I clearly still haven't figured out to write about here. But let me stop myself from veering into meta-blogging, at least for now.

Have you ever thought about Rock, Paper, Scissors strategies? It's pretty obvious and boring if we assume players have some way of choosing randomly. But there's a pretty strong psychological aspect when playing in real life. The notion that writing a program to play wouldn't necessarily be a boring exercise of using whatever random number generator you have occurred to me when I ran into a program at Mathcamp that would fairly consistently pull ahead of a human opponent. Rather than spend too much time trying to tinker with some sort of not-great solution, the aspect of this that I'm thinking about at the moment is testing RPS bots. Which is to say, coming up with programs that play poorly, but in interesting ways. So how do you make something be dumb like a human?