Accuracy
How accurate is the Creepz by OVERLORD 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)
90.8%
Mean accuracy (100−APE)
86.8%
Within ±5%
28%
Within ±10%
53%
Within ±20%
80%
Bias
-2.4%
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 | 1877 | 91.1% | 55% | -2.9% |
| middle | 71 | 73.0% | 17% | +12.0% |
| grail | 17 | 75.4% | 35% | -8.2% |
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
| Item | Tier / evidence | Date | Sale | Model | Error | Accuracy |
|---|---|---|---|---|---|---|
| #28 | floor · high | Aug 24, 2026 | 0.190Ξ | 0.244Ξ | +28.4% | 71.6% |
| #5461 | floor · high | Aug 24, 2026 | 0.192Ξ | 0.219Ξ | +14.1% | 85.9% |
| #7810 | floor · medium | Aug 24, 2026 | 0.196Ξ | 0.221Ξ | +12.8% | 87.2% |
| #9426 | floor · high | Aug 24, 2026 | 0.200Ξ | 0.215Ξ | +7.5% | 92.5% |
| #9105 | floor · high | Aug 24, 2026 | 0.195Ξ | 0.219Ξ | +12.3% | 87.7% |
| #9383 | floor · medium | Aug 24, 2026 | 0.201Ξ | 0.222Ξ | +10.5% | 89.6% |
| #8959 | floor · high | Aug 24, 2026 | 0.201Ξ | 0.240Ξ | +19.4% | 80.6% |
| #8750 | floor · medium | Aug 24, 2026 | 0.202Ξ | 0.230Ξ | +13.9% | 86.1% |
| #7976 | floor · medium | Aug 24, 2026 | 0.203Ξ | 0.234Ξ | +15.3% | 84.7% |
| #8841 | floor · medium | Aug 24, 2026 | 0.202Ξ | 0.235Ξ | +16.3% | 83.7% |
| #8799 | floor · medium | Aug 24, 2026 | 0.240Ξ | 0.274Ξ | +14.2% | 85.8% |
| #6582 | floor · high | Aug 24, 2026 | 0.220Ξ | 0.219Ξ | -0.5% | 99.6% |
| #5950 | floor · medium | Aug 24, 2026 | 0.242Ξ | 0.272Ξ | +12.4% | 87.6% |
| #9079 | floor · high | Aug 24, 2026 | 0.225Ξ | 0.219Ξ | -2.7% | 97.3% |
| #3166 | floor · high | Aug 24, 2026 | 0.320Ξ | 0.226Ξ | -29.4% | 70.6% |
| #9368 | floor · high | Aug 24, 2026 | 0.220Ξ | 0.219Ξ | -0.5% | 99.6% |
| #7394 | floor · high | Aug 24, 2026 | 0.281Ξ | 0.227Ξ | -19.2% | 80.8% |
| #7394 | floor · high | Aug 24, 2026 | 0.220Ξ | 0.219Ξ | -0.5% | 99.6% |
| #5654 | floor · high | Aug 24, 2026 | 0.213Ξ | 0.257Ξ | +20.7% | 79.3% |
| #4237 | floor · high | Aug 24, 2026 | 0.213Ξ | 0.254Ξ | +19.3% | 80.8% |
| #4867 | floor · medium | Aug 24, 2026 | 0.212Ξ | 0.248Ξ | +17.0% | 83.0% |
| #6582 | floor · medium | Aug 24, 2026 | 0.204Ξ | 0.235Ξ | +15.2% | 84.8% |
| #3363 | floor · medium | Aug 24, 2026 | 0.314Ξ | 0.308Ξ | -1.9% | 98.1% |
| #9361 | floor · high | Aug 24, 2026 | 0.219Ξ | 0.245Ξ | +11.9% | 88.1% |
| #9368 | floor · medium | Aug 24, 2026 | 0.218Ξ | 0.261Ξ | +19.7% | 80.3% |
| #9453 | floor · high | Aug 24, 2026 | 0.217Ξ | 0.268Ξ | +23.5% | 76.5% |
| #8892 | floor · high | Aug 24, 2026 | 0.215Ξ | 0.261Ξ | +21.4% | 78.6% |
| #9079 | floor · medium | Aug 24, 2026 | 0.216Ξ | 0.252Ξ | +16.7% | 83.3% |
| #7811 | floor · medium | Aug 24, 2026 | 0.217Ξ | 0.261Ξ | +20.3% | 79.7% |
| #8592 | floor · high | Aug 24, 2026 | 0.217Ξ | 0.266Ξ | +22.6% | 77.4% |
| #7394 | floor · high | Aug 24, 2026 | 0.218Ξ | 0.269Ξ | +23.4% | 76.6% |
| #9152 | floor · high | Aug 24, 2026 | 0.217Ξ | 0.275Ξ | +26.7% | 73.3% |
| #8463 | floor · high | Aug 24, 2026 | 0.238Ξ | 0.251Ξ | +5.3% | 94.7% |
| #9846 | floor · high | Aug 24, 2026 | 0.238Ξ | 0.248Ξ | +4.4% | 95.6% |
| #3166 | floor · high | Aug 23, 2026 | 0.231Ξ | 0.293Ξ | +26.8% | 73.2% |
| #871 | floor · high | Aug 23, 2026 | 0.296Ξ | 0.269Ξ | -9.1% | 90.9% |
| #871 | floor · medium | Aug 23, 2026 | 0.248Ξ | 0.300Ξ | +21.0% | 79.0% |
| #8978 | floor · high | Aug 23, 2026 | 0.278Ξ | 0.249Ξ | -10.4% | 89.6% |
| #7809 | floor · high | Aug 23, 2026 | 0.278Ξ | 0.249Ξ | -10.4% | 89.6% |
| #8605 | floor · high | Aug 23, 2026 | 0.261Ξ | 0.259Ξ | -0.8% | 99.2% |
Model genesis-creepz-1787613443-2026-08-03.2 · evaluated through 8/24/2026. Method details on the methodology page.