Way my dad and I used to play is that when somebody won they got to grab the hand of the loser and then, with the fore- and middle-fingers only, slap the loser on the wrist. My dad would lick his fingers (“to cut down on air resistance”) before whacking me.
Say, these days that would be child abuse. Well, the government knows best, right?
Briefly, you and simultaneously your opponent select rock, paper, or scissors. Rock breaks scissors, paper covers rock, and scissors cut paper. There are 9 outcomes: 3 of which are rock for you, and rock, paper, or scissors for your enemy; et cetera. Three of the 9 outcomes are ties, in which nobody gets slapped; there are 3 ways for you to lose, and 3 ways for you to win.
Now, given just those premises, and none other, what is the probability you don’t get slapped? Well, “don’t get slapped” means tying or winning; and since there are 6 ways for these things to happen, the chance is 2/3. Similarly, there is a 1/3 chance for you to lose, and 1/3 probability you get to slap.
These deduced probabilities are correct assuming only the premises describing the rules of the game. Which implies, and it is true, that the probabilities are not likely to be correct assuming other premises. What other premises? There are an infinite number of premises we might choose, but what we’re after are premises that help us win the game.
The thing to emphasize is that we know with certainty the non-human premises, and so know with certainty the non-human probabilities. Rock paper scissors is thus similar to poker where we have a good handle on the probabilities; but where in poker they are harder to memorize, yet in poker we know there are consistent and good players.
There are also consistently winning Rock Paper Scissors players. Like 2008 champion Sean “Wicked Fingers” Sears. That means, like poker, human premises must exist that change the odds.
In any number of places you’ll read that the way not to get beat is to make your pick “randomly.” This is impossible. No matter what, you must cause your pick, and no cause in the universe is ontologically “random.” Suppose you decide to divide up a minute into thirds and pick based on the secondhand of your watch. If your opponent does not know this, he has no way of guessing what you’ll do (except that you must choose, of course), and so to him your guess is “random”—which means only unknown. But to you, the choice is determined, caused by your decision and the state of the time.
If your opponent catches you sneaking peeks before each round, then he’ll too know what you’re going to do—creating a new and probative and deductive premise for him—and you’ll consistently lose.
That’s the way to win, too. By searching for patterns, i.e. premises, which your opponent is using, knowingly or not. Bin Xu and pals think they’ve discovered patterns many people use. In their arxiv paper “Cycle frequency in standard Rock-Paper-Scissors games: Evidence from experimental economics” they posit that winners often repeat their winning pick, and losers select the next object in the cycle (lose with scissors and move to rock). Armed (handed?) with these premises allows you to change the odds.
Until your opponent figures out his mistake; and when he does, he can use that knowledge against you. Figuring you are figuring on a rock (since your opponent just lost on a scissors), and thus you’d pick paper, your opponent upends the algorithm and sticks with scissors.
Noted RPS expert Sharisa Bufford, a member of the prestigious USA Rock Paper Scissors League, is sure that “Girls always throw scissors first. Guys always throw rock.” If that’s so, these are winning premises. Unless your opponent knows these rules, too. And you know they know. And you know they know you know…you know?
Now it’s true you cannot “pick randomly” but must always cause a choice. But that does not mean it’s impossible to discover a strategy which your opponent cannot guess—where your opponent may be some “sophisticated computer” (computers are only dumb unthinking distillations of fractions of human brains). All it takes to create impossible-to-discover picks are to create picks which your opponent cannot guess beyond the standard premises (a tautology, really).
Maybe that’s memorizing a stream of numbers generated by some process which remains unknown to your opponent. Or maybe that means changing your picking algorithm based on circumstance. All that matters is that your opponent cannot guess.
Incidentally and curiously, computer programs which test “randomness” (i.e. unpredictability) always turn a blind eye to the algorithm which generated the sequence they’re testing.
Lastly, to prove (as I’ve done before) that I can pick a number you cannot guess with certainty, no matter what your resources, I’m thinking of a number between 0 and 4. What is it? (It’s hidden in the code to this page.)