If you could cycle exclusively on cycle tracks, would you?

[vickster]

Very true. Last year I rode through Burgess Hill in Surrey. This is a town I've ridden through a number of times, normally by just heading in to the town centre and out again. This time I elected to try out the cycle path route. To my surprise it was really jolly good, and made a nice change.

But my usual reaction to such an option is: "cycle path? yeah, right. It will be awful".

Pedant alert... Burgess Hill is in West Sussex not Surrey

 

[Dogtrousers]

Pedant alert... Burgess Hill is in West Sussex not Surrey

Sorry

 

[mjr]

There's a problem with saying "Statistically it's safer to ride on the road, so do so. Off you go." That is, it excludes all of those potential cyclists who do not feel safe on the road, and if that's all that's on offer then no thanks.

Not only that, but it's also less fun and often farking depressing riding among faster, heavier vehicles spewing out fume and particulate pollution. When on the cycleways, I basically just don't have to care about them unless I hear the fateful sound of tyre on kerb or the distinctive revving of a motorist doing tight turns to get onto a cycleway that they shouldn't. Most of the time, all I have to think about is the next junction (and there are fewer of them than on a carriageway where basically we have to worry about junctions on both sides) and can just pedal along at my own speed nearish the left edge, ideally with occasional looks back to see if another rider will overtake soon. If I want to get past someone, we can usually talk to each other (noisy motorists permitting) or a bell can be far more expressive than a hooter. If I want to stop and look at the blossom or whatever, other cycleway users can get past me.

Some would-be cyclists will not become cyclists if they have no option but to mix it with the traffic. Those of us who are comfortable, or comfortable-ish with riding alongside other traffic need to recognise that ours is a privileged viewpoint and that there are those who find cycling in traffic intolerably stressful.

There can be a certain fun to sometimes going jousting in motor traffic, which I expect appeals to some (including some famous youtubers), but I'd hate to do that every ride. Too many drivers don't seem to give a shoot that their vehicle can kill others if they fark up, and they're on their phones, eating lunch or trying to light a spliff while hammering along at 50 in a 40 zone.

 

[vickster]

Sorry

don't want you getting lost

 

[bobzmyunkle]

Last year I was in Normandy and Brittany and tried some bits of cycleway - Velo Francette, EV1 etc. Very scenic with all sorts of French family groups out touring on all sorts of bikes - including those ones you see in Decathlon that don't look like a bike a sensible person would ride.
What I did notice is the routes bypass towns and villages, and even open countryside - i.e. the sort of typically French stuff I might want to experience. So exclusively - no.
Cycle tracks in this country - generally a no. Routes often have a Sustrans input so tend to vary in quality greatly and often meander 5 miles to avoid 100 yards of main road. The latter is ok in some cases and just plain daft in others (category can vary from person to person and according to my mood).

 

[matticus]

What I did notice is the routes bypass towns and villages, and even open countryside

Are they in tunnels?!?
(not a TOTALLY facetious comment, as I believe whole networks of indoor/underground cyclepaths have been proposed for cities with crazy winters ...

 

[mjr]

It's John Franklin (author of Cyclecraft) who has collated all the studies on cycle path safety. You can find the list on his website.

Far far far from all. For example, note that the most recent study is from 2001.

1) They increase the number of vehicles crossing one anothers paths:

How do you figure that out? The car driver is crossing literally every point of the paths of the two carriageway cyclists until the turning, after which it crosses the path of the cyclist on the cycleway at one point.

I'd like everyone to note that drawing shows a current, valid (but second-choice) junction layout, with the give-way markings behind the cycleway.

2) They create confusion by increasing the junction complexity:

Who cares how complex the aerial view is, if it works? I'd bet that crossroads has fewer lower cycling casualty rates than the 1970s-style mixed equivalent still being built in England until recently. Can we really not have nice things because drivers are too thick to cope with traffic lights and lane markings?

3) They dilute a driver's sense of responsibility by removing cyclists from "his" road

Any evidence for that one? I find it difficult to believe that it could get much more diluted.

4) They reduce visibility:

Foul! Unlike the drawing in point 1, whoever drew that had to exaggerate by putting the minor road give-way markings incorrectly between the cycleway and carriageway.

Even with the give-way markings in the correct place, a layout like that should never be built in the UK because the visibility would not meet the requirements of CD 195 of the Design Manual for Roads and Bridges. If that still exists anywhere, it's now an obvious easy-to-highlight safety hazard that should be corrected, same as any blind carriageway-carriageway junction. We could spend all day arguing how dangerous things are based on old mistakes, but it's not that informative beyond a conclusion "we need to stop mistakes being built and correct old mistakes".

5) They create needless conflict about who has right of way:

That conflict exists anyway, hence the "applies whether cyclists are using a cycle lane, a cycle track, or riding ahead on the road". Has anyone who has been riding regularly for more than, say, 5 years NOT been "left-hooked" by an idiot at some point?

6) In the case of painted lanes, they reduce passing distance because they turn the issue from one of leaving cyclists enough room to wobble into one of territory and borders:

I don't think anyone here is in favour of painted lanes, but I can see easily that those lanes are narrower than the "absolute minimum" now permitted and closer to parked cars than permitted, assuming typical 60cm flat handlebar width.

Even the "Baseline passing distance" looks at least 30cm less than the 1.5m minimum for 30mph in the current highway code and even that car is only about 1.6m wide, so more like a Fiat 500 than the sports car its shape suggests:

Let's discuss reality, not Fantasy Crayon World where the cars and bike lanes are too small and the carriageway lanes too wide.

So that's just three out of a total of 36 studies you've addressed, and then not with any evidence, you just make the assumptions that it suits you to make.

So? Your points above seem to be evidence-free assumption-based spin. Some don't even stand up to simple geometry, yet you want me to spend my time digging out some detailed evidence for you? I'm slightly irritated that I've even spent time replying, but I don't want to risk people reading this stuff and thinking it's accepted.

As Franklin says:
"This list is intended to be without bias, but little evidence has been found to suggest that cyclists are safer on paths than on roads. If you know of other research, please contact John Franklin."

If you have the evidence, why not send it to him?

I have. He doesn't care. Surely you can tell where he stands, if not from his self-authored dodgy Redways papers in the list, then from the utter garbage in the cycleways chapter of Cyclecraft that would cause conflict and probably cycle-cycle crashes if many people actually rode like that.

I think the Camden scheme is the only one that's been subjected to a Public Inquiry isn't it?

The Camden scheme? I'm pretty sure there's more than one cycleway in Camden and I'm not sure what you're quoting from there, so cannot answer.

This is probably why riding cycling infrastructure in the UK is statistically more dangerous than riding on the road - cos its sheet!

But is it "statistically more dangerous"? Most evidence I've seen is that it's basically little difference. Cycling is pretty safe in general.

 

[Dogtrousers]

Thanks @mjr I noticed a lot of the inconsistencies in those diagrams that you pointed out. But I couldn't be fussed to list them. Partly because I'm conflicted on the question of segregated infrastructure, I don't really have a dog in this fight. But those diagrams didn't convince me of anything. They look like straw men. "Look cycle lanes are crap. Here's evidence" (insert diagram of how a crap cycle lane would look).

 

[Drago]

@MKR indeed, as of a couple of years ago (was pre-lockdown, so possibly 2019) CUK was reporting in their chipwrapper that the roads were indeed slightly safer than dedicated infrastructure in terms of reported casualties.

Please don't ask me to report on what they said about journey-mile adjustments to the calcultation and stuff after so much time has passed.

Agreed. Cycling is pretty safe in general.

 

[bobzmyunkle]

Are they in tunnels?!?

Disused railway - often in cutting etc

 

[Mike_P]

Some cycle route planners have a cycle route mapping layer (Komoot) for one. It's a pity on one has done a layer for ease of cycling on A and B roads. At the top of the pile you could have those with a bridleway lane in each direction and at the bottom those that have fast moving traffic on a narrow twisting undulating carriageway.

 

[matticus]

Disused railway - often in cutting etc

Ah right - like a lot of UK ones. Gotcha. (I'm quite likely to try the 'Verte thing to Paris in August, but I'm not averse to using the roads in that part of France so we'll see ... )

Now you see that doesn't happen in NED. Their routes actually get you to/from real places. Such as towns where you can also cycle round safely, and so on ...

 

[mjr]

@MKR indeed, as of a couple of years ago (was pre-lockdown, so possibly 2019) CUK was reporting in their chipwrapper that the roads were indeed slightly safer than dedicated infrastructure in terms of reported casualties.

Please don't ask me to report on what they said about journey-mile adjustments to the calcultation and stuff after so much time has passed.

I suspect the key word there may be "reported". Even in Norfolk and Cambridgeshire, we sometimes hear of police refusing to accept reports of cycle casualties because no motor vehicle was damaged or motorist injured. And the current collision reporting form still doesn't explicitly ask whether a cyclist at a junction entered it from dedicated infrastructure or not, only "vehicle location at time of accident", so allocating mid-junction collisions to cycleway or carriageway is rather trying to rebuild the pig from the sausagemeat: a problem Franklin solved in one of his 20th-century papers by allocating all junction collisions to cycleways, ignoring that there are riders who never use them.

 

[mjr]

Ah right - like a lot of UK ones. Gotcha. (I'm quite likely to try the 'Verte thing to Paris in August, but I'm not averse to using the roads in that part of France so we'll see ... )

Now you see that doesn't happen in NED. Their routes actually get you to/from real places. Such as towns where you can also cycle round safely, and so on ...

Oh, there are ex-railway routes like this one, but they still go between small towns and, for some reason (🌊?), they don't do deep cuttings in most of the Netherlands, so you get more views of open country than you could shake a 🌷 at.

 

[mjr]

Some cycle route planners have a cycle route mapping layer (Komoot) for one. It's a pity on one has done a layer for ease of cycling on A and B roads. At the top of the pile you could have those with a bridleway lane in each direction and at the bottom those that have fast moving traffic on a narrow twisting undulating carriageway.

I'm a fan of the CyclOSM layer, also viewable on the main OSM and on cycle.travel at the moment. It tries to emphasise easy cycling in blue and deemphasise fast/big roads more than the OpenCycleMap or cycle.travel default layer. I don't think it includes amount of traffic or any measure of twistiness. I think @Richard Fairhurst uses traffic and gradient data as inputs to the cycle.travel routing engine, but I don't think traffic is used in its default layer map-drawing at all, which is probably why motorways and quasi-motorways are still far more visible than I'd like.