Skip to main content

METHODOLOGY

How we examine the evidence.

A useful comparison begins with the same player, year, brand and set, parallel, card number, autograph or memorabilia variant, grade, grading company, and league. Our catalog keeps those identity fields separate, merges exact duplicates, and preserves aliases. Missing identity fields are shown as missing.

The public catalog uses existing Card Hedge comp summaries. A provider-reported comp median is shown separately from an asking price or estimated value. USD is the existing source presentation; no currency conversion is applied. Invalid or non-positive prices are unavailable.

The latest reported sale date belongs to the observation. Snapshot publication records when an export was produced. Last successful ingestion and an exact comp sample window are left unknown when they were not recorded. Rebuilding the site never replaces the sale date.

For research, observations older than 30 days are historical. Recent evidence with fewer than three valid comp prices, or an unknown count, is limited. A recent observation does not itself establish a current buying recommendation. Missing individual timestamps mean no transaction chart. Provider summary counts are not silently converted into annual sales.

Allocation uses the existing deterministic graded-card and direct-comp rules, with a $250 median-price and capacity minimum. Core grails require a liquidity score of 7, at least 4 reported 90-day sales and 24 recent comps, and a latest sale no older than 90 days. Established stars use 8, 6, 36 and 60 days. Speculative cards use 9, 10, 50 and 45 days. The existing durable annual path requires 24 annual observations, 24 recent comps and a latest sale within 90 days. Saved approvals are rechecked against actual time. Anthony Richardson and ambiguous or patch-dependent variants remain excluded.

Observed sales activity describes the available sample. It does not guarantee an exit, a waiting buyer or a particular realized price. Allocation eligibility, freshness and editorial interest are separate. The portfolio tool may leave money unallocated when hard constraints cannot be met.

The selling-cost calculator uses integer cents and hundredths of a percentage point. Percentage fees round to the nearest cent; break-even rounds upward. The model assumes fees apply only to the sale price. Income taxes and costs not entered are excluded.

A failed or incomplete refresh must preserve the last valid snapshot and its original dates. The research interface qualifies that history instead of pretending it is a current update.

Marketplace search links are labeled as searches and do not verify individual listings. Affiliate relationships, when used, are disclosed beside the link and in our disclosures. Editorial comparisons are not ranked by commission.