Bay Area Motorcycle Crashes 2013-2017

DataDan

Mama says he's bona fide
If you just happened into this thread, spare yourself the tedious beginning and jump right to the conclusions here.


This thread will analyze in considerable depth motorcycle crashes in the Bay Area 2013-2017. It will include police-reported crashes of all severities (non-injury to fatal), and I will eventually present data on factors such as type of road, time and day of the week, other vehicles involved, pre-crash movements, fault as assigned by law enforcement, alcohol involvement, injury severity, and more.

I haven't done one of these tables-and-graphs threads in a long time, and I hope the local focus will be of interest. I chose a 5-year time period to strengthen conclusions about less-frequent occurrences, and for now, 2017 is the most recent complete data available.

If you have a suggestion for a topic you'd like to see, I'll try to include it if the data is available (cell phone usage is not).


First up is the trend in crashes and crash rate since 2001 (earliest detailed data I have):

attachment.php


The crash rate has been fairly flat at around 16 crashes per 1000 registrations for 15 years, except for the drop during the recession. Though the source I use for registrations by county hasn't been updated in two years, I would guess that, based on a slight statewide increase, the rate continues flat through 2017. It appears that the perceived increase in motorcycle crash risk due to increasingly careless drivers has not been severe.


Next is the breakdown by county. Pretty much what you'd expect.

attachment.php



Finally, a look at single- and multiple-vehicle crashes and rider injury severity.

[table=head]rider injury|single vehicle|multiple vehicle|all crashes|ped & cyclist
no injury|398|2,875|3,273|
minor injury|2,723|7,674|10,397|
major injury|725|1,370|2,095|
fatal|91|227|318|
total|3,937|12,146|16,083|240
||||
lethality|2.3%|1.9%|2.0%|

[/table]

The "ped & cyclist" column shows motorcycle crashes in which a pedestrian or cyclist was involved. "Lethality" is fatalities as a percentage of crash-involved riders. The difference between single- and multiple-vehicle crashes in that respect is not surprising; it is seen also in statewide and US data.

And because I wanted to do a pie chart, here's the injury severity distribution in that format:

attachment.php



The fine print:

Data for all crashes is from CHP's SWITRS database (Statewide Integrated Traffic Records System). Unfortunately, it is down for maintenance with no estimate of return, so no link for now. Fatal crash data is from NHTSA's FARS database (Fatality Analysis Reporting System).
 
Last edited:

GAJ

Well-known member
Interesting, I'd have guessed "lethality" would be higher.

For cars it is apparently 0.7%, so the bike rate is 3 to 4 times higher.
 

motomania2007

TC/MSF/CMSP/ Instructor
Is the data pretty similar Statewide?

Is there any "fault" indication from the data you are working with?

Obviously, the single vehicle crashes are almost 100% rider's fault. What is the percentage of multi-vehicle collisions that are rider's fault?
 

DataDan

Mama says he's bona fide
Interesting, I'd have guessed "lethality" would be higher.

For cars it is apparently 0.7%, so the bike rate is 3 to 4 times higher.
Bay Area motorcycle crash lethality is low. Statewide is about 3.0%, so some other regions in the state are much higher. US average is about 4% but includes non-helmet-law states that push it higher. BTW, I revised the lethality table in my OP after discovering an error.

My recent car crash data shows lethality of 0.16%, so the state motorcycle rate is about 20 times that. BTW, that car rate is for drivers only, since I didn't account for passenger exposure (1.0 drivers per car, but how many passengers on average? 0.5? 1.0? :dunno)
 

DataDan

Mama says he's bona fide
Is the data pretty similar Statewide?

Is there any "fault" indication from the data you are working with?

Obviously, the single vehicle crashes are almost 100% rider's fault. What is the percentage of multi-vehicle collisions that are rider's fault?
There's quite a bit of variation in motorcycle crash data around the state. Lethality, mentioned in my reply to Geoff, is one example.

Yes, the CHP database includes party deemed by law enforcement to be at fault. About 10% of the single-vehicle motorcycle crashes do not report the rider at fault. Could be a ped or cyclist or road conditions, I suppose. The multiple-vehicle motorcycle crashes are 41% rider at fault. I'll get into that, including pre-crash movement and crash type, in a later post.
 

DataDan

Mama says he's bona fide
Time and place

Around two-thirds of Bay Area motorcycle crashes occur during commute hours, Monday to Friday, 6:00am to 8:00pm:

attachment.php




In warm weather, twice as many crashes happen per month as in cold weather. But the commuting percentage is somewhat higher in winter--less recreational riding, I suppose.

attachment.php



Freeways are, well, freeways, including bridges (there's some fuzziness here because I had only roadway identifier to make the determination). A city street is a non-freeway in an incorporated city--handled by PD or Sheriff (city contracted). Unincorporated comprises non-freeway crashes handled by CHP--backroads, rural communities, and a few pockets near cities. It includes 17, BTW--not a freeway, mostly not in an incorporated city.

Roughly half of crashes happen on city streets, one-third on freeways, and less than one-quarter in unincorporated areas.

attachment.php
 
Last edited:

John999R

New member
Regarding the 10% of single-vehicle motorcycle crashes not reporting fault could be the fact the accident in under investigation for a longer period due to a possible lack of "witnesses". Therefore, fault has to be determined by physical evidence or lack thereof. I would venture to say most of these crashes occur in rural areas while canyon carving.

Good stuff BTW, thanks for the effort.
 

DataDan

Mama says he's bona fide
Regarding the 10% of single-vehicle motorcycle crashes not reporting fault could be the fact the accident in under investigation for a longer period due to a possible lack of "witnesses". Therefore, fault has to be determined by physical evidence or lack thereof. I would venture to say most of these crashes occur in rural areas while canyon carving.

There were 430 single-vehicle crashes with motorcyclist not found at fault. In a quick scan I didn't see many of the usual backroads. Half were on city streets, one-quarter on freeways.

I broke them down a little further... In 335 (78%) the motorcycle had been going straight before the crash. In 124 (29%), the type of crash was "hit object", and in 128 (30%) it was "overturned", often due to overbraking I expect.

In single-vehicle crashes where the motorcyclist was found at-fault on 9, 84, Mines, Redwood, and a few others I looked at, the primary collision factor is usually "unsafe speed" or "improper turning".

Good stuff BTW, thanks for the effort.
Glad you like it. More to come...
 
Last edited:

DataDan

Mama says he's bona fide
Crashes by city

I was kinda surprised at the number of city-street crashes in the third graph of my post #6 (up 3 ^). I had expected to see a more even distribution between city and rural, so I wondered: which cities have the most motorcycle crashes?

Mystery solved. I may later do an in-depth analysis of San Francisco crashes alone to get a better idea of the shit y'all have to deal with.

attachment.php
 

DataDan

Mama says he's bona fide
Age

While the aging riding population has been a big motorcycling story for the past 20 years, it actually began 60 years ago. The Baby Boom generation (born 1946-1964) started a Motorcycle Boom in the 1960s that drove US registrations from a half-million in 1959, when the Honda Super Cub was introduced to America, to five million in 1975. But the ardor cooled when family, career, and community came to dominate their lives. Then in the 1990s, as nests emptied and disposable income grew, the Motorcycle Boom echoed. Over ten years from 1998 to 2008, Baby Boomers rediscovered motorcycles, and registered bikes in the US doubled from under four million to nearly eight million as median owner age increased from 38 to 48. At the same time, some aging riders become aging crashers, though median crash age remained well below median rider age.

Now, Boomers are aging out of the sport as younger riders take it up. While I don't have ownership demographics for the state, evidence is seen in crash data. In California and the Bay median crash age has declined from a peak at 38 in 2010 to 34 statewide and 35 for the Bay in this 2013-2017 data:

attachment.php



An interesting relationship seen in the graph above is that older riders are a higher percentage of fatalities than they are of those who have crashed. This reflects greater crash lethality among older riders. In the event of a crash, a rider age 65+ is more than twice as likely to die as a rider < 25. This isn't a fluke of thin data in this limited analysis. It is also seen statewide, in US data, and in other states where I have found enough information to make the calculation.

attachment.php
 
Last edited:

DataDan

Mama says he's bona fide
Two-vehicle city crashes

As we have seen in this thread, 45% of Bay Area motorcycle crashes 2013-2017 occurred on city streets, that is, within incorporated city limits but not on a freeway. Of those 7000+ incidents, 80% were multiple-vehicle. To get a good idea how other drivers can ruin your day--and, with luck, prevent that from happening--this post will analyze 4200 two-vehicle, non-freeway crashes within city limits, between a motorcycle and a car, SUV, van, truck, or bus (collectively referred to as "car" for brevity).

I have not considered 3-vehicle (or more) crashes because they are a small percentage, and because it's hard to figure out who did what to whom. I excluded from the 2-vehicle crashes those that involved two motorcycles and those that also involved a parked vehicle, bicyclist or pedestrian (as shown earlier in the thread, there were few crashes between a motorcycle and a ped or cyclist). Narrowing further, I eliminated rare one- or two-of-a-kind crashes. Backing car vs. parking motorcycle and wrong-way motorcycle vs. right-turning car crashes don't hold practical lessons for us, yet would require of me a separate worksheet and paragraph.

What's left, 72% of multiple-vehicle Bay Area city street crashes 2013-2017, is, I believe, a representative set of collisions from which we can learn.


The following graph will give you the TL;DR of the rest of this post. One eyebrow-raising impression is that, in the city-streets environment, motorcyclists are at fault in multiple-vehicle crashes much less often than is generally believed. In fact, the rider was at fault in only 34% of the crashes in this analysis. That will differ in upcoming posts on freeway and rural crashes.

attachment.php



LEFT TURN--26% of crashes in this analysis
Motorcycle going straight, car turned left. In half of the cases, the car was initially oncoming, and in one-fourth each, the car turned left from a cross street on the left or right. Car was at fault in 84%, usually failure to yield right of way (abbreviated hereafter, FTYROW). Motorcycle was at fault in 16%, unsafe speed, running light or stop sign, FTYROW. Not included are passer vs. passee crashes, where a motorcycle passed on the left as the car turned left.

REAR END--17%
The result was a front-to-rear crash, but they came about in different ways. The most common was one party going straight, the other stopped or slowing. In others, one was stopped or slowing and the other had completed a turn or lane change. Motorcycle was at fault 56%, due to unsafe speed or following too close. Car was at fault 44%, half due to unsafe speed, the other half following too closely, unsafe starting, improper turn, FTYROW, unsafe lane change, et al.

CAR MERGE & PASS--12%
Motorcycle vs. car entering traffic, merging, changing lanes, or passing. In 88% of these crashes, the motorcycle was going straight. The car was at fault 88% of the time, due to unsafe lane change, FTYROW, improper turning. MC at fault 12%, usually unsafe speed.

CROSSING PATHS--9%
This is the red light-runner scenario. Motorcycle and car were both going straight and crossed at a 90-degree angle. Car was at fault in 65%, evenly divided between FTYROW and running a light or stop sign. When the motorcycle was at fault, it was usually due to running a light, but also FTYROW or excessive speed.

MC MERGE & PASS--8%
Car vs. motorcycle entering traffic, merging, changing lanes, or passing. Motorcycle at fault in 86%: improper passing, unsafe lane change, improper turning, unsafe speed, FTYROW, wrong side of road. When the car was at fault, it was due to an improper turn.

MC LEFT, RIGHT OR U-TURN--8%
Motorcycle turning vs. car going straight (half of cases) or turning left or right (one-fourth; remainder thinly distributed). MC was at fault 55% due to FTYROW, improper turning, unsafe speed, running a light, or improper passing. Car at fault 45% for same reasons.

CAR RIGHT TURN--5%
Includes two kinds of crashes: In the first, car turned right from street on the motorcycle's right. In some of these crashes, the car was at fault for FTYROW or improper turn, presumably not completing in the rightmost lane; in others the motorcycle was at fault for unsafe speed. In the second kind of crash, the car turned right from a lane on the motorcycle's left. Some were due to the car making an improper turn, others to the motorcycle making an improper pass.

CAR U-TURN--4%
These comprised U-turns made from the motorcycle's right (58%), from the oncoming lane (27%), and on a cross-street (15%). U-turns from the right were 80% the driver's fault, improper turning and FTYROW, while those that were the motorcyclist's fault were mostly due to unsafe speed. Fault distribution was about the same for oncoming U-turns.


Data source for this post is CHP's SWITRS database (see the fine print at the end of my OP).
 
Last edited:

ThinkFast

Live Long
good stuff - lots of data, Dan :)

So these are findings - facts, objective truths, call em what you like.

Are you going to develop any conclusions from these findings? Recommendations? Or?
 

DataDan

Mama says he's bona fide
good stuff - lots of data, Dan :)

So these are findings - facts, objective truths, call em what you like.

Are you going to develop any conclusions from these findings? Recommendations? Or?
No plans for that. The idea is to post facts, a detailed compilation of Bay Area crash data. But more than that, I want to present available data in ways not usually seen. Commute vs. non-commute hours, city vs. freeway vs. rural, kinds of crashes and at-fault party. These are useful facts I had never seen before tallying them up for this thread.

However, if I were pressured for conclusions so far... ;)

  • The perceived increase in motorcycle crash risk, due to distracted driving etc., is not seen in the data. Maybe it's less serious than claimed, but maybe motorcyclists are adapting to it, too.

  • The Baby Boom generation, now age 55+, is no longer a major factor in California motorcycling. They drove skyrocketing popularity twenty years ago but are aging out, down to 15% of crash-involved riders.

  • Based on reported crashes, most riding is commuting and city/freeway, with a only small proportion on backroads. I don't get it. For me, the appeal of a motorcycle has always been rural riding. While I favor mountains and canyons, even a straight road through farms, pasture, or vineyards gives me the peace and relaxation I'm looking for when I go for a ride. But whether I "get it" or not, urban riding now dominates.

  • We have been misled about parties at fault in multiple-vehicle crashes. Usually, it's not us but the other guy. If it were us, a more important safety emphasis would be to obey the law. Quit speeding and popping wheelies and running red lights you hoons, because you're causing your own crashes. But since it's usually the other guy, it seems that we ARE obeying the law, so a more important emphasis is learning how to play defense.

  • And finally, stay out of San Francisco. You may need to pass through, but ala Captain Willard who said never get out of the boat: Never get off the freeway.
 

ThinkFast

Live Long
No plans for that.
....

However, if I were pressured for conclusions so far... ;)

Nice analysis and summary of some potential takeaways. I like it. Following are your conclusions and some of my own thoughts about each.

The perceived increase in motorcycle crash risk, due to distracted driving etc., is not seen in the data. Maybe it's less serious than claimed, but maybe motorcyclists are adapting to it, too.

I wonder how long it's going to be before we start to see an uptick in motorcyclists crashing due to their own distracted riding?? Fumbling with bluetooth communicators, finding the next radio station, chatting with riding buddies while negotiating the twisties - can any of this really be a good thing, unless used with high levels of discipline?

The Baby Boom generation, now age 55+, is no longer a major factor in California motorcycling. They drove skyrocketing popularity twenty years ago but are aging out, down to 15% of crash-involved riders.

I recall seeing data from a few years ago that showed a high correlation between returning boomer riders, heavyweight bikes and alcohol-related fatalities and injuries. Did you look at alcohol-related crashes in your research? It's a pretty easy story to imagine: returning rider buys that heavyweight bike they've always dreamed about. Goes out with the buddies for a daytrip, and "just has one beer." And then on the way back - surprise surprise - discovers that even one beer can compromise your ability to keep a big heavy bike on the road. Hypothesis: as this demo ages out, we will see less of this particular relationship in the data.

Based on reported crashes, most riding is commuting and city/freeway, with a only small proportion on backroads. I don't get it. For me, the appeal of a motorcycle has always been rural riding. While I favor mountains and canyons, even a straight road through farms, pasture, or vineyards gives me the peace and relaxation I'm looking for when I go for a ride. But whether I "get it" or not, urban riding now dominates.

Probably just a function of numbers. Here in sunny California having a bike as your primary commuter transportation is so viable weather-wise, and so cost effective - HOV lanes, discounts on tolls, cheap and plentiful parking near your office in the city, 50+mpg, low maintenance and insurance costs - hard to resist. Oh - and yes, it's also a hoot on the weekends to get the heck out of town and go play in the hills. Compare that with, say, Wisconsin (where I'm from). Weather? Yeah - no. Fun riding on the weekends? OK, there is this one road...

We have been misled about parties at fault in multiple-vehicle crashes. Usually, it's not us but the other guy. If it were us, a more important safety emphasis would be to obey the law. Quit speeding and popping wheelies and running red lights you hoons, because you're causing your own crashes. But since it's usually the other guy, it seems that we ARE obeying the law, so a more important emphasis is learning how to play defense.

Bingo. Learn how to ride defensively. Yes. When other riders ask me what riding defensively means, I start with the old saw, "ride like you're invisible to other motorists." I've evolved my own, amped up version of that one: "Pretend they do see you and they're going to try to take you out." Now I know that isn't literally true - in fact most motorists I encounter in the Bay Area are very aware of bikes on the road, and they're courteous to me when I'm riding.But by thinking that way it helps me continually scan for potential threats and identify escape routes. (or SEE, from MSF=Search Evaluate Execute).

(Re "staying in the boat" - ie, out of SF - I don't feel that way myself. I actually like a little urban riding now and again. Keeps me sharp... But totally understand your point - if the goal is risk reduction, that advice is spot on).
 
Last edited:

DataDan

Mama says he's bona fide
My "never get out of the boat" comment was mostly tongue-in-cheek. But still, the graph in post #9 showing San Francisco with 2.5 to 3 times as many motorcycle crashes as Oakland and San Jose was a real WTF revelation. To any San Franciscan thinking about taking up motorcycling, my recommendation would be, first, to move.

I recall seeing data from a few years ago that showed a high correlation between returning boomer riders, heavyweight bikes and alcohol-related fatalities and injuries. Did you look at alcohol-related crashes in your research? It's a pretty easy story to imagine: returning rider buys that heavyweight bike they've always dreamed about. Goes out with the buddies for a daytrip, and "just has one beer." And then on the way back - surprise surprise - discovers that even one beer can compromise your ability to keep a big heavy bike on the road. Hypothesis: as this demo ages out, we will see less of this particular relationship in the data.
I found just that in a project last year. The following graph is derived from NHTSA's Crash Report Sampling System and its predecessor, the General Estimates System. It is an estimate of US crash involvement based on a statistical sample of police accident reports. It is not from the Bay Area crash data I'm using in this thread.

attachment.php


This is what I wrote in a post elsewhere:
Note the three humps at 1990, 2000, and 2010. Dominant age groups are: in the first 25-34; in the second 35-44 and 45-54; and in the third 45-54 and 55+. What do these groups have in common? Their birth year is mid-1960s and earlier--it's the same group of riders seen over a 20-year span. Unlike that group, the 1970-1990 birth-year cohort--now ages 50 and under--has never been a big part of the alcohol involvement group. At present, these riders--the < 25, 25-34, and 35-44 age groups--are a small and declining segment of the drinking-and-riding-and-crashing group. An eyeball projection out 5 or 10 years suggests that the pre-1960 cohort will be nearly gone and the overall rate of alcohol involvement in motorcycle crashes will be down around 3%.​
 
Last edited:

DataDan

Mama says he's bona fide
Freeway crashes by freeway

I've slowed down due to temporary burnout from work on 2-vehicle city crashes. The next topic will be freeway crashes, probably next week. But this morning I amused myself counting freeway crashes by freeway. It seemed to make most sense segmented by county. This covers about one-third of segments with reported crashes and 70% of total freeway crashes.

Bay Area Motorcycle Crashes on Freeways 2013-2017

[table=head]freeway..|county...........|crashes
I-880|Alameda|615
I-580|Alameda|543
US-101|Santa Clara|527
I-80|Alameda|361
US-101|San Mateo|313
I-80|San Francisco|274
SR-4|Contra Costa|268
US-101|San Francisco|246
I-280|Santa Clara|234
I-80|Solano|230
SR-85|Santa Clara|200
I-80|Contra Costa|192
[/table]

There's some fuzziness built in because I relied strictly on roadway name and didn't precisely locate each crash. US-101 in San Francisco probably includes some crashes on Van Ness, which is 101 north of the Central Freeway. I think SR-4 is not officially a freeway in a few sections between Martinez and I-80, and is definitely not a freeway east of Brentwood.
 
Last edited:
Love this data - very interesting =)
Thanks for doing all the work.

San Francisco city riding is definitely a mixed bag. The combination of bad roads and impatient drivers means you have to be incredibly cautious about everything and everyone around you. You could be humming along between cars and find yourself heading straight for a huge pothole that is awkwardly shaped and need to figure out a way around it :|

Additionally, with the incredible boom of electric scooters and bicycles has actually made riding a motorcycle X-times harder. They weave in and out of the cars very fast and most of the time are not paying attention to other vehicles (me) coming down between the lanes. Especially downtown you have to be aware of all of them in all lanes at all times.

Riding in the city sucks =) I stick to larger main roads that move to fast for the scooters/bicycles any chance I can get.
 
Last edited:

DataDan

Mama says he's bona fide
Single-vehicle city street crashes

Twenty percent of Bay Area crashes in incorporated cities, but not on freeways, were single vehicle. In some, the motorcyclist lost control and grip, and the machine went down. In others, it hit a stationary object. This post will identify maneuvers that preceded these crashes and causes that contributed.

This graph shows the two kinds of crashes, going down and hitting something. Collision types "overturn" and "other" are combined because some agencies code the type of collision when a motorcycle overturns as "overturn" while others code them as "other", distinguishing them from car and truck overturns.

attachment.php



In 58% of the single-vehicle city crashes, the motorcycle was going straight, not turning, before the loss of control. It seems likely that these were crashes under braking, but the data does not provide that detail.

attachment.php



In 14% of all and 18% of overturns, the rider was not found at-fault. I suspect that other vehicles often contributed, perhaps by violating the motorcyclist's right of way, but were not involved in a collision and were not identified. Speed contributed to more than half. Improper turning, 13%, is a factor when the rider runs wide in a corner. DUI accounted for 8%.

attachment.php
 
Last edited:
Top