Accuracy
How accurate is the Sappy Seals 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)
93.0%
Mean accuracy (100−APE)
90.0%
Within ±5%
37%
Within ±10%
66%
Within ±20%
88%
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 | 1816 | 93.1% | 67% | -1.0% |
| middle | 36 | 81.3% | 19% | -4.0% |
| grail | 11 | 58.0% | 18% | +3.4% |
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 |
|---|---|---|---|---|---|---|
| #5581 | floor · high | Sep 13, 2026 | 0.061Ξ | 0.065Ξ | +5.9% | 94.1% |
| #1959 | floor · high | Sep 13, 2026 | 0.061Ξ | 0.064Ξ | +4.1% | 95.9% |
| #7840 | floor · high | Sep 13, 2026 | 0.061Ξ | 0.067Ξ | +8.9% | 91.1% |
| #5788 | floor · high | Sep 13, 2026 | 0.061Ξ | 0.067Ξ | +8.9% | 91.1% |
| #6731 | floor · high | Sep 12, 2026 | 0.071Ξ | 0.063Ξ | -11.1% | 88.9% |
| #4800 | floor · high | Sep 12, 2026 | 0.065Ξ | 0.069Ξ | +6.7% | 93.4% |
| #2980 | floor · high | Sep 12, 2026 | 0.079Ξ | 0.072Ξ | -8.8% | 91.3% |
| #1779 | floor · high | Sep 12, 2026 | 0.063Ξ | 0.072Ξ | +14.8% | 85.2% |
| #3573 | floor · high | Sep 12, 2026 | 0.063Ξ | 0.069Ξ | +10.2% | 89.8% |
| #431 | floor · high | Sep 12, 2026 | 0.062Ξ | 0.070Ξ | +12.5% | 87.5% |
| #1602 | floor · high | Sep 12, 2026 | 0.065Ξ | 0.068Ξ | +4.3% | 95.7% |
| #2139 | middle · high | Sep 11, 2026 | 0.088Ξ | 0.102Ξ | +15.9% | 84.1% |
| #348 | floor · high | Sep 11, 2026 | 0.073Ξ | 0.075Ξ | +3.0% | 97.0% |
| #116 | floor · high | Sep 11, 2026 | 0.070Ξ | 0.066Ξ | -6.2% | 93.8% |
| #7727 | floor · high | Sep 11, 2026 | 0.071Ξ | 0.063Ξ | -10.7% | 89.3% |
| #7727 | floor · high | Sep 11, 2026 | 0.063Ξ | 0.063Ξ | +0.6% | 99.4% |
| #2980 | floor · high | Sep 11, 2026 | 0.072Ξ | 0.068Ξ | -5.0% | 95.0% |
| #7297 | floor · high | Sep 11, 2026 | 0.066Ξ | 0.065Ξ | -1.2% | 98.8% |
| #116 | floor · high | Sep 11, 2026 | 0.066Ξ | 0.066Ξ | +0.2% | 99.9% |
| #7066 | floor · high | Sep 11, 2026 | 0.072Ξ | 0.066Ξ | -8.2% | 91.8% |
| #6523 | floor · high | Sep 11, 2026 | 0.069Ξ | 0.068Ξ | -1.3% | 98.7% |
| #6011 | floor · high | Sep 11, 2026 | 0.072Ξ | 0.071Ξ | -1.3% | 98.8% |
| #777 | floor · high | Sep 11, 2026 | 0.069Ξ | 0.064Ξ | -6.7% | 93.3% |
| #5502 | floor · high | Sep 11, 2026 | 0.069Ξ | 0.064Ξ | -6.7% | 93.3% |
| #4718 | floor · high | Sep 11, 2026 | 0.069Ξ | 0.066Ξ | -3.8% | 96.3% |
| #4552 | floor · high | Sep 11, 2026 | 0.070Ξ | 0.061Ξ | -12.7% | 87.3% |
| #6037 | floor · high | Sep 10, 2026 | 0.078Ξ | 0.071Ξ | -8.6% | 91.4% |
| #1131 | floor · high | Sep 10, 2026 | 0.104Ξ | 0.078Ξ | -25.0% | 75.0% |
| #5391 | floor · high | Sep 10, 2026 | 0.068Ξ | 0.065Ξ | -4.5% | 95.5% |
| #4381 | floor · high | Sep 10, 2026 | 0.068Ξ | 0.063Ξ | -7.9% | 92.2% |
| #5677 | floor · high | Sep 10, 2026 | 0.075Ξ | 0.068Ξ | -9.1% | 90.9% |
| #4381 | floor · high | Sep 9, 2026 | 0.063Ξ | 0.064Ξ | +2.4% | 97.6% |
| #5502 | floor · high | Sep 9, 2026 | 0.063Ξ | 0.065Ξ | +2.9% | 97.2% |
| #8465 | floor · high | Sep 9, 2026 | 0.200Ξ | 0.072Ξ | -64.0% | 36.0% |
| #8238 | floor · high | Sep 9, 2026 | 0.066Ξ | 0.067Ξ | +1.8% | 98.2% |
| #777 | floor · high | Sep 9, 2026 | 0.063Ξ | 0.064Ξ | +2.2% | 97.8% |
| #287 | floor · high | Sep 9, 2026 | 0.062Ξ | 0.076Ξ | +22.2% | 77.8% |
| #4871 | middle · high | Sep 9, 2026 | 0.119Ξ | 0.094Ξ | -20.9% | 79.1% |
| #8238 | floor · high | Sep 9, 2026 | 0.063Ξ | 0.072Ξ | +13.9% | 86.1% |
| #287 | floor · high | Sep 9, 2026 | 0.076Ξ | 0.063Ξ | -17.1% | 82.9% |
Model sappy-seals-1789261451-2026-08-03.2 · evaluated through 9/13/2026. Method details on the methodology page.