Trading card pack sites have taken the "provably fair" label straight from the wider mystery box world, and it now sits on the front page of almost every Pokemon, One Piece, and sports card ripping site. This list ranks provably fair TCG sites on one practical question: can a normal user take a real result and confirm it, using the site's own published inputs and a standard hash, without trusting anyone's word? It is not a ranking of pull value, sealed product, bonuses, or star ratings. For provably fair Pokemon box sites, the only thing scored here is how far you can actually check the draw yourself.
How this list ranks sites
Every card site below is scored on four questions you can confirm for yourself, in this order of importance:
- Does it publish its algorithm? The exact hash function and how the server seed, client seed, and play count are combined into a result. Without this, there is nothing to recompute.
- Can you reach your own seeds? The hashed server seed before you rip, and the revealed server seed, client seed, and nonce afterward. No exposed seeds, no check.
- Is the result reproducible? Given those inputs, does re-running the published method produce the same outcome the site displayed?
- Can you map the result to the card? The best card sites publish per-item ranges, so you can confirm not just that the draw was honest but that the exact card you received matched your number.
CheckMyPull is not a pack site and does not sell cards, so it never appears as an entry in this list. Its role here is the neutral instrument: each site is ranked by whether its fairness claim actually verifies, using an independent verifier you can run yourself, which recomputes the outcome on your own device and sends nothing anywhere.
The comparison at a glance
| Site | TCG focus | Algorithm published? | Seeds you can reach | Independently verifiable? |
|---|---|---|---|---|
| Ripster.gg | Pokemon, One Piece, sports cards | Yes, with verifier code | Yes | Verified (cross-checked against its own verifier) |
| Boxed.gg | Pokemon, Magic, Yu-Gi-Oh | Main steps, plus a linked code sandbox | Yes, server seed revealed after 24h | Partial (one step lives off the page) |
| PackDraw | Sports cards, Pokemon | Yes, full method | Yes, in your account history | Partial (no confirmed real result yet) |
| Pullbox.gg | Pokemon, Magic, Lorcana, One Piece | States the scheme, exact code login-gated | On its own site, after login | Not with an outside tool today |
| ArenaClub | Sports cards, Pokemon | No seed-commit method published | No | No, not a provably fair system |
| Packz.io | Sports cards, Pokemon | No, publishes pull odds only | No | No, no seed system to check |
Why the card mapping matters more here
On a plain number game, verifying the draw is the whole job: reproduce the roll and you are done. Card ripping adds a second step that matters just as much, because the number has to land on a specific card. The strongest card sites publish per-item ranges, so a roll of, say, 96.31 out of 100 maps to a defined slice of the range that belongs to one card and no other. That two-part check, the roll is honest and the card it produced was the correct one for that number, is what separates a site you can fully verify from one where you can only confirm half the process.
This is also where trading card sites can be checked more thoroughly than most number games, and where a published odds table alone falls short. An odds table tells you a chase card is a 0.5 percent hit before you buy. A per-item range plus exposed seeds lets you prove, after the fact, that your specific pull fell inside or outside that card's slice exactly as claimed. Ripster.gg and Boxed.gg both publish those ranges, which is a large part of why they rank where they do. When a site shows odds but no ranges and no seeds, you are back to trusting the number, which is the thing provably fair is supposed to remove.
Fully verifiable: you can reproduce a real result
These are the card sites where the "provably fair" label survives an actual check.
Ripster.gg is the clearest example on the trading card side, and it is the strongest verifiable option aimed squarely at the TCG crowd, covering Pokemon, One Piece, and sports cards. Its fairness page publishes the complete scheme: your box result comes from an HMAC-SHA256 keyed by the server seed over a message of your client seed and nonce, the first 8 hex characters of that digest become a number, and dividing by 2 to the 32nd power and multiplying by 100 gives a roll between 0 and 100. The page ships copy-paste verifier code and its own in-browser check, and each item in a box carries a percentage range so you can map the roll to the card. Running the published method against Ripster's own verifier produces the same roll, so the draw is reproducible in the strict sense.
One honest caveat belongs next to that. Ripster.gg is very new and has little public track record: at the time of writing there is no established review history for it the way there is for older names. Verifiability and reputation are separate things, and this list scores only the first. On that measure it qualifies, and you can load it in the Ripster.gg verifier and check a real rip yourself.
If you want the single cleanest verified method anywhere in this space, JemLit publishes runnable code and per-item win ranges and reproduces a published example exactly, though its catalog is general merchandise rather than trading cards, so it sits outside the card focus of this list. The wider provably fair mystery box sites ranking covers it in full.
Verifiable, with a catch
These card sites give you almost everything, but one step takes extra effort or is not yet nailed down against a real result.
Boxed.gg is the most established trading card focused option here, covering Pokemon, Magic: The Gathering, and Yu-Gi-Oh, and it is the site most Pokemon rippers will reach for first. Its help center documents the inputs (a client seed, a server seed shown hashed before you open and revealed every 24 hours, a per-account nonce, and a permanent account salt), states that each open produces a number between 0 and 999,999 that maps to per-item ranges, and every box shows a drop table with a "provably fair range" for each card. The one catch is that the exact step turning the hash into that ticket number is not printed on the fairness page itself. The published sandbox code shows the HMAC-SHA512 stage but stops before the hex-to-ticket conversion, so a full independent check is possible but takes an extra hop off the page. You can still load a real Boxed.gg result in the Boxed.gg verifier and check the parts that are public.
PackDraw publishes its method plainly and it is well documented for sports cards and Pokemon rips. The server seed's hash is shown before you buy, and afterward the revealed server seed, your client seed, and a nonce are combined, hashed, and normalized to a ticket between 1 and 1,000,000, with battles using a public blockchain block hash in place of the client seed. Your seeds sit in your account and purchase history. The honest caveat, the same one that applies on the general list, is that PackDraw does not publish a worked example, so while the published method is implemented in the verifier, it has not yet been confirmed against a real published PackDraw result. Treat a check here as a test against the published method rather than a settled match until a real seed set confirms it end to end.
Publishes a system you cannot fully check yet
Pullbox.gg sits one notch below. It is a genuinely card-first site, covering Pokemon, Magic, Lorcana, and One Piece, and its fairness page states it uses HMAC-SHA512 over the server seed, client seed, salt, and nonce, converted into a number across the box's total weight, with an on-site verifier. The problem for an outside check is that the exact order, separators, and conversion live in a verification snippet that only opens from the login-gated Verify button, so an independent tool cannot reproduce a Pullbox result without an account. On its own site, logged in, the seeds and the built-in verifier are there; from the outside, the method is described but not fully reproducible today. That is a real distinction, and it is why Pullbox lands here rather than in the verified tier.
Not provably fair: publishes odds, not proofs
Two card sites that show up in "provably fair" searches do not actually run a seed-commit system, and it is worth being plain about that.
ArenaClub, the graded-slab pack site co-founded around the sports card and Pokemon audience, does not publish a provably fair algorithm at all. There is no seed-commit page, no exposed server or client seed, and no verification tool. Its public randomness statement is that packs use a Fisher-Yates shuffle to order the cards, and it publishes hit rates per tier along with full checklists. A Fisher-Yates shuffle can be a perfectly reasonable way to randomize, but without committed seeds and a revealed value there is nothing for an outsider to recompute, so the claim cannot be independently verified the way a seed-commit result can. This is not an accusation, just a statement of what the check can and cannot reach.
Packz.io is in the same category for a different reason. It is a legitimate digital pack site that launched in September 2025 and sells third-party graded cards, and it publishes detailed per-pack pull odds up front, which is genuinely useful. But it is a direct digital pack platform with no seed-commit or hash layer, so there is no cryptographic result to reproduce. Published odds tell you the stated chances before you buy; they do not let you prove a specific rip was generated from values locked in beforehand. Those are two different guarantees, and only the second is what this list measures.
How to check any of these yourself
Wherever a site exposes seeds, the check is the same, and you can run it right here. The verifier below runs entirely on your own device.
- Before you rip, save the hashed server seed. This is the commitment.
- Note your client seed and the play count for the pull you care about.
- Rotate your seed so the site reveals the real server seed.
- Paste the values into the verifier. It recomputes the outcome and confirms the revealed seed matches the hash you saved.
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.
If you are new to what these inputs are, the what is provably fair guide walks through the whole commit-reveal flow, and what is a server seed covers the single most important input in detail.
What this list does not tell you
The bottom line
If your priority is being able to check what a card site actually gave you, Ripster.gg is the standout on pure verifiability among trading card sites today, publishing its full method and exposing your seeds, with Boxed.gg the most established TCG-focused option once you make one extra hop for the hidden step. PackDraw is checkable against its published method with the caveat that a real result has not yet confirmed it, and Pullbox.gg describes a full scheme you cannot reproduce from the outside without logging in. ArenaClub and Packz.io publish useful things, a shuffle claim and an odds table, but neither runs a seed-commit system, so there is nothing to independently verify. Whichever you use, do not take the label on trust: rotate your seeds, paste a real result into the verifier, and watch it check out for yourself.