Your calculations failed to take into account physical number of humans involved. When calculating the odds of a fatality... you must include how many humans. Right now all your numbers indicate are the odds of there being 'a' fatality... not the odds of 'a person' being involved in a fatality.

When calculating the odds of a fatality... you must include how many humans.

In the whole discussion above we are after only one number - the 'odds' for a specific person.A specific person Mr. John Doe, SS# 012 34 5678, is making today a choice of mods of transportation between two points A and B and seeking to minimize his chance of death. He doesn't know the driver of a car too well, nor he knows much about a pilotof this white-blue Piper Seminole that is offering him a ride. These are the numbers discussed throughout this thread. There is nothing wrong with mathematical methods used or numbers derived by mgh in view of this objective.We are not discussing or calculating the odds of a fatality, that someone in the US will die today in an aircraft crash or will be killed on the road. Totally different calculations and these would be based on number of people engaged in air or road travel today.

Seems to me that if ,indeed, we are talking about Mr Doe trying to get from point A to point B, then we need a better comparison than hours flown and miles driven. Yes, I know that pilots keep track of hours flown, but that does not have any correlation in any other mode of transport that I can think of. One can fly around in circles for an hour and land. One can drive around in circles for an hour and arrive back home. One can also fly or drive 50 miles from point A to point B. Seems to me that we need to be able to compare like events. Is Mr Doe more likely die driving from LA to New York or flying? Since the distance covered by an airplan(somewhat direct route) is far different than distance covered by a car(must go around mountians and lakes), I'm just not seeing a valid correlation.How many people die over some period of time(1 year, 10 years...)driving between LA and New York? How many flying? How many dogsledding? I think Mr Doe wil get more out of that than wondering why, as his plane is going into a mountian, his calculator says this shouldn't be happening.Bob

Mgh-I never disputed your figures. I did dispute comparing hours in an aircraft to miles in a car. For a correct compare they have to be converted to similar units-either hours or miles.As Ga aircraft-you are right-the figures I posted were the only ones I found on the web (http://www.hearldfamily.com/FlyingSafetyDivingvsFlying.xls) . I tend to agree that the figures are not correct because in further research:There seem to be no mileage stats for GA aircraft-and that reason is simple-Vfr flight. The commercials are all on ifr flight plans and can be documented. Vfr flight simply can't as it is...Vfr.Add to the fact that GA traffic is around 59% of the traffic load in the United States (more than commercial traffic). (data from the 2005 aopa fact card) and there is a huge undocumented area of flight.Flying 50,000 miles in a lifetime-I haven't heard of many pilot's that have accomplished this but perhaps there are a few. Most extremely high time pilots I know are at 25k to 30k and that is typically more from flying commercially and in the military. I would expect a typical GA pilot flies 50 hours a year if even that. It would be interesting to look up the average Ga pilot's logbook time-another project..... I would expect the average GA time to be much lower over a lifetime-perhaps as low as 1200-1500 hours. That changes the math and the odds per hour.Michal-I am glad you approve of this Aopa article. You left out though the conditions of the "stats" you approve of (I quote here as non aopa members will not be able to access):"GA accident rates have always been higher than those of the airlines, partly because of GA's versatility. Less regulation offers cost-effective access to thousands of smaller airports where airlines can't go. Also, the definition of GA includes aerial application, law enforcement flights, and banner towing-all relatively high-risk operations.To say that "flying is much safer than driving" is popular at cocktail parties, but the AOPA Air Safety Foundation says that it's difficult to draw such a simple comparison. Transportation profiles, measurements, and operating environments are significantly different." ..and the conclusion.. "If you exercise good common sense while flying, avoid pushing weather, and stay proficient, your chances of being involved in a fatal accident are pretty slim."e.g. take away accidents that a typical GA driver would not be involved in-e.g crop dusting, stunt flying, law enforcement flights, banner towing, suicides, buzzing houses, perhaps even fairly obvious stuff like the usual highest causes-running out of gas and flying vfr into ifr weather-the stats can be hugely reduced.Which brings it back to how you interpret the stats-and the gamut runs both extremes on the web from what I can see.In the link I posted from the Department of transportation:http://www.phmsa.dot.gov/portal/site/PHMSA...vgnextfmt=printThe fatal rate for motor vehicles is quoted as 1.3 deaths per 100 million vehicle miles, yet-the general population risk of dying in a car crash is listed as 1 out of 7,700 as a risk of death!The fatal rate for commercial airliners is listed as 1.9 deaths per 100 million aircraft miles (higher than autos), yet!- the general poplulation risk is listed as 1 out of 2,067,000!How can that be?! Simple-how many hours does a member of the general population normally fly a commercial airliner per year vs. drive a motor vehicle? Now these stats that seem cut in dry have new meaning-they can and need to be interpreted. The risk for the airline pilot may match the numbers-but for the passenger and typical person-a different story.How about this logic I found on another site talking about "stats":"Flying is much safer than driving. If 140 people want to take a cross-country trip of 3000 miles, they could do that in a single plane, or in 70 cars, each with two people. That trip would be only 3000 plane-miles in the plane, but it would be a whopping 210,000 car-miles by driving. So the exposure in traveling by car is 70x higher per person. So even though cars produce fewer deaths per mile, there are more total deaths because there are so many more miles driven."You can of course quibble with this-but there are definately angles to the "stats"..Looking on the web-it seems everyone can justify and interpret the stats in many ways.I'll again state that I think I'll side with the insurance companies-after all their job is to make money based on risk assesment. Right now they say I am safer in my airplane than in my car by a premium cost and coverage compare of both policies...

Since the distance covered by an airplan(somewhat direct route) is far different than distance covered by a car

Bob, let's not invent problems where there are none. The distances are not "far" different. On a typical long route in the mountainous western US you save about 20-25% of distance flying rather than driving. In the east it is far less. So yes, you can account for this small difference if you wish - it doesn't change results in any meaningful way.

Mgh-I never disputed your figures. I did dispute comparing hours in an aircraft to miles in a car. For a correct compare they have to be converted to similar units-either hours or miles.

I am speaking for Mgh here - he expertly handled the problem of converting hours to miles. It is all above - for anyone to read and understand. Apart from Poisson distribution this topic should be accessible to anyone with high school diploma.

To say that "flying is much safer than driving" is popular at cocktail parties, but the AOPA Air Safety Foundation says that it's difficult to draw such a simple comparison.

I am frankly quite disappointed in AOPA in the way they spin facts. All this controversy exists because flying on commercial carriers is indeed much safer than driving in a car and AOPA wants to piggy back on this fact their own agenda. They are incapable of admitting that what works in the Part 121 world has no bearing on Part 91. The numbers are strikingly different, over 25 times, so they shift blame on crop dusting, police flying, home buzzing, etc. without presenting any believable accounting of the breakdown of hours flown in GA. I was always much more comfortable in the way FLYING handles this area, actually pretty much every aviation related topic. I look at AOPA as a political-lobby group, similar to NRA.

Seems to me that if ,indeed, we are talking about Mr Doe trying to get from point A to point B, then we need a better comparison than hours flown and miles driven. Yes, I know that pilots keep track of hours flown, but that does not have any correlation in any other mode of transport that I can think of. One can fly around in circles for an hour and land. One can drive around in circles for an hour and arrive back home. One can also fly or drive 50 miles from point A to point B. Seems to me that we need to be able to compare like events. Is Mr Doe more likely die driving from LA to New York or flying? Since the distance covered by an airplan(somewhat direct route) is far different than distance covered by a car(must go around mountians and lakes), I'm just not seeing a valid correlation.How many people die over some period of time(1 year, 10 years...)driving between LA and New York? How many flying? How many dogsledding? I think Mr Doe wil get more out of that than wondering why, as his plane is going into a mountian, his calculator says this shouldn't be happening.Bob

The following calculation shows that the probability of having a fatal accident flying by GA from New York to Los Angeles is 0.02603% while the same probability by car is 0.00391%. That means, although the probabilities of either are low, the probability of having a fatal accident by GA is almost 7 times greater than that by road. ( Note that the probability of a fatal accident (%) = 100 - probabability of no fatal accident(%)).New York - San Francisco...................2500 miles GA GA Speed....................................125 mph GA Time......................................20 hrs GA fatal accident rate........................0.00027120 fatalities/20 hrs GA Probability of no fatal accident..........99.97288% GA Probability of fatal accident..............0.02712% Car Distance...................................2900 miles (Driving distance by MapQuest) Car fatal accident rate.......................0.00003914 fatalities/2900 miles Car Probability of no fatal accident.........99.99609%Car Probability of fatal accident.............0.00391%Factor 0.0271/0.00391 = 6.928

"I am frankly quite disappointed in AOPA in the way they spin facts. All this controversy exists because flying on commercial carriers is indeed much safer than driving in a car and AOPA wants to piggy back on this fact their own agenda. They are incapable of admitting that what works in the Part 121 world has no bearing on Part 91. The numbers are strikingly different, over 25 times, so they shift blame on crop dusting, police flying, home buzzing, etc. without presenting any believable accounting of the breakdown of hours flown in GA. I was always much more comfortable in the way FLYING handles this area, actually pretty much every aviation related topic. I look at AOPA as a political-lobby group, similar to NRA."I couldn't disagree more. While part of what AOPA does is political-its' safety part does a great service to all flyers, including breaking down accident data.The fact is they are not shifting "blame" to crop dusting, police flying, home buzzing but pointing out that a large number of fatal accidents occur here. If you do not belong to this group your odds go up.Quite similar to the car industry charging a premium on less than 22 year old male drivers. The stats simply support this. If you are not part of this group your odds are considerably better.Insurance companies certainly go beyond just the vanilla stats-why I put more credence on their take on the odds than the flat stats themselves.

I couldn't disagree more. While part of what AOPA does is political-its' safety part does a great service to all flyers, including breaking down accident data.The fact is they are not shifting "blame" to crop dusting, police flying, home buzzing but pointing out that a large number of fatal accidents occur here. If you do not belong to this group your odds go up.

The accident data does need to be broken up, and not lumped into one sum. As I've read this thread for the last few days, I keep thinking "whatever"... when I see hard numbers compared to hard numbers. If you're looking for a meaningful answer, these hard numbers don't mean a thing, except for a possible "scare" tactic to make one's point. I'm one of those who reads the NTSB (aviation) accident reports three times a week, as they are released; and every week. Because I'm well aware of the accident statistics and multiple causes, I see that attemting to compare one number to one number .......is what it is. Which is not much........ if you're looking for a helpful answer.L.Adamson

The following calculation shows that the probability of having a fatal accident flying by GA from New York to Los Angeles is 0.02603% while the same probability by car is 0.00391%. That means, although the probabilities of either are low, the probability of having a fatal accident by GA is almost 7 times greater than that by road. ( Note that the probability of a fatal accident (%) = 100 - probabability of no fatal accident(%)).New York - San Francisco...................2500 miles GA GA Speed....................................125 mph GA Time......................................20 hrs GA fatal accident rate........................0.00027120 fatalities/20 hrs GA Probability of no fatal accident..........99.97288% GA Probability of fatal accident..............0.02712% Car Distance...................................2900 miles (Driving distance by MapQuest) Car fatal accident rate.......................0.00003914 fatalities/2900 miles Car Probability of no fatal accident.........99.99609%Car Probability of fatal accident.............0.00391%Factor 0.0271/0.00391 = 6.928

Mgh-Your math looks correct. However, I believe there is a variable that becomes important in a single event like this to well raise the odds-the training level and capability of the pilot flying the flight.If your pilot is ifr trained and current your odds go up greatly as flying vfr into ifr conditions is one of the biggest accident producers. Will the trip be flown ifr or vfr? If your pilot is in the 200-400 range-he is statistically the most dangerous. Does the pilot do regular training? How many hours in the last 90 days has he/she flown? Has your pilot planned appropriate and conservative fuel stops for this trip (the other biggest accident producer), or is he/she maximizing legs and oblivious to headwinds. Does the aircraft have xm weather, stormscope, or radar, and if the weather goes sour is your pilot willing to land and give the flight up for a day-or will he/she keep pushing to get you there? Will your pilot be buzzing your house on the way out, and is he/she familiar with mountain flying conditions - density altitude , mountain currents he/she will face when he gets in the Western United States? Likewise, is your pilot current with crosswinds and the usually much higher gusty wind conditions he/she may face in the Western United States? Is your pilot familiar with the poor visibility , thunderstorm, and icing conditions he/she may face in the Eastern US? If there is forecast icing in the weather briefing that hopefully your pilot will get-will your pilot scrub the flight or decide to risk it? Will your ifr pilot be flying ifr at night (statistically much more dangerous) or only in the daytime. Will your pilot do a careful weight and balance and especially in the high density altitude airports of the west, consult performance charts. Will he choose to do all high density takeoffs only in early morning hours when the temperature is cool, and make sure that the loading of the airplane gives a comfortable margin?I am sure there are car drivers you would refuse to take this trip with also to increase your odds on the road. For instance, not driving with a 20 year old male driver would increase your odds greatly here. With aviation, the judgement and capability of the pilot becomes even more crucial-and is mostly the deal maker on the odds. The recent Colgan crash certainly shows this even in the realm of commercial flying.So if you jump in any plane with any pilot-your odds count for sure. If you jump in a well maintained plane, with great equipment, with a well trained and experienced pilot who most importantly has good judgement, you can change those odds completely. A detailed look at the Nab report shows this.

From another Aopa article: http://www.aopa.org/members/files/pilot/1997/accs9706.html.This article is very balanced..

The article states:GA accident rates have always been higher than those of the airlines, partly because of GA's versatility. Less regulation offers cost-effective access to thousands of smaller airports where airlines can't go. Also, the definition of GA includes aerial application, law enforcement flights, and banner towing-all relatively high-risk operations.Why doesn't it mention the obvious reason - that many GA pilots have had less training and less experince than airline pilots?You say "I would expect a typical GA pilot flies 50 hours a year if even that" and "I would expect the average GA time to be much lower over a lifetime-perhaps as low as 1200-1500 hours". An Airline Transport Pilot requires 1500 hrs even to get a licence. I suspect you are referring to private pilots here though. Given the range of activities and types of pilot encompassed by GA I can't help suspecting that, while some classes of GA pilot must have accident rate lower than the average for GA, other classes will have a higher rate.

The article states:GA accident rates have always been higher than those of the airlines, partly because of GA's versatility. Less regulation offers cost-effective access to thousands of smaller airports where airlines can't go. Also, the definition of GA includes aerial application, law enforcement flights, and banner towing-all relatively high-risk operations.Why doesn't it mention the obvious reason - that many GA pilots have had less training and less experince than airline pilots?You say "I would expect a typical GA pilot flies 50 hours a year if even that" and "I would expect the average GA time to be much lower over a lifetime-perhaps as low as 1200-1500 hours". An Airline Transport Pilot requires 1500 hrs even to get a licence. I suspect you are referring to private pilots here though. Given the range of activities and types of pilot encompassed by GA I can't help suspecting that, while some classes of GA pilot must have accident rate lower than the average for GA, other classes will have a higher rate.

All in the nab report. A pilot with less training is not necessarily a danger. It is when he/she exceeds their capabilities that the danger occurs. It seems also by reading accident reports that increased ratings don't always equate to good judgement. Flying attracts its share of personalities like all walks of life. To contradict that however-insurance companies look very kindly on increased ratings...It really is very interesting going thru the stats in the nab report in great detail. Also the individual reports on individual aircraft also available from Aopa. What one finds is mechanical is usually incredibly small as a reason for an accident. The C172 is very safe-the pilot is the variable. Airlines have minimized this by automating, requiring at least 2, and higher standards of training for sure.Larry mentioned a few accident pilots he knew. In my case of flying 20 years I have never known anyone in my realm of flying who had a fatality, and only one who had an accident (I just had a well known flight instructor I knew die in a car crash though). The person who got me into flying-a WwII carrier pilot/korean war jet pilot, and one of those high time 30,000 hour pilots had his first "incident" in his career at age 75 on an instructional flight when the engine quit on short final in a Piper. In a masterful show of airmanship he landed the plane along a highway in a small grassy area with very little damage to the aircraft and none to the pilots.I have known thru the grapevine a number of fatalities though. One was when I was training-a plane stalled and spun in at my local airport. Concerned I asked my instructor about it. He said the guy was known as the "airshow pilot". He would always takeoff in a steep nose high attitude close to the stall. That day, in addition he loaded a bunch of logs into the baggage compartment-over rear c.g-didn't bother to do a weight and balance. Did his airshow stunt on takeoff and paid. During my training period, there was another crash-killed a father and son-very sad. The pilot had a layer of ice on the wing-didn't bother to remove it before takeoff.Then there was the guy at our local field that had Alzheimers...had long lost his medical. Decided one day his plane needed an annual, got in it-took off -forgot what he was doing and decided to land in a bunch of trees.I met a pilot at a social event. He proceeded to brag to me about all the incidents he had been in. One-a flight where he did not preflight-took off-and the engine quit requiring a forced landing. Turns out kids had drained the gas from his plane for their car during the night and he had no gas. He then told me about taking off once vfr-the ceiling getting so low that he had to fly low above the roads and read the road signs to find the airport.I listened and then thought-here comes a statistic...The reason pilot's study accident reports is to try to put oneself into the same situation, and then ask what they would have done differently. I will openly admit it could be a rationalization-or perhaps it is not. Sometimes you read a report and say this guys luck just ran out and there was nothing he could do about it. But a majority of the time, you ask yourself-what was this pilot thinking?It is hard to believe, but runnng out of gas, and flying vfr into ifr conditions when not rated are still two of the big ones. If you are one not likely to do that you have just changed the odds considerably.By the way-if a member of Aopa-they have a great re-creation of a vfr pilot flying into ifr conditions with the actual atc and entire flight recreated on fsx. Watching it is amazing-not only to wonder why the pilot did what he did-but a great use of fsx!http://flash.aopa.org/asf/acs_vfrimc/

It is hard to believe, but running out of gas, and flying vfr into ifr conditions when not rated are still two of the big ones. If you are one not likely to do that you have just changed the odds considerably.

But if you are making such 'adjustments' for pilots you should also make similar adjustments for drivers. How much your odds are improved if you don't speed, don't drink before driving, keep distances from other cars, don't attempt to run red lights or generally engage in something which is called "defensive" driving technique? There is only one point that potentially looks like a mitigating factor for GA - while flying you may be much more in control of your own destiny than while driving since driving still involves being in close proximity to other vehicles, sometimes driven by really bad drivers. But how does this translate into numbers?

But if you are making such 'adjustments' for pilots you should also make similar adjustments for drivers. How much your odds are improved if you don't speed, don't drink before driving, keep distances from other cars, don't attempt to run red lights or generally engage in something which is called "defensive" driving technique? There is only one point that potentially looks like a mitigating factor for GA - while flying you may be much more in control of your own destiny than while driving since driving still involves being in close proximity to other vehicles, sometimes driven by really bad drivers. But how does this translate into numbers?

I agree totally and I personally do make adjustments when driving. I won't drive with a single drink, and I refuse to drive with my mother in law-lol. So my odds should be a little better there too. Despite that, I was run into in a stopped position by a drunk driver going 50 miles an hour, leaving virtually nothing left of the car but me and the seat I was in which is why...I also agree with your second point-personal responsibility is much more a factor in flying. In driving unfortunately it is often trusting that the other person is as responsible as you are. If people in general drove with the same seriousness of purpose that we fly with the auto accident rate would also go down substantially. I never quite understand that one who flys a C150 going 95 miles an hour must have extensive training, an exam (bfr) every two years, a medical, currency experience but one who drives an auto 70 mph literally feet from another vehicle passes a test once, has no recurrent training, can drink , and even can text/talk on a cell phone at the same time they drive-or read a book (I see that all the time where I live).

The following calculation shows that the probability of having a fatal accident flying by GA from New York to Los Angeles is 0.02603% while the same probability by car is 0.00391%. That means, although the probabilities of either are low, the probability of having a fatal accident by GA is almost 7 times greater than that by road. ( Note that the probability of a fatal accident (%) = 100 - probabability of no fatal accident(%)).New York - San Francisco...................2500 miles GA GA Speed....................................125 mph GA Time......................................20 hrs GA fatal accident rate........................0.00027120 fatalities/20 hrs GA Probability of no fatal accident..........99.97288% GA Probability of fatal accident..............0.02712% Car Distance...................................2900 miles (Driving distance by MapQuest) Car fatal accident rate.......................0.00003914 fatalities/2900 miles Car Probability of no fatal accident.........99.99609%Car Probability of fatal accident.............0.00391%Factor 0.0271/0.00391 = 6.928

So in this case, since you listed GA time and speed, I must assume that those numbers are factored in. If that is true, why are car time and speed not listed? I would think that car time would be far greater than GA time and therefore risk would go up for the driver because he is in potential danger longer.If GA time and speed are not a factor, then I don't understand the point of listing them.Thanks for answering. I only have a college degree, and am trying to follow this.Bob

If that is true, why are car time and speed not listed? I would think that car time would be far greater than GA time and therefore risk would go up for the driver because he is in potential danger longer.

The driving time is obviously much longer than flying time but highway statistics are delivered in terms of driving distance, not driving time whereas typical GA statistics are delivered in hours flown. This has always been like this - distance is a more natural yardstick for driving while time is more natural yardstick for flying. Mgh above makes this adjustment by assuming an arbitrary speed of 125 Mph (Cessna 172?). You could put say 200 mph if you happen to fly a turbo Cirrus and have a good tail wind and the result would be slightly different. One way or the other you have to translate distance into time if you want to use official NTSB GA statistics in this comparison. It should be obvious why driving speed is not needed here.