The Hat DrawNames in. Winners out. ☕ Tip the hat

Fair drawing methods compared

Every fair drawing gives each entry the same chance and lets the audience believe it. Here is how the common methods measure up.

“Fair” in a drawing has two parts. The first is mathematical: each entry should have the same probability of being chosen, and earlier picks should not tilt later ones except by removing winners from the pool. The second is social: the people in the room should be able to see enough of the process to trust it. A method can be perfectly random and still feel shady if it happens out of sight, and a method can look festive while quietly favoring some entries. This guide compares the common options and explains exactly how The Hat Draw makes its picks.

Names in a hat

The classic. Write each name on a slip, fold, drop in a hat, stir, pick. It is cheap, instantly understood and fun to watch.

Where it goes wrong: slips are not identical. A slip folded into a tight square feels different from a loosely folded one, and a person who wants to can feel for it. Slips added last sit on top unless the hat is mixed well. Large or heavy slips settle differently. And the bigger the group, the harder it is to mix a hat thoroughly.

Making it fairer: cut slips from the same sheet to the same size, fold each the same way (once in half, then again), mix with a hand for a good ten seconds, and have the drawer look away or close their eyes while choosing. Count the slips before you start and announce the count.

Numbered tickets in a drum

A rotating raffle drum with double tickets is the standard for paid raffles. The buyer keeps one half; the other goes in the drum. Because the halves are machine-cut and identical, it avoids the folding problem.

Where it goes wrong: drums that are overfilled do not tumble well, and tickets dropped in at the end stay near the door. Static can make tickets cling together; occasionally two come out at once.

Making it fairer: fill the drum no more than about two-thirds, spin it several full rotations in both directions, and have the drawer reach in deep rather than grabbing from the top. If two tickets come out stuck together, return both and draw again; announce that rule in advance.

Spinning prize wheels

Wheels are great for games where the outcome is a category (a prize tier, a chore, a question topic). They are less good for picking one person from a long list, because you can only fit so many labeled wedges, and a real wheel has physical biases: a slightly heavy side, a sticky pointer, or a spinner who learns how hard to push. On-screen wheels avoid physical bias but still get crowded with more than about 20 to 30 names.

Use a wheel when: you have a handful of options and the spin is part of the entertainment. For a guest list or a stack of ticket numbers, a picker that shuffles the whole list is simpler to read and verify.

Pick a number

Some groups assign everyone a number and roll dice or ask a calculator for a random number. This is fair if the range matches the count exactly. Problems arise with uneven mappings: if there are 37 entries and you roll two six-sided dice, some totals are far more likely than others, and many numbers can never come up. Be careful with any method that converts a random number into a list position unless it handles the range evenly.

Digital random pickers

A good random name picker solves the physical problems: every entry is identical, nothing sticks, nothing sits on top. The questions to ask are where the randomness comes from and what the tool does with your list.

How The Hat Draw picks winners

The Hat Draw runs entirely in your browser. When you press Draw, it does the following:

  1. Splits your text into entries on new lines, commas or semicolons, trims extra spaces and drops blank lines.
  2. If “Remove duplicates” is on, keeps only the first copy of each name, ignoring capitalization. If it is off, a name that appears three times has three chances, which is what you want when a person bought three tickets.
  3. Shuffles the whole list with the Fisher–Yates algorithm. For each position from the end of the list to the start, it picks a random position at or below it and swaps the two. Fisher–Yates is the standard way to produce a shuffle in which every possible ordering is equally likely.
  4. Each random number comes from crypto.getRandomValues, the browser’s cryptographically secure random generator. The code also rejects values that would create a tiny bias toward lower positions (a technique called rejection sampling), so every position is exactly equally likely.
  5. Takes the requested number of winners from the front of the shuffled list. Those names leave the hat, so the same entry cannot win twice in one draw unless it was entered twice.

Drawing again with “Draw Next Name” reshuffles the remaining names each time. Nothing about the list is sent to a server. The full technical walkthrough is on How it works.

Making any method look fair

The audience cannot audit an algorithm or the inside of a drum, so the visible process matters as much as the math:

Which method should you choose?

For a small group in one room, a well-mixed hat of identical slips is charming and fair enough. For a paid raffle with hundreds of tickets, numbered double tickets in a drum are traditional and easy to audit against stubs. For lists you already have typed (sign-up sheets, staff rosters, class lists, online registrations), a browser-based picker with a cryptographic shuffle is the fastest and most even option, and it can project the moment on a screen. Many groups combine methods: sell paper tickets, then type the stub numbers into a picker so the drawing can be shown to the whole room.

More: all guides · how the draw works

Open The Hat Draw →

Last updated 10 October 2026.