AppraisrAppraisr
Accuracy

How accurate is the rektguy 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)
92.8%
Mean accuracy (100−APE)
90.1%
Within ±5%
35%
Within ±10%
66%
Within ±20%
89%
Bias
-1.9%

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
floor161092.9%68%-2.1%
middle3567.5%14%+4.2%
grail558.6%0%+17.2%

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

ItemTier / evidenceDateSaleModelErrorAccuracy
#2982floor · highSep 18, 20260.130Ξ0.170Ξ+30.8%69.2%
#6205floor · highSep 18, 20260.165Ξ0.145Ξ-12.1%87.9%
#6970floor · highSep 18, 20260.139Ξ0.139Ξ0.0%100.0%
#3035floor · highSep 17, 20260.133Ξ0.138Ξ+3.8%96.2%
#6237floor · highSep 17, 20260.139Ξ0.141Ξ+1.6%98.4%
#6970floor · highSep 17, 20260.134Ξ0.144Ξ+7.5%92.5%
#6970floor · highSep 17, 20260.143Ξ0.140Ξ-2.0%98.0%
#1347floor · highSep 17, 20260.135Ξ0.140Ξ+3.7%96.3%
#8143floor · highSep 17, 20260.135Ξ0.146Ξ+8.2%91.9%
#3035floor · highSep 17, 20260.137Ξ0.145Ξ+5.7%94.3%
#4907floor · highSep 17, 20260.137Ξ0.135Ξ-1.6%98.4%
#6997floor · highSep 17, 20260.158Ξ0.135Ξ-14.5%85.5%
#8100floor · highSep 16, 20260.135Ξ0.141Ξ+4.4%95.6%
#760floor · mediumSep 16, 20260.157Ξ0.165Ξ+5.1%94.9%
#4907floor · highSep 16, 20260.134Ξ0.147Ξ+9.7%90.3%
#6237floor · highSep 16, 20260.137Ξ0.153Ξ+11.7%88.3%
#6970floor · highSep 16, 20260.136Ξ0.142Ξ+4.4%95.6%
#2289floor · highSep 16, 20260.147Ξ0.133Ξ-9.5%90.5%
#4907floor · highSep 16, 20260.147Ξ0.137Ξ-6.8%93.2%
#4907floor · highSep 16, 20260.137Ξ0.152Ξ+11.0%89.1%
#4907floor · highSep 16, 20260.147Ξ0.128Ξ-12.9%87.1%
#5562floor · highSep 16, 20260.137Ξ0.160Ξ+16.8%83.2%
#2431floor · highSep 16, 20260.170Ξ0.150Ξ-11.7%88.3%
#7041floor · highSep 16, 20260.243Ξ0.136Ξ-44.0%56.0%
#6357middle · highSep 16, 20260.270Ξ0.216Ξ-20.0%80.0%
#7147floor · highSep 16, 20260.158Ξ0.125Ξ-20.9%79.1%
#6254floor · highSep 16, 20260.147Ξ0.132Ξ-10.2%89.8%
#7888floor · highSep 16, 20260.170Ξ0.149Ξ-12.3%87.7%
#7760floor · highSep 16, 20260.199Ξ0.132Ξ-33.7%66.3%
#7593floor · highSep 16, 20260.159Ξ0.125Ξ-21.5%78.5%
#3035floor · highSep 16, 20260.139Ξ0.134Ξ-3.6%96.4%
#6237floor · highSep 16, 20260.148Ξ0.132Ξ-10.8%89.2%
#7098floor · highSep 16, 20260.170Ξ0.151Ξ-11.1%88.9%
#3035floor · highSep 16, 20260.134Ξ0.130Ξ-2.7%97.3%
#7760floor · mediumSep 16, 20260.126Ξ0.140Ξ+11.1%88.9%
#7768floor · highSep 15, 20260.134Ξ0.125Ξ-6.4%93.6%
#5308floor · highSep 15, 20260.134Ξ0.126Ξ-5.9%94.1%
#7031floor · highSep 15, 20260.134Ξ0.125Ξ-6.7%93.4%
#2349floor · highSep 15, 20260.134Ξ0.125Ξ-6.4%93.6%
#1347floor · highSep 15, 20260.134Ξ0.125Ξ-6.4%93.6%

Model rektguy-1789708535-2026-08-03.2 · evaluated through 9/18/2026. Method details on the methodology page.