Findings
What the analyzed market actually shows, measured at page load — a read on what we scanned, not the whole hobby. Claims below their minimum sample say NOT MEASURED instead of pretending.
Do asking prices sit above what cards actually sell for?
19% of confirmed-priced live listings ask MORE than their own confirmed fair value; the median listing sits 45% below it.
n=89 live listings priced off confirmed sales, this view
Method: Each listing's delivered ask against its own confirmed-sales fair value. Modeled values excluded — this claim is only about real sales.
How much of the valued board rests on models rather than sales?
52% of valued cards carry a MODELED value; 48% are priced off confirmed sales.
n=185 cards carrying any fair value, this view
Method: Every card's fmvBasis flag. The split is printed on each card too — this is the same field, aggregated.
Are parallel premiums measured from sales, or assumed from print runs?
0% of parallel premiums in view are MEASURED from sold prices; the rest lean on the print-run table.
n=165 cards carrying a parallel premium, this view
Method: Each card's premium flag: measured (from sold prices of the parallel vs base) or assumed (print-run table). Disagreements over 2× set the suspect flag.
How many cards trade often enough to exit when you want to?
76% of cards with measured sale velocity are THIN markets (velocity under 4/10) — the paper gain is real, the exit is not guaranteed.
n=185 cards with measured sale velocity, this view
Method: Sale velocity from each card's own valuation evidence. Cards with no velocity data are excluded, not counted as thin.
How much of the live board is already slabbed?
54% of the analyzed view is graded; raw and slabbed are different markets and are never comped against each other.
n=241 analyzed live listings, this view
Method: Each listing's parsed grading fields. The cross-grader comp rule this feeds is on the methodology page.
The formulas behind these fields are on research & methodology and in plain English on how we score.