Hacker News new | comments | show | ask | jobs | submit login
A Type of Road Junction that Kills Cyclists (singletrackworld.com)
569 points by thebent 11 months ago | hide | past | web | favorite | 507 comments

I feel a lot of commenters are simply focusing on what signs can do or what sort of junction to build, rather than focus on the fact that on high speed primary country roads, bicycles shouldn't be sharing the road with the cars.

In Denmark, a lot of high speed roads (excluding motorways, of course) out in the country have separate dual-way bicycle lanes near it but not attached to it. (Example: https://i.imgur.com/dS6jqXS.jpg)

That way, the cyclists can cross the side roads on their own accord, where they are more visible and have their own junction with the side road. Additional, one can set up a traffic light that only turns red for the cars when a cyclist is crossing (i.e. activated by a button).

Furthermore, one might consider signage that warns drivers about cyclists in the junction: https://i.imgur.com/CX6SJdW.jpg

Also, a way to reduce speed of the motorists without putting in stop signs would be to add chicanes just before the junction, so they are forced to slow down.

Plus, as I've mentioned before (see https://news.ycombinator.com/item?id=15977162), I also think it is because UK drivers don't really have enough training with how to deal with bicycles and the fact that they are also participants on the road.

Most roads in the UK are perfectly well designed to handle bike traffic. But they are not designed for the massive quantity of vehicle traffic travelling at relatively high speed on roads originally designed for a different age. Danger to cyclists are a symptom of this wider problem which we have no real solution for.

Also, this particular junction is terribly designed for cars and will result in collisions.

I disagree vehemently. While my experience on UK roads isn't that big, I have not seen any well designed bicycle road. Even the roads in the residential areas I've seen are quite bad; small, many cars parked on them (bad for visibility) and of considerable lacking quality. The bicycle roads in London are just lines painted on top of whatever surface the route encounters: curbs, manholes, poles in the middle of the road, pedestrian areas... Sudden and unmarked bicycle road stops where cyclists can't make out where to go. See also: https://ssj3gohan.tweakblogs.net/blog/11985/city-cycling-in-...

No, if anything, UK roads are specifically not designed to handle bicycle traffic. I wonder what makes you think so?

I was referring to the normal road itself, not the bicycle specific infrastructure. London for example would be fantastic for bikes if vehicles were removed. Most of the roads (even the narrow ones) are exactly the right size to accomodate a large number of bikes, but instead it is used to accomodate a very modest number of private cars. Even if there were no bikes or pedestrians our roads would still be wholly unsuitable for the volume of traffic. The country is riddled with narrow and windy single lane roads with very high commuter traffic.

Yes cyclists in the main have to share the same network. And that network in many places is just overburdened. My fastest route to town is down a country road that has the national speed limit. Local drivers complain about cyclists even using this road - because it 'slows' them down. Anyway the road has become hellish, and I've witnessed enough after/accidents to just bypass the road these days for an inferior, but safer dirt track. The irony of course is that that very same dirt track could make for a very good cycle lane. Build them and they will come. The pedestrians/country side ramblers won't be too impressed mind.

Roads in the UK have generally had some work done on them since the automobile was invented.

Saying that they are designed for bike traffic, when they are continuously rebuilt to the needs of motorised vehicles, is a bit like complaining that it is problematic marching a centuria down the middle of the A59 out of York, given that is what the Romans designed it for.

And we do have a solution to this problem, it is that of creating dedicated bike routes, like the Bristol to Bath cycle path, or where they have to run alongside roads with cars doing 60mph and above, keeping them separated by more than just paint.

Dude, the A59 would be great! You could march XX Valeria down that thing abreast.

In fact, that may be what happened to IX Hispania: they were having so much fun, they just kept going.

> a bit like complaining that it is problematic marching a centuria down the middle of the A59 out of York, given that is what the Romans designed it for.

Thank you, this observation has brightened up an otherwise mundane Wednesday.

I hope I haven't ruined the mundanity. Is what Wednesdays are for.

I mean, you say that, but every time I see cyclists on an A road or pedestrian crossings on a 2-2 lane highway I die a little inside

Assuming you see them when you're driving, it's probably the sedentary living that's killing you rather than other road users.

You can live again! Get out of the car.

There’s a subset of drivers in the UK who have a resentment to sharing the road with cyclists. Their primary arguments follow; “car drivers pay road tax...” or “cyclists should be forced to sit a driving/road test...” blah blah.

There are certainly cyclists who do nothing to help the reputation for the rest but car drivers fail to realise how vulnerable cyclists and motorcyclists are.

Why shouldn’t cyclists need a license to ride on the road? Others road users must have one.


It's a shame I'm getting downvoted, but the current state of affairs is that in cities like London bicycles, cars and trucks will be sharing the road for the near-mid future.

I'd rather not have trucks or cars in the city, and ideally cyclists would have their own road separate from cars and pedestrians, but I think that's a pipe dream.

Until then what else can do we? At least if we require tests for cyclists, like we do for cars, buses and trucks, it might help reduce the number of cyclists doing risky things that perhaps they aren't even aware of?

Sure there are still reckless drivers, but at least they have a license that can be revoked and points that can be fined. How do we ban reckless cyclists from the road?

> Why shouldn’t cyclists need a license to ride on the road?

Thought experiment: if you didn't have any motorised vehicles on the road, would you still want cyclists to have a licence?

In a general case though, requiring cyclists to have a licence would reduce the number of cyclists, probably significantly. This would correspondingly increase the number of motorists.

This dual effect would have a number of negative impacts. More motorists lead to more congestion, more CO2 emitted and poorer air quality. Poorer air quality leads to health problems, up to and including deaths. CO2 causes climate change, and greater congestion has an economic impact. Fewer people cycling mean fewer people getting exercise, leading to more health problems, up to and including deaths.

I'm confident these problems would overshadow the small improvements in cyclist safety you might gain with a cycling licence requirement. There are better ways to improve cycling safety.

> Thought experiment: if you didn't have any motorised vehicles on the road, would you still want cyclists to have a licence?

Of course! In a world where there were only cyclists there would still need to be a set of rules that all cyclists are required to follow. Those might include how to behave at intersections, how and when you should overtake other cyclists, and so on. To be allowed to cycle, each cyclist should have demonstrated that they understand and follow these rules by taking some kind of test.

> In a general case though, requiring cyclists to have a licence would reduce the number of cyclists, probably significantly. This would correspondingly increase the number of motorists.

I don't think so. I expect the government to continue to penalize use of motor vehicles in cities such as London. The bicycle will remain the cheaper mode of transport.

> Of course! In a world where there were only cyclists there would still need to be a set of rules that all cyclists are required to follow.

What about walking? People can run into other people, knock them down stairs, it's a risk...

I think that most observers find the idea that bike use is a public hazard, and must be restricted to be obviously ludicrous.

I don't. In the United States, bicyclists must follow largely the same rules of the road that drivers of motor vehicles must follow. They can even be ticketed for violating those rules although usually the tickets don't have much in the way of teeth (except things like DUI while biking).

Here in the Bay Area, it's a rare bicyclist that actually follows the rules. Talking to other bicyclists, I've found a large number of them are simply ignorant to the fact that the rules apply to them as well; or they have adopted some bizarre version of the rules or a sense of entitlement.

I could talk anecdotes for hours, but one particularly egregious example is a bicyclist who showed up to a 4-way stop after I was already moving through the intersection, cut through my path, and gave me the finger, all without even slowing down, much less stopping.

Requiring bicyclists to pass at least a basic test would go a long way towards clearing things up or at least changing the "bicyclists don't have to follow the rules" culture that exists here, making the roads safer for themselves, walkers/runners, and drivers alike.

Forcing bicycles to follow rules designed for cars is often silly and sometimes actively dangerous. For example, the bike route from downtown to my house involves making a left turn on a busy street. For a mile and a half, the street is nice and wide, with generous bike lanes in both directions - sensible and safe. But the moment I have to turn left? By car laws, I need to stop in the middle of the road, with traffic coming up behind me (often at 40+mph), and wait until I have a clear path to turn left, which can take a while. This is deadly.

So I do the illegal. I turn right on the quiet side street and immediately u-turn to the stop sign, to wait to cross both sides. It's not a dangerous maneuver for me or anyone else, although it would be dangerous for a car to do so.

See the problem?

edit: A car waiting to turn left is in considerably less danger. It's much easier for traffic coming from behind to see, so it's less likely to be hit. And if it is hit, the driver is protected by the car itself. At worst, the car gets totaled. But a bicycle hit at 40mph? The cyclist is likely going to be killed. It's not just silly to follow the car law - it's hazardous.

That u-turn sounds legal for everybody.

I think one the purposes of traffic law is to make the behavior of all participants predictable. My only issue with cyclists is I feel I can never predict what they're going to do.

"That u-turn sounds legal for everybody."

There's usually laws about how far you can see, how far you have to be from an intersection etc.

"My only issue with cyclists is I feel I can never predict what they're going to do."

This is not entirely the fault of the cyclists, it's the fault of the law givers

The laws are not written in such a way that it appears safe to people using bikes to follow all the rules, so everyone is compelled to come up with their own rules. The build of roads is the same; a small detour for a person in a car can become quite substantial for a person on a bike.

At the moment the laws get written and the roads get design assuming almost everyone is going to use a car, and then they say "well, is it physically possible to follow it on a bike?" - if they consider bikes at all.

If governments wanted people to ride bikes predictably, they must consider them when crafting the laws. It's probably quite okay for people riding bikes to act differently than people driving cars: just like people walking around behave differently. It will then take some time to spread the word and regain the trust of cyclists. But it's a more helpful solution than saying "they must follow rules and infrastructure designed for cars in the same way as cars, except that they must not delay me: that way they will be predictable and safe".

Depends on the local law. In some cities, this is bike-legal, and even officially encouraged as "indirect left turn": see the side-street bike box, intended just for this: on your green, you cross the side street, turn right-then-left into the box and wait for a green signal to cross the main road. https://prahounakole.cz/wp-content/pnk/uploads/2013/03/pruhy...

Your u-turn sounds legal, even in a car, but I'd have to see the area and know the state to be sure. Regardless, bicyclists are already required to follow the law. Having a required license would simply help make sure bicyclists are aware of the law and provide a tool for law enforcement to keep the most egregious offenders (e.g. stop sign runners, cyclists traveling against traffic, etc.) from causing problems.

Why? Requiring a license doesn't seem to make car drivers any better. And most bicyclists already have a driver's license. Those who don't tend to be kids, or poor (having a driver's license is privilege, although a common one). Forcing them to need a license will just create more illegal bicyclists.

Law enforcement can already ticket a bicyclist, license or no license. Bicyclists already know when they're breaking the law.

I disbelieve your proposed solution would actually solve any problems.

> Why? Requiring a license doesn't seem to make car drivers any better.

Then let's advocate for getting rid of licensing since it has a cost and no benefit. I don't think you really mean that, do you? I've got plenty of ideas for a better licensing system for driving.

> Forcing them to need a license will just create more illegal bicyclists.

Illegal drivers face consequences that licensed drivers do not face. The same would be true for illegal bicyclists. I fail to see this as a bad thing.

> Bicyclists already know when they're breaking the law.

They don't know. They might know if they did it in a car they would be breaking the law. They also believe themselves to be incapable of creating a dangerous situation. (Quick anecdote: I was on a jury for a civil case where a bicyclist admitted on the stand he ran into a car that, by all evidence and even his own testimony, was stopped. The bicyclist had also run a stop sign. He felt the driver was at fault for the accident.)

Being a bicyclist myself, I assumed everyone knew as well. Back home, all my cycling friends knew and followed the rules. When I started cycling with people here, I found out they really don't know that the rules apply to them, they think the rules are different in some way, or they just have some really bizarre notions. AFAICT, it's baked into the culture. I mean, people don't know basic stuff like that you can get a DUI on a bike or that you can't ride on the sidewalk...

A bicycle-specific license (or maybe an endorsement on a DL) could focus on bicycle-specific issues, with basic questions like "I have to stop for a stop sign on a bicycle. T/F" and more 'complex' stuff like how bicycle lanes work at intersections with people making right turns. Drivers and cyclists alike seem to have no clue that the driver should enter the bicycle lane (after yielding to bicycles in the lane) for his turn and that the bicyclist should not try to pass him on the right.

Not sure where that is - out here (central Europe) the examples like "no DUI or sidewalks allowed" are known almost universally. Of course, knowledge won't stop you - a completely drunk (0.2 BAC) driver just killed a pedestrian a few days ago: he knew it was blatantly illegal, had a valid driver's license to certify that, but went out for a drive just the same.

Anecdotes are not data.

It's a rare driver that actually follows the rules. Speed limits are almost never followed. I frequently see drivers go through a light that has already turned red, roll through a stop sign without fully stopping or come dangerously close to hitting a pedestrian to whom they were supposed to yield the right of way.

Drivers routinely kill other drivers or pedestrians through carelessness. Few if any cyclists ever kill someone else. If anything we need more rigorous tests for drivers, not for cyclists.

What I loathe most about cyclists as a pedestrian, (I am also a cyclist and also a driver) is when they ignore pedestrians crossing a street and simply swerve between people --at speed. Kids are unpredictable and they don't have the same sense as adults, yet, I've had prickish cyclists ride on through as if nothing.

In a world where there were only pedestrians, would you want them to start having licences? Per mile, walking is exceedingly dangerous compared to other modes of transport.

No, I don't think so?

1. Pedestrians move at much slower speeds than bicycles. If two pedestrians walk into each at 5mph the damage is much lower than if two cyclists collide at 30mph.

2. Since pedestrians move much slower than bicycles they can manoeuvre much faster. The stopping distance for a walking pedestrian is less than 1m, the stopping distance for a bycicle travelling at 30mph is much greater.

3. If two pedestrians walk into each other the chance of collateral damage is quite low. It's quite easy to walk around two people that just walked into each other.

If two cyclists collide, there is a good chance that cyclists on each side of them will get caught up in the collision, crash and injure themselves too. A good example of this are the crashes seen on Tour de France.

Of course, if people started running as a mode of transport it would become increasingly dangerous, and you'd have to start separating the walkers from the runners as the damage from a potential collision would be much greater.

> A good example of this are the crashes seen on Tour de France.

Basing rules for cyclists on the Tour de France is like basing rules for automobiles on F1 or Nascar. Normal commuters just don't reach anything close to Tour de France speeds. Most would be lucky to even maintain a third of that speed without the help of a hill.

Most commuters will be going about 10 mph. Even getting hit with a car at that speed will only cause an injury at all about 25% of the time, a serious injury about 15% of the time and is almost never fatal[1]. A bicycle will barely do anything at that speed.

By all means add special rules for using bicycles at unusually high speeds, but don't act like normal commuters are going to be going that speed.

[1] https://nacto.org/docs/usdg/relationship_between_speed_risk_...

Ok, so licences for runners then, as you may need to have a test to see that people know the rules and presumably ban people from running if they break them.

Since running is largely recreational most runners avoid congested areas on purpose. If this changes it will have to be revisited. Nice strawman though.

Was trying more for reducto ad absurdum. I didn't come up with the idea of having laws to separate runners from walkers, you did. I was just exploring the enforcement mechanism for such a suggestion.

Well I think the issue is that unlike the pavement, the roads are much more stringently regulated with sign posts, markings and lights.

For as long as cyclists have to cycle on the roads as we know them a cyclist should be required to demonstrate that they understand the signs and rules of the road. That could be by presenting their official drivers license, or, should they not have one, be required to take a theory test. I think that would be a sensible first step.

You could then imagine a practical test for hazard perception, efficient use of gearing and safe filtering through queues of traffic.

Naturally, people don't walk on the road, and pavements don't have the same kind of mandatory control flow unless they intersect with the road, and most children are taught how to cross the road from a young age.

I walk on the road all the time, having grown up in an area with plenty of rural roads without pavements.

I would get into a ton of shit should I visit anywhere that had jaywalking laws, though in the places they exist and are enforced, they presumably get by without a dedicated licensing and testing regime.

Dude, you're missing the point. You do not walk on the road the same way a cyclist cycles on the road. You do not walk up to T junctions, indicate with your arms and turn left. You neither come to a set of traffic lights in the same way. Cyclists have to do all these things.

And they are somehow managing to do so currently without licences.

One thing to consider is how much effect licensing has on car safety anyway. Mexico city has no test requirement at all, you just pay for permission to drive rather than sitting any test, while Peru has more road deaths per capita, despite needing a practical test, written test and medical certificate before you get behind the wheel.

so what is the conclusion here? Road tests have no positive impact on road safety because Mexico City has fewer accidents than Peru?

No, just that the effect on safety seems minimal compared to other factors.

In so many discussions on HN I see this silly myth of 30mph bicycles.

Did you know that in Sweden, the legal maximum speed of a electric bike is 15mph? Anything faster and it is classified as a moped and you need a driving license. Did you also know that the average speed of a bike in a city in the Netherlands is around 9mph? In context, a common running speed is 8mph.

So here is the stage. Two bikes are traveling more than 300% faster than the average speed and about 200% the legal limit of a electric bike, and they collide. My question would not be if that situation is safer if two pedestrians walk into each other, but rather why two bikes is traveling that fast in the first place.

Not all cyclist ride one speed "gentleman/lady" bikes. Many travel upto 30mph. That's a safety concern. There have been a number of cases of cyclists running over pedestrians resulting in serious injury, so it's not inconceivable to require licensing.

Licensing doesn't stop people driving cars from charging into others, killing them. It's not going to stop bike riders who are inclined to drive with such recklessness that they endanger their own life from driving with that recklessness

> In a general case though, requiring cyclists to have a licence would reduce the number of cyclists, probably significantly. This would correspondingly increase the number of motorists.

Most cyclists are licensed drivers, and a test for cycling licensing would, at worst, probably be no more difficult than that for driving. The impact on the number of cyclists would be minimal.

A bicycle can injure or kill a pedestrian or cause a motor vehicle accident. What is the justification for licensing motor vehicle driving but not bicycling when done on public roads? I cannot think of anyone who fails a motor vehicle licensing test that I consider competent to not cause others harm while operating a bicycle on public roads. I would much rather suffer the 'dangers' of increased carpooling or public transit use.

No, they are not. Atleast outside the US they're not. I know a significant number of people, myself included, who are experienced cyclists who have never driven a car in their life. Cycling is far more common in urban areas where driving is impractical or amongst young people who can't afford or don't want to drive.

>Thought experiment: if you didn't have any motorised vehicles on the road, would you still want cyclists to have a licence?

No. Why do you think this is relevant? It doesn't seem to have anything to do with the rest of your argument.

You are correct in pointing out that it doesn't have much to do with the rest of my argument. I posed it as a question because I think a lot of the opinions around this issue (like many or most issues) stem from emotions. In this case I think this argument comes from a sense of fairness: motorists have to pay tax / have insurance / pass a test; so why shouldn't cyclists? I think if you consider whether we would even be asking that question if there were no motorists, you can get around the "fairness" point of view.

Interesting. I wasn't coming from that angle. I think cycling would be less risky if there weren't cars on the road, so I would see less of a reason to have cyclists take a test and get a license if they didn't have to mix with auto traffic.

You're suggesting that requiring bicyclists to be licensed would kill people?

Requiring licensing for cyclists could indirectly result in more deaths depending on the change in other behaviours.

Off the top of my head, walking is more risky per mile for the traveller than cycling while driving is safer per mile, but increases your risk of obesity and DVT and is more dangerous for everyone else in terms of pollution and accidents.

It would have that effect, obviously. Is this not something you've thought about before?

> Until then what else can do we? At least if we require tests for cyclists, like we do for cars, buses and trucks, it might help reduce the number of cyclists doing risky things that perhaps they aren't even aware of?

Statistics show that in the large majority of cases of accidents between cycles and motor vehicles, the driver of the motor vehicle is at fault. Requiring a license for cycles will do just about nothing to fix this. Just yesterday I was nearly run over by a car turning left into the oncoming cycle lane and by another one who thought that the pedestrians traffic light applies to bicycles. My knowledge that I was correct in both cases did not help the least bit.

So maybe yearly retraining for drivers of motor vehicles would be a better move.

>Statistics show that in the large majority of cases of accidents between cycles and motor vehicles, the driver of the motor vehicle is at fault. Requiring a license for cycles will do just about nothing to fix this.

As a regular cyclist, I would say that requiring a license from cyclists would certainly do something to fix this: make cyclists more aware of it.

Over here, about 2/3 of collisions between motor vehicle and bicycle are fault of the motor vehicle driver. But that leaves 1/3 where cyclist errors or breaches of traffic code are reason, which is also significant.

And then there are actually quite many cyclist accidents that do not involve a motor vehicle.

> Over here, about 2/3 of collisions between motor vehicle and bicycle are fault of the motor vehicle driver. But that leaves 1/3 where cyclist errors or breaches of traffic code are reason, which is also significant.

You are assuming that this 1/3rd is due to lack of knowledge - I question that assertion. And still, regular retraining of motor vehicle drivers would still be a better investment. They choose to move a ton of steel around in a manner that's potentially dangerous to others and they cause twice as many accidents, despite having a license requirement.

> And then there are actually quite many cyclist accidents that do not involve a motor vehicle.

Those tend to be less severe. Last years stats for Berlin: All cyclists killed were killed in an accident with a motor vehicle. Most common causes: dooring, right turn accidents, mostly trucks, one cyclist running a red light.

Likewise, that 2/3 is mostly not due to lack of knowledge. Over here (Finland), driving license training and testing nowadays has increased "attitude" training, as that is seen becoming more important than knowledge of traffic rules etc.

Here, a motor vehicle was involved in 62 % of all bicyclist deaths over the period 2011-2015 (116 cases). Motor vehicle was not the guilty party in all of those (but in roughly 2/3 of them). So, motor vehicle driver was at fault in about 40 % of cyclist deaths.

> So, motor vehicle driver was at fault in about 40 % of cyclist deaths.

Without any breakdown of who was at fault in the other accidents your stats still indicate that in comparison between bikes and motor vehicles, motor vehicle drivers are still twice as likely to be at fault. Retrain them before you retrain cyclists.

In a large majority of the remaining cases it is the cyclists who is at fault. There are a few cases where a pedestrian causes a cyclist death, but they are rare. The list includes those cyclists who are run over by a train, for instance, but we don't really expect trains to swerve to avoid cyclists, and such a collision is the cyclist's fault.

I don't necessarily say that requiring a license from cyclists is what I want. However, I do think that some training and education of cyclists would certainly help. I also see some cyclists who have clear attitude problems. So retraining - or perhaps re-orienting - everyone would help.

I'm pretty sure elementary schools in Finland have a cycling lesson or two. Mine certainly did.

> Retrain them before you retrain cyclists.

Why can't we retrain both? Do we have to pick one or the other?

Because it’s an inefficient use of time and money. All that you’ll achieve is less people cycling and that’s exactly the opposite of what you want. Think about the practicality of a cycle license: do you expect a 6 year old to pass a test before cycling? A ten year old to carry along a license when he goes out? You can’t even fine them because the statuary limit in germany is 14. You can’t fine the parents either because parents are not obliged to keep their kids in sight all the time. Adults are already very likely to have a drivers license, should they now need a cycling license on top.

So make better use of the time and money: create safe infrastructure or - as a better investment train those that already have a license requirement and that insist on moving around tons of steel. Maybe increase the number of hours in school that deal with safe conduct in traffic. Or - gasp - fine the people blocking cycle lanes and otherwise endangering cyclist hard.

So the conclusion here is that cycling is fine just how it is, based on a cost/benefit analysis, and that motor car users need to be regularly examined and educated to ensure good road safety for everyone?

no, that's not the conclusion. I didn't bring up the "let's have a license requirement for cyclists idea." My point is that this is not a change that would improve cycling in a way that outweighs the negative effects. It's a bullshit idea that gets floated from time to time.

Improve cycling by all means, but use something that is effective: better infrastructure, education, enforcement. The Berlin police introduced a police squad on cycles a few years ago in one district and that has proven to be a very effective method: For one, they enforce road rules for people on bicyclist, but the major effect is that they remove cars from infrastructure dedicated to bicycles which in turn actually enables cyclists to use that infrastructure and makes people less likely to ride on the sidewalk, for example.

Education is generally far more cost efficient than building new infrastructure like roads.

No, not in this case. It's not like people don't know the rules and the majority of people get killed despite being absolutely in the right and having done nothing wrong. How do you fix that with education? The dooring accident that nearly killed me happened on a cycle lane that I was by law required to use. The car driver just opened the door into the oncoming cycle traffic. How could you educate me better in a way that would have prevented me from making a hard impact with that door? Witnesses told me they didn't even expect me to get up again. How would educating me help prevent the three close passes I had today, especially the one where the car driver thought I'd have to squeeze into the dooring zone? The one that passed me when I was doing 30 in a max 30 zone? The close pass yesterday where the car driver didn't manage to pass me before the oncoming traffic was there, squeezing me against the parked cars? I was in the right, still I'd be the one dead. Should I yield and crawl? Cycling isn't a sport for me, it's my preferred mode of transport, I really want to get somewhere and I don't want to spend ages doing so.

Calling for education of cyclists has a strong smell of victim blaming. Oh, sorry, he's dead. If he'd been better educated, he might have known that he was in the right.

Because, while it does occasionally happen, the risk of a cyclist killing or seriously injuring another road user is much lower than the risk of a car driver doing so. Unlicensed cyclists are a much lower risk to other road users than licensed drivers; if anything we should be looking to impose stricter licensing requirements on drivers.

They're still a risk. Not knowing the rules can cause a motorist who did absolutely nothing wrong to kill said cyclist or swerve away from the cyclist and kill someone else not even involved, all because of a negligent cyclist.

That said, demanding a license for cyclists I agree is completely unreasonable.

In my country all school children are taught the rules of the road at traffic school where they ride their bikes around a miniature road layout that comprises of most road scenarios.

To be fair that happens in Britain too, or at least it did at both my primary and secondary school.

Could you add some more explanation here? What is unreasonable about it?

What is the actual gain to society from requiring a cyclist license? Today, without licenses, police can and do hand out moving violations to cyclists. Today, drivers who have their license revoked or car impounded still find ways to drive cars, at least in the US. The barrier to picking up a new bike after having yours impounded by police is far lower. If it's forcing a baseline requirement for skills and knowledge, well we've already seen how well that's worked out for drivers (most drivers suck at it).

Licensing cyclists seems like a feel good idea to make things "fair". The fact is, things can never be fair between to such disparate classes of vehicles. Can someone explain how a world with licensed cyclists would be fundamentally different than today?

> What is the actual gain to society from requiring a cyclist license? Today, without licenses, police can and do hand out moving violations to cyclists.

I could, as someone who doesn't have a driving license, or has never learned to drive, go out on the road with my bicycle and more or less ride whenever and wherever I like.

I don't know what any of the road signs mean and I have no experience or training in hazard perception and collision avoidance. The law doesn't require me to. If I tried to do the same thing in a car I'd be arrested.

Kinetic energy = 1/2 m v^2

In a typical city scenario, a car has KE of 1/2 (2000 kg) (50 kph)^2 ~= 2e5 Joules. A bicyclist has 1/2 (100 kg) (30 kph)^2 ~= 3.5e3 Joules. Almost 100x less capable of inflicting damage on a typical pedestrian. Add onto this that bikes are way more maneuverable, less collision surface area, are totally exposed to the same forces exerted on the pedestrian with which they may collide... your comparison with cars is intellectually dishonest because you are comparing a human on top of a 10-20kg machine with a human inside a 2000kg metal box with an incredible array of safety features for the driver and hardly any for pedestrians outside.

Also, 30 kph is practically the top speed for a majority of cyclists - there's a level, narrow, short (200 m), 30-max-speed, all-traffic (except pedestrians) tunnel on my usual route. Do I temporarily reach 30 kph there, in order to be polite and get out of there as fast as possible? Yes, with some difficulty. (Do drivers honk at me for being unbearably slow while going the max legal speed? Yes.)

So, I'd say a typical city scenario is <20 kph, more than halving the estimate.

> I could, as someone who doesn't have a driving license, or has never learned to drive, go out on the road with my bicycle and more or less ride whenever and wherever I like.

Sure. Realistically, how much danger do people doing this represent to society? There are plenty of dangerous things that people could do, but the heavy machinery of law should be reserved for actual problems rather than hypothetical ones.

> If I tried to do the same thing in a car I'd be arrested.

And quite rightly, because if you weren't there's a good chance you'd kill people. Car drivers kill a lot of people, they're something like the 4th biggest cause of death in the UK IIRC.

For 5-19 YO it is 3rd most common, but when you look at the whole population it doesn’t even make the list.

source: https://www.gov.uk/government/publications/health-profile-fo...

Ah ok, I guess that must've been specific to my age range. (3rd most common there now).

If you choose to venture out onto the roads with no idea what the signs mean or how to avoid obstacles, good luck to you.

Realistically, every 5 year old knows what the colors on a traffic light mean and what a stop sign tells you.

There are many more rules that just red, yellow green and the stop sign. It would be nice if the highway code was that simple though.

For example, how to navigate a busy multi-lane roundabout safely as a cyclist, such as Old Street roundabout in London, which looks like this: https://upload.wikimedia.org/wikipedia/commons/d/dc/Old_Stre..., or this roundabout in Bristol, UK: http://c8.alamy.com/comp/B2NMCH/aerial-view-looking-down-on-...

Driver or cyclist, if you fail to navigate that intersection safely and correctly you will be cited (or injured or killed). How does requiring a license for cyclists solve this problem? You mentioned that driving without a license will get you arrested. This is true, but only after you've committed a violation, unless your country has random driver's license checkpoints (mine doesn't). We can already cite cyclists for moving violations today. Licensing doesn't change anything.

> How does requiring a license for cyclists solve this problem?

Well, would you argue that requiring a license for motor vehicles is unnecessary? The benefits should be similar and identifiable.

The law does require you to understand road signs and behave accordingly. It doesn't require you to get a piece of paper from a bureaucrat to attest to it.

I very much doubt that you would be arrested if you tried to drive a car without a license. Anecdotally, I've never had my driving license checked by a traffic officer. The large number of undocumented residents who are presumptively driving without a valid license is another point.

> I very much doubt that you would be arrested if you tried to drive a car without a license.

Without a valid driving license you will not be able to get insurance, unless you stole an insured vehicle from your parents or off the street.

Patrol cars in the UK and Europe have cameras that read number plates and check a database against all insured vehicles. If the car is not insured you will get pulled over, upon which you will be asked to present your driving license.

The number of people on bikes running up the inside of vehicles on narrow streets is too high. I’d walk through Bank (a junction in London) 5 times a week and I’d see time and time again people riding up the inside of tipper trucks and buses indicating to turn left. It happens all the time. I’d witness 3-4 near misses a month.

Rule 73 of the Highway Code:

> Pay particular attention to long vehicles which need a lot of room to manoeuvre at corners. Be aware that drivers may not see you. They may have to move over to the right before turning left. Wait until they have completed the manoeuvre because the rear wheels come very close to the kerb while turning. Do not be tempted to ride in the space between them and the kerb.

Let's be careful not to conflate two things: conforming to red lights isn't a goal in itself, safety is the goal. A lot of drivers get exercised about "cyclists running red lights" in cases that are actually safer than the alternatives, e.g. cyclists crossing a junction ahead of motor traffic when the junction doesn't have a dedicated light phase for cyclists.

Cyclists running up the inside mostly happens because waiting in a queue of traffic at a light or junction is very dangerous for a cyclist; stopping and starting are inherently dangerous and made much more so when you're starting in the middle of a queue that's anxious to start moving. In many cases road design endorses this - you'll see junctions with a cycle lane on the left running up to the light, and cyclists are expected to pass on the left even when a vehicle in line is turning left.

The solution that I've seen work best is a marked cycle-only box immediately before the junction - that provides cyclists a place to wait for the light, but it also makes for a clear dividing line between the section where cyclists are expected to pass the queue on the left versus the entry into the junction where that would be unsafe. Unfortunately cars often occupy those boxes and are seemingly not penalised for doing so. More of those boxes and more enforcement of those that exist would help, IMO, as would giving cyclists a dedicated light phase (as is done at e.g. some of the junctions on CS2) to let them clear the junction before motorists enter.

In terms of Bank specifically, it's an outlier: it's an extremely complex junction (7 roads) that simply has too much traffic demand for the space available. The junction has now been restricted to buses, cycles and pedestrians only during the daytime, and there's talk of removing the buses as well. That's appropriate: it's both safer and more efficient in terms of person throughput than allowing cars to use the very limited space there.

  waiting in a queue of traffic at a light or junction
  is very dangerous for a cyclist
When I'm approaching traffic lights and there's a queue of traffic and no bike box, I stop my bike at the tail of the queue, in the middle of the lane - just like I would on my motorbike or in my car.

I'm curious as to what you think the danger is in that situation?

A driver coming up behind you as you're starting to set off ( or immediately after you've fallen down) when the queue ahead of you is already moving. Somehow you tend to blend into the queue visually - the driver doesn't see you as separate from the vehicle in front of you, and doesn't realise you're still there when the vehicle starts moving (or does realise you're there, but rushes past you unsafely) - and the risk of falling over is much higher as you're starting off than at any other point.

Personally I see running red lights as a much higher accident risk than drivers of stationary vehicles failing to see me when I'm stopped right in front of them.

It's not the stationary drivers I'd worry about, it's those arriving behind them.

"Running red lights" with no further context seems to describe something extremely dangerous, but many specific classes of running red lights are completely safe.

Depends on the length of the queue. I find if I'm too far back from the intersection, cars will get up to speed and want to overtake me before I get through. If I continue to take the lane the driver behind me is likely to start honking and/or pass me dangerously. If I move over then I risk getting hooked as I go through the intersection.

So I find that it's usually safer to filter to the front at intersections. Even in the absence of a bike box I can squeeze into the crosswalk and get ahead of the lead car, allowing me to go safely through.

I'm also surprised by that. I do not feel safe trying to filter through traffic unless I know the light periods very well. And in a queue is one of the places I feel safest starting and stopping. I can accept an argument about how a slow bicycle is wobbly, but from a pure power standpoint, there is much less acceleration involved in a queue than in many other cases. If motorists have a problem with cyclists, its often about acceleration.

I generally wait with the traffic also when there is no bike lane/box. The only thing I ever worry about is getting hit from behind by some idiot not paying attention. I do try to mitigate this by turning my bike at an angle so that I look a little bigger.

This is all dependent on the traffic/intersection though, things vary depending on the circumstances.

It's a mistake to conflate cyclists filtering through traffic and undertaking.

The Highways Code specifically mentions that motorcycles do this also Rule 88 [1] and tells drivers that that they should be aware of other road users including cyclists doing this too [2] Rule 21.

[1] https://www.gov.uk/guidance/the-highway-code/rules-for-motor...

[2] https://www.gov.uk/guidance/the-highway-code/road-users-requ...

How many drivers were killed or injured by these riders?

> All other road users must have one.

Not true. Horse drawn vehicles and horse riders don't.


Invalid carriages also need no license.


I'm sure other examples can be found.

In most of the US (it varies by state), moped riders (where mopeds are generally defined as motorcycles below a certain engine capacity and top speed, often 30 or 40 mph) do not need a license.

In the Netherlands, since about two decades, moped riders (<50 cc engine, <40 kph) need a license.

Quite possibly one of the reasons for this is that they share a lot of the infra with cyclists here.

I think having to ride a moped is a sufficient disincentive.

You're being silly, but this does impact the fastest growing segment of new bicycle riders: e-bikes. Right now, most states classify them as mopeds and may or may not require licensing accordingly, usually not. I don't know if there are any stats yet, since they are relatively new, but these e-bikes seem to me more dangerous in the hands of inexperienced riders than a traditional bike.

Indeed, pedestrians need no license.

> Why shouldn’t cyclists need a license to ride on the road?

I don't think cyclists should need licenses to co-habit the roads with other vehicles. However, I think road safety education should be mandatory, perhaps in schools? When I attended (ordinary state) primary school we were all put through the RoSPA[0] Cycling Proficiency Test[1] at around ages 8-10. This was back in ~1977/78 so I'm showing my age :)

As I remember it the course was good fun. You got to use your own bike - I had a Raleigh Chopper :) - and the instructor set up different road layouts in the school playground to negotiate - he even had traffic lights and road signs. We were then lead out on the actual road where the instructor shepherded us through various real life challenges. We were even taught to perform "life saver" looks over our shoulders before carrying out any manoeuvres. At the end of the course we sat a test and were given wee metal triangular RoSPA badges.

I'm not sure if this is still a thing in schools now (I don't have any kids of my own), but I think it should be. Even basic education such as this prepared me to be a better driver, biker (i.e. life saver looks) and pedestrian.

[0]: https://www.rospa.com/

[1]: https://en.wikipedia.org/wiki/Cycling_Proficiency_Test

Because they are not zooming around in a several hundred kilogram steel box? Also, I regularly walk on or over roads, with no license whatsoever. Outside, everyone's participating in traffic, one way or another.

Even so, a bike is capable of moving at speeds high enough to cause severe harm to the rider or pedestrians. What if the threshold for a test was based on the average speeds of the vehicle?

Have there been a lot of pedestrian deaths (or serious injuries) caused by cyclists in your area? How does it compare to deaths/accidents caused by collisions with cars?

I'm also pretty sure that most of those accidents could be avoided by having better cycling infrastructures.

There have been a couple. However, when it does happen it gets a lot of media attention, the most recent incident taking place on Oxford Street, a popular shopping street in London: http://www.bbc.co.uk/news/uk-england-london-41263926

I'd say the licensing regimes should be set at a level that's proportionate to the number of deaths of others caused by that kind of vehicle, i.e. consider the number of other road users killed per year by every 10000 cars/bikes/bus-passengers/pedestrians/.... (perhaps as a QALY calculation to take into account serious injuries as well). Modes where that number is higher should have their licensing requirements made stricter, modes where that number is lower should have their licensing requirements relaxed.

Speed makes no sense for measuring the potential danger of an object. Energy or impulse do.

If we'd be playing billiards you'd be right. But in traffic the objects are driven by humans. And then speed becomes a factor.

Speed and impulse. Speed is a factor in the risk of collision (since our brain has limited real-time processing chops), and then when a collision happens, impulse is indeed an important consideration, together with another factor: protection level.

In a world with just those three factors, eg no traffic rules or speed limits, a car makes lots of accidents (because it's fast) and when they happen, a lot of energy and thus destruction are involved (because they're heavy), and also, while the car driver is protected by his/her big heavy steel box, the collision counterpart - a cyclist - might not be.

This asymmetry trinity is exactly why we should have (and have, to some extent) road rules and license regulations to level the playing field so that the outcome of the equation moves a bit in favour of the slow, light, and less well collision-insulated.

So if one endeavours to go inside a heavy (dangerous to others) steel box (that nicely protects you) that goes fast (you will have less time to react, and moreover, you "steal" other drivers attention because they constantly have to be on the lookout for "is there something fast coming? is there something fast coming?") - then yes, the onus is on this car driver to be responsible and submit to road rules, licensing, and heavy fines.

Then there's other non-collision factors that make current man-driven combustion engine cars look very silly indeed: they are noisy, pollute, take up a lot of space even when not moving (which is most of the time) and require expensive infrastructure. That leads us to ask ourselves - who on earth designed this system? And in the question belies the answer: no one did - there were coaches, we put a petrol engine in it, and went from there without any vision. With today's technology and insight, given a blank slate (there are no roads, no cities, no nothing - just a can full of people that need to work and live and travel, to be poured out on the green fields where we will build our infra) , NO ONE would design a transportation system this silly.

You're getting downvoted but here in Denmark while you don't need a license to cycle, if you have a driver's license you can lose it for breaking rules while cycling. Cyclists being familiar with the rules is just as important as car drivers.

Who says they don't have one? I drive, cycle, walk, and even use public transport. (Yes, yes, I should pick one of them and stick to it regardless of applicability, I know ;o))

Haven't you heard? We have motorists, cyclists and pedestrians. Not "a person driving a car", "a person riding a bike", "a person walking in public". You must identify with and be one!

Because the speed and the weight of the vehicles driven by motorists are generally so large that they tend to kill people in a collision. The same is not true for cyclists.

A pedestrian struck by a bicycle can absolutely be killed. It's not an uncommon occurrence, and it doesn't happen more often only because it's easier for a pedestrian to get out of the way of a bicyclist.

It is a highly uncommon occurrence and it doesn't happen more often precisely because of the vastly diminished kinetic energy involved.

Operating a vehicle requires a licence because it is a lot of responsibility. You can kill people or do serious damage to property if you are unable to operate one safely. Riding a bike is something children can do. Do you suggest a licence to walk as well?

I would like to see a road bicycle test, though. But it would be for potential motorists, not cyclists. I think if you want to operate a motorvehicle on the road you must be able to demonstrate an ability to operate a bicycle on the road.

Wow... Children can also drive cars, ride motorbikes, fly planes, perform surgery. That doesn't mean it's in society's best interest to let them without some training or examination beforehand.

The difficulty in riding a bike in a city doesn't come from having to control the vehicle. Danger doesn't come from the ability to maim someone with a bike. It comes from hugely complex set of rules, dynamic nature of traffic, high speed high pressure decision making necessary, sensory overload and fatigue, self control, competition on the road etc.

That's what you need to take into account to navigate safely.

But children do manage to ride bikes safely all over the world every day without any formal training. On the other hand, adults who have had training and are licensed to drive still manage to cause accidents.

Not in city centers on busy multilane inrersections, at least not in my country. We arent talking about riding a bike in a park or in the suburbs. And kids cause a lot of traffic accidents as well, regardless of their mode of transportation... Nothing will eliminate accidents. Training and systems can minimize it though.

But children do manage to ride bikes safely all over the world every day without any formal training.

Certainly when I was kid, bike and traffic safety was taught in school from 1st grade.

I imagine a major reason cyclists don't require licenses is because it would cost a lot compared to the benefit it would confer.

You could push the cost onto cyclists that want the license of course.

In the US the "exam" that most drivers have taken is so stupidly easy that it doesn't really mean much that they earned their license. This is less true for the young people that have had to do more work to get a license, but they aren't the majority of drivers yet.

They should but the reason it will never happen is political. Just look at the comments here. Dozens of cyclists throwing a fit because you have the gall to suggest someone operating a vehicle on a public road might be subject to licensure. They'd never vote for a politician supporting that, and they'd work to defeat any who did. On the flip side, the folks who agree with this are not going to vote for a politician they otherwise wouldn't simply because they suppose bicyclist licensure.

It's the same reason why you need a dog license but not a cat license in many areas of the US. The cat people will freak out when you suggest a $10 license and nobody else gives a shit.

Ahh.... It's actually happened. Various governments have promised to introduce licensing schemes. I don't think there's any current ones around. New South Wales had a plan, but they abandoned it just before implementing it. Not because of the voting power of the lycra lobby, but because it's bad public policy.

As for how good an idea it is to have a licensing system for people travelling without lethal weapons, papers please! It's a fantastic idea. You probably also should have a chip in you so the government can track you and see if you've travelled to a part of town you don't have a licence to be in. After all, we've got to cut down on crime.

Statistically the best thing for cyclist safety is increasing the number of cyclists on the roads. So any safety rules that decrease the number of cyclists (like requiring a license) has to be offset against the increased risk caused by having less cyclists. This is the reason that mandatory helmet use makes roads less safe for cyclists.

The other aspect is that you still have pedestrians, who may have little or no knowledge of road safety, so you still have to train drivers to be extremely careful regardless.

Pedestrians don't.

Pedestrians are, with the exception of some rural areas, provided with dedicated carriage such as a pavement or sidewalk. The rules of crossing the road are well defined and in some European countries pedestrians are even arrested for J walking.

There is no such thing as jaywalking in the UK, it's neither an offence (with the exception of motorways, which also regulate the type of vehicle which may enter), nor a term that's widely used or understood.

Outside of cities, there are lots of roads without pavements (not just in super-rural areas). Walking on them is normal, and people know how (e.g. always walk on the right, facing oncoming traffic).

And I very much hope it stays that way. Cars should not rule our public spaces.

s/Pedestrians are/Pedestrians ought to be/ - "meh, just plan for cars, anything else is irrelevant" is all too common.

> in some European countries pedestrians are even arrested for J walking.

Not quite, however you may be ticketed, if you happen to do it in front of a grumpy officer. I have only heard of it happening to a friend of a friend and the fine was under 40 euros.

In Baltics you’re certainly fined the moment you take your first step on the pavement. It started with a war against reckless driving and ended up in this state as of now.

"Some rural areas" is in reality almost every mile of road outside of towns. You've clearly never walked in the countryside.

> You've clearly never walked in the countryside.

I've lived in the countryside for 14 years. I don't walk on the roads where I live. The roads are narrow and there is no place to walk as the side of the roads are raised embankments built from tree roots. Looks like this: http://c8.alamy.com/comp/C41G6T/country-lane-with-trees-form...

> The roads are narrow and there is no place to walk

The place to walk in that situation is on the road. As someone who has decades of experience as a pedestrian, that road looks eminently walkable.

The roads are narrow, so you don't walk? What? That makes no sense.

Living somewhere that you can't even walk around sounds like some kind of prison. How do you get around? It sounds like you are unable to get around without burning fossil fuels. How does that make you feel?

sad to see your comment down voted but there are laws that bicyclist have to obey and they should prove they know them. this would include showing they understand the protective gear requirements and how proper attire protects them.

it isn't uncommon near some bike parks where I am for groups to block intersections which only serves to aggravate drivers. The are required in many areas to pass through intersections like all other vehicles, alternating with those waiting on the cross roads.

the bikes should be tagged to indicate they have the proper reflectors and lighting. there are still too many riding without reflective clothing let alone any type of lighting in low light conditions. I ride, I hate the idiots that ruin the sport

I imagine your suggestions would cause more deaths than they would prevent by sucking all the joy out of cycling, causing more people to just use a car instead, leading to more dangerous accidents and more obesity.

I think a requirement of a year's worth of cycling on the road before you can have a driver's license would increase cycling awareness significantly.

Nothing raises your appreciation of the difficulties faced by a group more than belonging to that group in a real and physical manner on a daily basis.

Not that I think it would be implementable: Humans are remarkably good at rejecting constraints.

Interestingly I've become a much better cyclist after obtaining my driver's license. It made me very aware of the fact that you can't see anything from cars (compared to the elevated and pillar-free view from a bike) so you have to be aware of that fact as a cyclist and react accordingly. Some things cyclists do without thinking much are stupidly dangerous and I think this often is the result of cyclists and drivers being two different sets of people.

Also I got to know the rules of the road much better than before and could adjust by behaviour to follow the law. I didn't go over red lights before, but there are a bunch of other things where cyclists don't tend to follow the rules of the road and IMHO it creates a much less stressful environment for everyone if they just did so.

> you can't see anything from cars

This is a very important difference and that’s the main reason I yearn for self-driving cars. Current vehicles are horrible to see where you are driving and drivers are in denial about it. At least, when you have engineers trying to represent a scene, they gut reaction is “No way I am dealing with such a slim view, I”m putting the camera where it can see what’s happening.”

That was basically one of my first reactions in my first hour of driving. »I can't possibly make a turn here. I don't see where I'm going.« That feeling fades over time and one (at least I still do with just a bit over two years of driving) adapts by going slower, looking more carefully, and especially looking for potential dangers before it's too late. Often (not always, as the article details) it's possibly to see ahead of time what can become problematic, e.g. overtaking a cyclist that you might have to yield to 100 meters later when making a turn. But it takes practice and I'm sure different driving instructors and different drivers place varying amounts of emphasis on such things.

The amount of times whilst pulling out of my driveway that I've thought "I wish I had a cameras pointing left and right from about where my indicators are" is quite large.

It's always worse when large vans are parked up and down the road so visibility is obscured.

I live in Belgium. We have a lot of cyclists here, but I think 80% of them don't know their weak position on the road.

As my father said, "They think they are terminators, that if a car crashes into them, the car will break and they will be OK". And that's what I almost always see on the road.

We have a lot of cycling lanes, thank god for that, but sometimes they are not in a good state so all those people tend to share the roadway with other cars. I almost driven into 1 because he was riding a bicycle on a forest road where you can drive 90km/h in a turn.

I do realise how vulnerable cyclists are, I've driven my whole youth to school with one. As a cyclist, I always look when crossing the road, or just around me, to realise where I am and where other people are and do not trust other cars. But most people ignore logic and common sense and just cycle on a roadway.

I still think it's a good idea to have some education on how to use the road. I mean, if a car crashes into a cyclist, the car will have a minor dent, but you can break a lot of things or even die. It's so easy to think before driving on a road because the cycling lane is in bad state.

PS: I also understand that there are bad chauffeurs on the road too, but from my experience, far less than bad cyclists...

Whatever the speed limit, you should never go 90 km/h in a blind turn if it is too fast to avoid a cyclist. You're the one responsible as long as cyclists are allowed on the road (and for good reason). For example, on the tiny winding mountain roads of Corsica, the speed limit is generally 90 km/h but it would be crazy for a non-local to go more than 50 or 60 km/h (if only because of the random encounters with wild boars ;))

> Whatever the speed limit, you should never go 90 km/h in a blind turn if it is too fast to avoid a cyclist.

It might just as well be some other obstacle: slow moving vehicle, car stuck on the road, end of a traffic jam. People really really need to learn that you need to be able to come to a dead stop within the distance that you can see.

> Whatever the speed limit, you should never go 90 km/h in a blind turn if it is too fast to avoid a cyclist.

I suppose it depends on the country, but usually roads mandate minimum speeds (tipically half the max speed) in order to avoid issues like this. Anything lower than the minimum speed should be treated like a static obstacle, so maximum speed should always take into account unexpected hazards.

Given that cyclists cannot tipically maintain the minimum speed required in any country road, they cannot share the road with faster vehicles. It's either curbs or segregated roads. Anything else is plain craziness.

Which countries have minimum speed limits? Not the US or UK to my knowledge?

(Yes, there are plenty of other countries, but those seem to be the two most discussed here)

In Belgium on the highway, the minimum speed is 70km/h (max is 120km/h)

People conveniently forget to adjust the speed to road conditions.

Cycling is similar KSI (killed-seriously-injured) per hour travelled as driving and walking according to some measurements. The trick is in calculating exposure: per-trip, per-distance and per-time all give different results. Per-time often suggests that cycling is safer compared to other modes of transport[1]. Per-trip and per-distance flip between cycling and driving[2]. Interestingly, being a public transit user works out the safest under both of these metrics.

" I also understand that there are bad chauffeurs on the road too, but from my experience, far less than bad cyclists..."

Your comment, which to be fair is not unusual, is essentially concern trolling. This impression is reinforced by your retailing of your measureless anecdote about how cyclists are worse than car drivers for being "bad".

Essentially, cycling is safe enough. Cyclists are not worse than any other road users. However compared to other road users they are not contributing to CO2 and are substantially less likely to kill other road users or suffer health problems stemming from low exercise levels.

[1] http://careymcandrews.org/pdf/McAndrews_et_al_Accident_Anal_... [2] https://journal.cpha.ca/index.php/cjph/article/download/3621...

There's no such thing as road tax, roads are paid for out of general taxation [0]. What they're thinking is vehicle excise duty, which doesn't apply to all vehicles, so that point is moot.

[0] http://ipayroadtax.com/

I would assume drivers are also thinking of the fuel tax they pay, not solely the excise tax. The UK has some of the highest rates of fuel taxation in the world.

Which is something you pay, at least in abstract, for someone else to clean up the carbon dioxide you litter the atmosphere with. If you dpn't want to pay the cleaning fee, don't litter -- take the bike!

Is that same level of tax levied on heating oil? If so, that bolsters the case that it's a carbon tax. If, as in the US, heating oil is untaxed or is taxed at a fractional rate of automotive fuels, that suggests that it's a road tax and not a carbon tax.

This chart suggests that it's the latter (substantially a road tax more than a carbon tax): https://www.gov.uk/guidance/fuel-duty#rates


Please don't post unsubstantive comments here.

I think it should be the other way around - some hours of cycling on busy roads should be required to get you your driving license.

Can you do 30-40 km/h on those bicycle roads or is there slow traffic such as 10-20 km/h bicycles and runners? In the latter case fast bicycles will stay on the main road no matter what.

In my experience I also have much better visibility at junctions on the main road instead of on bicycle lanes that are closer to walls and hedges and have 3 or 4 meters less of buffer between me and incoming cars from the right. I'm also less visible to them. It means to slow down at every side junction and over 100 km it would make a huge difference in time and wear.

As a matter of fact I tend to stay on the road all the time. A consequence is that in the last years I started going over dirt roads along channels and rivers. It's very quiet and more relaxing than being close to the noise of cars.

> I also have much better visibility at junctions on the main road instead of on bicycle lanes that are closer to walls and hedges and have 3 or 4 meters less of buffer between me and incoming cars from the right. I'm also less visible to them.

More anecdata: I take the bike when I can, and I have had motorists run into me a few times.

Every single time has been on dedicated bike roads.

Don't get me wrong -- I have had many, many more near misses on regular roads, but regular roads are built in ways which make it easy to convert impending collisions into near misses. Bike roads tend to not have those margins.

Before dedicated bike roads can be considered a solution, they must be built in ways which make me feel safer when I choose them.

How did motorists end up on dedicated bike roads?

Often these "dedicated bike roads" become not-so-dedicated at junctions. Also motorists frequently ignore painted lines marking bicycle lanes. (You can argue for physically segregated bike lanes instead of painted lines, but as a cyclist I won't use physically segregated bike lanes unless I know how they end, because all too often you get a "bike lane to nowhere" with no way to get off it).

Yes, and suddenly you're in a tunnel frequented by characters you're likely to find in a crime thriller who like to find out-of-the-way places to practice their craft.

So you mean dedicated bike lanes, I think. Dedicated bike roads implies physically segregated to me.

An (anti)pattern I often see is a road that's physically segregated for plain road, but then merges into and out of the main roadway (sometimes becoming a dedicated lane) for junctions.

It's so that drivers would see the cyclists in junctions. If the bike path is separated until the junction, it's common that drivers don't pay attention at all to what happens on bike paths and turn right into crossing bike traffic. What makes things worse is that usually there are bushes, placards, bus stops and others between road and bike path, all of which block visibility and grab attention.

I like your optimism. I've heard that in the Netherlands, the solve this problem by having a red turning arrow when people (on foot, on bike or in cars) are allowed to go forward. So in drive-on-the-left terms, vehicle traffic is allowed to go from the south to the north or east or from the east to the south. Since people who are driving cars from the south and want to turn to the west are obliged to stop, there's no possibility they'll unintentionally charge into people walking or riding bikes across the western entrance. The disadvantage is it requires four cycles not two, and therefore reduces the time pedestrians can cross.

On Albert St and Swanston St in (East) Melbourne, the separated lane stops at intersections so that cars can use the bike lane to turn left. People riding bikes are supposed to give way to people driving cars and then filter around the stopped cars into the bike box ahead of the cars. I never really feel safe using the Albert St lane for this and other reasons. Fortunately I have other options.

You ask great questions and they have been answered by someone else already! These dedicated bike roads do not have grade separated junctions, so in intersections sometimes motor traffic has to use a part of the cycleway, and sometimes cycle traffic has to use part of the regular roadway.

Yes, you can do 30-40 on the bicycle paths, and bicycle paths are not used by pedestrians.

You can if they are maintained. Most dedicated bike paths around me have ripples in the surface from tree roots and have never been resurfaced since they were built. Using a road bike with skinny tires on them is bone jarring at low speed.

Is that in Denmark?

No, the United Kingdom.

I don't think that's a fact. We are not even talking about particularly high speed roads here. The same thing happens regardless of speed so that's really nothing to do with it. The cost of separate cycle lanes (everywhere?) is huge compared to simple junction alterations as in this particular case. Traffic lights are even worse! These are vastly over-engineered solutions compared to that proposed in the article. Sure it may not work everywhere but remember the majority of roads and junctions are accident free where the road is shared by all users without issue. It's the easily avoidable accidents we should target. The lack of driver training/awareness is also explicitly called out in the article. It's an interesting read.

> simple junction alterations as in this particular case

I'd propose an even-simpler alteration, which would require zero engineering & road closures: block some of the right sight lines from Dibden to Beaulieu when approaching with a sightscreen, trees, or billboards. Drivers approaching would notice the impaired visibility and switch to a more defensive technique or be more aware of a potential collision.

In theory, yes I agree, some evergreen bushes would be sufficient. But it technically still leaves it possible for a driver to shoot straight over whereas the road change is fool-proof. Plus you don't actually have to close any roads to implement the junction alteration (in this case). Of course it is more expensive and time consuming to implement, I grant you that.

Shooting straight through is safe if Beaulieu is empty, so why prevent it? Even a low wall or hedge would be enough to prompt a driver to check for traffic more carefully than a superficial scan.

Because of the constant speed/constant bearing problem that makes it impossible to scan while moving and having a blind spot.

> In Denmark, a lot of high speed roads (excluding motorways, of course) out in the country have separate dual-way bicycle lanes near it but not attached to it. (Example: https://i.imgur.com/dS6jqXS.jpg)

Yes, there are many many of these in southern Jutland. However I can count on one hand how many cyclists I've seen using them. And I've spent months down there.

I cycle every day and am pro-bike, but they seem wasteful of both the countryside and money.

I would be opposed to building similar in the UK since more strips of asphalt throughout the countryside is the last thing we need.

My experiences cycling on fast country roads have been hairy at times, but we also need to recognise that there are problem spots and tackle these rather than copy/paste concepts blindly.

Separation is not the issue here, the separated lanes have to cross just as well. They are only less likely to have cars run over cyclists, because now the guy with the perfectly unobstructed view is forced to always yield to the guy who is driving around huge, dangerous blind spots. On-demand lights are even worse, now the guy who isn't a danger to others is legally obliged to preventive defer to those who are even if nobody is there, that's like saluting the portrait of the dictator.

> I feel a lot of commenters are simply focusing on what signs can do or what sort of junction to build, rather than focus on the fact that on high speed primary country roads, bicycles shouldn't be sharing the road with the cars.

Yes, they are focusing on feasible mitigation, not abstract theorizing (to be fair, about half of this comment does discuss the same sort of feasible mitigations as are dismissed in the first paragraph.)

I'm not sure having bicycle paths is always feasible. The first reason is that building a bicycle path could take years and the problem can be solved much earlier and simpler in the way described.

Second, low-traffic roads may normally be safe enough for cyclists (traffic amount was never stated). The same issue can apply to remote roads of all sizes. Granted, it would be nice to have a bike road next to every high speed road, but for traffic in the range of 20cars/h I've seen that exactly once in my lifetime.

Strong disagree with Svip. Training doesn't make good drivers. Who is going to ensure the trainers are good? Who is going to test that the training is being effective? More restrictive licensing, with more consistent requirements, might make a difference, but is politically nearly impossible. As long as many people will be inattentive and unreflective humans, there will be bad drivers. Unless you're going to support a policy of refusing to allow a large percentage of the population to drive (like, say, how we restrict most people from performing surgery), you'll have bad drivers.

Bez's proposed solutions have the benefit of being politically palatable, and aligning actual driver behavior with public safety.

Uhm, training actually does make good drivers in Germany, tests are hard, enforcement is very consistent. That is one reason they don't need a speed limit on the autobahn.

The drivers shouldn't be punished for bad design of the road. A separate bike road should be constructed.

Indeed. The main type of accident involving cyclists in Denmark are so called 'right turn accidents', where a lorry turning right at a junction hits a cyclists because they are in their blind spot. This obviously happens exclusively in cities.

The media has made a big deal about how tragic it is for the cyclists - which it is, but much less about how tragic it is for the driver. Remember, it's not like drivers want to hit cyclists or other soft traffic participants.

That's why I propose a separate bike road, even if it is quite an expensive undertaking. But it will make all the road users feel safe. Or -- at least -- safer.

Unfortunately, the cycling organizations don't really like to admit that most right turn accidents happen because cyclists don't seem to understand that motor vehicles have blind spots, and big motor vehicles have bigger blind spots.

As someone who regularly drives, bikes, walks and uses public transport in Copenhagen, it is blatantly obvious that people just don't know how to handle right turns involving both cars and bicycles correctly. Car drivers will leave space open on their right (between the car and the curb), instead of putting the car as close to the curb as possible (which is taught in driver's ed). Bicyclists will pile into this space, and when the light turns green, it becomes almost impossible for the driver to turn right, because of the bikes already there, as well as the additional bikes coming from behind, completely ignoring the right blinker on the car.

The mitigation of this problem is two-fold. Firstly, the drivers should not leave open space on their right, when turning right. Secondly, cyclists need to grow a sense of self-preservation and not dive into that space, even if it's open. Large vehicles need to have some distance from the curb in order to complete the turn, but that should not be seen as an invitation to cyclists.

"when the light turns green, it becomes almost impossible for the driver to turn right, because of the bikes already there, as well as the additional bikes coming from behind, completely ignoring the right blinker on the car."

Isn't that how it should be ? I'm not in DK, but in LV traffic laws in such a situation they should ignore the blinker, the driver must wait and turn right only after everyone going straight e.g. those bicycles and also pedestrians crossing the road he's turning to; since the vehicle turning must yield to everyone else.

That sounds like an extremely dangerous law... In bike-crowded cities you will never be able to take a safe right-turn that way.

You simply wait for a gap in the flow of bikes, and pull all the way to the curb, so there's no gap. Then you only really have to worry about the bikes in front of you, and any pedestrians in the crossing.

Portland gets this wrong in every possible way. First, drivers turning right must yield to bikes in the bike lane passing on the right (not unreasonable IMO). However, state law also makes it illegal to pull into the bike lane and establish a position to turn right. Add in pedestrians and it's almost impossible to make a right turn on some roads at rush hour.

For some reason, we seem to have a lot of car/bike/ped conflicts.

Around here it's generally solved by having separate traffic lights for cars, pedestrians and the bike lane, and having some extra time every cycle of lights where it's red for bikes and pedestrians but green for cars to enable those turns.

E.g. this shot in google maps - https://www.google.lv/maps/@56.9578862,24.1189538,3a,75y,240... ; it's green for cars, but red for the bike lane on the right side.

No, the law clearly says that you can never overtake a vehicle on the right - and it applies to cyclist the same as to everyone else, even if the vehicle wasn't turning right it's illegal for the cyclist to overtake it on the wrong side.

Do I misunderstand your traffic laws (taken from https://www.retsinformation.dk/forms/r0710.aspx?id=158005) pt. 21 about overtaking, which says "... Cyklist og fører af lille knallert kan overhale køretøjer af andre arter til højre." which seems to translate that it does not apply to cyclists and they can overtake on the right?

And 26.6 "...Ved svingning til højre må den kørende ikke være til ulempe for cyklister og knallertkørere, der kører lige ud. ..." saying that vehicles turning right must yield to cyclists going straight?

Yes, the car must yield, which is why the best approach is to wait for a gap in the flow of cyclists, and then go all the way to the curb, leaving no gap.

Of course, bicyclists will just jump the curb, ride on the sidewalk or overtake on the left, because everyone wants to be in front.

No, bicycles and small scooters (limited to 30kph, allowed on bike lanes) are allowed to overtake on the right.

Which is precisely why right-turning drivers should leave no gap between the car and the curb, to claim their space, as it were.

In almost every country (at least in Europe) cars have to yield to any traffic going stright on their right side. This includes both pedestrians and cyclists. And, yes, these have the right to overtake on the right. If straight traffic would have to yield to turning traffic, that would be way too dangerous

Yes motor vehicles have blind spots that is why in driving school they teach you checking over shoulder when you turn right. Remember? And yes you should wait for pedestrians and cyclists first if you want to turn right. That is the traffic law. And maybe you can park your car on the curb so that you don't need to wait for pedestrians either!

Pulling right up to the curb is the correct way to prepare for a right turn in a car, while sharing the road with bicycles, as taught in driver's ed here in Denmark.

And yes, obviously you check your mirrors and do a good shoulder check, that goes without saying.

But you ALSO pull to the curb after checking mirrors+over your shoulder, to claim the lane while preparing for your turn. You pull in either in front of or behind the cyclists, because the dangerous situations happen when you're side by side.

If there's a bike lane, you obviously don't pull across that, this only applies for streets without bike lanes.

This could not apply to Copenhagen and Denmark but I guess that in most other countries cyclists are also car drivers and spend more time in their car than on their bicycle. However there are many car drivers that don't have a bicycle or seldom use it, maybe because they are scared by cars. If the driver is not a cyclist it's difficult to understand what's going on. I suggest that to get a license one should also get some practice in the traffic as a cyclist, to remember what not to do with a car.

About your mitigation strategy: that would solve the problem but unfortunately the cyclist won't stop behind the car. The cyclist would turn left to overtake the stopped car. I don't like to do that because it puts me in the middle of the road or to the right of another car. It also requires some space to maneuver and a car could be arriving from behind me and close the gap.

Usually stopped cars don't hit cyclists (but watch out for doors). I look at the traffic light and refrain from overtaking the car (both on the left and on the right) if it's about to turn green.

There is definitely an issue with crossing over to overtake on the left, which is why the best thing to do is actually to stay behind the car, but leave some distance to the curb, to claim your position in the lane.

That way, there will be no cyclists blocking the car from turning safely, and the driver will only have to yield to pedestrians in the crossing. To put it bluntly, it's because they're idiots.

> The main type of accident involving cyclists in Denmark are so called 'right turn accidents'

It's the same in London, a city with notoriously poor cycling infrastructure and narrow streets due to long history. Ok, it's the "left turn" mirror image version due to driving on the left not right side of the road, but otherwise the same.

It’s not helped by cyclists habit of going straight on from a left turn lane.

That is one of the things that annoy me most about new bike lanes. Why do they put them on the left side of left turn lanes!?? Look, we have had lanes for traffic like a century now. It shouldn't be hard. They just have ti do what they always did!

(Also the stupid unsafe cycling advice of staying to the left. Take the lane and take the correct lane, please.)

Many cars drivers have a habit of accelerating to get around cyclists as they are both coming to the junction where the car wants to turn left using the left turn only lane and the bike wants to go straight. I get cutoff at least once a month in this way. Very scary and usually the car drivers are oblivious to how dangerous their driving is.

If the bike lane must be next to a turn lane on an intersection, the dutch have a very simple hack [1].

All bike lanes get green, and get to sort themselves out, but crucially all car lanes get red.

[1]: https://imgur.com/a/VFElf

That is both legal and necessary to do

I absolutely disagree. If I, as a driver, see someone in a left-only lane, it should ba absolutely safe for me to assume that they are indeed turning left - so I can safely pull out at the junction. Obviously it leads to accidents if the person then goes straight and runs into me. Being in a left-only lane is the same as indicating a turn.

If I as a pedestrian/motorcyclist/car-driver assume such things, eg that people not using turn indicators aren't turning then I'd have an accident at least once a day.

Making a couple of second allowance to protect the lives of other road users seems reasonable to me.

That said I don't live in an area with high cycle use - in the small UK city I'm in you're more likely to be stopping to avoid a drunk or a teenage boy doing wheelies.

>and runs into me //

Cleverly worded, but if you use your mirrors properly ... what's that phrase "hell is other road-users"?

The highway code says that cyclists should stay to the left. Even on a roundabout with multiple lanes, cyclists are supposed to stay to the left and cycle across exits.

I don't ever follow that rule when cycling, it does seem dangerous, but it is the rule, regardless of whether you disagree. You should always check blind spots when turning left, and you should have been coached to do this when learning to drive.

> The highway code says that cyclists should stay to the left.

The uniform vehicle code (which is the basis of many state road rules in the US) states the following:

>> Any person operating a bicycle... shall ride as close as practicable to the right-hand curb or edge of the roadway except under any of the following situations

>> [...]

>> When reasonably necessary to avoid conditions including [...] substandard width lanes that make it unsafe to continue along the right-hand curb or edge. For purposes of this section, a "substandard width lane" is a lane that is too narrow for a bicycle and a motor vehicle to travel safely side by side within the lane.

A cyclist riding as close to the edge of a given lane as practicable will leave roughly 2 feet between the edge of the bicycle's handlebar and the edge of the lane. The cyclist is about 2 feet wide. In a lot of states in the US, there is a requirement that a least 3 feet of space is left between the cyclist and the overtaking vehicle.

2 + 2 + 3 = 7 feet. A passenger vehicle in the US is about 6 feet wide. A commercial vehicle (e.g., truck/lorry or bus) is about 8.5 feet wide. The typical lane width ranges from 10 to 12 feet. For a 10 foot wide lane, a cyclist taking up 7 feet of space only leaves 3 feet of space for a 6 to 8.5 foot wide vehicle (5 feet for 12 foot wide lanes). That means that the two vehicles cannot safely travel side-by-side within the lane.

So, according to the UVC, there isn't a requirement to keep toward the edge of the lane in most cases.

I haven't checked to see whether the UK highway code has a similar exception to the "stay to the left part of the lane" rule.

The other problem with keeping towards the edge of the lane as opposed to riding near the middle is that the cyclist is less visible to overtaking, cross, and oncoming traffic because they're hidden by other motor vehicles or not in the field of view of a lot of drivers.

I completely agree. There are similar rules over here about leaving a gap to the kerb, and that there should also be a gap left by overtaking vehicles. Both are often not observed.

Here's the (brief) rules for roundabouts: http://www.highwaycodeuk.co.uk/rules-for-cyclists-roundabout...

Most cycling proficiency courses here in the UK will tell cyclists to ride more centrally when in a narrow lane, which is good advice. A cyclist is also legally allowed to do that.

It gets a bit fuzzy though when there's a cycle lane provided.

> It gets a bit fuzzy though when there's a cycle lane provided.

A lot of times, the cycle lane is really too narrow to safely ride in. It really ought to be a minimum of 4 feet wide (allowing a 1 foot space on each side of the cyclist if they're centered in the lane. Some that I've seen aren't even as wide as the handlebars on the bike.

Where I live, a cyclist isn't required to use the cycle lane even if one is present. Do you know whether the same applies in the UK?

You don't have to use a cycle lane here, but there have been cases in recent years where police have incorrectly stopped/charged someone for not using a cycle lane.

Should be; and as a driver, rider, and cyclist, would absolutely love to transport myself through such a world. Alas, a twenty-minute drive yesterday got me: people turning right from a left lane (across the through and right lanes, no less), people driving straight through red lights, pedestrians nearly run over on their green light by drivers turning into them, not to mention no indicator lights at all, and 10% of cars driving without any lights at all at night. (Incidentally, "world has road standards and then just ignores them" is IMHO the #1 reason that self-driving cars are not in our near future)

> If I, as a driver, see someone in a left-only lane, it should be absolutely safe for me to assume that they are indeed turning left

The highway code is a bit more nuanced than you on this topic: http://www.highwaycodeuk.co.uk/road-junctions.html

it is safer for the vehicle driver to not make that assumption, it is safer for the cyclist not to be in that position as you point out; and primarily, it is safer if the road layout does not put these different vehicles in the same space.

It's not helped by drivers harrassing cyclists who are not in the, empty, left turn lane "because they want to get by".

Seeing bike riders in London a lot of those sort of accidents (90 % of fatalities in London) are caused by cyclists undertaking agresivly.

Also the way the UK's bike lanes are set up encourage cyclists to get in front of other traffic at traficlight controlled junctions - normalises undertaking.

You need to build the intersections properly though. My city, in the NW US, built a beautiful separated multiuse path alongside a 50mph road. The trouble comes at the southern end where two 50mph roads intersect and the separated bikeway turns into a standard on-road bike lane on the other side. This also happens to be the highest traffic volume intersection in my city of 100,000.

The intersection is designed so right turning cars can make the turn at about 15-25mph when they have the green. The same green that the cyclists have. Cyclists must wait at a position that puts them slightly behind the right turning driver and the design simply doesn't allow a cyclist to inch forward enough to change that. If you were designing an intersection that will get cyclists killed, this is how you would do it.

My approach is to stop at the green light and hit the pedestrian "beg button" and wait for the walk signal on the next light cycle (a minimum 2 minute wait). This is still no guarantee that drivers will actually yield to me, but at least it gives me a chance to be seen. Even waiting, and having the walk signal and right of way, drivers get irrationally angry at having to wait a few seconds for me to cross. A few don't wait and swing wide right in front of my path. Fun Times.

There is a move away from seperate bike roads because these have a much higher right turn accident rate at junctions and a move towards bike lanes which are safer due to the shared space (more awareness of bikers by drivers)

> The drivers shouldn't be punished for bad design of the road.

No excuse of "bad design" can work for blowing across a crossroaded "Give Way" at 37mph. That's reckless in any situation.

No it isn't, I do it daily in Prague together with police cars that are coming from a nearby police station. The difference is that I can be sure no bicycles are on the road. If I stopped or slowed down it would be very unsafe but this way I just merge into the traffic seamlessly.

I think you're referring to a different situation. It sounds like you're talking about a merge into a freeway or something similar. Your parent referred quite specifically to a cross road. The goal is not to merge, it's to get across. I agree with your parent; the action was necessarily dangerous. A person approaching a cross road must slow down, more so if there's a give way sign.

Nope I don't mean a merging lane, I literally mean a side road where you have to give way to a main road but the safest thing to do is to speed up to 60-70kmph and merge into like if it was a highway. It's completely legal and most safe. In Europe the "give way" sign doesn't say anything about your speed, just that you can't cause others to speed down. Btw the same sign is used on freeway merging lanes, having to slow down as you imply would be extremely dangerous; the correct way is to use the merging lane to match speed with the lane you're merging into.

I'm talking about this place: https://goo.gl/maps/2JwZGyzLPGG2

What's the punishment exactly? Being forced to slow down at a junction where you don't have right of way to ensure you give right of way to the road traffic you failed to see?

The road was obviously not designed with cyclists in mind, it was designed to enable cars to merge into the road quickly. The punishment is exactly what you describe: "we failed to anticipate bycicle traffic and so we'll make it impossible to go fast because we made a bad design."

Cyclists shouldn't be punished either. All too often these bike roads are built to be less safe than the main road.

There's a lot of math in this article about blind spots and car pillars and whatnot, but a Google Maps Street View appears to show the terrain and brush blocking the view on approach [1][2], making all of this math largely moot. That the drivers in question didn't even so much as slow down for this approach is beyond reckless, aside from the other lawbreaking that followed.

The 'Give Way' sign is idiotic, put a Stop sign on the other, flatter road, or make it a 4-way stop. The drivers causing accidents were blowing through the intersection, and shame on the wording of the law that they were able to weasel out of harsher punishment.

Altering the roadway geometry, either as proposed, or more drastically with a roundabout, is a brute-force solution on this country road when other societal measures don't suffice, but it seems that there's plenty of room to enact other changes before you dig up the road.

[1] https://www.google.com/maps/@50.8638239,-1.4511249,3a,75y,26... [2] https://www.google.com/maps/@50.8634141,-1.452595,3a,75y,269...

«make it a 4-way stop»

You may be unaware, but 4-way stop signs are very uncommon outside of the US. In fact, in the UK they have always been formally prohibited by the Department for Transport in 2002.

As a European driver, I discovered the existence of 4-way stop signs when I moved to the US, and I have always found them dangerous in a counter intuitive way: drivers are so used to 4-way stop intersections, that they may adopt the habit of (1) doing rolling stops and (2) assuming an intersection with a stop sign is almost always a 4-way stop. Both of these 2 habits are dangerous... At least that's my personal experience. I prefer the consistency of my home country where every stop is a 2-way stop, therefore I'm never surprised by non-stopping traffic at an intersection.

> rolling stops

What's special about a 4-way stop here? And a rolling stop is good enough to prevent a high-speed collision anyway.

> assuming an intersection with a stop sign is almost always a 4-way stop

Around here there is a good mix and I don't think that happens.

> I prefer the consistency of my home country where every stop is a 2-way stop

Well I personally hate going through 2-way-stop intersections that are over 25mph.

A rolling stop is only good enough to prevent a 'high-speed collision' as judged by people wearing steel armor boxes.

If you don't drop to walking speed, it's not a rolling stop, it's flat-out running the stop sign.

I think we can agree that 4mph is not high speed.

I always found it hard to understand what's so difficult in actually stopping when the sign clearly says "stop". The wording hardly leaves any room for misinterpretation, doesn't it?

I'd guess some people want to keep a bit of the momentum going for fuel efficiency? Going from standstill to moving, with a big heavy metal box, requires more energy than simply accelerating an already existing momentum, even if it's just a small momentum.

But I'm with you: Stop means stop

Many stop signs are placed such that stopping for them prevents you from safely observing oncoming traffic. So after stopping at a stop sign you still need to slowly inch forward checking that the path is clear making the stopping portion of the exercise absolutely pointless.

Further my habit of actually stopping has gotten me rear ended and tossed into an intersection. Doing things unexpected can be very hazardous.

Yup, but you still are supposed to, you know, "STOP". :) And isn't it done on purpose? This way, you are not breaking the knees of pedestrians that are crossing the road & you are not sending bicyclists flying. Stop, slowly inch forward (no broken knees, no flying people), enter the road when safe. I'm always wary when walking across the road with a stop sign, precisely because people don't stop at the line.

It's a shame that following signs is unexpected. Somehow, it's not acceptable to run a red light because "no-one is around anyway", but it's fine to do rolling stops?

Cyclists are supposed to come to a full stop as well. They're easy to see stopped right at the corner.

There is at least one state (Idaho [1]) in the US where cyclists are legally allowed to treat a stop sign like a yield sign.

[1] https://en.wikipedia.org/wiki/Idaho_stop

This doesn't happen, at least from my limited experience in Beverly Hills.

I ride several blocks to work sometimes and I take local streets which all have 4 ways stop signs.

9 out of 10 cars yield and refuse to go. I get the pedestrian treatment.

The problem of rolling stops is when people accidentally do that at a 2-way stop that they mistake as a 4-way stop. This greatly increases the chance of a collision.

The prevalence of rolling stops is much reduced in countries where all intersections are 2-way stops.

Yeah where I live (the outer sunset neighborhood of San Francisco) it feels like half the stops are 2-way and the other half are 4-way. As a cyclist this is dangerous for me, because if I go through an intersection where I am not stopped but cars are subject to the stop, they will stop, not see me, then start going. So consistency is key!

It happens the other way around too; I had a bicyclist roll through a 2-way where I had no stop and almost cause a collision. They yelled at me, probably thinking it was a 4-way.

It might be my imagination, but I seem to recall nearly all, if not all, 4-ways marked as such with a small rectangular placard below the stop sign in Virginia. I've seen them sometimes in California, but not even at the majority (not really any more common than the opposite sign "Cross traffic does not stop").

> It might be my imagination, but I seem to recall nearly all, if not all, 4-ways marked as such with a small rectangular placard below the stop sign in Virginia.

This is the key problem. Whether an intersection is a 4-way or 2-way stop sign changes the right-of-way of any person with a stop sign. (If it's 2-way, I must yield, but if it's 4-way and I've reached the intersection first, I have the right-of-way.) It is crazy that in the US, a person at a stop sign can't distinguish these two situations for certainly without looking to see if the other direction has a stop sign.

(Similar complaints can be made about flashing-red and flashing-yellow intersections.)

There's a 4-way give way junction here [1]. That's the closest you will get in the UK. It means give way to anything that is in the junction before you are, but the actual effect is for cyclists to frequently be aggressively hooted at and nearly run over by cars.

[1] https://www.google.co.uk/maps/@52.2190575,0.1178374,105m/dat...

Interesting construction. This wouldn't be legal in Nordics, there can be no direction in give way signs indicated, and there can be no 4-way give way junctions.

There is at least two of those four-way yields in Stockholm, and yes, it is of course absolutely insane. In a regular junction you already have to

1) Yield to traffic inside the junction,

2) Yield to traffic approaching from your right.

In a four-way yield junction, you have to

1) Yield to traffic inside the junction,

2) Yield to traffic approaching from your right,

3) ??????? to traffic approaching from the left.

It breeds confusion and has no advantage over a regular intersection. Unless, of course, you count confusion as an advantage. Which you do if you drive the vehicle with the largest inertia around.

To be honest, I was quite surprised that it even existed, and it may possibly not be legal in the UK either.

I think the UK has a lot of "mini roundabouts" where they just paint a circle on the road and mount roundabout signs. Traffic paths are similar to what they would be without the circle. Of course, the give-way rule is different: give-way to your left (easy), rather than give-way to those who arrived first (hard).

You give way to your right, not your left, on a mini-roundabout in the UK.

The "4-way" or "All-way" addition helps a little ( https://i.imgur.com/dzRksUw.jpg ), but I've made the mistake at least once or twice myself.

At least 4-way stops have the same basic rules throughout the US (cars go in order of arrival).

2-way stops do not:

Consider the case where the first car to arrive is turning left, but cannot do so due to traffic on the main road. Then a car arrives from the opposite direction going straight. Who is supposed to go first? In some states it's the car that reached the stop sign first, in others it's the car going straight.

All this assumes drivers who know and follow rules. Where I live, I have to wait in a suicide lane to make a left, and sometimes drivers at the stop sign will try to steal the right of way to make their left first. And then there's the problem that it's accepted practice here to point your car and inch toward traffic as a way to request they give you an opening, despite the very significant risk involved in doing so.

When I did my driving lessons many many years ago (UK), I was told that the car turning left would have right of way, because they weren't crossing a carriage way.

You have to flip it so the car is turnin right: across the path of the second car. In Australia the first car would have to wait. You can’t turn in front of a car going straight.

Sorry change "left" to "across traffic" (i.e. right in UK, Japan &c. left in mainland Europe and Americas)

That's the same as everywhere else, just on the other side.

Left is across traffic in the US. Different sides of the road.

> In some states it's the car that reached the stop sign first, in others it's the car going straight.

Do you have a cite/link about this? I've heard people say different things, but I've never seen evidence that the left-turning vehicle has legal right-of-way in any state.

I could not find a US state with that law, interestingly enough, despite being taught it in Driver's Ed. in Virginia, that doesn't appear to be the law on the books.

Most states' language is similar to California's 21801 which is must yield to vehicles "approaching frome the opposite direction which are close enough to constitute a hazard"

Not USA, but Ontario, CA actually does have this rule in 136.2[1]

Michigan found a 3rd option: If there are two cars going straight from one direction and one car waiting to turn left from the opposite, then it goes Straight, Left, Straight. This is regardless of the order that they arrive[2]


I also find it absurd that this situation is not covered in every single States' driver's manual. After googling this, it's clearly contentious, and has been for years. If there ever was a situation that needed clarification it's this one.

I think it was New York's manual that had the section and had a list of questions "you should be able to answer" before proceeding (without answers provided). One of the questions was this situation, but I could not figure out the answer from reading the chapter!

1: https://www.ontario.ca/laws/statute/90h08#BK230

2: http://www.legislature.mi.gov/(S(did15hikprdseomawdqif0jc))/...

This is evidence to the contrary, as you've said...

Utah law only talks about who must yield. I use to believe that a person going strait had the right of way. Some of the confusion may come from the wording from different ways of entering the roadway. In Utah it says the following.

41-6a-903 Yield right-of-way -- Vehicle turning left -- Entering or crossing highway other than from another roadway -- Merging lanes. (1) The operator of a vehicle: (a) intending to turn to the left shall yield the right-of-way to any vehicle approaching from the opposite direction which is so close to the turning vehicle as to constitute an immediate hazard;

This doesn't actually apply to a stop sign because it says "other than a roadway". It also talks about "opposite direction" and not cross traffic.

The other part of the law that actually seems to apply is the following.

41-6a-902 Right-of-way -- Stop or yield signals -- Yield -- Collisions at intersections or junctions of roadways -- Evidence. (b) After having stopped at a stop sign, the operator of a vehicle shall yield the right-of-way to any vehicle in the intersection or approaching on another roadway so closely as to constitute an immediate hazard.


That's crazy. I find myself frequently having to pause, look around, make sure, etc (and the pausing and looking around in itself is dangerous) so often because of the discrepancy. But as someone born and raised here I've never considered an alternative, and am shocked that the alternative is so simple!

The other thing about 4-way stop intersections in the US is that the right-of-way rules when two vehicles arrive at the intersection at the same time is not quite what one would expect. That is, the vehicle to the right has the right of way. If the intersection was controlled by a roundabout, then the vehicle to the left would have the right of way.

In the US right-of-way for a 4-way stop is order of arrival, with yield to your right if you arrived at the stop line at the same time as someone else.


In general it works fairly well even in areas notorious for bad driving (although sometimes people steal a turn, I've never seen someone blow right through one, even late at night).

Roundabouts are different. It's a one-way street that goes in a counter-clockwise circle, so naturally you if you are joining (effectively taking a right turn onto it) you must yield to traffic already on the roadway (that would be traffic coming from your left) as you would when turning right onto any street. Once you are on it, you now have right of way.

4-way stops are more deterministic, whereas traffic can get "starved" trying to enter a roundabout. However, you always must stop at a 4-way (even if it's a rolling stop), whereas you can blast through a roundabout if there's no-one to yield to.

> In general [4-way stop] works fairly well

If you're not considering intersection throughput, then that may be true, but 4-way stops are one of the worst forms of intersection control when there's even a moderate amount of traffic.

> Roundabouts are different.

Of course, but what I was implying when I made my post is that a 4-way stop sign controlled intersection right-of-way rule is unusual compared to most traffic situations where one has to yield the right of way. At a yield sign, one typically yields to traffic coming from the driver's left. When entering a highway, one yields to traffic on their left. When making a right turn on red, one yields to traffic on their left after coming to a full stop.

Only when coming to a 4-way stop are the right-of-way rules reversed it seems.

> whereas traffic can get "starved" trying to enter a roundabout.

One thing that's done in the UK is to add traffic lights at the entrances to busy roundabouts. When traffic gets heavy, the activate the traffic lights to regulate traffic flow to prevent the starvation scenario that you mention. To me, that's the best of both worlds where one isn't legally required to stop even when there's no traffic approaching.

They do exist in Germany.

I won't dispute it because of the general "If you've never seen a black swan, is that proof that none exists? (No)" fallacy, but speaking as a German, I cannot remember having seen one in Germany, ever. So even if one exists it's an exceedingly rare thing.

When I lived in the US I always found it unnecessary that everybody has to stop. In Germany we always have a "main" road and a "secondary" road, and those on the main road don't have to stop. That priority pattern is kept from major roads to tiny roads. Then there's the "right before left" rule when the roads are equal - creating a priority without signs and without "everybody has to stop" rule.

In the places where I see four-way stop, there is some constant but not heavy traffic during parts of the day, so if you had a primary / secondary road with only a two-way stop, the secondary road would not flow well. However, the traffic was not high enough to warrant a signal.

This makes sense to me, and I don't see how other systems would be anything but worse. A signal would have a short cycle and only delay people during large parts of the day, a two-way stop would cause problems in one direction, it seems excessive to tear out a bunch of road to put concrete in, etc.

It is of course very natural that the roads in the US are built to completely different standards than Germany, given how incredibly different the layout of US cities are from German ones, broadly speaking.

Almost every such intersection in Germany was replaced with roundabouts, afaik. Perfect use case for them.

Australia is more similar to the US than Germany, but we don't have four way stops here; they case is provided for by roundabouts. Even still perhaps we have less because we rarely have pure grids.

I went to check on google maps, I suspect I misremembered and it was a 4 way priority loosing (downwards triangle), but the road markings are mostly worn out (but I think they are 4 dashed lines, not continuous), and Germany being Germany, I can't use google street view.

St Peter Strasse / Von der Tann Strasse in Rorhbach in Heidelberg.

Sorry for causing a stir with my bad memory, I remembered that from my French point of view that would cause a deadlock.

I'm guessing they actually do exist almost everywhere, but they are rare, and only typical in very slow traffic areas such as residential block intersections (see google map in next post) where you can't motivate building a roundabout or adding lights. Anywhere else they would be replaced by either traffic lights or a roundabout.

Most importantly, you would avoid at almost any cost to cross two rural roads in a flat 4 way crossing. This is exactly where there is ample space to make a roundabout.

In the Netherlands, a 4 four stop would lead to an instant deadlock. I.e., if you reach a stop sign then all traffic on the road you approaching has priority.

There are no rules for solving this conflict. Because you cannot have two roads that have priority over each other.

The Dutch equivalent of a 4 way stop is a raised intersection. Basically an intersection that is one big speed bump. Traffic is forced to slow down but otherwise normal priority rules apply.

> In the Netherlands, a 4 four stop would lead to an instant deadlock. I.e., if you reach a stop sign then all traffic on the road you approaching has priority.

In a 4 way the same priority rules apply thatwould have applied if there were no stop signs. Obviously the signs have to make clear to everyone that it is a 4 way stop, so that you don't have 4 cars all thinkning they are in a 2 way stop. This is usually done with an additional sign under the STOP sign. After that it just works like any no-stop intersection, you yield to e.g. the car to the right (Sweden) or the first car to arrive (US) etc.

The issue is that you can get in a situation with a car on each road. One of the drivers then has to signal the car on his left that he gives up his right of way. This is the same across the EU.

That happens all the time. But the point is that it's no worse because of the stop signs. Even without 4 stop signs you can have 4 cars stopped. This isn't a "bad" situation or a deadlock, this is a very safe situation and just means cars will have to crawl out in the crossing while making eye contact and taking some kind of turn driving. Now, if this happens regularly then the junction is much too trafficed for being a planar 4 way - with or without stops. Should be a roundabout in this case. But 4 stopsigns does not make an automatic deadlock. You don't need a rule to decide who goes first in that situation.

The same can be seen in any intersection with red lights, as soon as the red lights don't work. You just crawl into the intersection, trying to take turns. It works.

The thing is, in the Netherlands, you could not use a stop sign for that, because a stop sign does not mean "stop and then crawl out into the intersection while maintaining eye contact with other drivers", it means "stop and give way to the road with the higher priority". A proper traffic code should always provide a consistent way to approach any given sign.

If there's no working traffic lights and no stop signs, a different set of rules come into play.

> The thing is, in the Netherlands, you could not use a stop sign for that, because a stop sign does not mean "stop and then crawl out into the intersection while maintaining eye contact with other drivers", it means "stop and give way to the road with the higher priority". A proper traffic code should always provide a consistent way to approach any given sign.

The normal solution (used e.g. in Sweden) is to simply add a supplemental sign saying "multi-way stop" under the stop sign. https://d3h4jqp4au0mzy.cloudfront.net/wp-content/uploads/201...

I was taught that the shape of the stop sign is octagon (and yield triangle) because you can recognize the shape from behind. The main road doesn't need to have the diamond priority sign. You know you have priority if the you see the octagon or triangular shape on adjoining roads.

Not sure I understand what you are saying here...

Imagine driving towards an intersection. There is no sign telling you if you have priority or not. There is a red car coming from the right. Then you notice that the red car has a stop sign in front of him (you recognize the shape). You infer that you are on the main road and you can ignore the right-hand rule.

Sorry, I understood what you meant, just not what the relevance was. What you are saying applies, and applies universally (regardless of whether there also exists multi-way-stops).

That there is no need for supplemental signs I guess.

Either you have a regular stop intersection, in which case there are octagonal stop signs on the roads that do not have priority. No supplemental signs needed. A driver stopping at the stop sign knows that when the intersection is clear he can drive. If it's a 4 way and he's turning left, he may have to yield to a driver on the opposite road, depending on the local laws (In a right hand yield priority the car turning left yields to the car in the opposite road).

Or it's a multi-way-stop (which is rare) in which case all roads have the supplemental sign. None of the roads have priority and all drivers can do to resolve deadlocks is be careful. All the supplemental sign does is inform drivers that OTHER drivers will also stop. It just removes the ambiguity we talked about above - where a driver at a stop sign might otherwise think he can't drive until the intersection is completely clear.

A stop sign means stop and give way to the cars on the main road, or if the other road has same priority, to the cars on your right. If there are no cars, you are free to enter the junction. A road with a yield sign and a road with a stop sign are of the same priority and right-hand rule applies. Unless you are in Croatia where people universally refuse to accept this simple rule and consider the stop sign as having lower priority "because that car has to stop".

It's the same situation when there are no signs. Then all roads have same priority and you give way to the vehicle on your right. If 4 vehicles come to the intersection at the same time then every vehicle has another vehicle on their right and it's a standstill. If I remember correctly in such situations drivers need to communicate and one vehicle should go first to resolve the situation.

I don't think so. The literal text from the StVo for the STOP sign (VZ 206) is

    Ge- oder Verbot

    1. Wer ein Fahrzeug führt, muss anhalten und Vorfahrt gewähren.
And that strongly implies that at least one of the other streets has the right of way. I've also never seen one.

This is the same in the Czech Republic and the behavior should be the same as if there are no stop signs (except that you have to stop). One of the drivers have to give way to the one on his left and after he goes through, it's all clear.

Really? I've driven through much of Europe, and I've never seen a 4/all-way stop junction.

Edit: That came across as if I was simply trying to dispute your statement, but I am genuinely curious to see an example of this in Germany.

We have at least one in Sweden, https://goo.gl/maps/kYZpAcQnD3n, it also has the sign subtext "flervägsstopp" /"multi-road-stop".

Never seen one and I suspect that it would be illegal. The rules that govern which sign can go where in Germany are very detailed. And they are broken all the time, so I'm really not disputing the existence of counterexamples.

As a motorist, the 4-way stop is the most disruptive and maddeningly inefficient traffic control device in the municipality's tool chest. How much fuel and how many brake pads are consumed because towns just default to this kind of intersection. There are many better options (as others have listed in this thread) for almost all traffic situations. Always reaching for the 4-way stop is lazy and irritates motorists.

There are a lot of situations where a two way stop would cause traffic to get backed up in the direction of the stop sign.

A great example is a 4 way stop at my former high school. Twice a day, everyone came and left causing a bit of a traffic jam. A two way stop would leave people in one direction waiting indefinitely, whereas a 4 way kept things chugging along rather nicely in both directions.

That is an intersection that needs traffic lights. And if it's just twice a day, default the less used direction to red and have it switch to green only if a car is actually waiting.

As a european 4 way stops seem really inefficient to me. If the intersection isn't that busy a 2 way stop (or no stop at all) favouring the more frequently used road is the obvious choice. If it's a busy 4 way stop however, can you really keep track of when other cars arrived at the stop? And if it's busy enough that you have to yield to the car to your right, chances are that everyone has a car to their right - so vague negotiation between drivers is required. Nobody wins in that scenario.

Also, roundabouts are the superior 4 way stops.

There often isn't a "more frequently used road". If there were, it would probably be a 2-way stop.

Figuring out who goes next is not remotely as difficult as you have made it out to be. Where I used to live, I would go through two four-way stops twice each day during peak traffic. There are generally two scenarios when you approach the intersection. In light traffic, you get to the intersection and nobody else is waiting to cross, or it is obvious who got there first. In moderate traffic, you will queue up behind one or possibly two cars. Once you get to the front of the queue, you wait for the cars on your left and right that are already at the front of the queue, and then you go. This is simple because each direction of traffic alternates, and it is reasonably fast because the queues are very short.

The real worst case scenario is that during light traffic four cars approach simultaneously, but not only is this extremely rare, but it is not really even a problem when it happens.

In my personal experience I have had far, far more problems in US cities on two-way stops when I am bicycling and a driver incorrectly yields when they have right of way.

I suspect that most Americans feel the same way about roundabouts that, say, Germans feel about 4-way stops. But please remember that roundabouts require more land and much more capital investment. The tradeoffs in the US are different.

> Once you get to the front of the queue, you wait for the cars on your left and right that are already at the front of the queue, and then you go. This is simple because each direction of traffic alternates, and it is reasonably fast because the queues are very short.

Sounds like a high frequency traffic light, but less explicit.

Well, admittedly my experience with 4 way stops is nonexistent (never saw one in europe) and it may simply be a case of not being used to them, but I consider give-way intersections, 2 way stops, roundabouts and finally traffic lights as superior solutions - pretty much in that order and depending on how busy the intersection is. The goal should be to reduce breaking and full stops.

> Also, roundabouts are the superior 4-way stops.

Except on the “compact real estate” axis. A plain intersection is always going to be smaller.

I'm sorry, have you seen some of the British roundabouts? Where the "roundabout" is just a tiny circle painted in the middle? I'm sure it's not bigger than your normal intersection by any measurable standard, and at least everyone knows exactly what the priorities are, you literally can't get it wrong.

We've built something like that on neighborhood streets in the US in the name of "traffic calming". Usually, it's a planter or an island with a single tree. We still put up 2-way or 4-way stop signs though because traffic circles make our heads explode. In the end, all it really accomplishes is making it hard to see other vehicles at the intersection.

Yes it's very easy to keep track of who goes first.

Have a roundabout instead. Easy, safer, and less traffic jams. Seems objectively superior to me.

Are you just conveniently omitting the fact that roundabouts are more expensive and require more land, which may already be used by something else?

That is why mini-roundabouts were invented.

America doesn't really have those... they would be as alien and confusing as a four-way stop in Europe. I can't imagine it not being dangerous.

They exist in some parts of New England and in a few shopping centers in Texas. You're right about them confusing motorists though. New Englanders generally understand what to do, but I've seen many Texans just stop in the middle of the roundabout for no good reason. Maybe if more were built behavior would change.

I was surprised by three new mini traffic circles in Florida on a new road extension. The approaches are embellished with speed bumps. The bumps took away all enthusiasm for the thing.

(Cattlemen Rd in Sarasota)

they would be as alien and confusing as a four-way stop in Europe

For the first year or two perhaps, but I wonder if the long term pay-offs might be worth it.

You can find a few of these small roundabouts in Arlington, VA. There don't seem to be issues with them


Because death is an acceptable trade-off for avoiding irritation - as long as it's not your death, of course.

What a bullshit response, he argues that there are a lot of better alternatives, not that there shouldn't be any safety precautions.

A roundabout is a much better solution than a stop sign. I still believe that the ubiquous stop signs in the US are there due to the oil lobby. Bringing a car to a full stop is incredibly wasteful.

For cyclists, roundabouts aren't that much better unless we're given separate lanes.

Motorists still tend to blow through roundabouts at speed if they can't see anything coming. From personal experience you can easily navigate a roundabout at 35+ mph.

"From personal experience you can easily navigate a roundabout at 35+ mph."

If a roundabout is small enough, you can't. I have seen many small roundabouts throughout Europe that you cannot navigate without swerving faster than 20-30 km/h (which is assumed safe speed for pedestrian).

Plus, at least in my region there always seems to be some sort of obstacle blocking sight in the center of larger roundabouts. Maybe it's just a convenient place for memorials, statues, etc., but I can't help but think that slowing down drivers is at least a welcome side effect.

The obstacles are intentional. Blocking sight forces people to slow down.

Roundabouts are the very last place where I would consider a bike lane to be a possible net win in safety. If the bike lane has priority at exits, then it will be constantly ignored, if it doesn't, then cyclists would be at the mercy of exciting drivers never forgetting to indicate, resulting in similar bloodshed. Besides being totally shitty towards cyclists who would then have to yield three times for going left at a crossroad-equivalent, at a very uncomfortable viewing angle and with permanent risk of un-indicated exits.

I've been curious as to why American roundabouts include unnecessary vegetation or a hill in the middle? Why obscure oncoming traffic?

If it's a roundabout, the traffic across the roundabout is not actually oncoming, it's the traffic approaching from your left around the roundabout that you need to watch out for.

The obstruction in the middle also makes the existence of a roundabout more obvious.

Here‘s an example of how they are handled in Germany:


The space it requires is not much bigger than the traffic light. That town replaced many of its intersections with roundabouts even in central and older parts of the city.

(You might say they went a little overboard with it, looking at the southern limits near that hospital.)

Look at what they have in Berkeley. It addresses exactly this situation

It might be wasteful, but so is killing people.

Stop signs and some speed bumps would soon slow them.

True, but as I pointed out, so would a roundabout. In the area where I grew up we also had a dangerous piece of road that got "fixed" by setting up a "dangerous curve" sign with a warning light on top. It went from multiple fatalities every year to zero.

Having lived near a very similar intersection, also with frequent accidents (it has a 2-way stop with heavy 50-70 mph cross-traffic and frequent left turns, depending on the time of day), cost may be a significant portion of the reason; a neighborhood campaign to replace the 2-way stop with a traffic light fell through due to a >$1 million price tag for adding a light to a countryside road, and it's doubtful a roundabout would cost less.

Anecdata: fixing frequent accident sites (e.g. 70 kph main road x 50 kph access roads) with traffic lights reduces minor accidents, but there was a major accident on one of them just recently...because "traffic lights be damned, nobody ever comes out that road anyw-BOOM." Traffic lights are a discretionary access control mechanism ("ignore at your discretion"), roundabout is a mandatory access control mechanism.

I wonder if that happened because the traffic light was just on a timer rather than taking actual traffic in account. If the smaller road hardly ever sees traffic the light would be red very infrequently and getting your attention when red.

It's a timer, yes - but if you do not have enough attention to see two RED lights on an otherwise empty intersection (one overhead and one right at your eye level), you have no business being behind the wheel, as you either a) are legally blind, or b) have the attention span of a proverbial goldfish.

The suggestion that anything important should be screaming and blinking (otherwise the poor driver couldn't have noticed) is patently ridiculous. There's quite a lot of other, less conspicuous things that a driver needs to be aware of (such as pedestrians or potholes).

A roundabout might cost less. I wonder how much of that signal price was getting grid electricity to the site.

Interestingly, there is a flashing red/yellow light there (similar to [1]) and plenty of houses/powerlines, so it shouldn't be the power costs. I imagine the price of buying and laying signal cable might be similarly expensive, however, as the nearest signaled intersection is about two miles away, if I recall correctly.

[1] https://safety.fhwa.dot.gov/intersection/other_topics/fhwasa...

A roundabout is always the best solution, but it's very expensive. Replacing the "Give Way" by a stop sign costs almost nothing and would probably help a lot, although not for truly reckless drivers.

Traffic lights chop up traffic flow, creating windows of opportunity for turns and crossings at minor intersections downstream. Roundabouts on the other hand create bottlenecks of relatively constant throughput, smoothing out any pulsation in traffic density (e.g. from upstream traffic lights), which makes minor downstream intersections much more difficult.

Determining the best option for a given case is more an at than a matter of simple preferences.

I agree. A roundabout is also a much better than solution presented in the article, whereby traffic would have to turn across oncoming traffic in order to continue on their original track.

I have roundabouts in my neighborhood. They have the exact same problem. If you don't come to a full stop and/or look around your A pillar deliberately, you may never see the car or bike hidden behind it.

I've also had the same problem turning onto a crossroad without coming to a complete stop at the stop sign first.

I believe the only real solution is driver training.

Training is important, but the idea behind roundabouts is "we'll have plenty of collisions here, but all of them will be 30km/h or less".

Obviously you want to having bicycles or pedestrians in your roundabout so adding a bike/pedestrian path next to the roads has to be part of the plan if you want to do it right.

But you modulate speed and rotate the car while navigating the roundabout, which solves the constant speed/constant bearing problem. At a roundabout, blind spots are all over the place and will never perfectly track another traffic participant.

Roundabouts and 4-way stops are used in different situations - a 4-way stop is typically used in residential neighborhoods or street grids where the roads are only single-lane and the lack of available space prohibits a roundabout.

In Europe the residential area would either 1. Have no signs resulting in the car to the right getting the right of way 2. One road gets a right of way sign and the other road had to yield.

I strongly prefer those setups because they are much more fuel efficient.

However, I don't think any of that would work in the UD because it would require real driving classes and tests. I had to drive around for close to 20 hours with a professional driving teacher on to of dozens of hours of theoretical classes. All that end in a 30min+ examination with a examiner in the back of the car putting me through as many challenging situations as possible.

In the urban eastern U.S. apartment buildings usually run right up to the road, except a ~ 1-3meter sidewalk. There isn't sufficient visibility to yield safely; both directions have to come to a stop anyway to check for cross traffic.

In the western U.S. (and eastern suburbs) the 4-way stops are largely vestigial; I'd agree that most of them could go away without any loss of safety. Many of the newer neighborhoods are signed that way anyway, though - I just drove through one this morning with 2-way yields.

My driver's ed class was IIRC 36 hours of instruction and 12 hours in the car, with a professional driving teacher. It had an hour long examination putting me through as many challenging situations as possible at the end.

It's also a matter of being used to bicycle traffic, and training. In the Netherlands in a residential area there are probably more bicycles than cars, so every car driver is used to being careful at junctions.

And on average a person needs 38 hours in the car with a professional teacher before they pass the exam, most people don't pass on the first try.

Most middle- and upper-class Americans take a driver’s education course (like what you’ve described) prior to receiving their driver’s license. In 10 years it won’t even be necessary because nobody’s going to be driving themselves anyway.

A lot of the cars on the road right now are ten years old or older. Even if self driving cars became a thing that anybody can afford right now it would take longer than ten years before nobody had to drive themselves.

In my country (Finland), the average age of cars used in traffic is 13 years.

Except pedestrians, bikers, cyclists, etcetera. (Plus you are overestimating the rate of change: I still see horse carriages on the road, 100+ years after the beginning of the car era)

And don't get me started on driver ed in PA, USA: in my case, it consisted of a video, signing an drivers' permit application - and voila, go out there and drive (under supervision). I was shocked, shocked.

In the UK you can get a provisional license that allows supervised driving with zero training. It's assumed your supervisor (anyone holding a driving licence for something like 7 years or more) will stop you doing stupid shit.

Interesting. I was so used to the continental idea "driver school takes at least weeks and is like any other training, except it's in a car" that the opposite approach was highly unexpected to me; it does make sense though.

>> In 10 years it won’t even be necessary because nobody’s going to be driving themselves anyway.

Try 50+ years. Driverless cars aren't happening anytime soon, and certainly not to a level where you literally don't need a licence to drive one.

And public transport definitely isn't going to be taking up the slack given the catastrophic decline in public investment therein.

There are plenty junctions that have been turned into a roundabout simply by painting a circle in the middle and putting up the necessary signs. They don't need that much space, especially in places that are low speed, low traffic volume like residential neighbourhoods.

Indeed, a stop sign would solve this, wouldn’t it? Simplist & cheapest solution first. The point about the geometry is that it doesn’t matter at which point the driver can see over the hill and brush; they still won’t see the cyclist.

Apparently not. I had a very similar junction outside my house where cars crashed constantly though usually non fatally as it was in a 30 zone. I wrote to the road engineer folk asking for a stop sign and apparently motorists don't take much more notice than of give ways. In the end the cross roads were replaced with two mini roundabouts which have fixed the problem. Physical obstacles are much better at making cars slow down than signs.

Most UK cities will have one single stop sign somewhere, just so that the driving lessons can go past one.

I thought only the US loved stop signs.

Poland also loves stop signs. Actually, Poland just loves signs in general which is a problem: https://d-w24.ppstatic.pl/g2/6b/6f/65/116789_1258901612_33ec...

Well, one reason for that amount of signage is that there are multiple roads from the right and some of those signs only extend to the next junction. There's simply no way to have fewer of those signs. Same for the pedestrian crossings.

Some of the signs are only relevant to pedestrians and cyclists, or motorists looking for a place to park, so it's not that bad.

It can be way better. Some ideas:

- pedestrian/cyclists signs can be made on the pavement instead

- which lane goes where is repeated 3 times - if it's so that you know them when you join the road, 2 can be moved to the side roads instead

- one of the right of way signs can go

- no parking can be replaced with a chain / bolards

- the end of pedestrian/cyclist path is useless since it's reopened after the crossing

Poland puts the repeated signs after each and every tiny side road. Many countries don't do that. Instead they rely on more informative big signs over the intersections that can be visible from further away.

I'm not sure the signs on the pavement have the same meaning as the actual signs. In many places (Germany for example) »signs« on the ground are merely a courtesy (e.g. the 30 in 30 zones), but not legally binding, and the actual signs define the rules.

They're called horizonal and vertical signs in Poland and they are binding. The solid line between lanes counts as a horizontal sign as well.

Wow, that is crazy! Now imagine those were US style signs that are covered in text instead abstract symbols...

Thanks for the link but it doesn't say why?

There are stop signs now (as of June 2017) and STOP markings in the road. Streetview shows old imagery from 2015 when viewed from the offending road but Beaulieu Rd. is newer.

In the US, they would put up a stop sign and park a cop near there for some instant, costly learning experiences. Although, if the speed is sufficient you would probably get an overpass, particularly if one road is a known truck route. It would be interesting to build a bicycle only overpass.

The US also tends to overuse stop signs when a yield (give way) sign would suffice. That is, if there is sufficient visibility of approaching traffic, coming to a complete stop is not necessary to safely yield to any conflicting traffic. Also, coming to a complete stop and then proceeding isn't sufficient to properly yield to approaching traffic.

In the end, stop signs end up being treated like yield signs. The only time they're properly treated like stop signs is when it's not possible to see cross traffic before you approach the intersection.

In the end, stop signs end up being treated like yield signs.

This definitely varies based on where in the US you are. In California almost no one comes to a complete stop. (That's probably why rolling through a stop sign is called a "California Roll"). But in many other parts of the country people do, even if there are no cars around.

Some of this may be that stop signs in California tend to be more superfluous.

In New England & NYC, a 4-way stop is usually used for urban neighborhoods where single-lane, low-traffic streets meet. Roundabouts and slip lanes are space-prohibitive in these situations (they'd have to go through privately-owned lots), traffic lights are overkill given the light traffic, and visibility is insufficient for a safe 2-way stop or yield.

In CA and other western states, these same neighborhoods are usually suburban houses with lawns, and even the apartment buildings have setbacks from the street. A yield would work fine, but either for habit or consistency, it's signed as a 4-way stop. Drivers feel like they have good visibility and it should be a yield, though, and so they roll through it.

While I may California Roll occasionally, at a yield sign I may not slow at all, or only slow enough to safely make a turn. So there's definitely a difference between the two, even if you treat the stop sign as "just a guideline". Stop means go slow enough that you can make a complete stop if you see police in the intersection. "Oops, missed the limit line by a couple feet, but I was totally always gonna stop."

>In California almost no one comes to a complete stop. (That's probably why rolling through a stop sign is called a "California Roll").

It probably depends on where in CA. Have been living in Bay Area, mid-Peninsula, for 18 years and i've so far seen most people (and myself included) do complete stops. And just recently saw police stopped and cited a cyclist for the "California Roll" in Los Altos.

I know that practice by the name "California Stop"[1]

[1] https://www.urbandictionary.com/define.php?term=california%2...

In my experience driving through many states in the eastern part of the US, it's not really any different compared to your description of how drivers in California deal with stop signs.

> The US also tends to overuse stop signs when a yield (give way) sign would suffice. That is, if there is sufficient visibility of approaching traffic, coming to a complete stop is not necessary to safely yield to any conflicting traffic.

From TFA there already is a give-way sign - drivers who should be yielding think they have good visibility, blast through the intersection at full speed, and kill somebody. The article suggests this could be due to the CBRD phenomenon.

Even if people treated the stop sign as a rolling stop sign, slowing down seems like it would help, at least WRT drivers not seeing people due to CBRD.

Of course there are fancier (more expensive) engineering solutions too.

Even then, they're still not treated like stop signs. Drivers blow past the sign up to the farthest possible point and only stop there, if at all. In particular they speed right over the pedestrian crossing without looking.

That's because in America you grandma can legally teach you how to drive and so you learn all her bad habits. It's a feature, not a bug. And your driving test might cover this type of intersection or it might not. The testing is like 50 out of 5000 possible questions. Again, it's a feature (of idiocy) not a bug.

I personally know maybe one other person, who knows that the location of the stop sign defines a line perpendicular to your lane and you're supposed to stop completely before this imaginary line. Almost no one knows this is the law. And sure as fuck no one actually does it.

I'd love to have a light saber for the sole purpose of cutting off everyone's goddamn front end of their car right along this imaginary line. Oh, your front seat is in the imaginary line? So sad...

> the location of the stop sign defines a line perpendicular to your lane and you're supposed to stop completely before this imaginary line. Almost no one knows this is the law.

I am inclined not to believe this. I have been told by a traffic cop before that if there is no marked stop line or crosswalk, there is no legally defined point where the stop must happen: it's just a "use reasonable judgment" kind of thing.

That was in NC. The following is apparently a quote from a statute in Washington State that contradicts your statement:

“every driver of a vehicle approaching a stop sign shall stop at a clearly marked stop line, but if none, before entering a marked crosswalk on the near side of the intersection or, if none, then at the point nearest the intersecting roadway where the driver has a view of approaching traffic on the intersecting roadway.”

Can you point to a statute somewhere supporting your claim?

driver of a vehicle shall stop at a stop sign or at a clearly marked stop line before entering the intersection, https://www.revisor.mn.gov/statutes/?id=169.30

And stopping right before the stop sign is what I was taught in driver's ed, not by my grandma.

One of the joys of driving in the U.S. is every state has it's own driving laws and while they're substantially the same, they are by far not identical (see lane splitting in California vs everywhere else, and turning right on red).

Also you'll find some states will define the intersection as including sidewalks, which is why you'll usually find the stop sign placed right before the sidewalk, not after it. http://www.ncsl.org/research/transportation/pedestrian-cross...

Actually, you can't pass the line with your wheels without coming to a full stop, so you have to account for the engine overhang, you can't just zap the cars at the line with your lightsaber... :-P

This CBDR problem would seem to imply that we actually don't underuse stop signs. I bet if you found a similar junction in the US you'd find a much higher rate of accidents because we're all so lousy at driving.

Americans don't do yield signs very well except in the most obvious situations that shouldn't even need yield signs, like highway on ramps, and intersections with islands creating curved right turn lanes. Americans barely have competency at four way stops, the main way who goes next is figured out is someone launching from a rolling stop before the others. There's also often hand waiving, honking, and lights flashing because people are just plain ignorant the actual law, or even common sense.

And traffic circles? Fuck that's super confusing! Americans would rather get into accidents at a four way stop than be confused in a traffic circle. That's why traffic circles are so uncommon.

I’ve personally rarely encountered issues with yielding. Four-way stops are another issue, and any driver in the U.S. has probably encountered them, but they’re solveable issues with basic driver courtesy. They assume basic understanding and cognizance to prevent incidents, but nothing more is strictly required.

The greatest number of accidents I’ve seen have been at stop lights rather than stop signs, weirdly. Driver apprehension and cautiousness at four-way stops may actually prevent more incidents than other systems.

I encounter problems with yielding nearly every day, but I think it has more to do with road design. The yield is on a very long acceleration ramp and it is placed towards the front of the ramp. Then there is a merge sign near the end of the ramp.

I think the idea is that you should yield and only go when it is clear. But because of the length of the ramp, I think they are legally required to have a merge sign there. Invariably people fail to yield, and, since it is uphill, 'merge' into 45 mph traffic at 20 mph. And by merge I mean they just drive on and expect people to stay out of their way.

I have way more problems with people turning on red at the worst possible moment (despite lots of previous safe opportunities) than I do with yield signs though. But I don't doubt the GP's experience either. Drivers are very different in different states and regions of the U.S.

When the sight lines are blocked, the false sense of safety which is undermined by theconstant speed/constant bearing blind spot simply would not happen. If you don't think that you know that the other road is empty (because you can't see it), a yield sign is enough.

Blocking the view? Strange, I think the views to the road that crosses this one are remarkably open here. Before looking at this, I thought there could be some hedges or such for which England is famous, but I can't think of any excuse for why a driver doesn't see a bike when approaching from [1] or [2]. The visibility is very good.

>> The 'Give Way' sign is idiotic, put a Stop sign on the other, flatter road, or make it a 4-way stop. The drivers causing accidents were blowing through the intersection, and shame on the wording of the law that they were able to weasel out of harsher punishment.

It seems the cyclists are blowing through the intersection as well. To me the problem is simply that it's an uncontrolled intersection where nobody is required to stop. Why blame the driver so much?

The cyclists were on the road that has priority, whereas the motorists were on roads were they are required to give way. Should the road user who has priority give way to the road user who doesn't?

100% agreed. This is just basic common sense: don't take a junction at a speed greater than 20km/h, no matter what. That's it. I slow down instinctively at junctions. I literally cannot see the other road because of trees/buildings. How can I assume it is clear?

Same is true for bicyclists. Slow. The. Fuck. Down. At. Intersections.

For low volume intersections, roundabouts are great at avoiding accidents.

The math is wrong. It ignores binocular vision and normal head motion.

4-way-stops aren't really 4-way stops. Nobody stops at them.


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact