The Magic 8-Ball has twenty answers and they are not evenly split: ten are affirmative, five are non-committal and five are negative. That two-to-one lean toward yes has barely changed since 1950, and it is the most revealing thing about the object — it was designed to be pleasant to consult rather than to be a fair coin.
That is worth knowing before treating any chance device as neutral, because the ones built for entertainment usually are not.
What is the skew for?
Repeat use. A device that says no half the time gets put down; one that leans encouraging gets asked again. The non-committal answers do similar work. "ask again later" defers rather than refuses, which keeps the interaction open.
Compare that with a coin, which is genuinely symmetric, or with a twenty-sided die, where each face is exactly five per cent. Those are instruments; the 8-ball is a toy with an opinion.
Rock, paper, scissors sits between the two. Play it against something that is genuinely random and the game evaporates: every throw wins, loses and draws a third of the time, and no strategy can do better, because there is nothing to exploit. All of the actual game lives in the opponent’s habits — which is why practising against a random generator is useful in the way a metronome is, as a baseline you can hear yourself drifting from.
Why does randomising a decision help at all?
Because your reaction to the answer is the information, not the answer. Randomising does not tell you what to do; it produces a moment in which your preference surfaces before you have had time to construct a justification for it.
The yes-or-no article works through that mechanism in detail. The short version is that relief and disappointment both arrive faster than reasoning, and both are data.
The other genuine use is ending a deliberation that has stopped producing new information. Choosing between forty films is a hard problem and choosing between two is easy, so the value of a picker is that it converts the first into the second, not that its choice is wise.
What does a draw without replacement change?
Whether the same result can come up twice, and it is the distinction that separates a shuffled deck from a single card draw.
| Draw | Replacement | Consequence |
|---|---|---|
| Shuffled deck, dealt | Without | A hand never repeats a card |
| Single card, each time | With | The same card can come up twice running |
| Lottery numbers | Without | Six distinct numbers |
| Dice, coins, letters | With | Every draw is independent |
A shuffled 52-card deck uses Fisher–Yates over the browser’s cryptographic generator, which makes all 52 factorial arrangements equally likely; a number so large that the shuffle is almost certainly unique in history.
A single-card draw reshuffles each time, so the same card can appear on consecutive draws. That is what "each card equally likely" means, and it is the thing people read as a fault.
Can a generator improve lottery odds?
No, and nothing can. In a six-from-49 draw the chance of matching all six is one in 13,983,816 regardless of how the numbers were chosen.
What a uniform generator does avoid is the clustering human choices produce. People pick birthdays, so numbers above 31 are under-chosen; people pick patterns on the ticket grid; people pick 7 far more than a nineteenth of the time. None of that changes the probability of winning, and it changes the expected share of a jackpot, because an unusual combination is less likely to be split.
That is the only honest claim available: a random pick does not win more often, it shares less often when it wins.
What are natural ones and twenties actually worth?
Five per cent each on a d20, which is far more often than the drama suggests. Across a three-hour session with sixty rolls, three critical successes and three critical failures is the expectation rather than a memorable night.
That frequency is why the twenty-sided die became the standard for the job. It is granular enough that a modifier of one matters and coarse enough that the extremes arrive often enough to be worth having rules for.
Questions people ask
Is the 8-ball distribution standard? The twenty answers and their split have been essentially unchanged since 1950. Ten positive, five neutral, five negative is the original design.
Why does a random time generator exist? Sampling, mostly; picking observation times, scheduling unannounced checks, or generating test data that is not clustered on the hour. There are 1,440 minutes in a day and each is equally likely.
Does rock, paper, scissors have a best opening? Against a random opponent, no. Against people, there are documented tendencies, and the moment you exploit one you become predictable yourself, which is why the game does not resolve.
Is a shuffled deck really unique? Almost certainly. 52 factorial is about 8 × 10⁶⁷, which is more orderings than there have been seconds since the Big Bang by a very wide margin.
Some of these are instruments and some are toys with an opinion, and it is worth knowing which you are holding. The Magic 8-Ball, decision maker and rock paper scissors cover the deciding ones, the dice roller, card shuffler and lottery number generator cover the drawing ones — a d20 is the dice roller with one setting changed, and a single playing card is the shuffler stopped after one — and the random time generator is the one that is genuinely a sampling tool.