{"id":2974,"date":"2013-07-19T09:56:44","date_gmt":"2013-07-19T15:56:44","guid":{"rendered":"https:\/\/iamtraffic.org\/?p=2974"},"modified":"2013-11-01T09:38:14","modified_gmt":"2013-11-01T15:38:14","slug":"china-cups-and-butterflies-options-and-ethics","status":"publish","type":"post","link":"https:\/\/iamtraffic.org\/news-views\/china-cups-and-butterflies-options-and-ethics\/","title":{"rendered":"China Cups and Butterflies; Options and Ethics"},"content":{"rendered":"

\"china-butterfly\"Vulnerability and risk. Statistics and ethics. Solutions or fixes. Top-down interventions or individual actions. These are the core issues in the long-running bike-lane (or cycle track)-versus-integration argument and in the book Antifragile<\/em> <\/a>by Nassim Nicholas Taleb (better known for his previous book, The Black Swan<\/em><\/a>). Antifragile<\/em> is a long and complex read, but the author managed to summarize it while metaphorically standing on one foot: \u201cEverything gains or loses from volatility. Fragility is what loses from volatility and uncertainty.\u201d<\/em><\/p>\n

We know what fragile means; the characteristic of being at risk of serious harm or destruction through sudden and extreme shocks. We also have the term robust, which means something is mostly immune to sudden extreme shocks. Taleb noticed there are also things, people, and systems that improve with such shocks, which required the new term, antifragile<\/em>. There is no hard line between fragile and antifragile; rather it is a continuum. Taleb explores this fragility to antifragility continuum throughout the book.<\/p>\n

We bicyclists can\u2019t make ourselves totally antifragile, to the extent that we can benefit from the volatility of traffic, but we can significantly reduce our fragility and become somewhat antifragile by learning to respond effectively to the many small violations we experience. These small violations are often indicators of the actual crash we\u2019re trying to avoid. The close pass indicates the risk for the sideswipe. The right hook close call indicates the potential for a right hook collision. And so on. We cannot and will not become less fragile or more antifragile by putting most of the responsibility for our safety on government or motorists. The best way to achieve it is by changing our own behavior.<\/p>\n

Butterflies are very vulnerable, but similar to bicyclists they have the\u00a0options<\/em>\u00a0of movement which reduces their fragility; so much so that they have survived as species for millions of years.<\/p>\n

Before we explore how to reduce our fragility as cyclists, it\u2019s important to differentiate between fragility and vulnerability. A china cup is both vulnerable and fragile, we can reduce its vulnerability by protecting it with bubble wrap, but it remains fragile. It\u2019s also useless when wrapped up in the bubble wrap. Butterflies are very vulnerable, but similar to bicyclists they have the options<\/em> of movement (as well as coloration and chemistry) which reduces their fragility; so much so that they have survived as species for millions of years. (Options<\/em> is a major theme of Antifragile<\/em>.)<\/p>\n

Bicyclists don\u2019t collide with motorists because they\u2019re vulnerable or slow; but mostly because they\u2019re unpredictable or inconspicuous. (Please don\u2019t misconstrue this as fault; predictability and conspicuity are relative, and of course some motorists are careless. But when we look at the data, it\u2019s clear many — if not most — crashes can be avoided by improving cyclist predictability and conspicuity.) As cyclists, our vulnerability cannot be significantly reduced, but our fragility can. To reduce fragility we must reduce the volatility and uncertainty of our environment, and increase our options.<\/p>\n

We must also bear in mind that we\u2019re not in this by ourselves; motorists also need a more predictable environment, because they are also fragile. Yes, motorists are fragile \u2013 emotionally and economically. A collision with a cyclist is often damaging to a driver\u2019s psyche or bank account, or both. Well-designed streets and adherence to well-reasoned rules reduce volatility and uncertainty for both cyclists and motorists. As cyclists our options are increased by enhancing (but not violating) the rules of the road with special strategies.<\/p>\n

\u201cA complex system, contrary to what people believe, does not require complicated systems and regulations and intricate policies. The simpler, the better. Complications lead to multiplicative chains of unanticipated effects. Because of opacity, an intervention leads to unforeseen consequences, followed by apologies about the \u201cunforeseen\u201d\u009d aspect of the consequences, then to another intervention to correct the secondary effects, leading to an explosive series of branching \u201cunforeseen\u201d\u009d responses, each one worse than the preceding one.\u201d<\/em><\/p>\n

\u2014Nassim Nicholas Taleb<\/p><\/blockquote>\n

Rules Trump Laws (or at least they should)<\/h2>\n

Adding more rules will not solve the problem, because adding more rules adds points of failure. Instead of more rules, cyclists need more options to avoid the errors of others.<\/p>\n

When it comes to rules it\u2019s essential to understand how they relate to laws and to risk. Rules precede laws; they usually arise out of the culture out of experience and consensus. Laws often have political motives and biases; sometimes they uphold the rules, but other times they usurp them. When it comes to risk, particularly in traffic, we must recognize that rules which are violated become points of failure. (This is not always so with laws, because some laws are frivolous or not intended to reduce risk.)<\/p>\n

The vast majority of traffic crashes — regardless of mode — happen because one or more parties violated one or more rules. What\u2019s more, even if one party did not violate any rule, they often failed to take effective action to prevent it. I\u2019m reminded of my early auto driving days as a teenager. I was driving along a collector street and another car was approaching from a side street. He was facing a stop sign and I wasn\u2019t. From the passenger seat my dad mildly chastised me for not covering my brake. If that other driver had pulled out and caused a crash he would have clearly been at fault and received the citation, but it would have also been because I failed to take appropriate action. I would have failed to take an option available to me.<\/p>\n

Adding more rules will not solve the problem, because adding more rules adds points of failure. Instead of more rules, cyclists need more options to avoid the errors of others. Our most important options are sufficient operating space and adequate line of sight. We can also learn ways to give motorists more options to avoid unnecessary conflicts with us.<\/p>\n

So we need a traffic system with the fewest number of rules (not laws) possible, with street designs that support those rules, with the lowest volatility we can manage, and with the greatest options for all. This means we need a system based on what people do best and most naturally, not on what they do poorly or need to be nagged into doing. Many bicycling advocates are calling for systems that increase the failure points and reduce options both for cyclists and motorists.<\/p>\n

The rules we have are based on the limits of human perception. Our eyes are at the fronts of our heads and we don\u2019t have x-ray vision. They are also based on years of practical experience, first on open waters where boat captains needed to avoid collisions, and then on roads and streets when cart and carriage drivers needed to do the same. It was only when the industrial revolution came that cities got large enough and traffic got thick enough that the rules needed to be formalized into laws. Eventually traffic control devices were created to help manage the movements and reduce delays. For those who think the rules of the road were created for motor vehicles, note that less than 10% of urban traffic was motorized when William Phelps Eno wrote the first formal traffic codes for New York City in 1909. Eno thought cars were a fad and would be gone in a few years. He wrote his code for horse-drawn vehicles, bicyclists, and pedestrians, with autos, the horse-drawn and bicycles all categorized as vehicles.<\/p>\n

The essential fairness of the rules of the road is based on the idea that all people are equal, no matter their mode. That\u2019s why the core rule is:<\/p>\n

\"Rules
The basic rules of the road are simple and designed to optimize human capabilities\u2014yield to what’s directly in front of you.<\/figcaption><\/figure>\n

First Come, First Served.<\/strong> You yield to the traffic in front of you because the person ahead of you in the lane was there first. Just like at the grocery store check out. You yield when entering a roadway from a driveway because the people on the road were there first. The same goes when you get to an intersection; you yield to the person who got there before you did. When changing lanes; you yield to the person who is already in that lane. Drivers yield to pedestrians who\u2019ve entered the crosswalk because the pedestrian was there first.<\/p>\n

Next comes Keep Right; Pass Left.<\/strong> We have to all agree to keep to the same side of the roadway when going in the same direction, or else it\u2019s chaos. And we want passing to be limited to one side to minimize the amount of scanning we need to do, and because, as I noted earlier, we don\u2019t have x-ray vision. When you pass on the right you can be hidden from view to left-turning drivers approaching from the front.<\/p>\n

Intersection Positioning<\/strong> reduces the number of conflicting movements at intersections by eliminating or minimizing converging movements, instead making all movements parallel or diverging.<\/p>\n

Yield To the Right at Intersections.<\/strong> If two or more drivers arrive at an intersection at the same time, they yield to the drivers to their right. Why the right? Because motorists sit at the left side of the vehicle and the line of sight is better to the right. (Roundabouts are a special case. They\u2019re really a set of T intersections around a circle, and with the bottom of the T yielding to the top.)<\/p>\n

Yield To On-coming Traffic When Turning Left.<\/strong> The purpose of this is to minimize delay and increase certainty. And as noted above, the turning driver needs to be able to see the traffic he\u2019s supposed to yield to.<\/p>\n

Notice that with the exception of lane changing, drivers are to yield to people who are in front of them or to the side but still ahead. Our eyes are at the fronts of our heads. Lane changing requires a rearward scan, but lane changing is only supposed to be done mid-block, not at intersections.<\/p>\n

In earlier times there was no rule requiring slower traffic to yield to faster traffic. That would conflict with the rule of first come, first served. Today minimum speed laws are in place in many states, but they are only intended to keep drivers of motor vehicles from unnecessarily slowing other motorized traffic. If such laws are applied to bicyclists or other non-motorized users, they are no longer about equality; they are effectively fascist, favoring the powerful over the powerless.<\/p>\n

Seeking Order; Gaining Failure Points<\/h2>\n
\"Rules
Channelizing one type of vehicle on the edge of the road adds complexity, breaks the basic rules of the road and defies human capabilities.<\/figcaption><\/figure>\n

When we add a bike lane or cycle track to a street we effectively add new rules and violate some of the original rules. With such facilities, for bicyclists we now say:<\/p>\n

Pass On the Right.<\/strong> The right rear section of a motor vehicle has the worst blind spots, but we expect drivers to check those spots with the same level of effectiveness as when they look forward or into their left side mirror.<\/p>\n

Bike lanes discourage proper intersection positioning; cycle tracks often prohibit it. \u201cWhen do I leave the bike lane to get into the left turn lane? Or should I just turn left directly from the bike lane?\u201d<\/p>\n

And for motorists we now say:<\/p>\n

When Turning Right, Yield to Through Vehicles Approaching From Your Right Rear<\/strong> (your blind spot), while also scanning ahead for pedestrians in the crosswalk.<\/p>\n

When Turning Left, Yield to Vehicles You May Not Be Able to See<\/strong> (because they\u2019re moving toward you while hidden behind other stopped vehicles).<\/p>\n

Motorists are often confused as to how to do intersection positioning. \u201cDo I move all the way to the curb as I normally do, or do I stay fully in my current lane?\u201d<\/p>\n

So we have added new rules, which means new points of failure. What\u2019s more, these rules are more difficult to obey because they require motorists to look back when they also need to be looking forward, and they require both cyclists and motorists be able to see through stopped vehicles. None of these new rules, or the bikeways that necessitate them, eliminate the need for adherence to any of the original rules, or eliminate the crashes caused by the violations of those rules.<\/p>\n

Types of bicyclists and types of bicyclist behaviors.<\/strong><\/p>\n

Bicycle planners are fond of systems that attempt to categorize bicyclists into \u201ctypes.\u201d The most commonly used type system uses \u201cA\u201d for \u201cexperienced\u201d adults, \u201cB\u201d for \u201cnovice\u201d adults, and \u201cC\u201d for children. The problem with this system is it assumes certain types of behaviors and preferences for each group, but with no data to support those assumptions. A far better approach is to categorize basic types of behaviors. Some behaviors are more successful at mitigating conflicts and crashes, and some are less successful. Whether a cyclist is a 40 year old man on a $2,000 road bike or a 12 year old girl on a bike from Target is beside the point. Successful behaviors work for everyone. Unsuccessful behaviors fail for everyone. (The presumption that successful bicycle driver behavior is not achievable by novice adults or by children in an appropriate environment is demonstrably false. It\u2019s been done.)<\/em><\/p>\n

On the other hand, the basic behavior types are driver, edge and pedestrian. Pedestrian behavior is bicycling on a sidewalk, a sidepath, or on the roadway against traffic. Edge behavior is cycling close to the roadway edge with the flow of traffic. Driver behavior is using the roadway in the same manner as any other driver of a slower vehicle, usually in the center of a lane. It\u2019s not a matter of whether a behavior is \u201cright\u201d or \u201cwrong,\u201d but whether it\u2019s successful or unsuccessful for a given type of street and situation. The well-trained cyclist understands the pros and cons for each behavior.<\/em><\/p>\n

For more about the bicyclist behavior spectrum, see Bicyclist Behaviors & Crash Risk<\/a><\/p><\/blockquote>\n

For some strange reason, the opinions of the untrained and less experienced cyclists are held in higher regard than those of the trained and experienced ones. I know of no other activity in which that is the case.<\/p>\n

A bicyclist whose default behavior is driver sees a bike lane or cycle track as adding volatility and limiting options. A bicyclist whose default behavior is edge or pedestrian sees a bike lane or cycle track as reducing volatility and increasing options. This is partly because he or she has little or no experience using driver behavior, and partly due to what the majority bicycling culture is saying. It also creates some cognitive dissonance on the part of the edge or pedestrian rider, because he still experiences conflicts while in the bike lane or cycle track.<\/p>\n

For some strange reason, the opinions of the untrained and less experienced cyclists are held in higher regard than those of the trained and experienced ones. I know of no other activity in which that is the case.<\/p>\n

\u201cWhat can we say about people who studiously avoid learning from the mistakes of others?\u201d<\/em> — Taleb<\/p><\/blockquote>\n

There are better ways to attempt to reduce volatility and improve options that don\u2019t involve violation of the rules of vehicular movement. The best is to reduce motor vehicle speeds in urban and suburban areas, and to limit the number of multi-lane arterials. The faster we travel the poorer our ability to track multiple conflicts and the more distance we need to react to them. Crash data from Florida shows that crash rates for all road users increase significantly when roads are widened from four lanes to six.<\/p>\n

We can accept the superiority of bicycle driver behavior over edge and pedestrian behavior by the simple observation that people who effectively learn bicycle driving do not revert to the other behaviors as their default; driving becomes their default because it clearly works better for them (though bicycle drivers will on occasion use edge or pedestrian behavior when it serves a specific purpose.)<\/p>\n

\u201cSuckers try to win the argument; non-suckers try to win.\u201d<\/em><\/strong><\/h2>\n

\u2013 Taleb<\/p>\n

This bike-lane-versus-integration argument has been running for 30 years at a stalemate, and statistically speaking it\u2019s unwinnable. The argument is in essence: overtaking crashes versus other crashes. The experienced and trained bicycle driver sees the risk of the overtaking motorist as exceptionally low and the risks of turning and crossing conflicts as significantly higher. (Overtaking crashes account for no more than 10% of urban and suburban crashes, while most of the rest involve turning and crossing conflicts.) Based on direct, real world experience they see bikeways that keep cyclists along the edge of the road as increasing their risk for turning and crossing conflicts. Virtually all cyclists who predominantly use driver behavior were once edge or pedestrian riders, so they understand the difference in effectiveness based on real world experience.<\/p>\n

Cyclists who predominantly use edge or pedestrian behavior have little to no experience with bicycle driving behavior, so they are unable to compare the effectiveness for themselves.<\/p>\n

Advocates of bike lanes, cycle tracks and sidepaths see the risk of overtaking motorists as unacceptably high and believe the turning and crossing conflicts that bikeways create can be mitigated with better design and motorist training and enforcement. They also tie the tenuous safety benefit of bikeways to the desire to get more people on bikes, making the equally tenuous claim<\/a>\u00a0that getting more people on bikes will reduce the crash rate (but not the crash numbers). Cyclists who predominantly use edge or pedestrian behavior have little to no experience with bicycle driving behavior, so they are unable to compare the effectiveness for themselves.<\/p>\n

If you\u2019re going to trade off one risk for another, you have to first know which risk is greater. Statistically speaking, that means knowing the absolute risk in a measure such as crashes per mile or per hour of exposure. What\u2019s more, you need to know how those risks change with different environments. For example, the risks of turning and crossing conflicts are relatively low on a high-speed rural road with few cross streets and driveways, but one could reasonably expect overtaking crashes to be a potential problem. Conversely, on a low-speed downtown street with many driveways and cross streets, turning and crossing conflicts are a significant concern and overtaking crashes are exceptionally low. Lane control greatly reduces the predominant type of overtaking crash, the sideswipe, in which motorists attempt to squeeze past cyclists within the lane. Therefor, the only type overtaking crash that lane control might not mitigate is the type in which the motorist is so seriously impaired or distracted that he or she does not see the cyclist directly ahead. Such crashes in the urban context are so rare they could be compared to being struck by lightning.<\/p>\n

In order to measure the risk for a particular type of crash and how it might change by the type of street, you need to measure quite a few variables, and be confident you\u2019ve measured them accurately. To put this to use, you also need to create a sound mathematical model that reliably predicts the risk for different situations.<\/p>\n

Let\u2019s take one type of crash \u2013 the right-hook \u2013 as an example. In order to predict how many right-hook conflicts per mile or per hour cyclists would expect if they are traveling along the edge of the roadway, we need to know:<\/p>\n