Does a digital random number generator actually count?

Not every charity lottery draw looks like the ones you've probably pictured: a barrel of numbered balls, a person of standing watching closely, a room full of onlookers. A growing number of NZ charity lotteries pick their winner with a random number generator instead. That's legitimate under the rules. It's also worth actually thinking through what those rules can even mean once the thing being witnessed isn't a physical object anymore.

What the rule actually says

DIA's Lottery Game Rules set out two separate safeguards for how a lottery draw has to happen. One applies to every class of lottery: "the drawing of the lottery must be open to the public." The other applies specifically to Class 3 lotteries: "the drawing of the lottery must be supervised by a person of standing in the community who is independent of the society conducting the lottery." Neither clause says anything about method. Nothing here distinguishes a physical draw from a digital one.

There's a separate section covering what happens once a ticket is identified as the winner, verifying it belongs to the right person and is genuinely the winning ticket, but that section says nothing about how the winner gets selected in the first place. The actual randomising step is only ever addressed through those two clauses above, and both of them were clearly written with a physical event in mind.

Why a physical witness actually works

It's worth being precise about why standing a person in the room is meaningful in the first place. At a physical draw, a witness is verifying something they can directly observe: the balls are all there, nobody's hand favoured one, the ball that came out is the one reported. The thing being verified, the actual mechanical act of randomising, is visible. The mechanism and the observable event are the same thing.

Why that breaks for software

With a random number generator, the thing that actually determines the outcome, the code executing, isn't visible to a human eye at all, no matter how closely someone watches. A witness standing next to a screen when someone clicks "generate" can see a button get pressed and a number appear. They cannot see whether that number was genuinely unpredictable, or whether the software was quietly set up to produce a specific result and simply display it as if it had been random. The mechanism and the observable event are no longer the same thing.

"Open to the public" has the exact same problem, just less obviously

Here's the part worth sitting with, because it's easy to assume a livestreamed digital draw solves this. It doesn't. A charity could genuinely satisfy the letter of "open to the public" for a digital draw: run the generator live, in front of an audience, or on a livestream at the stated time. People really can watch it happen. But watching a screen show a result isn't the same kind of "open" as watching balls get physically mixed and pulled. With a physical draw, the mixing itself is what's visible, and that's what makes "open to the public" meaningful. With a livestreamed digital draw, the audience only ever sees the output appear. What actually happened inside the software to produce that number stays completely invisible, no matter how many people were watching the screen.

For a physical draw, "open" and "verifiable" are the same thing. For a digital one, they come apart completely, and a charity can tick the box on both required safeguards, a supervised, publicly watched button-press, without having delivered any of the actual accountability those rules were built to guarantee.

So can a digital draw be witnessed in a way that's actually meaningful?

Yes, but only by redefining "witnessing" into things that are genuinely checkable, and none of them look like a person standing in a room:

Each of these verifies something real about the process itself. None of them is a person standing there watching, and none of them is currently required anywhere in DIA's rules.

What this looks like in practice for a small NZ charity lottery

Nothing found here suggests NZ's Class 3 charity lotteries are commonly held to, or use, that level of technical rigour. One real example: a major NZ raffle's own published terms state only that "the winning ticket will be selected using a random number generator." No named software, no published seed, no independently auditable method. That's not necessarily evidence of anything wrong. Most likely it's exactly what it looks like: a genuinely random draw, run in good faith. But the draw is being taken on trust in a way a properly run physical one, by its nature, isn't.

What's worth asking, if you want to check

A charity using a digital draw and being genuinely transparent about it will usually be willing to say more than "a random number generator was used." Worth looking for: whether they name the actual method, whether a specific winning ticket number gets published afterward the same way a physical draw's result would be, and whether there's any way, even informally, to ask how the process worked. None of this is currently spelled out by DIA in the same detail as the physical-draw rules. That's exactly why it's worth charities holding themselves to it anyway, and worth a ticket buyer knowing to ask.