Optimizer Scoring Roadmap

Build the desirability index as an upgrade path, not a black box.

The scoring path starts with explainable print recommendations, then adds live market movement, exact-print scarcity, collector premium, and risk gates until each card recommendation can say what changed and why it matters.

Formula Target
0-100 printing score
Mythiverse Exchange shield
Card demand
generic pull
Exact printing
version fit
Market timing
movement
Risk filter
wait or buy
Useful signal

Find printings becoming more desirable before price alone makes them obvious.

Score Stack

Start explainable, then tune with outcomes.

The index is the output. Each input gets normalized first, weighted second, and backtested later against real sale and price movement.

Demand

35%

Card-level pull from EDHREC rank, inclusion count, commander breadth, growth, staple status, and format relevance.

Market

25%

Live buying pressure from price trend, sales velocity, recent sold volume, listing depth, buylist price, and spread.

Scarcity

20%

Printing-specific supply pressure from set age, set sales, rarity, printings, sealed availability, and finish supply.

Collector

10%

Premium appeal from foiling type, frame, first printing, artist, lore, serialized status, old border, and condition sensitivity.

Risk

-10%

Penalty for reprint risk, ban risk, narrowness, volatility, hype decay, excess variants, and known finish quality issues.

Upgrade Path

Six milestones from simple scoring to predictive guidance.

Stage 01

Baseline Score

MVP

Separate card-level demand from printing-level desirability and normalize every input to a 0-100 range.

Inputs
EDHREC rankcurrent pricerarityrelease datefinish
A stable first-pass score that explains why a printing appears in the optimizer tree.
Stage 02

Market Momentum

Next

Connect historical prices, sold volume, listing count, and buylist spread so the score responds to real movement.

Inputs
30/90/365 day trendsales velocitylisting countbuylist spread
Early flags for printings getting tighter before the sticker price fully reacts.
Stage 03

Printing Scarcity

Next

Model exact-print supply instead of treating every version of a card as interchangeable.

Inputs
set salesset agefoil supplyvariant countsealed availability
Different scores for first printings, premium finishes, reprint versions, and mass-supply variants.
Stage 04

Collector Premium

Planned

Layer in the emotional reasons a player chooses this copy: art, frame, foil treatment, nostalgia, and table identity.

Inputs
foiling typeframe treatmentartistlore relevancecondition sensitivity
A version-aware premium score that supports style upgrade recommendations.
Stage 05

Risk Gates

Planned

Subtract the reasons a signal can break: reprints, bans, power creep, single-deck demand, and unsustainable hype.

Inputs
reprint riskban riskvolatilityformat dependencyhype spike age
Watchlist and wait signals that keep recommendations from chasing fragile spikes.
Stage 06

Backtest And Learn

Model

Compare historical scores to later price movement and sales outcomes, then tune weights with real misses.

Inputs
score snapshotsfuture salesfuture pricemanual labelsmiss reasons
A tuned desirability model that predicts opportunity quality, not just current price.
Signal Map

Keep card demand and print desirability separate.

Rhystic Study can be a high-demand card while each printing earns a different score. The model should preserve that difference so the optimizer can explain budget picks, style upgrades, and watchlist calls.

Card-Level Demand

Answers whether people want the card at all.

  • EDHREC rank and inclusion count
  • EDHREC 30/90 day growth
  • Commander and archetype breadth
  • Staple status versus narrow synergy
  • cEDH and constructed relevance

Printing-Level Supply

Answers whether this exact copy is meaningfully harder to find.

  • Set age and estimated set sales
  • Number of printings and variants
  • Time since last reprint
  • Foil, etched, surge, textured, or showcase supply
  • Sealed product still available

Market Liquidity

Answers whether the market is actually moving.

  • Sales velocity and recent sold volume
  • TCGplayer listing depth
  • Buylist price and buylist spread
  • Price versus all-time high
  • Volatility and trend stability

Collector Fit

Answers why a player would choose this version over another.

  • Foiling type and frame treatment
  • First printing, promo, or serialized status
  • Artist and character relevance
  • Old border or retro appeal
  • Condition sensitivity for older cards
Data Contract

Store the score at printing level.

oracle_idcard_nameset_codecollector_numberfinishtreatmentrarityrelease_datemarket_priceprice_90d_changelisting_countsales_velocityedhrec_rankprinting_countlast_reprint_daterisk_scoredesirability_index
Validation Rules

Backtest against behavior, not vibes.

  • Known EDH staples should rank high even when the cheapest printing is inexpensive.
  • Premium variants should not outrank playable copies when demand is narrow and supply is deep.
  • Cards with fast sales and falling listings should surface before price alone makes them obvious.
  • Recent hype spikes should move to watchlist when volatility is high and demand is concentrated.
Desirability scores are hobby guidance for ranking printings inside DeckSwap. They are not financial advice, investment advice, or a guarantee of future card value.

Privacy choices

DeckSwap uses essential cookies for sign-in and can use optional analytics to understand product performance. Analytics stays off unless you opt in. See Privacy Policy.