Introduction
Expected value, or EV, is the average result across possible outcomes, weighted by how likely each outcome is.
You can use EV to compare repacks. It depends on the values and probabilities used in the calculation and does not predict a single opening.
Calculating expected value
Multiply each possible outcome's value by its probability, then add the results.
EV = (Probability of Outcome 1 × Value of Outcome 1) + (Probability of Outcome 2 × Value of Outcome 2) + ... for every possible outcome
Subtract the cost to calculate net expected value:
Net EV = (Sum of all probability-weighted outcomes) − Cost
Positive net EV means an average gain across repeated outcomes. Negative net EV means an average loss.
Example 1: A coin flip
Say a friend offers you a bet: flip a coin, and if it lands on heads, you win $150. If it lands on tails, you lose your $100 stake.
- Probability of heads: 50%, value: +$150
- Probability of tails: 50%, value: −$100
EV = (0.5 × 150) + (0.5 × -100) = 75 − 50 = +$25
The average result is +$25 per flip, but a single flip can still lose $100.
Example 2: Roulette
A single-number bet on American roulette pays 35 to 1. The wheel has 38 numbers, including 0 and 00.
Your probability of winning is 1 in 38. If you bet $1:
EV = (1/38 × $35) + (37/38 × -$1) = 0.921 − 0.974 = −$0.053
The expected loss is about 5.3 cents for each dollar wagered. Previous spins do not improve the odds of the next one.
Example 3: Insurance
An insurer estimates claim costs and adds overhead and profit margin. A buyer may accept a cost above the expected payout to protect against a large loss.
Example 4: Poker
A poker decision can have positive EV and still lose a particular hand. One result does not establish whether the decision was sound.
EV and risk
Two choices can have the same EV and different risks.
A guaranteed $50 and a 50% chance of either $0 or $100 both have EV of $50. They differ in variance, which is the spread of possible outcomes around the average.
For repacks, compare the range of outcomes and what you can afford to lose as well as EV.
EV in sports card repacks
A repack is a sealed pack of cards assembled by a seller and sold at a fixed price. Its EV depends on the value and probability of each possible outcome.
To estimate a repack’s market-value EV, use its full checklist, the probability of each card, and each card’s current market value.
EV = Σ (Probability of pulling card X × Market value of card X) − Cost of the repack
Worked example
This example groups a $100 repack’s 500 possible cards into four tiers:
| Tier | Odds | Cards in Tier | Value per Card | Contribution to EV |
|---|---|---|---|---|
| Common | 90% | 450 | $2 | $1.80 |
| Mid-tier hit | 8% | 40 | $25 | $2.00 |
| Chase card | 1.8% | 9 | $150 | $2.70 |
| Grail card | 0.2% | 1 | $2,000 | $4.00 |
Add up the contribution column: $1.80 + $2.00 + $2.70 + $4.00 = $10.50 in expected value per pack, before subtracting the $100 cost.
Net EV = $10.50 − $100 = −$89.50
The example has negative net EV. A seller's costs and profit margin affect its pricing, so a negative estimate alone does not establish that a listing is misleading.
Checking the inputs
Repack EV depends on the quality and age of its inventory, odds, and card values.
Checklist accuracy matters. Check whether the list updates as cards are pulled and whether it describes the inventory available now.
Card prices change. Use current, comparable sales when making a market-value estimate.
Resale value is not the amount you keep after grading, shipping, marketplace fees, and taxes. Account for those costs when estimating a net return.
Without probabilities for each outcome, you cannot calculate a complete EV.
Comparing repacks
Use an EV estimate to compare the price with the possible returns.
A $100 pack with EV of $80 differs from a $100 pack with EV of $15, even if both advertise the same chase.
Compare estimates that use the same valuation method. A market-value estimate and a guaranteed buyback estimate measure different amounts.
Treat a repack purchase as entertainment spending. EV describes an average across outcomes, not what you will receive.
How PackScout calculates EV
The example above uses market values. PackScout EV uses the platform's stated values and guaranteed buyback terms instead.
PackScout starts from each outcome's platform-stated value and probability.
Apply each outcome's documented buyback terms to its stated value. A product-wide rate applies only when the same rate covers every eligible outcome.
Apply mandatory fees, caps, floors, and fixed offers in the documented payout order. An explicitly ineligible outcome contributes a $0.00 payout. A final buyback payout is never discounted a second time.
EV $ is the probability-weighted sum of those final guaranteed payouts, multiplied across a pack’s approved number of draws. With one documented uniform rate, $1.00 stated Outcome EV × 96% buyback = $0.96 EV $.
PackScout metrics
EV $ is the expected buyback payout. EV % is EV $ divided by Pack Price, multiplied by 100. A result of 100% matches the price.
EV % of 100.00% is break-even, when EV $ equals Pack Price. In the example, a $1.00 pack has EV $0.96 and EV 96.00%.
- Pack Price
- The listed price to open one pack, before any discounts or promo codes..
- EV $
- The average guaranteed buyback value of a pack. For supported platform reports, EV $ is reported underlying EV × Buyback %..
- EV %
- EV $ as a percentage of Pack Price. 100% matches the price..
- Trust Score
- How complete and up to date the pack data is. Not financial advice..
- Buyback %
- The share of collectible value the vendor pays when you sell it back. A single rate applies to every outcome..
- Vendor-reported EV
- The platform's EV estimate, adjusted by PackScout using the buyback rate for supported reports. Independent estimates stay separate..
- Top Chase
- The most valuable collectible PackScout has matched to this repack..
Buyback examples
These examples are hypothetical and do not describe current listings.
Odds and stated values
For a finite pool, current remaining-inventory odds take priority when the platform provides a complete list. Complete platform-published odds are the fallback.
The odds must cover every outcome and describe the same observation time.
PackScout uses exact stated values where possible. A closed platform range uses its midpoint.
Trust Score and freshness
Every pack in the catalog gets a trust score, whether or not PackScout can estimate its EV. The trust score describes how complete and how recent the data PackScout collected for that pack is. The published bands are Low (below 50%), Medium (50% to under 80%), and High (80% and above).
How complete and up to date the pack data is. Not financial advice.
The score starts from a checklist of the pack’s data. Each item PackScout collected adds its share: a current pack price (10%), complete odds (25%), values for every outcome (20%), buyback terms (15%), the cards in the pack’s pool (15% when the known list is complete, 8% when only part of it is known), chase cards (5%), and at least one pull recorded in the last 7 days (10%). A pack with every item collected starts at 100%.
Freshness then scales that checklist score. Data collected within the last hour keeps the full score. After that, the trust score falls in a straight line and reaches zero 3 days (72 hours) after collection unless fresh data arrives. The shown score is the checklist score multiplied by freshness, rounded down, so a refresh of the page never raises it.
Trust Score measures collected data independently of EV. Last-known EV remains visible and available packs stay eligible for rankings as data ages. A sold-out repack retains its last valid estimate and stays out of opportunities.
A Last-known estimate keeps its original calculation-time price. If the listing price changes, check the current listing before comparing that estimate with other packs.
Recent pulls
Recent pulls change EV only when they update the remaining inventory. Removed items leave the pool and the odds recalculate from what remains.
A hot or cold streak does not estimate future odds. PackScout does not infer EV from recent hit rates.
A pull recorded in the last 7 days counts toward the trust score as evidence of recent activity. Its result does not predict the next opening.
Platform-reported EV
Phygitals, Courtyard, and Collector Crypt report an underlying EV using their own inventory, odds, and valuation approach. When the platform supplies a USD EV, a USD pack price, and one uniform buyback rate, PackScout displays platform-derived EV $ as reported underlying EV × buyback rate, in place of its own estimate from the published odds. EV % compares that result with the pack price. Each pack shows one EV.
For Underdog, PackScout measures what paid pulls actually returned. EV $ is the average card value a pull returned per dollar paid over the last 7 days, applied to the current pack price and multiplied by the buyback rate. Once a pack has had a paid pull, observed EV is the only EV it shows; when the latest week has none, or the weekly report is overdue, it shows the last observed value marked last known.
For supported platform reports, PackScout multiplies EV by the uniform buyback rate. Independent estimates are calculated separately.
For example, a $100 pack with reported underlying EV of $104.21 and a 90% buyback has EV $93.79 and EV % 93.79%. This remains a platform-derived estimate. The pack’s trust score still comes from its collected data. A final buyback payout is never discounted again.
Costs outside the estimate
PackScout EV compares guaranteed buyback payouts. It excludes the following costs and adjustments:
Liquidity and resale friction
The time and difficulty of selling an item are not included.
Shipping, resale fees, and taxes
No shipping costs, marketplace or resale fees, or taxes are subtracted from any payout.
Personalized prices
Pack Price is the current public listed price before personalized, membership, or promo discounts.
Unsupported currencies and live FX
Calculations use USD or a documented USD-equivalent. PackScout does not apply live FX conversion.
Independent market valuation
PackScout does not independently value collectibles, use external sales comps, or apply a proprietary valuation model at launch.
Compare more than EV
Repacks with similar EV can have different odds, inventory, and buyback terms. Review the vendor listing alongside the estimate. PackScout does not independently verify every underlying data point.