Inspired by CommuteOrlando last December, I started counting and characterizing traffic interactions. Pretty soon, I realized I could use that information to make my bicycle commute less stressful and safer at the same time. I think you'll agree. And it's dirt simple.
I count bicycle/motor vehicle interactions. If I happen to ride on a MUP, I'll add bicycle/pedestrian/dog interactions (though I did have one armadillo interaction).
Here's how I peg them. Your system will adapt to your own tastes.
- Each car passing me on a two lane road counts as an interaction
- A car I happen to notice changing lanes to pass me on a multi-lane counts as one
- Other cars passing on a multi-lane don't count - unless some dweeb two lanes over feels he needs to honk because I'm on his road (remember -don't let your paranoia get out of hand here)
- Cars going the other way don't count, UNLESS there's some sort of real interaction
- Cars that wait for me, or that I wait for count as one each. Ditto if they should have.
- Cars at a stoplight when I shoot through on a green don't count for anything unless I notice a driver looking to make a turn that'd affect me or something similar.
- No more than a count of five at any intersection (otherwise, you could rapidly reach 20 or more at a busy intersection or even a clogged 4-way stop)
- No more than a count of five if you're at a stop sign, watching cars whiz by on a busy arterial you're going to cross or ride on.
- One interaction for each clump of pedestrians you pass on a MUP
- One interaction for each dog associated with any clump (that might run the count up a bit) with a max of five per clump.
- Anyway, try it, you'll rapidly get the hang of it. Remember, we're doing an engineering approximation here, not some science project.
- For a variation, you can count things like debris and fall risks, though I've not tried that myself. I've counted motorists making passes in no passing zones. I know that, after noticing a lot of female joggers in Keller over most of a week, I discovered that 80% of the joggers along Bear Creek Parkway were female, in contrast to male majorities elsewhere. However, back to the subject...
That's pretty much it. For a first trial, count the interactions between intersections versus at intersections. Compare with the count when you ride on a MUP (Multi-Use Path). Take a couple of experimental runs through a parking lot. Ride close to the store versus the far side. You can refine things further and compare interactions in all sorts of ways. Myself, I discovered I pick up about 3X interactions in the afternoon compared to morning. It's no wonder the motorists are crabbier in the afternoons. I also discovered that 80% of my interactions were at intersections, which was a surprise since I'd always noticed cars passing more - until I got quantitative data. I discovered that about 0.01% of my interactions resulted in some sort of motorist bad behavior. Finally, and relevant to a couple of upcoming posts, I noticed that parking lots, especially in the afternoon, shot the interaction count right off the charts. It changed my default view of parking lots and I began to treat them as potentially dangerous, if still useful in rare circumstances.
Remember, every interaction is a potential collision. Count them and you'll soon find the count dropping as you understand how they occur on your ride and how you can affect them.