Running a raffle by ticket number instead of by name

 Walk into any school fête, community hall, or charity dinner and the raffle works the same way it always has: you buy a strip of tickets, tear off the stubs, drop the other half in a box, and wait for a number to get called. Nobody collected your name at the point of sale. The ticket number is the entry, and that single difference changes almost everything about how you should run the draw compared to a typical name-based giveaway.

This is a practical guide to running a raffle by ticket number: why the ticket-number model works differently from a name list, how to set the range up correctly, the specific mistake that ruins more raffles than anything else (unsold tickets left in the pool), handling multiple prizes, and how to actually verify and announce a winning number.

Why raffles use ticket numbers instead of names

A few structural reasons this became the standard, and understanding them clarifies what you actually need from a raffle number generator.

Tickets are sold before anyone collects names. Physical raffle books are pre-printed with sequential numbers, and you sell them at a table, a door, or door-to-door, often to people who never give you their name at the point of sale. A name-based entry system requires you to capture and manage that information; a ticket-number system doesn't need it until a winner is actually drawn.

One person can legitimately hold several tickets. This is the biggest mechanical difference from a giveaway. In a typical Facebook comment giveaway, you generally want one entry per person and remove duplicates. In a raffle, someone buying five tickets is completely normal and by design, since ticket sales are how a raffle raises money. Each of their five numbers is a separate, equally valid chance, and there's no "duplicate" to remove, five tickets means five real entries for the same person.

The stub system is its own verification. The physical half kept by the buyer, matched against the half kept by the organiser, is a built-in way to confirm a winner is genuine without needing any database at all. A digital equivalent needs to replicate this somehow, covered further down.

It scales to events where a name list isn't practical. A raffle at a festival with thousands of attendees would be an administrative nightmare to run as a name-collection giveaway. Numbered tickets sidestep the whole problem.

Setting the range correctly

The core of a ticket-number draw is a number wheel or number picker set to the range your tickets actually cover, but "just set it to match the ticket book" is where the most common and costly mistake happens.

Match the range to tickets actually issued, not tickets printed. If your raffle book runs from 1 to 500 but you only sold or distributed tickets 1 through 350, the valid range for the draw is 1 to 350, not 1 to 500. Leaving the full printed range in the generator means roughly 30% of possible draws in this example would produce a number nobody holds, which isn't just wasteful, it's a real fairness problem if you redraw quietly until you land on a sold number, since that's a form of hidden manual intervention in what's supposed to be a transparent random process.

Handle scattered unsold tickets properly. The clean version of the above assumes unsold tickets are a solid block at the top of the range, which is common but not universal. If unsold tickets are scattered throughout the book, a simple min-max range won't exclude them correctly, and you'll need either a list of the specific valid ticket numbers rather than a plain range, or a manual redraw policy stated in advance for hitting an unsold number, announced and applied consistently rather than decided in the moment.

Confirm inclusive boundaries. A range set to "1 to 500" should give ticket 1 and ticket 500 exactly the same odds as every number between them. Test this on any tool you haven't used before, since a boundary error at either end is invisible until you specifically check for it.

Account for any voided or refunded tickets. If a ticket was sold and later cancelled or refunded, it needs to come out of the valid pool the same way an unsold ticket does, and this is worth tracking as sales happen rather than trying to reconstruct at draw time.

The unsold-ticket problem, in detail

This deserves its own emphasis because it's the single most common way well-intentioned raffles become quietly unfair, and almost nobody checks for it.

Picture a school raffle: 1,000 tickets printed, 640 actually sold by the time the draw happens. If the organiser sets a generator to the full 1–1000 range "because that's the ticket book," then 36% of possible draws land on a number nobody bought. When that happens, the honest options are to redraw immediately and transparently, stating clearly that the number wasn't sold, or to have pre-committed, in the published rules, to a specific policy for this exact situation.

What actually happens too often is worse: an organiser redraws quietly, without announcing why, which looks, from the outside, indistinguishable from picking a different result because they didn't like the first one. The fix costs nothing: set the generator's valid range to match exactly what was actually sold, not what was printed, and state your redraw policy for unsold numbers in the raffle's published rules before you start selling tickets at all.

Multiple prizes and drawing without replacement

Most raffles award more than one prize, and this is where a second common mistake shows up: letting the same ticket win twice when the rules didn't intend that.

If your raffle has a first, second, and third prize, and the rules imply or state that a ticket can only win once, you need to draw without replacement, removing each winning number from the pool before the next draw, so ticket 214 can't be pulled for both first and second prize. This is a deliberate setting, not the default behaviour of most simple generators, and it's the same underlying mechanism covered in more depth in draw without replacement, which applies to numbered tickets exactly as it does to named entries.

The flip side matters too: some raffles genuinely do allow one ticket to win multiple prizes, and if that's your intent, leaving replacement on (independent draws) is correct, not a bug. The only real requirement is deciding which rule applies and stating it clearly before the draw, since both are legitimate raffle structures and the dispute only arises when nobody said which one you're running.

Wheel or plain generator, for a raffle specifically

Raffles tend to involve larger ranges than a typical classroom or small giveaway wheel, often hundreds of tickets, which changes which tool actually fits.

A visual wheel stays genuinely readable only up to roughly 20 to 30 segments before the individual numbers become unreadable slivers, so a literal spinning wheel with five hundred numbered slices isn't a realistic option, technically possible in some tools but useless as a way for your audience to actually see the pool. For a raffle of any real size, a plain number generator that returns a result instantly, with the range set correctly as covered above, is the more honest and more practical choice. The full reasoning behind this trade-off, including exactly where the readability ceiling sits and why a wheel isn't automatically the more trustworthy-looking option once a range gets large, is covered in random number generator vs number wheel.

If your raffle happens to be small, under thirty tickets, a genuine wheel works fine and adds a bit of visible ceremony that suits an in-person draw. For anything bigger, don't force the range onto a wheel just for the visual, since an unreadable wheel doesn't actually demonstrate anything to your audience, and a plain, clearly-stated result does the job better.

Verifying and announcing the winning number

Drawing the number is the easy part. Making the result trustworthy and actually claimable is where a bit of process matters.

Have a way to match the number to a real ticket. For physical ticket sales, this is the classic stub system: the organiser's kept half should have the same number as the buyer's half, so when a number is drawn, you can check it against your retained stubs before announcing. For any digital sales, keep a simple record mapping ticket numbers to purchasers so you can confirm a claim quickly.

Announce the number, not a name, first. Since you likely don't have names attached to tickets until someone comes forward to claim, the announcement itself should be the ticket number: "Ticket 214 wins first prize," followed by a claim process ("bring your stub to the desk," "contact us with your ticket number") rather than naming a person you may not know yet.

Set and publish a claim window. Exactly as with any prize draw, state how long a winner has to come forward with a matching ticket before the prize moves to a redraw, and decide that policy in advance rather than improvising when someone doesn't show up.

Record the draw if you can. A short recording of the actual draw, showing the valid range being used and the number being generated, is straightforward proof of a fair process, and it costs nothing beyond a phone camera pointed at the screen for the few seconds the draw takes.

Legal basics worth a mention

One important point, kept brief because it's a separate topic from the mechanics covered here: a raffle where people pay for tickets is a fundamentally different legal category from a free-entry giveaway, since paid entry plus a prize plus chance is the classic definition of a lottery in most jurisdictions, and that generally means registration, licensing, or specific charitable-exemption rules apply depending on where you're running it. None of the mechanical advice above changes that requirement. If you're organising a real-money raffle, confirm your local rules before selling a single ticket, since the number-generator setup described here handles the draw itself, not whether you're legally allowed to run one.

A quick pre-draw checklist

Before you generate a number, confirm: the range matches tickets actually sold, not tickets printed. Any voided or refunded tickets are excluded. You know your policy for an unsold number coming up, and it's written down, not improvised. You've decided whether repeat wins are allowed or you need to draw without replacement. You have a way to verify a claimed ticket against your records. And you have a stated claim window for the winner to come forward.

The bottom line

Running a raffle by ticket number is a genuinely different task from a name-based giveaway draw, not just the same mechanic with numbers instead of names. The core discipline is getting the valid range exactly right, matching what was actually sold rather than what was printed, and having a stated, pre-announced policy for the numbers that fall outside it. For multiple prizes, decide deliberately whether one ticket can win more than once, and use a draw-without-replacement approach if not. For sizeable raffles, a plain number generator set to the correct range beats forcing hundreds of tickets onto an unreadable wheel, a trade-off covered further in random number generator vs number wheel, and the same range-handling principles that apply to a smaller draw like 1 to 100 scale up cleanly to a raffle of any size. Get the range and the policy right before you draw, and a raffle number generator does exactly what the physical tombola always did, just faster and with a cleaner record.