Is a screen-recorded randomizer enough? Provably fair vs a simple spin
A randomizer shown on stream is fine for picking, but it leaves buyers unable to check afterward. A provably fair randomizer is built so that they can.
Many breakers use a public randomizer, such as random.org, to match teams and buyers. Cardboard Connection's group breaking guide describes the usual pattern: enter team names and buyer names, randomize each list, then match them. That is a sound way to pick a result. The question is what a buyer can confirm later.
Side by side
- Screen-recorded spin. Buyers see the result on the stream. After the break they cannot recompute it. They rely on the video showing the same lists, one run, and no edits.
- Provably fair (commit, then reveal). The lists and a secret seed are locked first. Viewers add their seed. Afterward the seed is revealed and anyone can recompute the shuffle and confirm it matches. A changed list or a different run fails the check.
What BreakSeal does
- A SHA-256 commitment over the secret seed and the lists is shown before the break.
- A viewer seed is added after the commitment is locked.
- The shuffle uses an HMAC-SHA256 stream, rejection sampling to avoid bias, and a Fisher-Yates shuffle, so every order is equally likely.
- The verify page recomputes everything in the viewer's browser.
When a simple spin is enough
If your buyers already trust you, a recorded spin may be all you need. A verifiable shuffle helps most when you sell to people you do not know, when the stakes per spot are higher, or when someone asks "how do I know it was not rigged?"
Provably fair covers the randomizing step only. It does not replace honest rules, clear recordings of the break, or accurate shipping.