Understanding our data & methodology

In Value & Analytics

CollectorWorth’s mission is to bring transparency and accuracy to collectible pricing. Our market values are based on real sales from trusted marketplaces — not guesses, not estimates. Here’s how we do it.

Our Philosophy: True Market Value

We don’t believe in relying on a single sale to define value. Instead, we aggregate and clean data from multiple sources to calculate what a collector would reasonably expect to pay for an item today.

Step 1: Aggregating Sales Data

Our system pulls millions of completed listings every week from the most trusted marketplaces:

  • eBay (completed/sold listings)
  • TCGplayer (card market sales)
  • Specialty marketplaces and stores across TCG and Funko Pop niches

For each listing, we capture details such as name, price, sale date, grade, condition, and shipping.

Diagram of data aggregation from various marketplaces
Live sales data powers our values.

Step 2: Filtering & Cleaning Raw Data

Sales data is noisy. We apply a proprietary algorithm to clean listings before using them in calculations:

  • Lot Removal: Filters out multi-item listings (e.g., “Charizard + 50 other cards”)
  • Outlier Removal: Detects and removes price outliers using statistical analysis
  • Shipping Normalization: Removes inflated shipping that skews prices
  • Condition Matching: Ensures accurate comparisons across grades (e.g., PSA 10, Ungraded)
  • Offer Filtering: Excludes listings sold via “Best Offer” without a known final price
Note: Values reflect items in NM condition or the specific grade shown — not a mix of all conditions.

Step 3: Calculating Market Value

  • Weighted Average: Recent sales carry more weight than older ones
  • Daily Updates: Our models re-calculate values every 24 hours
  • Price Signal Count: Shows how many valid sales contribute to the current value
Screenshot showing price signal count and value trends
Values update daily based on real collector activity.

Special Cases: New or Rare Items

Newly released or extremely rare collectibles often have limited sales history. These values may fluctuate more until we collect enough high-confidence data. We’re constantly working on predictive improvements for these edge cases.

Have questions about our data model? Get in touch — we’re happy to walk you through it.