Accuracy
How accurate is the DeGods 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)
92.8%
Mean accuracy (100−APE)
91.3%
Within ±5%
37%
Within ±10%
67%
Within ±20%
93%
Bias
-3.8%
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 | 533 | 92.8% | 67% | -3.9% |
| middle | 2 | 75.5% | 50% | +24.5% |
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 |
|---|---|---|---|---|---|---|
| #4717 | floor · high | Oct 30, 2026 | 0.131Ξ | 0.125Ξ | -4.6% | 95.4% |
| #9721 | floor · high | Oct 29, 2026 | 0.134Ξ | 0.127Ξ | -5.2% | 94.8% |
| #1620 | floor · high | Oct 29, 2026 | 0.134Ξ | 0.131Ξ | -2.2% | 97.8% |
| #7700 | floor · high | Oct 29, 2026 | 0.134Ξ | 0.137Ξ | +2.2% | 97.8% |
| #1692 | floor · high | Oct 29, 2026 | 0.134Ξ | 0.133Ξ | -0.8% | 99.3% |
| #1911 | floor · high | Oct 29, 2026 | 0.155Ξ | 0.125Ξ | -19.6% | 80.4% |
| #4717 | floor · high | Oct 29, 2026 | 0.121Ξ | 0.134Ξ | +10.7% | 89.3% |
| #3254 | floor · high | Oct 27, 2026 | 0.149Ξ | 0.125Ξ | -16.1% | 83.9% |
| #5601 | floor · high | Oct 27, 2026 | 0.149Ξ | 0.143Ξ | -4.0% | 96.0% |
| #6895 | floor · high | Oct 25, 2026 | 0.136Ξ | 0.131Ξ | -3.7% | 96.3% |
| #3550 | floor · high | Oct 24, 2026 | 0.169Ξ | 0.133Ξ | -21.3% | 78.7% |
| #2106 | floor · high | Oct 24, 2026 | 0.150Ξ | 0.131Ξ | -12.7% | 87.3% |
| #2568 | floor · high | Oct 21, 2026 | 0.160Ξ | 0.138Ξ | -13.8% | 86.3% |
| #5105 | floor · high | Oct 20, 2026 | 0.233Ξ | 0.153Ξ | -34.3% | 65.7% |
| #6177 | floor · high | Oct 20, 2026 | 0.130Ξ | 0.144Ξ | +10.8% | 89.2% |
| #9721 | floor · high | Oct 20, 2026 | 0.125Ξ | 0.132Ξ | +5.6% | 94.4% |
| #9581 | floor · high | Oct 20, 2026 | 0.152Ξ | 0.151Ξ | -0.7% | 99.3% |
| #7254 | floor · high | Oct 18, 2026 | 0.160Ξ | 0.140Ξ | -12.5% | 87.5% |
| #1034 | floor · high | Oct 17, 2026 | 0.125Ξ | 0.147Ξ | +17.6% | 82.4% |
| #1953 | floor · high | Oct 16, 2026 | 0.159Ξ | 0.148Ξ | -6.9% | 93.1% |
| #3254 | floor · high | Oct 15, 2026 | 0.127Ξ | 0.154Ξ | +21.3% | 78.7% |
| #8345 | middle · high | Oct 15, 2026 | 0.240Ξ | 0.241Ξ | +0.4% | 99.6% |
| #8345 | floor · high | Oct 14, 2026 | 0.241Ξ | 0.149Ξ | -38.2% | 61.8% |
| #1911 | floor · high | Oct 13, 2026 | 0.126Ξ | 0.171Ξ | +35.7% | 64.3% |
| #1717 | floor · high | Oct 13, 2026 | 0.131Ξ | 0.163Ξ | +24.4% | 75.6% |
| #780 | floor · high | Oct 13, 2026 | 0.164Ξ | 0.158Ξ | -3.7% | 96.3% |
| #9731 | floor · high | Oct 11, 2026 | 0.180Ξ | 0.164Ξ | -8.9% | 91.1% |
| #733 | floor · high | Oct 4, 2026 | 0.163Ξ | 0.157Ξ | -3.7% | 96.3% |
| #5601 | floor · high | Oct 4, 2026 | 0.159Ξ | 0.159Ξ | 0.0% | 100.0% |
| #7613 | floor · high | Oct 3, 2026 | 0.187Ξ | 0.157Ξ | -16.0% | 84.0% |
| #1953 | floor · high | Oct 2, 2026 | 0.157Ξ | 0.155Ξ | -1.3% | 98.7% |
| #4378 | floor · high | Oct 2, 2026 | 0.188Ξ | 0.149Ξ | -20.7% | 79.3% |
| #1953 | floor · high | Sep 30, 2026 | 0.157Ξ | 0.150Ξ | -4.5% | 95.5% |
| #9823 | floor · high | Sep 27, 2026 | 0.189Ξ | 0.171Ξ | -9.8% | 90.2% |
| #7351 | floor · high | Sep 26, 2026 | 0.164Ξ | 0.156Ξ | -4.9% | 95.1% |
| #8482 | floor · high | Sep 26, 2026 | 0.163Ξ | 0.145Ξ | -11.3% | 88.7% |
| #2540 | floor · high | Sep 26, 2026 | 0.159Ξ | 0.143Ξ | -10.0% | 90.0% |
| #277 | floor · high | Sep 26, 2026 | 0.179Ξ | 0.141Ξ | -21.2% | 78.8% |
| #4549 | floor · high | Sep 26, 2026 | 0.179Ξ | 0.147Ξ | -17.9% | 82.1% |
| #680 | floor · high | Sep 26, 2026 | 0.164Ξ | 0.147Ξ | -10.1% | 89.9% |
Model degods-1793330795-2026-08-03.2 · evaluated through 10/30/2026. Method details on the methodology page.