Legal Research

The Friday Fillip

By Simon Fodden

Deciding disputes is a big thing for human beings — as readers of this blog are bound to know. Sometimes we opt for a facsimile of reason, other times we court fate with chance, as when we toss a coin or throw a die. (We’ll leave aside the urge to cause harm in order to decide, with such lovelies as ducking, trial by battle, and peine forte et dure.) A variation on the theme of chance is the game of rock, paper, scissors. I say variation, because although the game is apparently unskilled, it involves two people in harmless contest and for that reason leads participants to believe that they have a role in it, when in fact they are nothing more than coins being tossed.

Or are they? Is there some “skill” that will allow a player to “psych out” an opponent or guess an opponent’s move?

Not, I’d say, with a single round of rock, paper, scissors (RPS). But it seems that when we play iterated RPS — that is, repeated instances of the game with the same players — we do so with a detectible style or tendency. I can’t say I’ve ever been able to suss out what my opponents are planning in RPS, or, indeed, that I’ve ever played iterated RPS where I’ve kept score. But evidently a computer program can do it.

The New York Times Science section has a web app that lets you pit your skills against such a program. You can choose to play against a “novice” computer, one that knows nothing about how human beings are likely to choose and that must learn from playing you, or against a “smart” computer, which is one aided by “over 200,000 rounds of previous experience” with human opponents. And the thing keeps score.

As the intro material points out, if you were to make “truly random” choices, you’d eventually wind up in a tie with the machine. But, it seems, it ain’t so easy to be random. And you’ll almost certainly betray a pattern of choices that the program will recognize and exploit.

Doubt me? Go and give it a try. It’s easy enough to talk the talk, but when it comes to walking the random walk…

Full post as published by Slaw.ca on March 18, 2011 (boomark / email).