|Steve Avery generated this plot with a random number generator|
I do worry about taking the Minneapolis safety assertions without a grain of salt. Plotting commuters against crashes (second figure) makes the assumption that commuters are a proxy for cyclists exposed to the risk of crashes or of total cyclists. However, if you look at actual data, turns out that most crashes are during the evening rush hours while few are during the morning rush hours. If commuters really are the population that is crashing, why are they only crashing while going home? I would want to test whether after-work or after-school recreational riders are crashing, thus helping create the bulge in the return rush hour numbers.
|Actual Minneapolis data. Source document in pdf is here|
Smeed's law, as well as mechanistic attempts to explain his numerical observations, go back a long way. I am sure there are advantages to having more cyclists on the roads, i.e., motorists have more experience interacting with bicyclists, there is an expectation of seeing cyclists, more money is spent on bicycling education and awareness, etc.. But I wouldn't count on numbers alone to keep one safe. This article for example suggests just the opposite of "safety in numbers" for motorcyclists (and based on how I saw or read about young servicemen riding their motorcycles in Honolulu, I can understand why--young riders were often engaging in high risk behavior, were riding beyond their competence, often on extremely high performance vehicles, or were drunk.)
Although one doesn't have 50-100 horsepower to abuse on a bicycle, take it from my personal experience: doing something dumb on either a motorcycle or on a bicycle is an excellent way to get hurt, no matter how many other people are riding today. Don't let "safety in numbers" convince you to be complacent. Its not simply more butts on bikes that makes for safety, but more brains on bikes.