AppraisrAppraisr
Accuracy

How accurate is the CryptoPunks model?

Every score comes from the newest chronological hold-out; the target sale is excluded from training, comparables, and its causal floor proxy.

Median accuracy (100−APE)
94.2%
Mean accuracy (100−APE)
91.2%
Within ±5%
44%
Within ±10%
72%
Within ±20%
91%
Bias
-2.1%

Accuracy by market tier

Floor, middle-market, and grail sales are scored separately so aggregate accuracy cannot hide a weak segment

TierSalesMedian accuracyWithin ±10%Bias
floor175294.5%76%-2.1%
middle8976.9%21%+1.3%
grail2075.4%20%-12.5%

Weekly median accuracy

Chronological hold-out accuracy by sale week

Error distribution

Signed hold-out error; positive means over-predicted

Scored hold-out sales

60 out-of-sample predictions frozen at evaluation time

ItemtypeTier / evidenceDateSaleModelErrorAccuracy
#9912Femalefloor · lowSep 18, 202635.00Ξ39.22Ξ+12.1%87.9%
#3637Femalefloor · lowSep 17, 202629.38Ξ28.30Ξ-3.7%96.3%
#1607Malefloor · highSep 17, 202633.50Ξ29.69Ξ-11.4%88.6%
#1521Malefloor · highSep 15, 202635.00Ξ36.27Ξ+3.6%96.4%
#8557Malefloor · highSep 11, 202632.33Ξ32.68Ξ+1.1%98.9%
#2167Malefloor · highSep 9, 202628.16Ξ28.35Ξ+0.7%99.3%
#8450Femalefloor · lowSep 7, 202629.90Ξ28.52Ξ-4.6%95.4%
#1020Malefloor · highSep 7, 202638.99Ξ36.52Ξ-6.3%93.7%
#3797Malefloor · highSep 5, 202630.50Ξ33.38Ξ+9.4%90.6%
#2167Malefloor · highSep 5, 202628.33Ξ28.86Ξ+1.9%98.1%
#7366Femalefloor · lowSep 5, 202628.42Ξ32.12Ξ+13.0%87.0%
#4023Malefloor · highSep 5, 202628.20Ξ31.28Ξ+10.9%89.1%
#1549Malefloor · highSep 4, 202680.50Ξ41.68Ξ-48.2%51.8%
#8375Malefloor · highSep 4, 202632.50Ξ34.59Ξ+6.4%93.6%
#1895Femalefloor · lowSep 3, 202630.50Ξ33.57Ξ+10.1%89.9%
#4457Malefloor · highSep 2, 202630.00Ξ33.38Ξ+11.3%88.7%
#3498Femalefloor · lowSep 1, 202634.30Ξ31.50Ξ-8.2%91.8%
#7358Malefloor · highSep 1, 202634.50Ξ32.27Ξ-6.5%93.5%
#2687Malefloor · highAug 31, 202634.99Ξ35.02Ξ+0.1%99.9%
#8154Femalefloor · lowAug 31, 202630.50Ξ32.92Ξ+7.9%92.1%
#3218Malefloor · highAug 29, 202632.40Ξ32.62Ξ+0.7%99.3%
#9700Malefloor · highAug 28, 202631.49Ξ32.49Ξ+3.2%96.8%
#9700Malefloor · highAug 28, 202632.49Ξ32.07Ξ-1.3%98.7%
#2403Femalefloor · lowAug 27, 202631.79Ξ31.42Ξ-1.2%98.8%
#8607Malefloor · highAug 26, 202632.99Ξ32.14Ξ-2.6%97.4%
#8480Malefloor · highAug 25, 202637.49Ξ32.05Ξ-14.5%85.5%
#9799Malegrail · highAug 24, 2026117.0Ξ112.5Ξ-3.9%96.2%
#6570Femalefloor · lowAug 24, 202634.48Ξ31.81Ξ-7.8%92.2%
#5541Femalemiddle · lowAug 23, 202650.50Ξ66.31Ξ+31.3%68.7%
#3862Malefloor · highAug 23, 202630.80Ξ33.12Ξ+7.5%92.5%
#5151Malefloor · highAug 21, 202638.50Ξ37.32Ξ-3.1%96.9%
#9018Malefloor · highAug 21, 202636.95Ξ32.77Ξ-11.3%88.7%
#1326Malemiddle · highAug 21, 202651.00Ξ50.07Ξ-1.8%98.2%
#4711Femalefloor · lowAug 20, 202632.41Ξ30.50Ξ-5.9%94.1%
#4164Malefloor · highAug 20, 202635.60Ξ33.11Ξ-7.0%93.0%
#9546Femalefloor · lowAug 20, 202640.00Ξ41.25Ξ+3.1%96.9%
#5917Femalefloor · lowAug 19, 202631.45Ξ30.88Ξ-1.8%98.2%
#1930Femalefloor · lowAug 19, 202630.50Ξ31.11Ξ+2.0%98.0%
#9335Malefloor · highAug 19, 202633.50Ξ32.48Ξ-3.0%97.0%
#3298Malemiddle · highAug 17, 2026140.0Ξ84.22Ξ-39.8%60.2%

Model cryptopunks-1789697843-2026-08-03.2 · evaluated through 9/18/2026. Method details on the methodology page.