Safest Seat on a Plane, Or How to Survive a Crash 454
Ant writes "Popular Mechanics shares a short article on an exclusive look at 36 years' worth of National Transportation Safety Board (NTSB) reports and seating charts to determine the best way to live through a disaster in the sky. Move to the back of the Airbus."
BBC already did this... (Score:5, Informative)
The best place is "near an exit door".
Statistically, most crashes are survivable if you can get out. The biggest impediment to getting out is the number of other people between you and the door. The ones who don't get out die of smoke/fire.
Excuse me... (Score:5, Informative)
So, they are working off of a sample size of twenty??? Not sure if I would draw too many conclusions from this dataset.
Re:What are the odds? (Score:2, Informative)
Re:Anti-EU much ? (Score:4, Informative)
"Move to the back of the bus." is a common phrase in America.
Re:Not on Oceanic 815... (Score:3, Informative)
And, strictly speaking, he wasn't a tail section passenger. His assigned seat was midsection, he just happened to be in a tailsection bathroom when the excitement started. (Although IIRC he had time to strap himself into a seat which ended up in a tree.)
(I just got Season 2 from the library and watched the whole thing in about three days. Probably fried some neural circuits.
Bad Statistics (Score:2, Informative)
To quote the article, In 11 of the 20 crashes, rear passengers clearly fared better. What kind of statistical analysis is this? If they are going to publish something with as broad a claim about an enormous industry as airlines, they need to back their claims up. That means t-tests, multiple ANOVA, chi-square tests, etc.
But most importantly, that means stating their p-values! With a sample size as low as 20, I wouldn't be surprised if the statistical significance of this data is null (that is, the likelihood of being wrong is greater than the likelihood of being right).
Until they publish that number, I will take this study with a grain of salt.
Re:What are the odds? (Score:2, Informative)
1. the plane has to slow down dramatically to allow the jumpers a safe exit and even then when you're sitting in the door it feels like a hurricane is blowing outside - even from a small single prop aircraft
2. the plane must be at a safe altitude for jumping
3. there are very few jet airliners that it is considered possible to make a safe exit from - from conversations with some old skydiving buddies I recall that a 707 is one of them - with most airliners you are most likely to get sucked back into the rear of the aircraft - and that would be very terminal
So to summarize - jumping from an airliner is something that an expert skydiver would think twice about - and so its definitely not an option for your average passenger.
Re:Sit in the rear (Score:3, Informative)
Its not just a theory, it did happen this way for flight 901 in 1979. There were no survivors, all 257 passengers and crew died in the initial impact (with Mount Erebus) and fireball.
Re:What are the odds? (Score:5, Informative)
Jumping from 37,000 feet and hundreds of MPH requires training and equipment. At that altitude the ambient air temperature is -70 fahrenheit. If the average terminal velocity of a person skydiving is 250 ft/s then you'll take about 2m30s to get to the ground without a 'chute. At 250 ft/s the wind chill is really, really significant. You've then got a choice to make (any perhaps the airline would instruct you about the best action): open the 'chute immediately after exiting the plane or wait until you are nearer the ground.
Opening the parachute early means you are certain to hit the ground slowly but maximized your exposure to very low temperatures and low oxygen with all those inherent injuries.
Opening the parachute later means more wind chill and possibly more tissue damage. Your betting that you'll be conscious to pull the rip cord. You also have much less time to perform an maneuvering to get to a "good" landing spot.
That said, given the choice of almost certain death on a severely disabled airliner or possible death by parachute I'd probably choose the parachute.
Re:What are the odds? (Score:2, Informative)
Here is the source of the "1 in 5051" figure cited by the GP.
http://www.nsc.org/lrs/statinfo/odds.htm [nsc.org]
The methodology is also explained on that page. (Note, the NSC has many other interesting statistics and reports on this and related topics.)
Basically, the number of airplane crash deaths in the US was divided by the entire population of the US in the year of the study (2003). The data was presented in two forms, annual odds of dying a particular way and lifetime odds of dying a particular way. This means that all of the following discussion is directly relevant only to someone living (and/or dying!) in the US.
The airplane crash numbers were 1 in 391,981 (annual odds) and 1 in 5051 (lifetime odds). This means that the "1 in 5051" figure is the odds of a given person that died having died in a plane crash.
The odds of a person who died in a given year having died in a plane crash are 1 in 391,981. These numbers are NOT directly translatable into an individual's odds of dying each time they get onto an airplane. For that, you would have to know how many flights over US territory there are in a given year and how many plane crashes occur in that same time (since the odds of dying are roughly equal to the odds of a plane crash).
For an exact calculation, you'd need to know how many people flew on those flights (the aggregate would be ok), and how many people died in crashes (again, the aggregate is ok). From that, you could determine the odds of dying on any given plane flight.
The FAA also has some interesting data. The target safety rate for the U.S. is 0.010 fatal accidents per 100,000 departures (appears to include all flights, commercial and private, even though the statistic is called the "Commercial Airline Fatal Accident Rate"), though the current rate in 2007 is 0.023 fatal accidents per 100,000 departures.
http://www.faa.gov/data_statistics/accident_incide nt/ [faa.gov]
See the "Airline Fatal Accident Rate" PDF on the linked page.
For the below data, FAA/NTSB reports were used. Much more data is available at these sites for anyone who wants to do more analysis. For example, the commercial data below is a summary of Part 121, Part 135, and On-demand Part 135. The accident rates were much higher for the "On-demand Part 135" which not what we typically fly as commercial passengers.
Also: http://www.ntsb.gov/Publictn/A_Stat.htm [ntsb.gov] has annual summarized reports.
The data for 2003 is:
Commercial Air Carriers:
Background data in 2003 (rounded to nearest whole million/billion):
639 million passengers boarded commercial airplanes
8 billion miles were flown
11 million departures
23 million flight hours
Accidents:
Total: 130
Fatal: 21
Deaths: 66
Fatal accidents per departure: 1.9091x10^(-6) (1 in 523,810)
Fatal accidents per hour: 9.130x1-^(-7) (1 in 1,095,239)
General Aviation:
Total Accidents: 1739
Fatal Accidents: 352
Injuries:
Fatal: 632
Serious: 324
Minor: 523
Involved but Not Injured: 1697
Or an immortal jellyfish (Score:2, Informative)
"Turritopsis nutricula is a hydrozoan (jellyfish) with a life cycle in which it reverts back to the polyp stage after becoming sexually mature."
Re:I'll go for 56% (Score:2, Informative)