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)
91.2%
Mean accuracy (100−APE)
90.5%
Within ±5%
26%
Within ±10%
58%
Within ±20%
95%
Bias
-6.5%
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 | 514 | 91.2% | 58% | -6.5% |
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 |
|---|---|---|---|---|---|---|
| #8658 | floor · high | Jul 31, 2026 | 0.169Ξ | 0.185Ξ | +9.5% | 90.5% |
| #3681 | floor · high | Jul 27, 2026 | 0.303Ξ | 0.177Ξ | -41.6% | 58.4% |
| #7721 | floor · high | Jul 25, 2026 | 0.174Ξ | 0.183Ξ | +5.2% | 94.8% |
| #6859 | floor · high | Jul 24, 2026 | 0.219Ξ | 0.168Ξ | -23.3% | 76.7% |
| #980 | floor · high | Jul 21, 2026 | 0.300Ξ | 0.186Ξ | -38.0% | 62.0% |
| #8669 | floor · high | Jul 20, 2026 | 0.188Ξ | 0.193Ξ | +2.7% | 97.3% |
| #6362 | floor · high | Jul 19, 2026 | 0.208Ξ | 0.165Ξ | -20.7% | 79.3% |
| #5761 | floor · high | Jul 19, 2026 | 0.200Ξ | 0.182Ξ | -9.0% | 91.0% |
| #2765 | floor · high | Jul 19, 2026 | 0.200Ξ | 0.174Ξ | -13.0% | 87.0% |
| #441 | floor · high | Jul 19, 2026 | 0.195Ξ | 0.177Ξ | -9.2% | 90.8% |
| #7041 | floor · high | Jul 19, 2026 | 0.195Ξ | 0.180Ξ | -7.7% | 92.3% |
| #5601 | floor · high | Jul 18, 2026 | 0.207Ξ | 0.171Ξ | -17.4% | 82.7% |
| #1320 | floor · high | Jul 18, 2026 | 0.195Ξ | 0.185Ξ | -5.1% | 94.9% |
| #5236 | floor · high | Jul 18, 2026 | 0.220Ξ | 0.167Ξ | -24.1% | 75.9% |
| #4549 | floor · high | Jul 17, 2026 | 0.218Ξ | 0.164Ξ | -24.8% | 75.2% |
| #7351 | floor · high | Jul 17, 2026 | 0.194Ξ | 0.180Ξ | -7.3% | 92.7% |
| #5212 | floor · high | Jul 17, 2026 | 0.200Ξ | 0.164Ξ | -18.0% | 82.0% |
| #5171 | floor · high | Jul 17, 2026 | 0.181Ξ | 0.164Ξ | -9.4% | 90.6% |
| #9589 | floor · high | Jul 17, 2026 | 0.181Ξ | 0.187Ξ | +3.6% | 96.4% |
| #1039 | floor · high | Jul 17, 2026 | 0.181Ξ | 0.166Ξ | -8.0% | 92.0% |
| #6416 | floor · high | Jul 17, 2026 | 0.181Ξ | 0.191Ξ | +5.8% | 94.2% |
| #8840 | floor · high | Jul 17, 2026 | 0.181Ξ | 0.169Ξ | -6.7% | 93.3% |
| #7034 | floor · high | Jul 16, 2026 | 0.181Ξ | 0.176Ξ | -2.5% | 97.5% |
| #111 | floor · high | Jul 16, 2026 | 0.181Ξ | 0.163Ξ | -9.7% | 90.3% |
| #7598 | floor · high | Jul 16, 2026 | 0.181Ξ | 0.180Ξ | -0.3% | 99.7% |
| #5621 | floor · high | Jul 16, 2026 | 0.181Ξ | 0.165Ξ | -8.6% | 91.4% |
| #3629 | floor · high | Jul 16, 2026 | 0.181Ξ | 0.170Ξ | -5.8% | 94.2% |
| #5761 | floor · high | Jul 16, 2026 | 0.180Ξ | 0.170Ξ | -5.6% | 94.4% |
| #5236 | floor · high | Jul 14, 2026 | 0.164Ξ | 0.176Ξ | +7.3% | 92.7% |
| #983 | floor · high | Jul 14, 2026 | 0.160Ξ | 0.171Ξ | +6.9% | 93.1% |
| #111 | floor · high | Jul 14, 2026 | 0.161Ξ | 0.168Ξ | +4.4% | 95.7% |
| #9177 | floor · high | Jul 14, 2026 | 0.194Ξ | 0.173Ξ | -10.9% | 89.1% |
| #4549 | floor · high | Jul 13, 2026 | 0.165Ξ | 0.185Ξ | +12.1% | 87.9% |
| #6310 | floor · high | Jul 11, 2026 | 0.208Ξ | 0.174Ξ | -16.4% | 83.7% |
| #1039 | floor · high | Jul 10, 2026 | 0.171Ξ | 0.169Ξ | -1.2% | 98.8% |
| #3629 | floor · high | Jul 8, 2026 | 0.175Ξ | 0.199Ξ | +13.7% | 86.3% |
| #6310 | floor · high | Jul 8, 2026 | 0.169Ξ | 0.253Ξ | +49.7% | 50.3% |
| #5621 | floor · high | Jul 8, 2026 | 0.169Ξ | 0.210Ξ | +24.3% | 75.7% |
| #1894 | floor · high | Jul 8, 2026 | 0.175Ξ | 0.203Ξ | +16.0% | 84.0% |
| #6895 | floor · high | Jul 7, 2026 | 0.183Ξ | 0.174Ξ | -4.9% | 95.1% |
Model degods-1785493967-2026-08-03.1 · evaluated through 7/31/2026. Method details on the methodology page.