Transcript Measuring the Danger of Driving Motorcycles and the

Courtesy of Roadracing World Magazine, Greg Fryer during practice at Daytona. Photo by Rick Menatace
Is Greg a safe motorcycle driver?
2012 SMSA Conference - Nashville, TN
Measuring the Danger of Driving Motorcycles
and the Effectiveness of
Motorcycle Safety Programs
in the USA
Joseph Elliott
Executive Director
National Motorcycle Training Institute
Courtesy of
DeMatha HS
Yearbook
Teammate Vince (left) experiencing a collision
Apparently I was interested in dangerous and
enjoyable activities from a young age. I was cocaptain of my High School Rugby Team.
Fully Funded Graduate
Research in Nuclear
Physics at Oregon State
University.
Studied Electrical
Engineering at
University of MD.
Honors project: Microwave
Communications Test Apparatus
This is the team I worked with at Oak Ridge National Laboratory. We
used the Heavy Ion Collider to create exotic unstable isotopes to study.
Brotherhood Raceway, Terminal Island,
Los Angeles. I am on the left.
Found my way to University of Southern California in Los
Angeles in 1990 with fully funded research in Photonics.
Also studying and practicing motorcycle driving, and
motorcycle training with the “Doug Fitts” CMSP (and
Hoot Gibson).
A nonprofit IRS 501(c)3 Public Benefit Institute
with an independent Board of Directors
“We do not promote motorcycling, we promote making informed choices!”
Our mission is to reduce the fatality
rate and morbidity (disablinginjury) rate for motorcyclists.
Mission Specifics:
• Properly define and measure appropriate fatality and morbidity rates.
• Encourage behaviors that decrease these rates; discourage behaviors that increase
these rates.
• Provide and maintain motorcycle rider education curriculum that is open and is
free to copy, distribute, and use.
• Maintain objectivity through careful monitoring of funding, minimizing and
eliminating conflicts of interests of our funding sources with our mission. Our
Institute can receive Tax Deductible Bequests, Transfers, and Gifts. All donations
are tax deductible. The Board of Directors of NMCTI diligently enforces the IRS
Code 501(c)(3) for Charitable Organizations.
NMCTI uses
Scientific Method
Francis Bacon
Karl Popper
A scientific hypothesis cannot be proved, only disproved.
Example: “A motorcyclist can’t live forever.”
This is not scientific because it cannot be disproved. There
are many statements that are reasonable to say even if they
are not scientific.
Of course there are unreasonable and unscientific statements too.
Truthiness:
The quality of preferring facts one wishes to be
true, rather than facts known to be true.
Truth:
The quality of being factual.
Truthiness cultivates confirmation bias, the seeking
out of information that confirms the statement or
belief and discounting information that conflicts with
the statement or belief.
Truth accepts both information that confirms and
conflicts with the statement or belief.
Motorcycle Truthiness Example:
“Motorcyclists who ride a lot get into less crashes.”
This may be something we wish to be true and it has
good truthy feel to it.
Will this statement cultivate confirmation bias, the
seeking out of information that confirms the
statement or belief and the discounting of information
that conflicts with the statement or belief?
Who comes to mind for you?
A Scientist would make the following chart:
Then count and compare all the boxes. Not just box d.
Paris Street; Rainy Day, Gustave Caillebotte
We must also be careful with correlation and
causality.
Example: It is correlated that people carry
umbrellas more often when it is raining. Umbrella
carrying is correlated with rain.
Does umbrella carrying cause rain?
Today’s Goal: Properly Define and measure
appropriate Motorcycle Fatality Rates and relate
these rates to the danger of driving a motorcycle.
Images provided by HB
We use the Simplified definitions:
Risk - The chance of Loss
Danger - The chance of Harm
For motorcycling, we are using
the word “danger” this way
Danger - Chance of Bodily Harm
Morbid Injury - Gruesome or grisly injury
Fatality - the ultimate
morbid injury, the
ultimate bodily harm.
Fatality - the ultimate
morbid injury, the
ultimate bodily harm.
We use the fatality data to estimate and
model the danger.
A very brief intro to
Probability and Statistics
Probability is often used as the
“chance of an event occurring”
Statistics is not the same as
probability. We will use
statistics to create a model
for probability.
In science, we often use statistics to
Estimate or MODEL probability.
For example, I flip 100 coins and
observed that 53 coins came up
heads, 47 came up tails.
Scientists would say the statistics
observed from this experiment show
heads came up 53% of the time.
We use the statistics to estimate the coin flip in this
case: Initial Model of Probability Heads: 53% chance
The statistics gave us an estimation, a
Model of Probability: Heads 53% of the time.
Of course with more experimentation,
and independent verification from
others, our model would approach the
50% probability of a coin flip.
The probability model gets “sharpened” with
independent experimental repetition and
statistical analyses.
Modeling Danger, the chance of harm,
related to Motorcycles:
We use the fatality rate that include all
fatalities occurring in collisions that involve
at least one motorcycle to model the
danger to the population.
We use the fatality rate that include just
motorcycle drivers to model the
danger of driving motorcycles.
VS
2001-2010 USA
2001-2010 USA
Comparing the fatality rates due to
motorcycles with those due to
passenger vehicles can quantify the
relative danger.
VS
1991-2000 USA
2001-2010 USA
Comparing fatality rates for the same
locality in different years can quantify
changes in danger.
Gathering the Data
FARS
Fatality Analysis Reporting System
FARS is managed by
National Highway and
Traffic Safety Administration
Our thanks to Lorenzo Daniels of NHTSA
for providing the FARS data used in this
presentation.
I ask Mr. Daniels to extract the following categories
of fatalities. In addition to motorcycles, I asked Mr.
Daniels for identical data for Passenger Vehicles.
• Fatalities in motor vehicle crashes involving
at least one motorcycle. (ALL)
• Fatalities of motorcycle drivers. (MCD)
• Fatalities of licensed motorcycle drivers. (LMD)
• Fatalities of motorcycle occupants. (OCC)
Making Rates:
We need to divide the number of annual
fatalities with a count of a meaningful group
so we can make useful comparisons.
Some common categories used:
 Population (POP)
 Number Registrations issued (REG)
 Vehicle Miles Traveled (VMT)
Scientists like to Collect and Compare,
then reduce to manageable pieces.
How can we usefully combined
the available data?
The following chart makes
scientists happy.
Combining the Available
Fatality and Group Data
ALL
OCC
MCD
LMD
Box 1
Box 2
Box 3
Box 4
POP
Box 5
Box 6
Box 7
Box 8
REG
Box 9
Box 10
Box 11
Box 12
VMT
We will look at Box 1 and Box 11 today.
Box 1 = All/POP
We choose to use Box 1, fatalities in motor vehicle
crashes involving at least one motorcycle divided by
census population, to model the danger of
motorcycles to the population.
Box 11 = MCD/VMT
We choose to use Box 11, motorcycle driver fatalities
divided by Vehicle Miles Traveled, to model the
danger to the driver.
Looking closer at Box 1 = ALL/POP
ALL = All fatalities in motor
vehicle crashes involving at least
one motorcycle
Divided By
POP = census population
Using Population makes
comparison of big
States with small States
possible.
vs.
1990s
2000s
Also we can compare a particular
State’s danger with itself as it grows
or shrinks in population over time.
Applying these ideas to our subject:
Compare the relative danger:
What is the relative danger between motorcycles and
passenger vehicles for the USA Population?
We choose to use Box 1 =All/Pop fatality
rates, and compare.
What is the relative danger between motorcycles and passenger
vehicles for the USA Population?
Compare the motorcycle and passenger vehicle,
Box 1 =All/Pop, fatality rates.
1990s
9
145
>>
2000s
16
134
>>
% Change +74% -7%
>>
Units = Fatalities per Million Population
Passenger Vehicles kill many more
people than Motorcycles
in the 1990s and in the 2000s.
However Passenger Vehicle
danger decreased in the 2000s.
Motorcycle danger
increased in the 2000s.
A different look at the same data.
Comparing the Danger, Passenger Vehicles to Motorcycles, USA
Fatalities / Million Population
160
140
120
100
Passenger
Vehicles less
dangerous
80
60
40
20
0
1991 - 2010
Motorcycles
more
dangerous
Motorcycles got more dangerous
Passenger Vehicles got less dangerous
1990s
9
145
>>
2000s
16
134
>>
% Change +74% -7%
>>
Units = Fatalities per Million Population
Next: What is the relative danger to the Driver?
Box 11 = Driver Fatalities per
100 Million Vehicle Miles Traveled
1990s
21
0.9
>>
2000s
27
0.8
>>
% Change +28% -11%
>>
Units = Driver Fatalities per 100 Million Miles Traveled
Important! What if you don’t trust VMT?
You can you check and estimate the VMT
in your state!
The first step is to prepare
yourself for accepting the
enormous number of
passenger vehicle drivers
on the road.
Then you can determine the ratio of passenger vehicles to
motorcycles. This ratio would approximate the VMT
ratio. Warning, motorcycles are less than 1%.
Just start by trying to count the number of
passenger vehicles you see when driving
to work or school or the store, etc.
Note that it is better to have a passenger do the counting.
Comparing Motorcycle to
Passenger Vehicle Rates of Box 11
=MCD/VMT shows:
Driving a Motorcycle is much more dangerous
than driving a passenger vehicle.
Driving a Motorcycle is 34 times more
dangerous than driving a passenger vehicle.
34 times more dangerous is difficult
to understand.
Courtesy of Steve and HB.
Small Passenger Vehicle and Large Motorcycle.
Who is being attended to?
Common Follow-up Question: Are the
Motorcycle Drivers getting killed licensed?
Motorcycle Driver Fatalities as Percentage Licensed
90
Percentage Licensed
80
70
60
50
40
30
20
10
0
1991 - 2010
77% killed had a valid motorcycle license
Another look at the licensed/not licensed fatality data
Motorcycle Drivers Killed, USA
5000
4500
4000
3500
3000
2500
Licensed
2000
1500
1000
Not
Licensed
500
0
1991 - 2010
Quiz Question:
Comparing a sober motorcycle driver and a drunk
car driver, which situation is more dangerous?
Hint: Multiple answers, must answer “Dangerous to whom?” first.
Dangerous to the driver or to all?
All the statistics are also available by State.
We can use fatality statistics to
model the danger motorcycles pose
to the population by US State.
Comparison between 1990s and 2000s
of the danger of motorcycles to population for US States
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Mass…tts
Michigan
Minnesota
Mississippi
Missouri
1991-2000
2001-2010
% Change
7.9
9.5
15.0
8.5
9.0
13.4
11.4
10.8
12.1
7.6
14.3
12.3
9.2
10.6
10.8
8.9
7.8
7.8
13.0
7.4
5.5
7.1
7.8
5.6
7.0
15.8
11.6
18.8
22.4
11.6
17.2
13.4
14.0
24.1
13.7
16.8
16.5
10.8
15.2
16.3
14.0
18.4
18.1
14.2
12.5
7.9
10.4
10.7
13.7
13.7
100
22
25
164
29
28
18
30
99
80
17
34
17
43
51
57
136
132
9
69
44
46
37
145
96
National Average
Copyright 2012 National Motorcycle Training Institute
Prepared for 2012 SMSA Conference
1991-2000
2001-2010
% Change
Montana
Nebraska
Nevada
New Hamp.
New Jersey
New Mexico
New York
North Carol.
North Dak.
Ohio
Oklahoma
Oregon
Penn.
Rhode Is.
South Carol.
South Dak.
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West VA
Wisconsin
Wyoming
16.5
4.4
11.8
15.1
5.4
16.1
6.0
10.1
7.2
11.0
9.3
8.3
8.9
8.2
15.3
16.8
10.6
7.8
9.9
10.4
5.6
7.0
10.4
11.3
16.2
26.2
8.9
17.0
18.7
8.4
20.3
8.6
15.4
10.6
14.1
19.0
11.1
14.9
11.3
22.3
27.0
18.9
15.6
10.4
13.1
9.3
10.5
17.3
16.0
33.7
59
102
44
24
56
26
43
52
47
28
104
34
67
38
46
61
78
100
5
26
66
50
6
42
108
9
14.2
58%
Units = Fatalities per Million Population
Top Ten relatively most dangerous
1990s 2000s %Change
Wyoming
16.2 33.7 108
South Dak. 16.8 27.0
61
Montana
16.5 26.2
59
Florida
12.1 24.1
99
Arkansas
8.5 22.4 164
South Carol. 15.3 22.3
46
New Mexico 16.1 20.3
26
Oklahoma
9.3 19.0 104
Tennessee 10.6 18.9
78
Arizona
15.0 18.8
25
Units = Fatalities per Million Population
Top Ten relatively least dangerous
1990s 2000s %Change
Massachus. 5.5
7.9
44
New Jersey 5.4
8.4
56
New York
6.0
8.6
43
Nebraska
4.4
8.9 102
Virginia
5.6
9.3
66
Michigan
7.1
10.4
46
Utah
9.9
10.4
5
Washington 7.0
10.5
50
North Dak. 7.2
10.6
47
Minnesota 7.8
10.7
37
Units = Fatalities per Million Population
Next, model the effect of
motorcycle programs in States.
Measure the change in danger from the
1990s to the 2000s within each state.
Ten states with the biggest increase in danger:
1990s
Arkansas
8.5
Mississippi
5.6
Kentucky
7.8
Louisiana
7.8
Wyoming
16.2
Oklahoma
9.3
Nebraska
4.4
Texas
7.8
Alabama
7.9
Florida
12.1
2000s
22.4
13.7
18.4
18.1
33.7
19.0
8.9
15.6
15.8
24.1
%Change
164
145
136
132
108
104
102
100
100
99
Units = Fatalities per Million Population
Ten states with the smallest increase in danger:
1990s 2000s %Change
Utah
9.9 10.4
5
Maine
13.0 14.2
9
Hawaii
14.3 16.8 17
Illinois
9.2 10.8 17
Connecticut 11.4 13.4 18
Alaska
9.5 11.6 22
New Hamp. 15.1 18.7 24
Arizona
15.0 18.8 25
New Mexico 16.1 20.3 26
Vermont
10.4 13.1 26
Units = Fatalities per Million Population
Quiz Question:
Motorcycle Safety:
A. is about making the population safer.
B. is about making the motorcycle driver safer.
C. is an oxymoron.
NMCTI uses
Scientific Method
A scientific hypothesis cannot be proved, only disproved.
“Motorcycling can be safe” has been disproved.
Old Paradigm:
Motorcycling can be Safe and Enjoyable
New Paradigm:
Motorcycling is Dangerous,
and can be Enjoyable
Motorcycling is Dangerous, and can be Enjoyable
Start using this new paradigm and
begin reducing the danger of driving
motorcycles in your State.
Start reducing the danger to the population.
Start reducing the danger to the drivers.
2012 SMSA Conference - Nashville, TN
Measuring the Danger of Driving Motorcycles
and the Effectiveness of
Motorcycle Safety Programs
in the USA
Joseph Elliott
Executive Director
National Motorcycle Training Institute
Our mission is to reduce the fatality
rate and morbidity (disablinginjury) rate for motorcyclists.
Mission Specifics:
• Properly define and measure appropriate fatality and morbidity rates.
• Encourage behaviors that decrease these rates; discourage behaviors that increase
these rates.
• Provide and maintain motorcycle rider education curriculum that is open and is
free to copy, distribute, and use.
• Maintain objectivity through careful monitoring of funding, minimizing and
eliminating conflicts of interests of our funding sources with our mission. Our
Institute can receive Tax Deductible Bequests, Transfers, and Gifts. All donations
are tax deductible. The Board of Directors of NMCTI diligently enforces the IRS
Code 501(c)(3) for Charitable Organizations.