Michael Darnell writes on baddesigns.com about USB:
When you try to plug it in the wrong way, it doesn't go in. You just need to flip it over and plug it in. But the problem is persistent. I still catch myself doing this occasionally, even though I know it is a problem.
From my experience, I believe that this isn't entirely true. The situation is actually a lot worse. When the connection doesn't go in, the user doesn't just flip it over and solve the problem. The mathematical chance of plugging in a USB correctly the first try is 50%, because there are two ways the user may try to plug it in. However, the real-world chances of plugging it in correctly the first try are actually less than half because of real world conditions.
When plugging anything in, it usually does not go in perfectly the first try. This could be because the user tried plugging it in slightly incorrectly. It is uncommon to position and angle the cable in line with the port on the very first attempt.
This is even more difficult in situations where the connection is not entirely visible. USB connections are often placed on the side or back of a computer, not where the user is typically interfacing with the computer. The connection is also likely to be in a place that is poorly lit, such as a connection on the back of a computer. A user will have a more difficult time plugging it in if the correct plugging path is not the same as the user's vantage point. The connection may also be located between several other cables, obstructing the port.
People familiar with USB have incoming knowledge and experience that the USB connection is not reversible. As a result, after a quick first attempt, a user is very likely to give up and try the reverse. If you were to watch someone plugging in a USB, you'll see that they don't make just one or two attempts. Because of human factors and the conditions under which USB connections are put, users may make several attempts, constantly flipping the connection around to make it work.
No actual testing was done regarding this issue. These are just my observations. If there have been actual studies performed, I'd love to hear about them. Send them my way: @mdznr