@RespectTO on Twitter claimed that "Bloor has the worst rate for collisions in #biketo. Safe passage now. #topoli #BloorBikeLanes". They were rightfully celebrating the fact that we're one step closer to getting bike lanes on Bloor. Maybe I should have left well enough alone but I asked them: ".@RespectTO explain "worst". Gross numbers or per volume of cyclists compared to other streets?" It was an honest question but hard to explain intention in Twitter, thus my blog. How would we know Bloor Street is the worst and by what definition and measure? @RespectTO told me that it's according to City stats but they didn't have a link to share.

Let me just be clear that I'm quite happy that Bloor is getting this focus and deservedly getting bike lanes (hopefully) to help make people safer. But I think it's worth digging into this a bit. In the past I have dug up and analyzed statistics to help answer such questions, so let's do the same here. 

The claim: "Bloor is the worst for collisions". How can we make the claim testable? We'll need to compare Bloor Street to other streets, obviously, but we can't just use the entire lengths of streets because then we'd be comparing apples to oranges. We need equal measures. More importantly, the Bloor Street bike lanes pilot is only going from Shaw to Avenue so that's the only stretch that might become safer post-install. So let's compare the crash stats of Bloor from Shaw to Avenue to an equal length on an other street. I chose Queen St West, partly because it passes through the same cross streets as Bloor, going from Shaw to University, and because it's also a well-used route by people on bikes. (I could have also chosen Dundas, King or College for similar east-west streets.) Then I think if we just count all the collisions on Bloor and Queen along the same stretches we will know which one has had more collisions.

Note that there are also different ways to interpret "worst". In the gross sense, we might be focused on the total numbers of crashes and trying to reduce them. From the point of view of an individual person on a bike, however, the concern would be about the risk of a crash, that is, which street is more "dangerous". In this latter case, we're more concerned about the chance of a crash per person. And for that we'd need to divide our crashes by the volume of people cycling since that can be quite different depending on the street. But this will be too much work and it wasn't the original claim, so I'm just going to focus on the gross numbers.

The closest we have to crash stats by location is this map from a few years ago produced by the Globe and Mail, mapping out all collisions from 1985 to 2010.

crash stats -- which is the "worst" street?
Which street is the worst for collisions? Not so simple


So I just started counting the dots and this is what I got:

Total collisions on Bloor Street, Shaw to Avenue: 261 collisons. 1 fatal, 14 major injuries (1985-2010)

Total collisions on Queen Street West, Shaw to University: 308 collisions. 1 fatal, 19 major injuries (1985-2010)

So Bloor and Queen are pretty close with Queen just edging out Bloor. We can speculate to why: maybe the average cycling volumes on Queen are a bit higher? Or maybe the streetcar tracks make Queen a bit more dangerous? We can't really know unless we have more data.

I'd actually like to see if the numbers on Queen have dropped since the protected bike lanes on Richmond and Adelaide were installed, creating a continuous route from Shaw to University (and beyond). With people going to those alternative routes the volumes on Queen have dropped and hopefully the overall crashes. And I hope the same happens on Bloor.

At any rate, I think we can conclude that the original claim by @RespectTO might have been true at some point (though I'm not sure) but it doesn't seem to be true anymore. But I will celebrate with them that we're finally getting bike lanes on Bloor.

[Update: I decided to count College Street between Shaw and University and came up with 259 collisions. 0 fatalities, 11 major injuries from 1985 to 2010. There are painted bike lanes along about 2/3s of the stretch, no physical separation. I believe the bike lanes have been there for the whole period, I'm not certain. It's not a surprise that College numbers are this high since it is also the busiest street for bike traffic. So if the City is serious about reducing collisions they could upgrade College. A painted bike lane is obviously not enough.]

Bloor bike lanes pilot area

Last night the City presented the preferred option for the Bloor Street bike lanes (pilot) (see more coverage here, here). I couldn't make it but looks like it was option C, which was the one with the most extensive physical separation between people biking and driving. That's nice and quite a surprise actually. I had expected more compromises and we got fewer. It is indeed something to be celebrated given the large political barriers overcome (and still to be overcome). Councillors Cressy and Layton should be commended for pushing this and taking a political risk. A key part of this is the new cycling manager Jacquelyn Hayward Gulati who is willing work hard at taking advantage of this political opportunity. Previously the Cycling Unit has often been content at not upsetting the apple cart and were more than happy to delay the Bloor bike lanes by agreeing to unnecessary Environmental Assessments.

In terms of the design there are two main items which I feel are compromises:

  • the stretch between Bathurst and Spadina will only have flexiposts on the one side, which was given up to preserve car parking on one side
  • a number of people were asking the staff to take a risk of trying Protected Intersections on Bloor but looks like they've decided to stick to the substandard approach of forcing bikes to "intermingle" with cars at all intersections. This approach has already proved to be problematic on Richmond and Adelaide and does nothing to try to reduce the worst locations for collisions: intersections

That being said, this will still be a game changer as much as the protected bike lanes on Richmond and Adelaide were game changers, having a huge impact on people flowing into, through and out of the core. It's only been 40 years!

It may be winter still but the snow has left (again) and you can start thinking about doing some maintenance. My wife sometimes trusts me to work on her Dutch bike, which features an annoying to remove chaincase which also happens to be quite good at keeping most moving parts away from dirt, water, salt. A while back I was trying to find out how to remove the chaincase on my wife's Dutch bike. The info out there was sparse, even in Dutch. So I had taken some photos with the idea of posting them so they may this help someone else out there. Then I promptly forgot about them. Until now. So I'm reviving this for your benefit. You're welcome.

You don't need to flip the bike over, but for these photos I did to get a better view.

chaincase exposed


There are at least a couple varieties of chaincases on Dutch bikes. This one is traditional with frabric covering a metal frame. The frabric is held together with a wire woven between hooks alternating along the edges of the fabric.

Undo the wire by pushing it away from the hooks. Don't push so hard that you bend the wire. It has to maintain some stiffness so you can put it back together.

undo fabric on chaincase


Wire undone and exposes the chain and crank so it can clean them or replace the chain. A chaincase will keep things fairly clean, but eventually even a fully-covered chain will get dirty.

Undo the wire


Pull up the little doohickey that holds together the two sides of the fabric.

pull rear chaincase to undo


Pull back bend of the metal chaincase so it separates. It might need a bit of tapping from a wrench if it's stuck. Just look at that chain—still quite clean and no rust. It's just the cheap metal frame that's been rusting a bit. Salty Toronto winters are hard even on Dutch bikes, but particularly on bikes where all the bits are naked to the elements.

undo wheel axle


If you ever need to change the tire or wheel, remove the nut and pull the braces to get the wheel loose. You may also find you need to tighten the chain by tightening that small bolt that points towards the back. This is a chain tensioner.

undo coaster brake arm


Undo the arm for the coaster brake. At this point you'll be able to remove the wheel and change the tube, tire or fix something else such as a broken spoke (thankfully rare on a sturdy bike like this).

To get it rideable again, reverse the process. Put the wheel back on and tension it. Fix the coaster brake arm back to the chainstay. Slide the metal curve back onto the end of the chaincase. You may need to tap it gently with rubber mallet.

weave fabric back together


Pull the flaps back together and start weaving the wire back on between the hooks. It can be difficult to pull the plastic fabric all the way back. Be persistent but careful so that you can get the snap shut. The snap is the hardest. It isn't crucial but it will expose more of the chain to the elements.

Good to go -- fingers crossed!


All back together! If you did things right it'll look like this. (Notice I gave up on trying to close the snap.)