John Forester

You state that I should have stated something like: "inadequate for a general solution". Well, I did so, because I specified that the product was to be a "bicycle transportation system", meaning a means of getting from everywhere to everywhere. A design that is "inadequate for a general solution" therefore does not make such a system possible.

Bekologist

How unrealistic. no rider, NO rider expects a path network as "a means of getting from everywhere to everywhere".

that's a sophistic construct. John's absolutism falls short in his attempt to discredit the efficacy of path networks to bicycle transportation.

wphamilton

That would explain the discrepancies. Thanks.

genec

I have to agree... Just as no motorist expects an interstate highway to take them from door to door. But I have battled and battled with John on this ad nauseam... a bicycle path transportation network need be no more connected than our typical automotive highway network... cyclists should anticipate sharing low speed, low volume, residential streets with motor traffic.

However, back to the point... as in what are the main causes of biking fatalities... the answer was and remains: motor vehicles.

John Forester

This is the first time in these years of battling that you have stated the limitation you have just presented. Therefore, your ideal system has a grade-separated bike path along all streets that are not "low speed, low volume, residential streets". Politically, this might appear to be a bit easier, but it presents greater engineering difficulties whose solution might well increase the political unacceptability.

Furthermore, the major cause of cyclist fatalities is not motor vehicles. It is mistakes by cyclists, motorists, and bikeway designers.

B. Carfree

While this is true, if there were no motorists present when cyclists make their mistakes there would be very few cyclist deaths. I don't even want to talk about motorist mistakes since that seems redundant where I live.

invisiblehand

I think you clarify later, but the characterization of bicycling collisions/crashes on isolated paths as scratches and bruises is likely an understatement. If "safest" means the lowest likelihood of mortality, then it's probably right.

FWIW, just off the top of my head, I know of at least one fatality on the WO&D that was away from roads when two cyclists collided. Just based on anecdotes, I know several examples of people going through years of recovery from collisions with peds/dogs/bad construction on the MUPs.

genec

And we know of at least 2 incidents of peds being killed by cyclists last year, that were discussed here. But these are altogether quite rare events compared to the numbers of cyclists killed and maimed by motor vehicles while on public roadways.

The laws of physics alone make it safer to be on a bike path vice a roadway... getting hit with 3000 pounds of mass moving at 30 MPH will impart more force than being hit by 200 pounds of mass moving at 20 MPH.

John Forester

You have ignored the relative frequencies of these events, and you have also ignored the most reasonable ways of preventing them. Both of these considerations are necessary to make your physics assertions valid as applied to bicycle transportation.

invisiblehand

You're forgetting their frequency and how to weight different injuries/mortality. "Safer" in a broader context probably includes things other than mortality; although it is harder to measure. If we really wanted to do a good job, we would consider rider selection. That is, what would happen if you took the road cyclist and put him/her on the MUP and vice versa.

Anyway, if all you care about is mortality, then avoiding cars is a simple but generally a very limiting way of doing it.

genec

Simple thought experiment gents. Remove all the motor traffic on the roadways... so that there is nothing but peds and cyclists... now do you honestly think that the same number of deaths (remember "fatalities" is part of the title of this thread) would occur for cyclists and pedestrians?

The injuries that may occur on paths are often due to the poor design of the paths and cyclists trying to "over drive" these conditions. If paths were designed for typical bike travel at 20 MPH, had a line down the middle and were at least 8 feet wide... (10 would be better) I contend they would be far far safer than roadways with motor traffic. I am not talking about just avoiding death, but safe, as in minimal injuries.

But the fact is that paths are often nothing more than a narrow poorly laid patch of asphalt that often has roots growing through, and kids and dogs playing on them (as the kids can't play in the streets).

But as much hemming and hawing that some folks are doing here... I challenge anyone to show me real statistics on the deaths of cyclists on even the current poorly designed paths... and compare this to the deaths of cyclists on roads filled with motor traffic. The statistics will show that deaths and injuries occur where paths cross roads... and the deaths and injuries occur because cyclists and motor vehicles collide. Go ahead, prove me wrong.

Giro

Given the length of this thread, you probably missed my post #61 which has exactly this information for 2010 from the NHTSA Fatality Analysis Reporting System which anyone can go to and get the data (see post#61 for details).

Here is the table for the most common fatal crash-types for bicyclists age 18 years and older for 2010 in that data:


Note that the most common group of fatal crashes is the "Motorist Overtaking" group with 13.3 + 7.7 + 4.1 = 25.1%. Bicyclist ride-through sign and signalized intersections is 5.7 +4.2 = 9.9%.

The crash-type also varies with the setting. Motorist overtaking is relatively more common in rural than urban settings (see my post #80). This is of some help since you know what setting and situation you are bicycling in. In rural areas, an eye and ear and rearward conspicuity are presumably helpful.

Giro

I always encourage people to find out for themselves, and in this instance both North Carolina's online query of bicycle crash data from 1997 through 2009 and, as of 2010, cross-tabulation in FARS NHTSA's Fatality Analysis Reporting System have your answers.

For North Carolina, do a cross-tabulation of Statewide Bicycle Crash Data selecting "Bicyclist Position" for the rows and "Bicyclist Injury" for the columns for 1997 through 2009. Then you will know where the bicyclist was located just before the crash (e.g. sidewalk, multiuse path, etc.) for the 308 fatalities during this period (as well as the 12,363 non-fatal injuries).

The PBCAT Ver 2 Manual makes it very clear that the Bicyclist Position field is where the bicyclist was before the crash (Appendix F, Table 9 and the example in Appendix G Table 15).

You can make a similar cross-tabulation in FARS since starting with 2010 it has PBCAT crash-typing. Since these are only fatalities, you can do Crash Type by Bicyclist Position for the 640 fatalities in 2010.

I think you will like the answers. The North Carolina data have the advantage that fatal crashes not involving motor vehicles are included.

p.s. On the NHTSA FARS page, start out by clicking the "Submit" button in the upper right part of the page with "2010" selected for the year.

Giro

Acutally, PBCAT crash typing is very explicit about driveways: "Driveways are considered to be nonintersection locations. The exception is signalized commercial driveways which should be coded as intersections." PBCAT Ver 2 Manual, appendix F, Table 9 Thus you can separate driveway crashes from intersections from roadway etc. PBCAT's Bicyclist position is also helpful in this regard.

Between North Carolina's 1997-2009 PBCAT data and now with FARS having PBCAT data, you can get some idea of where and how the crash developed. The problem that always remains is what is the exposure in the different settings?

At least now you can get some notion of the relative risks for crash-types in different environments, such as rural at night, urban intersections, etc. This is helpful in knowing what to look out for in what environment and what might be plausible countermeasures, although I do not think there are that many unexpected findings.

genec

All I get is Perhaps you can summarize for us.

wphamilton

You have to start here https://www-fars.nhtsa.dot.gov/QueryT...electYear.aspx and then go through the wizard to select data. It takes a little fussing with to get the hang of it if you're used to SQL or other queries, so it's best at first to stick with cross-tab reports with single variables. For example, selecting bicycles and "crash location" the total line for all states:Total 209-interstection 42 -intersection-related 371-nonintersection 7-nonroad 11-other, which comes to 39% relating to intersections, in line with one of the stats I quoted. (My recalled 80% is looking even more shaky).

The problem is still that the numbers are not related to the overall amount of bicycle traffic. Crashes per bike-mile, or crashes per road mile etc. So it doesn't tell us much about relative safety but it's half-way there if we could also get overall traffic data reliable data, correlated in the same fashion.

Giro

Thank you, wphamilton. I've fixed the links & added a p.s. since where you click to get started is hard to find in the upper right part of the page.

invisiblehand

I don't think that the response addresses my comment; although we might just be on different wavelengths here.

If you're just looking at fatalities then based on the counts of fatalities and hospitalizations, you're omitting a major component of cycling risk. Just because "fatalities" is in the title of the thread doesn't mean that "safest" means lowest risk of fatality. The challenge here is finding good statistics instead of rejecting/failing-to-reject your hypothesis.

Daves_Not_Here

How Not To Get Hit By Cars https://bicyclesafe.com/

This site lays out 10 common collision scenarios and specific steps you can take to prevent them. Actionable intelligence very clearly presented.