Games of Chance: Lottery Picks, Bingo Cards, and Coin Flips

Lottery quick-picks, bingo cards, and coin flips all get called "random", but they are three different random objects. A lottery draw is a hand of cards dealt from one deck, a bingo card is a structure filled by rules, and a coin flip is a single independent binary event. Knowing which is which tells you what "fair" means — and which surprises are actually normal.

Lottery picks: deal from a deck, then sort

A "6 from 49" draw is sampling without replacement: each ball leaves the machine, so the same number cannot appear twice. The clean machine equivalent is a partial Fisher–Yates shuffle — walk the array once, swapping each position with a random later one — then take the last six numbers and sort them:

pool = 1..49, partially shuffled; take the final 6 and sort:
1, 9, 13, 17, 26, 35

The sort is presentation, not chance: a winning ticket does not care in what order the balls emerged, so the numbers are shown ascending. What the fairness guarantee does not change is the odds. The number of possible tickets is C(49,6) = 13,983,816 — a combination count, because order is irrelevant (the full arithmetic is in permutations vs combinations). Matching all six is 1 in 13,983,816, and a fair generator makes every one of those tickets equally likely — which also means the "unlucky-looking" ticket is exactly as likely as the tidy one.

Bingo cards: structure first, randomness second

A US 75-ball card is not "25 random numbers from 1–75". Each of the five columns owns a fixed range under its letter: B draws 1–15, I 16–30, N 31–45, G 46–60, O 61–75 — five numbers per column, sorted top to bottom. The center square is the FREE space, so the N column only draws four numbers. The ranges are what make the game playable: they spread calls evenly across the card, so every called number has somewhere it can go.

The UK 90-ball ticket has its own constraints: three rows by nine columns, fifteen numbers, exactly five per row, and each column drawing from its own decade — 1–9, then 10–19, 20–29, … up to 80–90. Blanks fill the rest, arranged so no column is entirely empty and every row holds exactly five. Generating a legal card is a small constraint problem: the blank positions are chosen from rotating four-column windows (so no column is blanked too often), then each column deals from its decade range. Compare this with the lottery: there, the only rule is uniqueness; here, the rules are the game, and a uniformly random 5×5 grid of 1–75 would be a different, broken one.

Coin flips: independence is the whole story

A fair coin has P(heads) = 0.5 on every flip, regardless of history. The coin has no memory; after five heads, the sixth is still 50/50 — believing otherwise is the gambler's fallacy. This site's simulator draws its randomness from the browser's crypto.getRandomValues(), the cryptographically strong source rather than the everyday pseudorandom stream, and marks a head whenever a uniform draw in [0, 1) lands below 0.5:

rand() < 0.5  →  heads      (1 to 10,000 flips per run)

Large numbers behave in a specific, slightly counterintuitive way. The percentage of heads converges to 50% — after 10,000 flips you will typically land within about one percentage point — but the raw gap between heads and tails tends to grow, roughly like √n. Being at 60 heads versus 40 tails after 100 flips is completely ordinary. Streaks follow the same logic: the longest run of identical results across n flips grows near log₂(n), so in 10,000 flips a run of about 12–13 heads in a row is expected, not suspicious. The simulator reports longest-run statistics precisely so you can watch "weird" streaks show up on schedule.

Seeds: when a draw needs a receipt

The lottery and bingo generators accept an optional seed, which makes the draw reproducible: the same seed yields the same numbers, so a raffle result can be re-verified by anyone holding the seed. With no seed, the output is non-deterministic as usual. The mechanism — a small Park–Miller generator standing in for true dice — is explained in how pseudorandom numbers work; for picking lottery numbers and shuffling bingo columns, uniformity and independence are what matter, and both hold.

Try it yourself

The Lottery Number Generator deals unique numbers from any range and count — the same partial-shuffle method for 6/49 or your office raffle — and sorts them for the ticket. The Bingo Card Generator prints US 75-ball cards with a FREE center or UK 90-ball tickets, up to a dozen at a time, each one a fresh constraint-solving draw. The Coin Flip Simulator runs 1 to 10,000 flips and tallies heads, tails, and the longest streaks, so you can watch the log₂ streak law happen.