As a lover of user experience - and, by extension, psychology and sociology - I love noticing when people engage in odd behavior, particularly when I see the same mistakes happen again and again.
Now that I live, work, go to the gym, and just-about-everything-else in downtown Seattle, I find I'm doing a lot more walking around the city. Not only is this a decent way to sneak some exercise in edgewise, but it also lets me watch people drive around a lot more, which is endlessly entertaining in that gallows humor sort of way. I get to watch people making the same mistakes over and over again, ad nauseam. Most of these are so common that it's hard to find a stand up comic who hasn't humorlessly lamented about them, but one case has proven to be surprising: driving the wrong way down a one way street.
Obviously, I know that this happens - I've been in the car when people have done it. What I wasn't aware of is that it happens with a startling regularity. At least once a week I'll hear a chorus of car horns singing the lament of a pair of headlights facing the wrong way. Granted, one car in a week's worth of cars is not a large portion, but this is a case where it only takes one car making a mistake to cause some havoc.
I wouldn't have expected that people frak that up as much as they do, but in retrospect I suppose it isn't so surprising. I mean, the one way street is an artificial construct. There is nothing about the physical medium of a street that suggests that you can only drive on it in one direction. Put another way, you can drive on it in more than one direction, you're just not supposed to. Sure, there are signs at every intersection where there's a one way street, but they can be missed. I'd guess that the most obvious indicator of a one-way street - the closest thing to an affordance it can offer - is a stream of cars that are only flowing in one direction. Short of that, it only takes a small bit of preoccupation for a driver to not notice signs and signals.
Even more interesting is that even people who don't regularly make wrong turns down one way streets will often be able to empathize with what a pain in the ass they can be. Have you ever been driving downtown trying to find an unfamiliar place (or even just been along for the ride)? Missing a turn becomes quite a hassle, doesn't it? Because now the go-back-around-the-block correction is non-trivial. (If you're still in doubt, bring up driving in Boston with anyone who's ever done it and I guarantee you one way streets will factor into the conversation, followed closely by fist shaking and/or chronic alcoholism.)
So, the one way street is a phenomenon that is not natural to the medium, is missed with some regularity, and is a frequent bitching topic around the water cooler. Ergo, bad UX, right? Well... yes and no. Yes, because people are clearly having bad (wait for it) user experiences driving around on one way streets. But the entire purpose of a one way street is to improve traffic flow, reduce congestion, thus improving the reliability of the street grid and, by implication, the experience driving on it.
This brings up some issues that I think are fairly pervasive when designing user experiences. First, users often can't see the forest for the trees. In fact, that's often the point of the design. In my overly-loquacious one way streets example, by and large it is impossible for users to judge that this experience is better than it would otherwise be if all the streets were two-way. They have no point of reference to make a comparison against, and that's by design. The worse way has been eliminated in favor of this better way, which, unfortunately, still kind of sucks at the individual level, even though it's better for the group as a whole.
And that's the other thing that's difficult to communicate through design: a design decision that improves the experience for the group as a whole, but is somewhat undetectable from an individual's perspective. Even worse, I think that it's often the case that even if you manage to effectively communicate this idea (either within the design itself or, as in the case of one way streets, tangentially through documentation/PR), users just don't care. Users have goals, and if the system/design gets in the way, they often really don't care if this personal inconvenience makes things better in general. If they had their way (and were being honest about it), they'd make you change it so that it was great for them even if it screwed over everyone else. But, in a system as complex as a traffic grid, where no man is an island and actions/reactions are very interconnected, such a design strategy is also a losing proposition; the negative side effects would likely more than outweigh the benefits of that action.
I have no idea when, if, or how people will figure this stuff out, but I hope they do. In the mean time, it'll be fascinating to watch them try and figure it out.