Do I need to record myself doing the draw?

A recording proves that a draw happened and that the result on screen is the one you announced. It does not prove that it was the first draw you ran, because nothing in the video shows what you did before you pressed record. Publishing the full ordered result rather than only the winner is stronger evidence than a video of the winner alone, and it costs nothing.

What if the winner never replies?

Decide the deadline before the draw and draw the reserves in the same pass, in order. A second draw held later looks like a second chance for someone, however honestly it was run, and it reopens an argument you had already closed. With a published reserve order there is nothing to decide when a winner goes quiet — the prize moves down the list.

Someone says the draw was rigged. What can I actually show them?

You cannot demonstrate after the fact that a number was random; no result looks more random than any other. What you can show is that the rules, the closing time and the entry count were published before the draw, that the entry list matches the count, and that the announced order includes reserves you would have had to predict in advance to fake. Fairness disputes are settled by prior commitment, not by the result.

When is a draw like this not good enough?

When there is enough money involved that someone would pay to challenge it, or when the draw falls under rules in your jurisdiction covering lotteries and prize competitions. A draw run in a browser produces no record an independent party can verify, which is exactly what an auditor or an adjudicator will ask for. Take advice locally rather than assuming an informal procedure will carry.

How to run a prize draw people will accept as fair

The hard part of a giveaway is not producing a random number. It is being able to answer, three days later, someone who is convinced their entry was never in the hat. That is not a randomness problem, and no generator solves it.

Fairness is a claim about the past, not the result

Here is the uncomfortable part. Once a draw is over, you cannot demonstrate that the result was random. Every possible outcome is equally consistent with a fair draw and with a rigged one — the winner being your cousin proves nothing either way, and neither would any other name. There is no property of a result you can point at and say look, that is what randomness looks like.

So a draw is not accepted because it was random. It is accepted because of what you fixed in place before the outcome existed, and how little room you left yourself afterwards. Everything below is a way of narrowing that room.

Before: publish the things people argue about

Almost every giveaway dispute traces back to a rule that was never written down. Write these down, in public, before entries close:

The closing time, with a timezone. “Midnight” has caused more arguments than any other word in this list.
Who is eligible, and who is not — staff, family, previous winners, entrants in particular countries if shipping is a constraint.
How many winners, and whether one person can win more than one prize.
Whether duplicate entries count. One entry per person and one entry per action are different competitions, and people will assume whichever suits them.
How long a winner has to respond, and what happens when they do not.

None of this is bureaucracy for its own sake. Each line is a decision you would otherwise have to make after seeing the result, which is precisely when any decision you make looks self-serving.

Freeze the list, and publish the count

At closing time, export the entries and stop touching them. Then publish one number: how many there are.

That single number does more work than anything else on this page. “Was my entry counted?” is the most common challenge to any draw, and a count published before the draw makes the entry list falsifiable — if someone can show a hundred and forty people entered and you announced a hundred and twelve, you have a problem to explain. Without a count there is nothing to check, and the honest organiser and the dishonest one look identical.

If entries can be edited or deleted after the fact, keep a plain export from closing time. You are not building a legal record — you are making sure the list you drew from still exists in a week.

During: draw everything in one pass

Draw your winners and your reserves in a single go, and keep the order.

This is the step people skip, and it is the one that removes the most suspicion. A single-winner draw invites a second attempt — not usually a dishonest one, just a hmm, let me check that worked that quietly becomes best of three. An ordered list of ten from a pool of four hundred is very hard to fabricate convincingly after the fact, because faking it means having decided in advance who your fake reserves would be, in order, and then living with that list when the first two do not reply.

It also disposes of the awkward case before it happens. When a winner goes quiet, the prize moves to the next name on a list everyone has already seen, rather than being settled by a second draw that looks like a second chance for whoever wins it.

After: announce more than the winner

Publish the entry count, the winner, and the reserve order. Say when the draw was run. If you recorded it, post the recording — but understand what it is worth: a video shows that the draw you announced is the draw that produced that result. It says nothing about what happened before you pressed record, which is where a rigged draw would have happened. The ordered list is the stronger artefact, and it is free.

When someone challenges it anyway, the answer is not to argue about randomness. It is to point at the rules published before entries closed, the count published before the draw, and the reserve order published with the result. If those three line up, there was no moment at which you could have intervened without leaving a trace.

Where this stops

This is a procedure for a draw run in good faith among people who mostly trust you already — a community giveaway, a club raffle, a competition on a small account. It is not an audit trail. A browser draw leaves no record an independent party can verify, and that is the first thing anyone paying a solicitor will ask for.

Prize draws are also regulated differently from one country to the next, and the line between a free draw, a skill competition and a lottery is narrower than most organisers expect. If money changes hands to enter, check what applies where you are first.

Questions organisers ask

Should I let people see the entry list?

Publish the count always; publish the list only if entrants knew it would be public. Names and email addresses attached to a giveaway are personal data, and a surprise public list is a worse problem than an unverified draw. A count, plus a willingness to confirm one person's entry privately, covers almost every real challenge.

Can I exclude someone after the draw?

Only for a reason you published before it. Disqualifying a winner on a rule invented after their name came out is the single fastest way to turn a quiet giveaway into an argument, and it is indistinguishable from rigging even when it is entirely justified. If you think you will want to exclude bot accounts, duplicate entries or ineligible countries, say so while people can still choose whether to enter.

Is a random draw always the fair choice?

No, and it is worth asking. A draw is fair when every entrant has an equal claim. Where effort, need or merit genuinely differ, a random draw avoids a judgement rather than making a fair one, and people can usually tell. If some entrants deserve a better chance, that is a weighting decision and it needs stating openly rather than smuggling in.

The mechanics of the draw itself sit on the tool pages: the random number generator explains where its numbers come from and how a seed turns an informal draw into one a sceptic can re-run, and the random name picker is the one to reach for when you are drawing from a list of names rather than numbered tickets.