Methodology
How a valuation is built
How Graded Card Value derives estimates: exact-match comparables, evidence tiers, confidence levels, quick-sale and replacement models, and the limits of each number.
The sequence
1. Fix the identity
Card, franchise, set, number, year, language, edition, variation, parallel, grader, grade, label and certification number are resolved first. Any field below confidence threshold is escalated for manual confirmation.
2. Gather candidates
Market records for the same identity are collected. Each candidate keeps its own metadata: source, date, sale type, grader, grade and whether the transaction actually settled.
3. Rank by match quality
Exact comparables outrank close comparables, which outrank different-grade or different-grader records. Sold transactions outrank asking prices. Recent records outrank stale ones.
4. Derive the estimate
The final market value is derived from the strongest supported evidence available. Low and high estimates express the observed spread, not a marketing range.
5. Show the workings
Every valuation exposes the comparables that were included, and where possible the candidates that were excluded and the reason for exclusion.
Evidence tiers
Evidence is never blended without a label. Raw-card evidence and graded-card evidence are never mixed silently.
| Tier | Treatment |
|---|---|
| Exact comparable | Same card identity, variation, language, grader, grade and label. Sold. |
| Close comparable | Same card and grade, minor identity difference such as print run or stamp. |
| Different grader | Same card and numeric grade from another grading company. Labelled. |
| Different grade | Same card, adjacent grade. Used only as directional context. |
| Asking price | Active listing, not a settled sale. Never treated as a sale. |
| Guide value | Published guide figure with no transaction attached. |
| Population only | Census data. Informs scarcity, never price on its own. |
Derived figures
| Figure | Default model | Meaning |
|---|---|---|
| Quick-sale estimate | 80% of final market value | Approximate proceeds from a fast sale. |
| Replacement estimate | 110% of final market value | Approximate cost to reacquire. |
These defaults apply only where no different configured methodology is supplied for the card in question.
Limitations
- Thin markets produce wide ranges and low confidence. We show both rather than one number.
- A valuation is an estimate of likely market outcome, not an offer, appraisal or guarantee.
- Condition estimates on raw cards are estimates only and never a predicted grade.
- Evidence ages. Every valuation carries the date it was produced.
- Nothing here is financial advice.