Random Sample Picker

How to use this random sample picker

Random sampling without replacement means picking N distinct items from a larger list — each item can be picked at most once. This is the right tool for giveaway winners, team draws, randomized interview question selection, and quick audits where you need to grab a fair subset of a population. Under the hood it uses a partial Fisher-Yates shuffle seeded by crypto.getRandomValues, so every possible subset has equal probability.

1. Paste your list, one item per line. Leading and trailing whitespace is trimmed; empty lines are skipped. Duplicates in the input are preserved — if two people have the same name, enter them twice so both have a chance.
2. Set how many to pick. The sample size cannot exceed the list length. For a giveaway with one winner, set it to 1; for a top-3 draw, set it to 3.
3. Draw. Click Pick Winners. The tool shows the selected items in the order they were drawn and a copy-all button.
4. Document the draw. For real giveaways, screenshot the input list and the result before you announce winners. The draw algorithm is public and the tool is open source in spirit — anyone can verify the process.

Why this instead of shuffling the whole list? For small lists it does not matter — shuffling all of it and taking the top N gives the same distribution. But on long lists (5,000+ entries) the partial Fisher-Yates only touches N positions, which is faster and avoids returning a fully shuffled copy you did not ask for.

On fairness. Math.random is biased in ways that do not matter for shuffling playing cards but can matter for a regulated contest. This tool uses cryptographic randomness specifically so the process is defensible if someone asks how you drew winners. For US state-regulated sweepstakes, pair with documentation of the draw timestamp and method.