How mystery box sites work is simpler than the animation suggests: you buy a box, the site publishes what is inside and the chance of each item, and a draw decides which one you get. On a provably fair site that draw is not a black box. It comes from values the site commits to before you click, and you can recompute it afterward in about a second. Of the 17 platforms we track, 9 have a game our tool can check and 8 do not.
The four steps of a box opening
Every platform in this space follows the same sequence, whatever the interface looks like.
1. You read the box before you buy
Each box publishes a drop table: every item that can come out, and the chance of each one. Read the table, not the animation. Confirm the item you are interested in is in the pool at all, and at what rate, because the animation that follows tells you nothing the table has not already said.
2. The site commits to a secret
Before the draw, the site generates a secret server seed and shows you a fingerprint of it rather than the value itself, a hash there is no way to run backward. A SHA-256 fingerprint is 64 hexadecimal characters and a SHA-512 one is 128. That fingerprint is the commitment. Once it is published the site cannot change the underlying value without the fingerprint no longer matching.
3. The draw runs
Your outcome comes from that server seed combined with your own client seed and a nonce, a counter that goes up by one with each opening. Those three go through a hash, and the result becomes a number in a published range.
4. You take the item or sell it back
Whatever the number lands on is yours. Most platforms let you either have the item shipped or sell it back for site credit to use on another box. The two routes behave differently, so know which one you are choosing: shipping puts a physical object in the post, with whatever delivery window and country restrictions that platform operates, while selling back converts the item to credit at the site's own valuation, which is not always what you would expect the item to be worth elsewhere.
Nothing in the fairness system speaks to either of those. The draw is settled the moment the number lands; everything after that is ordinary e-commerce and support, judged on ordinary evidence.
What actually decides your item
The number the draw produces is not the item. It is a ticket, and every item in the box holds a block of tickets sized to its published chance.
The arithmetic is deliberately plain. If a site runs a ticket space of 100,000 and an item is listed at 1%, that item holds exactly 1,000 tickets. At 0.1% it holds 100. Your ticket falls in one block and one only, and that block is your item.
That mapping is what makes a box check stronger than a plain number check. Reproducing the ticket proves the number was fixed before you clicked. Seeing which published block it falls in proves that number corresponds to the item you were actually handed. A site that publishes a formula but no per-item ranges gives you the first half only.
Ticket spaces differ by platform, which matters if you ever check a result:
| Platform | Outcome range |
|---|---|
| PackDraw | Ticket, 1 to 1,000,000 |
| Skin.club | Roll, 1 to 100,000 |
| Clash.gg | Ticket, 0 to 99,999 |
| JemLit | Roll, 0 to 100, to 4 decimal places |
| Ripster.gg | Roll, 0 to 100 |
Five different answers to the same question. A verifier set to the wrong platform hands back a plausible wrong number: it looks entirely reasonable, and it is false. That mismatch is the most common reason a check appears to fail when nothing is actually amiss. The full list for all 17 platforms we cover is on the algorithms by platform page.
Why the commitment matters
The commit-then-reveal step is the whole difference between a site you can check and one you have to trust.
What the fingerprint prevents
Because the fingerprint is published before you play, the site cannot see what you do and then pick a server seed that suits it. Because your own client seed is mixed in, and you can change it whenever you like, the site cannot aim a specific item at you either. Neither side controls the result alone.
Why the reveal is the other half
A commitment nobody ever opens proves nothing. The site has to eventually publish the real server seed, normally when you rotate to a new one. At that moment every opening made under the old seed becomes checkable: you hash the revealed value, confirm it produces the fingerprint you saved, then recompute the outcome.
If a site shows a fingerprint but never reveals what it was a fingerprint of, you have been handed a sealed envelope that never gets opened. Our guide to spotting a fake provably fair claim covers the six checks that separate a real claim from a decorative one.
Check an opening yourself
You do not have to take any of this on trust. Take the revealed server seed, your client seed and the nonce from an opening you already made, and recompute it here. Everything runs on your own device, so nothing you paste reaches us or the operator.
Step 1Where did you play?
Using Ripster box / upgrade: HMAC-SHA256 over clientSeed:noncedetails
HMAC-SHA256 keyed with the server seed over clientSeed:nonce. The first 8 hex characters become a 32-bit integer, divided by 2^32 and multiplied by 100 for a roll in [0, 100). Items are sorted by item ID and selected by cumulative probability: the first item whose cumulative probability is at least the roll wins.
Where to find your numbers on Ripster.gg:
- Before rolling, copy the Server Seed Hash from the Provably Fair modal (also recorded per roll in Account History).
- After the roll, open the Provably Fair modal to copy the Server Seed, Client Seed and Nonce.
- Change your Client Seed anytime in Account Settings; doing so reveals the current Server Seed so all past rolls become verifiable.
Source: Ripster.gg's own fairness page
Step 2Paste your numbers
Runs entirely on your device via your browser's built-in cryptography. Don't take our word for it: open DevTools → Network, click the button, and watch: zero requests.
Runs entirely in your browser. Nothing you paste is sent anywhere.
A passing check confirms three things at once: the revealed server seed matches the fingerprint published before you played, the outcome recomputes to the number the site showed, and on platforms that publish ticket ranges, that number lands on the item you actually received.
If you are new to the three values, what is provably fair walks through the whole flow, and how to verify provably fair covers the steps one at a time.
Runs of nothing are normal, and the maths says so
The single most common misreading of these sites is treating a cold run as evidence that something is wrong. On a provably fair site every opening is independent, and no item becomes "due" because you have missed it.
Low probabilities produce long empty stretches as a matter of course:
| Item listed at | Chance of seeing zero across 100 openings | Openings for 95% confidence of one hit |
|---|---|---|
| 1% | 36.6% | about 298 |
| 0.5% | 60.6% | about 598 |
| 0.1% | 90.5% | about 2,994 |
Read the middle row. On a one in two hundred item, opening 100 boxes and seeing none of it is not a bad night, it is the single most likely outcome. If you want to test your own history against a site's published rates rather than going on feel, the variance checker does that arithmetic for you.
What a mystery box site is not
Two comparisons come up constantly, and both mislead in their own way.
It is not a sealed pack
A sealed booster publishes pull rates too, but you can never test them on an individual pack: nothing was committed in advance and there is no value to recompute afterward. Open a pack with no hit and there is nothing to do but box the bulk; no arithmetic can tell you whether that particular pack was fair. The promise is statistical, testable only across hundreds of openings. A provably fair box makes a per-outcome promise instead, settled on one opening. We cover that difference in published odds vs verifiable odds.
It is not something you can predict
The server seed is committed as a one-way hash before your client seed even exists, so nothing can read a future outcome in advance. Not you, not the site, not a paid tool. Any product promising to forecast your next pull is selling something that cannot exist, which is the subject of our case opening predictor breakdown.
What checking does and does not settle
Hold onto that boundary rather than collapsing it into "safe" or "rigged". What provably fair does not prove covers it in full, and how to tell if a mystery box site is rigged covers the other signals that matter.
The bottom line
How mystery box sites work comes down to four steps: a published drop table, a committed secret, a draw from three values, and an item you can ship or sell back. The part that separates one site from another is whether you can check the third step yourself. On 9 of the 17 platforms we track you can, in about a second, on your own device. On the other 8 you are taking the same claim on trust.
Before you spend anything, read the drop table and check whether the site publishes seeds you can retrieve. If it does, run one small opening through the verifier and watch it check out. If it does not, at least know that is the position you are in. If any part of this has stopped being fun, take a break and get support.