BIKEPGH MESSAGE BOARD ARCHIVE

Nice! Let's see how nah-sayer drivers respond to this.

This comes in nicely with Stu's shared article on how cars hitting cyclists/pedestrians/other cars are not accidents:

http://transalt.org/newsroom/magazine/2012/Summer/2

A bike got a "too fast for conditions"? Damn. I can see how but still, damn.

What conservative middle aged white men see: "In Pittsburgh accidents, cyclists [???] always to blame".

The graphic is kind of shitty in that it doesn't identify "Bike running a red light" or "Car running a red light" etc. They will see it their way and we will see it ours. Resolves nothing. Share the road.

Yeah, that graphic is pretty terrible, it doesn't even add up to 100%.

Shitty graphic, shitty headline, only moderately bad article, and the usual blast of asshole comments...

I wonder if Samuel Hurst owns a white Jeep...

edmonds, I was just thinking the same thing. The only accident "cause" that MIGHT be seen as the driver (and not the cyclist) is the "other improper driving actions", meaning that someone could interpret 80% of accidents to be the cyclists fault, especially when all the cyclists they see are light-running tie-fighter traffic-filtering granny-knocking-over hooligans.

Article text indicates "about half" of car v. bike accidents attributed to cyclists. However chart only details the accidents attributed to cars (that's why it doesn't add up). Results in what appears to be a slanted article, and makes the cyclists appear to be the 800-pound gorilla in the equation.

Unfortunately and poorly written article, and as edmonds says: resolves nothing. Sadly, lost opportunity to make a better point.

The Transportation Alternatives article is outstanding.

I love the article's whole bewildered take on bike-car crashes.

Why on earth do they happen? Where do they happen? Who can know for sure? It's such a mystery, all we can do is rant about it afterwards, waving our hands in the air. There's clearly no way of improving any aspect of the situation. Oh, my, whatever will we do? There's nothing we can do, it's just the way life is...

So. Tired. Of. That.

I agree that this is a poorly written article with a head scratching graphic - but it does appear on the front page of the newspaper, albeit below the fold, and it at least keeps the dialog going. Half of the battle is letting people know that there is a problem and this story pushes us closer in that direction.

letter to the editor time

I dunno. I worry that because of the way the information is presented, it lends the appearance they are underplaying the "half" caused by cyclists, and presenting a skewed perspective, which may only feed the bias against bikes a big fat "see - the cyclists are at fault (too)" bone to chew on.

It is A story, and I agree that keeps the dialog going. Unfortunately I'm afraid the ongoing dialog will be: "bikes cause crashes too." And won't advance toward the goal of acceptance, respect on the road and ultimately greater safety.

Agreed Marko. I feel that ANY discussion about bike safety right now is a good thing. There will always be comments about scofflaw cyclists, ranging from "tsk-tsk" to "run 'em off the road" and I'm sure this will trigger a new round of letters to the editor. But that's OK, because it means people are talking about it.

I'm hoping the next article puts the "bicyclists don't pay taxes" myth to rest.

there should also be a "myth of the scofflaw cyclist" article much like this one: http://washcycle.typepad.com/home/2008/07/the-myth-of-the.html

I really enjoyed Aaron Naparstek's article on T.A.'s page. We were retweeting that the other day when it came out. In fact I sent it directly to a bunch of local traffic reporters for them to read.

I did notice in the article that the signs directing cyclists away from Penn are to become permanent. Creating a cyclist ghetto is not the solution to dangerous roads.

Jon, no one is saying we should ignore redesigning Penn. The signs are needed and the three streets running parallel will be getting bike lanes and sharrows. I wouldn't call that a cyclist's ghetto. the state owns Penn and the mid-long term goal is to work with them and the community on redesigning it. The City owns the streets in the vicinity. I still stand by this decision 100%.

"Drivers get - I feel - irrationally angry about this. I wondered why for so long; and then an anthropologist friend of mine helped me to understand. Running a red light is so dangerous for cars that it isn't just illegal, it's taboo. You're breaking a social construct. That means people find it objectionable and abhorrent. So if education is needed, maybe it's needed to explain why it's safer for cyclists to do it than for drivers."

I don't know that I buy that, but drivers definitely see shades of gray with their own legal responsibility to follow the laws of the road. They just don't afford cyclists their own, unique shades of gray too. Like speed limits - no driver sees them as hard and fast unless there's a cop on your tail on a holiday weekend. Bikes, on the other hand, are not allowed any flexibility.

This: "The premise that cyclists' behavior somehow voids their right to sharing the road is indefensible."

And it's amusing to realize that drivers have their own versions of the law as it applies to cyclists. I was crossing the bridge between Wilmerding and Turtle Creek the other day, staying in the left lane because I turn left at the end of the bridge, signaling my intent to turn so drivers behind me know they should pass on the right, when a guy in a pickup truck (naturally!) yelled at me as he passed "According to the law you should have ..." something about signals required on my bike, I think. It's like people who hate cyclists have some Internet forum where they make up and share cyclist laws.

I find it surprising -even shocking - that half the bike-car accidents are the bicyclist's fault. My guess would be more like 9:1 cars cause.

That will make me seriously look at how I travel on the streets. It's dangerous out there for bikes, but careful riding could reduce accidents up to 50%.

Now, what I need to know is what is the proportion of fault in fatal accidents?

For example, truckers regularly tout the fact that their accident-per-million-miles rate is far lower than cars. They fail to note that their fatal accident rate is far higher than that for cars and it is usually not the trucker who dies.

just submitted to PG:

In response to "In Pittsburgh accidents, bicyclists not always to blame", as well as many recent letters to the editor in the wake of the recent Penn Avenue cyclist deaths, I feel as though I've sat down at the Mad Hatter's tea party or otherwise gone insane.

There is absolutely nothing defensible or rational in the following arguments:

1. I don't have to share the road with cyclists because they break the law all the time.

2. Cyclists don't [insert motor vehicle requirement or financial cost] so they should be banned from the road.

3. Cyclists have trails to ride on so they shouldn't be on the road.

I may be of the opinion that I am the rightful queen of England, but it doesn't mean the opinion is worth publishing. Please stop lending credence to ill considered, irrational ideas that put cyclists lives in danger when acted upon. Your writers and readers deserve a higher caliber news paper than that.

Mick - that's half by what the police reports say. That's not necessarily half in reality. Check out the washcycle link Scott posted.

I teach intro stats so I am grateful for that POS graphic since I can use it in my lectures. Otherwise the article sucked. And some of the comments were terrifying. I am glad I didn't read them until after I got home from riding with my kid today.

(Focusing on the positive here...) that TA article is awesome. My significant other and I were just talking about the use of the word "accident" the other day and how misleading it is.

Now I'm curious: do crashes involving drunk drivers get the same "accident" treatment, typically? Being drunk seems to be the one factor that removes a car driver's free ticket to cause mayhem.

Jeez the more I look at that the worse it gets; One of the "top 5 causes of accidents" is "unknown" ?!? WTF? Unknown is not a cause. Who the heck proofed this thing?

I would blame the data for that, not the reporting. Police reports are pretty terrible. The only data my police report was that the driver made a "careless turn."

One commenter got it right, "worry about you". This argument between drivers and bikers boils down to a jealous feeling that "operators" on the road have when they perceive another "operator" doing something against the rules or against what they themselves would do.

Want to test my theory? Drive the speed limt or under in your car. Drivers will tailgate you, honk at you, and usually give you a hand gesture....and you are driving a car.....

It is all about the perception....

@KBrooks - I think so, yes. It is usually described "an accident caused by a drunk driver"...

An example:

"...A 19-year-old South Fayette woman who pleaded guilty to drunk driving and killing two brothers in a car accident.."

http://www.post-gazette.com/stories/local/washington/woman-sentenced-in-drunk-driving-crash-639931/?print=1

The "I don't want to hurt anybody but..." arguments are my favorite.

Thanks bikeygirl! I guess some more searches for "drunk driver" would provide plenty of data.

I sent an email to the reporter asking about the graphic and the difficulty in discerning who was doing what - was a Driver making the inappropriate turn? was a Bicyclist running the red light?

He sent me a nice (and timely) response and included these numbers breaking it down by Role/Action:

Driver's "fault"

Making improper or careless turn n=60

Other improper driving actions n=32

Unknown n=27

Careless passing or lane change n= 10

Proceeding w/o clearance after stop n= 10

Cyclist's "fault"

Other improper actions n=42

Unknown n=18

Running red light n= 17

Making improper or careless turn n= 16

Running stop sign n=16

and so I couldn't help myself and put it in a spreadsheet:

As others mentioned earlier, my big takeaway is the poor quality of the source data.

@KBrooks: I know that wasn't enough data. However, looking more into it, I'm finding that "drunk driving" and "accidents" are acceptable...

Just like cars hitting/killing predestrians, cyclists, or other cars; are also "accidents"...

From a lawyer's point of view:

http://www.edgarsnyder.com/drunk-driving/drunk-driving-statistics.html

From DUI/alcohol prevention sites:

http://www.alcoholalert.com/drinking-driving-accidents.html

From the news: http://www.recordonline.com/apps/pbcs.dll/article?AID=/20120731/NEWS/207310325

A pie chart has to sum to 100%.

Period.

It's strange that "running a stop sign" or "running red light" aren't listed as possibilities under Driver's "fault." Also, thanks Vannevar because 56% does not equal 44%, yet the author said "half are the bicyclists fault. 12% is significantly higher.

Sara that just proves your political bias

http://www.buzzfeed.com/ebroodle/fox-news-pie-chart-fail-68j

Did you see this:

http://newsinteractive.post-gazette.com/bikeAccidents/

and then the raw data link at the bottom?

I can't seem to make the search feature work, but the popups at each data location are very interesting.

I am trying to correlate it to the Bike Pgh crash map, but haven't really had time to play with it yet.

haha yes! Another pie chart for my stat 111 course, lol. Classic.

I tried to find a way to present the data and I ended up with this:

Fault and Cause of Car-Bike Collisions in Pittsburgh, 2006-2012

my apologies for being unable to draw a stick-figure.

sarah - re the stats class, I think the original chart was what Tufte would call chartjunk - easily formatted, macro-driven, meaningless

+1 for Tufte!

+1Vannevar!

But...

um...

Don't you think the angle on that saddle is a little off? (ducks and runs)

JOSEPHINE ST, 2009-09-21

SUBMITTED BY PENNDOT

PennDOT says bike is at fault for speeding.

((tried to post this earlier but the msgboard thought I was spamming...))

@KBrooks:   I know that wasn't enough data. However, looking more into it, I'm finding that "drunk driving" and "accidents" are acceptable...

Just like cars hitting/killing predestrians, cyclists, or other cars; are also accidents.

From a lawyer's point of view:

http://www.edgarsnyder.com/drunk-driving/drunk-driving-statistics.html

From DUI/alcohol prevention sites:

http://www.alcoholalert.com/drinking-driving-accidents.html

From the news: http://www.recordonline.com/apps/pbcs.dll/article?AID=/20120731/NEWS/207310325

@rsprake that's at the tight turn with 18th St. A cyclist could easily be going pretty fast there.

The math is off a bit between the numbers in the PG image (236 incidents) and the numbers Vannavar was given (248 incidents) but I don't think the % labels on this pie chart are supposed to add up to 100%. They are supposed to add up to the "just over half" percentage of accidents mentioned in the article where the bike is at fault - 64%. That contradicts the numbers Vannevar was given where more than half are automobile faulted.

So "other improper" by cyclists is 21% of the PG image total and shown in the label on the pie chart (17% in Vannevar data), but "other improper" by anyone is about 30% of the total 100% and denoted by the area of that colored section of the pie chart.

Vannevar's graph is great!

Before I saw Vannevar's graph, I also tried to graph the data - you can see my results at:

http://danieleweeks.com/bicycling/

Vannevar - kudos to an awesome graphic. Only two small things: 1. if the causes were color coded one could easily see how they track that way too, and 2. It needs fenders and a rack to be a good commuter

I am creating a summary of this information for other purposes. Not quite done yet, but have been struck by the number of accidents that occur outside of what I think of as the standard cycling season (mid April to Mid October, or so).

I should have those numbers shortly. Unless Vannever or ?? has them already.

@Swalfoort, no I don't have any other analysis BUT I would like to say - I think you're onto a very important track because the a bike community really needs some objective, quantified, authoritative data about where/when accidents are happening - seasonally, night/day, weekday/weekend, etc.

Swalfoort's been leading on this with the BikePgh accident map, and I'm really impressed at his work and his initiative. And I'm really disappointed that his bottom-up outreach to the various police authorities hasn't been embraced.

How do we take this onetime infodump (which is a gift) and turn it into the basis of an ongoing learning?

I'm wondering how we build a joint-effort database between BikePgh, the City, the County (since the County is really the government entity that contains these accidents), and PennDot so that (1) bike events are reported routinely as a requirement and (2) the results come with the imprimatur of a Gov-Org partnership.

Both of those would build the credibility of the message, and a public front-end would make bike safety transparent to the community and accessible to reporters students etc working on a story.

This could be a positive response to the no-good terrible very bad week we had recently, a feather in the cap / photo-op for the pols involved, etc.

Is this worthwhile? Is it pragmatic?

If yes/yes, how do we do this?

Swalfoort's a lady dude. hee hee.

But to add to Vannevar, as demonstrated in the Trans Alts article, at the most basic level, let's also collectively push to have these "incidents" officially and unofficially referred to as something, anything other than "accidents". Bicycle wrecks, crashes? Anybody? Who's a linguist?

edit - collision. I'm liking collision.

Please excuse my faux pas, Swalfoort. No disrespect intended.

I believe a couple of other times in other threads stefb has indicated that the hospitals have stopped using the term "motor vehicle accidents" and are instead now using "motor vehicle crashes." The implication is that "crashes" is more inclusive than "accidents" -- not are crashes or collisions are necessarily accidents.

Just saying there is precedent for edmonds59's line of thought. I think either term would work, but "crashes" may have already been chosen at least in the medical world.

stefb, can you refresh us on this?

Euphemistically they're "events", but what makes the event significant is the "collision".

You could go all 1984-Milspeak-hybrid and use "kinetic event".

"Crashes" sounded tooooo self-initiated, or something, to me.

First off.....no disrespect assumed. I sort of like being genderless on this board. Many people know me through my work, or through events, so have figured out I am female. But for those of you who assume otherwise, that's ok too. (BTW, my user name here is first initial, then last name, so Sara Walfoort leads to Swalfoort.

As for the data, I think this group is right on. We have a gift here. I have similar gifts from other sources. I am using the Bike Pgh Crash map as well, although that is sort of glitchy.

More importantly, I am trying to work with PennnDOT to identify when data becomes valid and usable.

PennnDOT reported accidents clearly are. But we all know that they only tell part of the story due to the particulars of when an accident is reportable. What about the self reported incidents on the bike pgh map? Not quite so clear.

Police nnow have a disclaimer on the top of their daily incident logs to the effect that this log if for informational purposes only, and they are not responsible for how the data might be used. I think this is new, but I need to go back a month or two to verify that. If it is new, it COULD be in response to my/our requests that accidents involving bicycles be reported more consistently and more accurately. But, I am going by memory here, and I might be wrong. It could be that I just don't remember the disclaimer being there all along.

BTW, we are stuggling with the same "when does data become valid and usable" question with the bike count program. We can only do a limited number of counts in a given year. We choose to do a Spring and a Fall count of rush hour cyclists. We opt for the "typical" situation, counting on one day at each location for the morning and afternoon rush hour. We do this for about 2 dozen locations throughout the City. This is an effort that could not be completed without the generous help of the Bike Pgh, SCA and other volunteers, BTW.

So, after five years or so, we have trendline data for a couple dozen locations. We can factor for weather, etc. We seem to be seeing more cyclists than we did two years ago, at pretty much all locations. But, this is based on 8-10 observations at a given intersection. It that representative of true conditions? As what point does the data become meaningful?

Any of you statisticians want to help with that question? I could really use your insights and experience on how to best use and develop confidence in a vey limited database.

Thanks! I will continue to worrk with the count data and the crash data that we have. I also monitor local news reports of incidents that make it to the media, but not police reports or the bike crash map.

To what purpose, I am not quite sure, but I know the direction I need to take it, and I think I know the ultimate goal. I think....

I am a statistician but I am going on medical leave through Oct 1. Can help out after I am cleared to work again.

squesen407 at yahoo dot com

sheesh, I missed a most interesting turn in the conversation. Statistics are awesome, I look forward to seeing the results.

(fwiw, Swalfoort, when I read your userid in my head, it's said by a male viking in a menacing way that somehow also makes me giggle. However I hear your messages in my head in your normal IRL voice.)

Back on the PG pages, they published two more LTE today, both by BPGH members (one was me):

http://www.post-gazette.com/stories/opinion/letters/bike-trail-trials-651276/

http://www.post-gazette.com/stories/opinion/letters/irrational-arguments-651252/

All hail the Queen!

We are thoroughly amused

Good letter ej.

I felt it necessary to comment on the other.

@ejwme - I love the voices in your head! I hope I did nothing in my explanation of my user name above to dispel that male viking voice. I like that imagery!

Swalfoort - oh no, everything you write/wrote simply encourages them to continue, a fact of questionable merit.

(for comparison, internally edmonds is introduced by The Fonze, and Mick is introduced by the Monty Python's HG "Tim"... maybe I should stop while I still have friends...)

I base my BikePgh head-voices on the avatars, not the names. So it was really disappointing to me, in a way, when Mr. Marvelous was identified in a recent photo on Tag-o-Rama and I found it was really unlikely that he had the high squeaky voice I'd assigned to him based on his avatar. Now I imagine a somewhat deeper voice. As we grow, our illusions are stripped away, one by one.

See, Mr. Marv is mighty mouse (the deep booming voice, likely inspired by the avatar). Jonawebb, yours is an average american male, but it has a huge echo because you're inside the whale. Sigh.