Wildcard · Digital Packs · How it works

How digital pack odds actually work

Every platform publishes the numbers. Almost nobody multiplies them together.

Every digital pack platform publishes the three numbers you need to work out whether its packs are worth buying. Almost nobody multiplies them together. Here's the whole method — it takes about two minutes per pack, and you can run it on any platform, including ones we haven't covered.

1. A pack has exactly one outcome

This is the part that makes the math easy. A sealed hobby box has many independent chances — you might pull two autos and six refractors — so its value is a messy sum. A digital pack gives you one card from a published list. Those outcomes are mutually exclusive, which means the odds add up to 100% and the expected value is a single multiplication:

Expected value = (odds of each outcome × what that card is worth), all added up.

If a $100 pack offers a 1% shot at a $900 card, 4% at $260, 11% at $95, 22% at $42 and a 62% chance of a $14 base card, its expected value is $9.00 + $10.40 + $10.45 + $9.24 + $8.68 = about $48. That's before you've paid a single fee.

2. The average is not what will happen to you

That $48 average is dragged upward by the 1% card almost nobody gets. The more useful number is the median — what half of all packs return or less. In the example above, the 62% base-card outcome means the median pack returns about $14. The gap between $48 and $14 is the entire psychology of the product.

Because the outcome list is published and finite, this isn't an estimate. The full distribution is known exactly, which is a stronger claim than anyone can make about a sealed box.

3. Subtract the frictions before you compare

A card's sticker value isn't what you get. Two exits, and you choose after seeing the card:

4. Before anything else: is the "cash offer" actually cash?

This is the question we got wrong ourselves, and it matters more than any percentage on the page. Some platforms pay the buyback in site credit that cannot be withdrawn — it can only be spent on more packs. Read the terms before you read the rate.

If it pays real money, the buyback is usually fine, and this surprised us. A 90% offer sounds like a 10% penalty, but the honest comparison isn't against the card's sticker price — it's against what you'd net selling it yourself, about 87% after marketplace fees. On that comparison, taking the cash usually wins.

If it pays credit, the same 90% is a different animal entirely. You haven't been paid 90 cents on the dollar; you've been handed a token that only buys another go at a game with a house edge. Its worth isn't 90% — it's 90% of whatever a dollar spent there actually returns, and that dollar faces the same edge again, and again. A house that keeps even a third of each dollar turns a "90% buyback" into something worth well under half its face value. The cut isn't taken once. It's taken every time the credit goes back around.

That's also why a credit-only platform can advertise a generous-looking rate without giving much away: the money never leaves. If you can't withdraw it, the only exit that produces real money is having the card shipped and selling it yourself — so check what shipping costs before you buy. Some platforms don't publish that figure at all, which tells you something in itself.

4b. Then ask what the platform says your card is worth

Whatever the buyback pays in, it's calculated from a value the platform assigns. Ninety percent of an inflated number is a small payout. So the question is never "what's the buyback percentage" — it's "what does this platform say this card is worth, and what does that card actually sell for?" Check two or three against recent sold listings and you'll know quickly.

5. Now compute the house edge — and then ignore all of it

The house edge is what's left: 1 − (expected value ÷ pack price). Our $100 example returns about $48 of realizable value, so the house edge is roughly 52%. For scale, a slot machine keeps 5–10%, and a roulette wheel keeps 5.3%.

But here's the shortcut that makes the whole calculation optional. Unlike a sealed box, everything inside a digital pack is already for sale on the open market right now. You are never forced to gamble for a specific card — you can just buy it. So the only question that ever really matters is: what does the card I actually want cost outright? If the answer is less than the pack, the pack is entertainment, and it's worth deciding in advance how much entertainment you're buying.

6. A live feed of other people's pulls is not evidence

Most platforms show a running feed of recent hits, and it is the most persuasive thing on the site: real cards, real names, scrolling past every few seconds. It is tempting to treat it as a free audit — if the feed is full of good cards, the packs must be good. Before you do, find out which pulls the feed is allowed to show.

Arena Club's Hit Feed is the documented example. Per their own support pages, a card that a collector sells back through an Arena Club Offer does not appear in the Hit Feed — the feed shows cards that were hit and kept (Arena Club support: "If I accept an Arena Club Offer, will the card appear in the Hit Feed?"). None of that is hidden; it's written down in public, and there are perfectly ordinary product reasons to build it that way.

But look at what it does to the sample. Which pulls get sold back? Overwhelmingly the ones not worth keeping — and, as section 4 explained, especially the ones where the platform's assigned value sits above what the card really fetches, because those are exactly the pulls where taking the cash is the rational move. So the cards most likely to be filtered out of the feed are the disappointing ones. What's left over-represents the good outcomes.

This is survivorship bias, the same shape as judging a fund by the performance of the funds that didn't close. It doesn't require anyone to have done anything wrong; the filter does the work on its own.

We can tell you the direction of the bias, not its size. Measuring it would need the sell-back rate per tier, which no platform publishes and we have not estimated. So treat this as a reason to discount a feed, not as a number. The rule generalizes: any platform that offers an instant cash-out and also shows a public pull feed has this problem, whether or not it has documented it as clearly as Arena Club has. If you want to know what a pack really returns, the published odds and the median are the honest inputs. The feed is marketing, and it is downstream of a filter.

What we can't tell you

We can't verify that the published odds are the odds actually used. Nobody outside the company can. A sealed box's odds are printed by a manufacturer that has to stand behind them and can't change them after the box ships; a digital pack's odds are a webpage that can be edited at any time. We check them for internal consistency, compare stated card values against real sold prices, and watch the pages for silent changes — but that's scrutiny, not an audit. Treat any published pull rate as a claim, not a fact.

Where we stand

Wildcard doesn't sell packs and takes no commission, affiliate revenue or paid placement from any platform that does. We're not neutral about the math, but we have no stake in your answer — and we couldn't referee this honestly if we did. If you spot an error in our numbers, tell us and we'll correct it in public.

Which brings you back to the only question that survives all of this: what does the card you actually want cost right now? That's a question with an answer, and it doesn't involve any odds at all.

Find out what that card costs — and get told when it drops → See the packs we've scored →