Expected value is the average secondary-market value of the cards you would pull from a sealed product, weighted by the probability of pulling each card. EV is the long-run average — a single box can come in well above or below it.

Is EV the same as the box being a good buy?

No. EV ignores liquidity, the time and friction of selling singles, grading and shipping costs, and the opportunity cost of capital. A box with EV slightly above MSRP is rarely actually profitable to crack — the slot covers your cost only after you sell out the chase cards.

Why does EV change between sites?

Three reasons. First, the price source — TCGPlayer market vs. lowest listing vs. eBay sold differ by 5-20% on chase cards. Second, the pull-rate model — analytic (assume independent slots) vs. Monte Carlo (sample real booster sheets) can disagree on collector boosters. Third, what counts as a "hit" — some calculators include lands and commons at full price, others zero them.

How often does EV change?

Card prices update hourly here. EV is recomputed at most every 12 hours and only inside a 1-3 AM PST window so the recompute does not contend with daytime traffic. The "as of" stamp on each EV row reflects when its constituents were last priced.

Which calculator is most trustworthy for new sets?

For sets less than 30 days old, no calculator is reliable. Singles prices are still discovering equilibrium and supply is constrained. EV at release tends to overstate; wait for the post-prerelease 4-week mark before treating it as signal.

Sealed EV, explained

Published 2026-05-05

Expected value (EV) is the most-cited number on every TCG forum thread that starts with "is this box worth cracking?" — and the most misunderstood. This guide walks through what EV actually computes, where the data comes from, and the three pitfalls that make EV look more decisive than it is.

The basic formula

For a single booster pack, EV is the sum, across every possible card you might pull, of the card’s market price times the probability of pulling it. For a booster box, multiply by the number of packs, then add any guaranteed fixed-slot inserts (box toppers, buy-a-box promos, sample cards) at their market price.

EV(pack) = Σ_card P(pull card) · price(card)
EV(box) = packs · EV(pack) + price(box-toppers)

That formula hides three judgment calls: which probability table you use, which price you plug in, and which cards count.

Probability — pull rates are not what you think

Modern Magic and Pokémon boosters are not uniform random draws. A pack has sheets — common slot, uncommon slot, rare/mythic slot, foil slot, wildcard slot in collector boosters — and each sheet has its own card list and weighting. Some sheets are themselves split: the rare slot is, on most modern Magic sets, 1/8 mythic and 7/8 rare, but if you pulled a bonus-sheet card it replaced the rare entirely. Pokémon collector boosters add a guaranteed alt-art slot with its own ~15-card sub-pool.

Two implementations:

Analytic — treat each slot as an independent draw from its sheet. Fast, closed-form, accurate enough for play boosters where all slots are independent.
Monte Carlo — simulate millions of packs respecting inter-slot constraints (no duplicates within a pack, exclusive bonus-sheet slots, foil-replacement rules). Required for collector boosters and any set with bonus sheets — analytic over-counts there by 5-15%.

CCG Index publishes both. The leaderboard sort defaults to deterministic (analytic) for speed and Monte Carlo for any product where the two diverge by more than 5%. See methodology for the exact crossover threshold.

Price — which price do you use?

TCGPlayer market price is the canonical input here. It’s the most traded venue and the price reflects recent sales rather than current listings, which avoids the “one $2,000 listing on a $40 card” problem. For cards TCGPlayer is missing data on (very new printings, vintage out-of-print) we fall back to Mana Pool, then eBay sold-listing aggregates from the past 14 days.

Other calculators use lowest current listing or aggregator medians. Lowest listing systematically overstates EV on hard-to-move cards (the bottom of the listing stack is the most desperate seller, which is not the price you'll get when selling). Aggregator medians lag — a fresh chase card spikes on TCGPlayer 6-12 hours before the median catches up.

What counts — the “hit” question

EV is the sum of every card’s contribution. In practice you will not sell every common and uncommon — they're worth less than the shipping envelope. Some calculators mark cards under $1 as zero; others count them at face value. Both are defensible:

Counting bulk reflects what the cards are technically worth and is easier to defend mathematically. It’s the right number if you’re bulk-selling commons by the long-box.
Zeroing bulk reflects what most operators will actually recover and is a more honest answer to “should I crack this?”

CCG Index reports gross EV (counting bulk at market) and a "realizable" EV (zeroing cards under $0.50, deducting 10% for selling friction) side by side. Use whichever matches your actual sell channel.

The post-release curve

New-set EV is highest on prerelease day and decays for the next 4-8 weeks as supply enters the market. Watch out for the “week one trap” — the chase card is at peak demand against minimum supply, which yields an EV number that will almost certainly be wrong by week six. Sets that retain or grow EV after that 8-week window are the ones to consider for long-hold sealed positions, not the ones that look good at release.

Where to see live EV on this site

/research/ev — Magic sealed EV hub
/research/pokemon-pack-ev — Pokémon pack EV leaderboard
/research/browse-sealed — sortable browse by EV ratio (EV ÷ price)

FAQ

What is sealed EV?: Expected value is the average secondary-market value of the cards you would pull from a sealed product, weighted by the probability of pulling each card. EV is the long-run average — a single box can come in well above or below it.
Is EV the same as the box being a good buy?: No. EV ignores liquidity, the time and friction of selling singles, grading and shipping costs, and the opportunity cost of capital. A box with EV slightly above MSRP is rarely actually profitable to crack — the slot covers your cost only after you sell out the chase cards.
Why does EV change between sites?: Three reasons. First, the price source — TCGPlayer market vs. lowest listing vs. eBay sold differ by 5-20% on chase cards. Second, the pull-rate model — analytic (assume independent slots) vs. Monte Carlo (sample real booster sheets) can disagree on collector boosters. Third, what counts as a "hit" — some calculators include lands and commons at full price, others zero them.
How often does EV change?: Card prices update hourly here. EV is recomputed at most every 12 hours and only inside a 1-3 AM PST window so the recompute does not contend with daytime traffic. The "as of" stamp on each EV row reflects when its constituents were last priced.
Which calculator is most trustworthy for new sets?: For sets less than 30 days old, no calculator is reliable. Singles prices are still discovering equilibrium and supply is constrained. EV at release tends to overstate; wait for the post-prerelease 4-week mark before treating it as signal.