Need to draw a giveaway winner, call on a student, or settle who takes out the bins? This free online random picker for Australia chooses fairly from any list you paste. Put your entries into the Items box — one per line, or separated by commas — set how many you want with Count, and press Generate. A list needs at least two entries and holds up to 1000, each up to 200 characters long.
Every pick is made directly in your browser using the Web Crypto API (crypto.getRandomValues), which provides cryptographically strong random values from your browser's underlying system randomness. Your list is never sent to our server, so we never see or log it — and that matters more here than on most generators, because what people paste is usually a customer list, a class register, or a set of colleagues' names. Load from file keeps the same promise: your browser reads the file and fills the box locally, with nothing uploaded anywhere.
The options are made for the lists people actually have. Separator sits on Automatic, which reads line breaks and commas and works out which of the two your list uses, so a list copied from a spreadsheet, a chat message or a document works without reformatting. Lock it to Line break, Comma or Space when your data mixes formats — and note that Space is always a deliberate choice and never automatic, because entries such as full names contain spaces of their own and splitting on them by default would break more lists than it fixed. Count decides how many entries to draw, from 1 to 100. Drawing three prize winners out of 200 entrants? Paste the entrants, set Count to 3, and press Generate. Two controls only apply once Count is 2 or more and stay greyed out until then: No duplicates, which decides whether the same entry can be drawn twice, and Output format, which sets how the picks are separated — Each on a new line, Space-separated, or Comma-separated.
A row of buttons inside the Items box does the editing you would otherwise do somewhere else. Load from file pulls a list straight out of a plain-text or CSV file. Shuffle reorders the entries in the box, Sort arranges them alphabetically with numbers ordered naturally and a second click reverses it, and Clear empties the box. Fill with sample data drops in a short example list so you can try the picker in one click, while Expand to fit content unfolds the box to show a long list in full and Collapse to default size folds it back. Under the result, Copy All puts every pick on your clipboard, Save to File writes them out, and History keeps your last 10 draws in your browser, so an earlier result is one click away.
How is the pick actually made? When you press Generate, the page asks the browser for random values through crypto.getRandomValues and uses them to choose positions in your list, rather than going through a JavaScript pseudo-random function such as Math.random(). Turning a raw 32-bit value into a position is where pickers quietly go wrong: taking the value modulo the length of the list is the obvious approach and the biased one, because unless the length divides evenly into 2^32, entries near the start of the list come up slightly more often than the rest. That flaw is called modulo bias, and it is measurable rather than theoretical. This picker instead discards any value beyond the largest exact multiple of the list length and draws again — a technique called rejection sampling. It costs a few extra draws and buys an exact guarantee instead of an approximate one. Position in the list changes nothing: the first entry and the last carry the same probability on every draw.
No duplicates separates two genuinely different kinds of draw, and statistics has a name for each. With it on, the picker shuffles a copy of your list with the Fisher-Yates algorithm and takes the entries off the top — the same thing as drawing tickets from a hat without putting them back, so no entry is picked twice and every group of that size is equally likely. That is sampling without replacement, and it is what you want for prize winners, teams and turn order. With it off, each pick is made independently from the whole list, so the same entry can come up more than once — sampling with replacement, which is what you want when everybody stays eligible in every round, as in repeated draws or a quick statistical sample. Neither mode is fairer than the other; they answer different questions.
A fair draw is not supposed to look fair. Draw five names out of ten and you will regularly see two picks that sat next to each other in the original list, or the same name twice in a row once duplicates are allowed, and every one of those is ordinary. With No duplicates off, each draw is independent: earlier picks have no influence whatsoever on later ones, and nobody ever becomes "due" for a win. The instinct that insists otherwise is the gambler's fallacy, and it is the same instinct that makes us poor at picking randomly ourselves — asked to choose at random, we avoid the first and the last entry, avoid repeating a name, and spread our choices far more evenly than chance ever does. A result that looks suspiciously tidy is more likely to have been made up by a person than drawn by chance; random draws produce clumps.
Why choose this random picker?
- Browser-only picking: your list is read on your device and never sent to our server
- Web Crypto API randomness: cryptographically secure, not Math.random()
- No modulo bias: rejection sampling keeps every entry equally likely
- Any list shape: one per line or by comma automatically, by space on request
- No duplicates mode: draw meat tray raffle winners or club teams with no entry picked twice
- Up to 1000 entries, and up to 100 picks in a single draw
- List tools built in: load from a file, shuffle, sort, clear, sample data
- History and Save to File: keep a copy of every draw
- Unlimited and free: no sign-up, no usage limits, nothing on our servers
In Australia, random picking turns up wherever a choice has to survive being questioned. Organisers draw winners for a surf club meat tray raffle or school fete in front of an audience that has to believe the result. Teachers call on students without playing favourites and split a class into project groups. Office teams settle who is on call, who takes the minutes, or who is drawing whom in the Kris Kringle at Christmas. Researchers pull a random sample of records to audit. Mates might even use it to settle ties on footy tipping ladders. The common thread is a selection anyone can explain afterwards: the draw uses cryptographic randomness, it happens entirely on your device, and History keeps the result for later — though for a public giveaway, save the list of entrants and the result before you announce it, because a browser history is a convenience, not proof.
This picker has a few honest limits worth knowing. It cannot draw more distinct entries than your list holds — three winners need at least three names — and it stops at 1000 entries of 200 characters each, with up to 100 picks in one draw. Shuffle and Sort work line by line, so a comma-separated list pasted on a single line is reordered only once it is one entry per line; and both of them rearrange the box rather than the draw, which is equally fair whatever order your entries are in. Every entry also carries exactly the same weight by design, so a draw where one name should count double needs that name listed twice. Inside those limits the guarantee is a strong one — at every pick, every entry still in the draw has exactly the same chance.