AppraisrAppraisr
Accuracy

How accurate is the Doodles 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)
95.3%
Mean accuracy (100−APE)
92.9%
Within ±5%
52%
Within ±10%
80%
Within ±20%
95%
Bias
-0.2%

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
floor199095.5%83%+0.2%
middle2951.9%3%+10.5%
grail9089.7%48%-13.2%

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
#5244floor · highSep 18, 20260.311Ξ0.304Ξ-2.3%97.8%
#5595floor · highSep 18, 20260.354Ξ0.322Ξ-9.1%90.9%
#2411floor · highSep 17, 20260.310Ξ0.294Ξ-5.2%94.8%
#5911floor · highSep 17, 20260.304Ξ0.294Ξ-3.2%96.8%
#4959floor · highSep 17, 20260.310Ξ0.289Ξ-6.8%93.2%
#1513floor · highSep 17, 20260.314Ξ0.287Ξ-8.6%91.4%
#4094floor · highSep 17, 20260.309Ξ0.281Ξ-9.1%90.9%
#4187floor · highSep 17, 20260.329Ξ0.294Ξ-10.6%89.4%
#5244floor · highSep 17, 20260.304Ξ0.280Ξ-7.8%92.2%
#5361floor · highSep 17, 20260.304Ξ0.301Ξ-0.9%99.1%
#1313floor · highSep 17, 20260.298Ξ0.297Ξ-0.2%99.8%
#7299floor · highSep 17, 20260.292Ξ0.300Ξ+2.8%97.2%
#8266grail · highSep 17, 20260.450Ξ0.321Ξ-28.7%71.3%
#2411floor · highSep 17, 20260.290Ξ0.316Ξ+9.0%91.0%
#1313floor · highSep 17, 20260.296Ξ0.291Ξ-1.7%98.3%
#7405floor · highSep 17, 20260.308Ξ0.299Ξ-2.9%97.1%
#5361floor · highSep 17, 20260.300Ξ0.318Ξ+6.0%94.0%
#5918floor · highSep 17, 20260.309Ξ0.303Ξ-1.9%98.1%
#7299floor · highSep 17, 20260.299Ξ0.300Ξ+0.3%99.7%
#6837floor · highSep 17, 20260.329Ξ0.301Ξ-8.5%91.6%
#4432floor · highSep 17, 20260.310Ξ0.276Ξ-11.0%89.0%
#1595floor · highSep 17, 20260.309Ξ0.279Ξ-9.7%90.3%
#4981floor · highSep 17, 20260.298Ξ0.287Ξ-3.6%96.4%
#1605floor · highSep 17, 20260.292Ξ0.284Ξ-2.7%97.3%
#2946middle · mediumSep 17, 20261.450Ξ0.801Ξ-44.8%55.2%
#7930grail · lowSep 17, 20261.270Ξ1.402Ξ+10.4%89.6%
#7320grail · lowSep 17, 20261.270Ξ1.401Ξ+10.3%89.7%
#9279grail · lowSep 17, 20261.180Ξ1.095Ξ-7.2%92.8%
#2870floor · highSep 17, 20260.286Ξ0.288Ξ+0.7%99.3%
#2069grail · mediumSep 17, 20261.110Ξ0.895Ξ-19.4%80.6%
#8033floor · highSep 17, 20260.304Ξ0.287Ξ-5.5%94.5%
#1842floor · highSep 17, 20260.282Ξ0.276Ξ-2.1%97.9%
#1605floor · highSep 17, 20260.278Ξ0.292Ξ+5.0%95.0%
#3811floor · highSep 16, 20260.286Ξ0.291Ξ+1.8%98.2%
#4981floor · highSep 16, 20260.284Ξ0.281Ξ-1.1%98.9%
#4092floor · highSep 16, 20260.286Ξ0.285Ξ-0.3%99.7%
#8033floor · highSep 16, 20260.282Ξ0.294Ξ+4.3%95.7%
#2064floor · highSep 16, 20260.288Ξ0.284Ξ-1.4%98.6%
#8291floor · highSep 16, 20260.286Ξ0.285Ξ-0.3%99.7%
#4959floor · highSep 16, 20260.282Ξ0.290Ξ+2.8%97.2%

Model doodles-1789698203-2026-08-03.2 · evaluated through 9/18/2026. Method details on the methodology page.