Test-Driving Preliminary Results of Statistical Analysis

[James Burleigh]

It seems that every police-investigated California traffic accident finds its way into the Statewide Integrated Traffic Records System (SWITRS). Two days ago I received from the very helpful Support Services Section of the California Highway Patrol (CHP) a disk I had requested only three days earlier containing all SWITRS data related to statewide motorcycle accidents over the last 10 years. The disc contains 110,528 motorcycle accidents and covers the years 1997 to available 2007.

The database comprises three data files ralated through a Case ID number: Collision, Party, and Victim. I have just completed analyzing the Collision file and wanted to share (test drive) my preliminary results. My plan is to complete my analysis, write up the results, and publish them in a motorcycle magazine. (BTW, if there are any LEOs out there familiar with the fields in the SWITRS, let me know 'cause some of them are downright mysterious.)

Here then are some preliminary observations related to M.C. traffic accidents:

• In the last 10 years motorcycle accidents have increased 1.5 times since the base year; however, M.C. fatalities have doubled.
  
• Accidents per 100,000 licensed riders have remained fairly consistent; however, deaths per 100,000 licensed riders have increased 1.5 times over the base year.
  
• A third of M.C. accidents occur on weekends (34%).
  
• The least amount of accidents occur on Mondays, with accidents increasing each day of the week, and leveling off at the highest levels on weekend days.
  
• One-third of accidents occur in a rural area.
  
• One-fifth (22%) of all M.C. accidents occur at an intersection.
  
• The overwhelming majority (95%) of M.C. accidents occur on dry roads that are in good condition (surfaces that have loose material, are slippery, have holes or ruts, or are under construction combined account for only 5% of all reported M.C. accidents).
  
• Weather is rarely a factor, with 88% of accidents occurring on clear days; 10% on cloudy days.
  
• For all accidents, M.Cyclists are at fault in 55% of accidents; cars and pickups in 32% of accidents.
  
• For accidents where the M.Cyclist is at fault, speed is given as the violation in 45% of accidents; improper turning in 17%; DUI in 9%.
  
• The M.Cyclist was determined to be at fault in 73% of accidents that involved a motorcyclist fatality.
  
• In accidents where the M.Cyclist is at fault, 40% involved hitting another vehicle (down 57% for all at-faults); 31% involved just dumping the bike ("non-collision"); and one-quarter involved hitting a fixed object (ouch).

Of course the interesting part of this kind of analysis is "what does it mean for us riders." Well, one thing seems clear: riders are making decisions that contribute to their accidents.

Jb

 

[Murph]

Very interesting info.

[*]For all accidents, M.Cyclists are at fault in 55% of accidents; cars and pickups in 32% of accidents.

What accounts for the other 13% ?

 

[James Burleigh]

Very interesting info.

[*]For all accidents, M.Cyclists are at fault in 55% of accidents; cars and pickups in 32% of accidents.

What accounts for the other 13% ?

Pickup or Panel Truck with Trailer
Truck or Truck Tractor
Truck or Truck Tractor with Trailer
Schoolbus
Other Bus
Emergency Vehicle
Highway Construction Equipment
Bicycle
Other Vehicle
Pedestrian
Moped
or blank - Not Stated
Blank
"Blank" was 6%; everything else was less than 1%.

 

[Duff]

JB,

As a MC Safety rep for my unit, I am VERY interested in the data you have. In fact, I think this thread will be important enough for me to figure out the "subscription" thingy on here.

If you don't mind, I would like to be able to use some of your findings to get people's attention when I brief. I'll give you more time to go through them before i start asking a bunch of questions.

Thanks for doing this!

 

[Jazzzzz]

Do you have any data available that would allow you to correlate accidents involving severe injury/death of the rider to type of safety equipment worn? I know helmet use is mandatory in California, but when my wife and I stopped at Neptune's Rest while driving up the coast during a visit a couple of years ago, it seemed that a LOT of the riders were wearing $15 half helmets that probably weren't even DOT certified. Most of them were in T-shirts and jeans, too.

 

[2WheelTiger]

Based on what I've read elsewhere, you might want to gather some fatality data on rider age, experience level, and motorcycle type (if any of this is available on the disc).

My guess is that the fatality numbers will suggest that older riders with relatively little experience are over-represented, and in fact, constitute the majority of the percentage increase in fatalities. I believe you will also see a high correlation between fatality rates and riders on cruisers.

(This could bring the summation that older newbies on cruisers are at great risk.)

Basically what I'm suggesting is that the squid death rate has not increased significantly as a function of accidents per 100K.

Data is a wonderful thing. You can generally find whatever you want if you look really hard.....

 

[Taff]

Based on what I've read elsewhere, you might want to gather some fatality data on rider age, experience level, and motorcycle type (if any of this is available on the disc).
My guess is that the fatality numbers will suggest that older riders with relatively little experience are over-represented, and in fact, constitute the majority of the percentage increase in fatalities. I believe you will also see a high correlation between fatality rates and riders on cruisers.

(This could bring the summation that older newbies on cruisers are at great risk.)

Basically what I'm suggesting is that the squid death rate has not increased significantly as a function of accidents per 100K.

Data is a wonderful thing. You can generally find whatever you want if you look really hard.....

+1 on your comments Andre!

Have you taken a look at the stats here in our State in 2006?

 

[2WheelTiger]

+1 on your comments Andre!
Have you taken a look at the stats here in our State in 2006?

No, I haven't. I'm just basing my thoughts on other analyses I've read.

 

[Kevin Daly]

JB,

interesting stats you have generated. I know you probably dont have all the needed data... It would be interesting to explore descritions a bit more deeply. For example you say 5% of accidents occur in construction sites/dirty conditions. I think it would be more interesting if placed within the context of what percent of the total road is under construction. So for example if road construction occurs on less than 1% of the road at any time on average, this would mean that for every mile traveled on torn up roads you are far more likely to dump. Same with dry streets. Placed into the context of a per capita value (accedents per mile traveled on wet vs dry) think the values would be strikingly different.

My point is really one of statistical interaction. That is a drunk riding at night in the rain hits a construction site and lets hope he has gear and the construction guys parked the front loader way off the road.

 

[odot]

Weather isn't a factor because noone rides in the rain.....proved by the stat that most accidents happen on dry pavement. If pups weren't such fair-weather riders then that stat would even out a bit.

 

[Kevin Daly]

Weather isn't a factor because noone rides in the rain.....proved by the stat that most accidents happen on dry pavement. If pups weren't such fair-weather riders then that stat would even out a bit.

TO be more accurate, very few people ride in the rain. But you hit the nail on the head the stats would even out a bit. however, you can statistically correct for differences in amount of time people ride in rain versus dry and make a stat like: accidents per mile (or hour) in rain vs sun. That would be more telling.

Statistical interactions, that is where the meat meets the road!

 

[charismaticmegafauna]

Here then are some preliminary observations related to M.C. traffic accidents:

• In the last 10 years motorcycle accidents have increased 1.5 times since the base year; however, M.C. fatalities have doubled.
  
• Accidents per 100,000 licensed riders have remained fairly consistent; however, deaths per 100,000 licensed riders have increased 1.5 times over the base year.

Of course the interesting part of this kind of analysis is "what does it mean for us riders." Well, one thing seems clear: riders are making decisions that contribute to their accidents.

Thanks jb,

This info correlates with what I've been reading about the Federal Govt. wanting to take funds from Motorcycle Safety and put them into mandatory helmet legislation. Apparently, whatever decisions riders are making are incorrect...?

'Andre': My guess is that the fatality numbers will suggest that older riders with relatively little experience are over-represented, and in fact, constitute the majority of the percentage increase in fatalities. I believe you will also see a high correlation between fatality rates and riders on cruisers.(This could bring the summation that older newbies on cruisers are at great risk.)

From my observations, I tend to agree. Too bad that motorcycling has been infused, for the past 10 years (or so), by a sub-group who's popularity has been inversely proportional to their concern for safety (or, even basic riding skills & techniques).

Stats can get all 'tilted' that way.....

 

[Pjturbo]

I know this is going to get me into trouble. IMO If we want to reduce MC accidents don't let people ride motorcycles until they KNOW how to. i.e. Tiered license and TOUGH testing.

OK pile on.

Phil

 

[Jazzzzz]

I know this is going to get me into trouble. IMO If we want to reduce MC accidents don't let people ride motorcycles until they KNOW how to. i.e. Tiered license and TOUGH testing.
OK pile on.

Phil

I'm inclined to agree with you, but this line of thinking runs afoul of the commonly held notion that a driver's license is a right, not a privilege, and by extension motorcycle endorsements are too. It doesn't help that public transportation in most cities and suburbs of the US is pretty sub-par, so you HAVE to have a vehicle to get around. I typically don't like the notion of enacting expensive legislation as a means to keep idiots from killing themselves, but I would vote for a tiered licensing system for both motorcycles and four-wheeled vehicles. It's my opinion that it's entirely too easy to get a license to operate a motor vehicle of any sort here in the States.

 

[Bounce]

I know this is going to get me into trouble. IMO If we want to reduce MC accidents don't let people ride motorcycles until they KNOW how to. i.e. Tiered license and TOUGH testing.

only if they do it to cage drivers first!

 

[Jetmaker737]

Good stuff... indicates that we motorcyclists are in control of our destiny to a large extent. And I would say even when the cager was tagged at fault some of those accidents probably could be avoided with better skills and observation (on the part of the rider).

 

[mikerider]

James, thanks for the info.

Could you share the raw data? If not, how do I get a copy of the disk myself?

Thanks.

-mike-

 

[mikerider]

It seems that every police-investigated California traffic accident finds its way into the Statewide Integrated Traffic Records System (SWITRS). Two days ago I received from the very helpful Support Services Section of the California Highway Patrol (CHP) a disk I had requested only three days earlier containing all SWITRS data related to statewide motorcycle accidents over the last 10 years. The disc contains 110,528 motorcycle accidents and covers the years 1997 to available 2007.
The database comprises three data files ralated through a Case ID number: Collision, Party, and Victim. I have just completed analyzing the Collision file and wanted to share (test drive) my preliminary results. My plan is to complete my analysis, write up the results, and publish them in a motorcycle magazine. (BTW, if there are any LEOs out there familiar with the fields in the SWITRS, let me know 'cause some of them are downright mysterious.)

Here then are some preliminary observations related to M.C. traffic accidents:

• In the last 10 years motorcycle accidents have increased 1.5 times since the base year; however, M.C. fatalities have doubled.
• Accidents per 100,000 licensed riders have remained fairly consistent; however, deaths per 100,000 licensed riders have increased 1.5 times over the base year.
• A third of M.C. accidents occur on weekends (34%).

James, thanks for the info. Do you have the actual numbers you could post?

As the % didn't really tell us much.

Thanks!

-mike-

 

[James Burleigh]

James, thanks for the info. Do you have the actual numbers you could post?As the % didn't really tell us much.

Thanks!

-mike-

Hi, Mike. Here you go.

* In the last 10 years motorcycle accidents have increased from 8,501 in 1997 to 12,350 in 2006; M.C. fatalities have doubled from 221 to 433 (more than one a day on average).

* Accidents per 100,000 licensed riders have remained fairly consistent at about 1,100; however, deaths per 100,000 licensed riders have increased from 27 to 39.

* 37,790 of a total of 110,526 M.C. accidents occurred on a weekend.

Please PM me regarding the data set, and we can sort that out.

Jb