Talk:Aviation safety/Archive 3

Different risk ratios for different people
Note, that the risk is the same for every passenger of the plane, but is completely different for every person using a car. For example a young 18 y.o. man returning from the party on the Saturday night has different risk, than a 40 years old man going to the office on Wednesday morning. So there is a need to show the various risk ratios for various types of people. The risk of driving a car would decrease dramatically for a "normal" person.

The same is with bikes and motors. Also the risk of swimming is different for a 18 years old boy after alcohol, than for a 35 years old woman playing with her child in the water near the beach with safeguards. This issue should be raised in the article at least as a point to consider. — Preceding unsigned comment added by Ultimaratioregum (talk • contribs) 15:02, 4 August 2013 (UTC)

Human Factors source text
Editors may find this free e-book, publshed by NASA, a useful source text: Martinevans123 (talk) 15:10, 9 August 2013 (UTC)

Accidents and incidents: Statistics
In the same vein, when considering the "deaths per billion journeys" statistic, it is important to consider that airliners, buses, and trains will carry far more passengers than a car or bicycle for example. This is a ridiculous statement. If there are 300 people on the plane, then 300 journeys occur. What is missing is the definition of journey as used for these statistics. 75.247.135.166 (talk) 17:13, 26 August 2013 (UTC)

In that vein, I corrected the Space Shuttle statistics to be normalized for person journeys rather than vehicle journeys (which reduces the apparent fatality rate by 7x).DavidSJ (talk) 09:07, 5 May 2014 (UTC)


 * Flight safety statistics follow the definitions of aviation accidents and incidents which refer to the aircraft. Statistic of casualties refer to death and injury of souls on board. "Journey" in fatality statistics is defined by known common endpoint.-Yohananw (talk) 23:55, 24 August 2020 (UTC)

Asiana Flight 214, lives saved because of past improvements?
There's an interesting one-page article in the Oct. 2013 issue of Popular Mechanics, Tech Watch, Aviation Safety, "How Airplanes Save Lives," Ben Iannotta, page 22.

Asiana 214 crashlanded, apparently with a fire breaking out shortly after landing and yet. . . all but three of 307 persons survived. A tragedy for these three persons and their families of course. A miracle for those who survived. Of course we also have to ask what kind of serious injuries.

The Popular Mechanics article attributes the surviving passengers primarily to three improvements:

1) seats that stay attached able to withstand 16 g,

2) floor lighting in the event of smoke and reduced visibility, and

3) improvements in less flammable materials for the cabin interior.


 * Now, I tend to work slow. That's both a plus and a minus.  I do think this potentially would make a good addition to our article.  If anyone else wants to jump in and run with this, please, be my guest.  Cool Nerd (talk) 18:05, 10 September 2013 (UTC)

Disputing accuracy of section Aviation_safety
The official ICOA safety report lists 173 fatalities and 32.1 million departures in 2013. This corresponds to 5.39 fatalities per one million departures. On the other hand, the statistics shown in Aviation_safety claims that there are only 0.117 fatalities per one million departures, which differs by a factor of 50 from the official numbers. The reference given in Aviation_safety is an obscure website rather than an official report or a scientific study. The issue that this is not a good source has come up several times before. Unless someone can explain to me the discrepancy of a factor 50, I will delete the subsection (and perhaps try to find a better source). ylloh (talk) 11:31, 24 July 2014 (UTC)


 * The correct statistic that anyone cares about is ( number of fatalities ) / ( number of user-events ). This is not the same as ( number of fatalities ) / ( number of flight departures ). It IS the same as ( number of fatalities ) / ( sum of number of passengers on every flight departure ). From the source (2014 ICOA safety report with 2013 statistics), "Scheduled commercial international and domestic operations accounted for approximately 3.1 billion passengers in 2013", and "173 fatalities in 2013". This equates to 0.056 fatalities per 1 million passenger departures. 2013 was a relatively good year for aviation safety, 2008-2012 each has more fatalities, 2014 will as well with the two high profile crashes so far. So the number checks out.


 * If anything, the problem is with the language on the page as it is. "per departure" is deceiving when each departure tends to carry hundreds of people. I was concerned the page did not do the math correctly, but I can confirm here with your source that they did. Thanks for providing that and I will remove the accuracy is disputed tag. Perhaps it needs a citation needed. I haven't looked at any other sources than this one to get a ballpark estimate. 2604:2000:E830:4D00:244A:942F:EA4D:9B2D (talk) 04:30, 25 November 2014 (UTC)


 * Thanks, this settles my concerns. ylloh (talk) 10:45, 7 April 2015 (UTC)

Bias in comment about statistic metrics Aviation_safety
Essentially, the faster you travel, the less important fatalities/kilometer becomes. I don't want to argue what statistic is ultimately the best, but neither should the article. If statistics need explaining, explain them. But don't weight one over the other. Wikipedia exists to provide facts, not opinions. — Preceding unsigned comment added by 80.171.100.0 (talk) 20:29, 24 March 2015 (UTC)
 * You removed 1,123 bytes of text here with the edit summaries: "Statistics description was subjective and possibly misleading. Let the statistics speak for themselves" and "The statistics are very clear. The description only served to put weight on the one that favors airlines when in reality it isn't really better than the others." Would you care to suggest an alternative way of explaining the limitations in the comparisons that the table provides? Martinevans123 (talk) 20:35, 24 March 2015 (UTC)
 * The things that are explained are not limitations. The statistics simply measure different things and the explanation right now clearly favors "fatalities/kilometer" over the other statistics, because it's focused on distance, but "fatalities/kilometer" is the only one that takes distance into account. — Preceding unsigned comment added by 80.171.100.0 (talk) 20:46, 24 March 2015 (UTC)
 * The limitation of the comparisons is that, as you say, the "statistics simply measure different things". I don't see how offering no explanation at all is better than offering some. Martinevans123 (talk) 20:54, 24 March 2015 (UTC) p.s. try to remember to sign your posts using four tildes like this: ~
 * Yes, the statistics show different things and the explanation could say that, even though it's obvious from the titles. But right now it only states that "distance" is the best way to measure transportation safety, which is not an explanation, but a subjective statement. It then says that the first two stats are insufficient to show the safety per distance - but the stats don't even try to show that, so why point it out negatively? Frankly, this is quite obvious and if this argument doesn't come to a close, I'll mark this for administrator attention, because I get the feeling that you are supporting a certain agenda here. 80.171.100.0 (talk) 21:11, 24 March 2015 (UTC)
 * I have no agenda. By all means "mark this for administrator attention" if you wish, but it's normal to wait for the views from more than just two editors. I've suggested that you might like to offer improvements to the explanation, but you seem reluctant to do this. Martinevans123 (talk) 21:22, 24 March 2015 (UTC) p.s. thanks for signing.
 * And you seem reluctant to tell us how exactly it's an explanation at all. It isn't. Like I said, the statistics themselves clearly state what they show and the text doesn't make this any clearer. It only serves to persuade the reader that "fatalities/distance" is the best way of measurement and air planes are super safe. 80.171.100.0 (talk) 21:36, 24 March 2015 (UTC)
 * Why do you think that trips, or hours, is a better metric than distance? Martinevans123 (talk) 22:21, 24 March 2015 (UTC)
 * What I think is the best metric is of no importance. The article should be neutral unless there is clear evidence that one is better than the others, which is not the case. 80.171.100.0 (talk) 22:37, 24 March 2015 (UTC)
 * They each have their pros and cons. The table presents all three. I agree the passage would benefit from sources. But I don't see that it distorts the statistics in any way or is expressing a biased point of view. I think the table benefits from a discussion of its metrics. That's quite standard practice. Perhaps a link to List of countries by traffic-related death rate would be useful? Martinevans123 (talk) 22:43, 24 March 2015 (UTC)
 * http://en.wikipedia.org/wiki/Wikipedia:Neutral_point_of_view Read this, especially "Avoid stating opinions as facts". An article is not the place for an opinionated discussion. Comparisons are fine, but saying one is better, especially without any kind of reasoning or evidence, is not. 80.171.100.0 (talk) 23:12, 24 March 2015 (UTC)
 * You think I haven't ever read that? If you can offer a less "biased" commentary on the shortcomings of each of the three metrics, by all means suggest one. Or better still - wait for some more opinions. Martinevans123 (talk) 23:16, 24 March 2015 (UTC)
 * I showed that the article violates Wikipedia guidelines, so until there are more opinions or a better version, the paragraphs in question should be removed. 80.171.100.0 (talk) 23:24, 24 March 2015 (UTC)
 * I'd agree that the paragraphs need sources. You seem to also have a problem with neutrality which I don't fully understand and don't accept. I think you should wait for at least a third opinion before you remove them for a third time. Martinevans123 (talk) 23:29, 24 March 2015 (UTC)

The previous section above (Statistics)explores this issue. As it very cogently points out, in an extreme example a journey in space across light years, would (factually) make the safety per mile stats utterly meaninglesss. Similarly, if to a lesser degree, the same applies to safety per mile comparisons on Earth. The fact that insurance companies use deaths per journey rather than deaths per mile says it all. They are experts in this field. From the deaths per journey it would appear that one is roughly three times more likely to die on any plane journey than on one's car journey to the airport. That is very significant. As someone who has studied statistics I am acutely aware that their relative significance is very often not obvious, and thus some sort of explanation or commentary is of great value. The current comments seem fine and objective to me.Cassandrathesceptic (talk) 17:23, 27 March 2015 (UTC)

Pilot's stress and pressure is also one of important factors of Aviation Safety
I believe pilot's stress and pressure is also one of important factors of Aviation Safety and I can add this section to this page. I have some websites that describes different types of Pilot's stress and how they handle those stress and pressure in order to increase the safety of their flight.

Dhur2 (talk) 03:37, 12 October 2015 (UTC)DHUR2

Airport design section is not related to survivability
Just passing through here, and I noticed this. The paragraph on airport design has to do with accident avoidance and aviation safety in general, but survivability assesses the ability of an average person to survive an accident, not to avoid one. Availability and response time for fire and rescue is a part of survivability, but that is not a major element in airport design, other than services should have easy access to runways. (Communication and systematic response procedures are more important than airport design per se.) What g-forces were imparted onto the passengers? How toxic was the atmosphere inside the cabin? Was there adequate oxygen inflight? To what extent did a fire penetrate the passenger cabin before evacuation was possible? Was access to exits obstructed? Did anything hinder effective emergency response? Were there any failures of systems or structures within the cabin (such as seats, seatbelts, supplemental oxygen, lighting, poor cabin design, etc.)? How effective were the actions of the cabin crew? These are survivability aspects. They assume an accident has already happened. How survivable was that accident, and what factors made it more or less survivable than predicted? Dcs002 (talk) 20:22, 9 December 2015 (UTC)

Statistically biased comparison on death between car and air travel per bjn
The 117 deaths per bjn of air travel include chartered flights, private and light aircraft, and airlines that are less capable in terms of fleet maintenance and pilot training. Indeed, the numbers of deaths by car also included these figures, but in a context for customers (or general Wikipedia readers) to choose the means of travelling, the correct comparison should be between cars and scheduled flights using jet airliners, operated by reasonably sophisticated airlines. This will significantly change the conclusion on deaths per billion journeys. 14.203.71.4 (talk) 20:25, 12 March 2017 (UTC)

Transport comparison
I added skydiving in the list, and I tried to add a "per journey" measurement for paragliding. I practice paragliding myself, and I think it is pretty hard to time how long a paragliding flight usually is, because it ranges for 5 minutes for beginners to several hours for seasoned practitioners. I assumed 15 minutes per flight. If someone finds a better source, I will be delighted.-- 10 August 2018 Chmduquesne
 * I removed it as WP:OR. Mixing two different sources for a result like you did for skydiving is pretty close to WP:OR too.--Marc Lacoste (talk) 12:12, 10 August 2018 (UTC)
 * I did not know about WP:OR, but I think that under this rule your deletion is fair. Now that you said that, I don't know what to think of the other statistic about paragliding deaths. The formula says 11/(27615*410)*1000000000 deaths per billion of hours. Reading the report quoted as a source, it is not obvious at all that it does not fall under WP:OR too. Here are were the figures come from: 11 is the number of deaths in one year, 410 is the average reported hours of experience the pilots who had accidents. Note that this encompasses non deadly accidents. 27615 is the total number of paragliding practitioners in France. So the formula seems to assume that the pilot who died had the same experience as the pilot who did not. Plus 410 hours would be the expected number of hours provided you have an accident, but we are trying to find out the number of deaths provided you fly one billion hours. Those are two different expected values. --Chmduquesne (talk) 09:08, 20 August 2018 (UTC)
 * No that's just the arithmetic mean, OK per WP:CALC. Honestly I don't really care about removing it, I updated someone addition with limited references with a WP:RS but it's not as nice as having a RS compiling data itself.--Marc Lacoste (talk) 11:59, 20 August 2018 (UTC)