Accuracy
How accurate is the rektguy 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)
93.1%
Mean accuracy (100−APE)
90.4%
Within ±5%
38%
Within ±10%
67%
Within ±20%
90%
Bias
+0.7%
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 | 1528 | 93.3% | 69% | +0.7% |
| middle | 33 | 64.3% | 24% | -7.6% |
| grail | 6 | 42.2% | 0% | +44.4% |
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 |
|---|---|---|---|---|---|---|
| #2269 | floor · high | Aug 2, 2026 | 0.210Ξ | 0.213Ξ | +1.4% | 98.6% |
| #738 | floor · high | Aug 2, 2026 | 0.202Ξ | 0.205Ξ | +1.5% | 98.5% |
| #660 | floor · high | Aug 2, 2026 | 0.192Ξ | 0.199Ξ | +3.7% | 96.4% |
| #4885 | floor · medium | Aug 1, 2026 | 0.196Ξ | 0.203Ξ | +3.6% | 96.4% |
| #7654 | floor · high | Aug 1, 2026 | 0.191Ξ | 0.215Ξ | +12.6% | 87.4% |
| #738 | floor · high | Jul 31, 2026 | 0.200Ξ | 0.197Ξ | -1.5% | 98.5% |
| #2832 | floor · high | Jul 31, 2026 | 0.205Ξ | 0.198Ξ | -3.4% | 96.6% |
| #5576 | floor · high | Jul 31, 2026 | 0.208Ξ | 0.235Ξ | +13.0% | 87.0% |
| #6133 | floor · high | Jul 31, 2026 | 0.206Ξ | 0.201Ξ | -2.4% | 97.6% |
| #1349 | floor · high | Jul 31, 2026 | 0.203Ξ | 0.195Ξ | -3.9% | 96.1% |
| #6530 | floor · high | Jul 31, 2026 | 0.213Ξ | 0.206Ξ | -3.3% | 96.7% |
| #5576 | floor · high | Jul 31, 2026 | 0.235Ξ | 0.207Ξ | -11.8% | 88.2% |
| #976 | floor · high | Jul 31, 2026 | 0.215Ξ | 0.198Ξ | -7.9% | 92.1% |
| #7654 | floor · high | Jul 30, 2026 | 0.215Ξ | 0.204Ξ | -5.2% | 94.8% |
| #2016 | floor · high | Jul 30, 2026 | 0.197Ξ | 0.214Ξ | +8.6% | 91.4% |
| #7654 | floor · high | Jul 30, 2026 | 0.202Ξ | 0.202Ξ | 0.0% | 100.0% |
| #8020 | floor · high | Jul 30, 2026 | 0.195Ξ | 0.209Ξ | +7.2% | 92.8% |
| #976 | floor · high | Jul 30, 2026 | 0.199Ξ | 0.191Ξ | -4.0% | 96.0% |
| #2914 | floor · high | Jul 30, 2026 | 0.199Ξ | 0.246Ξ | +23.7% | 76.4% |
| #7302 | floor · high | Jul 30, 2026 | 0.199Ξ | 0.220Ξ | +10.6% | 89.4% |
| #904 | floor · high | Jul 30, 2026 | 0.199Ξ | 0.198Ξ | -0.5% | 99.5% |
| #904 | floor · high | Jul 29, 2026 | 0.186Ξ | 0.204Ξ | +9.7% | 90.3% |
| #7001 | middle · high | Jul 29, 2026 | 0.187Ξ | 0.315Ξ | +68.5% | 31.6% |
| #904 | floor · high | Jul 29, 2026 | 0.200Ξ | 0.209Ξ | +4.5% | 95.5% |
| #1055 | floor · high | Jul 29, 2026 | 0.189Ξ | 0.213Ξ | +12.7% | 87.3% |
| #1717 | floor · high | Jul 29, 2026 | 0.199Ξ | 0.217Ξ | +9.1% | 90.9% |
| #1363 | floor · medium | Jul 28, 2026 | 0.360Ξ | 0.232Ξ | -35.6% | 64.4% |
| #4717 | floor · high | Jul 28, 2026 | 0.198Ξ | 0.187Ξ | -5.6% | 94.4% |
| #5198 | floor · high | Jul 28, 2026 | 0.190Ξ | 0.200Ξ | +5.3% | 94.7% |
| #4717 | floor · high | Jul 28, 2026 | 0.185Ξ | 0.203Ξ | +9.7% | 90.3% |
| #976 | floor · high | Jul 28, 2026 | 0.186Ξ | 0.195Ξ | +4.8% | 95.2% |
| #6731 | floor · high | Jul 27, 2026 | 0.198Ξ | 0.195Ξ | -1.6% | 98.4% |
| #8707 | floor · high | Jul 27, 2026 | 0.199Ξ | 0.200Ξ | +0.5% | 99.5% |
| #1622 | floor · high | Jul 27, 2026 | 0.198Ξ | 0.189Ξ | -4.7% | 95.3% |
| #6841 | floor · high | Jul 27, 2026 | 0.205Ξ | 0.203Ξ | -1.0% | 99.0% |
| #738 | floor · high | Jul 27, 2026 | 0.191Ξ | 0.205Ξ | +7.3% | 92.7% |
| #976 | floor · high | Jul 27, 2026 | 0.191Ξ | 0.202Ξ | +5.8% | 94.2% |
| #6377 | floor · high | Jul 27, 2026 | 0.206Ξ | 0.258Ξ | +25.2% | 74.8% |
| #7720 | floor · high | Jul 27, 2026 | 0.206Ξ | 0.201Ξ | -2.3% | 97.7% |
| #8298 | floor · high | Jul 27, 2026 | 0.206Ξ | 0.218Ξ | +6.0% | 94.0% |
Model rektguy-1785700595-2026-08-03.2 · evaluated through 8/2/2026. Method details on the methodology page.