Get a random rock, paper or scissors throw β useful when you need an impartial opponent, or a starting hand that cannot be read from a habit. Each of the three is equally likely on every throw, with no memory of what came before.
Settings
Recent results
Generated locally in your browser β your settings and results never leave this page.
A game people cannot play randomly
Rock paper scissors is a solved game in one specific sense: throwing each gesture one third of the time, independently, cannot be beaten in the long run. No strategy wins against it, because there is no pattern to exploit.
The difficulty is that humans cannot do it. Asked to produce a random sequence, people avoid repeating themselves more than chance would, alternate too regularly, and let the previous round influence the next in consistent, readable ways. These tendencies are reliable enough that competitive players β and the game has a genuine competitive scene β build strategies around them.
The patterns that give players away
The best-documented tendency is win-stay, lose-shift: players who won a round tend to repeat their gesture, and players who lost tend to change it. Knowing only that, an opponent has a meaningful edge over someone playing on instinct.
There is also a well-known opening bias β rock is thrown first more often than a third of the time, particularly by inexperienced players and particularly by men β and a general reluctance to throw the same gesture three times running, even though that sequence is exactly as likely as any other.
None of this applies to a generator. Each throw is an independent draw, so there is no previous round to react to and no pattern to read.
Unbeatable is not the same as winning
Playing randomly guarantees a draw in expectation against any opponent, however skilled. That is a strong property and a limited one: it means you cannot lose in the long run, not that you will win.
Against an opponent with readable habits, exploiting them beats playing randomly. The trade-off is that any exploitative strategy is itself a pattern, and a good opponent will eventually read it. Random is the safe floor you fall back to when you cannot read the other player, or when you suspect they can read you.
This is the same reasoning that appears throughout game theory: the unexploitable strategy is rarely the highest-scoring one, but it is the one that cannot be punished.
Lizard Spock and the tie problem
The five-gesture variant adds lizard and Spock, arranged so each gesture beats two others and loses to two. The practical effect is fewer ties: with three options, two random players tie a third of the time, which is a lot of replayed rounds. With five, ties drop to one in five.
The optimal strategy is unchanged in structure β throw each gesture equally often β but there are more ways to be predictable, which slightly favours whoever is better at avoiding patterns.
Streaks are supposed to happen
Three identical throws in a row occurs about once every twenty-seven sequences. It looks wrong and it is correct. A generator that suppressed repeats would be introducing exactly the bias that makes human play exploitable, and would be weaker for it.
The same reasoning applies to any run that looks meaningful. Independent draws produce clusters, and reading intent into them is the mistake the tool exists to avoid.
Privacy
Throws are generated in your browser. Nothing is transmitted, stored or logged.
Settling things between more than two people
The classic game handles two players cleanly and gets messy beyond that. With three or more, a single round frequently produces no winner at all β if all three gestures appear, everything beats something and nothing decides.
The usual conventions are to keep throwing until only two distinct gestures appear, or to play a knockout where players pair off. Drawing several throws at once here supports either: everyone gets an independent throw, and the group applies whichever rule it has agreed.
For deciding an order rather than a winner, a shuffle is a better fit than repeated throws, and the turn order generator produces one in a single step.
How to use the Rock Paper Scissors Generator
- Choose how many throws you need β one for an impartial opponent, more for several players.
- Turn on Lizard Spock for the five-gesture variant.
- Throw, and throw again for the next round.
Frequently asked questions
Why would I need a generator for this?
Two reasons. You may not have an opponent β you need an impartial third throw to settle something. Or you have one and want a throw that cannot be read, because human choices in this game are famously predictable.
Are the three options equally likely?
Yes. Each throw is an independent draw with the same chance for each gesture, and no memory of previous throws. That is what makes it unexploitable.
Is random really the best strategy?
Against a strong opponent, yes β it is the game-theoretic optimum, and no counter-strategy beats it in the long run. Against a predictable human, deliberately exploiting their pattern does better. Random is the strategy that cannot be beaten, not the one that wins most.
Why are humans bad at playing randomly?
People avoid repeating the same gesture, over-alternate, and are influenced by the previous round in consistent ways β losers tend to switch and winners tend to repeat. These tendencies are strong enough that competitive players exploit them deliberately.
What is Lizard Spock?
A five-gesture variant that adds lizard and Spock, giving each gesture two it beats and two it loses to. Its main practical benefit is that ties are less frequent, since there are five options rather than three.
Can I throw for several players at once?
Yes. Raise the player count and each gets an independent throw, which is a quick way to settle something between three or more people without a round-robin.
Does the generator remember previous throws?
No. Every throw is independent, so three rocks in a row is possible and happens about one time in twenty-seven. A generator that avoided repeats would be predictable in exactly the way the tool exists to prevent.
