Mystery pack ROI — the formula, and the mistake in it

Return on investment for a mystery pack is one short formula: (cash-out value − pack price) ÷ pack price. The arithmetic is trivial. Nearly every wrong ROI figure in this hobby comes from the same place — putting the wrong number into the first slot.

The formula, and the word that does all the work

The word is cash-out. A pull's fair market value is not money — it is Courtyard's algorithmic estimate of what the graded card is worth. What converts to money is the instant buyback, and that pays a measured 84.6% of FMV, verified across tens of thousands of recorded offers in the buyback guide.

So the formula in practice is:

Why counting FMV as cash overstates your return

Run the same pack both ways and the gap is stark. Pokémon Starter Pack carries 6,087 indexed pulls, a $24.72 expected value against a $25 price, and a measured 0.885 buyback ratio.

Measured at FMV the pack looks like a −1.9% proposition — near enough to break-even that most buyers would call it a fair price for the entertainment. Measured in money it is −12.3%. The pack did not change between those two lines; only the unit did.

That ten-point gap is the single most common error in mystery pack accounting, and it compounds silently. A buyer averaging four packs a week at a genuine −12% while believing they are running at −2% will conclude, after a month, that they have been unlucky. They have not been unlucky. They have been using a catalogue value as a bank balance.

Where shipping and import tax belong in the calculation

Only if you redeem. Most pulls are instant-sell outcomes — on a pack whose median card is worth $21.9, paying $16 to ship it home is obviously irrational — so the honest default is to compute ROI on the buyback path and treat physical redemption as a separate decision with its own arithmetic.

When you do redeem, the costs are real and they stack. US redemption shipping commonly lands near $16 per shipment, quoted at FedEx cost from Courtyard's vault in Delaware. Outside the US, import tax applies on top — 20% in the United Kingdom, 25% in Sweden, 10% in Australia, with most European countries offering no tax-free threshold at all. The shipping and redemption guide carries the full country table and the breakeven rule.

The practical consequence for ROI: redemption improves your return only on high-value pulls, because shipping is close to a flat cost while the buyback haircut scales with the card. A $400 card loses about $62 to the buyback and about $16 to shipping — redemption wins comfortably. A $30 card loses about $4.60 to the buyback and the same $16 to shipping — redemption loses badly. The crossover sits far higher than most people assume.

What realistic ROI looks like across the catalogue

Applying the correct formula to every Courtyard pack with a known price and at least thirty indexed pulls produces a distribution rather than a single verdict.

The average priced pack returns about -11.2% per rip. The best currently measured returns +41.2%; the worst returns -27.0%. That spread is the actual opportunity: the difference between the best and worst pack on the same platform, on the same evening, is more than sixty percentage points of expected return.

Three ways ROI gets quietly overstated

1. Counting FMV as cash. The big one, worth about fifteen points. Covered above and in the buyback guide.

2. Ignoring the mean-median gap. ROI is an average, and averages in long-tailed pools are carried by rare hits. A positive-ROI pack can still deliver a losing median rip — see why average EV lies. ROI answers “should I repeat this bet?”, never “what happens tonight?”

3. Remembering the wins. No formula protects against selective memory. The fix is recording every rip rather than the memorable ones, which is what the decision journal and portfolio tracker exist for — they compute realised ROI from what you actually logged, and compare it against what the pack's numbers predicted at the time you bought.

ROI over many rips, not one

A formula that averages is only meaningful over repetition, and this is where most published ROI figures quietly mislead. Buy one pack and you take a single draw from a long-tailed distribution: the likely result is the median, which on most Courtyard packs sits well below the price, while the ROI figure describes an average that only emerges over dozens of rips.

Concretely: a pack with a −5% ROI and a 25% hit rate will, over twenty rips, most often deliver fifteen losses and five wins, with the arithmetic landing near that −5% only because one of the wins was large. Over four rips it will usually deliver four losses and an experience that feels nothing like −5%. Neither run contradicts the number. Both are what a −5% pack looks like at different sample sizes.

The practical consequence is that ROI should govern pack selection, never bankroll expectations. Use it to decide which pack to buy; use the median and hit rate to decide what tonight will probably feel like, and how many rips you would need before the average means anything at all.

Using ROI to decide rather than to justify

The useful discipline is to compute ROI before the rip and write it down. A pack's cash-out ratio, median pull and hit rate are all visible in advance on its rankings entry; a decision made against those numbers is auditable later, while a decision made on enthusiasm is not.

And when the ROI is negative — which, four packs aside, it currently is — that is not automatically a reason to walk away. A −5% pack with genuine chase odds is entertainment at a known and modest price, which is a perfectly rational purchase. A −30% pack is entertainment at six times that price. The number does not tell you what to want; it tells you what it costs, which is the only thing a formula can honestly do.