Comparisons

Spinning wheels: why the animation proves nothing, and how to fix it

· · By Yuval Abudarham

A draw your entrants can recompute themselves

Seed published before the draw, public drand randomness, shareable video. Free for entrants.

Create an accountVerify an existing draw

The spinning wheel is the most shared moment in giveaways. It produces a second of suspense, a shout, a stop. Excellent for engagement, and it must be said plainly: it proves nothing at all.

The order of operations

What a viewer believes they see: the wheel spins, slows, stops at random on a name.

What actually happens, in virtually every implementation:

  1. a random generator picks the winner;
  2. the animation computes the final angle needed to stop on that name;
  3. the wheel is animated to that angle.

The animation is not the draw. It is its playback. This isn't deception — it's the only way to produce a smooth animation — but it radically changes what the video attests.

What the video proves, and what it doesn't

It proves that an operation took place at a given moment, and that the displayed name is the one announced.

It does not prove:

  • how many times the operation was run before the one that was filmed;
  • who was on the wheel, and above all who wasn't;
  • that the list of names wasn't edited just before;
  • that the result wasn't fixed, then the animation replayed to reach it.

These four blind spots share one thing: they all concern what happened before the camera rolled. Video is proof of existence; what you need is proof of integrity.

The real problem: the hand-typed list

This is the most serious blind spot, and the least discussed. On most wheels, names are pasted by hand into a text field.

An entrant watching the video sees twelve names. They have no way to know whether there were fifteen, nor whether theirs was removed before the paste. The tool doesn't know the entrant list — it only knows what it was handed.

In other words, the question of the draw's honesty moves entirely upstream of the tool, where nothing is observable.

Keep the show, add the proof

The conclusion isn't to abandon the animation. It's to stop asking it to do what it cannot, and to place the proof elsewhere.

The order that works:

  1. Before entries open: publish the fingerprint of a secret seed and announce which public randomness source will be used.
  2. At closing: seal the entrant list — the fingerprint is published, the list can no longer move.
  3. At the scheduled time: the public randomness appears, the winner is computed.
  4. Only then: the animation plays back that result, and the video becomes a communication object, openly.

On screen the animation is unchanged. What changes is that it is no longer the only thing available: behind it, an entrant can recompute the result without taking your word.

That is how DrawSeal handles video: it is produced after the draw, from already-published values, and it carries what's needed to verify what it shows.

Key takeaways

On a wheel, the winner is chosen before the animation starts: it plays back a result, it doesn't produce one. Video attests that an operation happened, not how many times it was re-run nor who was on the list — typed by hand in most tools. Keep the animation for what it does well, communication, and put the proof upstream: published commitment, sealed list, public randomness.

FAQ

Can a wheel be honest?

Of course, and most are. The problem isn't the organizer's honesty, it's their inability to demonstrate it. A blameless organizer with no proof stands in the same position as a cheat.

Does filming live change anything?

A little: live removes editing. It removes neither off-camera trial runs nor the question of the list. It makes the attack costlier, not impossible.

Should the list be shown on screen before spinning?

Better than nothing, and still weak: the list shown is the list someone chose to show. A fingerprint published at closing can no longer be adjusted.

Does the animation have legal value?

None as such. What has value is the rules, and the conformity of the process to them. Video can illustrate that conformity; it doesn't establish it.

Try it