Parchustist' equipment reports for fatalities

[EnricoPalazzo 1]

Just one more thing, in anticipation of a follow-up argument:
The data pool cannot really be increased over a long time, since all gear will see a lot of evolution over a time span that makes the aggregation of meaningful data futile.
Before you have a large enough sample size about helmets, goggles, containers etc., the manufacturer will have modified the gear so that there is not much to be learned from historical data.

[mxk 1]

Lies, damned lies, and statistics. I'm not sure why this post ended up being so controversial. Wingsuit type should absolutely be reported. Helmet and any other gear should also be reported if it's relevant to the incident.

I don't care that G3 is the most popular helmet. If it keeps showing up in head injury incidents, maybe the bad publicity will finally get helmet manufacturers to produce a helmet that actually offers some protection. When a helmet doesn't do it's job, that's relevant information even if everyone is wearing the exact same helmet model. It's not always about comparison with competition.

[mark 102]

mxk

Lies, damned lies, and statistics. I'm not sure why this post ended up being so controversial. Wingsuit type should absolutely be reported. Helmet and any other gear should also be reported if it's relevant to the incident.

I don't care that G3 is the most popular helmet. If it keeps showing up in head injury incidents, maybe the bad publicity will finally get helmet manufacturers to produce a helmet that actually offers some protection. When a helmet doesn't do it's job, that's relevant information even if everyone is wearing the exact same helmet model. It's not always about comparison with competition.

Using your reasoning, it's safer for wingsuiters to jump without a helmet (since there are no or very few wingsuit fatalities involving no-helmet jumpers), than it is to jump with a helmet (since nearly every wingsuit fatalitiy involves a jumper wearing a helmet).

-Mark

[lasharp 4]

If you look at the incident reports in Parachutist, you'll see that "helmet" is one of the gear items reported.

[jcd11235 0]

benlangfeld

It’s probably useful for better understanding of individual incidents. It’s not useful (indeed it’s possibly dangerous) for the purpose of spotting patterns.

Not necessarily. The distribution of causes of incidents would likely be very different for wingsuit pilots vs. non-wingsuit pilots. We might find similar differences in incident causes between those who fly cross-braced canopies and those who do not, or between those who fly camera and those who do not. It can be very useful to consider equipment when considering incidents.

We do, however, have to be mindful of our degrees of freedom when we are identifying possible features for consideration. Thus, we are (fortunately) not likely to see enough fatalities in a short enough time period to identify any significant safety differences between, for instance, a Sun Path H&C and a UPT H&C. (Note that I'm in no way suggesting Sun Path's containers are safer than UPT's, or vice versa.)

Math tutoring available. Only $6! per hour! First lesson: Factorials!

[jcd11235 0]

benlangfeld

The argument that's being presented is a purely mathematical one which states that anything that you might do with the data once gathered that you wouldn't have planned to do before gathering it would be an attractive un-scientific trap. "See what it tells us" sounds like code for "if X brand helmet shows up a lot, there MUST be something wrong with it", which would be lacking in rigour.

To rephrase, the same data that raises the question cannot be used to answer the question. We have to obtain a new sample to answer new questions.

Math tutoring available. Only $6! per hour! First lesson: Factorials!

[Bob_Church 7]

It's like everyone got a day at a time calendar of statistics terms so now everything is an excuse to use them.
Should we consider pin checks to be statistically insignificant and stop doing them?
Or recognize the fact that not everything has to fit the model you've decided that it isn't good for?

[LeeroyJenkins 2]

Bob_Church

It's like everyone got a day at a time calendar of statistics terms so now everything is an excuse to use them.
Should we consider pin checks to be statistically insignificant and stop doing them?
Or recognize the fact that not everything has to fit the model you've decided that it isn't good for?

Alright, this is getting ridiculous. The majority of the information you have been given is correct. You really do appear to be taking it personally.

I'm not against colleting more information when it comes to incidents. I am however against attempting to draw a conclusion from that data. You cannot, not a suggestion or opinion, you cannot conclude patterns from incomplete data. Negating degrees of freedom, the fact that we do not have data on the total number of jumps overall with a specific piece of gear make it impossible to conclude that specific gear is more or less likely to be involved in an incident, let alone be a factor in the incident.

The reason people are hesitant to collect the data is because people that do not know about stats will attempt to draw conclusions from that data. This is a valid concern as it happens often.

Source: BS Mechanical engineer, minor in math, and masters in policy analysis. I know stats.

[benlangfeld 1]

> Should we consider pin checks to be statistically insignificant and stop doing them?

No, that's precisely the point. Pin checks tell you "my pin is good on this jump", not "my pin has to be good because it's in a vector". It's the same sort of difference here. Knowing what wingsuit someone was wearing when they impacted might help you understand why they impacted, but it won't tell you reliably that a certain wingsuit model is more or less dangerous than any other, or even that wearing a wingsuit is more dangerous than not.

[Bob_Church 7]

lasharp

If you look at the incident reports in Parachutist, you'll see that "helmet" is one of the gear items reported.

And everyone keeps saying "it will just be G3" But what if it isn't? What if some odd rarely used helmet keeps showing up? If you look at the number of blast handle fatalities in ten years is that number statistically significant? It didn't have to be. We're humans and looked at what was going on and changed what we were doing.
I just think we should collect the data. Not statistics terms. Data. Get the information. People can use it for whatever it turns out to be useful for, if it does.

[Bob_Church 7]

mark

***Lies, damned lies, and statistics. I'm not sure why this post ended up being so controversial. Wingsuit type should absolutely be reported. Helmet and any other gear should also be reported if it's relevant to the incident.

I don't care that G3 is the most popular helmet. If it keeps showing up in head injury incidents, maybe the bad publicity will finally get helmet manufacturers to produce a helmet that actually offers some protection. When a helmet doesn't do it's job, that's relevant information even if everyone is wearing the exact same helmet model. It's not always about comparison with competition.

Using your reasoning, it's safer for wingsuiters to jump without a helmet (since there are no or very few wingsuit fatalities involving no-helmet jumpers), than it is to jump with a helmet (since nearly every wingsuit fatalitiy involves a jumper wearing a helmet).

-Mark

If wingsuiters jump without helmets and 'None' keeps showing up in fatalities it would show that wingsuiters really should think about getting a helmet. I'm not even suggesting something new. We've been collecting gear information after fatalities for decades. The gear has changed and I think the information we collect should reflect that. We've always asked "helmet?" Now, with the profusion of different types I think we should add "which one". And how there can be any question about asking wingsuit type.
Personally I don't see how much of any of the info we gather helps, but I don't mind. I consider it my part to collect the information not work out how it gets used. But I don't think we should avoid information based on "but someone might take it wrong"

[LeeroyJenkins 2]

Bob_Church

***If you look at the incident reports in Parachutist, you'll see that "helmet" is one of the gear items reported.

And everyone keeps saying "it will just be G3" But what if it isn't? What if some odd rarely used helmet shows up?


That would mean they weren't wearing a G3 and that is it.

Quote

I just think we should collect the data. Not statistics terms. Data. Get the information. People can use it for whatever it turns out to be useful for, if it does

This is exactly why the data shouldn't be collected. It will not turn out to be useful for anything because it is incomplete and therefore inaccurate and unusable.

[Anachronist 2]

I'm still loling

[Bob_Church 7]

Anachronist

I'm still loling

Should we write that down or something?

[bryanburke 10]

I've been an STA since they were ASOs. For the last dozen years I've been keeping records of all of our reserve rides, injury accidents, and even most incidents that could have resulted in accidents. And, unfortunately, a number of fatalities.

I agree with the people who say the amount of data isn't really enough to draw firm conclusions, but I think there is enough to suggest some things. Before I go into some examples, let me mention an area where accident reporting has done some good and I think we can learn from it - skiing, specifically the work of the National Ski Patrol to collect and analyze accident reports. Two examples that really stand out are binding/boot interactions and ski pole design. In the first case, when rigid plastic boots became the norm and bindings were not very good, there were a lot of leg fractures at the boot top. Bindings got better, and then the injury trend (over years, mind you) moved to torn ligaments in the knee. Bindings were releasing in most falls at this point, but not in backwards twisting falls, where the toe piece failed to release because they had only been designed to release sideways, and there was insufficient pressure on the heel piece to release. The answer was to design toe pieces that could release from upward pressure as well as sideways. People still get leg injuries in falls, but the bindings improved dramatically over twenty years in response to information about how legs were being injured.

With the case of ski poles, in the early 60s they were still basically cross country poles, with a webbed basket base and a thin handle. Besides the baskets catching on things like branches, there were a lot of eye injuries when people fell against the pole and the handle penetrated the eye. That's why all poles now have a top of the handle that is too big to penetrate the eye, and the baskets are designed not to catch on stuff.

How does this apply to skydiving? As an example, my data suggests wingsuits are more likely to have malfunctions. How do I know? If they make about seven percent of all jumps but are more than ten percent of the reserve rides, maybe there is a corrollation. I'll wait a couple of years and see. Does it matter? Maybe. Even if they don't get hurt they are more likely to lose a main. And does the type matter? I don't know yet, but this year we're starting to collect that.

I can say that the single highest cause of reserve rides at Skydive Arizona is high wing loadings on elipticals. (Of course, nobody jumps elliptical canopies at low wing loadings.) The second biggest cause is toggle problems. Premature releases, mostly, but also knots or similar problems from the loose lower section getting caught on the toggle, guide ring, or around the riser. This was a virtually unheard of malfunction when we still used velcro to secure the toggle and steering line loop. So if you don't want to ride your reserve, jump a seven cell at a moderate wing loading and get an extremely secure toggle system. These are very clear choices that would reduce our reserve ride rate by a lot. But many people weigh the cool factor of the fast canopy over the reserve ride factor.

Twice (in a couple million jumps) we've had canopy lines half-hitch around a side container flap during deployment, creating an un-releasable main malfunction. In one case the reserve inflated but was slowly choked by the main, resulting in two badly broken legs. In the second case the free-bag and pilot chute spun up with the streamering main, but somehow the reserve made it out of the bag and opened in time. Frankly I thought I was going to watch the guy streamer in. When I saw the free-bag and pilot chute almost hidden inside the main lines and fabric, it was just a miracle the reserve got out cleanly.

Both times, ten years apart, had the same configuration: worn pull-out deployment system AND the same container model where the side flaps are squared off at the bottom rather than angled or rounded. Does this make it a bad system? Not really, in fact it is very popular. But not the one I would chose because this is not how I want to die.

Another interesting statistic: women are less than half as likely as men to die in a skydiving accident. Changing attitudes would save a lot more skydivers than changing gear.

All that stuff about too small a sample, constant changes in gear... that is true. But we might be able to draw at least some conclusions.

However, there's another problem that hasn't been mentioned.

I have the data to determine what the main, reserve, and container distribution of Skydive Arizona's boogie population is, over many years. I have their jump numbers, preferred discipline in the sport, gender, and age. I have the accident reports and reserve ride reports. What I don't have is the time to put it altogether, and we can't afford to hire someone to do it. It might never be put to use because of that. But I keep it anyway, in the hope that someday I'll be able to use it, or find an eager student wanting to do a dissertation on skydiving accidents.

Bryan Burke
S&TA at Skydive Arizona

[Bob_Church 7]

"I have the data to determine what the main, reserve, and container distribution of Skydive Arizona's boogie population is, over many years. I have their jump numbers, preferred discipline in the sport, gender, and age. I have the accident reports and reserve ride reports. What I don't have is the time to put it altogether, and we can't afford to hire someone to do it. It might never be put to use because of that. But I keep it anyway, in the hope that someday I'll be able to use it, or find an eager student wanting to do a dissertation on skydiving accidents. "

I keep kicking myself about the timing. Now that I'm retired I have more time for coming up with these projects but now I don't have the dozens of work study students I had at Ohio University's photo art labs. Most of these students worked about 12 hours a week sitting at the checkout/checkin window. It was expensive equipment so someone had to be there but sometimes they'd go hours with nothing to do and would love to have some info to type into the computer just to kill the time. At least my typing is getting better.

[benlangfeld 1]

Quote

If they make about seven percent of all jumps but are more than ten percent of the reserve rides, maybe there is a corrollation.

And the important thing there is that you apparently know they make up 7% of all jumps. Most DZs don’t have that kind of data about non-incident jumps.

I’m curious, how do you gather that? If filling out accident reports is so onerous that people complain about it, imagine providing that data for every jump.

[Bob_Church 7]

I think technology will catch up in the near future. Right now we have very fast, very good scanner. My HP Scanjet 8300 along with Vuescan can capture an entire issue of Parachutist in something like 20 minutes and most of that is me feeding the pages in. The problem of course is that OCR isn't there yet, at least not in the un-classified versions we have access to but they're working on it.
It won't be long, I believe, before you'll have a device sitting on your desk that you can put a stack of your forms in and not only have them scanned but the information massaged into the correct format and placed into whatever database or other software you're using.
One piece of technology that I don't believe will exist is a time machine. If we don't gather the information then it's lost. Not having the information because we didn't think to collect it is unfortunate. Not having it because we deliberately avoided collecting it is downright disturbing.

[LeeroyJenkins 2]

bryanburke

I've been an STA since they were ASOs. For the last dozen years I've been keeping records of all of our reserve rides, injury accidents, and even most incidents that could have resulted in accidents. And, unfortunately, a number of fatalities.

I agree with the people who say the amount of data isn't really enough to draw firm conclusions, but I think there is enough to suggest some things. Before I go into some examples, let me mention an area where accident reporting has done some good and I think we can learn from it - skiing, specifically the work of the National Ski Patrol to collect and analyze accident reports. Two examples that really stand out are binding/boot interactions and ski pole design. In the first case, when rigid plastic boots became the norm and bindings were not very good, there were a lot of leg fractures at the boot top. Bindings got better, and then the injury trend (over years, mind you) moved to torn ligaments in the knee. Bindings were releasing in most falls at this point, but not in backwards twisting falls, where the toe piece failed to release because they had only been designed to release sideways, and there was insufficient pressure on the heel piece to release. The answer was to design toe pieces that could release from upward pressure as well as sideways. People still get leg injuries in falls, but the bindings improved dramatically over twenty years in response to information about how legs were being injured.

With the case of ski poles, in the early 60s they were still basically cross country poles, with a webbed basket base and a thin handle. Besides the baskets catching on things like branches, there were a lot of eye injuries when people fell against the pole and the handle penetrated the eye. That's why all poles now have a top of the handle that is too big to penetrate the eye, and the baskets are designed not to catch on stuff.

How does this apply to skydiving? As an example, my data suggests wingsuits are more likely to have malfunctions. How do I know? If they make about seven percent of all jumps but are more than ten percent of the reserve rides, maybe there is a corrollation. I'll wait a couple of years and see. Does it matter? Maybe. Even if they don't get hurt they are more likely to lose a main. And does the type matter? I don't know yet, but this year we're starting to collect that.

I can say that the single highest cause of reserve rides at Skydive Arizona is high wing loadings on elipticals. (Of course, nobody jumps elliptical canopies at low wing loadings.) The second biggest cause is toggle problems. Premature releases, mostly, but also knots or similar problems from the loose lower section getting caught on the toggle, guide ring, or around the riser. This was a virtually unheard of malfunction when we still used velcro to secure the toggle and steering line loop. So if you don't want to ride your reserve, jump a seven cell at a moderate wing loading and get an extremely secure toggle system. These are very clear choices that would reduce our reserve ride rate by a lot. But many people weigh the cool factor of the fast canopy over the reserve ride factor.

Twice (in a couple million jumps) we've had canopy lines half-hitch around a side container flap during deployment, creating an un-releasable main malfunction. In one case the reserve inflated but was slowly choked by the main, resulting in two badly broken legs. In the second case the free-bag and pilot chute spun up with the streamering main, but somehow the reserve made it out of the bag and opened in time. Frankly I thought I was going to watch the guy streamer in. When I saw the free-bag and pilot chute almost hidden inside the main lines and fabric, it was just a miracle the reserve got out cleanly.

Both times, ten years apart, had the same configuration: worn pull-out deployment system AND the same container model where the side flaps are squared off at the bottom rather than angled or rounded. Does this make it a bad system? Not really, in fact it is very popular. But not the one I would chose because this is not how I want to die.

Another interesting statistic: women are less than half as likely as men to die in a skydiving accident. Changing attitudes would save a lot more skydivers than changing gear.

All that stuff about too small a sample, constant changes in gear... that is true. But we might be able to draw at least some conclusions.

However, there's another problem that hasn't been mentioned.

I have the data to determine what the main, reserve, and container distribution of Skydive Arizona's boogie population is, over many years. I have their jump numbers, preferred discipline in the sport, gender, and age. I have the accident reports and reserve ride reports. What I don't have is the time to put it altogether, and we can't afford to hire someone to do it. It might never be put to use because of that. But I keep it anyway, in the hope that someday I'll be able to use it, or find an eager student wanting to do a dissertation on skydiving accidents.

Bryan Burke
S&TA at Skydive Arizona

Send me a PM

[wolfriverjoe 1,340]

benlangfeld

Quote

If they make about seven percent of all jumps but are more than ten percent of the reserve rides, maybe there is a corrollation.

And the important thing there is that you apparently know they make up 7% of all jumps. Most DZs don’t have that kind of data about non-incident jumps.

I’m curious, how do you gather that? If filling out accident reports is so onerous that people complain about it, imagine providing that data for every jump.

The data is there. Waivers usually have the info on H/C type, Reserve type & size, main type & size. Also gender, age & jump numbers.

Manifest will usually differentiate WS jumps from other jumps. Not so much for angle, belly or FF, but it wouldn't be all that hard to sort out with a bit of asking around.

SDAZ, where Bryan is the S&TA, is likely the busiest DZ in the world (it isn't 'most DZs', not by a long way). It's to their benefit to keep tabs on stuff, if for no other reason than to give themselves some indication of who to watch out for.
If they simply fill out a short report for every reserve ride, with who it was, the rest of the info is already available.

The big issue (as was noted) is taking all that raw data to get a baseline established, and then applying it to all of the incident data to develop something valid and usable.

"There are NO situations which do not call for a French Maid outfit." Lucky McSwervy

"~ya don't GET old by being weak & stupid!" - Airtwardo

[faulknerwn 36]

I don't know if you guys have gone to electronic waivers but if so it would be easy to pull that data in. And I bet for paper waivers you could easily find a ton of rookie wanna be student skydivers who will enter that data from paper waivers into an Excel spreadsheet for jump credit.

[Luke_Quichotte 0]

Very interesting discussion. First the disclaimer:
I'm brand spanking new into the sport, as such, do not have enough knowledge of the sport itself to understand the intricacies. That said, I'm a Health & Safety guy with 15+ years of experience in managing safety in multiple companies in multiple countries with all kinds of different processes and risks. As such, this discussion peaked my interest.

I think Drew Rae's podcast "We dont Kill enough people" from disastercast http://disastercast.co.uk/wp/2014/06/03/episode-33-we-dont-kill-enough-people/ might be an interesting listen as he briefly discusses this topic.

In general, it is my believe keeping track of all irregularities makes sense. The question is where do you start and where do you stop?
This would mean defining irregularities. Hard openings & small malfunction too, lack of knowledge/training?
What information on an event do you track exactly, when is it enough?
Do you need a governing body to force a certain way of tracking? (certanly hope not, a nightmare internationally and one doesnt want his/her information shared).
In the end, the level of information gathered might not even matter the most.

Bill Booth mentioned (in a youtube video I believe) that most incidents at the moment are related to the people, not the equipment. Tracking all small incidents, and having an open discussion on those, will most likely uncover some knowledge gaps in jumpers or processes, and might fix those before they lead to more serious incidents. As such, the data itself to use in material improvements might actually be secondary to the effect it has to discuss irregularities and create a culture of learning and improving quickly. (Although I am not suggesting that isnt here already, improvements can always be made) The remark that tracking data will not lead to workable data for material improvements might therefore be correct, but it is not the entire picture. We want triggers to review / discuss safety mechanisms (training, processes, material, etc) and catch anything before it leads too serious incidents.

As a note, a lot of people tend to thing "blame" when you have a discussion about something that didnt go as planned: As any safety system, it is about the culture of openly being able to discuss things that did not go as planned and adjust accordingly, quickly. It can never be about blame. These situations are learning exercises. If someone does something dangerous, maybe they are lacking information / training. As such the training failed, not the person. In all the time I've been in my job, seeing a ton of minor and major things go wrong, including permanent disabilities and death, not once have I allowed any of the root causes to list "Operator error". When someone tried, I've sent them back or assisted them to investigate decently. No jumper is doing this with the intention to die.

Anyway, I know I will be tracking all of my jumps including any and all irregularity and discuss those with my mentors / peers to see what I can learn, as fast possible. Having an ego about this stuff is counterproductive.