computed here are based on "all" accidents, not on accidental vs reckless deaths.

With all due respect to the death.. if you seen a cluster of 20 deaths within a small group, you need to look closely at the reason why those deaths happened.. Where they trained properly? Where they at the appropriate experience level to engage on the type of activity they where doing? (RW, CRW, fast landings etc) Where there other external factors involved? , was there a single significant accidental event that affected many members of the group at the same time? (Aircraft crash for example..)

> I bet the micromorts computed here are based on "all" accidents, not on accidental vs reckless deaths.

That is definitely true. These figures are dividing all deaths annually by number of participants. We are in murky water just by talking about risk rather than mortality rate. I'm guilty of doing that. We can't know the risk of a skydive per-se, we just know that 1/100k skydives in the past has ended in a death. We also know that the rate of deaths per skydive is going down over time, given the last 20 years of data.

> There is a hidden factor here that I don't think is being taken into account. Accidents vs Recklessness.

You're right on both counts. 1- some people take more risk than others in a given sport, and 2- that relative risk-taking within a sport is not taken into account in the mortality rates we have at hand.

Recklessness is a subjective, touchy-feely concept though. It's opinion and not something we can measure, so we will never be able to take it into account. We also don't know whether people within one sport are any more "reckless" than in another. Reckless behavior might be correlated with overall risk, or it might be heavily influenced by training and regulation and community.

Most deaths aren't due to crossing the line of any specific safety procedure. It's not recommended to hook turn, but it's also not a rule or law. A lot of people choose to do it for the extra thrill, and know that it's riskier than a straight upwind landing. All reckless deaths are also accidents. The line just might be too hard to draw.

Many people do call choosing to participate in a risky sport "reckless". Proximity wingsuit flying seems reckless to me, it's an activity that is very prone to deadly accident. You can certainly play it safer than some other people, but by choosing to participate you're accepting a level of risk that is unacceptable to some other people. I also acknowledge individual choice in this regard. It's okay to accept a higher level of risk than someone else.

The main thing that's probably bad to do is to rationalize a choice by thinking it's less risky than it really is. If you calculate your risk wrong, then you can't make an informed choice, which leads to unintentionally reckless decisions.

The other thing not being considered in this thread is injury rate vs fatality rate. I think it's pretty reasonable to weigh fatality much more heavily than recoverable injury, but I will openly admit that the big picture has a lot more gray area than the black/white numbers on fatalities per year.

For some perspective here, the maternal mortality rate in the USA is 28 per 100,000 births. So skydiving is about 50x safer than giving birth.

If not dying every time I jumped resulted in a new human being born, I might agree, but the death rate of births is a net negative. :P

Joking aside, that's an interesting comparison but a bit difficult to really put side by side in a way that's fair. Skydivers jump on average around 90 times a year, which is why they die about 1 in 1,600 per year. Mothers only give birth once per year and on average less than 2 times per life in the US.

A average skydiver's chance of dying in after 20 years is 1 in 80. The average mother's change of dying from birth after 20 years is 1 in 2,000.

>but the death rate of births is a net negative

That’s a seriously interesting point. Maternal death rate is 280 micromorts. Infant death (neonatal + perinatal) is about 1% in modern countries (10,000 micromorts), for a net of -989,700 micromorts, or +0.99 lives (not counting multiple births, etc).

Reply on news.ycombinator.com