Maybe we need to institute Apple's smart shuffle into the mix, so it creates random patterns that feel random.
I have been listening to the Radiolab and Hypercritical podcasts lately, so I heard about that just in the last few days even though that's something that happened years ago.
Human perceptions and misperceptions of randomness used to be one of my research interests in grad school, so I have to just share a couple little anecdotes on this subject.
Anecdote #1 - The RAND Corporation publishes a book called
A Million Random Digits. It is exactly what it sounds like: one million random digits, generated using a very thorough physical randomization process. I have a friend from grad school who now works for RAND, and he loves to joke that this is one of the more exciting books that RAND has ever produced.
Anecdote #2 - There is a common experimental procedure which researchers will use to test strategic thinking about random events. Imagine that there are two doors in front of you, Door A and Door B, and you are told that there is a reward behind one of the doors. You must choose one door to open -- if you pick the door with the reward behind it, you get to keep the prize. You are going to play this "game" 100 times.
Now, what you don't know is that the experimenter has rigged the game so that the prize appears behind Door A about 70 percent of the time, and Door B about 30 percent of the time. After enough iterations of the game, though, most players figure out that Door A is more likely to have the prize than Door B, and can even identify the ratio fairly closely.
Now, the strategic, "maximizing" behavior in this case is to always choose Door A, since that will maximize your chance of getting the prize. But people hate to do that, because it makes them feel like they are "giving up" on all the times when the prize is behind Door B. So people tend to settle into a pattern whereby they pick Door A about 70 percent of the time, and Door B about 30 percent of the time. They match their choices to the observed frequency of the prize drops, even though this is suboptimal behavior.
Now, here's the funny part. People have done this sort of experiment with a variety of different animals. Some animals behave very much like humans, whereas other kinds of animals appear to maximize.
I found a chart one time which had data points for various different animal species. You had your usual suspects -- mice, chimps, dogs, etc. The one that blew my mind was that there was a data point for crocodiles. Somewhere, some time, some poor grad student got tasked with running an experimental trial on *crocodiles.*
I always tried to keep that in mind when I felt like my professors were being unreasonable -- at least they weren't handing me a bucket of steaks and telling me to go run the crocodiles.
(In case the suspense is killing you, it turns out that crocodiles maximize. Good for them.)