Draw Lots

Set up the slips, shuffle once, then open one at a time

This draw lots tool puts the old paper method on screen: write results on slips, fold them, pass the bag around. Choose items and slip counts, such as 2 winners and 8 blanks, and the tool shuffles the whole set once and fixes the result for every slip number. Each time you press a folded slip, that slip's result unfolds. An opened slip is not put back, so after opening 2 slips out of 10, 8 remain.

Every time a slip opens, the slips-left bar and each item's next-draw chance update right away. If you list the people drawing, the order you type becomes the turn order, and the log records who opened which slip and what it said. The chart below shows how the chance moves from turn to turn, and that the chance without knowing earlier results is the same at every turn.

Why draw lots on a screen instead of on paper

Nobody can choose a result in advance, and everyone draws from the same bag. That is why people still draw lots for fair-feeling decisions. Moving the draw on screen keeps that, and removes some of the fuss and suspicion of the paper version.

• The counts are exact: for a door prize at an office party, you do not have to cut, label and count thirty scraps to get 3 winners and 27 blanks. The total on screen tells you at once whether they all went in.
• Slips cannot be told apart by sight: hand-folded paper differs a little in size, creases or how much the writing shows through, and a sharp-eyed player can notice. On screen, every slip looks the same apart from its number.
• Everyone watches the same board: put it on a classroom display or a video call and open slips one by one, so everyone sees each result at the same moment. In full screen, only the slips and the progress panel stay visible.
• You keep a record of who drew what: with names listed, each turn and result is added as one line. When you are done, Copy log gives you text you can paste straight into a group chat or a notice.

Typical uses: picking who gets the door prizes, handing out 1st, 2nd and 3rd prizes, assigning classroom chores or group roles, setting a speaking order, and settling who takes the next shift or draws the short straw. The four example buttons fill in the setup for these cases.

How to use it

1. Set the slips: enter item names and slip counts. Two lines such as Winner and Blank make a prize draw, several lines such as 1st, 2nd, 3rd and Blank make tiered prizes, and one slip per item makes a role assignment. If you paste a list separated by line breaks or commas into a name field, each line becomes one item.
2. List the people drawing (optional): type one name per line, and turns follow the list from the top. If the number of people and slips differs, the tool shows the difference and the chance a prize slip stays unopened, with a button that adjusts the blank count.
3. Open the slips: press a folded slip on the board and it unfolds to show the result. To check the number first and reveal later, turn on 'Pick, then open', choose a slip, then press Open. Pressing a slip that is already open shows its result large again but adds nothing to the log.
4. Follow along: the side panel shows whose turn it is, the result just opened, the slips-left bar and each item's next-draw chance. In the chart below, you can change which item to track.
5. Finish: press Open all to reveal whatever is left at once, or Reshuffle to start again with the same setup. To keep the results, press Copy log.

Drawing first or last makes no difference

The question that comes up most is whether drawing first is an advantage. If the order is fixed in advance and nobody has seen earlier results, the answer is that every turn has the same chance. With 2 winners among 10 slips, the first person has 2 ÷ 10 = 20.0%. The second person has to weigh the cases where the first slip was a winner or a blank, and (2/10 × 1/9) + (8/10 × 2/9) = 18/90, which is also 20.0%.

Counting arrangements makes the reason clearer. There are 45 ways to place 2 winning slips in 10 positions, and this tool shuffles so that all 45 arrangements are equally likely. The arrangements in which the third position is a winner are those with the other winner in one of the remaining 9 positions, which is 9 of them. The count is 9 for the first position and for the tenth, so every turn has 9 ÷ 45 = 20.0%.

Once you have seen earlier results, things change. The chance is recalculated as winners left ÷ slips left: if the board started with k winners among n slips, and x slips have been opened with w winners among them, the next slip's chance is (k − w) ÷ (n − x). The 'Chance of the next slip' in the side panel and the solid line in the chart show this value, and the dotted line shows the value when earlier results are unknown (k ÷ n).

Blanks raise the chance for those who draw later and winners lower it, but that only applies once you have seen the earlier results. That is why both "drawing first is better" and "drawing last is better" are wrong, and there is nothing to argue about when you set the order.

One round: seven people, two prizes

Say seven people draw one slip each to share two prizes. Set the slips to 2 winners and 5 blanks, and list the seven names in the people field. At the start, the chance that any one person wins is 2 ÷ 7 = 28.6%. Below is one round as an example; real results differ every time you shuffle.

Both prizes were gone by the fourth turn, so the last three people are certain to get blanks. If you agreed that everyone would open a slip, it is still easier for the group to accept the result when you open them all the way through. Copy log gives you one line per turn with the turn, person, slip number and result, for example '1. Chris — Slip 3 — Blank'.

If fewer people draw than there are slips, a prize can be left over

The number of people does not have to match the number of slips. But if seven people draw one slip each from 10, three slips stay unopened, and a prize may be among them. If every prize must be handed out, check this chance first. With c winners among T slips and P people drawing, the chance that at least one winner is left is 1 − (the chance that the P people draw all c winners), and the second term is hypergeometric: C(T−c, P−c) ÷ C(T, P).

Even with a single slip left, one time in five a winner is in it. For a class of 28 students drawing from 30 slips with 3 winners, two slips stay unopened and the chance that a winner is among them is 19.3%. When you list the people drawing, the tool shows this chance right away. Pressing 'Remove n Blank slips to match headcount' removes only blanks, so the slip count equals the headcount. When there are more people than slips, a button that adds blanks appears instead.

For tiered prizes, split the items into several lines

When prizes differ, as in one 1st prize, two 2nd prizes and three 3rd prizes, split the items by tier and make the rest blanks. The prize tiers example has 1 slip for 1st, 2 for 2nd, 3 for 3rd and 6 blanks, 12 slips in all. The chance for the first slip is the item's count ÷ 12, and it is recalculated from what is left after each opening.

The number of distinct arrangements of these 12 slips is 12! ÷ (1! × 2! × 3! × 6!) = 55,440, and all of them are equally likely. The chart starts by following the rarest item, 1st prize, so you can see its chance creep up until it comes out and drop to 0 for every turn after that. Under 'Track item' you can switch to another tier.

How it shuffles: once, when the board is made

When a new board is made, the tool lists each item number as many times as its slip count and shuffles that list with the Fisher–Yates method. Starting from the last position and moving toward the front, it swaps each position with a randomly chosen position among those not yet shuffled. The random numbers come from Web Crypto, the browser's cryptographic generator, and the bias that appears when turning them into integers of a fixed range is removed by discarding values past the range and drawing again (rejection sampling). This makes every arrangement equally likely.

The shuffled result is fixed for each slip number and is never redrawn when you select or open a slip. A number is only a label for telling slips apart, and has nothing to do with rank or chance. The unfolding motion and the small effect on a winning slip are only a display, and neither the speed nor the order of your presses can change or reveal a result.

Common mistakes

• Not matching people to slips: if every prize must be given out and there are more slips than people, a prize may be left over. List the people drawing and use the button that matches the headcount.
• Entering the same name on two lines: two 'Blank' lines in the setup produce an error. Merge them into one line and raise its count. For people drawing, tell people with the same name apart, as in 'Team 1 Alex', so the log stays readable.
• Changing counts partway through: if you change a slip count after some slips are open, the board pauses. Reshuffle clears the log, so copy it first.
• Changing the order after seeing results: if you make up rules on the spot, such as giving up a turn or stopping partway after seeing earlier results, the claim that every turn is equal no longer holds. Fix the rules before opening the first slip.
• Reading Open all results as turns: slips revealed with Open all are logged separately under 'Open all'. Read them apart from the turns people chose themselves.
• Treating a number as lucky: if slip 7 was a winner, that does not make 7 better next time. Every reshuffle decides the arrangement from scratch.

Limits of this tool

• A board holds 2 to 100 slips in total, and item names can be up to 40 characters. For a draw above 100, use another method such as picking a random number.
• To let one person draw twice, type that name twice in the people field. A rule such as skipping people who already won is not handled automatically, so edit the people field yourself for the next round.
• Input and results are handled only inside this browser and are not saved or sent to a server. Reloading or reshuffling clears the log, and no share link is made.
• The hidden results already exist inside the browser, so this is not a guarantee that nobody could technically look. A copied log can be edited and does not show whether the board was rebuilt, so it cannot prove that a draw was fair.
• In very old browsers without Web Crypto, the tool falls back to ordinary random numbers (Math.random).

As of 2026-10-23. The probability formulas and shuffling method were checked against the sources below, and the table values are computed as fractions and rounded to one decimal place.

Sources

• 01Penn State STAT 414: Lesson 7.4 Hypergeometric DistributionThe chance of x items of one kind when n are drawn without replacement from N, m of which are that kind: C(m,x)·C(N−m,n−x)/C(N,n)
• 02Grinstead, C. M. & Snell, J. L. Introduction to Probability (2nd ed.), AMS — 5.1 Hypergeometric DistributionFinding the probability by counting the samples that contain x red balls in a draw without replacement
• 03NIST Dictionary of Algorithms and Data Structures: Fisher-Yates shuffleThe procedure that shuffles N items by swapping each with a random item from the not-yet-shuffled range
• 04MDN: Crypto.getRandomValues()Cryptographically strong random numbers provided by the browser
• 05W3C: Web Cryptography APIThe browser standard that defines getRandomValues
• 06Lemire, D. (2019). Fast Random Integer Generation in an Interval. ACM TOMACS 29(1)How to turn random bits into integers in a fixed range without bias

Frequently asked questions

Tools you may also need