I saw an article about an upcoming Rock Paper Scissors tournament in my area. I’ve heard of such tournaments before and have always regarded them as tongue-in-cheek, even though they’re always described in very serious terms. But this one is being held at the local science center, no less, and offers a $1000 prize (or maybe it’s $1000 in prizes).
So, are these “tournaments” any more than a pseudo-serious random process of elimination? Can there really be a winning strategy?
Well, it’s one of those games which look random from outside but which have a high element of psychology.
I can outguess 21 of my 22 cousins. #22 and me get each other about 50:50
At the World Series of Poker last year Phil Gordon organized a R-P-S tournament to raise funds for cancer charities. The players spouted off about their winning strategies but they weren’t being entirely serious. Annie Duke made it to the semis or finals using the serial numbers off random bills to determine her throw, so that gives you an idea of how effective strategies are.
Last year Aussie radio station Nova 106.9 had the $6,000 Rock Paper Scissors Competition in various cities. I remember hearing the exciting live call of the final.
There was a pretty big one in Vegas organized by rec.gambling.poker at their annual convention a few years ago. They may still do it. The rule was that you weren’t allowed to use randomization methods.
How could that possibly be enforced? There are a lot of pseudo-random things you could use entirely within your mind. For example, take several numbers that are easily accessible, say your family members’ birthdays. A number 1-3 gets rock, 4-6 is scissors, 7-9 is paper. 0 is skipped.
There is a distinct tendancy for naive players to throw the hand that beat them last. If you throw the hand that beats the hand that just beat the other guy, you’ve got a better than 50-50 chance of winning.
Also, rock gets chosen a lot. Against a naive player choose paper more often than 1/3 of the time.
Of course, against a sophisticated player who knows these and other angles, you should do the exact opposite. Except he’ll know you’ll do the exact opposite, which means you should do the same. Except he knows you’ll do the same, so you should do the opposite. Which means that a random series of throws is best, unless you notice a bias in his selection. But a sophisticated person will eliminate this bias, so random is best against a sophisiticated person.
It was a big group of friends so it was the honor system. Mainly, because there were a lot of geeks in this group, they didn’t want people to write code for hand held devices to do their choosing for them.
Ah, yes, the wonderful world of Roshambo. [ulr=“Rock paper scissors (disambiguation) - Wikipedia”]Wikipedia has more than you ever wanted to know on this subject, including links to strategy guides and the such.
I remember there being online Roshambo links against strong computer players, but I can’t find them anymore. I found remnants of a programming competition, but that would require downloading and compiling code. There must be something better out there.
Got it. That would be pretty weak.
‘…rock beats paper.’
“I thought paper covers rock?”
‘Nah, rock flies right through paper!’
“Well what beats rock?”
‘Nothin’ beats rock’
-Kramer and Mickey playing Rock Paper Scissors.
Scissors beats everything, doesn’t it?
Not stone, Sir!
They’re very good scissors…
Unless, of course, you have spent the last five years building up a resistance to iocane.
One of my favorite wineries up here, Roshambo (natch), hosts an annual tournament, as well as other fun events (like a Pirate Prom!)
Every year I say I’m going to go check it out but we always end up making other plans. Maybe next year… It looks really fun.
Roshambo is very serious, and brings out the uber-geek in even the most geeky of computer scientists, for example. You can play an online Roshambot here (warning, it may humble you)
Also, for more analysis than you could possibly imagine going into the design and execution of roshambo computer programs, look at the write-up of the “The Second International RoShamBo Programming Competition”
If you get really interested, you can even download source code for some of the competitive 'bots.
Of course, the first versions of Iocane Powder would be easy to beat, given the source code. You’d just run your own copy of the program inside your own program, see what Iocane Powder would throw in a given situation, and always throw to beat that. Later versions, though, threw in a randomizer, which would make that almost impossible. At about the same time that your program figured out that it was playing against Iocane, Iocane would figure out what you were doing and switch over to a mostly uniform random strategy, and it’s well known that you can’t consistently beat a uniform random strategy (though you can’t consistently lose to it, either). So the best you could hope for would be to maybe get a couple of points ahead of Iocane, even knowing everything about the program.
As an example of why you would want to use any strategy other than uniform randomness, consider Bart Simpson. Bart’s strategy at paper, rock, scissors is “good ol’ rock”. Now, consider a round robin tournament between a computer playing a uniform random strategy, Bart Simpson, and a smart player (whether human or computer). The random player will fare evenly against both the smart player and Bart (since it’ll throw scissors at him about as often as it throws paper). But the smart player will very quickly figure out Bart’s pattern, and once that happens, will always throw paper (and therefore very soundly defeat Bart). So the random computer wins a third of its throws, ties a third, and loses a third; Bart wins about 1/6 (the times he’s playing vs. the computer and it happens to choose scissors), ties about 1/6, and loses 2/3 (1/3 of the time vs. random, and all the time vs. smart player), and the smart player wins 2/3, ties 1/6, and loses 1/6. So even though the random computer is unbeatable by any given opponent, the smart player still wins the tournament.