Draw a random-sized sample from a list, not just a fixed number

Most "pick from a list" tools ask you the same question first: how many do you want? One winner, three winners, five names off the roster. But there's a whole class of decisions where the count itself is the thing you'd rather not own. How many appetizers should we order for the table. How many warm-up questions before the real interview. How many test cases to spot-check before you trust the batch. You want "some, but not all, and don't make me argue about the exact number" — and that's a different draw than picking a fixed quantity.

Two random choices in one click

The random subset picker does something most pickers don't: it randomizes the size of the sample and then the members. You give it a list — one item per line — and a range, a minimum and a maximum. It rolls a number somewhere in that range, then deals out that many items without repeats and shows you the result, with a quiet "Picked 4 of 15" line so you always know how big the draw turned out to be. Set the minimum to 3 and the maximum to 7, and any given click might hand you three items or seven or anything between, drawn cleanly from whatever you pasted in.

That's the part worth slowing down on, because it's the whole point. A normal picker collapses two questions — "how many" and "which ones" — into one by making you answer the first yourself. The subset picker answers both at random, which means the size of the sample stops being a decision you can be accused of fudging. You didn't choose to grab exactly five; the tool did, within bounds you set in advance.

Set the range, and what the range means

The two numbers you control are Min size and Max size. The picker treats every whole number from the minimum to the maximum as equally likely, then draws that many distinct items. A couple of guardrails are baked in so the tool never does something nonsensical: the minimum can't drop below 1, the maximum is never allowed to fall below the minimum, and if your range asks for more items than the list actually contains, the draw is capped at the length of the list. Ask for "up to 7" from a list of four, and you'll get at most four — never a phantom fifth.

Set the minimum and maximum to the same value and the tool behaves like a plain fixed-count picker: "always exactly four." Spread them apart and you reintroduce the surprise. The wider the gap, the more the sample size itself becomes part of the randomness — which is great for variety and bad for anyone who secretly needs a predictable count, so pick your range with the real constraint in mind.

Why "without replacement" matters here

Every item the subset picker returns is distinct. It's drawing without replacement — like dealing cards off the top of a shuffled deck rather than spinning a wheel that can land on the same name twice. For a sample, that's almost always what you want: a spot-check of "these four records," a tasting flight of "these three dishes," a reading list of "these five chapters" where naming the same chapter twice would be useless. Each line on your list either makes the cut for this draw or it doesn't, and the ones that make it are genuinely different from one another.

It's worth being honest about what that doesn't give you: perfectly even long-run coverage. Across many draws, the items aren't guaranteed to appear the same number of times — randomness clumps, and one item can sit out three samples in a row by pure chance. If your goal is "make sure everything eventually gets sampled," you'd run the draw repeatedly and watch the results, or reach for a tool built to walk a whole list without repeats. The subset picker's job is one honest, unbiased grab, not a guarantee of fairness over time.

Where a random sample beats a fixed count

The sweet spot is any situation where committing to an exact number feels arbitrary or political. Picking which features to demo in a review, when you genuinely could show three or six. Choosing how many raffle runners-up to announce. Building a "surprise me" playlist of somewhere between a handful and a dozen tracks. Deciding how many people from a long volunteer list to tap this week. In all of these, the count is a knob nobody wants to be holding, and handing it to a bounded random draw makes the result feel like luck instead of favoritism.

When a different picker fits better

If you already know the count and just want that many winners, the multiple random picker is the simpler tool — you set the number, it draws exactly that many. If your real task is to split a list into balanced groups rather than pull one sample out of it, the random team / group divider carves everyone into teams in one pass. And when the answer is a single item — one winner, one next-up — the plain random single picker is all you need.

A small tool for a sneaky-hard decision

"How many" sounds like a trivial question until you're the one who has to defend the answer. The subset picker takes both the count and the contents off your plate at once: set a sensible floor and ceiling, click once, and accept the sample it deals. Paste your list, pick a range you can live with, and let the draw decide how big "a few" turns out to be today.