Pullbox

Load a balance, pick a box, open it, and one card comes out. From there you ship it, sell it back into your balance, or list it on the Exchange. Boxes span Pokémon, Magic, Lorcana and One Piece, and prices run from well under a dollar to the tens of dollars.

Around the boxes sits a layer of case-site mechanics: Battles and Brawls (players opening head-to-head, winner takes the cards), Contracts (trade several cards for one of higher value), daily rewards, a coin drop, and a leaderboard. There is also a published provably-fair page describing how outcomes are generated.

This is Pullbox's strongest suit. Open any box page and it lists the full contents: every card, the value Pullbox puts on it, and the exact chance of pulling it. Nothing is bucketed into value bands, and nothing is hidden behind a rarity label.

We verified the arithmetic on the boxes we could read in full, and the drop odds sum to exactly 100.00% — no missing probability mass, no unlisted remainder. Plenty of sites publish odds that quietly fail that test.

Because values sit next to odds, the expected return is arithmetic rather than a model. On Indigo Plateau — $10.35 to open, 38 cards — the expected value works out to $9.11, or about 88%. The top card is a Blaine's Charizard at $782.81 on 0.07% odds; the floor is an Erika's Oddish at $0.72, which comes up nearly 17% of the time. That shape is the honest picture of a case box: most opens return a card worth a fraction of the price, and the appeal rides on a thin tail.

Every figure above is Pullbox's own. It picks the cards, sets the values that the odds are calculated against, and those same values are what a sell-back is measured against. Nothing is independently priced.

There is a live win ticker running down the homepage with real card names, values and timestamps, which looks like the missing check — but it does not appear to show ordinary pulls. Over about 45 minutes the cheapest win it displayed was $21.55, while the box we examined floors at $0.72 on 17% odds. A feed that only broadcasts the good outcomes cannot measure what a box returns, and it makes the site look considerably luckier than it is.

That is why Pullbox is a review here rather than a ranked brand: we rank packs on the median return that real pulls delivered, and there is no complete public record of pulls to work from.

It does several things a site trying to hide the maths would not. It publishes the entire contents of every box with per-card probabilities, those probabilities add up correctly, and it runs a provably-fair page explaining how draws are generated.

The counterweight is the framing. Battles, Brawls, Contracts, coin drops, daily rewards and affiliate codes traded in chat are casino furniture, not card-collecting furniture. That is a matter of taste rather than honesty, but it shapes how easy the site makes it to keep opening.

Practical gaps: the sell-back rate is not published anywhere we could find, and card values have no third-party source behind them.

Read the box contents first — they are right there, in full, which is rare. Look at where the probability actually sits rather than at the headline card: in most boxes two or three cheap cards carry the bulk of the odds.

Do the multiplication before you spend. Values and percentages on the same page mean you can work out the expected return yourself in a minute, and it is the single most useful thing you can do here.

Ship the cards you actually want. Selling back turns a real card into site balance at Pullbox's own valuation, which is the point where you depend most on numbers nobody outside the site sets.

And treat Battles, Contracts and the coin drop as what they are — mechanics designed to keep you opening, not ways to improve your return.