MTGBAN Guide

F.A.Q.

• Understanding EV and SIM at BAN

  BAN gives you two complementary lenses for estimating the value of sealed MTG product: Expected Value (EV) and Simulated Box Openings (SIM).
  
  For products with a fixed decklist, the two methods produce identical results; there is nothing to randomize, so we simply sum the card values. For products with built-in randomness (booster packs and certain decks), the two numbers can diverge significantly, especially over a small number of openings.
  
  Looking at both helps you weigh the theoretical value of a product against the realistic spread you'd see if you actually opened it.

• How is EV calculated?

  EV is a weighted average. We take the print sheet distribution for each card, multiply each card's drop rate by its market value, and sum the results.
  
  The output is the theoretical long-run value of one unit of product: the number you would converge toward if you opened an effectively infinite quantity. Because each card contributes linearly based on its probability of appearing in a slot, EV is straightforward but blind to variance.

• How is SIM calculated?

  SIM models the actual experience of opening product. Using the same per-card probabilities that drive EV, we generate a randomized box, total its value, and repeat the process thousands of times.
  
  The resulting distribution surfaces information that a single average can't, most usefully the median and standard deviation.
  
  SIM is not run for preconstructed decks or any other product where the contents are deterministic.

• What do the median and standard deviation represent?

  The median is the middle value of the simulated dataset. If we run 5,000 box simulations and sort them from lowest to highest, the median is the value of the 2,500th box.
  
  On its own this number isn't especially meaningful, but comparing it to the EV is. When the median sits well below the EV, the average is being pulled up by a small number of high-value chase cards, meaning most boxes you actually open will fall short of EV and the upside is concentrated in rare hits.
  
  The standard deviation measures how spread out the box values are. A low standard deviation means consistent, predictable openings: if EV is $100 and standard deviation is $1, your boxes will almost all land near $100. A standard deviation of $40 on the same $100 EV means swings in either direction are common.
  
  The more boxes you open, the more your realized average will tend toward EV, but how much volume it takes to get there depends entirely on the product.

• Which one should I follow?

  Both, in combination. EV gives you the theoretical baseline; SIM tells you what variance you should brace for.
  
  The right emphasis depends on your situation. Large-volume openers can lean on EV, since over months and years their realized returns will track it closely.
  
  Smaller operators may want to avoid high-variance products even when they're technically +EV; a single bad run can hurt enough to outweigh the long-run expectation.

• Why are some products tagged with high IQR?

  The interquartile range (IQR) is a measure of how much the price data within a product deviates from the typical range. A high IQR means the product contains a meaningful number of outlier cards, usually a small set of very expensive chases relative to everything else.
  
  Hits on those cards are statistically rare, but when they do land in a simulation they can pull the prospect value up disproportionately, even across thousands of runs.
  
  We tag these products so users know the numbers are more sensitive to luck and data quality than usual, and we exclude them from certain aggregations where appropriate. The threshold scales with current market conditions, so it accounts for both older sealed product (where individual cards run into the thousands) and recent product with a handful of chase cards. Serialized cards are not included in any of our computations.
  
  Always understand the underlying data before making a financial decision, and apply extra scrutiny to anything tagged high IQR.

• How do you emulate booster openings?

  We pull from several open-source projects. Collation data comes from the Lethe project, in-booster card distributions come from the mtg.wtf search engine, sealed product contents come from MTGJSON, and decklists come from this repository.
  
  Each project's site has more detail on how its estimates work, and all of them welcome contributions.
  
  If you want to look at the BAN side specifically, the estimation model lives in api.go and the actual computation lives in sealedev.go.

• What data sources are available?

  Three price sources feed the EV and SIM calculations. TCG Low reflects the lowest available price on TCGplayer, excluding shipping. TCG Direct represents the Direct price net of fees, i.e. what you would actually realize selling into that channel. CK Buylist shows the price Card Kingdom is paying for the cards.
  
  TCG Direct can occasionally be skewed by sellers parking inventory at absurd prices; we cap values at twice TCG Low when we can detect this kind of distortion.
  
  The prices listed under Purchase from and Sell to are the sealed product prices themselves, sourced from the relevant stores.

• Why is this set missing?

  Because we depend on three upstream data sources, missing data anywhere in the chain blocks a set or product from showing up.
  
  Most of the time the information just hasn't been filled in yet and hasn't propagated to us. Try again in a few days, or, if you're able to code, contribute to one of the open-source projects above.

• What's going on with the prices for a just-released set?

  Our prices reflect the new printing of each card, not the value the card held before the reprint. It takes a few days for listings to populate and for prices to settle into the calculator.
  
  Direct prices in particular can look off until the full TCG Direct inventory shows up; in the meantime we substitute the market price as a best estimate.
  
  Also keep in mind that early-window numbers reflect preorder prices, which almost always run higher than the prices a set settles into post-release.

• How is Set Value computed, and why does it differ from buylist?

  Set Value is the daily sum of every printing of every card in a set, including foil and nonfoil where both exist. It represents the value of a complete NM collection of the set.
  
  It contains no random element; that's the domain of the SIM calculations on individual products.

• What are the custom buttons next to product titles for?

  They open different views of a product:
  * 🎁 (when available) simulates opening a pack from the underlying data. Each run is fully randomized.
  * 🔎 unpacks a product and lists every possible piece of content. A booster box, for example, will list a single booster pack; a booster pack will list every possible card it can contain.
  * (when available) downloads the decklist as a TCGplayer-formatted CSV, ready for upload.
  
  Once you're on the search page, the standard search syntax works as normal: you can filter by color, rarity, name, and so on.
  
  Depending on the content type, results may show one copy per card or list the same card multiple times. This is intentional: decklists and complete box sets need accurate counts for import, while other sealed types are easier to read in a deduplicated view.

• This product has multiple copies of the same card, but I only see one. What gives?

  To keep results readable, we hide duplicate listings of the same card in the search view.
  
  The EV and SIM calculations do still account for the possibility of opening multiple copies; the deduplication is purely a display choice.

• Is there a minimum or maximum card value that gets considered?

  Yes. We initially included every price we could find, but counting both worthless and extraordinarily expensive cards distorted the results across runs.
  
  We now exclude bulk (under $0.30) and serialized cards, since serialized cards are too rare to open in any meaningful sample but skew the math when they do land. This applies to both EV and SIM.
  
  For everything else, the variance from ordinary cards is handled naturally by repeated SIM runs.

Disclaimer

These EV calculations are estimates and should not be used as the sole basis for financial decisions.