Bay Area Motorcycle Crashes 2013-2017

DataDan

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

The 5700 Bay Area freeway crashes 2013-2017 were one-third of all police-reported motorcycle crashes in the region. Of those, 20% were single vehicle (the subject of an upcoming post), 60% were 2-vehicle (the subject of this post), and 20% were 3+ vehicle (no post since the crashes are much harder to evaluate). While I refer to the other vehicles involved in these crashes as "cars", they comprise every kind vehicle on the road: cars, SUVs, vans, trucks, buses, etc. However, I have not included bike-vs-bike crashes.

Fault in two-vehicle freeway crashes was nearly equally divided between the driver (51%) and the rider (49%).

attachment.php



Half of crashes were sideswipes, usually caused by the car. Close to 40% were rear-end, usually caused by the motorcycle. The remainder was distributed among several different categories.

attachment.php



This graph shows the pre-crash movement by the party at fault. When the motorcycle was at fault, it was going straight 68% of the time. When the car was at fault, it was changing lanes 85% of the time.

attachment.php



Finally, here are the primary factors behind the at-fault assessment. When the motorcycle was at fault, it was due to unsafe speed 70% of the time. For the car, it was an unsafe lane change 85% of the time

attachment.php
 

GAJ

Well-known member
Finally, here are the primary factors behind the at-fault assessment. When the motorcycle was at fault, it was due to unsafe speed 70% of the time. For the car, it was an unsafe lane change 85% of the time

Great stuff Dan.

I used to love lane splitting when I commuted before I retired.

But in a car watching some of these guys split at large deltas with traffic despite big gaps on both sides of them makes my skin crawl.

Same for excessive speed within urban environments.
 

DataDan

Mama says he's bona fide
Single-vehicle freeway crashes

Of 5700 motorcycle crashes on Bay Area freeways 2013-2017, 1100 or 19% were single-vehicle, a proportion similar to city-street crashes.

As with single-vehicle city-street crashes (see post #20), single-vehicle freeway crashes were usually overturns--68% freeway vs. 61% city. Most of the remainder in both was hitting a stationary object. The cause of the overturns was likely the same as on city streets--overbraking--though, again, that detail is not in the data.

attachment.php



In 49% of the single-vehicle crashes, the motorcycle was going straight before the crash, and in 35% it had left the freeway travel lanes.

attachment.php



The distribution of primary factors is similar in overturns and object strikes. For all single-vehicle freeway crashes, 42% were due to unsafe speed, 37% to improper turning. "Improper turning" in the freeway context appears to be a maneuver that ended in failure to stay in the travel lanes, causing the motorcycle to go down or hit something on the shoulder.

attachment.php
 
Last edited:

DataDan

Mama says he's bona fide
Single-vehicle rural crashes

With crashes on freeways and city streets covered in previous posts, one last location category remains, which I call "rural", comprising non-freeways in unincorporated areas. While many of these were on our favorite back roads, others were in unincorporated communities in urban areas. For example, Castro Valley, with population over 60,000, is in this group. Unfortunately, to separate the unincorporated urbans from the back roads would have required going through 3000 crashes one by one, since there is no data element to characterize them.

While freeway and city street crashes were about 20% single-vehicle, 45% of rural collisions were single-vehicle, the subject of this post. Of 1358 single-vehicle crashes, 65% were overturns, proportionally similar to single-vehicle city and freeway crashes.

attachment.php



In 43% of single-vehicle rural crashes, the motorcycle was going straight before the overturn or collision; in 52%, it was turning or ran off the road. As might be expected, it ran off the road much more often in rural crashes, 28%, than in city crashes, 5%.

attachment.php



The causes of single-vehicle rural crashes were improper turning, 47%, unsafe speed, 37%, and DUI, 8%.

attachment.php
 
Last edited:

DataDan

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

As discussed in my previous post, the "rural" category includes many different kinds of roads. Some are back roads, some are basically urban but in unincorporated communities (e.g., Castro Valley), and some are more like freeways (Highway 17 and parts of Lawrence, Montague, and San Tomas Expressways). Correspondingly, crashes were varied, too. Some were the expected twisty road crashes, but others were more like city street and freeway crashes. It is an unfortunately artificial grouping that doesn't really tell us much about crashes in a particular environment.


Two-vehicle rural crashes were only somewhat more common than single-vehicle, 55-45%, compared to 80-20% for city and freeway. At-fault division differed, too: The motorcyclist was at fault slightly more often than driver in two-vehicle rural crashes, 51-49%, while it was usually the driver on the freeway, 49-51%, and in the city 37-63%.

attachment.php



Crash types reflected the different environments included in the category. Broadsides (31%) were the most common, as they were in the city. Sideswipes (27%) and rear ends (21%) occurred in relatively high proportions, as they did on both city streets and freeways. Overturns (10%) and head-ons (8%) would seem to be more closely associated with loss of control in a curve.

attachment.php



When the motorcycle was at-fault in rural crashes, it was going straight before the crash in 50% of cases. But at-fault pre-crash movement also included passing (15%) and crossing the centerline (14%), actions more typical of crashes resulting from loss of control in a curve. When the car was at fault, it was turning left or changing lanes in 49%.

attachment.php



When the motorcycle was at fault, it was due to unsafe speed 44%, a common factor in all environments. Other factors that appeared frequently were wrong side of road (18%), improper passing (13%), and improper turn (8%). These would seem to be twisty-road errors. Primary factors when the car was at fault, however, were more typical of city street and freeway crashes: right of way violation (41%) and unsafe lane change (22%)

attachment.php
 
Last edited:

pobey

Well-known member
Thank You Dan for posting this useful information. I'm always processing such information for my publications. Thanks again from another Dan.
 

DataDan

Mama says he's bona fide
Crash Fatalities

Earlier posts in this thread have been about all Bay Area motorcycle crashes reported to police, non-injury to fatal, derived from the CHP-maintained database of collisions reported by all agencies. For the next few posts, I turn to fatal Bay Area crashes that appear in the US DOT's database of US traffic deaths since 1975, which has considerably more detail than the larger CHP database.


We've come a long way in 40 years. Though the number of registered motorcycles statewide (I have Bay Area registrations only for the past 15 years) is 25% higher now than in the late 1970s, annual California motorcyclist deaths have fallen by 40%. The Bay Area drop was even greater at 60%.

attachment.php


The age distribution reveals another significant change. In 1978 and 1979, 90% of deaths were of riders under age 35, annual deaths under age 25 averaged 105, and 55 teenagers were killed per year. Now, the < 35 group is less than half, only 104 under age 25 have been killed in the past TEN years, and in the past five years, 5 teenagers died in total.

The tragic loss of young lives would not be tolerated today, and it was not tolerated for long back then. In the 1980s, the horrific death toll drove two legislative efforts: The training requirement first for riders under 18 (1987) then under 21 (1991), and the all-rider helmet law (1992).
 
Last edited:

DataDan

Mama says he's bona fide
Fatal crashes by location

After the short detour into history, back to crashes 2013-2017. First, an overview.

Bay Area Fatal Motorcycle Crashes 2013-2017

No code has to be inserted here.

Non-motorists were 7 pedestrians and 1 bicyclist. More on that in a later post.


Distribution of fatal crashes by county is somewhat different than the distribution of all crashes seen in post #1. While Alameda County had more total crashes than Santa Clara, the Santa Clara crashes were more likely to be deadly. OTOH, San Francisco's crashes were much less likely to be deadly--dropping from third on list of all crashes to a distant seventh in fatal crashes.

attachment.php



Post #9 showed the extraordinary number of crashes in San Francisco. But they were less likely to be deadly than those in other cities.

attachment.php



In post #6, crashes of all severities were broken down by roadway type: non-freeway crashes within city limits, freeway crashes, and "rural" or non-freeway crashes outside of city limits. This is the same data in a slightly different graph.

attachment.php



Fatalities were distributed somewhat differently than non-fatal crashes.

attachment.php



The crash and fatality distributions make it possible to calculate crash lethality, the percentage of crashes that claimed lives, by roadway type. While city crashes were close to the average Bay Area lethality of 2.0% (see post #1), freeway crashes were less likely than average to be deadly, and rural crashes more likely.

attachment.php
 

DataDan

Mama says he's bona fide
Vehicles involved

The NHTSA fatal crash data includes VIN (first 12 characters), and an auxiliary file with make/model details is provided. In this post I will show the make, model, and style of the motorcycles in fatal Bay Area crashes and style of the other vehicles in those crashes.


Motorcycles in Fatal Bay Area Crashes by Style 2013-2017

No code has to be inserted here.

"Style" is provided by R.L. Polk, an auto industry information service. The first three are pretty much what you would expect, but "touring" is loaded with Harleys, 22 of them. "Traditional" is what we used to call "standard"--unfaired, no cruiser styling cues. "Enduro" was probably a good description at one time, but it is now adventure and dual-sport. For "dirt" I combined street-legal dirtbikes and motocrossers (in road accidents). "Scooters" are what you would recognize as scooters.


Motorcycles in Fatal Bay Area Crashes by Make and Model 2013-2017

[table=head]make|model|count|%
BMW||11|3%
|S 1000 RR|5|
Ducati||11|3%
Harley-Davidson||97|28%
|FL|56|
|FX|21|
|Sportster|15|
Honda||42|12%
|CBR600 RR|10|
|CBR1000 RR|6|
Kawasaki||35|10%
|EX250/300|7|
|EX500/650|6|
|ZX600/636|7|
|ZX1000|4|
KTM||4|1%
Suzuki||51|15%
|SV/SFV650|5|
|GSX-R600|12|
|GSX-R750|5|
|GSX-R1000|7|
Triumph||9|3%
Yamaha||66|19%
|YZFR6|20|
|YZFR1|12|
|XVS950|4|
other/unknown||17|5%
total||343|

[/table]

I didn't include all makes in this list, excluding the ones and twos. Count on the make line is the total for the make. I broke out some model counts that might be of interest. Many different Harley models appear, but I don't know much about 'em except the two basic kinds of Big Twins and Sportsters. I was surprised how few Japanese cruisers appear in the data, and only one numerous enough to make this list.


Other Vehicles in Fatal Bay Area Motorcycle Crashes by Type 2013-2017

No code has to be inserted here.

I didn't find much of interest in the make/model distribution so didn't include it (but will on request). For info, there were as many Priuses as Peterbilts--5 each.
 
Last edited:

DataDan

Mama says he's bona fide
Types of Fatal Crashes

Earlier in the thread, I covered types of crashes for all severities, non-injury to fatal, from the CHP database. For fatal crashes only, the NHTSA database includes more detail on the events of a crash, the subject of this post.


In Bay Area fatal motorcycle crashes 2013-2017, the motorcyclist was at fault in 75% of all cases and 62% of multiple-vehicle crashes. These percentages are higher than for all-severity crashes, 54% for all and 42% for multiple-vehicle.

In 73% of single-vehicle fatalities, the motorcycle ran off the road. In 14% it struck an object, and in 11%, the bike overturned. The assessment of these factors is made by NHTSA after examining the police crash report. One or more may occur--ran off the road, overturned, hit object--but the factor cited is the one judged to be the initial event of the incident.

attachment.php



Multiple-vehicle fatalities comprised many different kinds of crashes, as shown below. A collision with a left-turner, either oncoming or from the right, was the most common, 28%. This was also the most common non-fatal crash on city streets. Rear-enders, motorcycle striking, was the second most frequent fatal crash at 14%. These occur both on city streets and freeways. Sideswipes in the same direction of travel were the highest non-fatal freeway crash and third among fatalities.

attachment.php
 

budman

General Menace
Staff member
Really interesting stuff on the comparison to the 70's and on the types of bikes crashing. HD's are way up there. My mind wonders how many of those involved alcohol.

I am always shocked by the amount of riders going off the road. :(

The different application of why that happens would be interesting to know. Hit an object seems straight forward. Rock, debris, gravel an object?? Alcohol there would be interesting too.

Thanks Dan.
 

budman

General Menace
Staff member
That is pretty high for both. Drinking and driving is bad. Drinking and riding is just a really bad idea period.
 

DataDan

Mama says he's bona fide
Of the 343 riders in fatal crashes, 97 or 28% were reported to have been drinking. This is based on a blood test, if available, or the investigating officer's subjective assessment. Here's how alcohol use breaks down by age:

attachment.php
 

DataDan

Mama says he's bona fide
Non-motorists killed in motorcycle crashes

In the Bay Area 2013-2017, seven non-motorists were killed in motorcycle crashes. I am including only those killed by a motorcycle, not those killed by another vehicle in a crash that also involved a motorcycle.

[table=head]city|victim|mc at fault?
---------------------|-------------------|----------------
Oakland|bicyclist|no
San Francisco|pedestrian|DUI
San Francisco|pedestrian|speed
San Francisco|pedestrian|DUI
San Francisco|pedestrian|no
San Jose|pedestrian|no
Santa Rosa|pedestrian|no
[/table]
 

DataDan

Mama says he's bona fide
Lane-splitting deaths

There were 13 fatal lane-splitting crashes in the Bay Area 2013-2017. I identified them from news stories that quote a CHP officer or press release (none in this group were in the jurisdiction of another agency). Lane-splitting might be mentioned in a news story unsourced or by a 911 caller via the CAD log, but that is not sufficient to make this list.


Bay Area Lane-Splitting Fatalities 2013-2017 by Roadway

[table=head]roadway/county|count
-----------------------------------|---------
US-101 Santa Clara|5
I-880 Alameda|2
US-101 Marin|1
US-101 San Mateo|1
US-101 Sonoma|1
I-80 Solano|1
Altamont Pass Rd|1
San Tomas Expressway|1
[/table]


Bay Area Lane-Splitting Fatalities 2013-2017 by Traffic Speed

[table=head]traffic speed|count
-----------------------------------|---------
< 20|3
20-49|4
50-64|3
65+|3
[/table]


Bay Area Lane-Splitting Fatalities 2013-2017 by Speed Differential

In 3 cases motorcycle speed was not reported, so differential could not be calculated. In all 3, traffic speed was 55 or higher. In 2, unsafe motorcycle speed was cited as the cause of the crash, and in the third it was rider DUI with speed as a contributing factor.

[table=head]speed differential|count
-----------------------------------|---------
<= 15|5
16-24|2
25+|3
[/table]


Bay Area Lane-Splitting Fatalities 2013-2017 by Other-Vehicle Pre-Crash Movement

[table=head]other vehicle pre-crash|count
-----------------------------------|---------
going straight|9
slowing or stopped|2
changing lanes|2
[/table]


Bay Area Lane-Splitting Fatalities 2013-2017 by Type of Crash

"Loss of control" is an overturn, most likely a crash under braking.

[table=head]type of crash|count
-----------------------------------|---------
sideswipe|5
rear-end|5
loss of control|3
[/table]


Bay Area Lane-Splitting Fatalities 2013-2017 by Fault

[table=head]motorcycle at fault?|count
-----------------------------------|---------
speed|5
improper passing|3
DUI|3
unsafe lane change|1
not at fault|1
[/table]
 
Last edited:

DataDan

Mama says he's bona fide
Backroad crashes

In this thread I have reported on crashes by roadway type--city streets, freeways, rural--but the "rural" distinction didn't capture what I had hoped, the sport-riding environment. Instead, most of those 3000 crashes occurred on non-freeway commuter arteries and unincorporated city streets.

To identify sport-riding roads, I went through road names in the "rural" group, flagged candidates, and, if unfamiliar to me, checked them out in Google Earth. Selecting those with a minimum of 5 crashes in 5 years, I came up with 38 roads and a total of 954 crashes. To some extent, selection reflects my personal preference--roads where the demand is on cornering skill and where the fun happens well under triple digits. Here's is the top of the list, representing 64% of backroad crashes:

Bay Area Sport-Riding Road Crashes 2013-2017

[table=head]area|crashes
-------------------|-------------------
SR-1|160
Lake Berryessa|131
SR-9|128
SR-84|84
Redwood Rd|64
Mines Rd|42
|
total|609[/table]


  • SR-1 here is Marin and Sonoma Counties only; the Santa Cruz and San Mateo parts are not in the 38 roads.

  • Lake Berryessa includes several Napa County roads near the lake: SR-121 between the Napa city limit and SR-128; SR-128 between Silverado Trail and the Solano County line; Berryessa Knoxville Road; Chiles Pope Valley Road; Pope Canyon Road; Butts Canyon Road.

  • SR-9 from Santa Cruz to Saratoga.

  • SR-84 from the coast to Skyline (Niles Canyon is part of the 954 crashes but in a smaller Alameda County group).

  • Redwood Road between Castro Valley and the Oakland city limit.

  • Mines Road includes SR-130 between the San Jose city limit and the observatory.

In upcoming posts I will explore single- vs. multiple-vehicle crashes, lethality, type of crash, and primary factors--similar to the city and freeway crash posts.
 

DataDan

Mama says he's bona fide
Backroad fatalities

While the 954 Bay Area backroad crashes I have identified 2013-2017 (see previous post) were a small part of the region's 16,000 motorcycle crashes, they were a larger contributor to motorcyclist fatalities--33 of 327. This is mainly because a multiple-vehicle crash on a backroad was much more likely to claim a life than on a city street or freeway. Another vehicle was involved in 43% of all backroad crashes but 64% of fatalities.

attachment.php



Lethality--the percentage of crashes resulting in death--was 2.2% for solo incidents but much greater, 5.1%, for multi-vehicle.

attachment.php



Compare single- and multiple-vehicle lethality among city streets, freeways, and backroads:

attachment.php
 
Last edited:

DataDan

Mama says he's bona fide
Types of backroad crashes

Of the 954 backroad crashes I identified, 546 or 57% were single-vehicle.

attachment.php



The motorcyclist was found to be at-fault in all but a few single-vehicle crashes and in 56% of the multiple-vehicle crashes:

attachment.php



Primary factor in the single-vehicle crashes:

attachment.php



Collision type in the multiple-vehicle crashes where the rider was at fault:

attachment.php


In "over centerline" incidents, the motorcycle crossed the centerline either in a curve or while passing and hit an oncoming vehicle. In "passer vs passee" the rider hit a vehicle turning left while attempting to pass it. In "rear end" the motorcycle struck the vehicle ahead. "OV left turn" includes incidents where the motorcyclist was found to be at fault in a crash with a left-turner either oncoming or crossing from the right, most often due to excessive speed. In "overturn" the motorcycle crashed on its own but incidentally hit another vehicle.


Collision type in multiple-vehicle crashes where the other driver was at fault:

attachment.php


"Left turn etc" is mostly left turns, from the oncoming lane or either side, but also includes entering the roadway, U-turns, and right turns. Basically where the driver failed to yield right of way to the motorcyclist. "Over centerline" incidents are like the motorcycle at-faults, but the other driver ran wide in a curve or was passing. "Cutoff" is like the crashes anywhere else--merging into the motorcyclist's lane. Same with "rear-end", but also includes some incidents where the vehicle was stopped in the roadway and unavoidably hit by the motorcycle.
 

DataDan

Mama says he's bona fide
Conclusion

When I started this thread I didn't know where it would go. I had lots of data on Bay Area motorcycle crashes--so much that it didn't cohere into a useful view. But the process of breaking it down this way and that has helped me make sense of it as a motorcyclist trying to learn how crashes happen in order to prevent them.

In this finale, I'm going to present as simply as I can: by roadway environment, the most common kinds of crashes. Most riders are somewhat familiar with the findings of the Hurt Report, and we all know the folklore that has been around forever. But to get very specific: in the context of the San Francisco Bay Area 2013-2017, based on police reports compiled in the CHP's SWITRS database, how do motorcyclists crash? What emerges isn't exactly "everything you know is wrong", but there are some unexpected results.


If you've skimmed this thread, you know that I initially identified three roadway environments: city streets (non-freeways within city limits), freeways, and rural (non-freeways outside of city limits). The "rural" division proved unhelpful. I was trying to capture the sport-riding environment, but instead got two-thirds unincorporated city streets and one-third twisty roads. I eventually did the hard work of identifying the cool ones, and posted about them. Now I have grouped unincorporated city streets previously in "rural" together with the incorporated city streets, collectively calling them all city streets. Sport-riding roads are what remains of the "rural" group. Freeways remain as I originally defined them.

Starting with approximately 16,000 SWITRS crashes, I identified 9300 on city streets (58% of the total), 5700 on freeways (36%), and 1000 on twisty roads (6%). In the sections below, I rank crash types in each environment by frequency of occurrence.

A few explanations about crash type:
  • cutoff: a crash caused by right turn, lane change, merge, or traffic entry into another vehicle's path
  • signal & ROW: failure to obey sign or signal or to yield right of way
  • LOC speed: loss of control due to speed
  • LOC turning: loss of control while turning


City Streets

[table=head]single/multiple|fault|type|%
----------------------|-------------|-------------------------|-----------------
multiple|car|car left turn|20.1%
multiple|car|car cutoff|10.7%
single|mc|mc LOC speed|10.1%
multiple|mc|mc rear-end|7.3%
multiple|car|crossing vehicle|6.6%
multiple|car|car rear-end|5.6%
multiple|mc|other|5.2%
multiple|mc|mc signal & ROW|4.5%
single|mc|mc other|4.1%
multiple|mc|mc speed|3.7%
multiple|car|car U-turn|3.4%
multiple|mc|mc passing|3.2%
single|mc|mc LOC turning|2.6%
multiple|mc|mc cutoff|2.4%
multiple|car|car other|2.1%
multiple|mc|mc left turn|1.9%
[/table]

The notorious left-turner is right where expected. Less obvious is the car cutoff. Both the single-vehicle motorcycle loss of control due to speed and the motorcycle rear-ender are easily preventable.

Freeways

[table=head]single/multiple|fault|type|%
----------------------|-------------|-------------------------|-----------------
multiple|car|car cutoff|29.4%
multiple|mc|mc rear-end|22.1%
single|mc|mc LOC speed|8.0%
multiple|car|car rear-end|7.7%
single|mc|mc LOC turning|7.1%
multiple|mc|mc speed|6.3%
multiple|mc|mc cutoff|3.7%
multiple|car|car other|3.0%
single|mc|mc other|2.5%
multiple|mc|mc other|2.0%
multiple|mc|mc turning|1.8%
multiple|mc|mc passing|1.8%
[/table]

Maybe more rider education emphasis is needed on preventing freeway cutoffs, the most common freeway crash in the Bay Area. I suspect that many of these are lane-splitting crashes, but that data is not available. The two-second following distance rule can prevent many rear-enders. If you can't maintain that much space cushion, put yourself behind a small vehicle that gives you a good view to traffic ahead.

Backroads

[table=head]single/multiple|fault|type|%
----------------------|-------------|-------------------------|-----------------
single|mc|mc overturn|34.6%
single|mc|mc hit object|18.7%
multiple|mc|mc over centerline|10.0%
multiple|car|car left, right, U-turn|7.6%
multiple|car|car over centerline|3.9%
multiple|mc|mc passer vs passee|3.9%
multiple|mc|mc rear-end|3.7%
multiple|car|car cutoff|3.4%
multiple|mc|mc other|3.0%
multiple|car|car other|2.1%
multiple|mc|mc overturn|2.1%
multiple|mc|mc speed|1.8%
multiple|car|car rear-end|1.3%
[/table]

These crashes, a small percentage of the total, are far more likely to be caused by the motorcyclist than those in other roadway environments.
 
Last edited:
Top