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)
89.4%
Mean accuracy (100−APE)
85.8%
Within ±5%
28%
Within ±10%
48%
Within ±20%
76%
Bias
-0.0%
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 | 1806 | 89.8% | 49% | -0.5% |
| middle | 88 | 72.9% | 18% | +9.7% |
| grail | 20 | 78.9% | 35% | -4.1% |
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 |
|---|---|---|---|---|---|---|
| #10140 | floor · high | Aug 2, 2026 | 0.527Ξ | 0.394Ξ | -25.3% | 74.7% |
| #8486 | floor · high | Aug 2, 2026 | 0.387Ξ | 0.356Ξ | -8.0% | 92.0% |
| #9302 | floor · medium | Aug 2, 2026 | 0.411Ξ | 0.367Ξ | -10.7% | 89.3% |
| #3885 | floor · medium | Aug 2, 2026 | 0.410Ξ | 0.360Ξ | -12.2% | 87.8% |
| #1976 | floor · high | Aug 2, 2026 | 0.432Ξ | 0.456Ξ | +5.6% | 94.4% |
| #1918 | floor · high | Aug 2, 2026 | 0.490Ξ | 0.395Ξ | -19.4% | 80.6% |
| #11016 | middle · high | Aug 2, 2026 | 0.412Ξ | 0.679Ξ | +64.8% | 35.2% |
| #624 | floor · high | Aug 1, 2026 | 0.443Ξ | 0.409Ξ | -7.7% | 92.3% |
| #74 | floor · medium | Aug 1, 2026 | 0.700Ξ | 0.411Ξ | -41.3% | 58.7% |
| #10452 | floor · high | Aug 1, 2026 | 0.423Ξ | 0.411Ξ | -2.9% | 97.1% |
| #3913 | floor · high | Aug 1, 2026 | 0.450Ξ | 0.360Ξ | -20.0% | 80.0% |
| #10452 | floor · high | Aug 1, 2026 | 0.414Ξ | 0.362Ξ | -12.6% | 87.4% |
| #624 | floor · high | Aug 1, 2026 | 0.412Ξ | 0.358Ξ | -13.1% | 86.9% |
| #10370 | middle · high | Aug 1, 2026 | 0.450Ξ | 0.611Ξ | +35.8% | 64.2% |
| #3534 | floor · high | Aug 1, 2026 | 0.485Ξ | 0.382Ξ | -21.2% | 78.8% |
| #2810 | floor · high | Jul 31, 2026 | 0.417Ξ | 0.415Ξ | -0.5% | 99.5% |
| #3634 | floor · high | Jul 31, 2026 | 0.419Ξ | 0.417Ξ | -0.5% | 99.5% |
| #7946 | floor · high | Jul 31, 2026 | 0.494Ξ | 0.385Ξ | -22.1% | 77.9% |
| #6961 | floor · high | Jul 31, 2026 | 0.425Ξ | 0.470Ξ | +10.6% | 89.4% |
| #8628 | floor · high | Jul 31, 2026 | 0.426Ξ | 0.424Ξ | -0.5% | 99.5% |
| #1705 | floor · high | Jul 31, 2026 | 0.469Ξ | 0.376Ξ | -19.8% | 80.2% |
| #5656 | floor · high | Jul 31, 2026 | 0.480Ξ | 0.449Ξ | -6.5% | 93.5% |
| #2839 | floor · high | Jul 31, 2026 | 0.469Ξ | 0.427Ξ | -9.0% | 91.1% |
| #3689 | middle · high | Jul 30, 2026 | 0.412Ξ | 0.553Ξ | +34.2% | 65.8% |
| #7724 | middle · high | Jul 30, 2026 | 0.589Ξ | 0.574Ξ | -2.6% | 97.5% |
| #5543 | middle · high | Jul 30, 2026 | 0.599Ξ | 0.553Ξ | -7.7% | 92.3% |
| #7055 | floor · high | Jul 30, 2026 | 0.510Ξ | 0.355Ξ | -30.4% | 69.6% |
| #11014 | floor · high | Jul 30, 2026 | 0.569Ξ | 0.413Ξ | -27.4% | 72.6% |
| #3947 | floor · high | Jul 30, 2026 | 0.530Ξ | 0.429Ξ | -19.1% | 80.9% |
| #9837 | floor · high | Jul 30, 2026 | 0.529Ξ | 0.427Ξ | -19.3% | 80.7% |
| #7694 | floor · medium | Jul 30, 2026 | 0.510Ξ | 0.418Ξ | -18.0% | 82.0% |
| #3009 | floor · high | Jul 30, 2026 | 0.560Ξ | 0.353Ξ | -37.0% | 63.0% |
| #4606 | middle · high | Jul 30, 2026 | 0.625Ξ | 0.492Ξ | -21.3% | 78.7% |
| #10046 | floor · high | Jul 29, 2026 | 0.645Ξ | 0.420Ξ | -34.9% | 65.1% |
| #5207 | floor · high | Jul 29, 2026 | 0.450Ξ | 0.414Ξ | -8.0% | 92.0% |
| #11016 | floor · medium | Jul 29, 2026 | 0.690Ξ | 0.374Ξ | -45.8% | 54.2% |
| #5656 | floor · high | Jul 29, 2026 | 0.448Ξ | 0.413Ξ | -7.8% | 92.2% |
| #6961 | floor · high | Jul 29, 2026 | 0.470Ξ | 0.372Ξ | -20.8% | 79.2% |
| #2839 | floor · high | Jul 29, 2026 | 0.425Ξ | 0.388Ξ | -8.7% | 91.3% |
| #5810 | floor · high | Jul 29, 2026 | 0.425Ξ | 0.475Ξ | +11.8% | 88.2% |
Model genesis-creepz-1785703823-2026-08-03.1 · evaluated through 8/2/2026. Method details on the methodology page.