The Perception of Risk

[From Bruce Gregory (2000.0914.1124)]

The recent discussion of teachers as controllers (may it rest in peace)
suggest certain analogs to the perception of risk. People willingly
take great risks if they have the illusion of control. For example
driving is much riskier than flying, but drivers have the illusion that
they can control their fates while driving but they know they cannot
control their fates while in a commercial airplane. The risk of being
hijacked is minute, but we willingly spend hundreds of millions of
dollars each year and endure delays and inspection of our bodies and
belongings in order to restore some sense of being in control (albeit
indirectly) of our fates.

I heard on NPR yesterday that a survey of world class athletes in 1995
(no other details given) asked them if they would take a drug that would
insure them a gold medal in the Olympics if they new it would be fatal
five years later. Fifty percent said they would take the drug. Chocolate
or vanilla is a horrible choice if you think you might have strawberry.

BG

[From Bill Powers (2000.09.14.1129 MDT)]

Bruce Gregory (2000.0914.1124)--

I should probably leave this alone, but ...

People willingly
take great risks if they have the illusion of control. For example
driving is much riskier than flying, but drivers have the illusion that
they can control their fates while driving but they know they cannot
control their fates while in a commercial airplane.

So the idea that you're in control while driving a car is an illusion? I
would say a person has a great deal more control of his fate while driving
a car than while riding in an airplane. All you have to do is relax and
stop moving the car's wheel to counteract disturbances to see that this is
true. The risk of a serious or fatal accident skyrockets if you stop
controlling a car. On the other hand, in an airplane it doesn't matter
whether you're trying to control it or not; you simply have no influence on
the risk.

I wonder, incidentally, how the risk figures come out if you measure not in
terms of passenger-miles, but in terms of passenger trips. Airplane trips
tend to be much longer in distance than car trips. Once a trip starts, it
doesn't matter how many miles you're going; you're committed for the whole
trip. I have actually been on far more airplanes experiencing mechanical
difficulties than in cars having mechanical difficulties, including my very
first commercial plane trip during which one of the two engines quit
shortly after takeoff. How many cars have you been in when the engine just
suddenly stopped working? And in how many cars would losing the engine most
probably cause a fatal crash?

I quit flying small aircraft a long time ago, when my sense of risk became
unbearable. Perhaps it was having the carcase of a flying club airplane in
my carport for six months, after a member had a near-fatal accident trying
to land in a non-existent crosswind. Perhaps it was having to pull up
during a takeoff to hop over a plane that blindly taxied across the active
runway. Perhaps it was having all four spark plugs foul on takeoff in a
Stinson Station Wagon due to idling too long with an over-rich mixture, and
having to limp sputtering and popping around the airport boundary (a forest
preserve) to just make it back to a landing. Perhaps it was getting back to
the airport a little too long after sunset, with no lights. Perhaps it was
just wondering if the engine sounded really OK, while peering around
through Chicagoland murk, in vain, to pick a spot where I might put down
without killing myself. "Stretches of serenity and joy punctuated by
moments of stark terror," that's flying. And I got in only about 100 hours
in the air.

On the other hand, I have driven cars hundreds of thousands of miles and
many thousands of hours, with only a few dents and scrapes to show for it
(on the car, not me), and obviously no fatalities. There have been close
calls, but they would have been a lot closer if I hadn't been controlling
the car pretty well, and most potential close calls were avoided long
before they became serious.

Those are the "statistics" I think of when comparing flying and driving a
car. They don't jibe with what the airlines and flying afficianados come up
with.

Best,

Bill P.

[From Bruce Gregory (2000.0914.1514)]

Bill Powers (2000.09.14.1129 MDT)

Bruce Gregory (2000.0914.1124)--

I should probably leave this alone, but ...

I often have that feeling.

All you have to do is
relax and
stop moving the car's wheel to counteract disturbances to see
that this is
true. The risk of a serious or fatal accident skyrockets if you stop
controlling a car.

Definitely. Most people do not drink and drive. Yet it has been
estimated that 50% of all accidents involve alcohol. If the drinking
driver does not hit a fixed object, he most likely hits another car--one
driven by a sober driver who is completely in control of his car. My
wife drives back roads to and from school. The accident rates on these
roads are much higher than the accident rates on the highways I drive.
I'm sure the drivers on the back roads are at least as much in control
of their cars as I am in control of mine.

On the other hand, in an airplane it doesn't matter
whether you're trying to control it or not; you simply have
no influence on
the risk.

Thank goodness!

I wonder, incidentally, how the risk figures come out if you
measure not in
terms of passenger-miles, but in terms of passenger trips.

18 Year (1982-1999) Average Fatal Accident Rate for Major Airlines:

.0432 accidents per 100,000 departures

1 accident for every 2,300,000 departures

.0000000004% odds of being in an accident each time you fly

I quit flying small aircraft a long time ago, when my sense
of risk became
unbearable.

Death Rate Per 100,000 Hours

- Commuters to Major Airlines: 10.8 to 1

- General Aviation to Major Airlines: 55.1 to 1

- General Aviation to Commuters: 5.1 to 1

Looks like your sense of risk is right on target. I'd also stay away
from commuter airlines, if I were you.

On the other hand, I have driven cars hundreds of thousands
of miles and
many thousands of hours, with only a few dents and scrapes to
show for it
(on the car, not me), and obviously no fatalities. There have
been close
calls, but they would have been a lot closer if I hadn't been
controlling
the car pretty well, and most potential close calls were avoided long
before they became serious.

Type of Passenger Transport Death Rate per billion passenger miles

Passenger Cars .89

Intercity Buses .03

Transit Buses .01

Trains .02

Airplanes .01

Those are the "statistics" I think of when comparing flying
and driving a
car. They don't jibe with what the airlines and flying
afficianados come up
with.

Well, you know the difference between casting nets.....

BG

[From Bill Powers (2000.09.14.1914 MDT)]

Bruce Gregory (2000.0914.1514)--

18 Year (1982-1999) Average Fatal Accident Rate for Major Airlines:

.0432 accidents per 100,000 departures

1 accident for every 2,300,000 departures

.0000000004% odds of being in an accident each time you fly

Comforting figures. I wonder what the comparable statistics would be for
car trips. Just a guess: 60,000,000 families, 4 car trips per day per
family, average of one passenger per trip, 240,000,000 trips per day, say
40,000 fatal accidents per year, comes out to 109 per day or 1 fatal
accident per 2,201,834 trips (departures). I'm just dredging up figures
from nowhere, so this could be way off. Note that each airline accident is
likely to kill 30 to 300 people (though that doesn't affect the
individual's risk of being in a fatal accident).

Type of Passenger Transport Death Rate per billion passenger miles

Passenger Cars .89

Intercity Buses .03

Transit Buses .01

Trains .02

Airplanes .01

This is where the difference between passenger-miles and trips becomes
important. The average flight distance is probably 100 times the average
car trip distance, so we're talking almost equal numbers in terms of trips,
even though it's 89:1 in terms of passenger-miles. Also, passenger-miles
multiplies the distance traveled by the number of passengers; a single trip
of 1000 miles might count as 30,000 to 300,000 passenger-miles on a
commercial airline. So the accident rate seems a lot lower than it is, by a
factor or 30 to 300. One accident per 1000-mile trip (almost every trip!)
would come out as one accident per 30,000 to 300,000 passenger miles, which
sounds much safer.

This works for the other entries, too. Intercity busses have 0.03 accidents
per billion passenger-miles, but if you convert that even to bus-miles,
you have to multiply by the bus capacity of about 30 passengerz, so it
comes out to 0.90 per billion bus-miles, about the same as for cars which
carry only about 1 passenger each.

This is what comes from letting the interested parties make up their own
ways of presenting accident statistics.

Best,

Bill P.

[From Bruce Gregory (2000.0915.0449)]

Bill Powers (2000.09.14.1914 MDT)]

This is what comes from letting the interested parties make up their own
ways of presenting accident statistics.

Indeed. Or, worse yet, make up statistics.

BG

from Ray B (2000.09.15.2230 CST Aust)
Its a known fact that 45% of statistics are made up anyway.

Bruce Gregory wrote:

ยทยทยท

[From Bruce Gregory (2000.0915.0449)]

Bill Powers (2000.09.14.1914 MDT)]
>
>
> This is what comes from letting the interested parties make up their own
> ways of presenting accident statistics.

Indeed. Or, worse yet, make up statistics.

BG

[From Bill Powers (2000.09.15.)]

Bruce Gregory (2000.0915.0449)--

This is what comes from letting the interested parties make up their own
ways of presenting accident statistics.

Indeed. Or, worse yet, make up statistics.

I admitted that I was making up estimates that could be far off. I was
rather hoping that you might have better data with which to correct them.
Your comment seems to imply that I was trying to present my estimates as if
they were real statistics. Is that what you think I was doing?

Best,

Bill P.

[From Bruce Gregory (2000.0915.1219)]

Bill Powers (2000.09.15.)

Your comment seems to imply that I was trying to present my
estimates as if
they were real statistics. Is that what you think I was doing?

It was a general observation on CSGnet: first the model and then the
data. Which would not be so bad were it not for the fact that we rarely
seem to get to the second step.

BG

[From Bruce Gregory (2000.0915.1232)]

Ray B (2000.09.15.2230 CST Aust)

Its a known fact that 45% of statistics are made up anyway.

Oscar Wilde said something the effect that it is one's duty to
accurately describe that which never happened. Needless to say, I
heartily agree.

BG