Guide
How to prove a card break was random
By FairBreaks · Updated
A card break is provably fair when the buyer list was fixed before the random seed existed, the seed came from a source nobody in the break controls, and any buyer can re-run the randomization later and get the same result. A wheel spin on stream shows a result; it does not prove the result was the only one drawn.
Why a spin on camera is not proof
Most live breaks randomize with a wheel spinner or a list randomizer shown on stream. Buyers see one result, but nothing shows how many times the randomizer ran before the camera was pointed at it, whether the buyer list on screen matches the one that was sold, or which result was final. When a break is disputed, the only evidence is a stream replay, which answers none of those questions.
The four steps that prove a break was random
- 1
Fix the buyer list before any randomness exists
Publish a fingerprint of the full buyer and team list (a SHA-256 hash) before the random value that will shuffle it has been created. Once the fingerprint is public, changing a single name changes the hash, so the list cannot be quietly edited after the fact.
- 2
Use randomness nobody in the break controls
Take the random seed from a public beacon such as drand, where a network of independent operators jointly signs a new value every 3 seconds. Neither the breaker nor the software vendor can predict or choose it.
- 3
Commit to one draw, in advance
Name the exact future beacon round that will seed the shuffle at the moment the list is locked. There is then one possible result per lock. A re-roll would need a new lock, a new fingerprint and a new public receipt.
- 4
Give every buyer a receipt they can re-check
Publish the list, the beacon round and its signature, and the resulting assignments. Anyone can recompute the hash, check the signature against the beacon's public key and re-run the shuffle to get the same result.
FairBreaks runs these steps for you: the breaker pastes buyers and locks the list, the drand quicknet round published 30 seconds after lock seeds the shuffle, and each buyer gets a public receipt. drand is operated by the League of Entropy, independent organizations including Cloudflare, Protocol Labs and university labs. The exact algorithm is published in fb-v1 so anyone can reimplement the check.
What buyers should check before they buy a spot
- The breaker shares a receipt link, not just a screenshot of a wheel.
- The receipt shows the list was locked before the randomness round it uses.
- Your handle and spot count on the receipt match what you bought.
- The Verify button passes on your own phone, not only on the breaker's stream.
- The breaker has a history of verified breaks. You can check a breaker by handle.
How common card break randomizers compare
| Method | Randomness from | Can it be re-rolled off camera? | List fixed before the draw? | Can buyers check it later? |
|---|---|---|---|---|
| Wheel spinner app on stream | The app, on the breaker's device | Yes, nothing records earlier spins | No proof | Only by rewatching the stream |
| Random.org list randomizer (free) | random.org servers | Yes, every click is a new result | No proof | No public record of the result |
| Dice or shuffled cards on camera | Physical objects | Depends on what the camera shows | No proof | Only by rewatching the stream |
| FairBreaks | drand public beacon, run by independent operators | No, one result per lock | Yes, SHA-256 fingerprint published at lock | Yes, receipt re-verifies in any browser |
The randomness in a wheel app or random.org is not the weak point. What is missing is a public commitment: proof of which list was used and that the result shown was the only one drawn.
Questions about fair card breaks
- Is a random.org list randomizer fair enough for card breaks?
- The randomness itself is good, but the free list randomizer does not prove which result was used. Each click gives a new ordering and the tool keeps no public record, so buyers cannot tell whether earlier results were discarded off camera. Random.org does sell a separate Third-Party Draw Service that keeps public records.
- Can a breaker fake a verified result?
- Not without breaking the receipt. The buyer list is hashed before the drand round exists, the round is fixed at lock, and the shuffle is a fixed function of both. Editing a name, swapping teams or picking a different round makes the Verify check fail in the buyer's browser.
- Does it work with Whatnot, eBay Live and TikTok Shop?
- Yes. FairBreaks does not connect to the selling platform. The breaker pastes the buyer list, so it works with Whatnot, eBay Live, TikTok Shop, Fanatics Live and any other platform or in-person break.
- What should a buyer do if a breaker will not verify?
- Ask them to run the next break through a verifiable randomizer and share the receipt link before they rip. A breaker's public record on FairBreaks shows how many breaks they have verified and that none were re-rolled.
Run your next break through FairBreaks.
Free for unlimited verified breaks. Works with any platform because you paste the buyer list.