Accuracy
How accurate is the Milady Maker 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)
95.9%
Mean accuracy (100−APE)
93.7%
Within ±5%
58%
Within ±10%
84%
Within ±20%
96%
Bias
-1.0%
Accuracy by market tier
Floor, middle-market, and grail sales are scored separately so aggregate accuracy cannot hide a weak segment
| Tier | Sales | Median accuracy | Within ±10% | Bias |
|---|---|---|---|---|
| floor | 1462 | 96.0% | 85% | -1.3% |
| middle | 7 | 91.1% | 57% | -8.6% |
| grail | 6 | 44.6% | 17% | +87.7% |
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
| Item | Tier / evidence | Date | Sale | Model | Error | Accuracy |
|---|---|---|---|---|---|---|
| #1271 | floor · high | Nov 1, 2026 | 0.696Ξ | 0.758Ξ | +8.9% | 91.1% |
| #248 | floor · high | Nov 1, 2026 | 0.688Ξ | 0.741Ξ | +7.7% | 92.3% |
| #6673 | floor · high | Oct 31, 2026 | 0.701Ξ | 0.738Ξ | +5.3% | 94.7% |
| #1438 | floor · high | Oct 30, 2026 | 0.774Ξ | 0.771Ξ | -0.4% | 99.6% |
| #3089 | floor · high | Oct 30, 2026 | 0.775Ξ | 0.741Ξ | -4.4% | 95.6% |
| #6455 | floor · high | Oct 30, 2026 | 0.721Ξ | 0.779Ξ | +8.0% | 92.0% |
| #266 | floor · high | Oct 30, 2026 | 1.500Ξ | 0.997Ξ | -33.5% | 66.5% |
| #990 | grail · medium | Oct 30, 2026 | 2.690Ξ | 2.364Ξ | -12.1% | 87.9% |
| #764 | floor · high | Oct 30, 2026 | 0.741Ξ | 0.733Ξ | -1.1% | 98.9% |
| #3089 | floor · high | Oct 30, 2026 | 0.721Ξ | 0.787Ξ | +9.2% | 90.9% |
| #2731 | floor · high | Oct 29, 2026 | 0.735Ξ | 0.760Ξ | +3.4% | 96.6% |
| #6849 | floor · high | Oct 29, 2026 | 0.926Ξ | 0.829Ξ | -10.5% | 89.5% |
| #6849 | floor · high | Oct 29, 2026 | 0.822Ξ | 0.848Ξ | +3.2% | 96.8% |
| #8243 | floor · high | Oct 29, 2026 | 0.759Ξ | 0.774Ξ | +2.0% | 98.0% |
| #1525 | grail · high | Oct 28, 2026 | 2.900Ξ | 2.005Ξ | -30.9% | 69.1% |
| #9084 | floor · high | Oct 28, 2026 | 0.785Ξ | 0.782Ξ | -0.4% | 99.6% |
| #7945 | floor · medium | Oct 27, 2026 | 0.796Ξ | 0.798Ξ | +0.3% | 99.8% |
| #651 | floor · high | Oct 27, 2026 | 0.803Ξ | 0.773Ξ | -3.7% | 96.3% |
| #534 | floor · high | Oct 27, 2026 | 0.860Ξ | 0.807Ξ | -6.2% | 93.9% |
| #795 | floor · high | Oct 27, 2026 | 0.840Ξ | 0.763Ξ | -9.1% | 90.9% |
| #3771 | floor · high | Oct 26, 2026 | 0.800Ξ | 0.725Ξ | -9.4% | 90.6% |
| #534 | floor · high | Oct 26, 2026 | 0.772Ξ | 0.728Ξ | -5.7% | 94.3% |
| #692 | floor · high | Oct 26, 2026 | 0.750Ξ | 0.719Ξ | -4.1% | 95.9% |
| #3469 | floor · high | Oct 26, 2026 | 0.800Ξ | 0.690Ξ | -13.8% | 86.3% |
| #9424 | floor · high | Oct 26, 2026 | 0.755Ξ | 0.705Ξ | -6.6% | 93.4% |
| #5248 | floor · high | Oct 26, 2026 | 1.011Ξ | 0.788Ξ | -22.1% | 77.9% |
| #8359 | floor · high | Oct 26, 2026 | 0.889Ξ | 0.728Ξ | -18.1% | 81.9% |
| #4337 | floor · high | Oct 26, 2026 | 0.824Ξ | 0.743Ξ | -9.9% | 90.1% |
| #5385 | floor · high | Oct 26, 2026 | 0.750Ξ | 0.726Ξ | -3.2% | 96.8% |
| #6867 | floor · medium | Oct 26, 2026 | 0.950Ξ | 0.778Ξ | -18.1% | 81.9% |
| #5997 | floor · high | Oct 26, 2026 | 0.875Ξ | 0.764Ξ | -12.6% | 87.4% |
| #7075 | floor · high | Oct 26, 2026 | 0.780Ξ | 0.712Ξ | -8.7% | 91.3% |
| #2914 | floor · high | Oct 26, 2026 | 0.755Ξ | 0.678Ξ | -10.2% | 89.8% |
| #5119 | floor · high | Oct 26, 2026 | 0.872Ξ | 0.816Ξ | -6.4% | 93.6% |
| #5400 | floor · high | Oct 26, 2026 | 0.885Ξ | 0.702Ξ | -20.7% | 79.3% |
| #9618 | floor · high | Oct 26, 2026 | 0.760Ξ | 0.701Ξ | -7.8% | 92.2% |
| #2525 | floor · high | Oct 26, 2026 | 0.760Ξ | 0.674Ξ | -11.3% | 88.7% |
| #1620 | floor · high | Oct 26, 2026 | 1.490Ξ | 0.886Ξ | -40.5% | 59.5% |
| #3327 | floor · medium | Oct 26, 2026 | 0.890Ξ | 0.747Ξ | -16.1% | 83.9% |
| #365 | floor · medium | Oct 26, 2026 | 1.519Ξ | 0.947Ξ | -37.7% | 62.3% |
Model milady-1791605423-2026-08-03.2 · evaluated through 10/10/2026. Method details on the methodology page.